CONVERGENCE Five Critical Steps toward Integrating Lagging and Leading Areas in the Middle East and North Africa Convergence Convergence Five Critical Steps toward Integrating Lagging and Leading Areas in the Middle East and North Africa © 2020 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 23 22 21 20 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution —Please cite the work as follows: World Bank. 2020. Convergence: Five Critical Steps toward Integrating Lagging and Leading Areas in the Middle East and North Africa. Washington, DC: World Bank. doi:10.1596/978-1-4648-1450-1. License: Creative Commons Attribution CC BY 3.0 IGO Translations —If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations —If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content —The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third- party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1450-1 ISBN (electronic): 978-1-4648-1451-8 DOI: 10.1596/978-1-4648-1450-1 Cover image: Matt Collins, Zephyr, and Ahmad Fakih, Greenlines. Cover design: Bill Pragluski, Critical Stages, LLC. Library of Congress Control Number: 2019921087 Contents Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Memorandum to a Concerned Finance Minister. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii About the Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxv Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Why do so many place-based interventions fail?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 How can the region’s countries approach convergence? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Fragmented cities, stuck people, walled-off countries: The symptoms of institutional constraints on growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Place-based and centralized: How national policies and institutions in the Middle East and North Africa perpetuate economic inefficiency and spatial inequity. . . . . . . . . . . . . . 19 Five transitional steps to reduce institutional inefficiency, speed the Middle East and North Africa’s economic development, and enable convergent growth. . . . . . . . . . . . 25 The prospects for regional integration: Distant yet vital to the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 1 Fragmented Cities, Constrained Growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Rapid urbanization has not brought commensurate economic benefits to the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Modernist planning and informality play crucial roles in the fragmented urban fabric. . . . . 38 Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Annex 1A Methodology for calculating the agglomeration index . . . . . . . . . . . . . . . . . . . . . . 52 Annex 1B Methodology for developing indicators of urban form. . . . . . . . . . . . . . . . . . . . . . 53 v vi    C o n t e n t s Annex 1C Methodology for analyzing road and intersection densities. . . . . . . . . . . . . . . . . . 54 Annex 1D Comparison of Global Human Settlement Layers and Global Urban Footprint datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 2 Unequal Spaces and Stuck People. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 High disparities and low migration hinder economic mobility. . . . . . . . . . . . . . . . . . . . . . . . 60 Low migration suppresses labor mobility in the Middle East and North Africa . . . . . . . . . . . 69 Credential-oriented education systems offer one explanation for low internal mobility. . . . . 80 Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Annex 2A Data sources and coverage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 3 Walled Urban Economies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Large cities will remain important in the Middle East and North Africa landscape . . . . . . . 92 Regional integration can deliver large markets for the Middle East and North Africa’s cities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Annex 3A Methodology for analyzing productivity across regions . . . . . . . . . . . . . . . . . . . . . 114 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 4 How States Shape Markets through Spatial and Private Sector Development Bets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 Competition regimes in the Middle East and North Africa: How do they differ from other regions, and why? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Middle East and North Africa governments intervene in markets to shape economic geography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Government interventions cause varying magnitudes of spatial distortion in the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Implications and persistence of governments’ approach to shaping markets in the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Annex 4A Disaggregation of government expenditure, by government level, for each of the comparator countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Annex 4B Disclaimers regarding the classification of government expenditures . . . . . . . . . . . 149 Annex 4C Classification of the IMF database. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Annex 4D Reasons for excluding other Middle East and North Africa countries from the spatial analysis of government expenditures . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 5 Centralized Government: Contributor to Economic Geography . . . . . . . . . . . . . . . . 157 What do citizens expect of the state in the Middle East and North Africa?. . . . . . . . . . . . . 158 Centralized government responses reinforce spatial bias, undermining instead of encouraging convergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Decentralization has complex implications for spatial disparity . . . . . . . . . . . . . . . . . . . . . 165 Efforts to move from state-centric to citizen-centric approaches vary across the region. . . . 172 Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 C o n t e n t s   v ii 6 Five Steps for Enabling Growth through Thriving Cities and Towns in the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Transitional Step 1: Adopt new, evidence-based criteria to guide spatial interventions . . . . 180 Transitional Step 2: Devolve greater functional authority and resources for local revenue generation and service provision to local governments . . . . . . . . . . . . . . . . . . . . 188 Transitional Step 3: Step away from credentialist education and toward schooling that cultivates globally tradable skills. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192 Transitional Step 4: Renew the focus on nurturing urban agglomerations by streamlining land transfer procedures and relaxing zoning regulations in existing cities, lowering the regulatory barriers to their redevelopment. . . . . . . . . . . . . . . . . . . . . 194 Transitional Step 5: Expand market access for cities by thinning the “thick borders” that inhibit mobility across the Middle East and North Africa for both regional trade and migration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196 Concluding remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 Boxes O.1 Place-based policies have not led to spatial convergence. . . . . . . . . . . . . . . . . . . . . . . . . . 2 O.2 Economic density and agglomeration effects: The urban advantage. . . . . . . . . . . . . . . . . 3 O.3  Many signs point to one problem: The Middle East and North Africa’s economies are moving insufficiently . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 O.4 Drivers and results of high urbanization in the Middle East and North Africa . . . . . . . . 8 O.5 In the Middle East and North Africa, some cities are more spatially fragmented than others. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 O.6 Urban fragmentation as a legacy of conflict: Today’s polycentric Beirut . . . . . . . . . . . . . 12 O.7 Low spatial mobility—and high public employment—among university graduates suggest that the Middle East and North Africa’s higher education systems do not impart tradable skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 O.8 Large cities in the Middle East and North Africa show few spillover benefits from regional trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 O.9 Middle East and North Africa countries stand out in directing a large share of investment expenditures toward place-based interventions . . . . . . . . . . . . . . . . . . . . . 21 O.10 Place-based investments amount to risky development bets—and the stakes are high . . . . 22 O.11 How to make successful spatial bets? Build on natural advantage. . . . . . . . . . . . . . . . . 22 O.12 Remembering the forgotten: Institution-based policies for the urban poor . . . . . . . . . . 26 1.1 The impact of conflicts on urbanization in the Middle East and North Africa . . . . . . . 35 1.2 The impact of Lebanon’s civil war on Beirut’s urban form. . . . . . . . . . . . . . . . . . . . . . . 46 1.3 Refugee self-sorting and fragmentation in migration to urban areas . . . . . . . . . . . . . . . 51 2.1 The ongoing effects of conflict on people of the Middle East and North Africa . . . . . . . 68 2.2 Methodology for calculating the cost of barriers to migration. . . . . . . . . . . . . . . . . . . . 71 3.1 Economic growth can be contagious—but in the Middle East and North Africa, it is not . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 4.1 The process for building a housing unit in Jordan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 4.2 The IMF Government Finance Statistics database . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 4.3 Government expenditure data for a subset of Middle East and North Africa countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 5.1 Comparing Arab Barometer and World Values Survey responses on government’s role . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 5.2 Handling decentralization in fragile environments. . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 5.3 Spatial bias in Egypt’s subnational fiscal architecture. . . . . . . . . . . . . . . . . . . . . . . . . . 168 viii    C o n t e n t s 5.4 Recent advances in implementing decentralization agendas across the Middle East and North Africa, by region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 6.1 Spatially sensitive “last mile” education provision . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 6.2 Do cheap land and labor create opportunities for lagging regions? . . . . . . . . . . . . . . . 183 6.3 Big bottleneck or big opportunity: Targeted place-based policies in Afghanistan and Morocco. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 6.4 Industrial zones in Egypt: Suffering a lack of density and complements . . . . . . . . . . . 186 6.5 How the dynamics of large investors can justify government intervention . . . . . . . . . 187 6.6 Instruments to improve scale and coordination among local governments. . . . . . . . . . 189 6.7 Government-regulated private sector service delivery in the Republic of Yemen and Kenya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192 6.8 How special economic zones supported China’s incremental integration into global markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 6.9 Logistics, more than infrastructure, impedes trade in the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 6.10 Scaffolding for cross-border trade and migration in the Great Lakes Region. . . . . . . . 202 6.11 The gains to border regions in Egypt and Jordan from better regional integration . . . 204 Figures BO.4.1 The Middle East and North Africa is a highly urbanized region . . . . . . . . . . . . . . . . . . . 9 BO.5.1 Fragmentation varies significantly across urban areas of several capital cities in the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 O.1 Aggregated 1990–2004 urban expansion trends of capital cities in the Mashreq, Maghreb, and GCC subregions show divergent patterns . . . . . . . . . . . . . . . . 11 BO.6.1 After decades of conflict, Beirut became a polycentric city . . . . . . . . . . . . . . . . . . . . . . 12 O.2 Inequality within most Middle East and North Africa countries exceeds that of global peers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 O.3 In the Middle East and North Africa, companies located on the periphery face harsher constraints on business development than those in the capital city . . . . . . . . . . 13 O.4 Within-country migration is lower in Middle East and North Africa countries than in the rest of the world. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 O.5 Globally, the probability of migration tends to rise with education—but not in the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 BO.7.1 Human capital in the form of tradable skills increases spatial mobility. . . . . . . . . . . . . 15 BO.7.2 In the Middle East and North Africa, tertiary education diplomas are highly valued in the public sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 O.6 Urban populations in the Mashreq and GCC subregions are highly concentrated in the largest city and even more so in large cities. . . . . . . . . . . . . . . . . . . 16 O.7 Few Middle East and North Africa countries trade electricity. . . . . . . . . . . . . . . . . . . . . 17 BO.8.1 Spatial economic spillovers to large city economies based on deep trade agreements with neighbors, by region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 BO.8.2 Bangkok’s per capita GDP would have shrunk had it experienced the regional economic spillovers of the Middle East and North Africa . . . . . . . . . . . . . . . . 18 O.8 Place-based interventions distort urban markets—which then fail. . . . . . . . . . . . . . . . . 20 BO.9.1 Government expenditure distribution in selected Middle East and North Africa countries, by spatial category, differs greatly from international comparators . . . . . . . 21 O.9 Respondents in surveyed Middle East and North Africa economies identified jobs enablement—not public service provision or citizen representation—as the most essential function of a democratic state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 1.1 Urbanization in the Middle East and North Africa is catching up with Europe and Latin America, 1960–2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 C o n t e n t s   i x 1.2 Urban population growth rates in the Middle East and North Africa vary by subregion, 1990–2016 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 1.3 The Middle East and North Africa displays the world’s highest urban concentration as measured by the agglomeration index . . . . . . . . . . . . . . . . . . . . . . . . . 37 1.4 Economic growth per capita has not kept pace with urbanization in the Middle East and North Africa, 1990–2016 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 1.5 Cities in the Middle East and North Africa display a lower share of urban tradable employment than in other regions of the world. . . . . . . . . . . . . . . . . . . . . . . . . 38 1.6 Fragmentation varies significantly across urban areas in several capital cities in the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 1.7 The Middle East and North Africa as a whole displays average interaction potential compared with other regions but varies widely by subregion . . . . . . . . . . . . . 41 1.8 Gulf Cooperation Council countries seem to compensate for their urban fragmentation with lower commuting costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 B1.2.1 After decades of conflict, Beirut became a polycentric city. . . . . . . . . . . . . . . . . . . . . . . 46 1.9 In the Middle East and North Africa, historic centers and informal neighborhoods display higher road and intersection density than modernist neighborhoods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 1.10 Aggregated urban expansion trends in capital cities in the Mashreq, Maghreb, and GCC subregions show divergent patterns, 1990–2014. . . . . . . . . . . . . . 47 1.11 The urban expansion of Middle East and North Africa capital cities varies within subregions, 1990–2014 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 2.1 Inequalities within most Middle East and North Africa countries exceed those of global peers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 2.2 Access to electricity has converged except in low-income economies of the Middle East and North Africa, where the poorest regions remain underserved. . . . . . . 62 2.3 Primary school completion remains lower in the poorest regions of the Middle East and North Africa, except in the Islamic Republic of Iran. . . . . . . . . . . . . . 63 2.4 Projected infrastructure needs and financing in the Middle East and North Africa. . . . 64 2.5 Access to a safe water source lags behind in the poorest regions of the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 2.6 Far higher shares of population and economic activity are exposed to high or very high water stress in the Middle East and North Africa than in world averages. . . . . . . 65 2.7 Economic losses from inadequate water supply and sanitation in the Middle East and North Africa vary by economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 2.8 Violent events and water risk are associated with higher spatial inequalities in the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 2.9 Within-country migration is lower in the Middle East and North Africa than in the rest of the world. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 2.10 Net migration flows in Tunisia reflect the movement of people from high- poverty to low-poverty regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 B2.2.1 Consumption gap between leading and other regions. . . . . . . . . . . . . . . . . . . . . . . . . . . 72 B2.2.2 Share of the consumption gap explained by endowments. . . . . . . . . . . . . . . . . . . . . . . . 72 B2.2.3 Share of the consumption gap explained by returns to endowments. . . . . . . . . . . . . . . . 72 2.11 Migration to leading regions could increase consumption potential significantly in the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 2.12 Among the bottom 40 percent who migrate to leading regions, the migration benefits are restricted to the top . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 2.13 Consumption gaps between the metropolitan region and others vary across countries and are largely explained by differences in returns to endowments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 x    C o n t e n t s 2.14 Morocco shows signs of regional convergence in living standards. . . . . . . . . . . . . . . . . 78 2.15 Poverty rates at origin and destination influence migration in the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 2.16 Distribution of occupations among internal migrants and stayers in the Syrian Arab Republic, 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 2.17 Internal migration rates are higher among women than men in several Middle East and North Africa countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 2.18 Unemployment rates are higher in the Middle East and North Africa than in upper-middle-income countries of other regions, particularly for educated women. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 2.19 In the Middle East and North Africa, female migrants are more likely than male migrants to be employed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 2.20 Education has virtually no effect on migration in the Middle East and North Africa, in contrast with the rest of the world. . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 2.21 Higher education in Middle East and North Africa households is not reflected in daily expenditure as much as in other regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 2.22 In the Middle East and North Africa, tertiary education diplomas are highly valued in the public sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 3.1 Urban primacy rates are high in the Middle East and North Africa, driven mainly by population distributions in the GCC and Mashreq subregions . . . . . . . . . . . 92 3.2 Urban population distribution is skewed toward large cities in the GCC and the Mashreq, but concentrations are much lower in the Maghreb. . . . . . . . . . . . . . . . . . . . 93 3.3 Distribution of the urban population skews toward the largest cities in the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 3.4 High urban concentration in the Middle East and North Africa cannot be explained solely by fuel-export-driven consumption cities . . . . . . . . . . . . . . . . . . . . . . . 96 3.5 In the Middle East and North Africa, firms in the capital city have 6 percent higher productivity than firms on the periphery—the highest location-related effect of any region in the world . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 3.6 In the Middle East and North Africa, companies on the periphery are likelier than those in the capital city to face major constraints . . . . . . . . . . . . . . . . . . . . . . . . . 97 3.7 Selected countries, including Tunisia, show large gaps in access to public services between the primary city and the other urban areas. . . . . . . . . . . . . . . . 98 3.8 Fiscal transfers to local governments in the Middle East and North Africa are among the lowest in the world, only slightly above Sub-Saharan Africa . . . . . . . . . . . . 98 3.9 Fiscal decentralization in the Middle East and North Africa reflects larger transfers of fiscal autonomy in the Maghreb than in the GCC and Mashreq subregions . . . . . . . 99 B3.1.1 Sweden would be far poorer under Tunisia’s low regional economic spillovers . . . . . . 101 B3.1.2 Spatial spillovers based on deep trade agreements with neighbors, by region. . . . . . . . 102 B3.1.3 Bangkok’s per capita GDP would have shrunk had it experienced the Middle East and North Africa’s regional economic spillovers. . . . . . . . . . . . . . . . . . . . . . . . . . 102 3.10 Many Middle East and North Africa economies have higher average tariffs than their economic peers in other regions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 3.11 Merchandise trade as a share of GDP in many Middle East and North Africa countries is quite low. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 3.12 Only a small share of global intraregional merchandise trade occurs within the Middle East and North Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 3.13 As a share of GDP, the Middle East and North Africa’s intraregional service trade is small relative to the size of its economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 C o n t e n t s   x i 3.14 Overall, the Middle East and North Africa has a high degree of capital openness . . . . 107 3.15 FDI inflows to the Middle East and North Africa remain low despite the region’s relatively high capital openness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 3.16 Remittances are significant contributors to several Middle East and North Africa economies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 3.17 Use of digital technologies correlates closely with economic wealth. . . . . . . . . . . . . . . 112 3.18 Many Middle East and North Africa countries make it hard to visit, and their citizens also face difficulties traveling elsewhere. . . . . . . . . . . . . . . . . . . . . . . . . . . 113 4.1 Middle East and North Africa countries are less competitive than most countries of comparable income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 4.2 Business regulations in the Middle East and North Africa are rated distant from good practice with respect to efficiency and quality. . . . . . . . . . . . . . . . . . . . . . . . . . . 122 4.3 The Middle East and North Africa has consistently ranked lower than other middle- to high-income regions on the Worldwide Governance Indicators, 2007–17. . . 124 4.4 Private investment has responded less to reforms in the Middle East and North Africa than in other regions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 4.5 Government interventions can create varying degrees of spatial distortion . . . . . . . . . 132 B4.2.1 The public sector and its main components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 4.6 Distribution and changes in government expenditures of comparator countries reflect priorities through a spatial lens. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 4.7 Government expenditure distribution in selected Middle East and North Africa countries show priorities through a spatial lens for one year . . . . . . . . . 138 4.8 Government expenditure distribution in selected Middle East and North Africa countries, by spatial category, differs greatly from international comparators. . . . . . . 138 4.9 Subsidies made up 23–31 percent of Tunisia’s yearly budget, with the greater share linked to current expenditure, 2013–17. . . . . . . . . . . . . . . . . . . . . . . . 141 4.10 Subsidies on current expenditures, representing 19 percent of Tunisia’s total budget, are mainly for place-based interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 4.11 Subsidies on capital expenditures, representing 7 percent of Tunisia’s total budget, are less focused than current expenditures on place-based interventions. . . . . . . . . . . . 142 4.12 Jordanian current public expenditures, by spatial category, 2018 . . . . . . . . . . . . . . . . . . 145 5.1 Respondents in surveyed Middle East and North Africa economies identified functions to improve socioeconomic well-being as the most essential characteristics of democracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 5.2 Rural respondents were more likely to cite government’s role in job creation and public service provision as essential characteristics of democracy . . . . . . . . . . . . . 159 5.3 Across Middle East and North Africa economies, subnational surveys also show a preference for governments’ role in job creation over public service delivery . . . . . . 160 B5.1.1 WVS respondents from most Middle East and North Africa economies identified economic growth as the country’s “most important” goal . . . . . . . . . . . . . . . 161 5.4 The subregions of the Middle East and North Africa represent a spectrum in the degree of decentralization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 5.5 The fiscal decentralization of subnational governments in the Middle East and North Africa remains low compared with OECD countries . . . . . . . . . . . . . 165 6.1 Framework for effective spatial policy in the Middle East and North Africa, from foundations to final steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 B6.9.1 Among the Middle East and North Africa countries, about half improved their logistics performance between 2010 and 2018. . . . . . . . . . . . . . . . . . . . . . . . . . . 201 xii    C o n t e n t s Maps B1.1.1 Massive migration patterns to urban areas in the Mashreq region are visible through nighttime light changes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 1.1 Visual representations of urban expansion show the extent of development, by type, in selected Middle East and North Africa capitals, 1990–2014. . . . . . . . . . . . . . . 49 B1.3.1 Refugees are concentrating in three main neighborhoods in Tripoli, Lebanon, 2015. . . 51 1D.1 Comparison of Global Human Settlement Layers and Global Urban Footprint datasets for Cairo and Casablanca, 2016. . . . . . . . . . . . . . . . . . . . . . . 55 2.1 Middle East and North Africa populations are concentrated in the areas closest to international markets. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 2.2 Consumption gaps in the Islamic Republic of Iran, by region, 2014. . . . . . . . . . . . . . . . 76 2.3 Consumption gaps in Iraq, by governorate, 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 B6.8.1 China gradually increased special economic zones from 1980 through the 1990s. . . . 197 B6.11.1 Spatial welfare gains from partial integration of an Egypt–Jordan road corridor. . . . . 205 Photos O.1 In Greater Cairo, recent informal settlements share a basic spatial structure with medieval neighborhoods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.1 Developments in or near the capitals of United Arab Emirates and Egypt represent modernist urban planning theory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 1.2 Aerial views show that, in density and spatial patterns, recent informal settlements highly resemble historic districts across several cities of the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Tables 1C.1 Selected neighborhoods for analysis of road and intersection densities in the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 2.1 Economic activity benefits more from natural geography in the Middle East and North Africa than in Sub-Saharan Africa and Latin America but less so than in other regions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 B2.1.1 Violence in four current major crises has affected between one-third and two-thirds of the population. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 2A.1 Sources and years of global census data, by country. . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 2A.2 Countries and years of survey data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 3.1 City size relative to a country’s total urban population is associated with positive or negative effects on economic growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 3.2 Trade agreements are fewer and shallower in the Middle East and North Africa than in other regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 3.3 The Middle East and North Africa has few agreements with important future markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 3.4 Trade in the Middle East and North Africa is still dependent on natural resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 3.5 The Middle East and North Africa has higher service trade restrictions than any other region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 3.6 Migration in the Middle East and North Africa has been driven by both job seekers and refugees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 3.7 Large migration flows have led to equally large remittance flows in the Middle East and North Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 3.8 Largest intraregional estimated remittance flows in the Middle East and North Africa, 2016. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 C o n t e n t s   x iii B4.2.1 Classification of expenditure, by government function, within divisions and groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 4.1 On average, comparator countries invest more in people- and institution-based interventions, and less in place-based interventions, than do Middle East and North Africa countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 4.2 New cities in Saudi Arabia’s development plan and spatial strategy. . . . . . . . . . . . . . . 140 4.3 Spatial classification of Tunisia’s Finance Law 2017 . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 4.4 Concentration shares of indirect subsidies in Tunisia, by income decile, 2010 . . . . . . . 143 4.5 Off-budget economic authorities in Egypt are highly spatially distortive. . . . . . . . . . . . 143 4.6 On-budget and off-budget expenditure by economic authorities in Egypt, by spatial category. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 4.7 Jordanian public expenditures, by spatial category, 2018. . . . . . . . . . . . . . . . . . . . . . . 144 4A.1 Disaggregation of government expenditure for comparator countries, by spatial category. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 4A.2 Distribution and changes in government expenditures by comparator countries through a spatial lens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 4A.3 Expenditure distribution in the Middle East and North Africa is highly weighted toward spatially distortive policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 4B.1 Difficulties in classification of spatial categories of government expenditures . . . . . . . 150 4C.1 Classification of subcategories of the IMF Government Finance Statistics (GFS) database into eight spatial categories. . . . . . . . . . . . . . . . . . . . . . . . . . 150 4D.1 Estimated shares of public expenditures, by category, in Lebanon, the Republic of Yemen, and Iraq. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 B6.9.1 Logistics Performance Index (LPI) rankings of Middle East and North Africa countries, 2018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 B6.11.1 A road corridor connecting Egypt and Jordan would bring welfare gains from each scenario. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Acknowledgments This report was prepared by a team led thoughtful advice of our peer reviewers: by Somik V. Lall, co-led by Ayah Mahgoub, Nabila Assaf, Safaa El-Kogali, and Harris and comprising Paolo Avner, Julie Biau, Selod. The report also reflects feedback from A lex C hu ne t , Ol iv i a D’Aou s t , U we participants in various workshops in which Deich ma n n , K at ri n Heger, M at h ilde we presented intermediate drafts, including Lebrand, Sally Murray, Emiko Naomasa, the World Bank’s MENA Chief Economist Diana Tello, and Yuan Xiao. Victoria Bruce- seminar, the Sustainable Development Chief Goga supported production throughout. It Economist seminar, and the Global Solutions was initiated and prepared under the guid- Group on Territorial Development seminar. ance of Ayat Soliman and delivered under The report benefited from discussions and the guidance of Sameh Wahba and Jaafar thoughtful insights from many colleagues, Friaa. It benefited from the contributions of including Tahir Akbar, Axel Baeumler, Chorching Goh, Ellen Hamilton, Leila Kevin Carey, Tabea Dietrich, Ibrahim K aba la n , E l isa C as c a rd i , a nd Paola Elghandour, Marianne Fay, Nancy Lozano Cordovez. Gracia, Maha Hussein, Hind Kadiri, Julian The report was commissioned by the Lampietti, Guido Licciardi, Augustin Maria, Middle East and North Africa (MENA) Balakrishna Menon, Mohamed Nada, Office of the Chief Economist, and the team Noriko Oe, Jean Pesme, Björn Philipp, Salma extends many thanks to Shanta Devarajan, Rasem, Francesca Recanatini, Jade Salhab, Rabah Arezki, and Daniel Lederman for Anastasia Touati, Mohamed Yehia Abd El their commitment to shedding light on Karim, and Hoda Youssef. the nature and drivers of—and ways to The report was edited by Communications address—the core drivers of spatial inequal- Development Incorporated. The team thanks ity in the Middle East and North Africa. Bruce Ross-Larson, Nick Moschovakis, The team benefited tremendously from Matt Collins, Sarah Bridges, Ahmad Fakih, the guidance and pushback of the report and their teams. It was designed by Zephyr advisory group. The team thanks its mem- Incorporated, and the logo was designed by bers: Lamia Boutaleb, Paul Collier, Ishac Greenlines. Jewel McFadden, of the World Diwan, Hedi Larbi, Lant Pritchett, and Tony Bank’s Development Economics Strategy Venables. The team is also grateful for the and Operations unit, and Mary Fisk and xv xvi    A c k n o w l e d g m e n t s Deb  Appel-Barker, of the World Bank’s Andu  Shuai Liu, William Stebbins, and formal publishing unit, were responsible for ­ Isabelle Poupaert for their communica- managing the book throughout the publi- tions support. The report was cofinanced cations process. Mary Anderson was the by the U.K. Department for International copyeditor. The team thanks them all. The Development through the Multi-Donor Trust team also thanks Kristyn Schrader-King, Fund for Sustainable Urban Development. Memorandum to a Concerned Finance Minister Subject: Five critical steps toward environment in many cities and towns integrating lagging and leading restricts new firms from entering and grow- areas in your country ing, and the lack of complementary infra- st r uc t u re i nve st ment s hobble s lo c a l This memo introduces a report that you economies. Second, most residents in your may find useful and interesting. Focusing on lagging areas are stuck in place, unable to actions that can put countries in the Middle take full advantage of jobs that vibrant urban East and North Africa on a path to territorial economies can offer. So, what do your cities convergence, it concludes that governments need for more vibrancy and private sector can take the lead by tackling the economic jobs? They need larger markets—often and institutional causes of spatial exclusion. beyond national borders—to increase the Rising spatial disparities are threatening demand for goods and services, to increase economic growth and social inclusion in your the demand for human capital, and to create country and across the region. This report fulfilling jobs for young people. shows that opportunities for your citizens are How then, can you start building a con- shaped by accidents of where they were vergence machine for your lagging areas? born—much more so than in any other part You can reduce territorial disparities more of the world. Decision makers in your coun- immediately and effectively by taking five try and in other parts of the region have steps: taken steps to respond to the needs of people left behind in your cities and across your 1. Strengthen coordination and comple- regions. Even so, spatial disparities either mentarities across sectoral interven- continue to grow or are closing more slowly tions. Efforts to enhance job prospects in than would be expected given the volume of places left behind should operate across investment you have directed to those and address multiple development axes locations. simultaneously. Development strate- Why is territorial convergence so difficult? gies are more likely to succeed if they First, most lagging areas in your country are are multidimensional—including access limited by an inability to leverage the full to energy, transport, land, and markets returns to their endowments. The business in the same place, whether sequentially xvii xviii   M e m o r a n d u m t o a C o n c e r n e d F i n a n c e M i n i s t e r or concurrently. Because ­ starting anew which concentrates economic activity geo- is extremely difficult, there is little value graphically. For this, the fabric of cities in single-sector interventions. A  good needs to be spatially connected, dense with place to start is by anchoring investments people, and transit oriented—not sprawl- around cities, which have many of the ing, which perpetuates the dispersion of missing complements. Complementary people and jobs. Planners and regulators reforms that help get the prices right—for can attract firms to invest in cities by energy and for land—can go a long way reducing frictions such as zoning regula- toward creating the conditions for job tions; impediments to property acquisition creation in lagging areas. The good news and new construction (costs, height limits, is that you don’t have to pay more to see and density limits); challenges to local better results, because spatial coordination business registration and licensing; limits will generate cost savings in the medium to on news and information; and obstacles to longer term. developing local business networks. 2. Redistribute roles and responsibilities 5. Enhance market access for lagging areas, across tiers of government. Citizens in dif- nationally and regionally. Historically, the ferent parts of the country have varying region’s cities were part of economically needs, and local conditions require flexible central global trade networks. Many of service delivery models. Devolving respon- these cities persisted into modern times as sibilities for local revenue generation and large, often vastly populated urban areas. service provision to local governments can Yet with today’s thick national borders, make them better equipped and more their economic reach has been limited. accountable. Effective decentralization Countries across the region need to would also empower them to cover the enhance links across national borders— recurring costs of their investments. reducing tariffs and nontariff barriers 3. Enable greater mobility of your people (such as logistics and trade facilitation) between lagging and leading areas. Major and easing movements of goods and peo- gains in living standards can be reaped ple. They may also need to enact policies from greater domestic labor mobility. to strengthen domestic markets. Such Research for this report shows that living efforts will expand the size of urban econ- standards of people moving internally to omies, providing much-needed tax major cities can increase by 37 percent on resources to redistribute in areas left average across the Middle East and North behind. Africa region. Women are more likely to move and find jobs in urban areas, but The extent and sequence for implementing they need support to do so. One of the key each transitional step will depend on your constraints on greater mobility is the cre- assessment of your country’s readiness— dentialist education system prevalent political, technical, and administrative—to across the region. It needs to be more ori- implement these recommendations. ented toward marketable skills. These steps will allow you to promote the 4. Build dense and connected cities. Well- building blocks of a convergence machine functioning cities offer a wide variety of for spatial inclusion by pursuing economic jobs for women and men. Making land growth and inclusion rather than spatially markets in cities more efficient is critical targeted mandates. The only spatial require- for agglomeration and specialization— ments are that people across your country two dynamics that enhance job creation have access to high-quality basic services— and economic prosperity. Whether in and that economic development interven- larger or in smaller (secondary) cities, tions harness the spatial and economic agglomeration and specialization require dynamics of agglomeration, migration, and the benefits from high economic density, specialization. M e m o r a n d u m t o a C o n c e r n e d F i n a n c e M i n i s t e r    xix All levels of your government have roles: complementing every investment with institu- the national, the provincial, and the local. By tional measures that make the desired jobs requiring that all interventions be responsive more likely to appear. to the basic needs of all, you can put far more These steps may appear daunting, even of your nation’s forgotten people into jobs painful. But evidence from two centuries of than ever before. How? Not by trying solely experience from around the world has shown to bring jobs where people are but by also that the potential gains are worth the pain. focusing on where the jobs are most likely to The World Bank Group can help as you be and enabling people to move there, while decide on a way forward. About the Authors Somik V. Lall is the World Bank’s global Economics; and working papers. He holds a lead on territorial development solutions, a bachelor’s degree in engineering, a master’s lead economist for urban development, and degree in city planning, and a doctorate in the team leader of this report. He is a recog- economics and public policy. nized expert on development policy related to urban and territorial competitiveness, agglom- Paolo Avner has been working as an urban eration and clusters, and infrastructure, with economist at the World Bank for the past five over 20 years’ global experience, most notably years. His current work focuses on the links in Africa, Asia, and Latin America. He has between urban form, land uses, transport sys- been a core team member of the World tems, labor markets, and vulnerability to nat- Development Report 2009: Reshaping ural hazards in the cities of low- and Economic Geography and developed the pol- middle-income countries. He has worked on a ­ icy framework for development of lagging number of analytical products, including the areas within countries. He is the lead author World Bank’s Urbanization Reviews (Ethiopia, of the World Bank’s flagship report on urban- Guinea, Haiti, Kenya, and Mali) and flagship ization, Planning, Connecting & Financing reports, and is the author of several Cities—Now, as well as the recent Africa’s policy-oriented research papers. Prior to join- ­ Cities: Opening Doors to the World. Somik ing the World Bank, he worked as a researcher heads a World Bank global research program at Laboratoire d’Economie de la Production et on urbanization and spatial development and de l’Intégration Internationale in Grenoble previously founded the Urbanization Reviews before joining the International Research program. His research and policy advisory Center on Environment and Development in interests focus on place-shaping policies Paris, where he collaborated on the develop- around cities, clusters, and corridors and the ment of an applied land use and transport functioning of factor and product markets, interaction model (NEDUM-2D). His work with more than 40 publications featured in specifically focused on the ability of public peer-reviewed journals including the Journal policies and investments to curb greenhouse of Development Economics and Journal of gas emissions from urban transport while lim- Urban Economics; edited volumes, including iting the costs of these policies for urban the Handbook of Regional and Urban residents. xxi xxii    A b o u t t h e A u t h o r s Julie Biau works with the World Bank Olivia D’Aoust is an urban economist with office in Brussels, overseeing the lending and the World Bank’s Urban and Disaster Risk technical assistance portfolio in European Management team for Africa, based in Union countries and serving as a focal point Washington, DC. Her work cuts across issues with the European Commission on struc- of the economics of urbanization, territorial tural reform projects. She previously worked development, and conflict and fragility. She in the World Bank’s Middle East and North has worked on several analytical products Africa region, based in Washington, DC, focusing on the drivers of and impediments where she provided analytical work and to cities’ productivity and livability, including operational support in the urban sector. the flagship report, Africa’s Cities: Opening Between 2011 and 2013, she was a junior Doors to the World. As a core member of the professional associate, supporting project Global Solutions Group on Territorial design and supervision in urban and water Development, her work expanded to inform teams, particularly in Central America and spatial development strategies, identify pri- Haiti. Her experience outside of the World orities for action in lagging regions, and Bank includes work at the Inter-American explore avenues for operations to leverage Development Ban k’s (I DB) Of fice of spatial synergies and address missing comple- Evaluation and Oversight, where she led the ments. Olivia holds a doctorate in economics evaluation of IDB country portfolios in vari- from the European Center for Advanced ous sectors.  She also  spent a year as a Research in Economics and Statistics at the research analyst at the Brookings Institution, Solvay Brussels School of Economics and working on the sustainable development Management and a master’s degree in agenda and statistical analysis of topics such demog raphy f rom t he C enter for as global poverty trends and aid effective- Demographic Research at UCLouvain. ness in fragile states. She holds a master’s degree in public policy from Georgetown Uwe Deichmann is an independent consul- University. tant based in Berlin. Previously, he spent almost 20 years in the Development Research Alex Chunet is an urban economist, focus- Group at the World Bank, where his main ing on topics linked to urban fragmentation, research interests included the role of infra- transportation, and access to services in structure in promoting regional growth, Africa, Latin America, and the Middle East urbanization, and the impacts of natural and North Africa. In the past year, he has h azards and global change on economic ­ worked as a consultant for the World Bank development. He was a principal author and participated in several Urbanization of World Development Report 20 09: Reviews and one flagship report. Alex com- Reshaping Economic Geography and pleted his undergraduate and graduate stud- ­codirector of World Development Report ies in public policy at Sciences Po in Paris 2016: Digital Dividends. Before joining the before specializing in local economic devel- World Bank, he worked for the United opment, earning a master’s degree at the Nations Environment Programme and London School of Economics, which gave the UN Statistics Division. He holds a doc- him the opportunity to do World Bank- torate in economic geography and regional sponsored field research on cities’ competi- science from the University of California, tiveness in Cameroon. Previously, he worked Santa Barbara. in several other international organizations, including the Organisation for Economic Mathilde Lebrand is an economist in the Co-operation and Development and the World Bank’s Transport Global Practice. Her United Nations, as well as in a consulting current work focuses on the economic impacts firm specializing in sustainable development of transport investments, such as corridors in in Bogotá, Colombia. China’s Belt and Road Initiative or urban A b o u t t h e A u t h o r s    xxiii transport systems, the role of labor and firm development. She completed her master’s mobility, and the spatial impacts of trade degree in development studies (research) at opening in low- and middle-income countries. the London School of Economics and her She holds a doctorate in economics from the bachelor’s degree in philosophy, politics, and European University Institute in Italy. Her economics at the University of Oxford. research focuses on economic geography, international trade, networks, and political Emiko Naomasa is an economist in the economy. Mathilde previously taught at the Development Economics Vice Presidency of University of Montreal and has worked at the the World Bank. Her specialization includes World Trade Organization in Geneva. She is agricultural economics, natural resource also a research fellow of the Ifo Institute for management, and rural sector development. Economic Research, Munich. Her current research focus is on the adapta- tion of climate-smart agricultural manage- Ayah Mahgoub, a senior urban develop- ment. She also has extensive experience in ment specialist at the World Bank, is the co- macroeconomic monitoring and sovereign team leader of the Convergence report. She risk assessment. Before joining the World leads urban and territorial development lend- Bank, Emiko was an economist at the Japan ing and analytical projects in North Africa Bank for International Cooperation, where and leads the World Bank knowledge groups she was responsible for analyzing fiscal poli- on competitive cities and results-based cies and sovereign credit risks in European financing. Her focus areas currently are and North African countries. She holds a urban and territorial development, city com- doctorate in economics from the University petitiveness, intergovernmental fiscal sys- of Hawaii. tems, municipal finance, smart cities, and results-based financing. Ayah completed her Diana Tello is an urban economist work- undergraduate and graduate studies in eco- ing as a consultant at the World Bank. Her nomics and international development at work focuses on territorial development, Harvard University. Before joining the World local economic development, urban resil- Bank, she worked for the Center for Global ience, competitive cities, and the spatial Development, the United States Agency for dimension of government expenditures. Most International Development’s Development recently, her work has focused on building Innovation Ventures program, the Crown the evidence base for identifying instruments Prince Court of Abu Dhabi, and Phipps of territorial development that have the Community Development Corporation. She potential to boost economic activity in a has also worked on urban development in place and reduce gaps in the quality of life New York City, economic integration of between places, spanning Africa, Latin minorities in France, and peace-building ini- America and the Caribbean, and the Middle tiatives in Sudan. East and North Africa. She also has experi- ence in Malaysia, where she worked on an Sally Murray is a consultant for the World economic assessment at the subnational level Bank’s Global Solutions Group on Territorial to propose policies that could trigger devel- Development, where her work focuses on opment in Melaka state. Before joining the regional development and urbanization. World Bank, Diana worked in the financial Before joining the World Bank in April 2018, markets of New York City and Monterrey, she was senior country economist for the Mexico, advising on fintech solutions and International Growth Centre (IGC) in credit risk management. She completed her Rwanda (previously South Sudan), where undergraduate studies in economics at she coauthored several IGC policy papers Tecnológico de Monterrey, Mexico, and on urbanization, taxation, state effective- holds a master’s degree in local economic ness  and fragility, and technology for xxiv    A b o u t t h e A u t h o r s development from the London School of Bank’s Middle East and North Africa Region. Economics. Before joining the World Bank, Yuan was an assistant professor of urban planning at Yuan Xiao is an urban development spe- Columbia University. She holds a doctorate cialist in the World Bank’s Beijing office, f rom t he M assachuset ts I nst it ute of where she plays a key role in the lending port- Technology, a master’s degree from University folio and advisory services as well as analyti- of Toronto, and bachelor’s degrees from cal work in the urban field in China. Before Peking University. this assignment, she worked in the World Abbreviations AGOA African Growth and Opportunity Act CDD Civil Defense Department CEPGL  Economic Community of the Great Lakes Countries (Communauté Économique des Pays des Grand Lacs) COFOG Classification of Functions of Government COMESA Common Market for Eastern and Southern Africa CPER State-Region Plan Contract (contrat de plan État-région) DAI Digital Adoption Index EBA enterprise bargaining agreement EPCI  Établissement Public de Coopération Intercommunale (Public Establishment for Intercommunal Cooperation) EU European Union FDI foreign direct investment GAM Greater Amman Municipality GCC Gulf Cooperation Council GCI Global Competitiveness Index (WEF) GDP gross domestic product GFS Government Finance Statistics (IMF) GHSL Global Human Settlement Layers (dataset) GUF Global Urban Footprint (dataset) HIP Hawassa Industrial Park ICT information and communication technology IDA Industrial Development Authority (Egypt) IDP internally displaced person xxv xxvi    A b b r e v i a t i o n s IMF International Monetary Fund ISIL Islamic State of Iraq and the Levant LEI Landscape Expansion Index LGPA Local Government Performance Assessment LGU local government unit LPI Logistics Performance Index (World Bank) MNAPOV Middle East and North Africa Poverty database NASA National Aeronautics and Space Administration OECD Organisation for Economic Co-operation and Development OSM OpenStreetMap PAFTA Pan-Arab Free Trade Area PARAS Project to Restructure Local Government and Services PIRLS Progress in International Reading Literacy PISA Programme for International Student Assessment PVH Phillips-van Heusen SEDAC Socioeconomic Data and Applications Center SEZ special economic zone TEN-T Trans-European Transport Network TFP total factor productivity TIMSS Trends in International Mathematics and Science Study TMSA Tanger Med Special Agency TTN Tunisia TradeNet TVA Tennessee Valley Authority UDLGP Urban Development and Local Government Program UN-Habitat United Nations Human Settlements Programme VAT value added tax WEF World Economic Forum WGI World Governance Indicators WTO World Trade Organization WVS World Values Survey Overview T he Middle East and North Africa is people are marginal to the formal economy— suffering from spatially divergent or live outside it, seemingly forgotten. development. The uprisings of the Arab Spring in part reflected grievances of citizens who were (or were perceived to have Why do so many place-based been) left behind, particularly by accidents of interventions fail? where they were born. Although the trajec- Why have place-based spatial initiatives tory of every nation in the region varies, one largely failed in the region’s countries? Why stated objective is clear for them all: improve have they not yielded more sustainable jobs outcomes for people in areas that have been and growth? Although the challenges are left behind. many and vary across the region, recent Policy makers across the region have long work in economic geography shows that been trying to integrate their people spatially most of these place-based policies get one a nd econom ical ly. Wish i ng to bri ng thing wrong: they attempt to treat inequity’s communities together and narrow economic ­ spat ia l a nd physic a l sy mptom s , not gaps, governments have made large capital its causes. Thus, to add jobs in a country’s investments in transport corridors and “new poorer areas, policy makers try to push new cities.” Wishing to provide jobs in places with production facilities into these areas. And to little economic activity, governments have meet the need for decent homes and ameni- designated new industrial zones supported by ties in poor urban neighborhoods, funds spatially targeted business incentives and support mass housing projects. Neither subsidized land and energy. effort has succeeded widely—because the Yet the results of these place-based initia- causes of spatial exclusion are not them- tives in these countries are mostly disappoint- selves spatial and physical; they are eco- ing (box O.1). The disparities between capital nomic and institutional. cities and lagging areas, and between richer First, a lack of economic density in rural and poorer quarters of cities, remain stark. areas, and even in many smaller municipali- Across much of the region, a fortunate few are ties, makes them inherently less competitive connected to opportunity, while many more than large cities because they are less suitable 1 2    C ONVERGEN C E BOX O.1  Place-based policies have not led to spatial convergence Insufficient attention to the economic causes of spa- such remote locations can be far more expensive tial exclusion has led governments in the Middle than supplying them in a city (where the cost can East and North Africa to pursue spatially targeted be spread across a larger number of customers). ­ interventions—yet most countries have little to show And firms that locate far from large urban areas for these place-based policies for several reasons: cite constraints in their business environment— the main challenges being political instability, low • New cities have not yielded the hoped-for access to finance, and low access to electricity. returns. Governments are building new cities as • New transport corridors have not facilitated a respite from the chaos of today’s large and bus- regional trade. Most middle-income Middle East tling metropoles—but the main result is that res- and North Africa countries have invested heav- idents are marooned far from jobs. In the Arab ily in national transport infrastructure, yet firms Republic of Egypt in 2012, 30 percent of the encounter high nonphysical barriers to trade national built environment budget was allocated within the region. Outside the Gulf Cooperation to new cities, which host just 2 percent of the Council (GCC), a the region’s countries score nation’s people. Today, the people who moved to poorly on such trade facilitation measures as those new cities must rely on their own cars or the quality of customs and logistics procedures. fleets of buses to shuttle daily to the older urban As a result, less than 7 percent of global intra- centers where they work (especially the Cairo regional merchandise trade occurs within the agglomeration) (Sims 2015). Middle East and North Africa, compared with • New industrial zones and newly designated 40  ­p ercent within East Asia and more than growth centers lack promise. Governments have 50 percent within Europe.b undertaken large-scale investments and offered generous subsidies to create jobs in socially and a. The Gulf Cooperation Council (GCC) includes Bahrain, Kuwait, Oman, Qatar, economically excluded areas. But the proposed Saudi Arabia, and the United Arab Emirates. new industrial zones and growth centers often lack b. Global intraregional trade data are derived from the World Trade Flows database, Bilateral Data files, of the Center for International Data, University of the agglomeration benefits that larger cites already California, Davis,HYPERLINK https://cid.econ.ucdavis.edu/Html/WTF_bilateral​ have. Supplying infrastructure and amenities to .html. for large-scale investment. Large urban areas of competitiveness, business environment, have well-understood advantages in today’s and governance. Subnationally, firms global economy. Dense agglomerations favor report in surveys a range of impediments specialized, scaled-up production for interna- to doing business, and the challenges vary tional markets (box O.2). So, an industrial from city to city within countries. One zone set up far from a country’s main cities major challenge is the limited coordina- is, most likely, set up to fail. tion of complementary investments and Second, not only rural but also urban policies needed to make cities and regions economies in the Middle East and North attractive for entrepreneurs to establish Africa are constrained by widespread institu- and grow businesses. Sectorally siloed tional inefficiencies. Five main types will be interventions make cities and regions considered here: inadequate homes for businesses that need complementary factors such as good mar- •  Barriers to market entry and lopsided ket access, well-serviced land, and a rel- business environments, which var y evant talent pool. within countries. Several of the region’s •  Centralized control over local public countries significantly underperform rela- services. Outside the capital city in the tive to comparator countries on indexes region’s countries, smaller cities and O v e r v i e w   3 BOX O.2  Economic density and agglomeration effects: The urban advantage For firms participating in today’s regional and global The higher productivity of efficient urban agglom- value chains, economic density —the geographic con- erations can promote a virtuous circle of economic centration of economic activity—gives large cities an growth. As a rule, if firms and workers are mobile, edge over smaller cities and rural areas. Such density both will pursue productive opportunities across boosts efficiency by making workers and firms more space, preferring to settle near existing agglomera- productive. Economic density can make public ser- tions. Large cities become even larger and denser. vice provision less costly. And as production becomes Agglomeration effects imply that an efficient eco- more complex and diversified, the physical proxim- nomic landscape is not smooth—with equal economic ity of firms becomes more critical. Thanks to these density everywhere—but is instead lumpy, with large agglomeration effects, firms that specialize in produc- cities driving growth. However, economic distance ing tradables can operate more efficiently at scale and (the difficulty of access from lagging to leading areas) can thus compete more effectively. (Scale economies may limit economic actors’ mobility and location and specialization are two key drivers of firm and sec- choices by making migration costlier or less benefi- toral productivity. Scale allows a firm to reduce unit cial. Friction can also result from less obvious eco- costs, increasing productivity. Specialization allows nomic factors, such as divisions— ethnic, political, it to narrow its focus to a few products, making the religious, or linguistic barriers to interaction. most of its key advantages.) Source: World Bank 2009. other localities lack the authority to raise often lack the tradable skills demanded their own revenues and to manage local by the private sector. So, the region’s cities service provision. These functions can, are less attractive than they could be to however, generally be conducted more firms and investors, and citizens are con- efficiently by localities—and local gov- strained in their ability to leverage their ernment officials can be more easily held skills for employment across places. The accountable. emphasis on credentials, rather than on •  Urban regulatory frictions. Especially portable skills, also reduces the chances in urban areas, rigid and outdated regu- for young women and men to move to lations distort land markets and stymie cities where they can access better job development. For example, Tunisia pro- opportunities. hibits residential buildings higher than •  Barriers to the spatial mobility of goods three stories, and Jordanian regulations and people. Input and output flows are impose a minimum lot size of 100 square impeded within the Middle East and meters—restrictions that effectively limit North Africa by barriers, sometimes the supply of affordable formal housing. called “thick borders.” Among the main In addition, land transfer fees and build- barriers are limits on news and informa- ing permit fees tend to be high in the tion and practical constraints on travel region’s countries. and trade (such as visa difficulties, weak •  Credentialist education systems. In the infrastructure, and logistical hurdles). Middle East and North Africa, education Thick borders discourage firms in the is widely seen as providing a credential— region’s cities from creating jobs and and the credential is valued mainly as a expanding their economic reach—and ticket to public sector employment. As a also inhibit the growth of smaller special- result, even the most educated workers ized cities. In addition, within countries, 4    C ONVERGEN C E migration controls can distort labor mar- revenue and services. With greater f ­ unctional kets by reducing mobility from rural to authority, they can better tailor their service urban areas. delivery models to the characteristics of their territories and varied needs of their citizens, Based on the research for Convergence, including better leveraging ­ digital technolo- this overview outlines the roots of spatial gies for service provision. Decentralization of institutional inefficiencies across the Middle functions and finance is urgent in lagging East and North Africa—within cities, within areas that have been targeted for place-based countries, and across national borders—and investments, where local governments need it proposes institutional remedies and invest- to be empowered to cover the recurring costs ment priorities informed by economic geog- of these investments. The speed and design of raphy. The proposed reforms are ambitious. decentralization may vary based on current They will require strong leadership from city and historical institutional architecture, eco- and national governments. However, getting nomic geography, and other strategic deci- the region on a path to spatial convergence is sions. But regardless of the specific design worth being ambitious and taking on a chal- choices, decentralization is likely to deliver lenging agenda. better results when shifts in functional assignment are accompanied by commensu- rate shifts in resources and capacity. How can the region’s countries Step 3: Step away from credentialist edu- approach convergence? cation and toward schooling that cultivates The overview identifies five transitional globally tradable skills. Decentralizing steps that Middle East and North Africa school systems could help. But also needed is countries can take starting now, with the a shift in public sentiment: citizens who now urgent aim of getting the region to a path see the government as chiefly a job creator toward convergent development. must come to recognize that marketable Step 1: Enact new, evidence-based ­ criteria skills will attract investment and growth. to guide future spatial interventions. To Such skills will also make workers more move on from decades of failed place-based mobile between lagging and leading regions, policy, require that the next generation of and such mobility is economically productive spatial interventions ensure efficient access to and desirable. a large urban market either within the coun- Step 4: Renew the focus on nurturing try or across national boundaries. Also focus urban agglomerations by streamlining land on identifying major bottlenecks that spa- transfer procedures and relaxing zoning reg- tial initiatives could clear—whether in land ulations in existing cities, lowering the regu- markets, skill markets, service provision, or ­ latory barriers to their redevelopment. trade and labor mobility. And thoroughly Spatial inclusion is less likely to happen in consult all stakeholders in an area, including new cities, since they are often disconnected local firms, authorities, and residents as well from contiguous urban fabric. Instead, inclu- as potential investors. The key issues here are sion follows economic growth, which will spatial coordination across complementary most likely occur in existing urban areas. To sectoral investments and measures to reduce promote the agglomeration and specializa- subsidies to large firms that distort the price tion that drive growth, policies should make of labor relative to capital and discourage cities more economically dense. A first step is labor-­intensive ­economic activities. to make existing urban land markets more Step 2: Devolve greater functional author- efficient. ity over local revenue generation and service Step 5: Expand market access for cities by provision to local governments. To empower thinning the “thick borders” that inhibit lagging areas, make localities more respon­ mobility across the Middle East and North sible, equipped, and accountable for both Africa, for both regional trade and migration. O v e r v i e w   5 For regional trade, add transit links and other across the Mashreq, the Maghreb, and the critical infrastructure and reduce nontariff G u l f C o op erat ion C ou nc i l (G C C) barriers. For migration, streamline customs subregions:1 and reduce internal migration controls. Focus on areas near major cities—but also on any •  Cities are physically and economically lagging areas that have already been targeted fragmented, precluding the economic for place-based investment. Expanding cities’ benefits of agglomeration. In the large cit- access to markets will help to expand the ies of the Middle East and North Africa, ­ fiscal base that can support greater redistri- planners and policy makers have tended bution to lagging areas. to start anew rather than work within As Step 1 suggests, policy makers are the existing urban fabric. The resulting urged to enhance the design and implementa- new cities and modernist neighborhoods tion of place-based policies such that they do not support the amount of interac- enable markets to take full advantage of three tion seen in historic districts or in newer growth drivers associated with economic informal settlements. A comparison of geography: agglomeration, migration, and 20 neighborhoods across 8 of the region’s specialization. Rather than solely focusing on cities shows that intersection d ­ ensity—a attracting capital to places where it is scarce, proxy for interaction potential—is almost a refined approach should focus on develop- three times higher in historic city centers ing the broader ecosystem that can make and informal neighborhoods than in mod- places attractive for investment while provid- ernist neighborhoods. Planners who put ing the opportunities for people to seek physical form before economic and social opportunities wherever their talents provide function have thus inhibited agglomera- the highest returns. tion effects. In the long term, making the most of these •  People are spatially and economically growth drivers will require reforming core stuck in place, lacking tradable skills and institutions and, in particular, working therefore limited in mobility. Around the toward regional integration in the Middle world, migration from places not doing East and North Africa. But in the short to well to those doing well has been key for medium term, the five transitional steps economic integration and the reduction should speed economic growth and thus of spatial inequalities. And although dis- add to the momentum and demand for fur- parities between subnational regions in ther  spatial convergence. All levels of the Middle East and North Africa con- ­ g overnment—national, provincial, and tribute to a 63 percent larger share of local—have roles to play. total inequality in consumption than else- where (see chapter 2), people are stuck in place. In contrast with other parts of Fragmented cities, stuck people, the world where higher education gener- walled-off countries: The ally increases spatial mobility, earning symptoms of institutional a university degree in the Middle East constraints on growth and North Africa does not make a per- Today, Middle East and North Africa son more likely to migrate. One reason is countries are not notably benefiting from that higher education confers credentials the three spatial dimensions of for coveted local public s ­ ector jobs, not ­ m arket-driven economic growth: agglom- tradable skills that are in demand by the eration, migration, and specialization. private sector. Other ­reasons analyzed in Instead, their economic development is this report include weak private sector characterized by three striking symptoms dynamism limiting the economic “pull” of institutional inefficiency, all of which of cities and frictions in trade and use of appear widely (though not universally) land. 2 6    C ONVERGEN C E •  Economies are walled off from others, governments in the region tend to assume an regionally and globally, by many barriers activist role in shaping markets, whether at that governments in the Middle East and the national or the neighborhood scale. The North Africa have created—or failed to preference for place-based investments remove. Compared with Europe or East inspires various interventions, from indus- Asia, few countries in the Middle East trial location regulations to growth poles and and North Africa benefit from growth new cities. The bets have high stakes: they spillovers from their neighbors. Indeed, consume an outsize share of public expendi- most export facilities in the region ship tures. Yet these countries appear to have raw materials directly to global markets. worse outcomes than their global peers, with The automobile production chains that isolated exceptions. include Morocco are based in Europe, not Policies that distort markets, spatially or in the Middle East and North Africa. otherwise, usually fall short of their stated goals. Many place-based interventions fail to achieve even their redistribution objectives— Excess centralization, the nature of and few are well designed and managed for place-based investments, and weak economic growth. Today, the economic envi- institutions have created less ronment across the Middle East and North competitive economic environments Africa is less efficient and less competitive than elsewhere in the world than those of comparator countries and Why are the Middle East and North benchmark regions. On the World Economic Africa’s urban areas fragmented, its people Forum’s Global Competitiveness Index, most still stuck in place, and its economies walled of the region’s countries underperform for off from each other and from the world? their income level (Schwab 2017), as further These symptoms have historical roots in a discussed in chapter 4.3 legacy of centralized administration and inat- Most countries in the Middle East and tention to lagging areas, first under the North Africa have national strategies that Ottomans, then under the Western colonial aim to balance growth spatially, bringing powers (Brixi, Lust, and Woolcock 2015; economic activity to lagging areas—but at World Bank 2011). Since independence, most what cost? Even assuming that a given place- of the region’s countries have ineffectively based policy can meet its spatial equity and sought to address this imbalance with inter- redistribution goals, do those gains justify the ventionist social contracts that focus on state associated compromises in economic effi- planning and redistribution over market out- ciency? Such trade-offs presuppose careful comes. While keeping government central calibration. However, in the Middle East and and hierarchical, such social contracts make North Africa, the pursuit of spatial equity the state responsible not only for service pro- through central planning has skewed spend- vision but also for other aspects of welfare— ing toward risky supply-driven interventions including much employment (Yousef 2004). that do not yield compensating benefits for Yet because natural resource rents and for- growth or inclusion. Celebrated successes— eign aid have limited the need to collect tax such as Dubai and Tangier—are isolated revenues, citizen participation in governance exceptions that prove the rule: all have inher- is low across the region (Mills and Alhashemi ent geographic advantages, and all receive 2018). In short, institutions remain weak unusually well-coordinated support. even as the public sector dominates markets. The weakness of national institutions National governments in the Middle East across the region’s countries—combined and North Africa have long used public funds with the state’s activist economic role and and subsidies to bet heavily on sectors and compounded by high political risk—is deter- industrial locations—and these development ring outside investment and thus hobbling bets shape markets. More than elsewhere, economic growth and job growth. On the O v e r v i e w   7 World Bank’s World Governance Indicators, population, predominantly local, and Middle East and North Africa countries lacking in economic mobility. score lower across the board than other •  As long as the region’s countries are m iddle-income and high-income regions.4 ­ walled off— disconnected from their Executives in the region perceive corruption regional neighbors and from the world— and inefficient government interventions as their cities cannot fulfill their potential: the most significant barriers to doing busi- cities need large markets. ness there. And in the eyes of the world, Whereas people throughout the Middle despite high financial openness, the Middle East and North Africa still look to the public East and North Africa is economically a sector for social services and for jobs, the pariah region: its countries receive the lowest region would benefit far more from institu- inflows globally of foreign direct investment tional reforms to integrate markets—creating (FDI) relative to gross domestic product incentives for private investment, skill acqui- (GDP).5 sition, and urban specialization in tradable Across many of the region’s countries, a goods and services. Short of this set of ideal thriving informal sector attests to the eco- reforms, the region’s economies can realize nomic potential of the poor. The poor have some benefits by better aligning government few roles in the formal economy and no real interventions with policies for efficiency and stake in the formal sector, which is largely competitiveness (see chapter 6). This report overseen by the state. The vast importance of presents findings on many specific aspects of informal employment to the poor points to the region’s economic geography (box O.3). failed policies and to the urgent necessity of For regional policy makers, the challenge is institutional reforms. Without new policies to confront the picture thus revealed and to that enable formal markets to function more give due weight to this evidence in their efficiently, the energies of the poor will not be deliberations. turned to more economically productive uses, The moment is critical. Choices today can and the Middle East and North Africa will set the next generation on a course either not achieve shared prosperity.6 toward more spatially convergent economic When will the region’s countries begin to growth—or toward further spatial interven- see faster progress on economic growth and tions that, in many cases, are likely to fail. spatial inclusion? Not until governments see and harness the economic potential of all and enable factor markets while expanding them Fragmented cities through regional integration. Otherwise, the To document the institutional constraints following will occur: on economic growth and inclusion, this •  As long as the region’s cities and city sys- report looks at the Middle East and North tems remain fragmented— with masses Africa across three spatial scales: within cit- of potential workers constrained from ies, within countries, and across national accessing opportunities and unable to boundaries. Why start with cities? Not just enter the formal economy—they cannot because cities drive economic growth glob- form the economically dense, efficient ally but also because the region’s urban popu- agglomerations that equip them to create lation share is among the highest in the world productive jobs and produce goods and (box O.4). services for regional and global markets. In the Middle East and North Africa’s cit- •  As long as the region’s education systems ies, the built-up area is on average quite spa- keep producing stuck people—locally tially dispersed—a fact likely to increase credentialed workers with limited trad- future development costs and thus limit pro- able, portable skills—the regional labor ductivity and welfare. The cost of this frag- force will remain small relative to the mentation is borne by all city residents, but it 8    C ONVERGEN C E is especially costly to low-income households. government development Masā kin Ùthmā n Take the Abdallah family, forced to move on the outskirts of Cairo (TADAMUN from the home they built in Ìzbit Khayrallah 2015).7 The move proved hugely expensive (a well-located and dynamic informal neigh- for the Abdallahs. It involved transport costs borhood in Cairo’s center) to the new to access current jobs; service costs for BOX O.3 Many signs point to one problem: The Middle East and North Africa’s economies are moving insufficiently Many recent economic findings, reviewed in this policies have distorted factor markets. Specifically, report, raise pointed questions about the Middle East place-based policies have shaped inefficient land and North Africa’s slow economic and job growth— markets and discouraged labor-intensive job cre- and the relation of these indicators to national poli- ation, while spatial equity policies and a reliance on cies. For example, why do the region’s countries public jobs have reinforced a local orientation in the receive the world’s lowest net inflows of FDI relative labor market and reduced the regional demand for to GDP?a And why are people with university degrees tradable skills. For the region to achieve faster eco- in the Middle East and North Africa no more likely to nomic growth—and more convergent growth, with be hired away from their places of origin than people economic mobility and shared prosperity—national with only a primary education?b governments will need to reduce these distortions. Underlying these and many other symptoms is a basic problem that the region’s governments have a. The question of low FDI inflows is examined in chapter 3. yet to face: for generations, national institutions and b. For discussion of the returns on education, see chapter 2. BOX O.4  Drivers and results of high urbanization in the Middle East and North Africa Urbanization in the Middle East and North Africa 43  percent in Egypt to around 80 percent in the has been rapid, driven by economic development as most urbanized countries of the GCC and about well as by environmental and political crisis. The 87 percent in Lebanon.b Among three major Middle region’s urban population increased fourfold from East and North Africa subregions—the Mashreq, 1970 to 2010, and forecasts predict that it will double the Maghreb, and the GCC—the Maghreb now has again between 2015 and 2050 (UN-Habitat 2012). the lowest national urban growth rates (less than The pace of urbanization partly reflects economic 2 percent a year on average). development, geography, and migration to oil-rich The Mashreq is seeing faster urban growth, driven countries, especially the GCC countries. But it also in part by refugees. Economic migration to the cit- reflects drought and conflict: in 2018, the Middle East ies of the Mashreq and the Maghreb has been chiefly and North Africa contained an estimated 7.2 million internal and partly climate-driven as migrants are refugees, 10.5 million internally displaced persons, pushed out of rural areas where agriculture is suf- and about 15 million economic migrants.a fering from higher temperatures (UN-Habitat 2012). T he region’s cou ntries never theless display By contrast, migration to the GCC “city-states” (such varying urbanization shares and urban growth as Bahrain, Kuwait, and Qatar) has come mostly rates resulting from varying confluences of fac- from abroad, especially during the oil price boom of tors. Today’s urban population share ranges from 2003–13. box continues next page O v e r v i e w   9 BOX O.4  Drivers and results of high urbanization in the Middle East and North Africa (continued) FIGURE BO.4.1  The Middle East and North Africa is a highly urbanized region 100 90 80 OECD Urbanization (%) 70 60 East Asia and Pacific 50 40 30 20 10 0 Middle East and GCC Mashreq Maghreb North Africa Source: World Development Indicators Database 2016. Note: “Urbanization” is measured as the percentage of total population living in urban areas. The orange dashed line designates the Organisation for Economic Co-operation and Development (OECD) average. The yellow dashed line designates the East Asia and Pacific average. Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Jordan, Lebanon, the Syrian Arab Republic, and West Bank and Gaza; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. a. Refugee data from the UN Refugee Agency (UNHCR) 2018 database: http://reporting.unhcr.org/node/36. b. In some countries, the definition of what constitutes “urban” contributes to significant disparities in reporting of urbanization versus agglomeration index levels—the latter being a metric of urbanization that classifies economic density using standardized criteria for population density, the population of a “large” urban center, and travel time to that urban center. amenities that were limited and thus more afford formal housing, in part because plan- expensive in the new city; and the cost of los- ning and building standards are rigid and ing the family’s social network. Significant outdated. In Cairo, formal housing units are hardships resulted. reported to be 20–30 percent vacant—even Although cities in the Middle East and as 70 percent of the city’s population occu- North Africa are not equally fragmented pies informal housing (World Bank 2012). (box O.5), many are quite fragmented in pop- The spatial fragmentation of many cities ulation density and the layout of physical in the Middle East and North Africa reflects structures, especially when controlling for both long-term and short-term growth pat- population.8 terns. Thus, Amman’s layout has become Within a city, spatial dispersion and frag- more linear with the expansion of Zarqa, a mentation make networked infrastructure nearby industrial town. But low-density and service provision costlier. They also informal settlements have also increased spa- make job matching less efficient and formal tial fragmentation in Amman as well as in housing more difficult to provide. In a highly other refugee destinations, such as Baghdad. fragmented city, firms are less likely to In the long term, modernist urban plan- quickly find people with the right skills. And ning in the Middle East and North Africa has low-density neighborhoods may indicate lost increased spatial dispersion, fragmentation, opportunities for infill development, contrib- and infrastructure costs—a pattern that con- uting to dysfunction in the formal housing tinues today. The low-density urban plans of market. For example, most Egyptians cannot the modernist era assumed horizontal 1 0    C ONVERGEN C E BOX O.5 In the Middle East and North Africa, some cities are more spatially fragmented than others In the Middle East and North Africa, Casablanca panels a and b). In contrast, Amman and Tripoli and Baghdad have a fairly high potential for inter- have ­l inear layouts and less concentrated popula- action among residents: people are more spatially tions (figure BO.5.1, panels c and d). concentrated than fragmented (figure BO. 5.1, FIGURE BO.5.1  Fragmentation varies significantly across urban areas of several capital cities in the Middle East and North Africa a. Casablanca, Morocco (4 million population) b. Baghdad, Iraq (10 million population) c. Amman, Jordan (3 million population) d. Tripoli, Libya (1.5 million population) Source: Developed from LandScan Global 2012 dataset, Oak Ridge National Laboratory, https://landscan.ornl.gov/. Note: Fragmentation is represented by the distribution of population within urban areas. (Each square represents population density per square kilometer.) For a detailed analysis, see chapter 1. expansion and private car travel. Globally, districts. And urban expansion in the GCC is many cities now reject such plans and pro- nearly as likely to occur through leapfrog mote density through vertical and infill devel- development as through infill—while in the opment. Yet most of the region’s cities have Maghreb subregion, infill development is not adopted policies for density. also rare, and extension development is the In the GCC subregion, city centers are norm (figure O.1). Leapfrog and extension being replaced with business and commercial development both tend to increase a city’s O v e r v i e w   1 1 infrastructure costs—leapfrog development FIGURE O.1  Aggregated 1990–2004 urban expansion trends of more so. capital cities in the Mashreq, Maghreb, and GCC subregions show In the short term, the wars and turmoil divergent patterns that have shaken the region since 2010 are 100 4 4 transforming many cities, with consequences 13 for their spatial forms and patterns of service 90 delivery. As with past conflicts, such as the 80 Lebanese civil war (box O.6), recent waves of war and unrest have altered urban footprints. 70 Cities in the Mashreq subregion have seen 60 69 72 large informal settlements come into being as Percent 66 50 people flee conflict—in Iraq, Fallujah and Ramadi are two of the country’s fastest- 40 growing urban centers—or as people spill 30 over from refugee camps, such as the 58 established by the United Nations Relief 20 and Works Agency for Palestine Refugees in 28 10 23 21 the Near East (UNRWA) for more than 1.5 million such refugees in Jordan, Lebanon, 0 the Syrian Arab Republic, and the West Bank Mashreqa Maghrebb GCCc (without Manama) and Gaza (Serageldin, Vigier, and Larsen 2015). In recent years, Jordan’s population Infill Extension Leapfrog growth rate has more than doubled with the Sources: Datasets from the European Commission Joint Research Centre’s Global Human Settlement refugee influx. Satellite photos show how the Layers. See Annex 1D for sensitivity analysis. Note: The analysis uses the Landscape Expansion Index developed by Liu et al. (2010), in which extension city lights of Amman have visibly spread and and leapfrog expansion increase infrastructure costs, with leapfrog having a higher cost impact. brightened (see chapter 1, box 1.1). a. Mashreq comprises the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic. b. Maghreb comprises Algeria, Libya, Morocco, and Tunisia. For Morocco, Casablanca was analyzed Across the Middle East and North Africa, instead of the capital city, Rabat. informal settlements are spontaneously c. GCC = Gulf Cooperation Council, comprising Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. Manama, the capital of Bahrain, was excluded because of the bias created by stitching together the gaps in the urban fabric its geographical constraints and resulting in an exceptionally high share of infill urban expansion. created by modernist city plans—offering a possible model for formal, planned densifica- tion. All but invisible to their governments, geographic mobility: its people are stuck in the poor build economies of their own. Take place, both spatially and economically. the informal areas of Greater Cairo, where Socioeconomic inequality between areas of about 12 million people lived in 2010: these most of the region’s countries far exceeds that areas spatially resemble the city’s oldest seen in countries with comparable GDP neighborhoods (photo O.1). Often viewed as around the world (figure O.2). This excep- relics of medieval squalor, such compact tionally high spatial inequality can be related neighborhood structures may in fact repre- to low spatial mobility. Being located far sent an approach to reduce future urban spa- from a large city also increases business tial fragmentation and limit negative development constraints in the region’s coun- e x t e r n a l it i e s s u c h a s a i r p ol lut io n tries (figure O.3). (UN-Habitat 2012). In most parts of the world, people vote with their feet and move toward opportunity, but this happens less in the Middle East and Stuck people North Africa (figure O.4). While fragmented cities limit efficiency Similarly, around the world, higher edu- and thus reduce workers’ economic pros- cation generally increases spatial mobility— pects, the Middle East and North Africa’s but not in the Middle East and North high inequalit y may also ref lect low Africa, where earning a university degree 1 2    C ONVERGEN C E BOX O.6  Urban fragmentation as a legacy of conflict: Today’s polycentric Beirut In 1975, civil war splintered Beirut into sectar- owners’ religion and ethnicity. The result today is a ian neighborhoods. The city center, once vibrant, polycentric city with a Christian sector, a Muslim became a no-man’s-land. Economic activity relo- sector, and the now reoccupied historical center cated to new areas that were defined by firm (figure BO.6.1). FIGURE BO.6.1  After decades of conflict, Beirut became a polycentric city Source: Hanna 2016. Note: The darker the square, the higher the population density. PHOTO O.1 In Greater Cairo, recent informal settlements share a basic spatial structure with medieval neighborhoods a. Aerial view of Bab el Wazir (established 800 years ago) b. Aerial view of Fostat Plateau (established informally in the 1980s) Sources: UN-Habitat 2012, from Sims 2010; Google Earth satellite images. O v e r v i e w   1 3 FIGURE O.2 Inequality within most Middle East and North Africa countries exceeds that of global peers Level of inequality in relation to GDP per capita a. By GDP per capita b. By urban share of population 30 30 YEM 2014 YEM 2014 Between-region inequality (%) Between-region inequality (%) 20 IRQ 2006 20 IRQ 2006 EGY 2012 IRN 2014 EGY 2012 IRN 2014 IRQ 2012 IRQ 2012 TUN 2010 TUN 2010 TUN 2005 DJI 2012 DJI 2012 TUN 2005 YEM 2005 IRN 2009 YEM 2005 IRN 2009 MAR 2000 MAR 2000 10 10 JOR 2008 JOR 2008 JOR 2010 MAR 2006 JOR 2010 MAR 2006 JOR 2006 LBN 2011 JOR 2006 LBN 2011 0 0 10,000 20,000 30,000 20 40 60 80 100 GDP per capita (US$, PPP) Urban share of the population (%) Middle East and North Africa Rest of the world Source: Middle East and North Africa Poverty (MNAPOV) database, Team for Statistical Development, World Bank. Note: Each point corresponds to a country’s data for a particular year. (Countries represent all income levels. For a list of countries and survey years, see chapter 2, annex 2A.) Inequality between subnational regions was calculated based on a country’s first administrative level (for example, governorates, provinces, and so on). PPP = purchasing power parity. does not make a person more likely to FIGURE O.3 In the Middle East and North Africa, companies migrate (figure O.5). One explanation is located on the periphery face harsher constraints on business that, in this region, higher education confers development than those in the capital city Percentage-point change in probability of factor being a major constraint to credentials for coveted public sector jobs, company on the periphery instead of in the capital not the tradable skills in demand by the pri- vate sector (box O.7). Notably, women are Access to finance Political instability more likely than men to migrate in Algeria, Electricity Egypt, Jordan, Lebanon, Morocco, and Labor skills Tunisia, which shows that female migrants Corruption in those countries are on average more likely Telecommunications Tax administration to be employed at their destinations (as fur- Crime ther discussed in chapter 2). Access to land Across the Middle East and North Africa, Transportation rural residents face substantial barriers to –10 –5 0 5 10 15 migration or have preferences against it—or Percentage point change else they would have migrated already, Source: World Bank Enterprise Surveys, 2011–18. given the returns to welfare from doing so. Note: T-bars indicate 90 percent confidence levels. 1 4    C ONVERGEN C E FIGURE O.4  Within-country migration is lower in Middle East and North Africa countries (averaging 14 percent) than in the rest of the world (28 percent) 80 60 Internal migration rate (%) 40 20 0 Po an . Tu Iraq Le Ben e g il Ec han s Sl rtug a en l a za Ch ia ew b a In ui ue ne a a m ia Jo nia Su an kr n ba in B on El M eria ut lva o Th ela a C a nd M Ri d H alay ca du ia ua a er ay re d S ce zu ib in a, a T RB m ey te C ica St ile nz tes Bo ga ia ts nda M ana C wi M Pe a go u lia Af r itz gu r om ep ov a h do SwUru do Al raz G ra n i N m in do ne si U da So Sa xic Ir ric ta n G lan el eri ub on r Ar nis on s ne L pa U an a d h Ja urk e ai G iq rd n os ila a R R e Ta a al w e ab Ar pu Mo ni t, yp U Ve Eg Pa Sources: Arab Barometer Wave IV (2016–17) data; University of Minnesota’s Integrated Public Use Microdata Series (IPUMS) International database. Note: The internal migration rate is the number of lifetime within-country migrants per 100 population. Lifetime migration compares the place of current residence with the place of birth (within the same country). The data are from various years; see chapter 2, annex 2A, table 2A.1. The migration rates in Algeria, the Arab Republic of Egypt, Jordan, Lebanon, Morocco, and Tunisia are computed on the basis of the Arab Barometer Wave IV (2016–17) data; rates for other countries are from latest censuses available on IPUMS in which place of birth was available. FIGURE O.5  Globally, the probability of migration tends to rise An analysis performed for this work of sev- with education—but not in the Middle East and North Africa eral countries across subregions, using 2006–14 data, shows that living in a metro- 0.25 politan area brought welfare benefits inde- pendent of individual characteristics.9 The Marginal effect of migration on analysis studies individual characteristics probability of employment 0.20 and the returns to these characteristics from living in metropolitan areas of Djibouti (2012), Egypt (2012), the Islamic Republic of Iran (2014), Iraq (2012), Jordan (2010), 0.15 Morocco (2006), Tunisia (2010), and the Republic of Yemen (2014). The results show that people outside the metropolis were less 0.10 well-off not simply because of their age, gen- Less than primary Primary Secondary Tertiary der, education, ­m arital status, labor force Educational achievement status, water access, electricity access, or Middle East and North Africa World possession of a computer. Rather, a substan- Sources: Arab Barometer Wave IV (2016–17) data; University of Minnesota’s Integrated Public Use tial part of the welfare gap reflected higher Microdata Series (IPUMS) International database. returns to these individual characteristics in Note: Estimations employ a Probit model with controls for gender, age, marital status, metropolitan areas. Similarly, another anal- employment status, urban or rural residence, education level, and country fixed effects at 90 percent confidence intervals (dashed lines). Migration refers to lifetime internal migration, ysis suggests that across the entire region, if which compares the place of current residence with the place of birth (within the same country). all people were living in capitals—where the In this figure, “Middle East and North Africa” comprises Algeria, the Arab Republic of Egypt, Jordan, Lebanon, Morocco, and Tunisia. returns on their characteristics would be the O v e r v i e w   1 5 BOX O.7  Low spatial mobility—and high public employment—among university graduates suggest that the Middle East and North Africa’s higher education systems do not impart tradable skills A u n iver sit y e duc at ion i n M idd le E a s t a nd This suspicion is strengthened by employment North  Africa countries confers locally valuable data. In the Middle East and North Africa, a tertiary c redentials—but does it provide marketable skills? ­ degree increases the probability of getting a public Not according to international evidence. Globally, sector job far more than the probability of working spatial mobility among the educated is a sign of in the private sector (figure BO.7.2). If private sector human capital: people with tradable skills are likely jobs are lacking for graduates, one reason is that firms to move to places where those skills are in demand require complementarity between human and physi- (figure BO.7.1). Conversely, the low spatial mobility ­ cal capital investments; they will not invest in a place of the region’s university graduates suggests that their that lacks a trained, skilled labor force. human capital remains low. FIGURE BO.7.1  Human capital in the form of tradable skills increases spatial mobility Credentialistic Human capital education Mobility Immobility FIGURE BO.7.2 In the Middle East and North Africa, tertiary education diplomas are highly valued in the public sector Achievement a. Effect of education on Achievement b. Effect of education on probability of working in public sector probability of working outside public sector 0.4 0.4 Marginal effect of education Marginal effect of education 0.3 0.3 0.2 0.2 0.1 0.1 0 0 Primary Secondary Tertiary Primary Secondary Tertiary Educational achievement Educational achievement Source: Arab Barometer Wave IV 2016–17 data. Note: Marginal effects use Probit regressions controlling for urban or rural residence, gender, age, number of children, marital status, and country fixed effects. T-bars indicate 95 percent confidence levels. The regressions were calculated for six countries in the Arab Barometer data: Algeria, the Arab Republic of Egypt, Jordan, Lebanon, Morocco, and Tunisia. 1 6    C ONVERGEN C E highest—per capita consumption would skills (because of the distorted incentives increase by 37 percent (D’Aoust and Lall, shaping education systems) but also the forthcoming). nearby markets or infrastructural links to People in remote areas of Middle East and markets that they would need to succeed in North Africa countries—especially in the business. If not for barriers to migration, Mashreq and GCC subregions—are stuck many more of these people would have partly because their countries lack secondary moved to capitals already. cities and dynamic private sectors: they live in So why do many of the region’s countries the long shadow of the capital. Outside the lack secondary, midsize cities? One reason is Maghreb subregion, the Middle East and another sort of barrier: the “thick borders” North Africa is dominated by its metropoles. that wall off countries, blocking trade and People far from cities lack not only tradable investment. FIGURE O.6  Urban populations in the Mashreq and GCC subregions are highly concentrated in the largest city and even more so in large cities (of more than 1 million people) a. Share of urban populations in largest city 40 population in capitals (%) 35 Average share of urban 30 25 20 15 10 5 0 ric d ci d l A and ia eb be d C eq a Af an Pa an ib an ric C As hr hr a a fic an G Af si th t ia tra e ar a ag as or as h en p C ic As ut C ro M M e er N eE So Eu th Am st dl Ea id tin M La Regions Subregions b. Share of urban populations in large cities (exceeding 1 million people)a Average share of urban population 60 in cities > 1 million (%) 50 40 30 20 10 0 be d ric d ci d l A and ib an Af an Pa an ia ric n C eq eb Af ara an a fic a C As hr hr si a ar a th t ia tra e G or as C ic en p h as ag As h C uro Sa e er N eE ut M M th Am st So b- E dl Ea Su id in M t La Regions Subregions Sources: 2015 data from United Nations Department of Economic and Social Affairs (UN DESA) World Population Prospects database: https://population.un.org/wpp/. Note: For subregional figures, the estimates of a few countries were modified, considering the population of the whole agglomerations (which was not always the case in the World Development Indicators dataset). Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Jordan, Lebanon, the Syrian Arab Republic, and West Bank and Gaza; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. a. The country-level figures were computed using either (a) cities of 1 million inhabitants or more, or (b) the share represented by the capital city in countries that had no cities exceeding 1 million inhabitants. O v e r v i e w   1 7 Walled-off countries FIGURE O.7  Few Middle East and North Africa countries trade electricity Trade with nearby countries confers economic advantages. But Middle East and North Africa countries are encircled by thick borders that constrain regional trade: onerous customs regimes and visa difficulties, high tariffs and nontariff bar- riers, regional infrastructure bottlenecks, and customs and logistics inefficiencies. L ack i ng de ep t rade ag re ement s , t he region’s countries are walled off from one another. Accordingly, firms in these countries trade far less than the region’s development levels might imply—and trade among countries within the region is strikingly low. These trade constraints are all the more striking given that the Middle East and North Africa is dominated by its largest cities, which eco- nomically should benefit from higher trade volumes. Across the Mashreq and GCC sub- regions, the share of a country’s urban popu- lation concentrated in its largest city—termed urban primacy—is among the highest in the world (figure O.6, panel a). And in these sub- Source: Ji et al. 2016, based on the United Nations International Trade Statistics Database regions, cities of more than 1 ­ million people (UN Comtrade) electricity trade statistics (http://comtrade.un.org). Note: Figure illustrates the Eurasian electricity trade network in 2010. Arrows indicate an electricity contain even higher urban population shares trading relationship between two countries. Full country names correspond to ISO 3 country codes. (figure O.6, panel b). Yet because the region’s Colors indicate seven “communities” as defined by Ji et al. (2016):  “A community in the electricity trade network consists of a group of nations tightly connected by electricity trade. Changes in one large cities lack access to regional or global nation have more effects on nations within the same community than on nations outside the markets, they fall short of their economic community.” The Middle East and North Africa countries are spread over four of these communities. potential. As a rule, large cities need large markets to In addition, when firms in the Middle East be productive and drive economic growth. and North Africa trade across borders, the That is because agglomeration effects presup- returns are lower than expected: firm-level pose scale. So, the specialized production data show that the region’s exporting firms tasks that make up links in global value are no more productive than the firms that do chains tend to be concentrated in large urban not export. And when its manufacturing agglomerations. In the Middle East and firms import critical intermediate inputs, North Africa, though, the economic reach of they pay a large and statistically significant large cities appears to be confined to domes- productivity premium (Francis and Schweiger tic markets. 2017). These patterns hint at an economic Although trade levels averaged across environment where access is privileged; the Middle East and North Africa seem where firms face barriers to entry, even if roughly in line with the region’s share of their operations may be more efficient; and global GDP, this statistic is deceptive: the where countries impose high barriers to buy- region’s trade consists disproportionally of ing goods from abroad. fuel exports. Trade in electricity, by con- Logistics inefficiencies appear especially trast, is less widespread in this region’s challenging—more so, for example, than countries than throughout much of the infrastructure (as discussed further in world (figure O.7). chapter 6). In 2018, most Middle East and ­ 1 8    C ONVERGEN C E North Africa countries fell in the middle or capital, and knowledge. Nearby countries lower range of the World Bank’s Logistics should accelerate their growth—creating a Performance Index of 160 countries. Among virtuous cycle of spillovers—and expanding them, three Maghreb states (Tunisia, regional markets should enable firms in con- Morocco, and Algeria) scored between 105 nected countries to benefit from scale econo- and 117. Exceptions included the GCC coun- mies.11 Because Middle East and North tries and Israel, which ranked fairly high Africa countries lack deep trade agreements across logistics indicators, followed by Egypt with their neighbors and trade facilitation is and the Islamic Republic of Iran.10 limited, the region’s large cities are missing A country whose major cities are cut off out on these spillover benefits (box O.8). from regional markets loses the opportunity Other regions—notably Europe and to reap positive spillovers from neighboring East Asia—have made regional trade a key countries. Ideally, a favorably endowed and driver of economic growth, while Middle fast-growing country with good policies East and North Africa continues to suffer should add to regional demand for workers, from high  restrictions on movements of BOX O.8  Large cities in the Middle East and North Africa show few spillover benefits from regional trade The spatial economic analysis in chapter 3 shows in  the world, what would happen to economic that the positive spillovers to the Middle East and output in a city from another region with greater North Africa’s large urban areas from regional trade spillovers if it were instead in the Middle East and are negligible (figure BO.8.1). In contrast, such spill- North Africa? A counterfactual for Bangkok is overs are positive across all countries globally—and revealing. If Bangkok were in the Middle East and strongly positive in East Asia and Pacific countries North Africa, its per capita GDP is estimated to with regional integration. drop from approximately US$12,000 to US$9,000 Given lower economic spillovers for neighbors in (figure BO.8.2). the Middle East and North Africa than elsewhere FIGURE BO.8.1  Spatial economic spillovers to large FIGURE BO.8.2  Bangkok’s per capita GDP would have city economies based on deep trade agreements with shrunk had it experienced the regional economic neighbors, by region spillovers of the Middle East and North Africa GDP simulation for Bangkok without regional economic 0.6 spillovers, 2000–15 0.5 Estimated spillover 0.4 13 coefficient (constant US$, thousands) 0.3 12 0.2 11 GDP per capita 0.1 10 0 9 l ric d ci d ric n be nd Al Af an Pa an Af ara ib a a fic a an th t ia ar a h 8 or as C ric Sa As N E e e b- th Am st e Su dl Ea 7 id tin M La Region 6 2000 2005 2010 2015 Source: Lebrand 2019. Year Note: The dependent variable is city GDP. Figure excludes Europe and Central Asia because the analysis of that region is less relevant to the Middle East and Bangkok, actual growth North Africa given the economic differences between the two zones. Also excluded Bangkok in Middle East and North Africa is South Asia because of the small number of observations for the region, rendering the results insignificant. Source: Lebrand 2019. O v e r v i e w   1 9 goods and people. With increased market and centralized public goods management, access, nonfuel exports could grow. Given assessing their implications for prosperity the size of many national economies in the and welfare. region, the regional service trade could be much higher. Countries could buy and sell Place-based interventions in the Middle much more electricity. Unfortunately, conflict East and North Africa are distortive and vested interests in many of the region’s while yielding low returns countries make comprehensive regional integration unlikely in the short term. ­ Public interventions in the economy may Because of thick national borders, Middle be institution-based, people-based, or place- East and North Africa cities generally lack based. Those that are institution-based sup- access to the large markets they need. To gain port governance, trade, and functioning such access, the region’s countries that border markets. Those that are people-based sup- the European Union (EU)—a large world port human capital, from health and nutri- market—can enter its supply chain through t ion to h ig h e r e du c at ion a nd s k i l l trade agreements. So far, Morocco and development. In contrast to both these types, Tunisia have benefited from this (although place-based interventions are often designed trade volumes are still low). The need for to improve development outcomes in particu- such agreements arises from the absence of a lar locations—whether directly through large regional anchor economy, as Western infrastructure investments or indirectly Europe was for Eastern Europe after 1990 t h r o u g h i n c e n t i ve s , s u b s i d i e s , a n d or  as Japa n (succeeded later by t he regulations. Republic of Korea) was for China and other Place-based interventions, while often East  Asian  neighbors. In the long term, well-intentioned to overcome market and regional integration should be a goal of coordination failures, typically distort factor Middle East and North Africa policy makers. markets, limiting economic efficiency by Right now, more trade with Europe does not reducing the power of demand to drive spe- imply less trade within the Middle East and cialization and agglomeration. Land markets North Africa; the two are complementary. are hobbled by heavy regulations on land use, which bottle up demand for both residential and industrial structures. Labor markets Place-based and centralized: shrink with disincentives to migrate and to How national policies and acquire tradable skills: Why specialize and institutions in the Middle East migrate for a relatively high-productivity job and North Africa perpetuate in a distant city when you can hope that the economic inefficiency and spatial state will one day lavish its largesse on your inequity rural province? And private capital follows How does the Middle East and North the lead of policy makers rather than seeking Africa’s fractured economic geography and meeting demand from domestic consum- today reflect the impact of state interven- ers and international business partners. tions in factor markets, combined with the In the Middle East and North Africa’s cit- regional legacies of centralized administra- ies, many neighborhoods and livelihoods are tion and service provision? How have cen- strongly shaped by place-based policies—and tral governments shaped the region’s land, not for the better. Favored policies include labor, and skill markets economically and heavy zoning and location regulations as well spatially at the city, country, and regional as the construction of modernist neighbor- levels? What does the persistence of highly hoods and cities from scratch—often deter- centralized government contribute to these ring formal redevelopment. In response, poor spatially and economically distor tive urban residents and migrants develop resi- dynamics? This section looks at national dential neighborhoods informally. The result preferences for place-based investments is the spatially and socially fragmented city, a 2 0    C ONVERGEN C E symptom of market and coordination failures secondary cities that can specialize and trade (figure O.8). with neighboring markets—as well as that of Within Middle East and North Africa’s efficient agglomerations, which large cities countries, place-based investments typically need if they are to serve large international loom large in the national budget—not the markets. Rather than fostering specialized usual pattern for middle-income countries. production centers and economically dense The place-based, or spatially targeted, inter- agglomerations to meet market demands, the ventions often used in the region include supply-driven initiatives used by the region’s governments to attract private sector invest- •  Physical infrastructure investments; ment have done the opposite. Take Tunisia, •  C apit a l subsid ie s a nd ot her f is c a l which used incentives to attract domestic incentives; and foreign firms such as Benetton to lagging •  Subsidies to places such as growth poles interior areas: as soon as the incentives and industrial districts; expired, many firms closed their operations •  Regulation of where industries may locate (World Bank 2018a). Generally, place-based production; and firm location initiatives in the Middle East •  Public sector industrialization, with cen- and North Africa have delivered few jobs, tral planning of investment locations. and many apparent success stories have Compared with institution-based and proven elusive. p ­ olicies ­ eople-based policies, these place-based p Since the 1950s, as the region’s national consume an outsize share of country invest- governments have persisted in shaping mar- ment expenditures in the region (box O.9). kets directly through place-based policies, When resources flow so freely to place-based lagging areas have stubbornly remained interventions, vital investments in institutions behind. Governments initially adopted and people must be neglected or deferred. In activist territorial development policies to effect, governments place large bets on specific answer the needs and grievances of neglected locations and hope for a stroke of luck. Yet areas. Strongly spatial in their focus, such place-based interventions have a poor track policies had two main aims: to direct com- record in the region’s countries (box O.10). pensatory support to the areas left behind From a regional perspective, place-based and to create new geographic centers for investments discourage the formation of economic activity. But lasting successes— such as Dubai and Tangier (box O.11)—are few and far between. FIGURE O.8  Place-based interventions distort urban markets— In cities, within countries, and throughout which then fail the region, place-based policies have yielded low economic returns—yet in spite of lessons Fragmented like Tunisia’s, Middle East and North Africa Fragmented, opaque central government control development and countries continue to rely on such policies. underutilized land and of land and incomplete buildings (land market Indeed, several of the region’s governments land markets are increasing incentives and transfers for failures) investments in lagging areas, new cities, and industrial development zones without Onerous building and Mismatch in locations zoning regulations, of jobs and formal housing rethinking their approach. limiting building heights (land market failure); Why bet so much on policies with such and location of activities demand is met informally unpromising track records? Four reasons stand out: Preference for modernist •  Vested interests. Despite a spatial inter- Underutilized and ghost vention’s limited returns to the local and new cities through new city cities (land market failures) developments national economy, some groups may stand to gain considerably from it. O v e r v i e w   2 1 BOX O.9  Middle East and North Africa countries stand out in directing a large share of investment expenditures toward place-based interventions Across the Middle East and North Africa, a larger and North Africa spend less on spatially distortive share of public expenditures —including initial interventions and more on investments in institu- i nvest ment s a nd recu rrent ex pend it u res — a re tions and people. This pattern is illustrated in ana- channeled toward place-based interventions than lyzing expenditure distributions for one year for toward institution-based or people-based interven- several Middle East and North Africa and compar- tions. In contrast, states outside the Middle East ator countries for illustrative purposes. FIGURE BO.9.1  Government expenditure distribution in selected Middle East and North Africa countries, by spatial category, differs greatly from international comparators 100 Share of expenditure (%) 80 60 40 20 0 Moroccoa Egypt, Arab Jordanc Tunisiad Austria Spain South Africa Hungary Poland Rep.b Institution-based interventions People-based interventions 1. Broad-based governance and 2. Basic health, education, and related human capital improvements institutional reforms 3. Provision of basic public services 4. Skill development, worker training, wage subsidies Place-based interventions 5. Physical infrastructure for connectivity and to support local production 6. Subsidies and other incentives to capital 7. Growth poles, industrial districts, other location subsidies 8. Public sector industrialization and industrial location regulations Source: Government Finance Statistics (GFS) database, International Monetary Fund; government data for Middle East and North Africa countries (as described in notes below). Note: The GFS database does not include data for Middle East and North Africa countries except for partial data for the United Arab Emirates. For Middle East and North Africa countries, report team analysis was based on national data as described in notes a–d: a. Morocco information came from the 2017 Finance Law, specifically the “General Budget” and the “Public Establishments and Enterprises” segments. b. Egypt on-budget data came from the Financial Statement 2018–19, and the total expenses of the 48 economic authorities were retrieved from their 2016–17 financial statements from the Ministry of Finance website. c. Jordan information came from “Capital and Current Expenses (2018),” classified by all line ministries, reported by the General Budget Department of the Ministry of Finance. d. The Tunisia analysis considers the most recent five-year plan, the Development Plan 2016–20, and in chapter 4, further analysis is provided drawn from the 2017 Finance Law. •  Path of least resistance. Investments in territorial development policies in the vacant land may present policy makers Middle East and North Africa. with a course of least resistance, spark- •  Extrapolating from rare success stories. ing less opposition than reforms affecting Successes such as Dubai and Tangier may developed areas. inspire hope for replication. Yet the suc- •  Lack of credible evidence. Little concrete cess may not be clearly replicable. Or even information may be readily available on more likely, policies may not be carefully the failures (and the rare successes) of designed to account for challenges and complementary factors. 2 2    C ONVERGEN C E BOX O.10  Place-based investments amount to risky development bets—and the stakes are high Most place-based investments in Middle East and Even so, the region’s governments keep betting on North Africa countries have been supply-driven new cities and vast capital investments. Egypt allo- rather than demand-driven. As a result, they have cated 30 percent of its 2015–16 national built environ- led to inefficiencies. Consider the poor track record ment budget to new cities and zones that hosted only of Egypt’s 20th-century new cities. In the 23 new cit- 2 percent of the country’s population. In contrast, just ies created from 1979 to 2000, the total population 29 percent of that budget went toward existing cities today is less than 800,000—a fraction of the targeted and towns hosting 98 percent of the population (Sims 20 million—and some cities are still unoccupied (Sims 2015). And Saudi Arabia, with its balanced territorial 2015). Similarly, Saudi Arabia’s established “desert development policy, allocated about 30 percent of its cities” remain unoccupied or house a fraction of their 2017 budget to developing eight planned new cities target populations. (see chapter 4, table 4.2). BOX O.11  How to make successful spatial bets? Build on natural advantage Dubai and Tangier stand out as rare successes among well as Port Rashid (jointly managed by Dubai Ports the many spatial development bets made by Middle Authority until Port Rashid was closed). Today it is East and North Africa countries. What made these the biggest port between Rotterdam and Singapore two spatially targeted policies succeed? And what les- and dominates the Arabian Gulf. Based on its success, sons can the region’s policy makers take from them to the government launched many other free zones, each improve their own spatial bets? The answer, broadly, with institutional reforms oriented toward the specific is that both Dubai and Tangier have been blessed with industries Dubai sought to attract to its city. geographic advantages—and their governments intro- Likewise, the Moroccan government recognized duced institutional reforms, and public investments, substantial unrealized potential in Tangier, driven that enabled these cities to leverage those advantages. by its strategic geographic location at a historical Dubai benefited from its strategic location at the crossroads of the Middle East, Africa, and Europe. crossroads between continents. In the early 1900s— In 2002, the government of Morocco recognized even before the 1950s oil discovery in the United Arab that it could better leverage this locational advan- Emirates—Dubai was declared a free port, and Dubai tage. It established Tangier-Med Port Authority and Creek served as a trading post for deep-sea merchants. invested €1 billion to develop and expand the port and As demand for port services increased, the govern- develop the surrounding city and region. The project ment of Dubai embarked on an expansion of port was developed as part of an integrated framework, facilities that led to the establishment of Port Rashid supporting the development of four critical sectors: and later Jebel Ali Port at the end of the 1970s. • Transport infrastructure in roads and rail In parallel, the government launched Dubai’s first • Industrial and logistics free-trade zones free zone at Jebel Ali Port in the early 1980s. The zone • Training and education of the local workforce was governed by a legal framework that addressed a • Collaboration across levels of government agen- range of issues raised by investors and allowed local cies and private sector stakeholders to spur inter- as well as foreign investments to be made in a busi- national investment. ness environment comparable to any in the world for private sector development. This helped demand The Tangier area was also governed by an alterna- for port services to skyrocket at Jebel Ali Port as tive legal framework that addressed core institutional box continues next page O v e r v i e w   2 3 BOX O.11  How to make successful spatial bets? Build on natural advantage (continued) and business environment challenges elsewhere in Because the advantages of market access cannot the country. The development and expansion of the be spread equally across a territory, spatial initiatives port boosted manufacturing, resulting in 28 percent must be sensitive to a place’s distinct advantages employment growth over 2002– 04 and an annual and potential. Most important is to ask whether a increase of investment of 13.2 percent (Kulenovic place is suitable to agglomerations. Efficient econ- et al. 2015). Improvements in supply chain links, omies are not evenly distributed in space; rather, diversification of the local and regional economy, they are lumpy, with economic activity clustered in and production increases in manufacturing-intensive economically dense places— cities, leading areas, sectors such as the automotive industry all benefited and regional economic hubs—and places naturally from port growth. As a result, the region evolved situated at a crossroads between continents, such as from being one of Morocco’s most lagging in 2000 to Dubai and Tangier. one of the most leading today. Seen in a broader historical perspective, Public investment decisions and service the region’s governments have used place- provision in Middle East and North based policies to offer visible responses to Africa countries are still overly subnational grievances and citizens’ expecta- centralized—making spatial tion that the state will provide jobs. In a convergence through place-based region with some of the world’s worst rates of policy even more unlikely unemployment and labor force participation, Centralized decision making and budget- and with episodes of violent conflict and ter- ing can limit the efficiency of place-based rorism, governments have identified jobs and efforts to increase spatial equity. Spatial stability as central policy goals. Surveys show convergence policies are likely to fail that citizens across various Middle East and ­ w herever public investment allocation is North Africa countries have come to regard centralized—especially if budgeting is also the state’s main function as providing jobs— opaque, politicized, or unpredictable. not as providing services and not as provid- In  some Middle East and North Africa ing a voice to citizens (figure O.9). countries, sectoral and subnational invest- In practice, public policies that distort fac- ment expenditures flow to areas with the tor markets tend to undermine social stability greatest bargaining power, as in Egypt by inhibiting private sector and job develop- (World Bank 2012). Thus, grants and trans- ment. Policies to shape factor markets are fers are too often directed to leading rather likely to foster a vicious cycle of worse job than lagging areas. Even budget allocation and inclusion environments. Other distor- formulas risk being biased toward wealthier tions are less intentional—reflecting decisions areas, such as those with natural resources that are centralized, yet uncoordinated, and (World Bank 2016). insufficiently guided by a holistic view of Centralized service provision, too, can needs across areas. Unless territorial develop- limit spatial convergence—both within cities ment policy is guided by a strong territorial and within countries. Top-down arrange- planning framework (World Bank 2018b), ments typically cause service delivery to be lagging areas are unlikely to get the specific less efficient in places that are less connected help they need to develop their economies to the center, especially poor urban neighbor- through specialization and agglomeration. hoods (including informal settlements) or 2 4    C ONVERGEN C E FIGURE O.9  Respondents in surveyed Middle East and North Africa economies identified jobs enablement—not public service provision or citizen representation—as the most essential function of a democratic state Percentage of respondents selecting answers in multiple-choice format a. Responses, by economy b. All responses, urban versus rural Algeria Egypt, Arab Rep. Urban Jordan Lebanon Morocco Rural West Bank and Gaza Tunisia 0 20 40 60 80 0 10 20 30 40 50 Percent Percent Jobs Fair elections Public services Free speech Source: Arab Barometer Wave IV (2016) data. Note: Percentages add up to more than 100 because they reflect respondents’ selection of one statement from each of four sets of answers to the question (Wave IV, no. 515): “If you have to choose only one from each of the four sets of statements that I am going to read, which one would you choose as the most essential characteristics of a democracy?” Categories shown summarize each choice described. lagging areas of countries. In the Middle East schools— can reduce costs and improve and North Africa, most public services access (Boex et al. 2016; Chambers, Wild, remain overly centralized, including health and Foresti 2013). For such areas, policy and education (though the United Arab makers need to expand the connectedness of Emirates is a rare exception in education providers and clients: for instance, they could [World Bank 2018d]). improve accessibility to services through dis- When different neighborhoods or areas tance learning with radio and cellphones. contain citizens with heterogeneous prefer- Other options include having nonstate ser- ences, these differences make centralized vice providers fill service gaps and training administration even more inefficient. community members to deliver services Whether within cities or within countries, themselves. For water and electricity provi- residents of different areas may differ greatly sions, off-grid or off-network solutions, such in their preferences. If so, efficient service as local independent water providers or local delivery will require local involvement, as generators for power provision may be emphasized in World Development Report more feasible because of the different levels 2004: Making Services Work for Poor of scale  sensitivity of different types of People (World Bank 2003). The same is true infrastructure. for subnational areas. Although Middle East and North Africa Further, decentralized service provision countries are rightly concerned about spatial can allow for experimentation and for lever- disparities in service provision, the prevailing aging a variety of service delivery models response—territorial development policy that allow, through the use of digital tech- directed from the capital—may be perpetu- nologies and alternative institutional ating poor outcomes and distrust of the state. arrangements, more efficient service provi- Empirical studies link citizens’ trust in gov- sion to marginalized populations. For low- ernment to their perceptions of government density areas, providing mobile services—for performance (Brixi, Lust, and Woolcock example, mobile health units and mobile 2015). O v e r v i e w   2 5 Recently, countries including Lebanon and North Africa governments may remain and Tunisia have embraced initiatives to tied for the short term to spatially targeted decentralize local services—a welcome chal- initiatives—that is, to place-based policies. lenge to the Ottoman and colonial legacy of To contain the economic damage from these centralized administration. Still, the bulk of policies, the region’s decision makers can, service provision in the Middle East and through a new framework, impose criteria on North Africa that could be managed subna- place-based interventions that will make tionally is not yet decentralized and remains, them more efficient (though all are likely to at best, deconcentrated, with local decisions be distortive). assigned to remote arms of the central This begins with market size: picking the government. right policies for each place requires under- standing its existing and potential access to markets, which may be domestic or interna- Five transitional steps to reduce tional. Not all places have equal access to institutional inefficiency, speed domestic markets (including for labor and the Middle East and North skills), which are concentrated in economi- Africa’s economic development, cally dense urban agglomerations. And not and enable convergent growth all places are equally connected to external Can Middle East and North Africa gov- markets: some have natural advantages such ernments steer from today’s spatial symptoms as coasts or borders, while others have better and distortions—rooted in institutional inef- connective infrastructure. Because of varying ficiencies—toward a more productive eco- density and market access, policies that yield nomic landscape, with faster growth and a high returns in one place may have no prospect of spatial convergence? Yes, if policy impact—or even a harmful impact—in makers end their overemphasis on traditional another. place-based development, supply-driven Where place-based policies have already investments, and centralized investment been adopted, this framework would guide decisions and service provision. ­ decision makers toward supporting it with But that is not likely to happen all at complements—whether for markets and once. The following five transitional steps connectivity or for basic service provision. can help policy makers chart a gradual path And those complements should be priori- from territorial development policies that tized based on expressed demand from citi- are predominantly place-based and focused zens and firms on what is needed to make on addressing the spatial symptoms of weak those territories effective enabling environ- economies to ones that address the causes ments. Yet wherever possible, the approach of depressed growth. This would entail a should also involve transitional steps to shift away from single-sector, supply- reform institutions (such as the other four driven, place-based policies to an approach transitional steps outlined below). And it of first mapping a place’s inherent economic should focus on enabling agglomeration, advantages, then identif ying suitable migration, and specialization based on a interventions. place’s natural advantage (as discussed in box O.11). Thus, to enhance market access, locate industrial parks near large cities. Or Step 1: Adopt new, evidence-based identify a location’s most constrictive trade criteria to guide future spatial bottlenecks and then clear them with new interventions connective infrastructure, institutional Ideally, broad institutional and people- reforms. Or both. based reforms would make cities more effi- When considering whether to retain tradi- cient, people more mobile, and countries tional supply-driven, place-based policies or more connected. But in practice, Middle East to shift toward this approach, countries 2 6    C ONVERGEN C E BOX O.12  Remembering the forgotten: Institution-based policies for the urban poor The poor of the Middle East and North Africa are, productive potential within a city’s fragmented eco- all too often, the region’s forgotten people. They lack nomic ­ landscape—and should enable formal invest- formal housing and the credentials or connections ment in these hot spots, raising densities and loosening needed for a formal job. So, they live wherever they restrictive regulations. Informal settlements point to must—and work however they can. where labor is in demand and where potential formal The economic dynamism of the poor appears sector ­ productivity lies untapped. Policy makers can in the informal sector, especially among urban encourage formal investment in such neighborhoods migrants. Their restless economic activity attests to with ­ institution-based reforms: revising urban plans, their energy, effort, and ingenuity—but it is not eco- ­ r elaxing zoning regulations, and allowing higher nomically efficient. Informal markets are unlikely ­population densities. to generate the specialized scale economies that If the urban poor often seem invisible to policy spur faster, wider growth through trade in goods makers, the rural poor are even further from realizing and services. their potential productivity—especially those who live To bring poor urban residents and migrants into far from cities. In the Mashreq and GCC subregions, the formal sector and give them opportunity, coun- the scarcity of secondary cities means that many rural tries would first need to reform factor markets. One residents cannot access urban markets. More and reason that the region’s cities create few opportuni- more of them migrate to the metropolitan capital, ties for the poor in the formal sector is that these not simply because of conflict or climate change but cities have only weak links to regional and global because the investments needed at home are lacking. markets. So, governments should observe informal Once settled in cities, these internal migrants join the urban economies for hidden signs of demand and informal economy. should keep one principle firmly in mind: raise revenues, to make investment decisions spatial equity arises not from investments in and allocate expenditures, or to deliver local a place but from the economic growth that services. Lagging areas are often the least creates jobs for a country’s forgotten able to mobilize and manage local revenues, ­ people—wherever those jobs may be—paired because the central government wields more with policies focused on improving access to fiscal control over smaller municipalities basic services for people regardless of their than over larger ones. Further, transfers to location. Only faster economic growth, with these localities are often insufficient and lack transitional but positive steps toward more objective standards, transparency, and pre- efficient institutions, will allow the Middle dictability. This lack of budget authority, East and North Africa to harness the eco- together with the lack of local control over nomic energies of the poor (box O.12). service provision, can undermine the local execution and maintenance of place-based investments—the very investments that are Step 2: Devolve greater functional supposed to reduce spatial disparities. authority and resources for local In  Morocco, for example, a lack of local revenue generation and service capacity has caused many municipalities to provision to local governments execute less than 50 percent of their invest- Place-based policies to reduce spatial ment plans (World Bank 2018c). equity are often undermined by a lack of Further, centralized service delivery mod- local authority and resources—whether to els make it difficult to offer services tailored O v e r v i e w   2 7 to heterogeneous preferences of citizens in smaller (secondary) cities, agglomeration through the use of alternative delivery models and specialization require the efficiencies and digital solutions. The way that functional arising from high economic density, or the authority and resources are devolved can and geographic concentration of economic activ- should vary based on economic geography, ity (as discussed earlier in box O.2). Urban the current authorizing environment, and the economic density and efficiency generally are current institutional architecture and histori- high if development is spatially compact, cal trajectory. Such reforms, even if gradual dense with people, and transit-oriented—but and experimental, have proven most likely to low if development is sprawling, as with leap- succeed when they have ensured that the frog development and widely dispersed devolution of functional authority is accom- structures. panied by the critical complements of greater One way to make cities more efficient is to resources and capacity support. reduce local frictions—boosting demand and increasing economic density while also tak- ing care to monitor the externalities (side Step 3: Step away from credentialist effects) of economic and population density. education and toward schooling that Planners and regulators can attract firms to cultivates globally tradable skills invest in cities by reducing frictions such as Helping stuck people become more spa- •  Impediments to property acquisition and tially and economically mobile could make new construction (costs, height limits, the region’s countries more prosperous—and and density limits); more stable. Generally, people in underper- •  Challenges in local business registration forming areas around the world migrate to and licensing; jobs and opportunities. Their choices help to •  Limits on news and information; and manage excessive population densities in •  Obstacles to developing local business the lagging areas, and it can benefit their networks. ­ relatives in their places of origin. But in the Middle East and North Africa, All these frictions are pervasive in the fewer people migrate than would be pre- Middle East and North Africa and call for dicted from the welfare gains accruing to policy remedies. migrants. As noted earlier, this immobility Planners can also address friction affect- results in part from credentialist education ing households, especially formal housing that does not impart tradable skills. It also costs. And they can seek ways to internalize reflects the many frictions that inhibit labor the negative externalities, or social costs, of mobility—from land transfer fees to active urban density—for example, by levying con- controls. And it is compounded by strong gestion charges, repealing fuel subsidies, and local identities and identity-related divisions. improving traffic management. Efficient city plans must be able to reflect changing demand: if they cannot, frictions, Step 4: Renew the focus on nurturing negative externalities, and market failures urban agglomerations by streamlining will result. Governments must, for example, land transfer procedures and relaxing enable zoning and building regulations to zoning regulations in existing cities, evolve with changing social and economic lowering the regulatory barriers to their needs. Consider Jordan, where rigid zoning redevelopment and building regulations impose standards Making fragmented cities’ land markets that make housing unaffordable for all but more efficient will be critical for agglomera- the top 30 percent of the income distribution. tion and specialization—two dynamics that More than 70 percent of new housing devel- amplify factors of production and drive eco- oped in Jordan is done without a building nomic growth and jobs. Whether in larger or permit (CAPSUS 2018). 2 8    C ONVERGEN C E A city’s density creates positive externali- Egypt start trading more with each other ties in public service delivery, which tends to following a decrease in border restrictions. be more efficient with density—and less so As further discussed in chapter 6, a simula- with sprawl. Providing one cubic meter of tion exercise based on a quantitative eco- piped water costs US$0.70 –US$0.80 in nomic geography model examined the Tunisia’s dense urban areas, but it costs US$2 implications of reducing border crossing in sparsely populated areas. Similar differ- times from 50 hours to 20 hours. There are ences appear in education and health care major welfare benefits in South Sinai costs (World Bank 2014). And in Amman, (Egypt) and around Aqaba (Jordan) from Jordan, with its rapidly rising population, a increases in market access. Greater market World Bank study finds that the cost of new access can also introduce greater competi- infrastructure through 2030 will depend on tion by promoting technology upgrading whe t her u rba n g row t h is sprawl i ng and greater productivity, often facilitated (low  ­d ensity) or compact (high density). by cross-border investment. And it could The study concludes that sprawling growth— trigger the emergence of regional produc- the pattern of the recent past—will cost tion networks, leading to greater intrain- 14 times as much as compact growth (World dustry trade within the region and allowing Bank 2018b). Nevertheless, urban plans can- growth in one country to spill over to its not simply pursue density but must also neighbors. attend to service quality: where density is too Increasing the spatial mobility of people high, service quality can decline (a negative makes especially good sense for the Middle externality from congestion). East and North Africa, not just because of how migration benefits migrants and their families but also because the alternative— Step 5: Expand market access for cities growing productive jobs in lagging areas by thinning the “thick borders” that through place-based policy—is not broadly inhibit mobility across the Middle East feasible. Policies to increase both spatial and and North Africa, both for regional economic mobility in the region’s countries trade and for migration could adopt three central priorities: building Historically, Middle East and North skills and human capital, reducing frictions Africa cities were part of economically cen- in spatial movement, and finding ways to tral global trade networks. Many of these cit- manage social divisions. ies persist in modern times as large, often The most urgent priority for making peo- vastly populated urban areas. Yet with ple more mobile is, arguably, to reorient edu- today’s thick borders, their economic reach cation toward tradable skills (assuming first has been limited. Countries across the region that health care, water and sanitation, and today need to enhance links across national basic education are present). Next is to reduce borders—reducing tariffs and nontariff the frictions that hinder spatial and economic ­ barriers, easing movements of goods and mobility, including people—though at the same time they may •  Low job market information, with high also enact policies to strengthen domestic job search costs; markets. •  High fees for land or property transfers; Expanding the market regionally would •  High permit fees for construction; help firms in small countries benefit from •  Restrictive building regulations; and scale economies in production, including •  Government controls that limit internal better access to inputs. It is also likely to migration. support secondary cities that can be inter- locutors with cross-border economies. ­dentity—​ Finally, social divisions related to i Consider what can happen if Jordan and in some cases a strong local identity—can O v e r v i e w   2 9 pose the greatest challenge to policy makers. But the economic impacts of those efforts Where intractable social divisions prevent have been minor, especially when compared labor mobility, place-based policies could in with those in other world regions. Apart some cases be justified. from natural resource exports, trade and Some may see labor mobility as undesir- integration with the rest of the world have able, given the challenge the region’s cities also remained below potential, with such face in absorbing migrants (including refu- notable exceptions as Morocco’s automobile gees). But leaving people stuck is not a prag- sector. matic alternative. Efforts to induce growth in Regional integration and global trade are lagging areas face an even more unyielding complementary, so both should be pursued. challenge: lumpy economic landscapes are Both expand product markets: regional inte- persistent. For poverty to be rapidly and sus- gration also increases supply capacity in large tainably reduced, people must be able to part by enlarging input (factor) markets, move where the jobs are. and global integration facilitates investment and access to technology and know-how. Deeper trade agreements that the EU is nego- The prospects for regional tiating with Morocco and Tunisia could integration: Distant yet vital to ­ perhaps become templates for similar coop- the Middle East and North Africa eration between neighbors. This overview has recommended five tran- Ultimately, effective steps to closer sitional steps toward more efficient spatial regional integration will always require giv- investments and initiatives to promote faster ing up some sovereignty to regional institu- growth and eventual spatial convergence. tions that set and enforce certain rules. So In  the longer term, the Middle East and far, governments of the Middle East and North Africa should envision regional inte- North Africa have not been willing to trans- gration that will break down walls between fer regulatory powers to regional institutions countries; connect firms to larger markets; that would govern intraregional trade. But and foster ever-larger scale and agglomera- steps can be taken today that do not require tion economies that are specialized for trad- giving up sovereignty or giving it up only in able goods and services. All of the region’s limited areas. Such steps would help to build countries would gain from such integration— the trust and experience to promote more just as some are already gaining from trade comprehensive integration later. Initial exam- agreements with large markets elsewhere. ples involving different types of policy instru- Although the prospects for the Middle ments could initially involve just two or a few East and North Africa’s regional integration neighboring countries, as follows: may seem remote today, the subject bears dis- cussing because of the high stakes of regional •  Tariffs have mostly been removed within trade. Cooperation among the region’s coun- the Pan-Arab Free Trade Area, but trade tries today is strikingly weak, but not for any costs remain relatively high because of known structural reason. A more apparent nontariff measures and transport prices. cause lies in political tensions and violent Information about nontariff measures conflict. The upshot of today’s scant coopera- is scarce, but they mostly represent red tion is clear: the Middle East and North tape and cause major complaints. They Africa’s economies produce less than they contribute to the Middle East and North could by trading more—and more freely— Africa’s low scores on logistics perfor- with each other. mance. Transport prices would fall if Closer regional integration has been a more competition were allowed in logis- long-stated objective of the region’s countries, tics and restrictions to operations in and there have been a number of initiatives. neighboring countries were reduced. 3 0    C ONVERGEN C E •  Institutions could be established in limited elimination of regulatory barriers, and areas where contributions and benefits introduction of new customs regimes can be more clearly defined and a trans- suitable for logistics activities (Rouis and fer of some sovereignty is therefore more Tabor 2013). acceptable. For example, regional elec- •  R efor m cu sto m s reg i m es . Customs tricity pools, a regional transport author- p ro c e du r e s wou ld b e s t r e a m l i n e d ity, or, where willingness of countries is and customs regulations harmonized greater, even a competition authority or across borders with private sector par- dispute settlement mechanism would be ticipation (as in logistics), and border possible. control information would be shared •  Cross-border integration of physical internationally. infrastructure is generally quite accept- •  Fill infrastructure gaps. Connective able, but there is often little coordination infrastructure—perhaps on the emerg- in upgrading transport links on different ing model of a development corridor sides of a border. A clear, long-term vision anchored by transport links (roads, rail, for regional transport integration could and ports)—would involve not just paving guide future investments including, in the roads and laying rail but also ensuring an longer term, reestablishing or newly con- environment that benefits the wider econ- structing rail links in the region and con- omy, with incentives for development. solidating port operations where several •  Ensure compensation for relative losers. smaller ports currently operate. Gains from integration should more than suffice to compensate those who may end Although prospects for formal integration up worse off as a result of trade. may remain uncertain, spatially sensitive physical infrastructure investments that con- nect neighbors can create conditions for later growth to spread. Again, places with access Notes to large markets—whether in existing urban 1. The Maghreb subregion includes Algeria, agglomerations or potentially through Libya, Morocco, and Tunisia. The Mashreq regional and global trade—are the most subregion includes Egypt, Iraq, Jordan, f av o r a b l e t o s c a l e e c o n o m i e s a n d Lebanon, the Syrian Arab Republic, and West specialization. Bank and Gaza. Regional integration for the Middle East 2. These and other reasons are being further and North Africa can be broadly envisioned analyzed in forthcoming World Bank Middle as a six-point plan. East and North Africa reports. 3. The GCC countries are no exception: most •  Make the business climate more favor- score below their income peers. able and state functions more efficient. 4. Data from the World Governance Indicators Institution-based reforms to speed gov- Database, https://info.worldbank.org/­governance​ ernment processes—and to make doing /­​wgi/. business easier across the region—would 5. Data from the World Development Indicators be a prerequisite for integration. Database: https://databank.worldbank.org​ •  Reduce tariffs and nontariff barriers . /­source/world-development-indicators. Tariff reductions would be accompanied 6. Even middle-income citizens in Middle East and North Africa countries participate exten- by the removal of inefficient subsidies sively in the informal economy. Their (such as energy subsidies) with care to resources, too, could be used more produc- address regressive impacts of removal. tively if factor markets in the formal sector •  Liberalize the logistics sector. Private sec- were more efficient. tor logistics firms would replace today’s 7. Actual names of the interviewees in the inefficient state monopolies. Morocco TADAMUN (2015) report, “The Hidden offers a model with its logistics zones, Cost of Displacement,” have been withheld. O v e r v i e w   3 1 8. The potential-of-interaction method (De la De la Roca, Jorge, and Diego Puga. 2017. Roca and Puga 2017) measures the average “Learning by Working in Big Cities.” Review number of people within a given distance of E c on o mic S tu di es 84 (1): 106 – 42 . from any spot in the city. To capture the prac- doi:10.1093/restud/rdw031. tical likelihood of interpersonal interaction, Francis, D. C., and H. Schweiger. 2017. “Not So one can control for city population, and one Different from Non-Traders: Trade Premia in can also control for GDP—an inverse proxy the Middle East and North Africa.” Economics for the cost burden of commuting. Controlling of Transition 25 (2): 185–238. for both, cities in the Mashreq and GCC sub- Hanna, J. 2016. “Changing Realities: Traumatic regions have some of the world’s highest Urbanism as a Mode of Resilience in Intra-War potentials for interaction, making the average Beirut.” International Planning History across all Middle East and North Africa cities Society Proceedings 17 (3): 383–88. still quite high. But controlling for population Ji, L., X. Jia, A. S. F. Chiu, and M. Xu. 2016. alone (that is, disregarding the commut- “Global Electricity Trade Network: Structures ing advantage conferred by higher GDP), the and I mplications.” PL oS ON E 11 (8): average city in the region appears more e0160869. fragmented. Kulenovic, Z. Joe, Alexandra Cech, Drilon Gashi, 9. The welfare aggregate (in most countries Luke Jordan, Austin Kilroy, Megha Mukim, comprising total food and nonfood expendi- and Juni Tingting Zhu. 2015. “Six Case tures) was temporally and spatially deflated Studies of Economically Successful Cities: to account for price differences over time and What Have We Learned?” Companion Paper 3 between areas. for “Competitive Cities for Jobs and Growth: 10. See the World Bank Logistics Performance W hat , W ho, a nd How,” World B a n k , Index (LPI) 2018: https://lpi.worldbank.org/. Washington, DC. 11. Close proximity remains advantageous in Lebrand, Mathilde. 2019. “International Growth buyer-supplier networks despite the rapid Spillovers from Deep Agreements for Countries decline in global transport costs. and Cities.” Unpublished manuscript, World Bank, Washington, DC. Liu, X., X. Li, Y. Chen, Z. Tan, S. Li, and S. Ai. 2 010 . “A N e w L a nd s c ap e I nd e x for References Quantifying Urban Expansion Using Multi- Boex, Jamie, Ammar Malik, Devanne Brookins, Temporal Remotely Sensed Data.” Landscape and Ben Edwards. 2016. “Dynamic Cities? Ecology 25 (5): 671–82. T he Role of U rb a n S e r v i c e D el ive r y Mills, R., and F. Alhashemi. 2018. “Resource Performance in Africa and Asia.” Working Regionalism in the Middle East and North paper, International Growth Centre, London. Africa: Rich Lands, Neglected People.” Brixi, Hana, Ellen Lust, and Michael Woolcock. Brookings Doha Center Analysis Paper No. 2015. Trust, Voice and Incentives: Learning 20, Brookings Institution, Washington, DC. from Local Success Stories in Service Delivery Rouis, Mustapha, and Steven R. Tabor. 2013. in the Middle E ast and Nor th Afric a . Regional Economic Integration in the Middle Washington, DC: World Bank. East and North Africa: Beyond Trade Reform. CAPSUS (CAPSUS Sustainable Capital). 2018. Directions in Development Series. Washington, “Urba n Grow t h S cena rios. H ashem ite DC: World Bank. Kingdom of Jordan.” Project brief for the S c hw a b , K l au s , e d . 2 017. T h e G l o b a l World Bank, Washington, DC. Competitiveness Report 2017–2018. Geneva: Chambers, Victoria, Leni Wild, and Marta World Economic Forum. Foresti. 2013. “Innovations in Service Delivery: Serageldin, Mona, François Vigier, and Maren International Experience in Low-Density Larsen. 2014. “World Migration Report Countries.” Research repor t, Overseas 2015: Urba n M ig rat ion Trends i n t he Development Institute (ODI), London. M iddle  E ast and Nor th A frica Region D’Aoust, Olivia, and Somik V. Lall. Forthcoming. and  the Challenge of Conf lict-I nduced “ U n e qu a l Pl ac e s a nd S t u c k Pe opl e .” D i s p l a c e m e n t .” B a c k g r o u n d p a p e r , Unpublished manuscript, World Ban k, International Organization for Migration Washington, DC. (IOM), Geneva. 3 2    C ONVERGEN C E Sims, David. 2010. Understanding Cairo: The ———. 2016. “Syrians in the Middle East: The Logic of a City Out of Control. New York: Lives and Livelihoods of Refugees and Their American University in Cairo Press. Hosts.” Unpublished manuscript, World Bank, ——— . 2 01 5. E g y p t ’s D e s e r t D r e a m s : Washington, DC. Development or Disaster? Oxford: Oxford ———. 2018a. “From Disfavored to Unique: University Press. Tunisia’s Territorial Variety as an Asset. TA DA M U N. 2015. “T he H idden Cost of Policy Note on How Tunisia Can Leverage Displacement: The Move from Ìzbit Khayrallah Its Regions in a Differentiated Manner for to M a s ā k i n Ù t h m ā n .” R ep or t of t he Shared Prosperity.” Report No. 130184, TADAMUN Initiative, American University, P r o j e c t N o . 1 5 9 0 7 2 , Wo r l d B a n k , Washington, DC. Washington, DC. UN-Habitat (United Nations Human Settlement ———. 2018b. “Jordan—Urban and Municipal Programme). 2012. The State of Arab Cities Program for Balanced and Inclusive Growth: 2012: Challenges of Urban Transition . Concept State Program Information Document Nairobi, Kenya: UN-Habitat. ( P I D) .” C o n c e p t N o t e , R e p o r t N o . World Bank. 2003. World Development Report PIDC151796, World Bank, Washington, DC. 2004: Making Services Work for Poor People. ———. 2018c. “Maroc: Tirer avantage de Washington, DC: World Bank. l’urbanisation pour promouvoir un nouveau — — —. 2009. World Development Report 2009: modèle de croissance et réduire les disparités Reshaping Economic Geography. Washington, territoriales.” [Morocco : Take advantage of DC: World Bank. urbanization to promote a new growth model ———. 2011. Poor Places, Thriving People: How and reduce territorial disputes]. Policy Note, the Middle East and North Africa Can Rise World Bank, Washington, DC. Above Spatial Disparities. MENA Development ———. 2018d. “Unleashing the Potential of Report Series. Washington, DC: World Bank. Education in the Middle East and North — — —. 2012. “Arab Republic of Egypt: Reshaping Africa.” MENA Education Flagship Draft Egypt’s Economic Geography: Domestic f o r D e c i s i o n M e e t i n g , Wo r l d B a n k , Integration as a Development Platform.” Report Washington, DC. No. 71289-EG, World Bank, Washington, DC. Yousef, Tarik M. 2004. “Development, Growth ———. 2014. “Tunisia Urbanization Review: and Policy Reform in the Middle East and Reclaiming the Glory of Carthage.” Report, Nor th A frica since 1950.” Jour n al of World Bank, Washington, DC. Economic Perspectives 18 (3): 91–116. Fragmented Cities, 1 Constrained Growth U rbanization is strongly and positively cities—through either low-density or leapfrog linked to economic growth. This is development—increases the costs of providing because cities, by reducing the distance infrastructure and services to firms and popu- between people, allow ideas to spread easily, lations. Low-density, noncontiguous urban improve matching between employers and spatial development affects the costs of employees, enable firm specialization and the ­ networked infrastructure (such as electricity, emergence of value chains, allow for cost shar- sewerage, roads, and water) because it ing of expensive but growth-enabling infra- increases the distances that need to be covered structure, and, finally, create consumer markets by the infrastructure per capita. In capital-­ large enough to sustain scale economies. constrained economies, this can result in lower Although urbanization levels in the coverage or maintenance of these services. It is Middle East and North Africa are among the also likely to hinder the functioning of urban highest in the world, economic performance labor markets by limiting matching between has on average been disappointing and below employers and employees. average for countries with comparable urban- However, spatial fragmentation dis- ization rates. In particular, the share of urban plays high variations across subregions jobs in the tradable sector is surprisingly low of the Middle East and North Africa. compared with South Asia, East Asia and the While the Maghreb displays high fragmen- Pacific, and Europe and Central Asia, while it tation, the Mashreq on the whole has dense is on par with the Sub-Saharan Africa region. built-up areas in part stitched together by By largely limiting themselves to local mar- recent informal neighborhoods that authori- kets, firms in the Middle East and North ties have over recent decades sought to elimi- Africa impede returns to scale, economic nate, following the (somewhat unrealistic) growth, and job creation. dream cities of some urban planners.1 In the This chapter uncovers a complementary, Gulf Cooperation Council (GCC) countries, and possibly more important, explanation for state-of-the-art transport infrastructure— the poor economic performance in urban geared toward personal vehicles and financed areas of the Middle East and North Africa th roug h nat u ral resou rce revenues —­ region: the spatial fragmentation of its cities. currently overcomes urban fragmentation The underutilization of land in the region’s through speedy transport. 33 3 4    C ONVERGEN C E But none of these coping strategies encour- Over 1970–2010, the region’s urban pop- ages the densification of urban areas. They ulation almost quadrupled. Forecasts predict rely on fiscally intensive workarounds to mit- that, compared to 2015, it will double by igate (in some cases) the nascent costs of 2050, with the urbanization level reaching urban land-use fragmentation. This is both 71  percent (UN DESA 2019). 3 This rapid costly and unlikely to foster the economies of urban growth has been driven by economic agglomeration needed for firms of the Middle development, migration to oil-rich countries, East and North Africa to make urbanization drought, and conflict. In 2018, the Middle work for all. East alone hosted an estimated 7.2 million refugees, 10.5 million internally displaced persons (IDPs), and 15 million economic Rapid urbanization has not migrants.4 brought commensurate Urbanization varies across subregions, economic benefits to the Middle reflecting different drivers, as follows:5 East and North Africa •  The Mashreq is moderately to highly Urbanization is high across the region urbanized (from 43 percent in the Arab The Middle East and North Africa region Republic of Eg ypt to 87 percent in has become one of the most urbanized Lebanon). As a result of wars and ongo- regions in the world, with 65 percent of its ing conflicts, it shelters half the world’s population living in cities in 2018.2 Although registered refugees, adding to the pres- only 34 percent of these economies’ popula- sure on large cities (box 1.1). With lim- tions were living in cities back in 1960, the ited options for expansion in the larger region has been progressively closing the gap cities, governments are emphasizing with Europe and following the urbanization development of secondary cities and sat- rhythm of Latin America (figure 1.1). ellite locations. FIGURE 1.1  Urbanization in the Middle East and North Africa is catching up with Europe and Latin America, 1960–2015 90 80 70 Urbanization rate (%) 60 50 40 30 20 10 0 90 70 00 10 68 96 98 72 78 04 08 12 60 80 82 84 02 16 66 76 64 74 88 92 06 86 62 14 94 19 19 20 20 19 19 19 19 19 20 20 20 19 19 19 20 19 20 19 19 19 19 19 19 20 19 19 20 19 Middle East and North Africa Europe and Central Asia Latin America and the Caribbean South Asia Sub-Saharan Africa Source: World Development Indicators database. Note: The “urbanization rate” is the percentage of total population living in urban areas. F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    35 BOX 1.1  The impact of conflicts on urbanization in the Middle East and North Africa The wars and turmoil that have shaken the Middle More than 60 percent of the West Bank and Gaza East and North Africa since 2010 have also shaped population has migrated, and most of the migrants patterns of urbanization. They created an estimated have settled in neighboring Arab countries. a An 2 million refugees and over 6 million internally dis- estimated 1.5 million people from the West Bank placed persons (IDPs), most of whom have settled and Gaza are living in the 58 recognized United in large urban areas (Serageldin, Vigier, and Larsen Nations Refugee Agency (UNHCR)-run camps for 2014). The number and location of displaced popula- this population in Jordan, Lebanon, Syria, and the tions shift regularly in response to the changing lev- West Bank and Gaza. b Over the years, the popu- els of violence in Iraq, Libya, Sudan, the Syrian Arab lation has overflowed the camp boundaries and Republic, and the Republic of Yemen (UNHCR 2014). contributed to the formation of large informal set- A comparison of urban footprints (proxied by tlements in the urban centers of the region, espe- nighttime lights) in the Mashreq subregion reveals cially in Jordan and Lebanon. a clearly visible change between 2012 and 2015 as I nternal displacements since 20 03 from the luminosity increased in urban areas in Jordan and war in Iraq (more than 1.8 million people) have Lebanon and decreased in Syria (map B1.1.1). Jordan’s also affected urbanization patterns, leading to the population growth rate, which averaged 3.2 percent expansion of a few cities. Baghdad, one of the cit- over 2000–16, rose to 7 percent a year from 2011 ies closest to the conflict-affected areas, has been onward as a consequence of the large influx of refu- the major recipient of IDPs. Many other cities in gees (World Bank 2017). the country attract large numbers, contributing to Similarly, the ongoing crisis in the West Bank unplanned urbanization (Serageldin, Vigier, and and Gaza from conflict with Israel has led to uncon- Larsen 2014). In the A nbar region, 66 percent trolled urban expansion in neighboring countries, of IDPs have remained in the province, making while the Iraq war led to large waves of internal Fallujah and Ramadiyah the two fastest-growing displacement (Serageldin, Vigier, and Larsen 2014). urban centers of the country. MAP B1.1.1  Massive migration patterns to urban areas in the Mashreq region are visible through nighttime light changes a. 2012 b. 2015 Source: Images from the Visible Infrared Imaging Radiometer Suite (VIIRS) instrument of the National Aeronautics and Space Administration (NASA) Applied Sciences Program, 2015. Note: Map shows the intensity of light in urban areas of the region surrounding the war-torn Syrian Arab Republic (Damascus), including in the countries of Lebanon (Beirut) and Jordan (Amman), to which many war refugees have fled. a. West Bank and Gaza migration data from “Estimated number of Palestinians in the world by country of residence, end year 2010,” Palestine Central Bureau of Statistics website: http://www.pcbs.gov.ps/Portals/_Rainbow​/Documents/PalDis-POPUL-2010E.htm. b. West Bank and Gaza refugee data from “Palestine Refugees” web page of the United Nations Relief and Works Agency for Palestine Refugees in the Near East (UNRWA): https://www.unrwa.org/palestine-refugees. 3 6    C ONVERGEN C E •  The Maghreb has an urbanization rate In the Mashreq and Maghreb subregions, typically above 50 percent, but cur- urbanization has resulted largely from the rent urban growth rates are low (below internal migration of rural people seeking percent a year on average over 2000–16, 2 ­ better opportunities as rising temperatures as shown in figure 1.2). reduced ag ricult u ral produc tivit y by •  The GCC countries are among the rich- 10–40 percent (UN-Habitat 2012). Improved est and most urbanized in the world, in public services, better education, and a high part owing to environmental factors. concentration of jobs in urban areas have Several (Bahrain, Kuwait, and Qatar) are attracted rural populations into the main cit- like city-states, with an urbanization rate ies of these countries. The Mashreq area has of around 80 percent and a foreign-born also received large waves of war refugees who population of around 40 percent. moved within and between countries and settled in major cities, contributing to consis- The main drivers of urbanization vary tently higher urban growth rates than in considerably by subregion, leading to dif- Maghreb countries over the past 20 years ferent urbanization patterns. Urbanization (box 1.1). in the GCC “city-states” is driven predomi- The Middle East and North Africa nantly by external and internal migration region overall displays a spectacular urbaniza- to cities, whose development is fueled by tion rate. When measured with the agglomera- the wealth accumulated from oil exports. tion index—a metric of urbanization that uses This pattern resulted in spectacularly high consistent criteria for all countries (Uchida u r b a n g r o w t h r a t e s ov e r 2 0 03 –13 and Nelson 2008, see Annex 1A)—the Middle (5–8.5 percent a year on average), when East and North Africa is by far the world’s international oil prices were hig h most urbanized region, with an f igure  1.2). The GCC cities have thus (­ index of 68; followed by Europe and Central become major nodes in the flow of people, Asia, whose high agglomeration index seems global capital, and services. to be driven by high-quality transport FIGURE 1.2  Urban population growth rates in the Middle East and North Africa vary by subregion, 1990–2016 9 8 7 Urbanization rate (%) 6 5 4 3 2 1 0 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Mashreq Maghreb GCC Source: World Development Indicators database. Note: Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    37 FIGURE 1.3  The Middle East and North Africa displays the world’s highest urban concentration as measured by the agglomeration index 90 80 70 Agglomeration index 60 50 40 30 20 10 0 fic a be d eq ia eb l A and ric d C ric ib an Af an As C ci hr hr an a a Af G Pa si ar a as ag tra e th t h or as C ic en p ut M d M C ro e er N eE So an Eu th Am dl ia id As tin M La st Ea Regions Subregionsa Source: World Bank 2009b. Note: The agglomeration index is a metric of urbanization that classifies the economic density of economies using standardized criteria for population density, the population of a “large” urban center, and travel time to that urban center. a. Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. infrastructure; and Latin America and the FIGURE 1.4  Economic growth per capita has not kept pace with Caribbean, where it is driven by high urban urbanization in the Middle East and North Africa, 1990–2016 primacy (figure 1.3).6 The GCC and Mashreq subregions display even higher scores, driving 50,000 Saudi Arabia the high regional agglomeration index. GDP per capita (PPP 2011 US) Iran, Islamic Rep. Lebanon Algeria Yet the region’s economies have not Egypt, Arab Rep. reaped the full benefits of urbanization 10,000 Jordan Urbanization in the Middle East and North Tunisia Africa has not been accompanied by commen- Morocco surate economic growth. The relationship between urbanization and economic growth per capita in the region is below the trend observed in the rest of the world (figure 1.4). 500 Cities allow workers to be closer to jobs, 0% 20% 40% 60% 80% 100% increasing opportunities and fueling produc- Urban share of population tivity. They bring people together physically, All countries MENA facilitating the exchange of ideas and bringing about innovations. So, urbanization has usu- Source: World Development Indicators database. Note: Country curves represent the evolution of the urbanization-to-gross domestic product (GDP) ally gone hand in hand with sustained eco- per capita ratio in selected countries from 1990 to 2016. MENA = Middle East and North Africa; nomic growth, as measured by gross domestic PPP = purchasing power parity. 3 8    C ONVERGEN C E FIGURE 1.5  Cities in the Middle East and North Africa display a lower share of urban tradable employment than in other regions of the world 60 Share of tradable sector in urban 50 employment (%) 40 30 20 10 0 South Europe East Asia Latin Sub- Middle East Mashreq GCC Maghreb Asia and and America Saharan and Central Pacific and the Africa North Africa Asia Caribbean Source: Oxford Economics data 2015. Note: The nontradable sector is defined as “consumer services + public services” and the tradable sector as “agriculture + financial and business services + industry + transport.” Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. product (GDP) per capita. However, when Faced with limited demand, urban firms in compared with international standards, GDP the Middle East and North Africa benefit in the Middle East and North Africa is below from limited scale economies and, all else its potential. Although urbanization is indeed equal, create fewer jobs (Lall, Henderson, associated with economic growth, GDP per and Venables 2017). Conversely, specializa- capita remains, on average, below what could tion in tradable goods such as manufactured be expected for any given urbanization level products allows for a larger and elastic poten- based on the world average. tial demand that goes well beyond local mar- Notably, however, the patterns vary kets. Catering to such international needs widely by country. Egypt, for example, dis- implies making the most of firm-level scale plays astonishing economic growth with a economies and beyond, often boosting near constant urbanization rate, whereas agglomeration economies through a buoyant Algeria is associated with rapid urbaniza- ecosystem of local suppliers. In the Middle tion but slower growth of GDP per capita. East and North Africa, cities and their firms Among the possible factors in the lower risk being unable to reverse the current trend GDP per capita is the domination of the non- and break into tradables because cities are tradables (services) sector in the region’s too costly. Expensive cities typically require urban employment, which limits returns to wages that would compensate workers for scale, economic growth, and job creation. high urban costs and so would entail produc- The share of the tradable sector (such as tion costs above the international price for manufacturing) in urban employment is only any given good (Venables 2017). around 41 percent (figure 1.5), in line with Sub-Saharan Africa (42 percent) but well below South Asia (53 percent), Europe and Modernist planning and Central Asia (50 percent), and East Asia and informality play crucial roles in Pacific (49 percent). the fragmented urban fabric Specializing in nontradable goods means Urban compactness drives economic that the demand for the region’s urban out- growth by increasing the efficiency of match- put is local and necessarily limited in volume. ing jobs and workers. From a pure economic F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    39 standpoint, the ideal city can be viewed as an public services (again, such as hospitals) efficient labor market, matching employers require a minimum density to be economically and job seekers through connections (Bertaud viable (Burton 2000a). So compact cities pre- 2014). When the matching is efficient, cities sent some advantages, especially for low- benefit because by increasing the size and income households, many of whom lack access diversity of their labor force, employers and to a car and would find public transport to be job seekers are more likely to find appropriate more efficient in nonfragmented urban areas. matches that make the best of workers’ skills Newman (1992) has provided strong evi- and aspirations. As such, compact cities have dence that higher-density cities are associated the potential to reduce the separation with higher use of public transportation. between people and jobs by reducing com- Such impact also inherently promotes social muting time and cost (Burton 2000a). equity, because the disadvantaged rely pri- Fragmentation—understood as either low- marily on public transportation. A nd density development or spatial development although the investment needed to build pub- broken up by areas of underdeveloped land lic transport is high, the usage costs, provided use—implies higher costs to ensure access to sufficient ridership, are low compared with public networked infrastructure. To service large road infrastructure. Rode et al. (2017) fragmented urban areas with networked show that although the up-front costs of col- infrastructure (such as roads, sewerage, and lective transportation infrastructure are on water) is typically costlier per capita for a the same order of magnitude as—or even given service quality because it requires longer higher than—the costs of high-capacity high- and more intricate networks. For example, in ways, collective transportation has the edge the context of U.S. cities, Burchell et  al. when cost estimates take carrying capacity (2005) found that urban sprawl increases into account. Capital costs divided per capita local road lane-miles by 10 percent and public (after dividing by capacity) would be in the services and housing costs by 8 percent. range of US$300–US$500 for a bus rapid Besides, in financially constrained urban transit system, compared with US$5,000– areas, these high costs may be difficult to US$10,000 for a dual-lane highway. match with resources, decreasing service Finally, it is widely argued that compact quality. Libertun de Duren and Guerrero cities offer more opportunities for social inter- Compeán (2016) examined the relationship action and reduce segregation. Compactness between municipal spending on services and encourages interaction between urban dwell- population density in 8,600 municipalities in ers from different economic strata and origins Brazil, Chile, Ecuador, and Mexico. They (locals and migrants). Sprawling development, find that optimal expenditures for a given characterized by scattered gated communities quality of services are obtained for high pop- or informal development in the peri-urban ulation densities (around 9,000 persons per region, tends to increase residential segrega- square kilometer). The authors also find that tion (Zhao 2013). This sometimes gets rein- coverage is strongly and positively associated forced by unevenly distributed public services with population density. Urban planning for and unbalanced investments. higher density and lower land consumption can also decrease infrastructure costs, as fur- Do the region’s cities truly function as ther described below. cities? Urban fragmentation typically impedes public transport and other public services, Can cities in the Middle East and North because minimum densities are needed for Africa qualify as well functioning? Arguably, those infrastructures to be viable. In a com- a primary raison d’être of cities is to reduce pact city, households are not only close to the economic distance between people and facilities (such as hospitals) but also likely to economic opportunities and to reduce the enjoy more facilities per capita. Localized costs of providing networked basic services. 4 0    C ONVERGEN C E Cities thrive, in particular, because they which measures spatial fragmentation in the enable matchmaking—between people, urban built environment. between firms, and between people and job The Puga index was calculated for the cap- opportunities (Avner and Lall 2016). With ital cities of the Middle East and North Africa this framework in mind, we ask whether cit- using a methodology that measures the num- ies in the Middle East and North Africa ber of people an average person can interact enable matching and provide networked with in a specific city within a given travel dis- infrastructure at reduced costs. In other tance (De la Roca and Puga 2017).7 The words, are cities in the Middle East and results show important gaps in many cases. North Africa functioning as efficient cities? For example, while Casablanca and Baghdad Cities’ urban footprints and fragmentation display fairly concentrated population pat- vary considerably. Using satellite imagery, this terns with high interaction potential, Amman section examines the growth patterns of the and Tripoli rank much lower in terms of inter- region’s urban areas and their degree of spatial action potential because of their more linear fragmentation. Fragmentation is approached layout with higher fragmentation (figure 1.6). here through two different metrics: (a) the In Amman and Baghdad, although low-­ interaction potential index (­ hereafter called density informal settlements have contributed the Puga index), which measures fragmenta- to urban expansion, the urban footprint of tion in the spatial distribution of the urban each city largely results from long-term urban population; and (b) the noncontiguity index, patterns. Amman’s linear layout reflects FIGURE 1.6  Fragmentation varies significantly across urban areas in several capital cities in the Middle East and North Africa a. Casablanca, Morocco (4 million population) b. Baghdad, Iraq (10 million population) c. Amman, Jordan (3 million population) d. Tripoli, Libya (1.5 million population) Source: Developed from LandScan Global 2012 dataset, Oak Ridge National Laboratory, https://landscan.ornl.gov/. Note: Fragmentation is represented here by the distribution of population within urban areas. (Each square represents population density per square kilometer.) F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    41 largely the growth of Zarqa, the industrial However, the interaction potential is signifi- town on the northeast side of the city. cantly lower in Organisation for Economic Although Middle East and North Africa Co-operation and Development (OECD) cities display average potential for interaction countries and in Latin America and the compared with other regions, its subregions Caribbean. Among the subregions, the have high variations (figure 1.7).8 On aver- Mashreq displays a relatively high potential age, controlling only for population, an indi- for interaction (exceeding 1.5 million), while vidual living in a city of 2 million inhabitants the GCC and Maghreb subregions lag behind. in the Middle East and North Africa is The Maghreb subregion’s high intraur- exposed to around 1.13 million people— ban  fragmentation—with lower interac- which, taking confidence intervals into tion  potential than in the comparator account, is below South Asia’s potential for ­ subregions—increases the costs of providing interaction and not significantly different networked infrastructure. This high urban from East Asia’s or Sub-Saharan Africa’s. fragmentation (as measured by the Puga FIGURE 1.7  The Middle East and North Africa as a whole displays average interaction potential compared with other regions but varies widely by subregion Interaction potential, controlling for population, by region and subregion 2,000,000 1,750,000 Interaction potential within 10 kilometers, 1,500,000 number of people 1,250,000 1,000,000 750,000 500,000 250,000 0 eb ia c an a a q D C fi ric r ic re As EC C ci hr be h G Af Af Pa ag as h O ib ut th n M ar M d ra So an or C ha N e a Sa th d si an A d b- an st Su st Ea Ea a ic er e dl Am id M tin La Regions Subregionsa Sources: World Bank calculations using the European Commission Joint Research Centre’s Global Human Settlement Layers 2014; Henderson and Nigmatulina 2016. Note: OECD = Organisation for Economic Co-operation and Development. “Interaction potential” is measured here as the number of people an average person can interact with in a specific city within 10 kilometers. (The selected cities include the primary [largest] city of each country and, depending on data availability, all cities above 500,000.) As the Puga index equation illustrates (annex 1B), an increase in a city’s population mechanically increases the index of interaction potential. Therefore, to compare several cities, it was necessary to control for at least the city population (the T-bars indicating the confidence intervals controlling for population). As such, the method used to produce the graph consisted of predictions from the statistical model based on the Henderson and Nigmatulina (2016) methodology, setting city population at 2.019 million and controlling for city population. a. Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. 4 2    C ONVERGEN C E index, controlling for population) indicates land consumption, CAPSUS (2018) finds that low-density or leapfrogged urban fabrics. To implementing the master plan would decrease service such urban areas with networked new infrastructure costs by 42 percent (from infrastructure (roads, electricity, sewerage, JD 231.67 million to JD 135.23 million).9 and water) is typically costlier per capita, for The Jordan study is a partial costing exercise a given service quality, because it requires of urban development trajectories—partial longer and more intricate networks. because it does not account for the barriers to In financially constrained urban areas, accessing jobs, for example. these high costs may be difficult to match Urban fragmentation can also hinder with resources and can lead to a decrease in accessibility to labor market opportunities service quality. Prospective scenario model- (by lengthening trips and travel times), but ing exercises in Jordan, for example, find that higher investments in transportation seem future urban growth patterns significantly to compensate for this effect in the Middle affect the costs of providing infrastructure. East and North Africa. When controlling for Comparing a business-as-usual urban growth both GDP and population, the Puga index s c e n a r io w it h one t h at i mple m e nt s results tell a different story than when con- Amman’s master plan and results in lower trolling for population alone (figure 1.8). FIGURE 1.8  Gulf Cooperation Council countries seem to compensate for their urban fragmentation with lower commuting costs Interaction potential, controlling for population and GDP per capita, by region and subregion 2,000,000 Interaction potential within 10 kilometers, 1,750,000 1,500,000 number of people 1,250,000 1,000,000 750,000 500,000 250,000 0 ia eb q D fic a be d ric d C re ric As EC ib an Af an C hr ci an a h Af G Pa ag as h O ar a th t or as ut C ic n M M d e er N eE So ra an th m ha dl A ia Sa id As tin M b- La st Su Ea Regions Subregionsa Sources: World Bank calculations using the European Commission Joint Research Centre’s Global Human Settlement Layers 2014 and World Development Indicators database; Henderson and Nigmatulina 2016. Note: OECD = Organisation for Economic Co-operation and Development. “Interaction potential” is measured here as the number of people an average person can interact with in a specific city within 10 kilometers. (The selected cities include the primary [largest] city of each country and, depending on data availability, all cities above 500,000.) As the Puga index equation illustrates (annex 1B), an increase in a city’s population mechanically increases the index of interaction potential. Therefore, to compare several cities, it was necessary to control for both city population and gross domestic product (GDP) (the T-bars indicating the confidence intervals controlling for population and GDP per capita). Predictions for representative cities—from the statistical model based on the Henderson and Nigmatulina (2016) methodology—set the city population at 2.019 million inhabitants and GDP per capita at US$4,280 as a proxy for commuting costs (averages of the sample cities in the Middle East and North Africa). a. Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    43 A lthoug h  the d if ference bet ween the Subregional differences reflect tension Middle East and North Africa and most between modernist urban planning regions remains or becomes insignifi- theory and the traditional “compact city” cant (apart from Latin America and the The subregional differences embody a Caribbean, Sub-Saharan Africa, and the much larger tension between two paradigms: OECD countries), the subregional pic- the modernist approach to urban planning ture changes: the interaction potential and the return of the “compact city” pattern. decreases in the Maghreb and Mashreq Many historians point to the medieval Arab countries, but it strongly increases in the cities (such as Fes in Morocco and Aleppo in GCC countries—from about 1 million Syria) as supreme representations of dense, potential interactions (figure 1.7) to about compact, and livable urban centers. In the 1.2 ­m illion (figure 1.8). As such, the differ- past century, those urban systems progres- ence in potential interactions between the sively became representations of underdevel- GCC and Mashreq subregions becomes oped urban systems as modernist planning less significant, and the gap between GCC theory prevailed, focusing on urban func- and the Maghreb subregion increases tional segregation, private car orientation, significantly. and low-density spatial layouts—the interna- This change occurs because countries tional norm. However, many unforeseen with higher GDP tend to display high urban impacts of the modernist city such as depop- fragmentation, as their populations can ulated centers, traffic congestion, urban air afford higher commuting costs than those pollution, or fragmentation have led Western in lower-income countries. Therefore, the planners and urban theorists to rediscover richest countries compensate for urban dis- the viability of compact cities. persion with faster transport within cities. The compact city concept, however, is still Adding GDP as a control (in addition to not one of the drivers of most of today’s plan- population) is a proxy for increased motor- ners in the Middle East and North Africa, and ized mode shares (that is, higher shares of many countries have taken the modernist people driving instead of walking) and planning theory to its extreme. Gulf cities are accounts for the compensation of faster among the greatest examples. Abu Dhabi, transport for more fragmented urban United Arab Emirates, for example, displays a fabrics. very fragmented urban fabric (photo 1.1, panel Similarly, an urban footprint analysis a) of modern low-density neighborhoods, such shows that the Mashreq subregion has the as the Mohammed Bin Zayed City neighbor- lowest noncontiguity, which is a measure hood (photo 1.1, panel b). Even downtown of bu ilt-up spatial frag mentation. areas of GCC countries created in the 1950s (See  annex 1B for methodology.) When and 1960s are being pulled down and replaced controlling for city size, the urban foot- with large business buildings surrounded by print noncontiguity of cities of around parking garages and vacant lots. 2 million inhabitants in the GCC is twice Similarly, in Egypt, a strategy was imple- as high as in the Mashreq subregion and mented in the 1970s based on the construc- 30  percent higher than in the Maghreb tion of functionally segregated and isolated subregion. However, when controlling for new towns (photo 1.1, panel c), which ended both city population size and GDP per cap- up attracting few people. New Cairo, built ita, the difference between subregions dis- on t he s out he a s t er n e d ge of C a i ro appears. This reinforces the conclusion Governorate, is a good example (photo 1.1, that the Mashreq subregion displays the panel d). In Morocco, the state developer lowest fragmentation and that the GCC Omrane has been proceeding to construct subregion compensates for its fragmenta- satellite dormitory towns miles away from tion through lower commuting costs. 4 4    C ONVERGEN C E PHOTO 1.1  Developments in or near the capitals of United Arab Emirates and Egypt represent modernist urban planning theory a. Aerial image of fragmented urban b. Aerial image of Abu Dhabi’s Mohammed Bin fabric in Abu Dhabi Zayed City neighborhood c. Aerial image showing New Cairo’s isolation d. Aerial image of detail of New Cairo’s layout from Cairo metropolitan area Source: Google Earth 2018. existing cities. Although wealthy countries centers (photo 1.2). In Cairo, for example, compensate for fragmentation with lower although the informal neighborhood of commuting costs, the trend creates many Fostat Plateau was established in the 1980s problematic aspects, as discussed earlier, (photo 1.2, panel a.2), its structure is very such as providing urban public services. similar to the neighborhood of Bab el Wazir, established more than 80 0 years ago (photo 1.2, panel a.1). Informality stitches gaps in the urban million In 2010, in Greater Cairo alone, 12 ­ fabric inhabitants lived in informal areas, and Nonetheless, in some Middle East and 75 percent of the demographic increase found North Africa countries, informal settlements homes in informal areas. Informal neighbor- have been constituting a new path to emerg- hoods in other cities have been stitching gaps ing compact neighborhoods. Although in the urban fabric as well, filling vacant or ­ informal settlements are considered as over- abandoned spaces. Still, this does not crowded and illegal, in many cities, they are mean that urban informality is the desired dense, inclusive, energy-efficient, and multi- future of the Arab city, but rather that urban functional urban neighborhoods, displaying dwellers are rejecting the modernist city spatial patterns similar to medieval historic and  that urban planners must consider F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    45 PHOTO 1.2  Aerial views show that, in density and spatial patterns, recent informal settlements highly resemble historic districts across several cities of the Middle East and North Africa a. Cairo, Egypt a.1 Bab el Wazir (established 800 years ago) a.2 Fostat Plateau (established informally in the 1980s) b. Beirut, Lebanon b.1 Beirut historic center b.2 Jnah slum c. Rabat, Morocco c.1 Rabat historic center c.2 Douar el Garaa slum photo continues next page 4 6    C ONVERGEN C E PHOTO 1.2  Aerial views show that, in density and spatial patterns, recent informal settlements highly resemble historic districts across several cities of the Middle East and North Africa (continued) d. Tyre, Lebanon d.1 Tyre historic center d.2 Borj el Chmali (West Bank and Gaza settlement) Sources: UN-Habitat 2012; Google Earth 2018. BOX 1.2  The impact of Lebanon’s civil war on Beirut’s urban form In 1975, violence erupted in the streets of Beirut, FIGURE B1.2.1  After decades of conflict, Beirut and in a matter of days, the city became a battle- became a polycentric city field, with clashes among different Lebanese fac- tions and g lobal political players. T he u rban warfare continued almost uninterrupted for the next 15 years and resulted in massive destruction and widespread displacement of people. The city eventually became divided between Christians and Muslims—and the city center, once cosmopolitan and vibrant, became an uninhabited green buffer between the rivals. Each community started gathering in its respective sector of the city, and economic activity progressively shifted from the center to the community sectors (Hanna 2016). This resulted a few decades later in a polycentric city with three centers: the traditional center, the Christian sector, and the Muslim sector (figure B1.2.1). Source: Hanna 2016. Note: The darker the square, the higher the population density. F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    47 readapting their framework for higher effi- FIGURE 1.9 In the Middle East and North Africa, historic centers ciency and welfare (UN-Habitat 2012). and informal neighborhoods display higher road and intersection Informal neighborhoods, being similar in density than modernist neighborhoods structure to the medieval historic centers of 50 350 Middle East and North Africa cities, display Number of intersections per km2 45 300 much denser road lengths and intersections, Kilometers of road per km2 40 characterizing higher compactness. In a sub- 35 250 set of 20 historic centers and informal neigh- 30 200 borhoods in the region, the road density is on 25 average higher than 40 kilometers per square 20 150 kilometer, and intersection density is higher 15 100 than 300 per square kilometer. In contrast, 10 50 modernist neighborhoods (such as New Cairo) 5 have an average road density of 27 kilometers 0 0 per square kilometer and intersection den- Historic centers Informal Modernist neighborhoods neighborhoods sity of 131 per square kilometer (­ figure 1.9). (For the methodology used to compute road Road network density (left axis) Intersection density (left axis) intersection density, see annex 1C.) Source: World Bank calculations based on OpenStreetMap data 2018, available under the Open Data Commons Open Database License (ODbL) by the OpenStreetMap Foundation. Note: km2 = square kilometer. Urban spatial expansion is trending toward compactness in the Mashreq, FIGURE 1.10  Aggregated urban expansion trends in capital cities less so in other subregions in the Mashreq, Maghreb, and GCC subregions show divergent The trend of compactness has been espe- patterns, 1990–2014 cially strong in the Mashreq subregion. 100 4 4 Between 1990 and 2014, the urban popula- 13 tion increased four times quicker than the 90 built-up areas in Mashreq capitals, which 80 highlights a strong densification. By compari- 70 son, the urban population has increased on average only 1.3 times quicker than built-up 60 69 72 expansion in the Maghreb subregion, which Percent 66 50 implies that Maghreb cities have been densi- fying more slowly.10 The Mashreq’s faster 40 densification can be linked to its higher urban 30 population growth and strong migratory flows to the cities generated by conflict and 20 lack of economic opportunities. 28 23 10 21 Based on the methodology developed by Liu et al. (2010) and used in Baruah, Henderson, 0 and Peng (2017), we classified the 1990–2014 Mashreqa Maghrebb GCCc (without Manama) urban expansion in the capital cities of each subregion (with the exception of Casablanca, Infill Extension Leapfrog which we used for Morocco instead of Rabat) Sources: Datasets from the European Commission Joint Research Centre’s Global Human Settlement Layers. See Annex 1D for sensitivity analysis. in the three following categories: infill, exten- Note: The analysis uses the Landscape Expansion Index developed by Liu et al. (2010), in which extension sion, and leapfrog. As expected, the Mashreq and leapfrog expansion increase infrastructure costs, with leapfrog having a higher cost impact. a. Mashreq comprises the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic. countries showed the largest share of infill b. Maghreb comprises Algeria, Libya, Morocco, and Tunisia. For Morocco, Casablanca was analyzed expansion (28 percent) and the lowest share instead of the capital city, Rabat. c. GCC = Gulf Cooperation Council, comprising Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the of leapfrogging (3.7  ­ p ercent) compared United Arab Emirates. Manama, the capital of Bahrain, was excluded because of the bias created by with the two other regions (figure 1.10).11 its geographical constraints and resulting in an exceptionally high share of infill urban expansion. 4 8    C ONVERGEN C E The  divergence between the Mashreq and city growth is constrained and contained GCC subregions corresponds to the conse- because of its location on the peninsula of the quences of extensive modernist urban plan- island. Similarly, Tunis and Casablanca in the ning policies in GCC cities and trends of Maghreb subregion have higher shares of compactness in the Mashreq subregion. The infill expansion than Cairo, Damascus, and larger shares of extension in the Maghreb and Amman in the Mashreq subregion. leapfrogging in the GCC subregion reflect the These variations within subregions and trends previously highlighted in our analysis of city-specific scenarios often reflect discrete noncontiguity and interaction potential. place-based policy interventions (map 1.1). However, the cities of the same subregion For example, in Cairo, a large share of exten- vary in their development patterns. Although sion can be attributed to a government policy: each subregion, on average, displays different the construction of New Cairo, a settlement trends, some countries display distributions on the east side of the city created to unclog similar to the averages from other subregions downtown Cairo. In Dubai, the leapfrog and (figure 1.11). For example, in the GCC subre- extension patches at the south and southwest gion, the cities of Riyadh and Manama dis- of the city can also be attributed to the gov- play relatively high shares of infill and low ernment’s industrial policy through which the shares of leapfrog expansion. In the case of Dubai Investments Park was created in 1997, Manama, this is the consequence of high geo- which in turn incentivized the placement of graphical constraints considering that the industries around the urban area. FIGURE 1.11  The urban expansion of Middle East and North Africa capital cities varies within subregions, 1990–2014 100 90 Share of development type (%) 80 70 60 50 40 30 20 10 0 . n an ep co lic s a n ya a ria n ar q t te ai da bi ai si no Ira ub at oc b m R ge uw i ira ra hr un Li or ba Q O ep b or d, A Ba Al m ,J ,K ,T ra i, Le a, da t, ,M R i ol E ud s, ca ,A an oh a, ity s ip gh ab er t, ab ni a Sa am us m Tr iru pt tC D nc Tu gi Ba Ar Ar Am gy M Al Be an a h, ai bl n d ,E ad w M ia te a Ku ro as yr iy ni R ,S ai ,U C C s ai cu ub as D am D Maghreba Mashreqb GCCc Infill Extension Leapfrog Source: Datasets from the European Commission Joint Research Centre’s Global Human Settlement Layers. Note: Uses the Landscape Expansion Index developed by Liu et al. (2010), in which extension and leapfrog expansion increase infrastructure costs, with leapfrog having a higher cost impact. a. Maghreb comprises Algeria, Libya, Morocco, and Tunisia. For Morocco, Casablanca was analyzed instead of the capital city, Rabat. b. Mashreq comprises the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic. c. GCC = Gulf Cooperation Council, comprising Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. Manama, the capital of Bahrain, was excluded because of the bias created by its geographical constraints and resulting in an exceptionally high share of infill urban expansion. F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    49 MAP 1.1  Visual representations of urban expansion show the extent of development, by type, in selected Middle East and North Africa capitals, 1990–2014 Source: Datasets from the European Commission Joint Research Centre’s Global Human Settlement Layers. Urban fragmentation and sprawl foster The positive impact of urban compactness on socioeconomic segregation, with broad social interaction and the reduction of segre- consequences gation has been documented in several stud- Urban sprawl and spatial fragmentation are ies. As Burton (2000b) stated, “Lower levels linked to social segregation, while urban com- of social segregation are more strongly asso- pactness is found to reduce this phenomenon.12 ciated with higher housing density than with 5 0    C ONVERGEN C E any intervening variable.” As such, social For individuals, the lack of opportunities segregation can decrease with a dense urban and the spatial and skills mismatch limit form, where the communities are more mixed members of the marginalized population to and not spatially segregated (Burton 2001).13 low-paying or informal jobs with unstable However, the link between urban form and income streams (World Bank 2009a). Lack of the extent of segregation has mostly been dem- a formal address or identity may preclude onstrated in high-income countries. access to formal jobs, while lack of education Spatial fragmentation is intertwined with and poor health can also restrict access to social and economic dimensions that foster higher-paying jobs. Additionally, unafford- segregation. Space is a critical factor when able transit fares and lack of connection to seeking to make cities more inclusive because the public transportation network further it is a core determinant of either segregation suppress access to economic opportunities. or inclusion. Usually, socially and economi- Low incomes and multiple necessary expen- cally marginalized groups inhabit physically ditures leave individuals and households with deprived spaces, where basic infrastructure is a very limited capacity to save or invest and lacking and economic opportunities are dis- little ability to withstand shocks (Shah et al. tant (World Bank 2009a). Often these popu- 2015). lations live in informal settlements. Within cities, the proliferation of informal Urban land underpins the spatial dimen- settlements has resulted in more pressure on sion of segregation. Informal land develop- urban infrastructure and less access to health, ment occurs because formal land options are sanitation, and public services for newcom- unaffordable or inconveniently situated and ers. Except in the GCC “cit y-states” public housing is too isolated (Serageldin (Bahrain, Kuwait, and Qatar), cities in the 2016). This situation exacerbates a lack of Middle East and North Africa have fragile access to basic services and amenities as urban administrations, with urban services well  as difficulties in accessing economic often overwhelmed by the needs of growing opportunities. populations. Slums and informal settlements The broader consequences of exclusion for have proliferated in almost all cities except urban areas or even at the national level raise the urban centers of the GCC city-states. a wide set of concerns for democracy, crime, Given the region’s instability, many cities and economic growth (Shah et al. 2015). The have been receiving increasing numbers of social and economic exclusion of low-income refugees and IDPs and are becoming increas- populations may render governments unable ingly “campified” (box 1.3) The Manshiet to act in the interest of all, including segre- Nasser neighborhood in Cairo, for example, gated people, because of the social distance is home to 1 million inhabitants living in pov- between those populations and decision mak- erty under precarious conditions. Similarly, ers. The policies implemented will probably large parts of the Beirut suburbs have become not favor those populations, consequently informal settlements predominantly housing increasing inequality and undermining dem- refugees, IDPs, and lower-economic-strata ocratic and participatory processes. Besides, nationals. People living in these settlements the feeling of being left out, polarization, and suffer systemically from poor access to health spatial segregation are root causes of violence care and poor housing quality, according to and crime in urban areas. In Honduras, research from the American University of where 54 percent of the urban population is Beirut (Yassin 2016). poor, social exclusion and a lack of public In Tunisia, most migrants from the inte- services and economic opportunities underlie rior to the coast, particularly to the Greater violence. Such conditions affect social inter- Tunis area, settle in peripheral urban neigh- actions and economic growth, directly influ- borhoods, with associated higher commuting encing the investment climate and business costs and often poorer access to services. opportunities (Shah et al. 2015). Migrants to the coastal area have longer F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    51 BOX 1.3  Refugee self-sorting and fragmentation in migration to urban areas Migrants living outside camps tend to self-sort spatially waves of refugees have been concentrating in the same into low-income neighborhoods with high access to neighborhoods. As such, Palestinian refugees from services, creating a spatial and social divide between Syria and Syrian refugees have been concentrating in migrants and host communities. Refugees and inter- the same neighborhoods that were already inhabited nally displaced persons (IDPs) settle according to their by Palestinian refugees from Lebanon, who arrived economic and social preferences, which often differ before 2011 (see map B1.3.1). from those of the host communities. These displaced Displaced populations are often forced to settle populations usually sort into areas with high popula- in parts of the city that were previously uninhabited. tion density, high service availability, easy access to aid, This is the consequence of several constraints such high potential for employment in the informal sector, as high rents and housing shortages. The presence and high availability of informal housing. Thus, they of refugees in urban areas has reportedly caused rent tend to concentrate in low-income neighborhoods prices to increase so much, particularly in Amman, within cities and informal settlements close to the city that some Jordanians cannot afford the higher rents center, where they can find informal jobs while enjoying ­ (CAPSUS 2018; World Bank 2016). Refugees will lower rents than in the city center (World Bank 2017). therefore settle in informal settlements on the edges In Lebanon, however, refugees and IDPs living in of urban areas in hazard-prone locations such as low- standard housing have better access than the refu- lying areas or landfill sites with substandard or unsan- gees living in informal neighborhoods (World Bank itary housing. This puts additional stress on urban 2016), as found by sorting results in spatial segrega- services; increases the risk of conflict, exclusion, and tion between refugee-majority neighborhoods and competition; and worsens the potential for urban fra- host communities. In Tripoli (Lebanon), the different gility (Zetter 2015; Zetter and Deikun 2010). MAP B1.3.1  Refugees are concentrating in three main neighborhoods in Tripoli, Lebanon, 2015 Source: Adapted from UN-Habitat Lebanon 2016, 40. Note: PRS = Palestinian refugees from Syria. PRL = Palestinian refugees from Lebanon. SR = Syrian refugees. Syrian refugees are presented by location as registered with the United Nations Refugee Agency (UNHCR) as of May 31, 2015. 5 2    C O N V E R G E N C E commutes than workers who remain in the infrastructure and reducing the risk that interior. They also have longer commutes, on urban areas will be underserved. average, than longer-term residents on the Two contrary forces seem to be at play in coast. Migrants often live in areas with less the Mashreq urban areas: First, informal set- access to basic services than the areas serving tlements limit urban noncontiguity by being longer-term residents (but comparable access located in central areas. Second, however, to that of inhabitants in the interior of the urban policies aiming to decongest the large country). The widespread perception that urban areas increase noncontiguity by choos- migrants to the coast are disadvantaged com- ing to develop areas far from current urban pared with longer-term residents may restrain footprints, increasing infrastructure provi- migration from the country’s lagging regions sion costs. The fragmentation of urban space to its leading ones (World Bank 2013). segregates marginalized populations, reduc- As this discussion suggests, spatial segre- ing welfare outcomes and contributing to a gation is a vicious circle, since lack of access vicious circle that traps households in to basic services impedes future prospects for poverty. earnings and upward mobility. Although documentation of this dynamic is scarce in the Middle East and North Africa region, Annex 1A Methodology for extensive studies in the United States show calculating the agglomeration that to partially overcome the handicaps index posed by racial segregation,14 one of the most The agglomeration index is the first mea- powerful tools is access to good schools, sure allowing comparisons of urban concen- which are usually located in wealthier parts tration across countries using a common of urban areas (Fryer and Katz 2013). definition. It is based on a uniform definition of what constitutes an “urban” or agglomer- Concluding remarks ated area, using the technique elaborated in Uchida and Nelson (2008). The United As this chapter has established, urban Nations maintains the World Urbanization areas in the Middle East and North Africa Prospects database, a treasure trove of infor- region display high potential for interaction mation on urban areas that provides urban when population and wealth are controlled shares and population data for 229 countries for. This means that they are acting as cities, back to 1950.15 But these data are based on enabling people to meet and exchange and country definitions, which can differ greatly. allowing labor markets to function ade- The agglomeration index adjusts these data quately. However, when only population is to make them comparable. accounted for, they show high fragmentation, The agglomeration index is calculated implying larger costs to provide infrastruc- using the following steps: ture, which could lead to low coverage rates in fiscally constrained settings. •  Specify thresholds. To be classified as There are also large differences across the “urban,” an area must satisfy three crite- region, with cities in the Maghreb subregion ria: (a) minimum population size used to comparing unfavorably in terms of interac- m inimum define a sizable settlement, (b) ­ tion potential and built-up contiguity to other population density, and (c) maximum subregions of the Middle East and North travel time, by road, to the sizable Africa and even of Sub-Saharan Africa when settlement. only demographics are accounted for. Cities •  Locate the centers of sizable settle- in the Mashreq and, to a lesser extent, the ments. This mapping is done for cities GCC countries show lower noncontiguity that meet the minimum population size and leapfrog development patterns than in criterion, using data from the Global the Maghreb, limiting the costs of providing Rural-Urban Mapping Project (GRUMP) F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    53 human settlements database produced indexes, weighted by the fraction of the total by the National Aeronautics and Space population living in each pixel area, as Administration’s (NASA) Socioeconomic follows: Data and Applications Center (SEDAC). ∑ N ∑ n (d ) •  Determine the sizable settlement’s bor- ni Puga10 = j ij < 10km , (1B.1) i j der. The border surrounding a sizable set- tlement center is calculated based on the where n represents the population of a pixel maximum travel time to the center. area; N represents the total population in the •  Create population density grids. These urban area; i represents the pixel area for are created at a 1 kilometer spatial resolu- which the population residing within a tion using two global grid-based popula- 10 kilometer radius is calculated; j represents tion data sources: GRUMP and LandScan the other pixels; and dij represents the dis- (of the Oak Ridge National Laboratory). tance between pixels i and j  . The condition •  Identify the areas . Identify the grid dij < 10km returns a binary value of 0 or 1 to cells that satisfy thresholds for all three enable accounting for or eliminating people criteria. living within or outside of the 10 kilometer •  Aggregate grid cell populations. The radius of a given pixel. result is analogous to urban population. The ratio of this number to that coun- try’s total population is the agglomeration Measuring the contiguity or index, a summary measure of the propor- leapfrogging of urban development tion of the population living in areas of The contiguity of urban development—or high density. more precisely its opposite, noncontiguity— This report uses the results presented in aims to quantify the phenomenon of leap- the World Development Report 20 09 frogging, a measure of the fragmentation of (World Bank 2009b). To calculate the index built-up area proposed by Amindarbari and for the World Development Report 2009, a Sevtsuk (2013). The measure accounts for base set of thresholds was used: 50,000 for (a) the number of isolated built-up patches, minimum population size of a settlement, and (b) their relative size. The higher the 150 people per square kilometer for popula- number of isolated built-up patches, the tion density, and 60 minutes for travel time higher the noncontiguity (or fragmentation). to the nearest large city. The density and The size of the isolated patches is also travel time thresholds are those employed in important: the larger the isolated patches, Chomitz, Buys, and Thomas (2005). The the higher the fragmentation. An urban area density threshold is the same as the one used with four isolated patches of the same size by the OECD. would be more fragmented than one with a single large isolated patch and three much smaller ones. Annex 1B Methodology for The noncontiguity of urban development developing indicators of urban is calculated as follows: form Measuring the interaction potential  ∑ Ai   ∑ Ai  N N ∑ N DI =  i = n +1  i=n  , (1B.2) The interaction potential between people n =1  An   Atotal  is measured through an indicator, the Puga       index, developed by De la Roca and Puga (2017). The number of people within a given where DI is the discontiguity index; N is the radius (here, 10 kilometers) is calculated for number of isolated patches; An is the area of each pixel of built-up area. The Puga index urban patch n; and Atotal is the total surface over the city is a sum of these pixel-level of the built-up area. 5 4    C ONVERGEN C E Urban isolated patches are sorted by their specific surfaces were analyzed. To identify surface areas in decreasing order. Therefore those anomalies, the data were compared A 1 is the surface area of the largest urban with another dataset, the German Aerospace patch so that the first right-hand term in the Center’s Global Urban Footprint (GUF), equation measures, for example, the ratio which is only available for 2011. The GUF between surface areas of the second and the dataset uses radar technology, which enables first patches (ordered by size). more precise and accurate detection of urban built-up areas. Comparing both datasets reveals that Annex 1C Methodology for some cities suffer from much larger anoma- analyzing road and intersection lies than others (map 1D.1). For example, densities although the GHSL and GUF layers align in The analysis was executed using the Casablanca, there are wide disparities in OSMnx library, which enables downloading Cairo. The following recurrent anomalies of OpenStreetMap (OSM) data and calcula- were detected: tion of a set of statistics. •  The GHSL layer tends to overestimate The neighborhoods were selected based on built-up areas, identifying wider areas existing literature and qualitative observation than the GUF dataset and sometimes spe- of satellite imagery (table 1C.1). cific open areas within the city (such as the airport in Cairo). Annex 1D Comparison of Global •  In many cases, rivers crossing through cit- Human Settlement Layers and ies are incorrectly identified as built-up Global Urban Footprint datasets areas in the GHSL dataset (such as the Nile River in Cairo). The analysis of urban built-up area was based on the European Commission’s Global When possible, those anomalies were cor- Human Settlement Layers (GHSL) dataset, rected manually. For example, built-up area which uses optical sensors technology. This associated with the Nile River in Cairo was dataset was chosen because data are available erased before running the analysis. However, for multiple years (1975, 1990, 2000, and it is important to highlight that any anoma- 2014), which enables comparing built-up lies that could not be corrected (such as the areas over time. overestimation of random built-up areas) However, some anomalies appeared in the might result in moderate margins of error in identification of urban built-up areas when the analysis. TABLE 1C.1  Selected neighborhoods for analysis of road and intersection densities in the Middle East and North Africa City Historic center Informal neighborhood Modernist neighborhood Cairo, Egypt, Arab Rep. Bab el Wazir Fostat New Cairo Abu Dhabi, United Arab Emirates n.a. n.a. Mohammed Bin Zayed Dubai, United Arab Emirates n.a. n.a. Al Warqa 2 Beirut, Lebanon Bourj Abi Haydar Borj-el Barajneh Jnah Rabat, Morocco Medina Kariat Secteur 10 Tripoli, Libya Old City Al Hadba Al Khadra Asahabah Tunis, Tunisia Ez-Zitouna Cité Khalid Ibn al-Walid El Mourouj 5 Amman, Jordan Ras al-Ain Awajan Al Zohour Source: Compilation based on Google Earth. Note: n.a. = not applicable. F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    55 MAP 1D.1  Comparison of Global Human Settlement Layers and Global Urban Footprint datasets for Cairo and Casablanca, 2016 Sources: European Commission Joint Research Centre’s Global Human Settlement Layers 2015 and German Aerospace Center’s Global Urban Footprint 2016 datasets. Note: GHSL = Global Human Settlement Layers. GUF = Global Urban Footprint. Notes second largest) in a country or region. In other words, urban primacy can be defined as 1. In this report, the Maghreb refers to Algeria, the central place in an urban or city network Libya, Morocco, and Tunisia; the Mashreq to that has acquired or obtained a great level of the Arab Republic of Egypt, Iraq, Jordan, dominance. Lebanon, the Syrian Arab Republic, and West 7. The urban form indicators used in this chap- Bank and Gaza; and the Gulf Cooperation ter are discussed in annex 1B. Council (GCC) to Bahrain, Kuwait, Oman, 8. By replicating the methodology established by Qatar, Saudi Arabia, and the United Arab Henderson and Nigmatulina (2016), we ana- Emirates. Middle East and North Africa lyze the difference in fragmentation controlling regional aggregates typically include all of first for city size (population) and then for both these countries as well as Djibouti and the city size and national GDP per capita (which is Republic of Yemen, with some notable excep- used as a proxy for commuting costs). tions. 9. For the findings of the CAPSUS (2018) study, 2. Population data from the World Development see the “Urban Growth Scenarios: Hashemite Indicators database. Kingdom of Jordan” website: http://jordan​ 3. Urbanization here refers to the percentage of .capitalsustentable.com.mx/. total population living in urban areas. 10. Densification figures were calculated using 4. Refugee estimates from the United Nations the European Commission Joint Research High Commissioner for Refugees (UNHCR) Centre’s Global Human Settlement Layers 2018 data portal: http://reporting.unhcr.org​ datasets. /node/36. 11. The LEI (Landscape Expansion Index)—a 5. Urbanization rates are from the World spatial metric developed by Liu et al. (2010)— ­ Development Indicators database: https:// enabled us to capture the information of the data.worldbank.org/. formation processes of a landscape pattern. 6. Urban primacy indicates the ratio of the pri- Classification: 0 = Leapfrog, 0–50 = Extension, mate city to the next largest (that is, the 50–100 = Infill. 5 6    C ONVERGEN C E 12. “Segregation” in this chapter is interpreted as Burchell, R. W., A. Downs, B. McCann, and the opposite of inclusion (loosely based on S. Mukherji. 2005. Sprawl Costs: Economic Shah et al. [2015]): the marginalization of I m pa c t s of Un c hec ke d D e ve l o p m e n t . groups and individuals on the basis of socio- Washington, DC: Island Press. economic status, gender, age, caste, ethnicity, Burton, E. 2000a. “The Compact City: Just or and other categories and where space acts as Just Compact? A Preliminary Analysis.” Urban catalyst for such exclusion. Studies 37 (11): 1969–2001. 13. Compactness is also associated with negative ———. 2000b. “The Potential of the Compact effects, such as an increase in criminality Cit y for Promoting Social E quit y.” I n (Burton 2001), which can produce contradic- Achieving Sustainable Urban Form , edited tory effects. For example, in Mexico City, by K. Williams, E. Burton, and M. Jenks. growing criminality and violence have deeply Abingdon, U.K.: Routledge. influenced social interactions. Privileged classes ———. 2001. “The Compact City and Social have abandoned public spaces, defected from Justice.” Housing Studies Association Spring public schools and health services, and are Conference, “Housing, Environment and using the car for transport, which has resulted Sustainability,” University of York, April 18–19. in a drastic reduction in interaction opportuni- CAPSUS (CAPSUS Sustainable Capital). 2018. ties with strangers. All this has rendered multi- “Urban Grow t h S cena rios , H ashem ite class interactions virtually nonexistent (Bayón, Kingdom of Jordan.” Project brief for the Saraví, and Breña 2013). World Bank, Washington, DC. http://jordan​ 14. But only partially, as Chetty et al. (2018) .capitalsustentable.com.mx/index.php/brief. show that, for black males in the United Chetty, Raj, Nathaniel Hendren, Maggie Jones, States, differences in upward mobility cannot and Sonya Porter. 2018. “Race and Economic be entirely explained by socioeconomic or Oppor t u nit y in the United States: A n spatial dimensions. I ntergenerat iona l Persp e c t ive.” N BE R 15. For the UN World Urbanization Prospects Working Paper 24441, National Bureau of database, see https://population.un.org/wup/. Economic Research, Cambridge, MA. Chomitz, Kenneth M., Piet Buys, and Timothy S. Thomas. 2005. “Quantifying the Rural–Urban Gradient in Latin America and the Caribbean.” References Policy Research Working Paper 3634, World Amindarbari, Reza, and Andres Sevtsuk. 2016. Bank, Washington, DC. “Measuring Grow th and Change De la Roca, Jorge, and Diego Puga. 2017. in Metropolitan Form.” Unpublished. Working “Learning by Working in Big Cities.” Review Paper, City Form Lab, Massachusetts Institute of E c on o mic S tu di es 84 (1): 106 – 42 . of Technology, Cambridge, MA. doi:10.1093/restud/rdw031. Av ner, Paolo, a nd S om i k V. L al l. 2016. Fryer, Roland G., and Lawrence F. Katz. 2013. “Matchmaking in Nairobi: The Role of Land “Achieving Escape Velocity: Neighborhood Use.” Policy Research Working Paper 7904, and School Interventions to Reduce Persistent World Bank, Washington, DC. Inequality.” American Economic Review Baruah, N., J. V. Henderson, and C. Peng. 2017. 103 (3): 232–37. doi:10.1257/aer.103.3.232. “Colonial Legacies: Shaping African Cities.” Hanna, J. 2016. “Changing Realities: Traumatic Discussion Paper 226, Spatial Economics Urbanism as a Mode of Resilience in Intra-War R e s e a r c h C e nt r e , L o n d o n S c ho ol of Beirut.” International Planning History Economics, London. Society Proceedings 17 (3): 383–88. Bayón, M., G. Saraví, and M. Breña. 2013. “The Henderson, V., and D. Nigmatulina. 2016. “The Cultural Dimensions of Urban Fragmentation: Fabric of African Cities: How to Think about Segregation, Sociability, and Inequality in Density and Land Use.” Draft manuscript, Mexico City.” Latin American Perspectives London School of Economics. 40 (2): 35–52. Lall, Somik V., J. Vernon Henderson, and Bertaud, Alain. 2014. “Cities as Labor Markets.” Anthony J. Venables. 2017. Africa’s Cities: Working Paper 2, Marron Institute on Cities Opening Doors to the World . Washington, and the Urban Environment, New York DC: World Bank. doi:10.1596/978-1-4648​ University, New York. -1044-2. F r a g m e n t e d C i t i e s , C o n s t r a i n e d G r o w t h    57 Libertun de Duren, N., and R. Guerrero Urbanization Prospects: The 2018 Revision. Compeán. 2016. “Growing Resources for New York: UN DESA. Growing Cities: Density and the Cost of UN-Habitat (United Nations Human Settlements Municipal Public Services in Latin America.” Programme). 2012. The State of Arab Cities U r b a n S t u d i e s 5 3 (14 ) : 3 0 8 2 – 10 7. 2012: Challenges of Urban Transition . doi:10.1177/0042098015601579. Nairobi, Kenya: UN-Habitat. Liu, X., X. Li, Y. Chen, Z. Tan, S. Li, and S. Ai. UNHCR (United Nations High Commissioner for 2 010 . “A N e w L a nd s c ap e I nd e x for Refugees). 2014. “2014 Syria Regional Quantifying Urban Expansion Using Multi- Response Plan: Strategic Overview.” Sixth Syria Temporal Remotely Sensed Data.” Landscape Regional Response Plan, UNHCR, Geneva. Ecology 25 (5): 671–82. Venables, Anthony J. 2017. “Breaking into Newman, P. 1992. “The Compact City: An Tradables: Urban Form and Urban Function in Australian Perspective.” Built Environment a Developing Cit y.” Jour n al of Urban 18 (4): 285–300. Economics 98 (March): 88–97. doi:10.1016/j​ Rode, Ph ilipp, Gra ha m F loater, Ni kolas .jue.2017.01.002. T homopou lo s , Ja me s Do cher t y, Pe ter World Ban k. 20 09a. “Systems of Cities: Schwinger, Anjali Mahendra, and Wanli Fang. Harnessing Urbanization for Growth and 2017. “Accessibility in Cities: Transport Poverty Alleviation.” Working Paper No. and Urban Form.” In Disrupting Mobility: 51860, World Bank, Washington, DC. Impacts of Sharing Economy and ———. 2009b. World Development Report 2009: Innovative Transportation on Cities, edited by Reshaping Economic Geography. Washington, Gereon Meyer and Susan Shaheen, 239–73. DC: World Bank. Cham, Switzerland: Springer International ———. 2013. “Bridging the Spatial Divide: Labor P u b l i s h i n g . d o i : 1 0 . 1 0 0 7 / 9 7 8 - 3 - 3 1 9​ Market Outcomes in Urban Tunisia.” Survey -51602-8_15. report for the Republic of Tunisia, Report No. Serageldin, Mona. 2016. “Inclusive Cities and 89132, World Bank, Washington, DC. Access to Land, Housing, and Services in ———. 2016. “Syrians in the Middle East: The Developing Countries.” Urban Development Lives and Livelihoods of Refugees and Their Series Knowledge Papers 22, World Bank, Hosts.” Unpublished manuscript, World Bank, Washington, DC. Washington, DC. Serageldin, Mona, François Vigier, and Maren ———. 2017. “Cities of Refuge in the Middle East: Larsen. 2014. “World Migration Report 2015: Bringing an Urban Lens to the Forced Urban Migration Trends in the Middle East Displacement Challenge.” Policy Note, Report and North Africa Region and the Challenge of No. 121515, World Bank, Washington, DC. Conflict-Induced Displacement.” Background Yassin, N. 2016. “Urbanization in the Arab World: paper, I nter nat iona l Orga n i zat ion for Between ‘Dubaification’ and ‘Campification.’” Migration (IOM), Geneva. Human & Health 35 (Spring 2016). Shah, Phoram, Ellen Hamilton, Fernando Zetter, R. 2015. “Protection in Crisis: Forced Armendaris, and Heejoo Lee. 2015. “World: Migration and Protection in a Global Era.” Inclusive Cities Approach Paper.” Report No. Report, TransAtlantic Council on Migration, AUS8539, World Bank, Washington, DC. Migration Policy Institute, Washington, DC. Uchida, H., and A. Nelson. 2008. “Agglomeration Zetter, R., and G. Deikun. 2010. “Meeting Index: Towards a New Measure of Urban Humanitarian Challenges in Urban Areas.” Concentration.” Background paper for Forced Migration Review 34 (February 2010): World Development Report 2009: Reshaping 5–7. Economic Geography. Washington, DC: Zhao, P. 2013. “The Impact of Urban Sprawl on World Bank. Social Segregation in Beijing and a Limited Role U N DESA (United Nations Department of for Spatial Planning.” Journal of Economic and Economic and Social Affairs). 2019. UN Social Geography 104 (5): 571–87. Unequal Spaces and Stuck People 2 W hile variation across space is a fea- associated with greater likelihood of migration ture of development the world over, across most of the world, that does not hold the Middle East and North Africa true in the Middle East and North Africa, displays greater spatial disparity than where completing tertiary education is not expected. Cities and regions in a country will associated with a greater likelihood of migra- vary in their productivity and the average tion (D’Aoust and Lall, forthcoming). income produced within their boundaries, but Greater internal migration in the region could as countries’ per capita gross domestic product increase consumption expenditures among (GDP) increases, spatial disparities tend to migrants. Estimates vary across countries in the decline. The countries of the Middle East and region, but for the broader region, migration North Africa are outliers to this trend, dis- from lagging areas to leading areas could enable playing some of the world’s highest between- a 37 percent increase in migrants’ consumption region spatial inequality relative to comparator expenditures (D’Aoust and Lall, forthcoming). countries of similar per capita GDP. And Given this substantial potential, low internal although the region’ countries exhibit spatial migration rates present a further puzzle. disparities across varying dimensions, virtu- One possible explanation is the credential- ally all of them stand to benefit from greater focused education system prevalent through- convergence across all dimensions (consump- out the Middle East and North Africa. The tion expenditure, access to basic services and region’s education systems have been oriented infrastructure, and so on). toward offering credentials that are key for Across the world, many individuals and accessing public sector jobs, which are often households move from their birthplaces to seek local. Relative to countries of comparable opportunities for improving their lives. But the income in other regions, students in the Middle East and North Africa shows anoma- Middle East and North Africa underperform lies here, too. Internal migration rates within in international exams that measure skill these countries are considerably lower than in acquisition. That education systems focus on comparator countries. In the region, on aver- offering these credentials but not necessarily age, 14 percent of adults have moved from their on the quality and caliber of skills needed to places of birth, compared with 28 percent else- compete for private sector jobs across the where (D’Aoust and Lall, forthcoming). country could be limiting individuals’ abilities And although higher educational attainment is to acquire jobs outside of their home region. 59 6 0    C ONVERGEN C E High disparities and low migration consumption in the Middle East and North hinder economic mobility Africa than in the rest of the world. Djibouti, the Arab Republic of Egypt, the Islamic Regional inequality is high in the Republic of Iran, and the Republic of Yemen Middle East and North Africa show the starkest regional inequalities. Despite relatively low within-country Most of the region’s 444 million people live inequality, the Middle East and North Africa near coasts, where they are closer to interna- is among the most spatially unequal regions tional markets and where economic activity is in the world. Although most of its countries concentrated. People and economic activity are middle income, their spatial inequality is have naturally concentrated in places that are high and rising compared with countries else- better endowed: closer to coasts and markets where in the world, where inequality and farther from remote deserts (map 2.1). generally decreases with rising income and ­ These natural endowments—land suitability, urbanization (figure 2.1).1 resources, or coastal access—mattered to the Spatial variables consistently account for at concentration of people and activity. Natural least half of all the reported variation in eco- geography in the Middle East and North Africa nomic opportunities in most Middle East and explains 42 percent of the variation in economic North Africa countries (World Bank 2011). activity, as measured by nighttime lights Disparities between subnational regions (table 2.1). This is higher than in Latin America account for a larger share (63 percent or and Sub-Saharan Africa but quite a bit lower 6 ­ percentage points) of inequality in household than in Europe and Asia, suggesting that the FIGURE 2.1 Inequalities within most Middle East and North Africa countries exceed those of global peers a. By GDP per capita b. By urban share of population 30 30 YEM 2014 YEM 2014 Between-region inequality (%) Between-region inequality (%) 20 IRQ 2006 20 IRQ 2006 EGY 2012 IRN 2014 EGY 2012 IRN 2014 IRQ 2012 IRQ 2012 TUN 2010 TUN 2010 TUN 2005 DJI 2012 DJI 2012 TUN 2005 YEM 2005 IRN 2009 YEM 2005 IRN 2009 MAR 2000 MAR 2000 10 10 JOR 2008 JOR 2008 JOR 2010 MAR 2006 JOR 2010 MAR 2006 JOR 2006 LBN 2011 JOR 2006 LBN 2011 0 0 10,000 20,000 30,000 20 40 60 80 100 GDP per capita (US$, PPP) Urban share of the population (%) Middle East and North Africa Rest of the world Source: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Note: Each point corresponds to a country’s data for a particular year. (For a list of countries and survey years, see annex 2A.) Inequality between subnational regions was calculated based on a country’s first administrative level (for example, governorates, provinces, and so on). PPP = purchasing power parity. U n e q u a l S p a c e s a n d S t u c k P e o p l e    61 MAP 2.1  Middle East and North Africa populations are concentrated in the areas closest to international markets Lithuania UK Belarus Ireland Poland Germany Czech Republic Ukraine Kazak Hungary France Romania Croatia Serbia Monaco Italy Bulgaria Uzbekistan Spain Portugal Greece Turkey Turkmenistan Algiers Tunis Malta Syrian Arab Republic Iraq Tehran Tunisia Beirut West Bank and Gaza Lebanon Damascus Rabat Tripoli Israel Baghdad Iran, Amman Islamic Rep. Jordan Morocco Cairo Algeria Kuwait Libya Bahrain Egypt, Saudi Arabia Qatar Arab Rep. Abu Dhabi Riyadh United Arab Muscat Emirates Mauritania Oman Mali Niger Sudan Yemen, Rep. Chad Eritrea Sanaa Guinea–Bissau Guinea Djibouti Benin Nigeria 0 500 1000 km Togo Ethiopia Ghana South Sudan Central African Republic Cameroon Somalia Population density per km2 Equatorial Guinea Uganda (0–10] São Tomé and Príncipe Kenya (11–50] Gabon Congo (51–2,000] Rwanda Congo, Dem. Rep. Burundi (2,001–5,000] Seychelles (5,001–25,000] Tanzania Source: Center for International Earth Science Information Network (CIESIN) of Columbia University, “Gridded Population of the World, Version 4 (GPWv4): Administrative Unit Center Points with Population Estimates, Revision 11.” doi:10.7927/H4BC3WMT. TABLE 2.1  Economic activity benefits more from natural geography in the Middle East and North Africa than in Sub-Saharan Africa and Latin America but less so than in other regions percent Region Share of economic activity explained by natural geography (%) East Asia and Pacific 56 Europe and Central Asia 65 Latin America and the Caribbean 33 Middle East and North Africa 42 North America 59 South Asia 58 Sub-Saharan Africa 31 World 53 Source: World Bank calculation based on Henderson et al. 2017. Note: Lights at night are used as the measure of economic activity because such data are measured consistently worldwide at the same spatial scale. Henderson et al. (2017) applied a set of 24 physical geography characteristics—including those primarily important for agriculture and those primarily important for trade—in the analytical model. 6 2    C ONVERGEN C E region defies the global association of coastal Only 14 percent of Tunisia’s bottom 40 ­percent proximity with better and cheaper access to live in the largest urban metropolitan area of global markets (Malik and Awadallah 2013). Grand Tunis; in Morocco, only 6 percent live in Even so, those who could most benefit its effective economic capital, the Casablanca- from concentrated economic activity in the Settat region;2 and in the Islamic Republic of Middle East and North Africa tend not to Iran, only 4 percent live in Tehran. live in those areas. Within countries, the bot- tom of the welfare distribution is concentrated Spatial disparities also exist in access to in particular areas—often lagging areas. For public services and infrastructure example, almost 85 percent of Egypt’s bot- tom 40 ­ percent lives in rural Upper and Lower The spatial disparities extend to service cov- Egypt. In Iraq, most of the bottom 40 percent erage rates, which are lower in areas with lower lives outside the capital city region of Baghdad consumption expenditures as well as in low- and the relatively prosperous Kurdistan region. income countries in general (figures 2.2 and 2.3). FIGURE 2.2  Access to electricity has converged except in low-income economies of the Middle East and North Africa, where the poorest regions remain underserved a. West Bank and Gaza, 2011 b. Yemen, Rep., 2014 c. Djibouti, 2012 Population with access Population with access Population with access 100 100 100 to electricity (%) to electricity (%) to electricity (%) 80 80 80 60 60 60 40 40 40 20 20 20 0 10 20 30 0 10 20 30 0 10 20 30 Average daily consumption Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) d. Morocco, 2006 e. Jordan, 2010 f. Egypt, Arab Rep., 2012 Population with access Population with access Population with access 100 100 100 to electricity (%) to electricity (%) to electricity (%) 80 80 80 60 60 60 40 40 40 20 20 20 0 10 20 30 0 10 20 30 0 10 20 30 Average daily consumption Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) g. Tunisia, 2010 h. Iraq, 2012 i. Iran, Islamic Rep., 2014 Population with access Population with access Population with access 100 100 100 to electricity (%) to electricity (%) to electricity (%) 80 80 80 60 60 60 40 40 40 20 20 20 0 10 20 30 0 10 20 30 0 10 20 30 Average daily consumption Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) Source: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Note: Each point corresponds to a subnational region. Of the economies shown, only the Republic of Yemen is a low-income economy (US$1,025 or less in gross national income [GNI] per capita) according to World Bank classifications. Djibouti, the Arab Republic of Egypt, Morocco, Tunisia, and West Bank and Gaza are lower-middle-income economies (US$1,026–US$3,995). Jordan, the Islamic Republic of Iran, and Iraq are upper-middle-income economies (US$3,996–US$12,375). PPP = purchasing power parity. U n e q u a l S p ac e s a n d S t u c k P e o p l e    63 FIGURE 2.3  Primary school completion remains lower in the poorest regions of the Middle East and North Africa, except in the Islamic Republic of Iran a. West Bank and Gaza, 2011 b. Yemen, Rep., 2014 c. Djibouti, 2012 100 100 100 Population completing Population completing Population completing primary school (%) primary school (%) primary school (%) 80 80 80 60 60 60 40 40 40 20 20 20 0 10 20 30 0 10 20 30 0 10 20 30 Average daily consumption Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) d. Morocco, 2006 e. Jordan, 2010 f. Egypt, Arab Rep., 2012 100 100 100 Population completing Population completing Population completing primary school (%) primary school (%) primary school (%) 80 80 80 60 60 60 40 40 40 20 20 20 0 10 20 30 0 10 20 30 0 10 20 30 Average daily consumption Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) g. Tunisia, 2010 h. Iraq, 2012 i. Iran, Islamic Rep., 2014 100 100 100 Population completing Population completing Population completing primary school (%) primary school (%) primary school (%) 80 80 80 60 60 60 40 40 40 20 20 20 0 10 20 30 0 10 20 30 0 10 20 30 Average daily consumption Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) Source: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Note: Each point corresponds to a subnational region. Of the economies shown, only the Republic of Yemen is a low-income economy (US$1,025 or less in gross national income [GNI] per capita) according to World Bank classifications. Djibouti, the Arab Republic of Egypt, Morocco, Tunisia, and West Bank and Gaza are lower-middle-income economies (US$1,026–US$3,995). Jordan, the Islamic Republic of Iran, and Iraq are upper-middle-income economies (US$3,996–US$12,375). PPP = purchasing power parity. This lagging coverage—in electricity, education, networks and therefore lower access to mar- road networks, health services, and even ade- kets, schools, clinics, and other public quate water and sanitation—­ potentially exacer- services. Only 22 percent of the region’s rural ­ bates inequality of opportunity and creates population live within 2 kilometers of an all- poverty traps that can have lifetime and inter- weather road (World Bank 2010). These generational repercussions. For example, in areas with limited connections to the centers rural areas of the Middle East and North Africa, of economic activity are left out from the an estimated 28 million people lack access to world of opportunity and growth, and this electricity (Krishnan et al. 2016). could potentially end in political discontent. In addition, people living in poor regions In regions with unrealized economic of the Middle East and North Africa (often potential, these connectivity constraints rural areas) have limited access to road also reduce the chance of developing local 6 4    C ONVERGEN C E agglomeration  economies. In Egypt, for There are therefore large spatial disparities instance, shipping goods for short distances and deficiencies in the quality of infrastruc- within the nation’s most lagging region is ture services, with lagging regions suffering more expensive than shipping to more-distant particularly from insufficient investment and urban centers. Shipping general cargo from low quality. And dissatisfaction with the qual- Aswan to Qena or Fayoum, for instance, can ity of delivery is pervasive across all services. be more expensive than shipping to Cairo or Overloaded electricity networks, traffic con- Alexandria, both farther away. In addition, gestion, and port overcrowding are just a few even where transport networks are relatively examples of overwhelmed infrastructure net- extensive, their poor quality and limited works (Devarajan 2016; World Bank 2019a). capacity hinder the impact of economic In Iraq, for instance, electricity provision varies growth (World Bank 2010). greatly. In 2012, although 99 percent of house- Even where access to basic services is con- holds were connected to the public electricity verging, quality remains a major challenge. To grid, less than 10 percent of households in be sure, the Middle East and North Africa has Baghdad and the central and southern gover- made marked progress in building physical norates received more than 12 hours of power infrastructure for service delivery and in a day (World Bank 2014a). As for the local or expanding basic access to health and site-specific infrastructure, the quality of water education. The Gulf Cooperation Council ­ and sanitation services is poor, and localities (GCC) countries and other oil exporters will face gas and other energy shortages (World be able to meet their national infrastructure Bank 2014a). Needs vary across the region. needs if they maintain investment spending at Access to safe water is an example where the rates prevailing in the 2000s, although oil large spatial disparities remain despite some importers—such as Egypt, Morocco, and convergence in the share of people with Tunisia—will fall short (figure 2.4).3 access to a n i mproved water sou rce However, the region has been unable to figure 2.5). The water and sanitation sector (­ ensure quality in the delivery of these services. is critical for survival and growth, and the water crisis is one of the greatest threats to FIGURE 2.4  Projected infrastructure needs and financing in the the region (WEF 2015). In Middle East and Middle East and North Africa North Africa countries, 61 percent of their populations live in areas with high or very 14 high levels of surface water stress—­ exceeding 12 the global average of about 35 percent (World Average share of GDP (%) 10 Bank 2017a).4 And more than 71 percent of 8 the regional GDP is generated in areas with high or very high levels of surface water 6 Gap stress, more than triple the world average of 4 22 percent (figure 2.6). Economic losses from 2 inadequate water supply and sanitation in the 0 region’s countries are stark (figure 2.7). Middle East and Oil importersa Non-GCC GCCc Water quality matters. If not managed prop- North Africa oil exportersb erly, improved infrastructure can become an Infrastructure needs, average through 2020 Total investment spending, average 2000 efficient distributor of disease. Systematic con- tamination of improved sources, which Source: Estache et al. 2013. Note: “Total investment spending” refers to public expenditures. More recent data for the region includes all protected sources,5 is common project similar needs in terms of GDP per capita but slightly lower spending. because of inadequate maintenance of infra- a. The oil importing economies include Djibouti, the Arab Republic of Egypt, Jordan, Lebanon, Morocco, Tunisia, and West Bank and Gaza. structure, interrupted supply, and improper dis- b. The non-GCC oil exporting economies include Algeria, the Islamic Republic of Iran, Iraq, Libya, infection. But contamination also occurs during the Syrian Arab Republic, and the Republic of Yemen. c. GCC = Gulf Cooperation Council countries, comprising Bahrain, Kuwait, Oman, Qatar, Saudi transport to and unsafe storage within house- Arabia, and the United Arab Emirates. holds (Bain et al. 2014; World Bank 2017d). U n e q u a l S p ac e s a n d S t u c k P e o p l e    65 FIGURE 2.5  Access to a safe water source lags behind in the poorest regions of the Middle East and North Africa a. West Bank and Gaza, 2011 b. Yemen, Rep., 2014 c. Djibouti, 2012 Population with access Population with access Population with access to improved water (%) to improved water (%) to improved water (%) 100 100 100 80 80 80 60 60 60 40 40 40 20 20 20 0 5 10 15 0 5 10 15 0 5 10 15 Average daily consumption Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) d. Morocco, 2006 e. Jordan, 2010 f. Egypt, Arab Rep., 2012 Population with access Population with access Population with access to improved water (%) to improved water (%) to improved water (%) 100 100 100 80 80 80 60 60 60 40 40 40 20 20 20 0 5 10 15 0 5 10 15 0 5 10 15 Average daily consumption Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) g. Tunisia, 2010 h. Iraq, 2012 Population with access Population with access to improved water (%) to improved water (%) 100 100 80 80 60 60 40 40 20 20 0 5 10 15 0 5 10 15 Average daily consumption Average daily consumption per capita (US$ 2011, PPP) per capita (US$ 2011, PPP) Source: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Note: Each point corresponds to a subnational region. An “improved” drinking-water source is defined as one that, by nature of its construction or through active intervention, is protected from outside contamination, in particular from contamination with fecal matter (World Bank 2017a). Of the economies shown, only the Republic of Yemen is a low-income economy (US$1,025 or less in gross national income [GNI] per capita) according to World Bank classifications. Djibouti, the Arab Republic of Egypt, Morocco, Tunisia, and West Bank and Gaza are lower-middle-income economies (US$1,026–US$3,995). Jordan and Iraq are upper-middle-income economies (US$3,996–US$12,375). PPP = purchasing power parity. FIGURE 2.6  Far higher shares of population and economic activity are exposed to high or very high water stress in the Middle East and North Africa than in world averages a. Share of population living in areas with b. Share of GDP generated in areas with high or very high water stress high or very high water stress Middle East and 61% 71% North Africa World 36% 22% Source: World Bank 2017a. Note: Water stress arises when water withdrawals for human, agricultural, and industrial uses are relatively high compared with the level of renewable water resources—in other words, when the ratio of water withdrawal to water availability is high. Estimates of surface water stress do not account for withdrawals from groundwater and nonconventional water supplies. 6 6    C ONVERGEN C E FIGURE 2.7  Economic losses from inadequate water supply and sanitation in the Middle East and North Africa vary by economy Libya Yemen, Rep. Iraq West Bank and Gaza Djibouti Algeria Tunisia Oman Egypt, Arab Rep. Morocco Saudi Arabia Kuwait United Arab Emirates Iran, Islamic Rep. Jordan Israel Qatar Lebanon Bahrain Regional GDP 0 1 2 3 4 5 Economic losses from inadequate water supply and sanitation (% of GDP) Sources: Hutton 2013; Sadoff et al. 2015. Note: Data are from 2010. Insufficient water provision impedes the water scarcity, water service fees in the growth of lagging regions through two Middle East and North Africa are very low, channels: First, it reduces quality of life and its water subsidies are the highest in the and directly affects human resources and world (Berglöf and Devarajan 2015). These capacities. People cannot effectively partici- policies not only promote resource degrada- pate in economic production when they tion and aggravate fiscal deficits but also have insufficient or unsafe water to drink. compound vulnerabilities and further disad- Second, lack of safe water impedes already vantage the poor and marginalized. Water li m ited econom ic ac tivit y i n lag g i ng subsidies typically benefit wealthier house- regions. When factories face limited or holds more than poor households, and unreliable water supply (similar to the wealthier areas benefit more than poorer effect of unreliable energy supply), it greatly neighborhoods from subsidized water hinders productivities and makes it eco- (Berglöf and Devarajan 2015). nomically unviable for companies to oper- As such, poor households may be located in ate in lagging regions. T hese lagging areas unserved by utilities, requiring residents regions with limited local infrastructure to buy water of dubious quality from vendors usually have difficulty in attracting and in the informal sector at prices much higher retaining private investment. than those paid by the rich. Even when the In this regard, policies followed by govern- poor have access to piped water, they capture ments in the region tend to benefit wealthier a smaller share of the benefits from subsidies households more than the poor. Despite because they use less water (Whittington et al. U n e q u a l S p ac e s a n d S t u c k P e o p l e    67 2015). Social inclusion protection of the poor the shortage in infrastructure services may and marginalized populations must be central slow down economic growth. If they continue to the delivery of water services and protection to come to lagging or poor regions, the exist- from water-related risks. ing problems of limited infrastructure invest- More important, water scarcity could lead ment and public services could be exacerbated to political instability and conflict if not dealt and regional gaps in income and economic with properly. In addition to the existing performance further widened. challenges, sudden population spikes due to a large refugee influx have further strained the Spatial inequality sits at the heart of infrastructure in many of the region’s cities. conflict and climate fragility Additional demand for basic services by refu- gees puts heavy pressure on public spending, Violence and climate risks are associated further stresses infrastructure services, and with higher spatial inequalities (figure 2.8). may create potential resentment between the Instability and turmoil in the Middle East host communities and the refugees. For and North Africa have reached new levels in example, in the northern border towns of the past decade (box 2.1). Because flood and Jordan, water supply used to be available drought risks are also increasing, challenges three times a week, each time for two hours. surrounding economic and regional integra- Since the refugee influx, service has been tion are piling up. Compounded violence and reduced to once a week for only one hour.6 climate shocks result in instability, under- If refugees and internally displaced persons mining existing institutions and affecting ser- (IDPs) flow into leading subnational regions, vice delivery. The demand for services cannot FIGURE 2.8 Violent events and water risk are associated with higher spatial inequalities in the Middle East and North Africa a. Violence b. Water risk 3 3 2 2 Spatial inequality, residuals Spatial inequality, residuals 1 1 0 0 –1 –1 –2 –2 –5 0 5 –2 –1 0 1 2 Violence, residuals Water risk, residuals Sources: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank; Armed Conflict Location & Event Data (ACLED) database, https:// www.acleddata.com/; and World Resources Institute’s Aqueduct Water Risk Atlas, https://www.wri.org/resources/maps/aqueduct-water-risk-atlas. Note: The figures show residual-on-residual plots with 95 percent confidence intervals. The x-axis depicts the residuals from an ordinary least squares (OLS) regression on the log number of violent events from the ACLED database and water risk from Aqueduct data controlling for population, area, time to access nearest city, economic density, and country fixed effects. The y-axis corresponds to mean log deviation of poverty rate (as an indicator of inequality [Haughton and Khandker 2009]) on the same controls. Each point is an administrative region corresponding to the lowest administrative area with poverty rates available. Most correspond to the first administrative level; Morocco and Lebanon are the second administrative level, province and district respectively; and Jordan is the third administrative level (subdistrict). No data were available for the Gulf Cooperation Council (GCC) countries, Libya, and the Syrian Arab Republic. 6 8    C ONVERGEN C E BOX 2.1  The ongoing effects of conflict on people of the Middle East and North Africa The region is suffering from different types of conflict, severe destruction of public infrastructure, and cata- ranging from terrorism, the rise of the Islamic State strophic humanitarian emergencies (table B2.1.1). In of Iraq and the Levant (ISIL),a and open civil war and modern conflict, violence is a minor cause of mortal- violence—causing large-scale displacement of popula- ity. The burden of losses in modern conflict has shifted tions—to sporadic violence stemming from political from combatants to civilian populations, now the main instability that is largely contained within countries. target of hostilities. Civilian casualties have doubled In the wake of the 2011 Arab Spring revolts, some between 2010 and 2016 alone, with many more deaths countries experienced gains while others encountered caused by unmet medical needs, inadequate shelter, increased conflict and instability. Several countries are and famine because of conflict (World Bank 2017b). experiencing open civil war, such as Libya, the Syrian Civil conflict also leads to breakdowns in food sys- Arab Republic, and the Republic of Yemen. In Iraq, tems, in health and water infrastructure, and in access Syria, and the Republic of Yemen, civil war is further to these services. These breakdowns often increase the aggravated by the presence of terrorist groups such as death toll. The frequent outbreaks of measles and chol- ISIL and al-Qaeda. Other countries such as Tunisia era in refugee camps and among internally displaced and Egypt have more-contained political unrest with persons (IDP) populations is another important cause sporadic violence. Finally, in Jordan and Lebanon, of death. In the Republic of Yemen, food insecurity and social tensions are growing because of the influx of the threat of famine placed 17 million people at risk refugees and increasing strains on service delivery. (UNOCHA 2017). From April 2017 to July 2017, more The recent conflicts in Libya, Syria, and the than 2,300 people have died in the Republic of Yemen Republic of Yemen caused large numbers of casualties, from cholera (WHO 2018). TABLE B2.1.1 Violence in four current major crises has affected between one-third and two-thirds of the population Within 2 kilometers from event (2016–18) Displaced Country Total population (millions) Total (millions) Share (%) Total Share (%) Libya 6 2 32.2 197,000 3.2 Iraq 36 13 35.4 2,648,000 7.3 Yemen, Rep. 27 13 49.4 2,014,000 7.5 Syrian Arab Republic 19 13 69.1 6,784,000 36.7 Source: Center for International Earth Science Information Network (CIESIN) of Columbia University, “Gridded Population of the World, Version 4 (GPWv4): Administrative Unit Center Points with Population Estimates, Revision 11”; Armed Conflict Location & Event Data (ACLED) database, https://www.acleddata.com/; and UN Refugee Agency (UNHCR) population statistics, http://popstats.unhcr.org/. Note: An “event” refers to battles, explosions, and violence against civilians. Events are georeferenced, and a radius of 2 kilometers was taken around each event to count the population within 2 kilometers. a. ISIL is alternatively referred to as the Islamic State in Iraq and al-Sham (ISIS) or as the Islamic State group (IS). be met when infrastructure or the provision building, and uncoordinated and fragmented of water, education, electricity, and transpor- planning in the Tigris and Euphrates basins. tation are disrupted by war or flooding Given the strains placed on host communi- (Brixi, Lust, and Woolcock 2015). ties through protracted displacement, refugees In Iraq, for example, the failure to preserve can potentially catalyze social conflict that water resources in the marshes in the south is diminishes trust in public institutions. There not only driving displacement but also rein- can be a perception that displaced persons forcing local perceptions of marginalization compete with the poorest hosts and push them and exclusion and perpetuating fragility deeper into poverty. Yet the reality is more (Mahdi and Fawzi 2014). Iraqi marshes have a nuanced. In many areas, the challenges for long history of upstream diversions, dam host communities already existed before the U n e q u a l S p ac e s a n d S t u c k P e o p l e    69 influx of forcibly displaced persons. Refugees Low migration suppresses labor and IDPs may provide convenient scapegoats mobility in the Middle East and for deep-rooted issues, but they are often not North Africa the main cause of all the difficulties facing host countries (World Bank 2017c). The pres- Labor mobility is key for economic inte- ence of refugees and IDPs typically increases gration and the reduction of spatial inequali- demand and creates jobs, but it also adds new- ties. Globally, inequalities in regional comers to the labor force. The impact on economic activity tend to persist. There is a prices is also unevenly distributed, with prices lag in the convergence process, which is not of land and housing typically going up and fast enough to offset initially faster change in owners benefiting to the detriment of renters. leading areas. Spatial inequalities are likely to Water scarcity is a key source of tension in increase if there are barriers to labor flows the Mashreq subregion. Displaced people liv- between regions and benefits do not spill over ing in camps or among host communities can to less fortunate regions.7 struggle to access basic water supply and san- Compared with the rest of the world, how- itation services, and they place significant ever, few people in the Middle East and burdens on both service delivery systems and North Africa have moved from their birth- underlying water resources in host communi- places. Unlike international migration, labor ties. The most impoverished and disenfran- mobility within countries is largely free of chised northern region of Jordan is now faced legal constraints. Yet, internal migration with the influx of refugees from the Syrian rates in the region are lower than in the rest war. Forced displacement has accelerated an of the world. Lifetime migration compares unsustainable exploitation of groundwater place of birth (within the country) with the and pollution of aquifers. Water prices are place of current residence. On average, rising, and supply systems are straining to 14 percent of the region’s people have moved meet increased demand (World Bank 2017a). from their places of birth, compared with an Most refugees live among the host commu- average of 28 percent elsewhere (figure 2.9).8 nities instead of in refugee camps, placing an People seem to be staying in the region where unplanned burden on water resources and ser- they were born, reducing the potential gains vices. Jordan’s average annual population for migrants and their families. growth rate, which was 3.2 percent between The differences between countries are 2000 and 2010, has increased to 7 percent since partly driven by the presence of a primate the start of the Syrian Civil War in 2011 city. The Maghreb countries (as well as (as seen in nighttime images since 2011)—more Syria [Khawaja 2002]) have regional capi- than half of which (3.8 percent) is estimated to tals that could lead people to migrate to a be from the influx of forcibly displaced persons city without going too far (see chapter 3), (World Bank 2017b). In Lebanon, the popula- which explains those countries’ high shares tion has increased by an estimated 30 percent of popu lation livi ng i n u rba n a reas. since the start of the Syrian Civil War, in turn Countries in the Middle East and North increasing water demand and the volume of Africa are “urbanizing but not metropoli- sewage requiring treatment (Farajallah 2016). tanizing” (World Bank 2011)—that is, Domestic water demand in Lebanon has their urbanization is more fragmented and increased by an estimated 43 million cubic less agglomerated than in other regions, as meters (to 70 million cubic meters a year), further discussed in chapter 1. Among the which corresponds to an 8–12 percent increase Mashreq countries, Jordan’s and Lebanon’s in national water demand. Similarly, an influx systems of cities—dominated by Amman of refugees has led to an 8–14 percent increase and Beirut, respectively—push internal in wastewater volume, placing an unplanned migrants to cross regional borders. Hence, burden on limited wastewater treatment facili- interregional migration is higher in these ties (World Bank 2017a). countries. 7 0    C ONVERGEN C E FIGURE 2.9  Within-country migration is lower in the Middle East and North Africa than in the rest of the world 80 60 Internal migration rate (%) 40 20 0 Tu Iraq Ec han s R Re o Po an . Sl rtug a en l a za Ch ia ew b a In ui ue ne a a m ia Jo nia Su an kr n Le Ben e ba in B on g il El M eria ut lva o Af r Th ela a C a nd M Ri d H alay ca du ia ua a itz gu r er ay re d Sp ce zu ibe in a, a Tu RB m ey te C ica Ta ta le nz tes Bo ga ia ts nda M ana C wi M Pe a go u lia ov a om p h do SwUru do Al raz G ra n ab cc i N m in do ne si U da So Sa xic Ir ric ta n G lan el ri ub on r Ar nis on s ne L a S i U an a d h Ja rk e ai G iq rd n os ila a e al Ar ro w e t, Mo pu Mo ni U yp Ve Eg Pa Sources: Arab Barometer Wave IV (2016–17) data; University of Minnesota’s Integrated Public Use Microdata Series (IPUMS) International database. Note: The internal migration rate is the number of lifetime within-country migrants per 100 adults. Lifetime internal migration compares the place of current residence with the place of birth (within the same country). The data are from various years (see annex 2A, table 2A.1). The migration rates in Algeria, the Arab Republic of Egypt, Jordan, Lebanon, Morocco, and Tunisia are computed on the basis of the Arab Barometer Wave IV (2016–17) data; rates for other countries are from latest censuses available on IPUMS in which place of birth was available. FIGURE 2.10  Net migration flows in Tunisia reflect the movement The United Nations estimates that approxi- of people from high-poverty to low-poverty regions mately 760 million people are internal 60,000 0 migrants—almost four times as many as Net number of internal migrants, those who have moved internationally –2 40,000 (Skeldon 2013). –4 Human mobility is beneficial for house- Poverty rate (%) 20,000 –6 1999–2004 holds’ welfare, health, and education pros- 0 –8 pects. Globally, workers move in search of –20,000 –10 welfare-improving opportunities. Domestic –12 migration can offer a way for people in areas –40,000 with limited opportunities to significantly –14 –60,000 –16 improve the use of their human capital and the quality of both their lives and the lives of is t t t t t t es as es es as as n the families and communities they leave Tu lW lE hw w e he th th r ut tra ut te tra or or So behind. International evidence shows that the So a N en N en re C C G Region more workers can move across regions to Net migration, 1999–2004 (left axis) take advantage of economic opportunities, Poverty rate, 2011 (right axis) the better the skill matches and the higher Source: World Bank 2014c. welfare gains from trade, employment, and wages. Migrants also continue to interact with their places of origin, increasing trade The returns to migration are high and transferring capital, ideas, and institu- Given high regional inequality, citizens tions (Hollweg et al. 2014; Testaverde et al. may be expected to migrate across regions in 2017). In Tunisia, for example, internal pursuit of better opportunities. Globally, migration flows indicate that people are mov- migration within countries constitutes a fun- ing from high-poverty to low-poverty regions damental process of socioeconomic change. in search of opportunity (figure 2.10). U n e q u a l S p ac e s a n d S t u c k P e o p l e    71 We can estimate the cost of barriers to on average if migration to leading regions migration in the Middle East and North were to increase.9 (For the methodology of Africa in terms of the large benefits that calculating the costs of barriers to migration, increased migration would bring: consump- see box 2.2.) The higher returns to residents’ tion per capita could increase by 37 percent endowments in the leading regions explains BOX 2.2  Methodology for calculating the cost of barriers to migration Decomposing the sources of welfare gaps between where yi is the log of consumption per capita in region individual endowments and returns to these endow- i —denoted as either leading (M, for metropolis) or as ments demonstrates the difference between observed nonleading (O, for other regions)—and Xi is a set of welfare and its counterfactual if migration had endowments mentioned above. occurred. The average consumption gap can be expressed as In a world where migration is unrestricted and equation (B2.2.2): free, nobody would move if standards of living were determined only by households’ nongeographic, por- yM − yO = XM β M − XOβO , (B2.2.2) table attributes. If living in another location would improve a worker’s welfare, given his or her profile, the worker would move to that location.a Spatial dis- which can be rewritten as equation (B2.2.3): parities in living standards can be explained by the sorting of people by endowment levels (for example, ( ) yM − yO = XM − XO β M + XM ( β M − βO ) education levels) or by differences in the returns to (B2.2.3) such endowments. If large differences in the wel- ( ) + XM − XO ( β M − βO ). fare gap are attributable to the returns of household endowments in a particular location, it suggests If residents outside the metropolis are the worse off that there are labor mobility barriers across regions and we consider the normal returns to be bM , the last (Ravallion and Wodon 1999). term of equation (B2.2.3) can be dropped (see Blinder Based on a geographic analog of Blinder-Oaxaca 1973; Oaxaca 1973; Cotton 1988; Jann and Zurich decomposition across locations,b spatial disparities 2008), and we have in living standards can be measured and decomposed between endowments and returns to endowments, controlling for gender, age, education level, marital ( ) yM − yO = XM − XO β M + XM ( β M − βO ). (B2.2.4) status, status in the labor force, access to electricity and water, and possession of a computer. Figures 2.11 and 2.12 show the gap of consump- ( ) The first part XM − XO β M measures the effect of the differential in endowments, while the second tion expenditures between the metropolitan leading XM ( β M − βO ) captures the differential in returns to regions and other regions and its decomposition into endowments. endowments and returns from endowments. The Graphically, figure B2.2.1 shows that for a set consumption gap is based on a welfare ratio that of endowments XM , the consumption of the resi- deflates expenditures per capita by a spatially differ- dents in the leading area is estimated to be YM . entiated price index to account for price differences The same applies to residents elsewhere, denoted by between regions. c The share of the gaps explained the subscripts O. There are two drivers of the gap: by returns to endowments indicates that returns to the endowments themselves (on the x-axis) and their one’s endowments would be higher in the metropolis. returns, captured by the slope of the income curves, The decomposition is obtained by estimating equa- which is steeper for residents who live in the leading tion (B2.2.1) for both the country’s metropolitan region. leading region and all other regions: In figure B2.2.2, the endowments of residents in other regions are shifted to reach the ones of leading yi =  Xi β i + ε i , (B2.2.1) regions, and the resulting consumption increases are box continues next page 7 2    C ONVERGEN C E BOX 2.2  Methodology for calculating the cost of barriers to migration (continued) FIGURE B2.2.1  Consumption gap between leading shown along their income curve (from YO to Y ′O). and other regions Closing the income gap in these regions will require closing the gaps in endowments between residents in Consumption all regions, which requires investing in education and Metropolis _ _ _ yM = aM + bM XM health as well as promoting gender equality in access, yM Other for example. _ _ yO = aO + bO XO In figure B2.2.3, the slope of the curve increases to Consumption offer the same returns to all regions, showing a con- gap sumption increase from YO to Y ′O if people were _ to move. yO The estimation of welfare gaps based on the ordinary least squares (OLS) version of the Blinder- Oaxaca decomposition focuses on mean effects. _ _ The effect of the covariates is limited to an aver- XO XM Endowments age effect that remains constant across the welfare distribution. However, the effect of each covariate might in fact vary across the welfare distribution. FIGURE B2.2.2  Share of the consumption gap Using quantile regression allows for the estimation explained by endowments of the welfare gap and extends the decomposition of endowments and returns effects across the wel- Consumption Metropolis _ fare distribution. We apply the method proposed in _ _ yM = aM + bM XM Chernozhukov, Fernández-Val, and Melly (2013) yM to estimate the welfare decomposition for each Other _ _ yO = aO + bO XM quantile. Consumption gap a. The welfare aggregate (total expenditures, which includes food and nonfood _ expenditures such as clothing, services, furniture, vehicles, medicines, y'O transportation services, and education) was deflated spatially to account for the _ Consumption yO gap difference in costs of living across space. b. The source of inequalities is typically analyzed using the Blinder-Oaxaca decomposition method. See, for example, Fortin, Lemieux, and Firpo (2011) for a review of studies decomposing the sources of gender or race _ _ _ inequality and a more recent stocktaking of the explanations advanced in XO X'O = XM Endowments the literature for the gender wage gap. In the Middle East and North Africa, Bouassida and El Lahga (2018) explore the sources of the wage distribution between the public and the private sectors in Tunisia. The geographic analog of the Blinder-Oaxaca decomposition decomposes average wage or FIGURE B2.2.3  Share of the consumption gap consumption across space and has been pioneered by Ravallion and Wodon (1999) to highlight geographic differences in living standards in explained by returns to endowments Bangladesh. Given the limitations of estimating differences only at the mean, Machado and Mata (2005), Melly (2005), and Chernozhukov, Consumption Fernández-Val, and Melly (2013) developed an estimation procedure Metropolis _ Other _ transforming each observation into a counterfactual based on quantile _ _ yM = aM + bM XM y'O = aO + bM XO regression, enabling the analysis of inequalities across the welfare _ yM distribution. Applications in the Middle East and North Africa include Other _ analyses of the sources of spatial disparities in consumption in Egypt _ _ yO = aO + bO XM (World Bank 2012) and Tunisia (World Bank 2014c) as well as regional wage Consumption y'O differentials in Colombia, Portugal, and Spain (Herrera-Idárraga, López- gap Bazo, and Motellón 2016; Motellón, López-Bazo, and El-Attar 2011; Pereira and Galego 2014). c. Except in Djibouti, where the poverty line is national. In the Islamic _ Consumption yO gap Republic of Iran, the welfare aggregate excludes expenditures on health and durables for technical reasons and is intertemporally and spatially deflated to account for changes in prices during the survey period and spatial variation in prices. A detailed explanation _ _ of methodology to construct the welfare aggregate is available from XO XM Endowments Atamanov et al. (2016). U n e q u a l S p ac e s a n d S t u c k P e o p l e    73 two-thirds of the consumption gap between cost of barriers to migration as well as the those regions and others. The other third is gains from migration. The gains from (and explained by lower endowments in the non- hence implied costs of barriers to) migration leading regions. Improving the endowments are highest for the richest (most productive) of residents in nonleading regions to reach people—varying between a 32 percent and ­ the levels of residents in leading areas would a 52 percent increase in consumption—­ increase consumption by 17 percent—half of suggesting heterogeneity in the gap across the consumption benefit from migrating. the distribution, which is increasing with Figure 2.11, panel a, shows the estimated wealth. welfare gap between the leading regions Comparing the bottom 40 percent living and the others for each decile of the welfare in leading regions against those who do not, ­d istribution.10 The gap increases with con- a different picture emerges. First, the con- sumption expenditures. In other words, rela- sumption gaps between leading and nonlead- tive to their consumption group, richer people ing areas are much lower for the bottom are worse off from living in the nonleading 40 percent than for the entire distribution, regions than poor people. A decomposition and they are higher for the poor than for the of the difference between endowments and less poor (for example, urban informal returns to endowments (figure 2.11, panel b) employment is more of a better choice than confirms the previous conclusion: income farming for the poorest than for the not-so- differences across the consumption distri- poor), as shown in figure 2.12, panel a. bution are more strongly driven by returns Second, endowments matter more only for than endowments—suggesting the existing the bottom of the bottom 40 percent FIGURE 2.11  Migration to leading regions could increase consumption potential significantly in the Middle East and North Africa a. Consumption gap between leading and b. Endowments and returns from migration-related nonleading regions, by decilea consumption increase, by decile 75 60 50 70 Consumption increase (%) Consumption gap (%) 40 65 60 30 55 20 50 10 45 0 20 40 60 80 100 0 20 40 60 80 100 Consumption quantile Consumption quantile Returns Endowments Source: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Note: Consumption expenditures are per capita, spatially deflated and converted in terms of US$ (2011) purchasing power parity. Solid lines designate point estimates; dashed lines designate 95 percent confidence intervals. “Leading regions” are Great Tunis in Tunisia; Amman in Jordan; Cairo, Alexandria, Suez, and Port-Said in the Arab Republic of Egypt; the Kurdistan governorates (Duhouk, Erbil, and Suleimaniya) in Iraq; Djibouti City in Djibouti; Tehran in the Islamic Republic of Iran; the Casablanca-Settat area in Morocco; and Sana’a in the Republic of Yemen. a. The total consumption gap is the sum of consumption effects from endowments and returns. 7 4    C ONVERGEN C E FIGURE 2.12  Among the bottom 40 percent who migrate to leading regions, the migration benefits are restricted to the top a. Consumption gap between leading and b. Endowments and returns from nonleading regions, by decilea migration-related consumption increase, by decile 40 60 30 40 Consumption increase (%) Consumption gap (%) 20 20 10 0 0 –20 –10 –40 0 20 40 60 80 100 0 20 40 60 80 100 Consumption quantile Consumption quantile Returns Endowments Source: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Note: Consumption expenditures are per capita, spatially deflated and converted in terms of US$ (2011) purchasing power parity. Solid lines designate point estimates; dashed lines designate 95 percent confidence intervals. “Leading regions” are Greater Tunis in Tunisia; Amman in Jordan; Cairo, Alexandria, Suez, and Port-Said in the Arab Republic of Egypt; the Kurdistan governorates (Duhouk, Erbil, and Suleimaniya) in Iraq; Djibouti City in Djibouti; Tehran in the Islamic Republic of Iran; the Casablanca-Settat area in Morocco; and Sana’a in the Republic of Yemen. a. The total consumption gap is the sum of consumption effects from endowments and returns. figure 2.12, panel b). The top of the bottom (­ substantial, and they matter significantly in 40 percent is actually better endowed outside explaining regional differences, explaining of the metropolitan area (think about being on average 74 percent of the gap—the high- unemployed in metropolitan areas relative to est among all Middle East and North Africa being employed elsewhere), but they could countries. gain from moving, particularly if they are In Egypt, living standards are highest in unemployed. urban Lower Egypt governates, which are in Overall patterns hide differences across the fertile Nile Delta and closest to the urban consumption groups in different regions. agglomerations of Alexandria, Cairo, Port- Welfare gaps vary across and within coun- Said, and Suez.11 In the poor region of Upper tries (figure 2.13). Egypt, urban residents are expected to con- In Djibouti, for example, 26 percent of sume on average 25 percent more if they relo- the gap in welfare between regions is cate to major urban agglomerations, and explained by barriers to migration, the elim- rural residents are expected to consume ination of which which would translate to a 36 percent more. The potentially high return 51 percent increase in consumption on aver- to endowments for people from Upper Egypt age. Djibouti shows one of the highest gaps who move to major cities elsewhere suggests between metropolitan areas and other that barriers to mobility are the highest in regions. It is a low-income country that has Upper Egypt. yet to improve convergence in access to basic In the Islamic Republic of Iran, people’s services (as discussed in the previous sec- consumption in southernmost provinces12 tion). As such, differences in endowments would increase by 75 percent if they moved (e du c at ion , h e a lt h , a nd s o o n) a r e to Tehran (map 2.2). In general, the farther U n e q u a l S p ac e s a n d S t u c k P e o p l e    75 FIGURE 2.13  Consumption gaps between the metropolitan region and others vary across countries and are largely explained by differences in returns to endowments Ali Sabieh Arta Djibouti, 2012 Dikhil Tadjourah Obock Borders Urban Egypt, Arab Rep., Lower Urban Upper Urban 2012 Lower Rural Borders Rural Upper Rural Alborz Isfahan Bakhtiari Mazandaran Gilan Qazvin Kermanshah Yazd Khorasan Razavi Fars E.Azarbaijan S. Khorasan Kohkiloyeh Iran, Islamic Rep., Khuzestan Markazi 2014 W.Azarbaijan Bushehr Zanjan Ardebil Semnan Kurdistan Lorestan Hormozgan Qom Hamadan N. Khorasan Ilam Golestan Kerman Sistan Suleimaniya Erbil Duhouk Al-Najaf Karkouk Babil Basrah Salahuddin Iraq, 2012 Kerbala Diala Al-Anbar Wasit Nineveh Thi-Qar Missan Al-Qadisiya Al-Muthanna Karak Zarqa Irbid Balqa Jordan, 2010 Madaba Aqaba Tafiela Mafraq Ajlun Ma’An Regions Sahariennes Marrakech-Tensift-Haouz Fès-Boulemane-Taounate Souss- Massa-Draâ Morocco, 2006 Oriental Chaouia-Ouardigha Tadla-Azilal Méknès-Tafilalet Taza-Hoceima Gharb-Chrarda-Beni Hssen Doukkala-Abda Centre East North East Tunisia, 2010 South East South West North West Centre West Aden Dhamar Shabwah Sana’a Region Taiz Al-Baida Abyan Socatra Remah Yemen, Rep., 2014 Al-Jawf Al-Mahweet Al-Dhale Ibb Al-Hodeida Al-Maharh Hadramout Hajja Laheg Amran Saadah –40 –20 0 20 40 60 80 100 120 140 Consumption gap (%) Endowments Return to endowments Source: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Note: Consumption expenditures are per capita, spatially deflated and converted in terms of US$ (2011) purchasing power parity. The consumption gap is the difference in expenditures between each region and the leading region, disaggregated to distinguish the share explained by endowments versus returns to endowments. 7 6    C O N V E R G E N C E MAP 2.2  Consumption gaps in the Islamic Republic of Iran, by northern neighbors. These are governorates region, 2014 where consumption is 35–50 percent what it is in Baghdad, largely explained by low returns to endowments. Kirkuk likely ben- efits from links with Kurdistan, and Najaf is well connected to Baghdad. In Jordan, barriers to migration explain a large share of the welfare difference across regions. Outcomes are best in the capital, Amman, and the top-performing regions are those along the connected des- ert highway—Amman, Karak, and Jerash. Only Jerash offers returns to endowments similar to Amman. People in other regions would gain from moving toward Amman, especially those in governorates far from the desert highway corridor. Residents of Ma’an, in southeast Jordan, could increase household consumption by more than 30 percent if they were to move to Amman. Ma’an is the furthest from Amman on the consumption distribution and is one of Jordan’s disconnected and sparsely popu- lated lagging regions. M igrants from Aljoun, in north Jordan, would realize sim- ilar potential gains, although Aljoun is much closer and bet ter connected to Amman. In Morocco, convergence is slowly hap- pening. The country’s living standards have shown convergence between 2001 and 2014, Source: Iran Household Income and Expenditure Survey 2014 via the Global Monitoring Database but the pace of convergence in consumption (GMD), World Bank Team for Statistical Development, using the Datalibweb Stata Package, including the correction detailed in Atamanov et al. 2016. remains slow (figure 2.14). In 2006, there Note: Consumption expenditures are per capita, spatially deflated and converted in terms of US$ was no difference in consumption expendi- (2011) purchasing power parity. The consumption gap is the difference in expenditures between each region and the leading region. tures between Grand Casablanca (Morocco’s largest region)13 and the regions of Rabat- from Tehran, the starker the drop in con- Salé-Zemmour-Zaër (hosting the Moroccan sumption, meaning that in the south of the administration) and Tanger-Tétouan (hosting country, the average consumption is less than the Tanger-Med port, Morocco’s logistics half consumption in the capital. Apart from gateway on the Strait of Gibraltar). The gov- the two southern provinces of Kerman and ernment estimated that, at current condi- Sistan, endowments explain a minimal share tions, it would take 24 years for the process of the consumption difference. of convergence to reduce the disparity in In Iraq, consumption levels and returns to regional consumption by half (HCP and endowments are better than in Baghdad in World Bank 2017). The large share of the several governorates (map 2.3). Endowments gaps explained by the returns that people are similar. Located at the border of Kuwait, would have had in Casablanca signals barri- Basra Governorate consumption levels are ers to mobility in locations where the starker lower than in Baghdad but higher than its gaps exist. Where the gaps are lower, U n e q u a l S p a c e s a n d S t u c k P e o p l e    77 MAP 2.3  Consumption gaps in Iraq, by governorate, 2012 Source: Iraq Household Socio-Economic Survey 2012 via the Global Monitoring Database (GMD), World Bank Team for Statistical Development, using the Datalibweb Stata Package. Note: Consumption expenditures are per capita, spatially deflated and converted in terms of US$ (2011) purchasing power parity. The consumption gap is the difference in expenditures between each region and the leading region. improving endowments should be enough to increase in consumption. In the West and accelerate convergence. North-West regions, the largest share of the In Tunisia, the leading East region con- consumption gap is explained by households’ trasts with the West region, which lags endowments, contrasting with the South-East behind. T he Center-East subregion—­ and South-West subregions, where returns grouping the Mahdia, Monastir, Sfax, and explain the largest share of the divergence in Sousse governorates—has standards of liv- living standards (above 60 percent). On aver- ing similar to Greater Tunis. The gap is a bit age, migrating in Tunisia would translate to larger in the country’s North-East region, an increase of 22 percent in consumption. largely explained by endowments. (Improving In the Republic of Yemen, households in education, which is lower there than in Aden (where the port and international air- the Center-East and Greater Tunis, would port are located) have slightly better endow- translate to an increase of consumption by ments but greater differences in returns than 11 percent and bridge 43 percent of the gap.) in Sana’a. In the rest of the country, lower Migrating from the North-East subregion endowments explain a large share of the to Greater Tunis would lead to a 10 percent difference. 7 8    C ONVERGEN C E FIGURE 2.14  Morocco shows signs of regional convergence in living standards 20,000 18,000 Eddakhla-Oued Eddahab Consumption per capita (dirhams, current terms) 16,000 14,000 12,000 Laayoune-Sakia El Hamra Casablanca-Settat 10,000 Tanger-Tetouan-Al Hoceima Guelmim-Oued Noun Rabat-Salé-Kénifra 8,000 Souss-Massa Fès-Meknès Oriental 6,000 Béni Mellal-Khénifra Marrakech-Safi 4,000 Drâa-Tafilalet 2,000 0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 Per capita expenditure growth (%) Source: HCP and World Bank 2017. Note:. Consumption expenditures are annual, per capita, spatially deflated and in annual dirhams in current terms. What causes such low migration? shown to give people the ability to migrate. From an economic perspective, the low These factors often increase migration not internal migration rate is surprising given the only directly but also through accompanying high disparities between regions and the processes of social and cultural change. Most potential returns to migration. Why do so people tend to migrate domestically first, and few people in the Middle East and North in later stages, they cross borders (Sabadie Africa migrate to pursue better standards of et al. 2010). living? Social networks also play a role in facili- Social, regulatory, and economic barriers tating moves. Migration networks can be a can prevent people from migrating despite source of information, thus increasing the economic gains. Structural community fac- certainty about returns in potential desti- tors, social capital, and cultural attachment nations. Globally, internal migration is to a particular region affect people’s likeli- higher in countries where more people have hood of m ig rat i ng (Z el i nsk y 1973). access to a mobile phone, because it facili- Government policy (such as land-use man- tates information flows and connections agement and valuation information systems) (Bell et al. 2015). or migration costs (whether driven by trans- As observed in the rest of the world, port, risk, or relocation costs) can also migrants in Algeria, Egypt, Jordan, Lebanon, impede migration (Ravallion and Wodon Morocco, and Tunisia are significantly richer 1999; World Bank 2014b). Overcoming these than the nonmigrants living in their destina- barriers should therefore promote mobility. tions after they move, and they are most Better educational achievement, incomes, likely to have moved to urban areas, while access to information, and transport and married residents are less likely to migrate.14 communication infrastructure have been Migrants typically originate from regions U n e q u a l S p ac e s a n d S t u c k P e o p l e    79 FIGURE 2.15  Poverty rates at origin and destination influence migration in the Middle East and North Africa a. Probability of migration, by origin b. Probability of migration, by difference in and destination poverty rate poverty between origin and destination 100 100 80 80 Probability of migrating (%) Probability of migrating (%) 60 60 40 40 20 20 0 0 20 40 60 80 100 –50 0 50 Povetry rate (%) Difference in the poverty rate Orgin Destination between origin and destination (%) Source: Arab Barometer Wave IV (2016–17) data; and Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Note: Predictions are based on a Probit model with the dependent variable being whether the person is a migrant and covariates including gender, age, marital status, educational attainment, number of children, poverty rate at destination and origin and the log of population density at origin, and country fixed effects at 90 percent confidence intervals (dashed lines). Average poverty rates from World Bank staff at the first administrative level were matched to place of residence and place of birth. Population density was estimated based on Center for International Earth Science Information Network (CIESIN) of Columbia University, “Gridded Population of the World, Version 4 (GPWv4): Administrative Unit Center Points with Population Estimates, Revision 11.” with higher poverty rates and move to regions FIGURE 2.16  Distribution of occupations among internal migrants where poverty rates are lower (figure 2.15, and stayers in the Syrian Arab Republic, 2002 panel a). The larger the difference in poverty headcount, the more likely migration is to 100 occur (figure 2.15, panel b). Poverty traps associated with the agri- culture map dampen migration. In Egypt, 80 those who do not migrate are predomi- nantly from low-productivity regions and 60 work in agriculture. The reliance on agri- Percent cultural activities can push individuals into an agricultural poverty trap, in which a 40 large share of total consumption comes from the food they produce, and little is left to finance and bear the risks of seeking new 20 opportunities, such as through migration (Herrera and Badr 2012). In Syria, migrants were less likely than 0 nonmigrants to be engaged in agricultural Nonmigrant Internal migrant work in 2002. Those who moved were Other Construction Crafts Industrial operators clerks, salespeople, or worked in profes- Agriculture Services Sales Clerks Technicians Manager sional occupations (Khawaja 2002), as Professional shown in figure 2.16. Source: Khawaja 2002. 8 0    C ONVERGEN C E FIGURE 2.17 Internal migration rates are higher among women member to move, and only those with than men in several Middle East and North Africa countries strong networks will move (Zuccotti et al. 25 2018). In Syria, individuals who move are more likely to receive help from relatives (Khawaja 2002). 20 Internal migration rate (%) In Algeria, Egypt, Jordan, Lebanon, Morocco, and Tunisia, women are on aver- 15 age 5 percent likelier than men to migrate (figure 2.17).15 Among the reasons could be 10 marriage or job-seeking opportunities, given the high unemployment rate among edu- 5 cated women, which is particularly high in the Middle East and North Africa compared with middle-income countries in other 0 Algeria Egypt, Jordan Lebanon Morocco Tunisia regions (figure 2.18). Although information Arab Rep. is not available about the reasons for migra- Males Females tion, it is worth noting that at any level of Source: Arab Barometer Wave IV (2016–17) data. education, female migrants are on average Note: The internal migration rate is the number of lifetime within-country migrants per 100 adults. more likely than male migrants to be Lifetime internal migration compares the place of current residence with the place of birth (within the same country). For the Probit model estimating probability, see figure 2.15. employed at their destinations in Algeria, Egypt, Jordan, Lebanon, Morocco, and Tunisia (figure 2.19). FIGURE 2.18  Unemployment rates are higher in the Middle East and North Africa than in upper-middle-income countries of other regions, particularly for educated women Credential-oriented education systems offer one explanation for low internal mobility No education International evidence shows a strong Primary incomplete positive relationship between human capi- tal and migration. Economic, geographic, Primary complete or policy-imposed barriers make it harder for low-skilled people to move and suc- Secondary complete ceed in new locations (U N DP 20 09). Census data suggest a strong relationship Some tertiary or postsecondary between educational attainment and the probability to migrate when demographics 0 10 20 30 and employment status are controlled for. Unemployment rate, females (%) For instance, the probability of moving is Middle East and North Africa much higher for university graduates than Other regions for adults with less than primary educa- Source: Global Monitoring Database (GMD), Team for Statistical Development, World Bank. tion. This migration effect is one of the Note: In this figure, “Middle East and North Africa” comprises Djibouti (2012), the Arab Republic of Egypt (2012), the Islamic Republic of Iran (2014), Iraq (2012), Jordan (2010), Lebanon (2011), Morocco major channels through which education (2006), Tunisia (2010), and the Republic of Yemen (2014). supports growth, and education of the poor ensures equitable growth and pov- In Egypt, individuals who move usually erty reduction. only do so once in a lifetime, and those In the Middle East and North Africa, who do so tend to move at the time of however, having no education is no differ- m a r r i a ge ( H er rera a nd B ad r 2 012). ent from having a university degree in terms In Tunisia, families frequently identify one of the probability of migrating (­ figure 2.20). U n e q u a l S p ac e s a n d S t u c k P e o p l e    81 Elsewhere in the world, the probability FIGURE 2.19 In the Middle East and North Africa, female migrants of migrating increases with every addi- are more likely than male migrants to be employed tional year of schooling. In the Middle 0.10 East and North Africa, the likelihood Marginal effect of migration on remains the same regardless of highest probability of employment degree acquired. This is true at all levels 0.05 of income. One explanation for this anomaly is the 0 credential orientation of most educational systems in the Middle East and North Africa. Despite the region’s high educational comple- –0.05 tion rates relative to comparator countries, the extent of learning is relatively low. Not a –0.10 single Middle East and North Africa coun- No education Primary Secondary Tertiary try’s students performed near the interna- Education level tional median on the recent Trends in Male Female International Mathematics and Science Study (TIMSS) and Progress in International Source: Arab Barometer Wave IV (2016–17) data; Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, World Bank. Reading Literacy (PIRLS) assessments Note: In this figure, “Middle East and North Africa” comprises six countries covered in the Arab (World Bank 2019b). Learning-adjusted Barometer Wave IV: Algeria, the Arab Republic of Egypt, Jordan, Lebanon, Morocco, and Tunisia. Predictions are based on a Probit model with the dependent variable being whether the person is schooling years in the region reflect, on aver- employed and covariates include being a migrant interacted with gender, age, marital status, age, approximately three fewer years of edu- educational attainment, number of children, poverty rate at destination and origin, the log of population density at origin, and country fixed effects at 90 percent confidence intervals. Average cational completion than the world average poverty rates from World Bank staff at the first administrative level were matched to place of (World Bank 2019b). residence and place of birth. Population density was estimated based on Center for International Earth Science Information Network (CIESIN) of Columbia University, “Gridded Population of the Postsecondary education in the region’s World, Version 4 (GPWv4): Administrative Unit Center Points with Population Estimates, Revision 11.” households is not reflected in higher daily expenditure. Consumption expenditure usually rises with educational attainment, FIGURE 2.20  Education has virtually no effect on migration in the but not in the Middle East and North Middle East and North Africa, in contrast with the rest of the world Africa (­figure 2.21). Nevertheless, demand remains strong for credentials to access 0.25 public sector jobs in the region, but the demand for skills remains relatively weak, Probablity of migrating partly because of weak signals for the need 0.20 for skills from the region’s private sector (World Bank 2019b). The effect of education on the likelihood of getting a public sector job is significantly 0.15 higher than the likelihood of simply being employed in the region. In the 1990s, the central government’s wages equated to 0.10 10 percent of GDP in the Middle East and Less than primary Primary Secondary Tertiary North Africa—almost double the world Educational attainnment average (Makdisi, Fattah, and Limam Middle East and North Africa World 2006). As governments rapidly raised Source: Arab Barometer Wave IV (2016-17) data; University of Minnesota’s Integrated Public Use requirements for employment, the desired Microdata Series (IPUMS) International database. Note: Both estimations are based on a Probit model (as in figure 2.15), restricting controls to gender, terminal degree that was once a high school age, marital status, educational attainment, and country fixed effects for comparability perspective diploma became a university education at 90 percent confidence intervals. In this figure, “Middle East and North Africa” comprises six countries covered in the Arab Barometer Wave IV: Algeria, the Arab Republic of Egypt, Jordan, (Salehi-Isfahani 2009). A diploma in hand Lebanon, Morocco, and Tunisia. “World” comprises all countries listed in annex 2A, table 2A.1. 8 2    C ONVERGEN C E increases the chances of working in the choice. Nevertheless, the wage bill has not public sector (figure 2.22). changed much since then and remains the By the mid-1990s, the region’s govern- highest in the world at 9 percent of GDP, or ments faced fiscal contractions, limiting 30 percent of expenditures (Baddock, their ability to serve as the employer of Lang, and Srivastava 2016). As such, cre- dentials have historically been viewed as FIGURE 2.21  Higher education in Middle East and North Africa critical to access local public sector jobs, households is not reflected in daily expenditure as much as in but those credentials have not necessarily other regions endowed individuals with portable skills and may play a part in limiting internal migration. No education Primary incomplete Concluding remarks In the Middle East and North Africa, Primary complete countries exhibit within-country spatial dis- parities that are generally higher than those Secondary complete in countries of comparable income in other regions. In some cases, such as Tunisia, spa- Some tertiary or postsecondary tial disparities have increased across certain dimensions over time instead of declining, as 0 10 20 30 40 is the pattern elsewhere in the world as Daily expenditure per capita (US$ 2011 PPP) incomes rise. Internal migration rates within many of Middle East and North Africa Other regions the region’s countries are significantly lower Source: Middle East and North Africa Poverty database (MNAPOV), Team for Statistical Development, than in comparator countries—and all while World Bank. Note: PPP = purchasing power parity. the potential gains from moving to relatively FIGURE 2.22 In the Middle East and North Africa, tertiary education diplomas are highly valued in the public sector a. Effect of education on probability b. Effect of education on probability of of working in public sector working outside public sector 0.4 0.4 Marginal effect of education Marginal effect of education 0.3 0.3 0.2 0.2 0.1 0.1 0 0 Primary Secondary Tertiary Primary Secondary Tertiary Educational attainnment Educational attainnment Source: Arab Barometer Wave IV (2016–17) data. Note: The marginal effects are based on Probit regressions controlling for area of residence (urban or rural), gender, age, number of children, marital status, and country fixed effects at 90 percent confidence intervals. U n e q u a l S p ac e s a n d S t u c k P e o p l e    83 leading regions could be significant with indexes, can be decomposed across these par- respect to consumption expenditures. Unlike titions in an additive way (while the Gini elsewhere in the world, the likelihood of index cannot). migration does not increase with attainment The Theil T Index can be written as of tertiary education. One explanation for N lower internal migration is the credential-­ 1 yi  yi  orientation of most of the region’s education T= N ∑ y  ln  y  , (2A.1) i =1 systems, whereby higher diplomas histori- cally enabled greater access to local public with yi denoting the expenditure of individ- sector jobs but did not necessarily confer ual i in a country, y the average expenditure skills needed to compete for private sector of the population, and N the total popula- jobs outside of home regions. tion. The country is divided into M regions j = 1...M. We can rewrite as Annex 2A Data sources and M Nj coverage 1 yij  yij  T= N ∑∑ y  ln   y   . (2A.2) This chapter relies on various sources of j =1 i =1 data. Socioeconomic surveys cover nine countries and 69 percent of the Middle East Following Sali-i-Martin (2002), equation and North Africa population. Consumption 2A.2 can be rewritten as expenditure and access to services and M  yj  M  y ji  related spatial inequality measures are cal- culated based on these surveys. Poverty T= ∑ s j ln   +  y ∑s ln  y j  , (2A.3) j  j =1 j =1 rates were also made available for Lebanon and Algeria, which increases coverage from where the first term represents the between- 69 percent to 80 percent of the region’s group inequality, and the second term repre- population. sents the within-group inequality. Typically, Wave IV (2016–17) of the Arab Barometer at least three-quarters of inequality in a includes six countries, covering 45 percent of country is due to within-region inequality, the region’s population: Algeria, Egypt, and the remaining quarter to between-group Jordan, Lebanon, Morocco, and Tunisia. differences. Empirical analyses on the links between ­ credentialistic education and mobility under- Census data taken in this chapter are based on this dataset and complemented by existing literature. No The author wishes to acknowledge the sta- data were available for the GCC countries, tistical offices that provided the underlying Syria, and Libya. census data making this research possible (table 2A.1). Measuring inequalities in consumption expenditures within countries Survey data The Theil T index can be used to assess The household survey data were gathered the major contributors to inequality by differ- using the Global Monitoring Database ent subgroups of the population, such as (GMD) and Middle East and North Africa between regions. Average expenditures vary Poverty database (MNAPOV) ex post har- not only between regions but also inside each monization of the World Bank’s Team for region, adding a “within-group” component Statistical Development (2018) via the to total inequality. The generalized entropy Datalibweb Stata Package for the countries class of indicators, including the Theil listed in table 2A.2. 8 4    C ONVERGEN C E TABLE 2A.1  Sources and years of global census data, by country Country Year Source Armenia 2011 National Statistical Service Benin 2013 National Institute for Statistics and Economic Analysis Botswana 2011 Central Statistics Office Burkina Faso 2006 National Institute of Statistics and Demography Brazil 2010 Institute of Geography and Statistics Cameroon 2005 Central Bureau of Census and Population Studies Cambodia 2008 National Institute of Statistics Chile 2002 National Institute of Statistics China 2000 National Bureau of Statistics Colombia 2005 National Administrative Department of Statistics Costa Rica 2011 National Institute of Statistics and Censuses Cuba 2002 Office of National Statistics Dominican Republic 2010 National Statistics Office Ecuador 2010 National Institute of Statistics and Censuses Egypt, Arab Rep. 2006 Central Agency for Public Mobilization and Statistics El Salvador 2007 Department of Statistics and Censuses Ghana 2010 Ghana Statistical Services Greece 2011 National Statistical Office Haiti 2003 Institute of Statistics and Informatics Honduras 2001 National Institute of Statistics Indonesia 2010 BPS Statistics Indonesia Iraq 1997 Central Organization for Statistics and Information Technology Ireland 2011 Central Statistics Office Jamaica 2001 Statistical Institute Kenya 2009 National Bureau of Statistics Kyrgyz Republic 2009 National Statistical Committee Liberia 2008 Institute of Statistics and Geo-Information Systems Malawi 2008 National Statistical Office Malaysia 2000 Department of Statistics Mali 2009 National Directorate of Statistics and Informatics Mexico 2015 National Institute of Statistics, Geography, and Informatics Mongolia 2000 National Statistical Office Mozambique 2007 National Institute of Statistics Papua New Guinea 2000 National Statistical Office Paraguay 2002 General Directorate of Statistics, Surveys, and Censuses Peru 2007 National Institute of Statistics and Informatics Portugal 2011 National Institute of Statistics Romania 2011 National Institute of Statistics Rwanda 2012 National Institute of Statistics Senegal 2002 National Agency of Statistics and Demography Sierra Leone 2004 Statistics Sierra Leone Slovenia 2002 Statistical Office of the Republic of Slovenia South Africa 2011 Statistics South Africa Spain 2011 National Institute of Statistics Sudan 2008 Central Bureau of Statistics Switzerland 2000 Federal Statistical Office Tanzania 2012 Bureau of Statistics Thailand 2000 National Statistical Office Turkey 2000 Turkish Statistical Institute Uganda 2002 Bureau of Statistics Ukraine 2001 State Committee of Statistics United States 2015 Bureau of the Census Uruguay 2011 National Institute of Statistics Venezuela, RB 2001 National Institute of Statistics Zambia 2010 Central Statistics Office U n e q u a l S p ac e s a n d S t u c k P e o p l e    85 TABLE 2A.2  Countries and years of survey data Notes Country Survey year(s) 1. Inequality in consumption is calculated based Argentina 2012, 2014 on the Theil index (see annex 2A for data and Benin 2015 methodology). Expenditures have been spa- Brazil 2009, 2014 tially deflated to account for price differences Burkina Faso 2009 across regions. The regional comparison con- Cameroon 2007 trols for income, population, the share of Chile 2011, 2013 urban population, and the share of popula- Colombia 2014, 2016 tion in the largest city. Congo, Dem. Rep. 2012 Costa Rica 2012, 2014, 2016 2. The Casablanca-Settat region of Morocco was Djibouti 2012 so named in 2015 after the region formerly Ecuador 2014, 2016 referred to as Greater Casablanca or Grand Egypt, Arab Rep. 2012 Casablanca annexed several provinces. Ethiopia 2010 3. More recent data are available for the Ghana 2012 entire region and portray a similar picture. Haiti 2012 Unfortunately, there is no disaggregation by Honduras 2012, 2013, 2016 country. In 2014, estimated infrastructure India 2009, 2011 spending in the Middle East and North Africa Indonesia 2005, 2011, 2014, 2016 was 6.9 percent of GDP (Fay et al. 2019). Iran, Islamic Rep. 2009, 2014a Iraq 2006, 2012 Needs were also assessed in various scenarios. Jordan 2006, 2008, 2010 In the preferred scenario, the Middle East and Kenya 2005 North Africa needs to spend 7.1 percent of Lao PDR 2007, 2012 GDP to develop infrastructure between 2015 Lebanon 2011 and 2030 to reach ambitious goals. This sce- Madagascar 2010 nario assumes high spending efficiency, which Malawi 2010 depends on the quality of complementary Malaysia 2016 policies and on measures to reduce unit costs Mali 2010 (like better procurement, planning, or execu- Mexico 2014 tion) (Rozenberg and Fay 2019). Mongolia 2010, 2011, 2012, 2014 Morocco 2000, 2006 4. Water stress arises when water withdraw- Mozambique 2008 als for human, agricultural, and industrial Niger 2011 uses are relatively high compared with the Nicaragua 2014 level of renewable water resources—in other Papua New Guinea 2009 words, when the ratio of water withdrawal Paraguay 2014 to water availability is high. It is quantified Peru 2014, 2016 as the ratio of annual water withdrawals to Senegal 2011 average annual surface water availability, Sierra Leone 2011 driven by either climate change under a high South Africa 2010 emission scenario or socioeconomic change Sudan 2009 Tanzania 2011 under a business-as-usual scenario for popu- Thailand 2006, 2009, 2012 lation growth and the economy. Estimates of Tunisia 2005, 2010 surface water stress do not account for with- Turkey 2014 drawals from groundwater and nonconven- Uganda 2012 tional water supplies. Ukraine 2016 5. An “improved” drinking-water source is Uruguay 2012, 2014 defined as one that, by nature of its construc- Vietnam 2006, 2010, 2012, 2014 tion or through active intervention, is pro- Yemen, Rep. 2005, 2014 tected from outside contamination, in Zambia 2010 particular from contamination with fecal a. For the Islamic Republic of Iran in 2014, the international poverty rate is matter (World Bank 2017a). slightly different from the poverty rate reported by the World Bank in World Development Indicators and PovcalNet. The difference comes from the way 6. Task team interviews with local residents. welfare aggregate is created. This welfare aggregate excludes expenditure on 7. Chapters 1 and 3 explore other key contribu- health and durables for technical reasons and is inter-temporally and spatially tors to convergence, such as agglomeration, deflated to account for changes in prices during the survey period and spatial variation in prices. Detailed explanation of methodology to construct welfare specialization, and interregional links. aggregate is available in Atamanov et al. (2016). 8 6    C ONVERGEN C E 8. Internal migration rates are based on cen- References sus data from 38 countries outside the Middle East and North Africa, whose cen- Atamanov, A., M. Mostafavi, D. Salehi-Isfahani, suses include birthplace at the first adminis- and T. Vishwanath. 2016. “Constructing trative level for comparability purposes. Robust Poverty Trends in the Islamic Republic The only Middle East and North Africa of Iran: 2008–2014.” Policy Research Working countries with a census were Egypt and Paper 7836, World Bank, Washington, DC. Iraq. Random respondents ages 18 and Baddock, Emily, Peter Lang, and Vivek Srivastava. above were selected from available censuses 2016. “Size of the Public Sector: Government (see annex 2A, table 2A.1), with half men Wage Bill and Employment.” Summary Note, and half women to be comparable with the World Bank, Washington, DC. Arab Barometer data—which are available Bain, R., R. Cronk, J. Wright, H. Yang, T. Slaymaker, for Algeria, Jordan, Lebanon, Morocco, and J. Bartram. 2014. “Fecal Contamination of Tunisia, and West Bank and Gaza. The lat- Drinking Water in Low- and Middle-Income ter was excluded given the peculiarity of Countries: A Systematic Review and Meta- that economy. For Egypt, the Arab Analysis.” PLOS Medicine 11 (5): e1001644. Barometer was used instead of that coun- Bell, M., E . Charles-Edwards, P. Ueffing, try’s census because that survey was more J.  Stillwell, M. Kupiszewski, and recent (2016 versus 2006). In Egypt, the D. Kupiszewska. 2015. “Internal Migration Arab Barometer did not survey the Sinai and Development: Comparing Migration Peninsula, Matruh, New Valley, and Red Intensities Around the World.” Population and Sea governorates where collectively less Development Review 41 (1): 33–58. than 2 percent of the population lives. Berglöf, E., and S. Devarajan. 2015. “Water for 9. Leading regions are Djibouti City in Djibouti; Development: Fulfilling the Promise.” In Water Cairo, Alexandria, Suez, and Port-Said in for Development: Charting a Water Wise Egypt; Tehran in the Islamic Republic of Iran; Path, edited by Anders Jägerskog, Torkil Jønch the Kurdistan governorates (Duhouk, Erbil, Clausen, Torgny Holmgren, and Karin Lexén, and Suleimaniya) in Iraq; Amman in Jordan; 23–27. Stockholm: Stockholm International the Casablanca-Settat area in Morocco; Great Water Institute (SIWI). Tunis in Tunisia; and Sana’a in the Republic Blinder, A. S. 1973. “Wage Discrimination: of Yemen. Reduced Form and Structural Estimates.” 10. The sum of endowments and returns effects Journal of Human Resources 8 (4): 436–55. gives the total welfare gap. Bouassida, I., and A.-R. El Lahga. 2018. “Public- 11. These governorates include Sharkiya, Behira, Private Wage Disparities, Employment, and Dakahliya, Domiyat, Gharbiya Ismailia, Labor Market Segmentation in Tunisia.” In Kafr-elsheikh, Menofiya, and Kaliobiya. The Tunisian Labor Market in an Era of 12. The southernmost provinces exclude Kerman Tra n si t i o n , e d ite d by R . A s s a ad a nd and Sistan. M.  Boughazala, 86–112. London: Oxford 13. The survey was conducted before the regions University Press. were redrawn, after which the region’s name Brixi, H., E. Lust, and M. Woolcock. 2015. Trust, changed to Casablanca-Settat. Voice, and Incentives: Learning from Local 14. All data here from the Arab Barometer unless Success Stories in Service Delivery in the otherwise stated. This is the only dataset with Middle East and North Africa. Washington, comparable information on lifetime internal DC: World Bank. migration in six countries in the region. Chernozhukov, V., I. Fernández-Val, and B. Melly. Although the data and sample are limited in 2 013 . “ I n f e r e n c e o n C ou nt e r f a c t u a l terms of representation and of sector of Distributions.” Econometrica 81 (6): 2205–68. employment and social network information, Cotton, J. 1988. “On the Decomposition of Wage the dataset contains the main migration driv- Differentials.” Review of Economics and ers such as age, income, and education. Statistics 70 (2): 236–43. 15. Figure 2.17 is a simple bar graph, while the D’Aoust, O., and V. S. Lall. nd. “Unequal Spaces estimated probability is based on the Probit and Stuck People.” Unpublished manuscript. model estimated in figure 2.15. World Bank, Washington DC. U n e q u a l S p ac e s a n d S t u c k P e o p l e    87 Devarajan, Shantayanan. 2016. “An Exposition Hutton, G. 2013. “Global Costs and Benefits of of the New Strategy, ‘Promoting Peace and Reaching Universal Coverage of Sanitation Stability in the Middle East and North and Drinking-Water Supply.” Journal of Water Africa.’” Working paper, Report No. 102936, and Health 11 (1): 1–12. World Bank, Washington, DC. Jann, B ., and E . Zurich. 20 08. “A Stata Estache, A., E. Ianchovichina, R. Bacon, and I mplementation of the Blinder- Oaxaca I .  S a l a mon . 2 013. In f ra s t r u c t u re a n d Decomposition.” Stata Journal 8 (4): 453–79. Employment Creation in the Middle East Khawaja, M. 2002. “Internal Migration in Syria: and North Africa. Washington, DC: World Findings from a National Survey.” Report Bank. 375, Fafo Research Fou ndation, Oslo, Farajalla, N. 2016. “Water Resources and Norway. Conflict in Lebanon.” In Losing Paradise: The Krishnan, N., G. L. Ibarra, A. Narayan, S. Tiwari, Water Crisis in the Mediterranean, edited by and T. Vishwanata. 2017. Uneven Odds, Gail Holst-Warhaft and Tammo Steenhus, Unequal Outcomes. Inequality of 125–40. Abingdon, U.K.: Routledge. Opportunity in the Middle East and North Fay, M., S. Han, Y. Lee, M. Mastruzzi, and Africa . Directions in Development Series. M. Cho. 2019. “Hitting the Trillion Mark–A Washington, DC: World Bank. Look at How Much Countries Are Spending M achado, J. A ., a nd J. M at a. 20 05. on Infrastructure.” Policy Research Working “Counterfactual Decomposition of Changes in Paper 8730, World Bank, Washington, DC. Wa g e D i s t r i b u t i o n s U s i n g Q u a n t i l e Fortin, N., T. Lemieux, and S. Firpo. 2011. Regression.” Journal of Applied Econometrics “Decomposition Methods in Economics.” In 20 (4): 445–65. Handbook of Labor Economics, Vol. 4A: Mahdi, B. A., and N. A. M. Fawzi. 2014. “Iraq’s 1–102. Amsterdam: Elsevier. Inland Water Quality and Their Impact on the Haughton, Jonathan, and Shahidur R. Khandker. North-Western Arabian Gulf.” Marsh Bulletin 2009. Handbook on Poverty and Inequality. 9 (1): 1–22. Washington, DC: World Bank. Makdisi, S., Z. Fattah, and I. Limam. 2006. HCP and World Bank (Haut-Commissariat au “Determinants of Growth in the MENA Plan and World Bank). 2017. «Pauvreté et Countries.” In Explaining Growth in the prospérité partagée au Maroc du troisième Middle East, edited by Jeffrey Nugent and m i l lé n a i re , 2 0 01– 2 014.» ( I n Fre n c h .) Hashem Pesaran, 32– 60. Bingley, U.K.: R e p o r t , H C P a n d Wo rld B a n k , Emerald Group Publishing Limited. Washington, DC. Malik, A., and B. Awadallah. 2013. “The Henderson, J. V., T. Squires, A. Storeygard, and E conomics of the A rab Spring.” World D. Weil. 2017. “The Global Distribution of Development 45: 296–313. Economic Activity: Nature, History, and the Melly, B. 2005. “Decomposition of Differences in Role of Trade.” Q u ar te rly Jour n al of Distribution Using Quantile Regression.” Economics 133 (1): 1–50. Labour Economics 12 (4): 577–90. Herrera, S., and K. Badr. 2012. “Internal Motellón, E., E. López-Bazo, and M. El-Attar. Migration in Egypt: Levels, Determinants, 2011. “Regional Heterogeneity in Wage Wages, and Likelihood of Employment.” Distributions: Evidence from Spain.” Journal Policy Research Working Paper 6166, World of Regional Science 51 (3): 558–84. Bank, Washington, DC. O a x ac a , R . 1973. “M a le – Fem a le Wage Herrera-Idárraga, P., E . López-Bazo, and Differentials in Urban Labor Markets.” E.  Motellón. 2016. “Regional Wage Gaps, International Economic Review 14 (3): Education and Informality in an Emerging 693–709. Country: The Case of Colombia.” Spatial Pereira, J., and A. Galego. 2014. “Interregional Economic Analysis 11 (4): 432–56. Wage Differentials in Portugal: An Analysis Hollweg, C., D. Lederman, D. Rojas, and across the Wage Distribution.” Regional E. Ruppert Bulmer. 2014. Sticky Feet: How Studies 48 (9): 1529–46. Labor Market Frictions Shape the Impact of Ravallion, M., and Q. Wodon. 1999. “Poor International Trade on Jobs and Wages. Areas, or Only Poor People?” Journal of Washington, DC: World Bank. Regional Science 39 (4): 689–711. 8 8    C ONVERGEN C E Rozenberg, J., and M. Fay. 2019. “Beyond the Countries, with Illustrative Simulations.” G ap: How C ou nt r ie s C a n A f ford t he Utilities Policy 34: 70–81. Infrastructure They Need while Protecting the W HO (World Health Organization). 2018. Planet.” World Bank, Washington, DC. “Epidemic and pandemic-prone diseases. Sabadie, Jesus A lquezar, Johanna Avato, Outbreak update: Cholera in Yemen, 19 July Ummuhan Bardak, Francesco Panzica, and 2018.” Release, WHO Regional Office for the Natalia Popova. 2010. Migration and Skills: Eastern Mediterranean (EMRO), Cairo. The Experience of Migrant Workers from World Bank. 2010. “Transport. Sectoral Notes, Albania, Egypt, Moldova, and Tunisia . Middle East and North Africa-Regional Directions in Development Series. Washington, Transpor t A nnual Meetings.” I nternal DC: World Bank. document, World Bank, Washington, DC. Sadoff, C. W., J. W. Hall, D. Grey, J. C. J. H. ———. 2011. Poor Places, Thriving People: How Aerts, M. Ait-Kadi, C. Brown, A. Cox, et al. the Middle East and North Africa Can Rise 2015. Securing Water, Sustaining Growth: Above Spatial Disparities. M ENA Report of the GWP/OECD Task Force on Development Report Series. Washington, DC: Water Security and Sustainable Growth. World Bank. Oxford: University of Oxford. ———. 2012. “Arab Republic of Egypt: Reshaping Sala-i-Martin, X. 2002. The Disturbing Rise of Egypt’s Economic Geography: Domestic Global Income Inequality. NBER Working Integration as a Development Platform.” Paper No. 8904, National Bureau of Economic R ep or t N o. 71 289 - E G , World B a n k , Research, Cambridge, MA. Washington, DC. Salehi-Isfahani, D. 2009. “Education and ———. 2014a. “Iraq: The Unfulfilled Promise of Earnings in the Middle East: A Comparative Oil and Growth: Poverty, Inclusion and Study of Returns to Schooling in Egypt, Iran, Welfare in Iraq, 2007–2012.” Report No. and Turkey.” Working Paper Series 504, 92681-IQ, World Bank, Washington, DC. Economic Research Forum, Giza, Egypt. ———. 2014b. “Tunisia Urbanization Review: S ke ld o n , R . 2 013 . “G lob a l M i g r at io n : Reclaiming the Glory of Carthage.” World Demographic Aspects and Its Relevance for B a n k , Wa s h i n g t o n , D C . h t t p s : / / b i t​ Development.” Technical Paper 2013/6, .ly/2WqQKaT. Popu l at ion D iv i sion , U n it e d N at ion s ———. 2014c. “The Unfinished Revolution: Department of Economic and Social Affairs Bringing Opportunity, Good Jobs and Greater (UN DESA), New York. Wealth to All Tunisians.” Development Policy Testaverde, M., H. Moroz, C. Hollweg, and Review, Report No. 86179-TN, World Bank, A .   S c h m i l l e n . 2 0 17. M i g r a t i n g t o Washington, DC. Opportunity: Overcoming Barriers to Labor ———. 2017a. Beyond Scarcity: Water Security in Mobility in Southeast Asia. Washington, DC: the Middle East and North Africa. MENA World Bank. Development Report Series. Washington, DC: U N DP (United Nations Development World Bank. Programme). 2009. Human Development ———. 2017b. “Cities of Refuge in the Middle Report 2009: Overcoming Barriers: Human East. Bringing an Urban Lens to the Forced Mobility and Development . New York: Displacement Challenge.” Policy Note, Report UNDP. No. 121515, World Bank, Washington, DC. U NOCH A (United Nations Office for the ———. 2017c. Forcibly Displaced: Toward a Coordination of Humanitarian Affairs). 2017. Development Approach Supporting Refugees, “ Ye m e n : 2 018 H u m a n i t a r i a n N e e d s the Internally Displaced, and Their Hosts. Overview.” Report, UNOCHA, New York Washington, DC: World Bank. and Geneva. ———. 2017d . “Reducing Inequalities in Water WEF (World Economic Forum). 2015. “Global Supply, Sanitation, and Hygiene in the Era of Risks 2015, 10th Edition.” Insight Report, the Sustainable Development Goals: Synthesis WEF, Geneva. Report of the WASH Poverty Diagnostic Whittington, D., C. Nauges, D. Fuente, and X. Initiative.” Report No. 127135, World Bank, Wu. 2015. “A Diagnostic Tool for Estimating Washington, DC. the Incidence of Subsidies Delivered by Water ———. 2019a. “Economic and Social Inclusion Utilities in L ow- and Mediu m-I ncome for Peace and Stability in the Middle East and U n e q u a l S p ac e s a n d S t u c k P e o p l e    89 North Africa: A New Strategy for the World Zelinsky, W. 1973. The Cultural Geography of Bank Group.” PowerPoint presentation, the United States . Englewood Cliffs, NJ: World Bank, Washington, DC. http://pubdocs​ Prentice-Hall. .worldbank.org /en /335571450333797609​ Zuccotti, C. V., A. Geddes, A. Bacchi, M. /121815-middle-east-and-north-africa-seminar​ Nori, and R. Stojanov. 2018. Drivers and .pdf. Patterns of Rural Youth Migration and its ———. 2019b. Expectations and Aspirations: I m p a c t o n F o o d S e c u r i t y a n d R u ra l A  New Framework for Education in the Livelihoods in Tunisia . Rome: Food and Middle East and North Africa . Washington, Agriculture Organization of the United DC: World Bank. Nations (FAO). Walled Urban Economies 3 E conomies are walled off from others, The region’s largest cities may be too large regionally and globally, by many barri- for their local markets but may be rightsized ers that Middle East and North Africa if they could more easily serve much larger governments have created—or failed to regional markets. The largest cities in well- remove. These walls limit regional market integrated regions attract human and finan- access and dampen the potential for econo- cial capital from their neighbors and produce mies of scale and specialization for the goods and services for the entire region. region’s cities. To improve local business eco- London, Paris, or Vienna are large relative to systems, the large cities need large markets, their countries’ populations but not relative and secondary cities need complementary to the markets they can serve. In the Middle policy and investment. East and North Africa, in contrast, most cit- Many of the Middle East and North ies that could play a much greater regional Africa’s largest cities appear too large for their role are prevented from doing so by ­ numerous domestic markets. Especially in the Mashreq barriers to regional economic integration. and Gulf Cooperation Council (GCC) coun- The region’s secondary cities suffer from a tries, urban primacy—the share of the urban variety of complementary policy and invest- population in the largest city—is higher than ment constraints. Unlike in other regions of in any other world region.1 This may be due to the world, firm productivity appears to be region-specific differences in urban manage- most influenced by location rather than firm ment or policy as well as public investment size. While not problematic per se, such a bias toward politically important large cities, finding raises the possibility that the lack of where a better business environment and sup- an enabling environment in the periphery2 — ply of urban services might attract more firms either through limited fiscal decentralization, and people. Alternatively, these large cities suboptimal spatial distribution of infrastruc- may reflect a long history of agglomeration ture, or cronyism—might be skewing the economies that continue to serve as attractive distribution of economic activity toward the ­ areas for colocation. The largest cities thus capital. become magnets for migrants. There is evi- It is important here that secondary cities dence for both explanations, but another and large cities not be viewed as substitutes; aspect must also be considered. they should perform as complements. 91 9 2    C ONVERGEN C E Three decades of research worldwide high- integration agreements also reflects these bar- light that businesses and people can exploit riers. Although such agreements are relatively economies of scale and agglomeration if strong among the Gulf countries, the rest of their cities perform their intended functions. the region is linked internally by shallow Policy makers should start by treating their regional integration agreements that have not cities as a portfolio of assets, each asset dif- promoted the emergence of regional produc- ferentiated by characteristics that include tion networks supported by strong urban size, location, and population density. agglomeration economies. This chapter dis- This chapter documents the unbalanced cusses the state of regional integration, chapter urban size distribution in most Middle East 4 discusses some of the main reasons explain- and North Africa countries and its implica- ing limited regional integration, and chapter 6 tions for regional integration. On average in suggests some of the steps toward greater inte- this region, the country’s largest city is home gration in the face of existing barriers. to about a quarter of the national urban population, while most secondary cities are small. Historically, the region’s largest cities Large cities will remain were closely connected and served as centers important in the Middle East for production, innovation, and learning for and North Africa landscape the entire region. But, as this chapter shows, Urban primacy—the share of the urban Middle East and North Africa countries population residing in a country’s largest today are poorly integrated, with lower urban area—is a defining characteristic of the interaction between countries than in eco- Middle East and North Africa’s urban system. nomically more dynamic world regions. In the region, on average, around 25 percent The chapter provides evidence that the of the urban population is concentrated in a flows of goods, services, capital, and labor— country’s largest city (figure 3.1). Sub-Saharan what the European Union (EU) calls the “four Africa is the only region with a higher urban freedoms”—face high barriers, as do the flows primacy, at 28 percent. Urban primacy—the of ideas, for instance, through media or per- share of the urban population residing in a sonal exchanges. The status of regional country’s largest urban area—is a defining FIGURE 3.1  Urban primacy rates are high in the Middle East and North Africa, driven mainly by population distributions in the GCC and Mashreq subregions 40 Largest city as percentage of 35 urban population (%) 30 25 20 15 10 5 0 a a ric d ci d C eb be d l A and eq i As ric Af an Pa an C ib an hr a fic hr an a G Af h th t ia si ag ar a tra e as or as ut C ic en p As n M M C uro N E So ra e er st e ha th Am E dl Ea Sa id tin M b- La Su Regions Subregionsa Source: UN DESA database: https://population.un.org/wup/ a. States composed of a single urban settlement and with a primacy equal to 1 (100% of urban population in the largest city) were excluded from the analysis. Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Israel, Jordan, Lebanon, and the Syrian Arab Republic; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates.  Wa l l e d U r b a n Ec o n o m i e s    93 characteristic of the Middle East and North population in the largest city ranges from Africa and its subregions’ urban system. On 24 percent in Saudi Arabia to 65 ­ percent in average, around 32 percent of the urban popu- Kuwait. In the Maghreb subregion, all coun- lation is concentrated in a country’s largest tries are included in an interval between 9 per- city in the region, which places it slightly cent (Algeria) and 23 percent (Libya), and in below Latin America and the Caribean the Mashreq subregion, in an interval between (34  percent) and Sub-Saharan A frica 22 percent (Jordan) and 64 percent (Israel). percent) (figure 3.1). Urban primacy is sig- (36 ­ The primate city in each of the region’s nificantly lower on average in other regions countries is the main area of wealth produc- such as East Asia and Pacific (28 percent), tion as well as population. In most countries, Europe and Central Asia (26 percent), and it is gross domestic product (GDP) concentration especially low in South Asia (only 23 percent). is higher than population concentration because agglomeration economies tend to raise residents’ productivity. The region’s urban population is heavily More genera l me a su re s of u rba n skewed toward its largest cities ­ concentration—the share of population living The Middle East and North Africa in cities larger than 1 million—are also higher region’s high average urban primacy is largely in the GCC and Mashreq than in the Maghreb due to the high concentration of urban popu- figure 3.2). While the Middle East and North (­ lation in the largest cities in the Mashreq and Africa as a whole is close to the overall average GCC subregions—36 percent and 35 percent, urban concentration (around 40 percent), the respectively. It is far lower in the Maghreb Mashreq and GCC subregions display very subregion, with the largest cities only high concentration patterns (43 percent and accounting for an average of about 19 ­ percent 46 percent, respectively). Conversely, the of the urban population. Maghreb subregion displays low urban con- Across countries within subregions, urban centration, with only 20 percent. primacy varies significantly. In the GCC The evolution of urban population distri- ­ subregion, the concentration of the urban bution in the past 35 years shows the FIGURE 3.2  Urban population distribution is skewed toward large cities in the GCC and the Mashreq, but concentrations are much lower in the Maghreb 60 Urban concentration (%) 50 40 30 20 10 0 ci d ia C ric d a q eb be d l A and Pa an Af an ric re ib an C As fic a hr an G h a Af ia th t ar a as ag si h or as tra e As C ic ut n en rop M M N eE e er ra So st th Am ha Eu Ea dl Sa id tin C M b- La Su Regions Subregionsa Source: UN DESA database, https://population.un.org/wup. Note: “Urban concentration” is the share of population living in large cities. The country-level figures were computed using either (a) cities of 1 million inhabitants or more, or (b) the share represented by the capital city in countries that had no cities exceeding 1 million inhabitants. a. Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Israel, Jordan, Lebanon, and the Syrian Arab Republic; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. 9 4    C ONVERGEN C E increasing divergence between subregions of m illion, as has around 45 percent in the 1 ­ the Middle East and North Africa. As in Mashreq subregion and less than 20 percent Europe and Central Asia and in Sub-Saharan in the Maghreb subregion. Africa, about 40 percent of the urban popu- This urban hierarchy is skewed more lation growth in the Middle East and North toward the largest cities in the Middle East Africa has been in cities with more than and North Africa than it is in other regions of 1 million inhabitants. Although this points to the world (figure 3.3). In the Maghreb subre- a significant increase in the share of urban gion, cities above 1 million concentrate only population living in secondary cities, the about 20 percent of the urban population, speed of this expansion is rather slow (UN which fits a rather low urbanization growth DESA 2019). And the dynamics vary signifi- rate in the past decade (figure 3.3, panel b). cantly from one subregion to another: more This pattern is similar to the one observed in than 60 percent of the GCC countries’ urban European countries, where urbanization has expansion has occurred in cities above happened gradually over time and the urban FIGURE 3.3  Distribution of the urban population skews toward the largest cities in the Middle East and North Africa Share of urban population, by city population class, 2015 a. GCCa b. Maghrebb c. Mashreqc 13.8% 20.0% 19.6% 10.7% 6.6% 37.3% 4.9% 20.5% 8.3% 12.0% 8.7% 64.3% 7.5% 31.9% 33.8% d. Europe and Central Asia e. Latin America and the Caribbean f. East Asia and the Pacific 5.7% 3.9% 13.8% 13.4% 16.6% 6.3% 10.6% 10.4% 24.8% 20.4% 7.8% 8.1% 11.2% 5.9% 6.6% 55.5% 41.2% 37.8% City population class <300,000 300,000–500,000 500,000–1 million 1–5 million 5–10 million ≥10 million Source: UN DESA database, https://population.un.org/wup. a. The Gulf Cooperation Council (GCC) comprises Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. b. The Maghreb subregion comprises Algeria, Libya, Morocco, and Tunisia. c. The Mashreq subregion comprises the Arab Republic of Egypt, Iraq, Israel, Jordan, Lebanon and the Syrian Arab Republic. Wa l l e d U r b a n Ec o n o m i e s    95 system is fairly well balanced. The urban sys- urban concentration with the share that fuel tems of the Mashreq and GCC regions, in exports represent in each country’s total imbalances, with cities above contrast, present ­ merchandise exports, we observe that oil ­ 1 million representing a disproportionate economies tend to display higher urban con- share of the urban system pyramid (figure 3.3, centration. High fuel exports tend to crowd panels a and c). out the private sector and foster the emergence Medium-size and small cities represent sig- of consumption cities. The resource-poor nificantly lower shares of the urban system. In Maghreb countries are concentrated in the this sense, the urban systems of the GCC and lower part of figure 3.4 given their low urban Mashreq countries look much like the urban concentration. All countries of the Mashreq systems of the Latin America and Caribbean subregion but Iraq, in contrast, display high region (figure 3.3, panel e), which is known urban concentration but low fuel exports. The for its imbalances and relatively high primacy. correlation between resource richness and While in Europe and Central Asia, the ratio urban concentration is thus weak, which between cities above 1 million and cities below means that such concentration is not the con- 1 million is around 0.74 (­ figure 3.3, panel d), sequence of oil exports but of other causes. in the Mashreq it is about 1.84. In the Mashreq subregion, the excessive urban primacy and underdevelopment of sec- ondary cities seem to be predominantly the Why do the region’s urban systems so consequence of a political bias favoring the favor primate cities? primate cities. High population growth, migration-driven rapid urbanization, non- There are several possible explanations democratic and centralized political govern- why urban systems in the Middle East and ment, and colonial history have all been North Africa could be fragmented and identified as factors driving further urban skewed toward larger cities: concentration (Faraji 2016). Politically cen- •  Resource-rich countries tend to have less tralized regimes in low- and middle-income diversified economies because private countries tend to provide better services and sector initiatives get crowded out. Major safety in the capital city and give more atten- cities become consumer cities rather than tion to the local population. As such, there is producer cities, attracting large shares of evidence that migrants settling in the primate the population. city come not only from rural areas but also •  Exceptional urban management attracts from small towns and medium-size cities a disproportionate share of firms and (El-Din Haseeb 2012). people. If city leaders of the largest urban Primate cities that are also political capi- areas can reduce congestion costs and fos- tals are on average 25 percent bigger than ter urban agglomeration economies, their primate cities that do not concentrate politi- cities will continue to grow. cal power (Henderson 2002). Besides, as •  Political choices and biases benefit the Bosker, Buringh, and Van Zanden (2013) main urban areas relative to the rest of highlight, the larger size of cities in the Arab the urban system through investments world can be linked to several centuries of and incentives. Populations living in sec- predatory state rule and low trade openness, ondary cities are left behind and have a which also contributed to making them con- greater reason to migrate. sumer cities instead of producer cities.3 These political biases seem to have deeply affected The high primacy and imbalanced urban the development of urban systems in the systems observed in many countries of the Mashreq subregion, resulting in high urban region could be expected in the case of small, concentration and the underdevelopment of oil-dependent economies. When comparing secondary cities. 9 6    C ONVERGEN C E FIGURE 3.4  High urban concentration in the Middle East and North Africa cannot be explained solely by fuel-export-driven consumption cities 70 Kuwait Egypt, Arab Rep. United Arab Emirates High concentration High concentration Saudi Arabia Low fuel exports High fuel exports Urban concentration (%) Lebanon Qatar Jordan Iran, Islamic Rep. Iraq Oman 35 Syrian Arab Morocco Bahrain Republic Tunisia Yemen, Rep. Libya Low concentration Low concentration Low fuel exports High fuel exports Algeria 0 0.5 1.0 Fuel as a share of all exports (%), avg. 2005–15 Source: World Development Indicators and United Nations Department of Economic and Social Affairs (UN DESA) databases. Note: As earlier, “urban concentration” here is calculated as the share of the population concentrated in cities above 1 million. However, for smaller countries with no cities above 1 million, the measure consisted in the share represented by the largest city. Policy and investment constraints are is the only region where location has a strong limiting the productivity and and significant effect on firms’ productivity, development of secondary cities with companies in the capital city having a Compared with other regions of the world, productivity 6 percent higher than firms the Middle East and North Africa is where located in the periphery (figure 3.5). geographical location has the highest impact This gap can be explained by higher con- on firms’ productivity, systematically giving an straints on many dimensions affecting firms’ advantage to companies in the capital city. To activity in geographically peripheral areas in analyze the determinants of firms’ productivity the Middle East and North Africa. With a across regions, we established a model control- similar model controlling for location, firm ling for firm size, sector of activity, location size, and sector of activity, we estimated the (either periphery or capital city), and country increase or decrease in the probability of fac- effects. (For this methodology, see annex 3A.) ing a major constraint for a firm located in the In other regions of the world (excluding the periphery compared with a firm located in the Middle East and North Africa), firms’ size is capital city. As such, firms in the periphery the highest predictor of productivity. For systematically experience a higher probability example, in the East Asia and Pacific region, of facing major constraints in terms of access large companies systematically enjoy a produc- to finance (+9 percent), political instability tivity 14 percent higher than small firms. On (+6 percent), skills (+3 percent), and corrup- the contrary, the Middle East and North Africa tion (+3 percent), as shown in figure 3.6. Wa l l e d U r b a n Ec o n o m i e s    97 FIGURE 3.5 In the Middle East and North Africa, firms in the capital city have 6 percent higher productivity than firms on the periphery—the highest location-related effect of any region in the world 17 13 9 Change in TFP 5 1 –3 –7 –11 –15 Middle East and Europe and Sub-Saharan East Asia Latin America and South Asia North Africa Central Asia Africa and Pacific the Caribbean Increase for large firms versus small firms Increase for firms in capital city versus firms in periphery Source: World Bank Enterprise Surveys 2017. Note: TFP = total factor productivity. T-bars indicate 90 percent confidence levels. “Large firms” = above 100 employees. “Small firms” = fewer than 20 employees. “Periphery” refers to the peripheral areas outside of the capital city. For the methodology, see annex 3A. FIGURE 3.6 In the Middle East and North Africa, companies on the T hese hig her constraints could be periphery are likelier than those in the capital city to face major explained by two underlying factors: the constraints unequal spatial pattern of access to basic services and low fiscal decentralization. The ­ Access to finance unequal spatial pattern of access to basic Political instability services—when considered either as a bundle ­ Electricity or when quality is factored in—supports the Labor skills idea that spatially biased public policies might be leading to imbalanced systems of Corruption cities throughout the Middle East and North Telecommunications Africa (figure 3.7). At first view, access to ser- Tax administration vices seems to be relatively equally distrib- Crime uted through space with only relatively small Access to land differences between access levels when each service (electricity, water, and education) is Transportation considered individually. But if basic services –10 –5 0 5 10 15 are considered as complements, these small Change in probability of dimension being differences add up to large gaps. a major constraint if firm is located on periphery instead of in capital city (%) Equally, if the quality of services is ­ considered—such as number of daily hours Source: World Bank Enterprise Surveys 2011. Note: “Periphery” refers to the peripheral areas outside of the capital city. T-bars indicate 90 percent that electricity is available rather than the confidence levels. simpler metric of the proportion of house- holds ­ connected—then the spatial inequity infrastructure, when opportunity is defined also appears even larger. As a 2016 World as access to a bundle of basic services, or Bank report on access to opportunities in when the quality of services is taken into the region concludes, “Opportunities are account. Coverage of infrastructure services far from universal in access to basic such as improved sources of drinking water 9 8    C ONVERGEN C E FIGURE 3.7  Selected countries, including Tunisia (the most decentralization index elaborated by Ivanyna centralized Maghreb state), show large gaps in access to public and Shah (2012), the Middle East and North services between the primary city and the other urban areas Africa is the second least fiscally decentral- 60 ized region in the world (slightly above Sub-Saharan Africa), which implies that sec- ­ between primary city and others Gap in public services access 50 ondary cities have a lower margin of maneu- ver there than in other regions of the world (percentage points) 40 (figure 3.8). 30 Across countries, the Mashreq subregion is significantly less decentralized than the 20 other two subregions, which could explain its 10 higher urban primacy and concentration (­figure 3.9, panel a). This could support the 0 argument that the Mashreq subregion is Jordan Iraq Tunisia Morocco more concentrated because of its historical Piped water Networked sanitation lack of margin for maneuvering for second- Source: Demographic and Health Surveys (DHS) since 2005. ary cities given the concentration of resources in the capital city. At the country level, we can argue that FIGURE 3.8  Fiscal transfers to local governments in the Middle in most of the Mashreq countries—Egypt, East and North Africa are among the lowest in the world, only Iraq, and eventually Syria and Jordan—the slightly above Sub-Saharan Africa level of fiscal centralization explains urban North America conc ent rat ion at t he n at ion a l level. However, Lebanon displays a relatively Europe and Central Asia high score on the decentralization index East Asia and Pacific (figure 3.9, panel b). Therefore, it is possi- South Asia ble that, in Lebanon, high urban concentra- Latin America and the Caribbean tion results partly from factors other than political biases. Middle East and North Africa However, the productivity premium Sub-Saharan Africa could also be linked to cronyism for firms 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 in the capital city. Cronyism is the practice Fiscal decentralization index of partiality in awarding jobs and other advantages to friends, family relatives, or Source: Ivanyna and Shah 2012. Note: The fiscal decentralization index considers numerous variables: local government autonomy trusted colleagues, especially in politics in rate and base setting for local revenues, self-financing of local expenditures, local responsibility and between politicians and supportive for and control over municipal and social services, and many others. “North America” here comprises Canada and the United States. organizations. For example, firms that were politically ­connected to the Ben Ali family in Tunisia enjoyed an important profit and (especially) adequate sanitation is mark- and market share premia (Rijkers, Freund, edly lower, with significant disparities in and Nucifora 2014). Similarly, Diwan and how they are distributed. Location of resi- Chekir (2015) estimate the market valua- dence explains more than half the variation, tion of political connections to be 20–23 with spatial advantages accruing to capital percent of the value of connected firms. cities and large metropolitan areas that Those firms are naturally more likely to be coincide with poles of economic activity” in the capital city to be closer to the politi- (Krishnan et al. 2016, 49). cal power, which could explain a certain The extent of fiscal decentralization is s h a r e of t h e p ro du c t iv it y p r e m iu m another influential factor. Based on the fiscal observed. Wa l l e d U r b a n Ec o n o m i e s    99 FIGURE 3.9  Fiscal decentralization in the Middle East and North Africa reflects larger transfers of fiscal autonomy in the Maghreb than in the GCC and Mashreq subregions a. Decentralization by subregiona b. Decentralization, selected countries 0.3 0.6 Fiscal decentralization Fiscal decentralization 0.5 0.2 0.4 0.3 0.1 0.2 0.1 0 0 Maghreb GCC Mashreq Tu es Al sia M eria Ar Le cco ep n Jo lic Q n a ud Ira . iA q O ia an Li r ab it ep a R no a by Ar wa b at ub t rd m ni ira ra R o g ab ba u or K Em ab Sa t, Ar yp Eg n d ria te ni Sy U Maghreb Mashreq GCC countries Source: Ivanyna and Shah 2012. Note: The fiscal decentralization index considers numerous variables: local government autonomy in rate and base setting for local revenues, self-financing of local expenditures, local responsibility for and control over municipal and social services, and many others. a. Maghreb refers to Algeria, Libya, Morocco, and Tunisia; Mashreq to the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic; and the Gulf Cooperation Council (GCC) to Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. Regional integration can deliver Rodríguez-Pose (2018) find that larger cit- large markets for the Middle East ies tend to drive the fastest growth in larger and North Africa’s cities countries, while smaller cities were associ- Regional integration policies do not only ated with higher growth in smaller coun- guide trade and cross-border factor mobility; tries (table 3.1). In contrast, several Middle they also shape economic geography within East and North Africa countries have small countries by guiding the placement of infra- domestic markets relative to the size of structure, the location of investments, and their cities. Kuwait, Oman, and Qatar all the movement of people. Regional integration have fewer than 4 million urban inhabit- therefore also influences national urban ants while their primate city populations are dynamics. Just as economic development has larger than 500,000. meant a sectoral reorientation of economies Economic history has shown that greater from agriculture to manufacturing and ser- regional integration goes together with eco- vices, accelerating growth also requires a nomic growth and regional stability. Deep geographical reorganization of production integration entails the EU’s four freedoms4 within and between countries. Most impor- (plus one): the movement across borders of tant, regional integration raises cities’ access goods, services, capital, people, and ideas. to markets, unlocking the scale of demand Regional integration efforts often only touch needed to support the Middle East and North a subset of these, even though they are inter- Africa’s largest and most productive agglom- dependent. Falling costs of transport, com- erations, which appear oversized relative to munication, and data transfer have allowed their domestic markets. trade to become increasingly specialized. A lthough agglomeration economies Trade in components and parts, for instance, favor large cities, large cities may also has allowed firms to organize in international need large markets to flourish. Frick and production networks, specialize in specific 1 0 0    C ONVERGEN C E TABLE 3.1  City size relative to a country’s total urban population is associated with positive or negative effects on economic growth Effect on growth, by total urban population City population < 4 million 4–12 million 12–30 million > 30 million Small cities (< 0.5 million) Positive Positive Negative Negative Medium-size cities (0.5–3 million) Negative Positive Positive Positive Megacities (> 10 million) Negative Negative Negative Positive Source: Frick and Rodríguez-Pose 2018. products or components as links in a value behind, regional integration has generated chain, and benefit from economies of scale. the resources to do so. In fact, interindustry and intraindustry trade The Middle East and North Africa has a (as opposed to trade in final goods) has long history of regional trade; a common cul- increased tremendously over the past half ture and language in most countries; abun- century, especially in East Asia, North dant labor, capital, and natural resources America, and Western Europe. within the region; and few geographic barri- A consequence of these trends has been ers. Yet while other world regions have prof- the agglomeration of production in special- ited from pooling input and output markets ized clusters that benefit from shared, spe- across borders, the Middle East and North ciali zed labor markets; prox i m it y to Africa lags on many aspects of integration. specific input or output markets; and learn- Compared with Europe or East Asia, few of ing across companies, which fosters inno- the region’s countries benefit from growth vation. Trade between nearby countries spillovers from their neighbors. The region has increased significantly as regional produc- “thick borders” restricting the flow of critical tion networks developed. Final demand inputs (labor, capital, and ideas) and outputs may be global, but there are still benefits in (goods and services). It also scores poorly on sourcing important inputs from nearby integrating output markets for goods and ser- countries. An example is the consumer vices and is the most restrictive region when it technology cluster in East Asia where comes to the exchange of ideas—the founda- China has done most of the assembly, but tions of the modern economy. high-tech inputs come from Japan or the This section documents the low integra- Republic of Korea and some lower-tech tion of the region’s countries across these ones from Southeast Asia. four-plus-one dimensions: trade in goods, Manufacturing, but increasingly also ser- trade in services, capital flows, flow of peo- vices, has become more about “specialized ple, and the flow of ideas. All countries that tasks” than finished products. A precondition have reached high-income status allow a high for such cross-national production networks degree of openness to these flows. has been countries’ willingness to open their markets for inputs and outputs and to reduce Trade in goods barriers to trade in goods and services as well as to the flows of capital and labor. Not all Tariffs countries that have actively pursued regional International borders reduce trade by an integration strategies have benefited equally estimated 20–50 percent and increase trade from globalization—a convenient shorthand costs by an average of about 40 percent for the processes just described. But practi- (Anderson and van Wincoop 2003; Arvis et cally all have raised their economic fortunes. al. 2016; Coughlin and Novy 2016). Tariffs And where there has been the political are a significant share of these costs; others w illingness to compensate those who fell ­ arise from currency exchange and language Wa l l e d U r b a n Ec o n o m i e s    101 BOX 3.1  Economic growth can be contagious—but in the Middle East and North Africa, it is not When a country with favorable endowments and FIGURE B3.1.1  Sweden would be far poorer under good policies grows faster, its neighbors benefit Tunisia’s low regional economic spillovers when there is a high degree of economic integration. (constant US$, thousands) Growth spills over to other countries as growing 45 areas take advantage of regional pools of workers, GDP per capita capital, and knowledge. Larger regional markets 40 increase demand and let firms benefit from scale 35 economies. 30 Quantifying the benefits of growth spillovers 25 across countries 1990 1995 2000 2005 2010 2015 Year Trade integration has become deeper over time, often Sweden, actual growth going beyond traditional trade policy to encompass Sweden in Middle East and North Africa areas such as investment, competition, and intellectual Source: Lebrand 2019. property rights protection. This makes ­ implementation more demanding. But the “deeper” the agreement, the more a country will benefit from the growth spillovers lower (figure B3.1.1), with a cumulative GDP loss of from other members. Even with the rapid fall of global US$1.375 billion (2010 constant U.S. dollars). transport costs, close proximity in buyer-supplier net- The analysis also suggests that openness by itself works is still an advantage. Analysis of regional growth would not automatically raise growth significantly in trends over the past ­several decades shows that from the Middle East and North Africa. Simulation of eco- 1990 to 2015, membership in trade agreements was nomic growth trends assuming spillovers consistent with associated with a growth spillover of about 36 percent a deep regional integration agreement raises growth (Lebrand 2019). Associated with this is a spatial multi- in some countries only moderately—and would even plier of 1.56, with deep trade integration increasing the reduce growth in others where poor neighboring perfor- effectiveness of growth-promoting domestic policies by mance would impose a drag on the local economy. 56 percent. Deeper integration will pull neighbors along only However, analysis of regional growth trends over if domestic reforms create faster growth at least in the past several decades shows that economic spill- some countries. But although deepening a regional overs have been strong in Europe and in Organisation integration agreement by itself will not raise a for E conomic Co-operation and Development region’s prospects, the reform pressure that builds (OECD) countries but largely absent in the Middle through more open markets and the opportunities East and North Africa. Among OECD countries, that come with greater integration can stimulate where deep trade integration has been the strongest, growth. the benefits over the past few decades have been even larger. In the Middle East and North Africa, growth rates among countries tied by weak integration agree- Spillover benefits from countries to cities ments have been uncorrelated (Lebrand 2019; see also Trade depth between countries allows cities to ben- Roberts and Deichmann 2011). efit directly from the growth of other cities. However, these spillovers between cities are absent in the Putting Sweden in the Middle East and North Middle East and North Africa, while they are strong in the East Asia and Pacific as well as Latin America Africa would have cost it US$1.375 billion and the Caribbean (figure B3.1.2). If Sweden had been subject to the lack of regional spill- Most notably, East Asia and Pacific cities have overs experienced by Tunisia between 1990 and 2015, experienced much larger spillovers from cities in coun- its GDP per capita in 2015 would have been 30 percent tries with deep trade agreements than have cities in the box continues next page 1 0 2    C ONVERGEN C E BOX 3.1 Economic growth can be contagious—but in the Middle East and North Africa, it is not (continued) FIGURE B3.1.2  Spatial spillovers based on deep trade FIGURE B3.1.3  Bangkok’s per capita GDP would have agreements with neighbors, by region shrunk had it experienced the Middle East and North Africa’s regional economic spillovers 0.6 Estimated spillover coefficient 13 (constant US$, thousands) 0.5 12 0.4 GDP per capita 11 0.3 10 9 0.2 8 0.1 7 0 6 2000 2005 2010 2015 l ric d ci d ric n be d Al Af an Pa an Af ara ib an Year a fic a an th t ia ar a h or as C ic Sa As e er N eE Bangkok, actual growth b- th Am st Su dl Ea Bangkok in Middle East and North Africa id tin M La Source: Lebrand 2019. Region Source: Lebrand 2019. was growing quickly in per capita terms, but its GDP per capita would have been shrunk significantly in the Middle East and North Africa for lack of growth Middle East and North Africa. Figure B3.1.3 shows spillovers. the level of city GDP per capita if Bangkok had been Note: Two regions are excluded from the analysis. The results of the regional in the Middle East and North Africa instead of in East analysis were not significant for South Asia or not of interest for comparison for Asia and Pacific. Over the 2000–15 period, Bangkok Europe and Central Asia. FIGURE 3.10  Many Middle East and North Africa economies have and information barriers. In the Middle East higher average tariffs than their economic peers in other regions and North Africa, countries’ tariffs vary con- Average tariff rates in relation to per capita income, 2015 siderably. The region’s high-income countries (that is, the GCC countries) have relatively QAT 50,000 ARE low average tariffs, but these are still about KWT twice those of their peers in other regions ISR BHR SAU (figure 3.10). The middle- and low-income GDP per capita (US$) OMN 10,000 countries typically have higher tariffs as TUN DZA IRN well— exceeding 10 percent in Algeria, 5,000 LBN MAR JOR Djibouti, the Islamic Republic of Iran, Syria, EGY DJI and Tunisia. 1,000 500 YEM Trade agreements As in other regions, Middle East and North Africa countries have specific agree- ments that allow some trading partners 0 10 20 Tariff rate, applied, simple mean, all products (%) ­ preferential market access, but those agree- ments are relatively weak. Three of these Source: World Development Indicators Database. Note: Full country names correspond to ISO 3 country codes. No data were available for Iraq, Libya, preferential trade agreements are registered the Syrian Arab Republic, and West Bank and Gaza. with the World Trade Organization (WTO), Wa l l e d U r b a n Ec o n o m i e s    103 TABLE 3.2  Trade agreements are fewer and shallower in the Middle East and North Africa than in other regions Average number of provisions in intraregional trade agreements, 2015 Region Total provisions Total enforceable Core enforceablea Number of agreements East Asia and Pacific 18.3 14.4 11.4 32 Europe and Central Asia 14.4 12.2 8.8 85 Latin America and the Caribbean 21.7 15.3 12.7 30 Middle East and North Africa 16.0 9.7 6.3 3 North Americab 22.0 21.0 17.0 1 South Asia 4.0 2.8 2.8 4 Sub-Saharan Africa 22.2 8.4 6.9 9 Source: World Bank calculations based on Hofmann, Osnago, and Ruta 2017. a. “Core provisions” are those included under WTO+ as well as those concerning competition policy, investment, capital movement, and intellectual property rights. Under the World Trade Organization (WTO), WTO+ provisions include areas such as customs regulations, export taxes, antidumping, countervailing measures, technical barriers to trade, and sanitary and phytosanitary standards. b. “North America” here comprises Canada and the United States. binding the partners within the Middle East TABLE 3.3  The Middle East and North Africa has few agreements and North Africa: with important future markets Number of agreements and average number of core legally enforceable provisions •  The GCC, which includes six high-income between Middle East and North Africa countries and those in other regions, 2015 Gulf countries Number of Average number of •  The Pan-Arab Free Trade Area (PAFTA), Region agreements provisions which was initiated in 1997 under the East Asia and Pacific 2 7.5 umbrella of the Arab League and covers Europe and Central Asia 23 10.0 almost the entire Middle East and North Latin America and the Caribbean 1 14.0 Africa region North Americaa 7 10.6 •  The Common Market for Eastern and South Asia 0 0.0 Southern Africa (COMESA), which con- Sub-Saharan Africa 2 5.5 nects Egypt, Libya, and Djibouti as well Source: World Bank calculations based on Hofmann, Osnago, and Ruta 2017. as 16 Sub-Saharan African countries. Note: “Core, legally enforceable” agreements are those included under WTO+ as well as those concerning competition policy, investment, capital movement, and intellectual property rights. The strength or depth of an agreement is Under the World Trade Organization (WTO), WTO+ provisions include areas such as customs regulations, export taxes, antidumping, countervailing measures, technical barriers to trade, and determined by its comprehensiveness and the sanitary and phytosanitary standards. extent to which its provisions are legally a. “North America” here comprises Canada and the United States. enforceable. On average, the Middle East and North Africa’s intraregional agreements eight such provisions. However, PAFTA, include 16 provisions—not too different from which covers most of the region, is a “shal- those in other world regions (table 3.2). They low” agreement, with only two legally are shallower, however, as the lower number enforceable core provisions, because it of total and core enforceable provisions shows. focuses mostly on tariff liberalization for “Core” provisions are those included industrial and agricultural goods only. under WTO+ as well as provisions concern- The largest number of Middle East and ing competition policy, investment, capital North Africa countries’ interregional bilat- movement, and intellectual property rights.5 eral or multilateral agreements are with Among the Middle East and North Africa’s countries of the Europe and Central Asia three intraregional agreements, COMESA is region—mostly with Western European fairly comprehensive, with nine legally countries and the EU (table 3.3). Such agree- enforceable core provisions, but it covers only ments with countries of East Asia and the three Middle East and North Africa coun- Pacific—the most important world market in tries. The GCC is also comprehensive, with the future—tend to be relatively shallow. 1 0 4    C ONVERGEN C E Europe and the Americas have more compre- The importance of trade is highest in the hensive agreements with that dynamic region. United Arab Emirates, where exports plus imports are about 140 percent of GDP, high- Exports and imports lighting the importance of Dubai and Abu At 4.1 percent each, the Middle East and Dhabi as global trading hubs. Trade is impor- North Africa’s share of world exports and tant in fuel-exporting countries but also in imports matches its share of global GDP Jordan, Morocco, and Tunisia, where trade is almost exactly (table 3.4). The export share is 60–80 percent of GDP (figure 3.11). largely due to a high share of raw materials Although the region’s overall trade levels exports—mostly oil and natural gas, which roughly correspond to its share of the global make up about a third of all exports.6 The e c onomy, t rad e — e sp e c i a l ly non f u el shares of other product types, especially capi- exports—could be significantly larger (Behar tal goods exports, are relatively small. and Freund 2011; Bhattacharya and Wolde 2010; Bourdet and Persson 2014; Hoekman TABLE 3.4  Trade in the Middle East and North 2016). One study estimated that a typical Africa is still dependent on natural resources percent Middle East and North Africa country could double to quadruple its exports given eco- Global share, type Exports Imports nomic and geographic characteristics and Trade assuming best-practice trade policies (Behar Total 4.1 4.1 and Freund 2011). Imports, in contrast, are Raw materials 10.5 3.3 more in line with expectations. Intermediates 3.0 4.7 Intraregional trade has been a key driver Consumer goods 3.4 3.6 of trade and, in turn, of economic growth in Capital goods 0.9 2.9 other world regions, notably Europe and East General Asia. The Middle East and North Africa’s GDP 4.1 share of total intraregional trade had been Population 5.9 gradually increasing from the 1980s through Source: World Bank, World Integrated Trade Solution (WITS) database. the mid-2000s before dropping during the 2008– 09 world financial crisis and only FIGURE 3.11  Merchandise trade as a share of GDP in many Middle East and North Africa countries is quite low recently recovering (figure 3.12). 160 Trade in services 140 Share of GDP from total exports Trade in services has increased globally, 120 commanding a substantial share of GDP, and imports (%) 100 including in the Middle East and North 80 Africa (Loungani et al. 2017). Services—such as financial and legal, telecom, and transport 60 services—are critical intermediate inputs, 40 and their trade can be an important mecha- 20 nism to transfer know-how and innovations. Access to high-quality services consequently 0 has a strong influence on manufacturing pro- . ductivity, importantly through its effect on ic . . en es ab Tun ia Ku an or n O o Sa Q n Em isia ep m ep ep Ira pt, A jib n Ba ria q ba it i A tar Is b ti Jo el M no c a y D rai Le wa Ira n, ra ou b m at oc ra rd m ,R la R R ge ra ud a h Ye ir Is foreign direct investment (FDI) f lows Al (Hoekman and Shepherd 2017). Ar Eg d In the Middle East and North Africa, the te ni U Country average share of GDP from services trade Source: World Development Indicators Database. has  fluctuated since 1990 between about Note: No data were available for Libya, the Syrian Arab Republic, and West Bank and Gaza. 20  ­p ercent and 25 percent (figure 3.13). Wa l l e d U r b a n Ec o n o m i e s    105 FIGURE 3.12  Only a small share of global intraregional merchandise trade occurs within the Middle East and North Africa 70 Share of global intraregional trade, 1984–2016 (%) 60 50 40 30 20 10 0 20 1 19 4 19 5 19 6 19 7 19 8 19 9 19 0 19 1 19 2 19 3 19 4 19 5 19 6 19 7 19 8 20 9 20 0 20 1 20 2 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 12 20 3 20 4 20 5 16 1 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 0 0 0 0 0 0 0 0 0 0 1 1 1 1 19 20 East Asia and Pacific Europe and Central Asia Latin America and Caribbean Middle East and North Africa North America South Asia Sub-Saharan Africa Source: Derived from the World Trade Flows database, Bilateral Data files, of the Center for International Data, University of California, Davis, https://cid.econ​ .ucdavis.edu/Html/WTF_bilateral.html. Note: Regions include all countries regardless of income status. “North America” comprises Canada and the United States. FIGURE 3.13  As a share of GDP, the Middle East and North Africa’s intraregional service trade is small relative to the size of its economies Share of GDP from intraregional services trade (%) 40 35 30 25 20 15 10 5 0 01 13 09 11 00 02 98 99 90 03 10 14 91 92 96 97 04 08 05 93 06 94 07 12 16 15 95 20 20 20 20 20 20 19 19 19 20 20 20 19 19 19 19 20 20 20 19 20 19 20 20 20 20 19 East Asia and Pacific Europe and Central Asia Latin America and Caribbean Middle East and North Africa North America South Asia Sub-Saharan Africa Source: World Development Indicators Database. Note: The data shown are moving three-year averages of countries in each region for which numbers were reported, including high-income countries. “North America” here comprises Canada and the United States. 1 0 6    C ONVERGEN C E TABLE 3.5  The Middle East and North Africa has higher service trade restrictions than any other region Average service trade restriction index Restriction East Asia & Europe & Latin America & Middle East & North South Sub-Saharan category Pacific Central Asia Caribbean North Africa Americaa Asia Africa Overall 32.6 19.1 21.0 45.3 19.7 43.9 32.0 Financial 24.4 9.8 18.5 42.1 21.1 38.1 26.7 Professional 55.7 43.7 38.0 61.6 47.5 60.7 48.7 Retail 20.8 7.3 8.3 30.8 0.0 30.0 22.8 Telecommunications 34.4 8.1 20.8 44.2 25.0 45.0 38.6 Transportation 33.5 26.1 22.1 50.1 12.0 50.4 29.9 Source: Borchert, Gootiiz, and Mattoo 2013. Note: The table displays the average index scores of countries in each region for which numbers were reported, including high-income countries. a. “North America” here comprises Canada and the United States. Shares are highest in Lebanon (where they uses, encourages financial development, pro- represent more than 50 percent of GDP) and motes portfolio diversification, and imposes in Jordan, Kuwait, and Qatar. greater discipline on monetary policy. Fi rm-level data most clea rly show Estimates of the economic benefits of capital that restrictions to services trade affect pro- liberalization range from a 1 percentage duc tivit y and ex por t performance point increase in annual GDP to as much as (Hoekman and Shepherd 2017).7 When a 14 percent permanent increase in con- countries reduced their service sector restric- sumption in the most capital-scarce coun- tions, the performance of firms relying on tries (Hoxha, Kalemli-Ozcan, and Vollrath such services improved. Given that such 2013; Kose et al. 2006).9 services inputs can be up to two-fifths of the ­ After North America10 and Europe and gross value of manufactured exports, this is Central Asia, the Middle East and North not surprising.8 Africa region has the highest average level of Countries in the Middle East and North financial openness, although it has retreated Africa for which data are available have, on somewhat over the past 20 years (figure 3.14). average, larger restrictions than other world Openness varies by country: several GCC regions (table 3.5). Egypt, the Islamic countries as well as Israel, Jordan, and the Republic of Iran, and Qatar have the highest Republic of Yemen have open capital mar- service restriction index scores, while Algeria, kets, while eight other economies in the Morocco, and the Republic of Yemen have region maintain significant restrictions on the lowest. The highest restrictions are in capital flows. professional and transportation services. The region’s relatively high (average) capi- Retail sector restrictions are the lowest but tal openness does not seem to have led to cor- still considerably higher than in Latin responding FDI inflows. Among international America and the Caribbean or in Europe and capital flows, FDI is arguably the most Central Asia. Thirteen Middle East and important for development. It is usually North Africa countries are in the dataset; in aimed at new, productive investments and seven of those countries, professional services also often comes with know-how for technol- are the most restricted sector. ogy transfer and innovation. Yet the average net inflows of FDI relative to GDP across Middle East and North Africa countries were Capital flows lower than for any other region in 2016 Financial openness can generally be good (­ figure 3.15). After an increase in relative FDI for growth. It tends to reduce the cost of flows in the mid-2000s, they reverted to lev- capital that can be invested for productive els last seen in 2000. Wa l l e d U r b a n Ec o n o m i e s    107 FIGURE 3.14  Overall, the Middle East and North Africa has a high degree of capital openness 1.5 1.0 Capital openness index 0.5 0 –0.5 –1.0 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Years East Asia and Pacific Europe and Central Asia Latin America and Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Source: World Bank calculations from the Chinn-Ito Financial Openness Index (KAOPEN) 2016: http://web.pdx.edu/~ito/Chinn-Ito_website.htm. Note: North America (Canada and the United States), which has the maximum value, is not shown. KAOPEN draws on the International Monetary Fund’s (IMF) Annual Report on Exchange Arrangements and Exchange Restrictions (AREAER). It is the first principal component of four binary variables: existence of multiple exchange rates, restrictions on current account, capital account transactions, and requirement of the surrender of export proceeds. FIGURE 3.15  FDI inflows to the Middle East and North Africa remain low despite the region’s relatively high capital openness 18 Average share of GDP from FDI net inflows (%) 16 14 12 10 8 6 4 2 0 00 03 04 05 09 11 12 90 91 92 93 94 95 96 97 98 99 01 02 06 07 08 10 13 14 15 16 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 East Asia and Pacific Europe and Central Asia Latin America and Caribbean Middle East and North Africa North America South Asia Sub-Saharan Africa Source: World Development Indicators Database. Note: FDI = foreign direct investment. “North America” comprises Canada and the United States. 1 0 8    C ONVERGEN C E Among the region’s countries, Djibouti persistently large wage gaps between coun- has received by far the largest relative FDI tries, continued slow growth in many poorer inflow. Large investments, notably from countries, and rising demand from high- China, boosted Djibouti’s relatively small income countries. When poorly managed, economy, increasing strongly around 2003 however, migration can encourage human and averaging about 9 percent of GDP trafficking and create social conflict as local between 2010 and 2015. Relative FDI inflows workers compete with newcomers or as have also been large in Israel, Jordan, migrants become scapegoats for other social Lebanon, and Morocco, where they were ills. 3 percent or higher during 2011–16. The low- The Middle East and North Africa has est FDI shares (less than 1 percent of GDP) high cross-country migration (table 3.6).12 during this time were in Algeria, the Islamic More than 40 million migrants live in the Republic of Iran, Kuwait, Qatar, and the region, of whom about a third (14 million) Republic of Yemen (which experienced net are from within the region. So, the Middle outflows).11 East and North Africa, which has about Institutional weakness may explain why 6 percent of the world’s population, hosts capital openness has not translated into FDI 15 percent of the world’s migrants. It is also in the Middle East and North Africa (Okada a large source of migrants: about 28 million 2013). With higher-quality institutions, capi- of its citizens live in other countries, about tal openness has a greater positive effect on half of them in other Middle East and North international capital flows. A study of 17 of Africa countries. For migration within the the region’s economies from 1989 to 2009 region, Egypt, Syria, and West Bank and shows that capital liberalization is effective Gaza are the largest source economies, only in an environment with good institu- accounting for three-quarters of all regional tions and low political risk (Gammoudi and migration. Jordan, Lebanon, Saudi Arabia, Cherif 2015). and the United Arab Emirates combined host about t wo-thirds of all regional migrants. Flows of people Moving from places with a low demand for labor and low wages to one where TABLE 3.6  Migration in the Middle East and North demand is higher is one of the best ways for Africa has been driven by both job seekers and people to improve their welfare. Open bor- refugees ders could more than double the wages of an Largest intraregional bilateral migrant stocks, 2017 estimates average worker in a less-developed country Migrants (including nonmigrants) without significantly Origin Destination (thousands) affecting real wage rates in high-income West Bank and Gaza Jordan 2,047 countries (Kennan 2013). Well-managed Syrian Arab Republic Lebanon 1,209 migration benefits both sending and receiving countries (EIU 2016; Pritchett 2006). Sending Egypt, Arab Rep. Saudi Arabia 872 countries get remittances and investments Egypt, Arab Rep. United Arab Emirates 858 from migrants and also benefit from the Syrian Arab Republic Saudi Arabia 746 expertise and wealth brought back through Syrian Arab Republic Jordan 725 circular migration. Receiving countries Yemen, Rep. Saudi Arabia 697 increase their labor supply in areas where West Bank and Gaza Syrian Arab Republic 630 there are shortages, which will be increas- West Bank and Gaza Lebanon 507 ingly needed where demographic dynamics Egypt, Arab Rep. Kuwait 423 shrink the labor pool. Source: World Bank Migration and Remittances Database 2017, Global The gains from (and demands for) Knowledge Partnership on Migration and Development (KNOMAD): m igration are likely to remain high given ­ https://www.knomad.org/data/migration/emigration. Wa l l e d U r b a n Ec o n o m i e s    109 The nature of migration varies greatly about 9 percent. Saudi Arabia and the United between countries by reasons for migration, Arab Emirates are the world’s second and dominant destination, and source of third largest origins of remittances globally, migrants, as these examples show: behind only the United States. These two GCC countries account for two-thirds of all •  Several Maghreb countries (Algeria, remittances from the Middle East and North Morocco, Tunisia) and the Islamic Africa, most of which go to South Asian and Republic of Iran have sent many migrants, East Asian countries. but most seek employment in Europe or For eight recipient economies in the North America. Middle East and North Africa, remittances •  Most migrants from Egypt and Jordan, in are equivalent to more than 3 percent of contrast, move to other countries in the GDP—and are more than 10 percent of GDP region, mostly in the GCC. in Lebanon, West Bank and Gaza, the •  Among the largest migration flows are Republic of Yemen, and Jordan (figure 3.16). refugees from Iraq, Syria, and West Bank Egypt, Lebanon, and Morocco are the largest and Gaza, who have mostly moved to recipients of such flows in the region neighboring countries. (table 3.7). •  The largest destination for economic Remit tance f lows in 2016 between migrants is the GCC, where migrant Middle East and North Africa countries shares in the population are often well totaled about US$31.6 billion, or about over 50 percent. Although the GCC is 5 percent of global flows. About 77 percent an important destination for migrants of these come from GCC countries, partic- from some Middle East and North Africa ularly Saudi A rabia, Kuwait, and the countries, most of its migrants come from Un ited A rab E m i rates , wh i le E g y pt , South Asia. the Republic of Yemen, and Jordan are the Remittances are second in magnitude only largest recipients of intraregional remit- to FDI. Middle East and North Africa coun- tances (table 3.8). tries are the source of about 20 percent of Remittances are an important social pro- global remittances and the destination of tection mechanism for the region’s economies FIGURE 3.16  Remittances are significant contributors to several Middle East and North Africa economies 16 remittances received (%) Share of GDP from 12 8 4 0 s ria n ti . . q l n t an a Em a a . on co ar e ep ep ai ep te ou ai Ira a bi si az ra at w oc ge rd n m ni ira hr ra R R ,R jib G Is ba Q Ku O Jo Tu or Al Ba iA ab ic D d en Le M m an ud Ar m la ab Sa k Ye Is t, an Ar yp n, tB Eg d Ira te es ni W U Sources: World Bank Migration and Remittances Database 2017 (Global Knowledge Partnership on Migration and Development [KNOMAD]: https://www​ .knomad.org/data/migration/emigration); World Development Indicators Database. Note: No GDP data are available for Libya and the Syrian Arab Republic. Because remittances are negligible for Bahrain, Kuwait, Oman, Saudi Arabia, and the United Arab Emirates, their totals amount to 0 percent of GDP in the figure. 1 1 0    C ONVERGEN C E TABLE 3.7  Large migration flows have led to equally large remittance flows in the Middle East and North Africa Estimated remittance flows and global rank, 2016 Sending countries Receiving countries Remittances Remittances Country (US$, millions) Global rank Country (US$, millions) Global rank Saudi Arabia 46,725 2 Egypt, Arab Rep. 19,983 7 United Arab Emirates 32,978 3 Lebanon 7,955 18 Kuwait 11,733 13 Morocco 7,467 20 Qatar 10,677 14 Jordan 4,418 34 Oman 4,486 24 Yemen, Rep. 3,351 39 Israel 2,926 31 Algeria 2,093 56 Jordan 2,857 34 West Bank and Gaza 2,034 57 Bahrain 2,705 38 Tunisia 1,903 61 Lebanon 1,167 55 Syrian Arab Republic 1,623 66 Libya 1,004 61 Iran, Islamic Rep. 1,379 72 West Bank and Gaza 915 65 Iraq 1,035 84 Syrian Arab Republic 418 85 Israel 990 85 Egypt, Arab Rep. 400 88 Qatar 668 92 Iran, Islamic Rep. 296 96 Saudi Arabia 287 117 Algeria 200 118 Djibouti 62 147 Iraq 193 122 Oman 40 155 Morocco 102 142 Kuwait 4 177 Tunisia 71 158 Bahrain .. 182 Yemen, Rep. 62 163 Libya .. 182 Djibouti 7 200 United Arab Emirates .. 182 Total 119,922 n.a. 55,291 n.a. Source: World Bank Migration and Remittances Database 2017, Global Knowledge Partnership on Migration and Development (KNOMAD): https://www​ .knomad.org/data/migration/emigration. Note: .. = negligible. n.a. = not applicable. TABLE 3.8  Largest intraregional estimated remittance flows in the with large shares of out-migrants (Devereux Middle East and North Africa, 2016 2016). Their impact on labor markets and eco- Origin Destination Remittances (US$, millions) nomic growth are somewhat more ambiguous Saudi Arabia Egypt, Arab Rep. 7,739 (Chami et al. 2018). There is evidence that Kuwait Egypt, Arab Rep. 3,156 remittances reduce labor force participation in Saudi Arabia Yemen, Rep. 2,062 receiving countries and raise informality. But United Arab Emirates Egypt, Arab Rep. 1,953 they also reduce unemployment in some Saudi Arabia Jordan 1,713 ­ sectors, such as construction and real estate, Saudi Arabia Lebanon 1,624 while reducing employment in manufacturing. Jordan Egypt, Arab Rep. 1,293 These findings do not hold in fragile states, Qatar Egypt, Arab Rep. 1,070 however, where remittances support wage Jordan West Bank and Gaza 936 growth and do not have the same negative United Arab Emirates Jordan 864 effects on labor force participation. Countries can develop strategies to counter these poten- Source: World Bank Migration and Remittances Database 2017, Global Knowledge Partnership on Migration and Development (KNOMAD): https://www.knomad.org/data/migration​ tially negative impacts to take full advantage /emigration. of these external financial resources. Wa l l e d U r b a n Ec o n o m i e s    111 Flow of ideas almost all of the region’s countries rank among the lower half of countries, with sev- The four freedoms—of the movement of eral near the bottom, indicating severe goods, services, capital, and people—are core restrictions on the media. In Freedom principles of regional integration. But they House’s 2017 index of press freedom for 199 also reflect an economic structure that has countries and territories, the average rank of been changing. In the traditional economy, Middle East and North Africa countries was manufacturing tends to be highly concen- 155, with only one country in the top 100: trated to exploit scale economies, services are Israel, ranking 64th (Freedom House 2017a). delivered in person through local branch Such rankings persist in other press freedom offices or traveling consultants, access to cap- indexes as well (RSF 2018a, 2018b). Similar ital often requires face-to-face contacts, and restrictions also exist on communication via people must move to where the jobs are. The the internet. Of 11 Middle East and North largest share of the economy in most coun- Africa countries included among 65 world- tries still follows this pattern. wide in Freedom House’s Freedom on the However, digitization of almost every aspect Net index, six ranked as “not free,” six as of the economy is disrupting the way we trade “partially free,” and none as “free” (Freedom and work (World Bank 2016). Automation will House 2017b). enable more decentralized production, even in In the use of digital technology more gen- higher-wage locations. Many services become erally, the region’s countries do somewhat tradable and can be delivered over the internet better (figure 3.17). The World Bank’s Digital from anywhere. Capital flows freely around Adoption Index (DAI) is a summary measure the world, a trend that could be boosted by of the availability and use of digital technol- digital currencies. And although most jobs are ogy among businesses, people, and govern- still place-specific, people can perform more ment.14 The average rank of Middle East and and more tasks remotely. North Africa economies among 183 econo- With these trends, a “fifth freedom” is mies globally was 82 in the 2016 DAI. becoming more important: the free move- Bahrain, Israel, and the United Arab Emirates ment of ideas across borders. Growth will ranked in the top 20, while Iraq, Djibouti, come from knowledge-intensive activities, Syria, and the Republic of Yemen were near and new ideas and data will become ever the bottom. more important factors of production. However, as the World Development C ou nt r ie s do not genera l ly blo ck Report 2016: Digital Dividends has argued, ­ information that has a clear economic use, technology adoption is only one aspect of such as technical publications, patent docu- digital development (World Bank 2016). At mentation, or financial data. Many block least as important are the so-called analog other types of information that they see as complements that allow firms, workers, and inconvenient or challenging to existing power officials to use these tools effectively: regula- structures. Yet the most innovative and tions that ensure that all firms can connect productive places allow all information to ­ and compete; skills that complement technol- flow freely and promote discussion and ogy so workers can be more productive; and exchange of ideas. By contrast, where infor- a high degree of accountability among public mation exchange is constrained, governments officials so e-government is deployed to often try to induce innovation through plan- improve service delivery, not to increase gov- ning, intervention, and subsidies—frequently ernment control. with limited success. A f ur ther channel for information Middle East and North Africa countries exchange is face-to-face interaction, which do poorly when it comes to the free flow of suffers in the Middle East and North Africa information.13 In surveys of press freedom, because of visa restrictions. The number of 1 1 2    C ONVERGEN C E FIGURE 3.17  Use of digital technologies correlates closely with economic wealth Digital Adoption Index, 2016 QAT 50,000 KWT ARE ISR SAU BHR OMN 10,000 LBN GDP per capita (US$) IRN 5,000 IRQ JOR DZA TUN PSE EGY MAR YEM DJI 1,000 500 0 0.25 0.50 0.75 Digital Adoption Index High income Upper middle income Lower middle income Low income Source: World Bank 2016. Note: Insufficient data were available for Libya and the Syrian Arab Republic. The World Bank’s Digital Adoption Index (DAI) includes business, people, and government components, each comprising several subindexes. Full country names correspond to ISO 3 country codes. other countries that require visas from the The region’s port cities can amplify their region’s residents is high, even for wealthy role as regional hubs and gateways GCC countries (figure 3.18).15 A significant Global and regional integration are com- number of its economies also require visas plementary. Regional integration enlarges from most other countries’ nationals. In markets for outputs. Perhaps even more contrast to most other regions, Middle East important, it also increases access to critical and North Africa countries also require inputs including labor, capital, and compo- visas from residents of neighboring coun- nents. This makes it easier for firms to break tries more often than one might expect. into global value chains such as those that Within the region, visas are required in stretch from Europe to East Asia. 60 percent of all country-pairs, as opposed The nonphysical barriers to closer integra- to less than 20 percent in Latin America and tion in the Middle East and North Africa are the Caribbean, a region where most coun- formidable, as argued elsewhere in this tries also share a similar culture and report. But upgrading of transport infra- language. structure will still be important in much of Wa l l e d U r b a n Ec o n o m i e s    113 the region. For regional integration, ports FIGURE 3.18  Many Middle East and North Africa countries make it play a central role (Arvis et al. 2018).16 hard to visit, and their citizens also face difficulties traveling elsewhere Transshipment ports are hubs where cargo Visa requirements, 2017 (usually containers) is exchanged between 200 Hard to get around (number of countries for which visa is needed) large vessels serving global trunk trade routes and smaller ones serving regional ports. IRQ Smaller ports serve as gateways to a local or IRN LBN YEM SYR PSE national hinterland and sometimes also serve DJI JOR LBY 150 neighboring countries. Ports compete for EGY TUN DZA MAR cargo. They can focus on transshipment busi- OMN SAU ness if they are in a favorable location. They can concentrate on so-called captive cargo, BHR KWT QAT which is produced or consumed in and 100 around the port location. Or they can serve a larger hinterland if it is well connected to the ARE port. Arvis et al. (2018) analyze port develop- 50 ISR ment in the Mediterranean, which borders eight Middle East and North Africa coun- tries. Port development strategies need to consider three elements: (a) maritime connec- 0 tivity (the ease of reaching other ports); (b) port efficiency (the time and money cost 0 50 100 150 200 Hard to get in (number of countries requiring a visa to visit) of processing cargo); and (c) hinterland con- nectivity (the size and proximity of economic Source: Passport Index database, http://www.passportindex.org. Note: For instance, citizens of Jordan require a visa when visiting about 150 countries (“hard to get and population centers). Connectivity within around”), but citizens from only about 60 countries need a visa to visit Jordan (“hard to get in”). the Mediterranean has been declining; it Circles are proportional to per capita GDP. Visa on arrival and eTA (electronic travel authority) are considered visa-free in this chart. dropped from 2,279 interport links in 2009 to 1,532 in 2016. Seventy percent of Mediterranean cargo traffic is between transshipment only, with few spillovers to European ports, 15 percent between Europe local or hinterland development. Tanger and North Africa, and 10 percent between Med, in contrast, was built as part of an Europe and West Asia. Only 5 percent is overall development strategy for northern between Middle East and North Africa coun- Morocco. It serves not just as a transship- tries. Port efficiency depends largely on size. ment but also as a gateway port connecting a There are large economies of scale in port large and populous area with many European operations and related services, so larger ports. Its hinterland is well connected by rail ports will attract larger ships and more cargo. and road and has become a large center for Smaller ports will struggle if they cannot automobile component manufacturing and scale up traffic by developing local economic assembly among other industries. In Port clusters or by expanding their hinterland. Said, after a change of governance structure Hinterland connectivity is not required for in 2014, there is also now a development plan transshipment hubs, but such ports can to better connect to Cairo and the rest of exploit scale advantages when also serving as Egypt. However, the port will compete with gateway ports. Port Said (Egypt) and Tanger several other Egyptian ports that aim to serve Med (Morocco) are transshipment ports at the same domestic markets. t he easter n a nd wester n end of t he For most Mediterranean ports in the Mediterranean, respectively. Port Said had Middle East and North Africa, becoming a limited land connectivity to Egypt’s popula- transshipment hub is unrealistic, especially tion centers and has largely focused on for those that are farther from the trunk 1 1 4    C ONVERGEN C E trade route. But many ports can enhance or personal  exchanges. Further, second- their roles as national or even regional gate- ary cities can have a more significant role way ports. The best would benefit if in their national economies if coordinated Mediterranean economic regions become less complementary policies and investments sup- fragmented and if further concentration of port improvements in the business ecosys- maritime networks occurs. This requires pol- tem alongside providing critical services and icies to strengthen both local economic infrastructure. Recommendations to reduce potential and hinterland connectivity. Port the barriers to these four freedoms and the development thus needs to be integrated with flow of ideas will follow in chapter 6. broader regional and trade strategies, includ- ing infrastructure investments, logistics development, and broader governance Annex 3A Methodology for reforms that ease doing business. Although analyzing productivity across governments cannot directly influence mari- regions time networks, they can make their ports Firm total factor productivity (TFP)—the more attractive through local development ability to generate greater outputs with lower and by expanding their regional reach. inputs—is one of the key elements of eco- nomic growth. Considerable scholarly analy- si s h a s b e e n de vot e d to m e a su r i ng Concluding remarks productivity, especially since Robert Solow’s Although urban primacy in Middle East groundbreaking work (Solow 1957). To esti- and North Africa countries is higher than in mate the productivity of firms in lower- any other world region, the region’s major income economies, researchers have turned cities appear to be too large, in part because to analysis using survey-based data, often in of the constraints on market access confront- the absence of comparable census data. The ing them. They may be too large for their World Bank’s Enterprise Surveys, which pro- local markets but would be rightsized if they vide detailed firm-level data collected by the could more easily serve a much larger Bank’s Enterprise Analysis unit, is well suited regional market. for such an inquiry. The largest cities in well-integrated regions To estimate TFP, we begin with a Cobb- attract human and financial capital from Douglas production function in the following their neighbors and produce goods and ser- form: vices for the entire region. London, Paris, or a a Vienna are large relative to their countries’ VAi = Ai Ki k Li l , (3A.1) populations but not relative to the markets they can serve. Historically, this was the case where VAi is firm-level value added, a function in the Middle East and North Africa, of inputs of capital (Ki), and labor (Li). whereby the largest cities were once closely Firms’ efficiency of production is mea- connected and served as centers of produc- sured by the term Ai, which is the portion of tion, innovation, and learning both region- output that cannot be directly attributed to ally and elsewhere in the world. Today, in the utilized inputs. We refer to the above contrast, most cities that could play much model, or rather its version with natural loga- greater regional roles are prevented from rithm applied on both sides, as VAKL. VA, doing so by numerous barriers to regional K, and L are proxied using the questions economic integration. available in the Enterprise Surveys. More pre- The chapter has provided evidence cisely, VA is proxied by the difference that the flows of goods, services, capital, between the total annual sales of the estab- and labor—what the EU calls the “four lishment (variable d2 in the data) and total ­ f reedoms”—face high barriers, as do the annual cost of inputs (variable n2e); K is flows of ideas, for instance, through media prox ied by the replacement value of Wa l l e d U r b a n Ec o n o m i e s    115 machinery, vehicles, and equipment (variable Notes n7a); and L is proxied by the total annual cost 1. The Gulf Cooperation Council (GCC) coun- of labor (variable n2a). tries are Bahrain, Kuwait, Oman, Qatar, To analyze differences in productivity Saudi Arabia, and the United Arab Emirates. across regions, regression analysis was used, The Mashreq subregion comprises the Arab including controls for sector of activity, firm Republic of Egypt, Iraq, Jordan, Lebanon, size, and location, as follows: and the Syrian Arab Republic. The Maghreb subregion comprises Algeria, Libya, Morocco,  log (TFP_VAKL )i = α + β Loci + γ countryi and Tunisia. 2. “Periphery” refers to the peripheral areas + α s + σ g + εi ,  outside of the capital city. (3A.2) 3. This typology comes from Weber (1958). 4. One of the original core objectives of where TFP_VAKL is the TFP estimate the  European Economic Community  (EEC), obtained from the VAKL model for each set out in the 1957 Treaty of Rome, was the firm; Loci is the firm location (either periph- development of a common market offering ery or capital city); country i are country free movement of goods, service, people, and dummies; a s are sector dummies; s g are size capital. dummies (small, medium, and large); and e i 5. WTO+ provisions include areas such as cus- is the discrepancy term. toms regulations, export taxes, antidumping, This regression was run individually for countervailing measures, technical barriers to each region to identify the effect of location trade, or sanitary and phytosanitary standards. and firm size on TFP, controlling for other 6. Jaud and Freund (2015) estimate that nonpe- troleum exports were as low as 64 percent factors. below what would be expected given the Then, to explore the factors leading to the region’s characteristics, and imports were as high importance of location in the Middle high as 22 percent over what one would East and North Africa region, a second expect. In other words, the region seems to regression, consisting of a linear probability underexport and overimport. model was used to identify the increase in the 7. Examples of service trade restrictions probability for firms to face a constraint if include limits or discrimination in licens- they are located in the periphery instead of in ing, constraints on foreign equity, or vari- the capital city. The equation was specified as ous requirements or limitations applied to follows: foreign providers. Not all of these are explicitly designed to restrict foreign firms’ Const i = α + β Loci + γ countryi access; some apply to domestic firms as well (3A.3) and may be designed to ensure safety and + α s + σ g + ε i ., quality. But de facto these often favor domestic firms. where Consti is a dummy qualifying the con- 8. The nature of service trade restrictions has straint as being either a major constraint or a not been studied as extensively as those for minor constraint; Loci is the firm location goods trade, but a recent effort generated a (either periphery or capital city); countryi are first overview for 103 countries (Borchert, country dummies; a s are sector dummies; s g Gootiiz, and Mattoo 2013). OECD and other are size dummies; and e i is the discrepancy high-income countries maintain restrictions in sectors such as professional services and term. some types of transportation, while some It is important to note that, in the case of low- to middle-income countries are surpris- the constraint associated with access to elec- ingly open. tricity, the Loci was not defined as a dummy 9. Recent financial crises have challenged the pre- differentiating periphery and capital city vailing notion that free capital flows support but  a dummy differentiating cities above low- and middle-income countries (Rodrik million and cities below 1 million. 1 ­ 2018). But this is mostly an argument about 1 1 6    C ONVERGEN C E the benefits of temporary capital account Bhattacharya, M. R., and H. Wolde. 2010. restrictions as a “second best” instrument “Constraints on Growth in the M ENA to  avert severe financial or macroeconomic Region.” Working Paper 10-30, International imbalances. Monetary Fund, Washington, DC. 10. In this chapter, North America refers to Borchert, I., B. Gootiiz, and A. Mattoo. 2013. Canada and the United States. “Policy Barriers to International Trade in 11. No FDI data were available for Libya and Services: Evidence from a New Database.” Syria. World Bank Economic Review 28 (1): 162–88. 12. Migration statistics are difficult to collect. Bosker, M., E. Buringh, and J. L. Van Zanden. These numbers are estimates from the World 2013. “From B ag hdad to L ondon: Bank Migration and Remittances Database Un raveli ng Urba n Development i n 2017, which are largely based on data com- Europe, the Middle East, and North Africa, piled by the United Nations Statistics Division. 80 0 –180 0.” Re vie w of Economic s and 13. Form of government strongly correlates with Statistics 95 (4): 1418–37. freedom of information access. Of the 20 Bourdet, Yves, and Maria Persson. 2014. Middle East and North Africa economies, 14 “E xpanding and Diversif ying South are considered authoritarian, 4 are hybrid Mediterranean E xpor ts th roug h Trade regimes, and 2 are flawed democracies. None Facilitation.” Development Policy Review is a full democracy (EIU 2018). 32 (6): 675–99. 14. Each component of the Digital Adoption C ha m i , R a lph , E k keha rd E r nst , C on nel Index comprises several subindexes (World Fullenkamp, and Anne Oeking, 2018. “Are Bank 2016): The business component Remittances Good for Labor Markets in LICs, includes the share of firms with a website, the MICs and Fragile States?” IMF Working number of secure servers, download speeds, Papers 18/102, International Monetary Fund, and 3G coverage. The people component Washington, DC. includes mobile access and internet access at Coughlin, C. C., and D. Novy. 2016. “Estimating home. The government component includes Border Effects: T he I mpact of Spatial core administrative systems, online public Aggregation.” Working Paper 2016-006C, services, and digital identification. Federal Reserve Bank of St. Louis, St. Louis, 15. Visa requirement data are from the Passport MO. Index database, http://www.passportindex.org. Devereux, S. 2016. “Social Protection and Safety 16. This subsection is based largely on Arvis et al. Nets in the Middle East and North Africa.” (2018). World Food Prog ra m and I nstit ute of Development Studies, Brighton, U.K. Diwan, I., and H. Chekir. 2015. “Crony Capitalism in Egypt.” Journal of Globalization References and Development 5 (2): 177–211. Anderson, J., and E . van Wincoop. 2003. E I U (E conomist I ntelligence Unit). 2016. “Gravity with Gravitas: A Solution to the “Measuring Well-Governed Migration: The Border Puzzle.” American Economic Review 2016 Migration Governance Index.” Study, 93 (1): 170–92. EIU, London. Arvis, J. F., Y. Duval, B. Shepherd, C. Utoktham, ———. 2018. “Democracy Index 2017: Free and A. Raj. 2016. “Trade Costs in the Speech under Attack.” Study, Economist Developing World: 1996–2010.” World Trade Intelligence Unit (EIU), London. Review 15 (3): 451–74. El-Din Haseeb, K. 2012. The Future of the Arab A r vis, J.-F., V. Vesin, R. C . Carruthers, Nation: Challenges and Options. Abingdon, C.  Ducruet, and P. W. De Langen. 2018. U.K.: Routledge. Maritime Networks, Port Efficiency, and Faraji, S. 2016. “Urban Primacy in Urban System Hinterland Connectivity in the Mediterranean. of Developing Countries: Its Causes and Washington, DC: World Bank. C o n s e qu e n c e s .” H u m a n R e s e a rc h i n Behar, A., and C. Freund. 2011. “The Trade Rehabilitation 6 (1): 34–45. Performance of the Middle East and North Freedom House. 2017a. “Freedom of the Press Africa.” Middle East and North Africa 2017: Press Freedom’s Dark Horizon.” Annual Working Paper Series 53, World Bank, report on global media independence, Freedom Washington, DC. House, Washington, DC. Wa l l e d U r b a n Ec o n o m i e s    117 ———. 2017b. “Freedom on the Net 2017: and North Africa. Directions in Development Manipulating Social Media to Undermine Series. Washington, DC: World Bank. Democracy.” Booklet, Freedom House, Lebrand, Mathilde. 2019. “International Growth Washington, DC. Spillovers from Deep Agreements for Countries Frick, Susanne A., and Andrés Rodríguez-Pose. and Cities.” Unpublished manuscript, World 2018. “Big or Small Cities? On City Size and Bank, Washington, DC. Economic Growth.” Growth and Change Loungani, Prakash, Saurabh Mishra, Chris 49 (1): 4–32. Papageorgiou, and Ke Wang. 2017. “World Gammoudi, M., and M. Cherif. 2015. “Capital Trade in Services: Evidence from a New Account Openness, Political Institutions and Databas e .” Wor king Pape r WP17/77, FDI in the MENA Region: An Empirical International Monetary Fund, Washington, DC. Investigation.” Economics Discussion Papers Okada, K. 2013. “The Interaction Effects of 2015-10, Kiel Institute for the World Economy, Financial Openness and Institutions on Kiel, Germany. International Capital Flows.” Journal of Henderson, V. 2002. “Urbanisation in Developing Macroeconomics 35: 131–43. Countries.” World Bank Research Observer Pritchett, L. 2006. Let Their People Come: 17 (1): 89–112. Breaking the Gridlock on International Labor Hoekman, B. 2016. “Intraregional Trade: Mobility. Washington, DC: Center for Global Potential Catalyst for Growth in the Middle Development. East.” Policy Paper 2016-1, Middle East Rijkers, B., C. Freund, and A. Nucifora. 2014. Institute, Washington, DC. “All in the Family: State Capture in Tunisia.” Hoekman, B., and B. Shepherd. 2017. “Services Policy Research Working Paper 6810, World Productivity, Trade Policy and Manufacturing Bank, Washington, DC. Exports.” The World Econom y 40 (3): Rob er t s , M ., a nd U. D eich m a n n. 2011. 499–516. “International Growth Spillovers, Geography Hofmann, C., A. Osnago, and M. Ruta. 2017. and Infrastructure.” The World Economy “Horizontal Depth: A New Database on the 34 (9): 1507–33. Content of Preferential Trade Agreements.” Rodrik, D. 2018. “What Do Trade Agreements Policy Research Working Paper 7981, World Really Do?” Journal of Economic Perspectives Bank, Washington, DC. 32 (2): 73–90. Hoxha, I., S. Kalemli-Ozcan, and D. Vollrath. RSF (Reporters without Borders). 2018a. “RSF 2 013. “How B ig A re t he G a i n s f rom Index 2018: Journalism Sorely Tested in North I nt er n at ion a l Fi n a nc i a l I nt e g rat ion?” Africa.” Regional analysis, World Press Journal of Development Economics 103 (C): Freedom Index 2018, RSF, Paris. 90–98. ———. 2018b. “RSF Index 2018: Middle East Ivanyna, M., and A. Shah. 2012. “How Close is R iven by Conf licts, Political Clashes.” Your Government to Its People? Worldwide Regional analysis, World Press Freedom Index Indicators on Localization and Decentralization.” 2018, RSF, Paris. Policy Research Working Paper 6138, World Solow, Robert M. 1957. “Technical Change and the Bank, Washington, DC. Aggregate Production Function.” The Review of Jaud, M., and C. Freund. 2015. Champions Wanted: Economics and Statistics 39 (3): 312–20. Promoting Exports in the Middle East and North U N DESA (United Nations Department of Africa. Washington, DC: World Bank. Economic and Social Affairs). 2019. UN Kennan, J. 2013. “Open Borders.” Review of Urbanization Prospects: The 2018 Revision. Economic Dynamics 16 (2): L1–L13. New York: UN DESA. Kose, M. Ayhan, Eswar Prasad, Kenneth Rogoff, Weber, Max. 1958. The City. English translation, a nd Sha ng -Ji n Wei. 20 0 6 . “Fi na ncia l edited by Don Martindale and Gertrud Globalization: A Reappraisal.” Working Paper Neuwirth. New York: The Free Press. 06/189, I nternational Monetar y Fu nd, World Bank. 2016. World Development Report Washington, DC. 2016: Digital Dividends. Washington, DC: Krishnan, Nandini, Gabriel Lara Ibarra, Ambar World Bank. Narayan, Sailesh Tiwari, and Tara Vishwanath. World Bank. 2017. Enterprise Surveys, Firm level 2016. Uneven Odds, Unequal Outcomes: TFP estimates and Factor Ratios Data, World Inequality of Opportunity in the Middle East Bank, Washington, DC. How States Shape Markets 4 through Spatial and Private Sector Development Bets J obs and stability: these have been two The chapter addresses five overarching stated drivers of several governments’ pol- questions: icy priorities in the Middle East and North Africa. As a region with some of the 1. How does government intervention in world’s highest unemployment rates and low- markets and competition in the Middle est labor force participation rates, with a East and North Africa compare with youth bulge and rising demand for housing, other regions and why? We compare these and with episodes of violent conflict and ter- countries’ performance on competitiveness rorism, these stated policy orientations are and governance indexes, and we offer a unquestionable. But several of the region’s political economy analysis to help explain governments are making policy choices and the current state of affairs. adopting market orientations that are con- 2. How are Middle East and North Africa tributing to a vicious cycle of worse job and governments intervening to shape mar- inclusion environments. kets at each of the three spatial scales Distinct patterns in many of these govern- described in chapters 1 through 3— ments’ orientations toward private sector-led within cities, within countries, and development and citizen welfare help explain across countries? the current economic geography of these 3. How does the extent of the governments’ countries. The patterns converge on an obser- spatially distortive interventions in the vation—that governments in the Middle East Middle East and North Africa compare and North Africa are more inclined than with those in benchmark countries? We many of their peers to take activist roles in analyze government intervention prefer- physically shaping markets at every spatial ences by analyzing their expenditure pat- scale. And for many, their outcomes appear terns—both across different types of worse than those of their peers, with some interventions and with respect to different exceptions. This chapter analyzes how gov- territories. We compare these patterns ernment orientations toward shaping markets with expenditure patterns in comparator through spatial and private sector develop- countries. ment bets has limited the region’s private 4. What are the implications of these inter- ­ sector and jobs development potential. vention choices? 119 1 2 0    C ONVERGEN C E 5. Finally, given the limited impact of some of many of the region’s countries have reverse the main territorial development policy expenditure patterns: They allocate a larger instruments in the region, why do coun- share of expenditures (investment and recur- tries continue to use them? rent expenditures) to spatially distortive tar- geted interventions such as public sector Recent research has found that most industrialization and industrial location reg- Middle East and North Africa countries ulations; growth poles, industrial districts, have significant barriers to competition. and other location subsidies; subsidies and Relative to most benchmark countries and other incentives to capital; and physical infra- regions, they have less competitive, less effi- structure for connectivity and to support cient environments. This outcome is likely local production. Conversely, they allocate a linked to the development model many of smaller share toward people-based and the region’s governments chose after inde- ­ institution-based interventions. Among most pendence and the evolution of those systems comparator countries across the world, the since then, as further discussed later in the expenditure emphasis is reversed: they allo- chapter. cate most of their expenditures toward The interventionist approach is not limited p e opl e - b a s e d a nd i n s t it ut io n - b a s e d ­ to competition in product and service mar- interventions. kets; it also dictates the spatial development Place-based interventions are associated of the region’s countries at all three spatial with relatively high expenditures, limited scales: results, and physical and economic geogra- •  Within cities, several Middle East and phies that are falling short of stated national North Africa countries have onerous goals. In the Middle East and North Africa, land market and building development the focus on directly shaping and controlling regulations that contribute to patchy markets beginning in the 1950s contributed and sometimes underdemanded formal to an economic geography that left many development and to considerable infor- areas behind. To respond to the needs and mal development except in the Gulf grievances of those regions, countries took Cooperation Council (GCC) subregion.1 on activist territorial development policies •  Within countries, most allocate consider- with a strong spatial focus. These policies, able resources toward building new cities which directed compensatory support to lag- and zones and toward incentives to divert ging territories and created new geographic economic activities to those new places. centers for economic development, saddled With few exceptions, these efforts have the region’s governments with high expendi- had large costs with minimal returns, tures and limited returns. despite the stated intent of reducing con- Why do most Middle East and North gestion and spreading population and Africa governments pursue the same territo- economic activities. rial development interventions despite the •  Across countries, historical barriers to limited impact? First, even if the overall trade have stifled cross-country trade and impact of these interventions is limited, some contributed to the Middle East and North groups benefit considerably from them. Africa’s position as one of the least inte- Second, starting an initiative on new land grated regions in the world. may face less resistance than addressing ongoing challenges in an established area. As for the governments’ emphasis on dif- Third, concrete information is scarce on the ferent instruments for enhancing economic relative successes and failures of the region’s development with spatial implications, the territorial development policies. And fourth, Middle East and North Africa countries are the occasional successes inspire hope for rep- outliers in their policy and intervention orien- lication, even if the recipe for success is not tations. Relative to comparator countries, replicable. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    121 Competition regimes in the underperform relative to their develop- Middle East and North Africa: ment levels (figure 4.1). The region’s coun- How do they differ from other tries have an average score of 4.28 out of 7 regions, and why? (Schwab 2017). 2 Half of the world’s coun- tries (out of 137 included in the GCI) score The region stacks up poorly on world higher, including countries at earlier stages competitiveness and governance of development. 3 The Middle East and indexes North Africa countries’ rankings range Global indexes of countries’ competitive- from last (the Republic of Yemen) to 17th ness, ease of doing business, and governance (the United Arab Emirates). Further, they indicate that the Middle East and North face a challenging duo of the most signifi- Africa as a region offers a less competitive cant barriers to doing business as per- and more constrained market environment c eive d by exe c ut ive s i n t he re g ion : than comparator regions. The latest scores in inefficient government interventions and the World Economic Forum’s (WEF) 2017–18 corruption. Global Competitiveness Index (GCI), the The WEF attributes the region’s weak WEF Executive Opinion Survey on doing performance on the GCI to several factors. business, and the World Bank’s Worldwide First is the extent of the state’s presence in Governance Indicators (WGI) reveal some several sectors, including manufacturing, consistent patterns. construction, finance, transport, and infra- structure. Over 2010–12, the central gov- Global Competitiveness Index ernment wage bill as a percentage of gross According to the 2017–18 GCI, most domestic product (GDP) for Middle East Middle East and North Africa countries and North A frica countries averaged FIGURE 4.1  Middle East and North Africa countries are less competitive than most countries of comparable income 120 100 LUX GDP per capita (US$, thousands) 80 CHE NOR QAT 60 USA SWE AUS 40 GBR ARE BHR KWT 20 OMN CHN LBN IORIRN DZA TUNEGY MAR YEM 0 20 40 60 80 100 120 140 GCI rank Sources: World Development Indicators Database; Schwab 2017. Note: The World Economic Forum’s 2017–18 Global Competitive Index (GCI) ranked 137 countries, including 15 in the Middle East and North Africa (Schwab 2017). GDP per capita uses latest available data from 2016 or 2017. 1 2 2    C ONVERGEN C E approximately 8.1 percent versus about business environment in the World Bank’s 5.9 percent for Organisation for Economic Doing Business Index. In 2017, the region’s Co-operation and Development (OECD) countries ranked above South Asia in how countries and 7.1 percent for Europe and the private sector functions but below Latin Central Asia (WEF and World Bank 2018). America and the Caribbean, Europe and The region’s GCC countries average a 2-to-1 Central Asia, and the OECD countries. For ratio of public to private sector employment, instance, to start a business in the Middle one of the highest in the world. East and North Africa takes 7.8 procedures Second are several investment climate over 20.1 days, and costs 26.3 percent of factors that constrain private investment: income per capita, compared with South political instability, a complex regula- Asia at 8.1 procedures over 15.4 days and tory environment, skills mismatches, and 13.4 percent of income per capita. The barriers to competition in critical factor region’s businesses also pay high and fre- markets. One of the only areas in which quent taxes to operate—17.8 payments a the region’s countries have made signifi- year on average —whereas Europe and cant progress relative to OECD countries Central Asian countries incur 17.6 pay- over the past few decades is infrastruc- ments, and OECD countries, 10.9 pay- ture, but the WEF report notes that this ments. Again, South Asia ranks lower, at has not been associated with gains in 31.8 payments a year on average (World competitiveness. Bank 2017a). Business regulations in the Middle East and North Africa are rated Doing Business Index distant from good practice with respect to The Middle East and North Africa has efficiency and quality, relative to other also underperformed with respect to its regions (figure 4.2). FIGURE 4.2  Business regulations in the Middle East and North Africa are rated distant from good practice with respect to efficiency and quality 80 70 Average distance to frontier score, 2017 60 50 40 30 20 10 0 OECD Europe and East Asia and Middle East and Latin America South Asia Sub-Saharan high income Central Asia Pacific North Africa and Caribbean Africa Regulatory efficiencya Regulatory qualityb Source: World Bank 2017b. Note: OECD = Organisation for Economic Co-operation and Development. The distance to frontier score in the World Bank’s Doing Business Index measures the gap between an economy’s performance and the “frontier,” representing good practice across the entire sample of indicators (excluding labor market regulation indicators) across all economies since 2005 (World Bank 2017b). a. The distance to frontier score for “regulatory efficiency” is the aggregate score for the procedures (where applicable) and time and cost indicators from the following indicator sets: starting a business (also including the minimum capital requirement indicator), dealing with construction permits, getting electricity, registering property, paying taxes (also including the postfiling index), trading across borders, enforcing contracts, and resolving insolvency. b. The distance to frontier score for “regulatory quality” is the aggregate score for getting credit and protecting minority investors as well as the regulatory quality indexes from the indicator sets on dealing with construction permits, getting electricity, registering property, enforcing contracts, and resolving insolvency. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    123 Worldwide Governance Indicators government license, and access to these Concerning five governance indicators licenses has been restricted to insider firms. that affect the local business environment, Firms with licenses to operate in this field the Middle East and North Africa scores benefit from large energy subsidies. This lower across the board than other middle- state of affairs limits both competition in to high-income regions of the world: East energy-intensive industries and the growth Asia and Pacific, Europe and Central of labor-intensive industries. Asia, and Latin A merica and the In Tunisia, there is evidence that industrial Caribbean. The  region’s scores in the policy was oriented toward rent creation for World Bank’s Worldwide Governance former President Ben Ali and his family Indicators (WGI) have been declining con- (Rijkers, Freund, and Nucifora 2014). sistently since 2007, with the exception of Differences in performance between firms brief positive upticks regarding regulatory connected to the former president and com- quality and rule of law in 2012. In the petitors are much larger in sectors with WGI 2017, higher regulation. Further, the sectors in which connected firms are active are dispro- •  O n go ve r n m e n t e f fec t i ve n e s s , t he portionately more regulated than other Middle East and North Africa scores sectors. approximately 10–25 percentage points In both Egypt and Tunisia, there is evi- lower than the other regions (figure 4.3, dence that barriers to entry in trade protect panel a); connected firms from foreign competition •  On regulatory quality, the region scores and orient them toward domestic markets about 10 –15 percentage points lower (Schiffbauer et al. 2015). These barriers figure 4.3, panel b); (­ include privileged access to factors of produc- •  On rule of law, the region scores about tion, inconsistent implementation of rules 10–20 percentage points lower (figure 4.3, and regulations within a given sector, and panel c); exclusive operating licenses in profitable ser- •  On political stability, the region consis- vice sectors. tently scores about half the score of the other regions, and this score declined markedly from 2007 to 2017 (figure 4.3, Competitive constraints stem from the panel d); and region’s political economy •  On control of corruption, the region The Middle East and North Africa’s scores about 10 –20 percentage points underperformance in creating an environ- lower (figure 4.3, panel e). ment that promotes competition and ease Adding to the evidence of governance- of doing business does not seem driven by related effects on competitiveness, several an absence of efforts to reform the legal recent studies have found that industrial and regulatory environment. The region’s policy and regulatory environments in governments have introduced about as Middle East and North Africa countries many market-enhancing reforms as com- are highly correlated with firm connected- parator regions, but the private sector ness (Rijkers, Freund, and Nucifora 2014; development response has been lower—a Schiff bauer et al. 2015). I n the A rab contrast attributed partly to the distinc- Republic of Egypt, for example, a small tion between legal reform and practical number of politically connected firms have implementation (Benhassine et al. 2009). been permit ted to operate in energ y-­ That contrast (illustrated in figure 4.4) intensive industries (Schiffbauer et al. offers insight into a finding that is devel- 2015). Any firm seeking to operate in oped further below—that adding layers e n e r g y- i nt e n s ive i ndu s t r i e s n e e d s a of reforms appears to be insufficient in 1 2 4    C ONVERGEN C E FIGURE 4.3  The Middle East and North Africa has consistently ranked lower than other middle- to high-income regions on the Worldwide Governance Indicators, 2007–17 a. Government effectiveness b. Regulatory quality 2007 2007 Pacific Pacific East Asia East Asia and and 2012 2012 2017 2017 2007 2007 Central Central Europe Europe Asia Asia and and 2012 2012 2017 2017 Africa Caribbean Africa Caribbean 2007 2007 East and America East and America Latin Latin and and 2012 2012 2017 2017 2007 2007 Middle Middle North North 2012 2012 2017 2017 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Percentile rank (0 to 100) Percentile rank (0 to 100) d. Political stability and absence c. Rule of law of violence or terrorism 2007 2007 Pacific Pacific East Asia East Asia and and 2012 2012 2017 2017 2007 2007 Central Central Europe Europe Asia Asia and and 2012 2012 2017 2017 Africa Caribbean Africa Caribbean 2007 2007 East and America East and America Latin Latin and and 2012 2012 2017 2017 2007 2007 Middle Middle North North 2012 2012 2017 2017 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Percentile rank (0 to 100) Percentile rank (0 to 100) e. Control of corruption 2007 Pacific East Asia and 2012 2017 2007 Central Europe Asia and 2012 2017 Africa Caribbean 2007 East and America Latin and 2012 2017 2007 Middle North 2012 2017 0 10 20 30 40 50 60 70 80 90 100 Percentile rank (0 to 100) Source: World Governance Indicators Database, https://info.worldbank.org/governance/wgi/. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    125 the Middle East and North Africa con- FIGURE 4.4  Private investment has responded less to reforms in text, where some fundamental distortions the Middle East and North Africa than in other regions remain unaddressed. Private investment (as % of GDP) in relation to reform episodes 20 Private investment as a History of the postcolonial development share of GDP (%) model 15 The development model most Middle 10 East and North Africa countries pursued in the 20th century after independence from 5 the British and French was both a reaction 0 to, and in some ways a perpetuation of, the t–15 t–10 t–5 Reform t+5 t+10 model pursued by colonial governments. Five-year time intervals Governments established social contracts Middle East and North Africa that were framed as correcting for spatial East Asia and Pacific Latin America and the Caribbean disparities and widespread poverty main- South Asia tained under colonial rule (Nabli et al. Source: Benhassine et al. 2009. 2006). They were designed to depart from Note: Private investment rates are from the World Development Indicators Database, national accounts, and the International Monetary Fund and have been averaged over the five-year periods. the colonial model, in which foreign private Episodes of reforms are based on the Economic Freedom Index of the Fraser Institute firms had lucrative deals to extract natural (http://www.freetheworld.com), and a reform episode is defined as a five-year episode during which the 0–10 index permanently improved by at least one unit. resources. As such, core attributes of the new social contract across the region comprised (a) a The new governments worked mostly shift from markets to states as key shapers of through state-owned enterprises controlling national economies; (b) a shift toward states oil, gas, phosphates, and other natural to determine economic priorities; (c) a focus resources. on redistribution through economic and With revenues assured for the medium social policy (and an assignment to the state term from control of natural resources and a of the provision of welfare and social ser- commitment to redistribution and social wel- vices); and (d) the establishment of the politi- fare, the region’s governments pursued public cal arena as an expression of national unity sector job creation as a centerpiece of their instead of as a platform for political contesta- development models. Natural resource and tion or a means for managing conflicting foreign aid rents limited the governments’ preferences (Nabli et al. 2006). needs for tax collection and therefore for The implementation of this social contract active citizen participation in governance was similar across the region, especially in (Mills and Alhashemi 2018). Further, public Algeria, Egypt, Iraq, and the Syrian Arab sector employment limited the likelihood of Republic, and to a lesser extent in Jordan, citizen contestation against the government. Morocco, Tunisia, and the Republic of Yemen. It included single-party or ruling Legacy of spatially biased policies and party governments, constitutions that development ­ established interventionist and redistributive Although their stated objectives were to principles, agrarian reform programs, change the colonial development model and ­ c entralization of professional associations redistribute to the needy, the legacy of spa- and trade unions, and state provision of tially biased systems inherited from colonial social services and subsidies. rule persisted in most countries of the Middle To execute this new social contract, gov- East and North Africa (Mills and Alhashemi ernments across the region took control of 2018). National administrations and fiscal their natural resources from foreign firms systems were spatially biased toward capital (which had negotiated long concessions dur- cities and coastal metropolises. Colonial rul- ing World War II) (Mills and Alhashemi 2018). ers had focused on resource extraction in less 1 2 6    C ONVERGEN C E connected areas and focused the investments subsidies, trade protection, and other incen- on the coast with little to no representation of tives for production. local interests. That system was maintained Historically, while a stated pillar of gov- or mirrored after independence. ernment development models in the Middle Regions that felt taken advantage of have East and North Africa has been to redistrib- increasingly risen in protest in a form of ute development gains to needier popula- resource regionalism that mirrors resource tions, their choice of vertical instead of nationalism (Mills and Alhashemi 2018). horizontal industrial policies is consistent These lagging regions where resource with pursuit of control. Vertical policies have extraction has been heavy include southern concentrated winners and diffused losers and Algeria, southern and interior Tunisia, east- therefore help secure government power ern and northeastern Syria, and the interior bases. In contrast, horizontal policies that are of Oman. And although some company not industry- or territory-specific have diffuse towns in these regions represent enclaves of benefits, concentrated losses, and often lon- development, even urban centers in resource- ger time frames for impact. So, they can be rich areas often lag behind their national perceived as less politically rewarding. Given peers. Such cities include Basra and Kirkuk the initial orientation to reward a limited few in Iraq, Assaluyeh in the Islamic Republic of at the expense of many, shifts from vertical to Iran, Gafsa in Tunisia, and Khourigba in horizontal instruments have become increas- Morocco. ingly difficult to make because insider eco- Instead of correcting the biases in their fis- nomic actors are resilient to reforms. Reforms cal and administrative structures that rein- often reorganize opportunities for rent seek- forced spatial disparities, the region’s ing rather than eliminating them (Nabli et al. governments sought to overcompensate for 2006). them with spatially targeted industrial poli- cies. The industrial policies were largely ver- tical, focusing both on specific industries or Middle East and North Africa sectors and on regions. Policies cultivated governments intervene in industries in specific sectors, while spatially markets to shape economic targeted industrial policies were oriented geography toward distributing jobs and services to lag- Across several countries in the Middle ging regions. More than in comparator East and North Africa, there is a tension regions, the Middle East and North Africa between governments facilitating and plan- governments pursued vertical (sector and ning for organic growth, akin to historical spatially targeted) industrial policy instead of city development, and adopting what they horizontal industrial policy (Nabli et al. perceive as more modern approaches to 2006). enabling housing development and managing More generally, most of the region’s gov- the location and nature of development. As ernments seemed to perceive industrial poli- outlined in chapter 1, informal development cies, including subsidies, as a necessary part across many countries in the Middle East and of industrial development because of the North Africa mirrors medieval city develop- externalities of their instruments of redistri- ment and features the compact development bution and because of their heavy reliance on that many industrialized countries aspire to natural resource revenues and foreign aid. today. Price controls in the Middle East and North However, in the Middle East and North Africa divorced prices from production costs, Africa, this compact, mixed-use develop- and the region’s governments pursued com- ment had been regarded as problematic for pensatory measures accordingly. Given the much of the past several decades. Planners focus from the 1950s onward on food subsi- across several countries aspired to develop dization, the agriculture sector received large cities with an orientation they perceived as H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    127 more modern—that is, as new cities with few decades relied on independent sectoral large new developments and zoning that authorities affiliated with the ministries of separates uses. In the process, they have in agriculture, irrigation, tourism, housing, many instances created neighborhoods and industry, and defense, which have been given cities that were not desired by citizens and control over large areas outside of the Zimam firms, while a large share of their citizens (the boundary of historically surveyed agri- and firms chose an informal route that bet- cultural lands that are subject to the land ter responded to their needs. This section tax). This has created segmented and isolated outlines the approaches many of the region’s land markets driven by administrative fiat governments have taken to overregulate and supply-side considerations. physical development and expansion at all The sectoral authorities responsible for spatial scales. developing industry, tourism, housing, and new urban communities and for agriculture and land reclamation control more than Within cities, centralized and 5 million feddans of public land (2.1 million fragmented ownership and hectares), equivalent to 2.5 percent of Egypt’s management distort land markets territory and about half of the land area Fragmented and opaque central govern- occupied by Egypt’s 80 million inhabitants ment control of land as well as dysfunctional (World Bank 2012). Moreover, less than land markets have contributed to fragmented 10 percent of urban land and property in and underused supply-driven development. Egypt is registered, owing to cumbersome Well-functioning land markets positively sup- and complex procedures prone to rent seek- port equity and efficiency in several ways. ing plus a backlog of an estimated 19 million For example, they can transfer land from less disputes to be settled. There are already to more productive activities and thus 12  million inhabitants living in informal increase productivity. Also, transferable land areas of Greater Cairo. rights make it possible to get jobs that are not Israel offers one of the most extreme land-intensive, increasing the labor pool for examples by which state control of cities nonfarm activities. Additionally, land rights introduces inefficiencies and reduces pro- incentivize investments by providing cer- ductivity and well-being. Since 1967, Israel tainty that investments will provide benefits has been implementing a broad range of even if investors do not personally use the policies aimed at controlling the movement land. Furthermore, well-functioning land of populations in the West Bank and Gaza, markets undergird equity, providing cer- causing fragmentation of the urban fabric in tainty to those owning the land where dwell- these territories. These policies include road- ings are located (Deininger 2003). blocks, checkpoints, residency requirements, Israeli-only highways, and the separation How state controls have led to fragmented wall. This strategy has severely affected the development quality of daily life for many citizens by However, land markets in Middle East increasing unemployment, poverty, home and North Africa’s cities are mostly incom- demolitions, and business failures and there- plete and inefficient, driven in part by cen- fore causing a major decline in GDP and tralized and fragmented ownership and average income (UN-Habitat 2012). Zoning, management. In Egypt, for example, as chap- planning, and land classification policies are ter 1 described, most of the population is all enforced by various Israeli administrative concentrated in a small portion (about 5 per- departments with little attention to the effi- cent) of the country, and the remaining land ciency of the urban system in the West Bank is mainly desert that is publicly owned and, and Gaza. for the most part, undeveloped. To develop Two important occupation phenomena these vast state lands, Egypt has in the past have had far-reaching consequences: First, 1 2 8    C ONVERGEN C E the creation of the seam zone adjacent to the can more freely respond to market forces. separation wall will separate 50,000 Arab This subsection outlines the way in which residents (living in 38 villages and towns) onerous building and zoning regulations from the rest of the West Bank and separate reinforce this duality. 60,000 people living in East Jerusalem from Several countries restrict floor area ratios their families and places of employment. and thus limit the formal supply of residential Second, the division of the West Bank into and commercial units in their cities. In three areas (A, B, and C) will result in further Tunisia, for example, the default height limit fragmentation. Areas A and B are mainly is three stories in residential areas, and this built-up areas, while area C constitutes standard is often applied to unplanned devel- 60 percent of the West Bank area. However, opments as well. In theory, urban master Israeli authorities have been thwarting devel- plans can assign higher density to registered opment and built-up expansion in those areas plots within planned areas. But that is despite provisions of the Oslo Accords restricted to large plots, and the cumbersome (UN-Habitat 2012).4 application process involves many steps and In Morocco, a cumbersome and opaque stakeholders. land regulatory environment has contrib- In several Middle East and North Africa uted to lower-density cities. Moroccan cit- countries, inefficient and outdated zoning ies face land challenges mainly around the regulations and building codes discourage complexity and cumbersomeness of legal efficient land use and contribute to low-­ provisions, the multiplicity of land tenure, density, high-cost housing developments. and the weakness in the proportion of reg- In Jordan, only 30 percent of households can istered real estate (about 10 percent in rural afford houses larger than 100 square meters municipalities). without spending more that 30 percent of An analysis carried out in Morocco at their monthly income (World Bank 2018b). the level of the Greater Casablanca, Rabat- These high costs are driven by a shortage of Salé-Témara, and Greater Marrakech land and inefficient use of land due to inef- regions for 2010–16 reveals the extent to ficient zoning regulations and building which physical fragmentation is associated codes. with declining urban densities (Lall et al. Where formal housing costs are unafford- 2019). The dropout between urban sprawl able and supply is restricted, a large share of and population growth was considerable: urban populations opts to live in informal the growth rate of urbanized areas was settlements. Between 2004 and 2015, half of more than 5 times that of demography in Amman’s new housing units were permitted Marrakech (12.25 percent and 2.21 per- before construction, more than 14 percent cent, respectively); more than 3 times for were registered after completion, and 37 per- the whole Rabat-Salé-Témara region (5.85 cent are still unpermitted (World Bank percent and 1.96 percent); and more than 2018b). In most of Jordan’s governorates, 1.7 times for Greater Casablanca (2.78 per- fewer than 30 percent of the new housing cent and 1.64 percent). units have a building permit (30 percent in Maan, 26 percent in Kerak, 24 percent in How regulatory schemes have increased Irbid, 23 percent in Madaba, 21 percent informality in Tafila, 19 percent in Zarqa, 11 percent in Onerous building and zoning regulations Mafraq, 10 percent in Jerash, and 8 percent are creating a mismatch bet ween the in Ajlun). In those governorates, Jordanians ­ location of jobs and formal housing, and in have been building unlicensed housing units t u r n , dem a nd is b ei ng me t t h roug h without. In Aqaba, the housing permit regu- ­ i nformality. Chapter 1 highlighted how lation is better followed, with 61 percent of holes in the formal urban fabric are being the new housing units permitted before con- “stitched” by informal sector activity that struction (box 4.1). H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    129 BOX 4.1  The process for building a housing unit in Jordan In Doing Business 2018 , Jordan ranked 72nd of clearances may be required by other referral agencies 190 countries in registering property and 110th in (Public Security Department) and GAM units (Traffic issuing construction permits (World Bank 2018a). Control Department or Special Projects Department) On average, national-level building permits require depending on project type and complexity. Once the 62  days to be issued. (In Greater Amman, it can structure is complete and third-party inspectors from take up to four months, because the auditing bureau GAM and CDD have cleared the project, a comple- has only seven representatives for the entire city.) tion or occupancy permit, issued by the municipality Typically, housing development follows these steps: to occupy the building, is necessary to get utility con- In areas zoned for urban use, land is subdivided and nections (electricity and water). sold. The owner (developer or individual) then con- The owner bears the cost of bringing the trunk sults with the municipality on the land-use options infrastructure from the nearest available network to and may have to pay “improvement taxes” if the use the site, which can be very high (JD 300 per elec- has been changed from residential to commercial. tricity pole at 30-meter intervals). The land owner Land-use changes as well as amendments on the must pay a building tax of JD 0.75 per square meter building terms for each land plot are frequent, and to the municipality and build an underground safe as a result, consultations with the authorities before space or “refuge area” not smaller than 10 percent any development are necessary. Subsequently, the of the surface. The new municipal law adopted in owner seeks approvals from several referral agencies 2015 includes a JD 150 fine per square meter built such as the Jordanian Engineers’ Association, the in contravention in order to push citizens and build- Civil Defense Department (CDD), and Ministry of ers to follow zoning and building norms, including Tourism and Antiquities as well as from utility pro- parking lots. viders before applying for a building permit at the Greater Amman Municipality (GAM). Additional Source: World Bank 2018b. Within countries, interventions to move Despite good intentions, the evidence that jobs and people have had limited success investment incentives have helped generate New cities, zones, and spatially distortive substantial jobs in targeted areas in the industrial regulation are some of the mecha- region’s countries is limited. For example, nisms governments in the Middle East and Syria tried to influence the location of private North Africa have used to shape economic sector investment by subsidizing specific sec- geography. The motivations for these inter- tors in some cities. During the early 2000s, it ventions vary by country but comprise sev- focused on supporting the textile and cloth- eral major drivers. The first is the desire to ing sectors, incentivizing the clustering of rel- move jobs to places where they are lacking evant manufacturing firms. These policies and people are in need. To do so, countries in were implemented without a private sector the Maghreb and Mashreq subregions have demand assessment, and there is no evidence deployed investment incentives and a range of that these policies increased efficiency or ben- zones facilitating the relocation and new efited Syria’s economy (Chahoud 2011). development of economic activity. The sec- The recent history of new city develop- ond is a desire to leverage unused land. In ment in several Middle East and North some cases, the focus is to create a new Africa countries has shown the high cost and growth engine (such as Neom in Saudi often limited returns to noncontiguous city Arabia), and in others, it is to correct for development. An estimate of the direct costs major challenges in existing cities (such as of developing these cities in Saudi Arabia and Egypt’s new capital city). Egypt (outlined below) reveals the enormous 1 3 0    C ONVERGEN C E outlays by the region’s governments to build But because of high housing costs and new desert cities. The low occupancy of most poor housing policies, few workers settled in of these cities relative to projections reveals the new towns, with vacancies in newer pub- that they are not delivering the returns lic housing units exceeding 50 percent (Sims expected. 2015). The result is huge fleets of buses shut- Although zones—industrial, economic, tling workers daily from the Cairo agglom- and free—could be a useful way to offer the eration and other urban centers to the new suite of services and infrastructure needed for towns. Employment growth has also been certain types of economic activity to develop, low. The largest five industrial zones in new the limited success of these zones across the towns together employ fewer than 300,000 region demonstrates that location always workers, and the contribution of the remain- matters. For example, Egypt has used indus- ing new towns has been minimal. Overall, trial zones and new towns to divert firms and government-sanctioned industrial estates in people from major cities and toward less- Egypt’s new towns and governates by 2006 developed areas. As a result, however, access had space for 2.5 million workers but hosted to markets has been a major impediment to just 483,000 jobs. By comparison, in 2009, their success and utilization. Of 139 indus- there were a total of 1.8 million workers in trial zones created from 1975 to 2012, almost registered factories5 outside formal industrial all were in desert locations disconnected zones, mainly within current urban agglom- from agglomerations (Sims 2015). They are erations (World Bank 2012). far from input sources and services, and mar- kets find it difficult to mobilize and retain a Across countries, resistance to regional large enough labor force. Manufacturing cooperation constrains trade firms in industrial estates commonly provide bus transport for workers who must travel The Middle East and North Africa’s poor long distances, adding to recurrent costs. In record on regional cooperation is somewhat some governorates where concerted efforts surprising. There are no good structural rea- were made, occupancy rates of these zones sons for it. The region has a history of inten- are less than 25 percent of developed sive trading relations going back centuries. capacity. Like Latin America, it is bound by a common In some cases where governments built language and similar culture. 6 Its endow- zones in distant locations and observed lim- ments should make it far more economically ited occupancy, they attempted to create vibrant. Like East Asia, the region has abun- new towns around the zones for people to dant and increasingly well-educated labor. live in and agglomerations to develop. Egypt Like Eastern Europe and Central Asia, it is created planned new towns in three waves. rich in resources. With the European Union Eight cities in the late 1970s and early 1980s (EU) to the north and the GCC within its were planned as major industrial towns with region, it has nearby sources of capital. The industrial zones (Sims 2015). They were dis- region also has a favorable geography. It is connected from existing agglomerations but close to the world’s largest economy—the within reasonable freight distances to major EU’s single market—and located on major ports and Cairo’s agglomeration in Egypt’s trade routes to the rising markets of East Asia north. Nine more new towns were con- and to the continent with the fastest growing structed in a similar area between 1982 and population, Africa. There are few major 1995, five of which were in Greater Cairo. physical barriers dividing the region, and no A final wave of new towns was established Middle East and North Africa country is since 2000, concentrated in Upper (south- landlocked. ern) Egypt, where access to international Clearly, the region is fragmented because markets and local agglomerations is much of history and policy, not geography (Malik weaker. and Awadallah 2013). Several factors have H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    131 put the brakes on regional integration: con- the onset of the Arab Spring and the conflicts flict, political economy (vested interests), and in Syria and Iraq, these regional integration the recent broader backlash against trade and efforts came to a halt. Syria and Iraq directly globalization. lost 14–16 percent of per capita welfare The trade barriers put in place over several because of war, with neighboring countries decades by the region’s governments con- losing 1–1.5 percent. If the forgone benefits strain trade and are difficult to remove. of regional trade integration are included, Significant tariff and nontariff trade barriers the costs for Syria and Iraq almost double mean that there is low complementarity to 23 percent and 28 percent, respectively, between these economies: many countries and rise to 10 percent for Egypt and 9 per- produce the same things their neighbors pro- cent for Jordan. Such estimates are obviously duce. Removing these protections will affect tentative, but conflict undoubtedly imposes a country’s less-competitive sectors. A sudden high costs on Middle East and North Africa opening of markets could cause severe adjust- economies. It is difficult to see how deeper ment shocks, including employment losses. In regional integration can be achieved without principle, the gains reaped by more competi- lasting peace and security. tive sectors should enable compensation for the losers and the creation of new job oppor- tunities. But in practice, adjustment is rarely Government interventions cause smooth. In many industrialized countries, varying magnitudes of spatial those that lost from globalization did not distortion in the Middle East and have the right skills to switch to industries North Africa that gained.7 The industries that lost and Governments in the Middle East and gained also often varied in different parts of North Africa and beyond have a spectrum of the country, making job switching more dif- intervention options they can leverage to ficult. These concerns do not void the case for address constraints to development across greater integration, but they need to be taken space. These interventions may introduce seriously by policy makers, especially in spatial distortions (figure 4.5). From the most countries with weak institutions and under- to the least spatially distortive, government developed social safety nets. interventions can (a) influence the exact loca- Conflict—both violent and political, driven tion of economic activities (place-based inter- by a range of frictions—constrains cross-bor- ventions); (b) influence access to opportunities der trade. Territorial disputes reduce beneficial for people (people-based interventions); or (c) trade between neighbors. Conflict leading to reduce distortions that constrain markets military confrontation raises the cost. For through broad-based national interventions instance, development economist Paul Collier (institution-based interventions). estimated in 2003 that a typical civil war This section analyzes government expen- imposes costs of about US$64 billion and that diture distribution in select Middle East and civil wars’ annual average cost far exceeds North Africa countries and in comparable total development aid (Collier 2003). countries with a focus on understanding the Before 2011, Turkey and five Levant extent to which they intervene in spatially countries (Egypt, Iraq, Jordan, Lebanon, distortive manners. Characteristics such as and Syria) pursued the establishment of the GDP per capita, population size, and level of Levant Economic Zone. 8 The agreement development were used to define the set of would have considerably deepened trade benchmark countries. Moreover, regional integration between the countries not just balance was sought in the choice of compara- by reducing tariffs but also by liberalizing tors, so they include countries from the East transport and services trade and by impos- Asia and Pacific, Sub-Saharan Africa, ing reforms that would have made the par- and  Europe and Central Asia regions. ticipating economies more competitive. With Although the analysis looks at snapshots in 1 3 2    C ONVERGEN C E FIGURE 4.5  Government interventions can create varying degrees of spatial distortion More Spatial Public sector industrialization and Distortion industrial location regulations The Policy Challenge: Growth poles, industrial districts, and • Do these instruments improve local other location subsidies growth and welfare in a particular region? • Do these instruments improve national growth and welfare? Subsidies and other incentives to capital • Are these interventions feasible given fiscal contraints? Physical infrastructure for • Are these instruments consistent with connectivity and to support local political objectives? production Skill development, worker training, and wage subsidies Provision of basic public services (water supply and sanitation) Basic health, education, and related human capital improvements Broad-based governance and institutional reforms Less Spatial Distortion Place-based interventions People-based interventions Institution-based interventions Source: Lall 2009. ©World Bank. Further permission required for reuse. time and relies on different data sources with government units of the central, regional, and some variation in coverage, the findings offer local levels of government,11 excluding public a preliminary indication of broader trends corporations. Expenditure data include bud- and should be interpreted as such. getary and extrabudgetary expenses only for nonmarket activities (box 4.2).12 Government functions are classified in 10 Most benchmark countries allocate the broad categories, which are then disaggre- largest shares of government gated into 69 subcategories. (Data availability expenditure to developing human in this database also influenced the choice of capital—and low shares to place-based comparator countries.) The methodology used interventions to obtain a spatial perspective of the govern- Data for benchmark countries were ment expenditure distribution was to assign obtained from the International Monetary each of the 69 subcategories into one of the Fund’s (IMF) Classification of Functions of eight spatial dimensions shown in figure 4.6,13 Government (COFOG) in the Government where the spectrum ranges from the least spa- Finance Statistics (GFS) system.9 General gov- tially distortive (broad-based governance and ernment units around the world can opt in to institutional reforms) to the most spatially dis- follow this classification and publish their tortive (public sector industrialization and expenses in the GFS system.10 In the GFS industrial location regulations). For example, database, COFOG adds both current and expenditure on law courts, one of the 69 sub- ­ capital expenditures, executed on each of the categories, is classified within broad-based f u nc tional categories by the general governance and institutional reforms.14 H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    133 BOX 4.2  The IMF Government Finance Statistics database Data in the IMF’s General Finance Statistics (GFS) a government unit, a public corporation, or a public system comprise general government expenditures, quasi corporation. In general, the decision is based excluding those related to public corporations. on whether the unit sells its output at market prices. The GFS system covers all entities that affect fiscal Any unit that sells all or almost all of its output at policies, which are usually implemented by entities market prices is a corporation or quasi corporation, focused on the economic functions of government, and all other units are government units. With public such as a government ministry, and not on market units, however, market prices are not always easy to activities (­fi gure B4.2.1). In addition to those enti- identify. ties, however, fiscal policy may be undertaken by The GFS system classifies expenditures of the government-owned or controlled enterprises that general government by function in 10 broader cat- engage primarily in commercial activities. These egories, disaggregating them further into 69 sub- enterprises, such as the central bank or national categories according to the OECD classification railroad, which are referred to as public corpora- published by the United Nations Statistical Division. tions, are not considered part of the general govern- Table B4.2.1 was extracted from the Government ment in the GFS system. Finance Statistics Manual 2014, which describes in When a unit sells some or all of its output, it can be more detail each of the four-digit categories listed difficult to decide whether to classify the unit as either below. FIGURE B4.2.1  The public sector and its main components Budgetary Central Extrabudgetary Governmenta Social Security Funds State Governmentsa Subsectorsc General Government Local Governmentsa Subsectorsc Social Security Fundsb Public Sector Public Nonfinancial Central Bank Corporations Public Deposit- Public Taking Corporations Corporations Public Deposit-Taking Public Financial Corporations except Corporations Other Public the Central Bank Financial Corporations Source: IMF 2014, 20. a. Includes social security funds. b. Alternatively, social security funds can be combined into a separate subsector, as shown in the box with dashed lines. c. Budgetary units, extrabudgetary units, and social security funds may also exist in state and local governments. box continues next page 1 3 4    C ONVERGEN C E BOX 4.2  The IMF Government Finance Statistics database (continued) TABLE B4.2.1  Classification of expenditure, by government function, within divisions and groups 7 Total expenditure 701 General public services 706 Housing and community amenities 7011 Executive and legislative organs, financial and fiscal affairs, external affairs 7061 Housing development 7012 Foreign economic aid 7062 Community development 7013 General services 7063 Water supply 7014 Basic research 7064 Street lighting 7015 R&D General public services 7065 R&D Housing and community amenities 7016 General public services n.e.c 7066 Housing and community amenities n.e.c. 7017 Public debt transactions 707 Health 7018 Transfers of a general character between different levels of government 7071 Medical products, appliances, and equipment 702 Defense 7072 Outpatient services 7021 Military defense 7073 Hospital services 7022 Civil defense 7074 Public health services 7023 Foreign military aid 7075 R&D Health 7024 R&D Defense 7076 Health n.e.c. 7025 Defense n.e.c. 708 Recreation, culture, and religion 703 Public order and safety 7081 Recreational and sporting services 7031 Police services 7082 Cultural services 7032 Fire protection services 7083 Broadcasting and publishing services 7033 Law courts 7084 Religious and other community services 7034 Prisons 7085 R&D Recreation, culture, and religion 7035 R&D Public order and safety 7086 Recreation, culture, and religion n.e.c. 7036 Public order and safety n.e.c. 709 Education 704 Economic affairs 7091 Pre-primary and primary education 7041 General economic, commercial, and labor affairs 7092 Secondary education 7042 Agriculture, forestry, fishing, and hunting 7093 Postsecondary nontertiary education 7043 Fuel and energy 7094 Tertiary education 7044 Mining, manufacturing, and construction 7095 Education not definable by level 7045 Transport 7096 Subsidiary services to education 7046 Communication 7097 R&D Education 7047 Other industries 7098 Education n.e.c. 7048 R&D Economic affairs 710 Social protection 7049 Economic affairs n.e.c. 7101 Sickness and disability 705 Environmental protection 7102 Old age 7051 Waste management 7103 Survivors 7052 Waste water management 7104 Family and children 7053 Pollution abatement 7105 Unemployment 7054 Protection of biodiversity and landscape 7106 Housing 7055 R&D Environmental protection 7107 Social exclusion n.e.c. 7056 Environmental protection n.e.c. 7108 R&D Social protection 7109 Social protection n.e.c. Source: IMF 2014, 143. Note: R&D = research and development. n.e.c. = not otherwise classified. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    135 FIGURE 4.6  Distribution and changes in government expenditures of comparator countries reflect priorities through a spatial lens 100 Share of expenditure (%) 80 60 40 20 0 1995 2016 1995 2016 2012 2016 1995 2015 2002 2015 Austria Spain South Africa Hungary Poland Institution-based interventions People-based interventions 1. Broad-based governance and 2. Basic health, education, and related human capital improvements 3. Provision of basic public services institutional reforms 4. Skill development, worker training, wage subsidies Place-based interventions 5. Physical infrastructure for connectivity and to support local production 6. Subsidies and other incentives to capital 7. Growth poles, industrial districts, other location subsidies 8. Public sector industrialization and industrial location regulations Source: Government Finance Statistics (GFS) database, International Monetary Fund. Note: The period considered for each country uses the oldest and most recent years available in the GFS database. For specific percentages by country, category, and year, see table 4A.1 and table 4A.2. Most of the governments, except for South interventions to which comparators assigned Africa, allocate the largest share of public the highest allocations. expenditure to developing human capital Over time, the share allocated toward figure 4.6). They are investing more than 60 (­ ­ people-based and institution-based interven- percent on people-based policies (such as skill tions remained stable or increased (figure 4.6; development and worker training, basic pub- also see annex 2A, table 2A.3), as follows:15 lic services, and health and education), with Spain investing the most in 2016: 75 percent •  In Spain, the government reduced by more of public expenditure. than half its investment on infrastructure South Africa invested a higher percentage between 1995 and 2016 while increasing in institutions (broad-based governance and its human capital-related expenditures. institutional reforms) than the other coun- This may be the case because right after tries, and it also directed a high share of pub- joining the EU in 1986, Spain invested lic expenditures toward people-based heavily in connective infrastructure. policies. South Africa allocated 52 percent of •  In Austria, on the other hand, the govern- its budget to categories that relate to improv- ment reduced by 2 percent its expenditure ing regulations. This is expected in low- to on institutions between 1995 and 2016 middle-income countries where there is still while increasing by 3 percent expenses on room for structural reforms. people-based interventions. Benchmark countries spent less than 15 •  In South Africa, the government reduced percent of their budgets on place-based inter- its expenditures on place-based interven- ventions, with less than 2 percent on the most tions and people-based interventions to spatially distortive ones. Hungary has the allocate 8 percent more of its budget on highest percentage among comparators with institutions between 2012 and 2016. a total of 12 percent, followed by Poland South Africa has been investing during with 10 percent. Excluding South Africa, recent years in enabling an environment infrastructure is the category of place-based for trade and businesses. 1 3 6    C ONVERGEN C E •  In Hungary, between 1995 and 2015, the different from comparator country alloca- government decided to invest less in insti- tions.16 One country from each subregion in tutions, while increasing by 2 percent its the Middle East and North Africa was cho- expenditure on people-based and 3 percent sen for a deeper analysis (box 4.3): Morocco on place-based interventions. This could (from the Maghreb subregion), Jordan (from be the case because, during the 1990s, the Mashreq subregion), and Saudi Arabia Hungary needed to invest in institutions as (from the GCC subregion), including addi- part of its post-Soviet Union transition. tional countries where data were more read- •  In Poland, expenditures represented the ily available,17 such as no Egypt and Tunisia. only exception in our benchmark cases. (For details on why other countries in the The government increased its expenditure region were excluded from a deeper analysis, on place-based interventions from 2002 see annex 4D.) to 2015 by 4 percent, presumably a result Results of the expenditure distribution of its increased focus on regional develop- analysis18 are as follows (as shown in ment policy. figure 4.7, figure 4.8, and table 4.1; also see ­ annex 2A, table 2A.3): •  Spatially distortive expenditures in Government expenditure allocations in the Middle East and North Africa differ Morocco represent about 77 percent of greatly from those in comparator government’s expenditure in 2017, and in countries Tunisia, about 50 percent, while in bench- mark countries, the percentage allocation Based on our indicative analysis, govern- of public expenditures to spatially distor- ment expenditure allocations in the Middle tive interventions was consistently below East and North Africa appear strikingly 15 percent. BOX 4.3  Government expenditure data for a subset of Middle East and North Africa countries Unlike the analysis for comparator countries, no The state budget is divided into five categories, where comparable and comprehensive data source exists autonomously managed state services and special that classified all government expenditures in a funds are special entities that have some flexibility in given time period for Middle East and North Africa their budgetary allocations. countries (except for partial data for the United In 2017, 69 percent of the state budget was allo- Arab Emirates in the IMF GFS database). Instead, cated through the general budget, which the team c ountry-specific sources for this information were ­ decided to zoom in on for the spatial analysis using analyzed and details on the scope of each are out- the Law of Finance of 2017, which ministries are lined below and in footnotes to the figures included obliged to follow but which may differ from actual in this subsection. As such, in this section, the analy- expenditure.a Moreover, data on expenditures of pub- sis offers a snapshot at one point in time. It may vary lic establishments and enterprises are also included in from the medium term distribution pattern for these the spatial analysis to give a more robust view of total countries but offers an illustration of how funds government expenditures in Morocco. were distributed in each case country. Tunisia Morocco The Government of Tunisia also votes on and approves The national budget in Morocco is split between the annual expenditures through annual finance laws. state budget and public enterprises and establishments. Tunisian finance laws use the nation’s five-year box continues next page H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    137 BOX 4.3  Government expenditure data for a subset of Middle East and North Africa countries (continued) development plans as reference for their yearly expen- The only data we do not include in the analysis ditures. Expenditures listed in the finance laws include are those of the public business sector, within which expenditures of state-owned enterprises and public there are 18 holding companies under Public Business agencies and authorities. This is a binding document for Sector Companies Law No. 203 (of which 8 report to all public expenses. However, data in the Finance Law the Ministry of Public Business Sector and the other of 2017 were in some cases ambiguously categorized or 10 to various line ministries); 2 companies under lacked the detail needed for the primary analysis. As the General Companies Law (Telecom Egypt and such, we did not rely on data from the Finance Law for Electricity Holding Company); and 4 corporations the primary analysis, but we did analyze it as part of under Public Sector Law 97 (3 state-owned banks a robustness check. For this robustness check, we also in addition to the Arab Contractors Company). We integrated World Bank’s Open Budgets Portal (BOOST). did not include them because financial data for the Instead, the primary analysis is based on Tunisia’s 24 companies were not readily available. However, overarching five-year development strategy. The most based on data we obtained on the capital expendi- recent five-year plan in Tunisia, the Development Plan tures of a subset of 13 of these companies, we found 2016–20, is used for the data allocation in our primary that their total represented less than 1 percent of total analysis in this chapter. This plan listed 262 projects expenditures (on- and off-budget). (divided into 16 sectors) that Tunisia plans to focus on during that period (and which are used in the spa- Jordan tial classification described in the main chapter text) Data used for Jordan are the “Summary of Functional at the national and at the governorate levels.b The plan Classifications for Estimated Public Expenditures includes expenditures of state-owned enterprises and According to Functional Divisions and Groups for the public agencies and authorities. The Tunisia analysis in Fiscal Year 2018” in the Law No. (1) for the Year 2018, this chapter drawing from the five-year plan is the only of the General Budget Law for the Fiscal Year 2018. one that spans more than one year. Saudi Arabia Egypt For Saudi Arabia and all other GCC countries, granu- The analysis for Egypt draws from multiple sources. lar data are not publicly available. Instead, we synthe- National public expenditures are executed by the gen- sized the relevant orientations in Saudi Arabia’s Ninth eral government, by economic authorities, and by the Development Plan 2010–2014 and analyzed budget- public business sector. However, there is no document ary allocations to new cities from publicly available that includes the three different sectors or entities. documentation. General government expenses, also called on-budget a. Morocco’s recent Finance Law does not provide budget data for some expenses, can be found in the financial statements categories (such as health services and most religious activities in the “General published by the country. The financial statement for Budget” segment) and for some spatially distortive interventions (such as funding of roads within “Public Establishments and Enterprise”). These exclusions could 2018–19 was used for the analysis in this chapter.c result in overestimation of the allocations for other sections and underestimations Economic authorities and public business sector are of the allocations for “Basic Health, Education, and Related Human Capital the entities performing the off-budget expenses. The off- Improvements” and “Physical Infrastructure for Connectivity.” b. The 20 sectors are agriculture, fishing, and water resources; industries; budget expenditures in Egypt are more than twice the information and communication technology (ICT); transportation; roads and on-budget expenditures. Our analysis incorporates the bridges; housing and urban development; environmental protection; education; expenditures of the 48 economic authorities that execute higher education and scientific research; job readiness; employment; culture; health; women, family, and the elderly; finance; public employment, governance, off-budget expenditures. They represent 14 sectors: agri- and corruption; communal investment; rehabilitation and housing integration culture and irrigation; industrial; mineral wealth and programs; regional development programs; and programs for inclusive development. petroleum; electricity and power; transport; commu- c. On-budget data for Egypt are available only at the aggregated level among nications and aviation; trade and supply; financial and nine functional categories: public services; public safety; economic affairs; economic; housing and construction; health services; environmental protection; housing and social infrastructure; health; youth, culture, and religion; education; and social protection. This limits our ability to religious and public forces; culture and media; tourism; complete the same analysis for Egypt as we have done for the focus countries of and social security and social affairs. this analysis. 1 3 8    C ONVERGEN C E FIGURE 4.7  Government expenditure distribution in selected Middle East and North Africa countries show priorities through a spatial lens for one year 100 80 Share of expenditure (%) 60 40 20 0 Moroccoa Egypt, Arab Rep.b Jordanc Tunisiad Institution-based interventions People-based interventions 1. Broad-based governance and 2. Basic health, education, and related human capital improvements 3. Provision of basic public services institutional reforms 4. Skill development, worker training, wage subsidies Place-based interventions 5. Physical infrastructure for connectivity and to support local production 6. Subsidies and other incentives to capital 7. Growth poles, industrial districts, other location subsidies 8. Public sector industrialization and industrial location regulations Sources: Government data (as described in notes below). Note: For specific percentages by country, category, and year, see table 4A.3. a. For Morocco, the information source is the 2017 Finance Law, specifically the “General Budget” and the “Public Establishments and Enterprises” segments. b. For Egypt, the Financial Statement 2018–19 was used for on-budget data, and the total expenses of the 48 economic authorities were retrieved from their 2016–17 financial statements from the Ministry of Finance website. c. For Jordan, the information source is “Capital and Current Expenses” classified by all line ministries, reported by the General Budget Department of the Ministry of Finance. d. The Tunisia analysis draws from the most recent five-year plan, the Development Plan 2016–20. FIGURE 4.8  Government expenditure distribution in selected Middle East and North Africa countries, by spatial category, differs greatly from international comparators 100 Share of expenditure (%) 80 60 40 20 0 Egypt, Arab Moroccoa Jordanc Tunisiad Austria Spain South Africa Hungary Poland Rep.b Institution-based interventions People-based interventions 1. Broad-based governance and 2. Basic health, education, and related human capital improvements 3. Provision of basic public services institutional reforms 4. Skill development, worker training, wage subsidies Place-based interventions 5. Physical infrastructure for connectivity and to support local production 6. Subsidies and other incentives to capital 7. Growth poles, industrial districts, other location subsidies 8. Public sector industrialization and industrial location regulations Source: Government Finance Statistics (GFS) database, International Monetary Fund; government data for Middle East and North Africa countries (as described in notes below). a. For Morocco, the information source is the 2017 Finance Law, specifically the “General Budget” and the “Public Establishments and Enterprises” segments. b. For Egypt, the Financial Statement 2018–19 was used for on-budget data, and the total expenses of the 48 economic authorities were retrieved from their 2016–17 financial statements from the Ministry of Finance website. c. For Jordan, the information source is “Capital and Current Expenses (2018),” classified by all line ministries, reported by the General Budget Department of the Ministry of Finance. d. The Tunisia analysis draws from the most recent five-year plan, the Development Plan 2016–20. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    139 TABLE 4.1  On average, comparator countries invest more in people- and institution-based interventions, and less in place-based interventions, than do Middle East and North Africa countries Average percentage of government expenditure Intervention category Comparator countriesa Middle East and North Africa countriesb Institution-based interventions 27 8 People-based interventions 65 42 Place-based interventions 9 50 Source: Government Finance Statistics (GFS) database, International Monetary Fund; government data for Middle East and North Africa countries (as described below). a. Comparator countries are Austria, Hungary, Poland, and South Africa, b. Middle East and North Africa countries are the Arab Republic of Egypt, Jordan, Morocco, and Tunisia. For Egypt, the Financial Statement 2018–19 was used for on-budget data, and the total expenses of the 48 economic authorities were retrieved from their 2016–17 financial statements from the Ministry of Finance website. For Jordan, the information source is “Capital and Current Expenses (2018),” classified by all line ministries, reported by the General Budget Department of the Ministry of Finance. For Morocco, the information source is the 2017 Finance Law, specifically the “General Budget” and the “Public Establishments and Enterprises” segments. The Tunisia analysis draws from the most recent five-year plan, the Development Plan 2016–20. •  Expenditures on institutions represent Strategy reveals the weight the government less than 1 percent of public expendi- places on balanced territorial ­ development. tures in Morocco and only 15 percent in The Ninth Development Plan 2010 –2014 Tunisia in the year studied, while com- outlines the nation’s development orienta- parator countries’ share of expenditures tions across the range of sectoral and range from 16  percent (in Poland) to ­ s ubnational dimensions. It explains that 52 percent (in South Africa). the nation has historically striven for bal- •  People-based policies represent 22 per- anced territorial development with a focus cent of public expenditure allocation on i n f ra s t r u c t u re prov i sion , t a k i n g in Morocco and 35 percent in Tunisia. reg ional potential into consideration While Tunisia seems to allocate more (MEP 2010). resources toward skill development, The regional development focus has basic public services, and basic health aimed to reduce internal migration to large and education, this allocation reflects a cities and to redirect economic activities to sharp difference from benchmark coun- areas that appear to have insufficient tries. In the European Union compara- demand for economic activities. The plan tors, this allocation is around 74 percent elaborates broad priorities for achieving bet- (Austria and Poland) and 75 percent ter territorial balance and references lever- (Spain). aging investment grants oriented toward •  Expenditures on basic health and edu- lagging regions as established through the cation showed the greatest gap with the Council of Ministers’ Resolution No. 359 of benchmark countries within the people- 2008. The National Spatial Strategy is based expenditure allocations. anchored around development corridors and growth centers. A ke y pi l l a r of S aud i t er r itor i a l Country analyses development policy is the development of ­ Saudi Arabia new cities. Our analysis reveals the estimated Although detailed budget data are not expenditure to deliver eight new cities publicly available for Saudi Arabia, an (table 4.2). These large outlays follow the analysis of the nation’s Ninth Development establishment of predecessor new desert Plan 2010 –2014 and its National Spatial cities. 140 TABLE 4.2  New cities in Saudi Arabia’s development plan and spatial strategy Anticipated Total Start completion estimated cost Surface Current Type of City date date (US$, millions) area (km2) Region Location Target population population Industries development King Abdullah 2005 2020 88,000 168  Makkah 90 km north of 2 million 10,000 Heavy industries, residential, Leapfrog Economic City Jeddah recreational Jazan 2009 2030 13,200 103 Jizan 100 km north of 300,000 0 Energy, residential; light Leapfrog Economic City Jazan industry; air and land transportation hub Al Faisaliah City 2017 2050 246,400 2,450 Makkah From the Haram 5.6 million 0 Residential, industrial, Leapfrog Boundary to the residents and 10 diplomatic quarters, resorts, Red Sea Coast million hajj visitors yacht club, hajj accommodations Neom 2017 2025 500,000 26,500 Tabuk West of Tabuk — 0 Biotech, food, advanced Leapfrog with 468 km of manufacturing, media, Red Sea education, tourism, sports, coastline energy, and water Red Sea Project 2018 2022 6,160 — Madinah 50 Red Sea — 0 Luxury tourism Leapfrog islands; 100 km north of Yanbu Qiddiya 2018 2022 (Phase 1) 6,160 334 Riyadh 40 km west of 1.5 million visitors — Recreation, entertainment, Leapfrog Entertainment 2035 (Phase 2) Riyadh (Phase 1); 17 theme parks City million (Phase 2) Prince 2006 2025 6,160 — Ha’il 8 km North of 140,000 0 Agriculture, residential, Leapfrog Abdulaziz Bin Ha’il  industrial, transportation Mousaed services, logistics, educational Economic City services, agricultural services, (PABMEC) industrial and mining services, recreational services Al-Madinah’s 2009 2024  6,600 4.8 Madinah 5 km east of 150,000 0 Religious tourism, shopping, Contiguous Knowledge Madinah medical facilities, residential Economic City Sources: World Bank estimates from the following data sources: Economic Cities Authority (2018) database on economic cities; Saudi Projects Network (2018) project database; King Abdullah Economic City official website (2018) Annual Report; Qiddiya official website (2018) press release on the establishment of Qiddiya Investment Company; Clyde and Co. (2016) Tax Incentives in the Kingdom of Saudi Arabia; Sports Venue Business (2018) interview with Almamoun Alshingiti, executive director of development, Qiddiya Investment Company; Gulf Business (2018), “Saudi Hires Ex-Canary Wharf Exec for Red Sea Tourism Project.” May 28; Sabq Online (2017), “Change in the Leadership of PAMBEC Creates New Opportunities.” Note: km = kilometers. km2 = square kilometers. — = unavailable. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    141 TABLE 4.3  Spatial classification of Tunisia’s Finance Law 2017 Policy option Share of total expenditure (%) 1. Broad-based governance and institutional reforms 5 2. Basic health, education, and related human capital improvements 32 3. Provision of basic public services 5 4. Skill development, worker training, wage subsidies 5 5. Physical infrastructure for connectivity and to support local production 25 6. Subsidies and other incentives to capital 3 7. Growth poles, industrial districts, other location subsidies 4 8. Public sector industrialization and industrial location regulations 21 Note: Classification excludes expenses related to salaries and categories that lacked enough information to classify in one of the eight categories (other subsidies). Tunisia FIGURE 4.9  Subsidies made up 23–31 percent of Tunisia’s Robustness check yearly budget, with the greater share linked to current expenditure, To check the extent to which the distribu- 2013–17 tion of spending categories is consistent with 35 Share of budget for subsidies (%) the foregoing analysis, we analyzed the 30 expenses of Finance Law 2017 within the 7 eight categories (table 4.3). 25 8 7 7 Before the classification, two special cases 20 6 6 were excluded: First, salaries were excluded. 15 In Tunisia, 53 percent of total expenses are 25 10 related to salaries. Of the international com- 20 17 17 19 19 parators, South Africa has the highest allo- 5 cation of expenses on salaries, with about 0 35 percent, far below Tunisia. Poland, 2013 2014 2015 2016 2017 Average Spain, Hungary, and Austria have even Subsidies on capital investments Subsidies on current expenditure lower shares, ranging from 26 percent to Source: Tunisia’s Finance Law of 2017; data from World Bank’s Open Budgets Portal (BOOST). 21 percent. Second, 32 percent of Tunisia’s total budget could not be classified in the catego- (such as cash transfers to producers to boost ries for lack of information or because of an economic sector or to promote employ- ambiguous categories. Among the remain- ment for people with disabilities) or indirect ing expenses, Tunisia’s inclination toward (such as a tax relief in the form of payable tax spatially distortive interventions is con- credit for enterprises or a price reduction for firmed, with 53 percent of total expenses health services) to influence the level of pro- allocated within the space-based categories duction, prices, or the profits of market and (5 to 8) and only 5 percent toward broad- nonmarket enterprises.19 Subsidies go to pro- b a s e d gove r n a n c e a nd i n s t it ut ion a l ducers, not final consumers. In the case of reforms. Tunisia, as in most countries, indirect subsi- dies could not be readily quantified. Subsidies In 2017, 24 percent of the total budget in Tunisia’s Finance Law of 2017 allocated Tunisia was allocated to subsidies in four cat- 26 percent of the total budget to subsidies, egories. First was public intervention expen- with 19 percent toward current expenses and diture by program, which contributed 7 percent toward capital expenditures—the 95  percent of total subsidies on current same as the average for 2013–17 (figure 4.9). expenditures. The other three contributed Governments in general categorize their 87 percent of subsidies toward capital expen- expenses as subsidies when they are direct ditures: interventions in the economy 1 4 2    C ONVERGEN C E FIGURE 4.10  Subsidies on current expenditures, FIGURE 4.11  Subsidies on capital expenditures, representing 19 percent of Tunisia’s total budget, representing 7 percent of Tunisia’s total budget, are mainly for place-based interventions are less focused than current expenditures on Percentage of total subsidies on current expenditures place-based interventions Percentage of total subsidies on capital expenditures 9 5 28 11 29 6 5 10 4 53 40 2. Basic health, education, and related human capital improvements 5. Physical infrastructure for connectivity and 1. Broad-based governance and institutional reforms to support local production 2. Basic health, education, and related 6. Subsidies and other incentives to capital human capital improvements 7. Growth poles, industrial districts, other 3. Provision of basic public services location subsidies 5. Physical infrastructure for connectivity and to support local production Sources: Tunisia’s Finance Law of 2017; data from World Bank’s Open 6. Subsidies and other incentives to capital Budgets Portal (BOOST). 7. Growth poles, industrial districts, other location subsidies (47 percent), interventions in the social field 8. Public sector industrialization and industrial location regulations (28 percent), and investment in services and Sources: Tunisia’s Finance Law of 2017; data from World Bank’s Open infrastructure (11 percent). Budgets Portal (BOOST). Using the same spatial lens as before, we can identify how spatially distortive direct subsidies are in Tunisia.20 In 2017, 95 percent Analyzing the three capital expenditure of total subsidies on current expenditures categories, subsidies are less spatially distor- were allocated to a category called “public tive than current expenses. Twenty-six per- intervention expenditure by program,” of cent of total subsidies on capital expenses are which 75 percent were focused on place- on place-based interventions, while 44 per- based interventions (categories 5, 6, and 7), cent are on people-based interventions, and while only 25 percent were focused on 29 percent on institution-based interventions people-based interventions (figure 4.10). ­ (figure 4.11). Forty percent of the subsidies Almost half of total subsidies on current are focused on housing and buildings for expenses are for food support and market con- social support (basic health, education, and trol (34 percent) and energy (14 percent), related human capital improvements), while which are classified as subsidies and other the 29 percent on institutional interventions incentives to capital (category 6), while 18 per- is split between grants to Tunisian Bank of cent is split in half between truck transport Solidarity, a National Advertising Program, (category 5) and local affairs (category 7).21 and a Program for Regional Development. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    143 However, as noted above, subsidies on capital shows the concentration share of indi- investments only represent 7 percent of the rect subsidies in each income decile of the total budget, whereas subsidies on current population. Accounting for 2.3 percent of investments represent 19 percent. government spending in 2010, indirect sub- Moreover, indirect subsidies are tar- sidies were  directed to the rich: the bot- geted to the richest people in Tunisia, leav- tom 20 ­ percent of the population received ing those in need further behind. Table 4.4 only 11.7 percent, and the richest 10 percent received 18.3 percent. TABLE 4.4  Concentration shares of indirect Egypt subsidies in Tunisia, by income decile, 2010 As mentioned in box 4.3, government Decile Indirect subsidies (%) expenditures in Egypt include on-budget and 1 5.2 off-budget expenditures. On-budget data, 2 6.5 however, are available only at the aggregated 3 7.6 level among nine functional categories. 22 4 8.3 This limits our ability to complete the same 5 8.7 analysis for Egypt as for Jordan, Saudi 6 9.3 Arabia, and Tunisia—the focus countries of 7 10.7 this analysis. 8 11.8 Approximately 59 percent of the off-­ 9 13.7 budget expenditure assigned to economic 10 18.3 authorities goes to place-based interven- Source: Jouini et al. 2017, using data from 2010 Tunisia National Survey of tions, with 42 ­p ercent on direct interven- Consumption and Household Living Standards (latest available). Note: Calculations from Commitment to Equity (CEQ) Institute’s “Tunisia tion in public sector industrialization Master Workbook 2015,” where 1st decile is the poorest and 10th decile is and industrial location regulations (cate- the richest. “Indirect subsidies” are energy, food, and other general or targeted price subsidies. gory 8). Table  4.5, panel a, shows the TABLE 4.5  Off-budget economic authorities in Egypt are highly spatially distortive a. Including public sector social insurance fund (% of expenditure) Spatial category Current Capital Total 1. Broad-based governance and institutional reforms 0 6 2 2. Basic health, education, and related human capital improvements 1 1 1 3. Provision of basic public services 3 8 5 4. Skill development, worker training, wage subsidies 36 25 33 5. Physical infrastructure for connectivity and to support local production 1 2 2 6. Subsidies and other incentives to capital 9 7 8 7. Growth poles, industrial districts, other location subsidies 5 12 7 8. Public sector industrialization and industrial location regulations 44 39 42 b. Without public sector social insurance fund (% of expenditure) Spatial category Current Capital Total 1. Broad-based governance and institutional reforms 0 7 3 2. Basic health, education, and related human capital improvements 2 1 2 3. Provision of basic public services 4 10 6 4. Skill development, worker training, wage subsidies 15 6 12 5. Physical infrastructure for connectivity and to support local production 2 3 2 6. Subsidies and other incentives to capital 12 8 11 7. Growth poles, industrial districts, other location subsidies 7 15 9 8. Public sector industrialization and industrial location regulations 58 49 55 Source: 2016–17 financial statements for each economic authority, Ministry of Finance of Egypt. 1 4 4    C ONVERGEN C E skewedness of off-budget expenditure in place-based policies (table 4.6). However, as Egypt toward spatially distortive policies. mentioned before, public business sector Moreover, a large share of category 4 (skill expenditures are not included in the off- development, worker training, wage subsi- budget data. dies) is due to government incentives to public sector employees, where the social Jordan insurance fund for employees in the gov- The current expenditures and capital ernmental sector accounts for 24 percent expenditures of all Jordanian line ministries of total off-budget expenditure. If we for 2018 were classified within the eight spa- remove the social insurance expenditure, tial categories (table 4.7). As in the Tunisian numbers skew even more toward spatially example, three categories were excluded from distortive interventions (table 4.5, panel b), the analysis: accounting for 78 percent of total off-­ budget expenditure. As noted earlier, •  Public sector wages, salaries, and social Egypt is one of the Middle East and North security contributions , which stood at Africa countries with the highest share of 58 percent public employment, hindering private sec- •  Use of goods and services by line min- tor activity. istries, including security, cleaning, and In both on-budget and off-budget data other contracts, which accounted for from economic authorities, 37 ­ p ercent of 23 percent (also including subsidies and the national budget is allocated toward grants to a wide range of uncategorizable TABLE 4.6  On-budget and off-budget expenditure by economic authorities in Egypt, by spatial category Percentage of annual expenditure Spatial category On-budget Off-budget Total 1. Broad-based governance and institutional reforms 43 2 8 2. Basic health, education, and related human capital improvements 21 1 13 3. Provision of basic public services 31 5 23 4. Skill development, worker training, wage subsidies 0 33 19 5. Physical infrastructure for connectivity and to support local production 0 2 1 6. Subsidies and other incentives to capital 5 8 8 7. Growth poles, industrial districts, other location subsidies 0 7 4 8. Public sector industrialization and industrial location regulations 0 42 24 Source: Ministry of Finance of Egypt. Note: For Egypt, the Financial Statement 2018–19 was used for on-budget data, and the total expenses of the 48 economic authorities were retrieved from their 2016–17 financial statements from the Ministry of Finance website. Percentages within each type of budget (on-budget and off-budget) consider only its own category. The “Total” column considers both. The off-budget data shares include the public sector social insurance fund. TABLE 4.7  Jordanian public expenditures, by spatial category, 2018 Policy option Share of total expenditure (%) 1. Broad-based governance and institutional reforms 8 2. Basic health, education, and related human capital improvements 46 3. Provision of basic public services 4 4. Skill development, worker training, wage subsidies 4 5. Physical infrastructure for connectivity and to support local production 12 6. Subsidies and other incentives to capital 21 7. Growth poles, industrial districts, other location subsidies 4 8. Public sector industrialization and industrial location regulations 1 Source: “Capital and Current Expenses (2018),” classified by line ministry, General Budget Department of the Ministry of Finance, Hashemite Kingdom of Jordan. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    145 administrative committees such as Jordan policies, the majority goes for building Motor Sports Committee, political party schools, hospitals, and other social amenities subsidies, and the Joint Jerusalem Fund in regional governorates. subsidies) The vast majority of current expenditures •  Military spending and nuclear energy in Jordan (85 percent) goes toward basic projects, which accounted for 0.2 percent health, education, and related human capital of total expenditures. improvements, standing at about JD 500 mil- lion (figure 4.12). Within that category, Excluding those categories, public expen- 93   p e rc e nt i s f u n n e l e d towa rd s i x ditures fall from about JD 9 billion to about organizations: Social Aids for the Royal JD 1.7 billion, of which 38 percent fall into Hashemite Court, Cash Subsidies under the spatially distortive interventions, 54 percent Ministry of Finance, Jordanian Government into people-based interventions, and approxi- Universities subsidies under the Ministry of mately 8 percent into institution-based Education and Scientific Research, the interventions. Kidney Failure Fund and Civil Health Distinguishing capital expenditures and Insurance Fund under the Ministry of current expenditures provides a clearer pic- Health, and the National Aid Fund under the ture of the state’s spatial intervention. Of Ministry of Social Development. capital expenditures, 56 percent go toward spatially distorting policies. Of the capital expenditures going toward people-based Implications and persistence of governments’ approach to shaping markets in the Middle FIGURE 4.12  Jordanian current public East and North Africa expenditures, by spatial category, 2018 Limitations on competition and an emphasis on spatially distortive territorial 5 development policies in the Middle East and North Africa have created a high fiscal 10 burden with limited results. At the city level, the new-city orientation of several of the region’s governments has resulted in the development of many expensive and unpro- ductive new cities and zones at the expense of enabling existing centers of growth to thrive. A mainstay of Egypt’s overarching devel- 85 opment vision for several centuries—and now part of the “Egypt Vision 2030” strategy—has been a model whereby Egypt ­ leverages its desert land to create new cities for growth (Sims 2015). An analysis of the 2. Basic health, education, and related government’s fiscal year 2015–16 budget by human capital improvements a nongovernmental organization, 10 Tooba 4. Skill development, worker training, wage subsidies Collective, found that about 30 percent of 6. Subsidies and other incentives to capital the nation’s built environment budget was allocated to new cities and zones (which Source: “Summary of Functional Classifications for Estimated Public Expenditures According to Functional Divisions and Groups for the Fiscal host about 2 percent of the nation’s popula- Year 2018,” Law No. (1) for the Year 2018, General Budget Law for the Fiscal tion), and about 29 percent of the budget Year 2018, Hashemite Kingdom of Jordan. Note: Chart excludes spatial categories 1, 3, 5, 7, and 8 because their value was allocated to existing cities and towns is less than 1 percent. (which host roughly 98 percent of the 1 4 6    C ONVERGEN C E nation’s population) (Sims 2015). Although those incentives could be perceived as both it is not possible to access GDP data for new a means of showing attention to regions in cities and zones in Egypt, an analysis of 23 relative need and also perhaps as a way of new cities established between 1979 and shifting competition away from economic 2000 found that despite a projected popula- hubs that are home to firms that govern- tion of over 20 million Egyptians, the 2006 ments want to protect from competition. In population was 783,103, representing less Tunisia, firms benefiting from those incen- than 4 percent of projected population. tives have located in a band parallel to the Several new cities were completely unoccu- western coastline in the zones bordering pied several decades after their development leading areas and distant from the areas of (Sims 2015). weak economic activity. Worse yet, many Although the vast majority of the desert of the success cases—foreign and domestic zones and cities have been underoccupied or firms, notably Benetton, that relocated to unoccupied, the government of Egypt has Tunisia’s lagging regions to benefit from announced a scaling-up of its new zone and those incentives—closed their operations as city development over the past half decade. soon as their subsidies expired (World The Industrial Development Authority, Bank 2018c). which is responsible for most zones in As with new cities and zones, several of Egypt, announced in 2017 that it would the region’s governments are expanding assign 60 million square meters of industrial their use of incentives and transfers for zone space between 2017 and 2020 (Sims investments in lagging regions, with few 2015). It had assigned 11 million square changes to their approach despite evidence meters during 2006–15 and another 11 mil- of limited results. In Tunisia after the lion square meters in 2016. In April 2017, Jasmine Revolution, the government revised t he M i n ister of Trade a nd I ndust r y the Investment Incentive Code, which gov- announced the establishment of new zones erns regional subsidies, to further distin- in all seven governorates of Upper Egypt, guish between degree of need, and it one of the country’s most lagging regions. increased its allocation of subsidies. Also, in Meanwhile, the occupancy rates of existing response to the worsening conditions in lag- z on e s i s low : t h e ave ra ge i n Q e n a ging regions after the Jasmine Revolution, Governorate was 14 percent in 2016, and the government increased the volume of the average in Sohag Governorate was financing it assigned to the Regional 34 percent (World Bank 2016). In addition, Development Program—a transfer program over the past four years, the Ministry of established several decades ago to support Housing announced 12 new cities (Sims regional development. 2015). At the cross-country level, many govern- At the within-country level, most Middle ments’ concerns about undermining their East and North Africa countries offer con- local firms—and thus the insiders support- siderable investment incentives to firms that ing their administrations—have led them to locate in high-priority regions identified by maintain barriers to trade and hence also governments. Those incentives cost govern- the barriers to increasing revenues from ments and deliver limited numbers of jobs greater international market access. The in lagging regions. In Tunisia, the priority low level of international trade observed in of a territory is negatively associated with the Middle East and North Africa relative its development level and is viewed as ter- to comparator regions affects overall eco- ritorial affirmative action—a subsidy for nomic development and spatial develop- firms to create jobs in regions that other- ment in two ways: First, it reduces the wise have an undersupply. In some cases, demand for production and for the further H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    147 development of the system of cities in the overcome the resistance from influential region’s countries. Second, some border rent-seeking beneficiaries of the status quo. areas that would otherwise be thriving are Since these cases represent situations of relatively depressed or preoccupied with concentrated winners and diffuse losers, informal activity because of the hard for- the pressure on governments to sustain mal borders they face. these interventions is high. Despite the limited impact of their Second, it can be easier to start fresh primary territorial development interven- ­ than to fix an existing challenge. In the tions, several of the region’s governments case of desert development policies in continue to take the same approach to ter- Egypt, Saudi Arabia, the United Arab ritorial development for several reasons. Emirates, and others, decision makers First, despite limited economic returns to advocate for developing new cities and their spatial bets, some groups receive zones as a way of navigating around rather them. Investors that receive no-cost or low- than working th roug h constraints to cost access to land that will appreciate in growth. They pitch new city and zone value benefit from new city and zone poli- development as ways to develop new cen- cies. Likewise, investors that benefit from ters of growth that are unfettered by poor generous locational subsidies have an inter- planning and informal development. est in sustained and even augmented subsi- Third, concrete information is scarce on dies linked to their location choices. the relative successes and failures of territo- Regarding international trade, there will be rial development policies in the region. In strong support for regional integration general, it is difficult to accurately quantify when local elites, especially the business the impact of different territorial develop- c om mu n it y, s e e opp or t u n it ie s f rom ment policies. In the Middle East and North enhanced trade and interaction. There will Africa context, this methodological challenge be little support, or outright obstruction- is exacerbated by the paucity of data and ism, if they benefit from the status quo accurate information about the costs, bene- (Malik and Awadallah 2013). Even though fits, and trade-offs of various development fragmentation of regional markets means interventions. that companies forgo benefits from scale Fourth, there is always hope. Some territo- and agglomeration, in many countries, they rial development experiments in the region gain from high levels of monopoly power in have succeeded—such as Dubai, Tangier, and their domestic markets. As long as market Marrakesh—which have inspired decision entry is restricted and political patronage makers elsewhere to attempt to replicate their serves incumbents, there will not be a siz- successes. What separates the success stories able constituency for greater openness. from the larger number of relative failures are A high degree of state ownership; resource natural geography advantages (as in Dubai rents that reduce reform pressures; and and Tangier); a focus on reinforcing market rents from licensing, quotas, contracts, or demand and organically and contiguously preferential access to land further reduce expanding development; up-front efforts to efficiency and international competitive- make the business environment extremely ness (Diwan, Keefer, and Schiffbauer 2015; welcoming; and coordinated efforts and Gatti et al. 2013). It will thus be difficult to high-level vision and support to deliver these initiate greater regional integration and projects. Few of the territorial efforts have related reforms without finding ways to met these requirements. 1 4 8    C ONVERGEN C E Annex 4A Disaggregation of government expenditure, by government level, for each of the comparator countries TABLE 4A.1  Disaggregation of government expenditure for comparator countries, by spatial category a. Spain Share of public Spatial category expenditure (%) 1. Broad-based governance and institutional reforms 45 2. Basic health, education, and related human capital improvements 42 3. Provision of basic public services 5 4. Skill development, worker training, wage subsidies 5 5. Physical infrastructure for connectivity and to support local production 1 6. Subsidies and other incentives to capital 1 7. Growth poles, industrial districts, other location subsidies 0 8. Public sector industrialization and industrial location regulations 0.2 b. South Africa Share of public Spatial category expenditure (%) 1. Broad-based governance and institutional reforms 59 2. Basic health, education, and related human capital improvements 22 3. Provision of basic public services 11 4. Skill development, worker training, wage subsidies 1 5. Physical infrastructure for connectivity and to support local production 4 6. Subsidies and other incentives to capital 3 7. Growth poles, industrial districts, other location subsidies 0 8. Public sector industrialization and industrial location regulations 2 c. Hungary and Poland Hungary (share of Poland (share of public Spatial category public expenditure, %) expenditure, %) 1. Broad-based governance and institutional reforms 24 15 2. Basic health, education, and related human capital improvements 58 69 3. Provision of basic public services 5 8 4. Skill development, worker training, wage subsidies 3 2 5. Physical infrastructure for connectivity and to support local production 8 4 6. Subsidies and other incentives to capital 2 1 7. Growth poles, industrial districts, other location subsidies 0 0 8. Public sector industrialization and industrial location regulations 0.4 0.2 Source: Government Finance Statistics (GFS) database 2015, International Monetary Fund. In addition, data to show changes over a Several Middle East and North Africa 10-year period among benchmark countries countries were further analyzed concerning were obtained from the International their government expenditures in the same Monetary Fund’s (IMF) Classification of eight spatial areas relative to those in Functions of Government (COFOG) in the c omparator countries (table 4A.3). the  ­ Government Finance Statistics (GFS) system The  expenditure distribution analysis (table 4A.2). These changes are shown graph- revealed striking differences with the com- ically in figure 4.6. parator countries—toward more spatially H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    149 TABLE 4A.2  Distribution and changes in government expenditures by comparator countries through a spatial lens Percentage of government expenditure Austria Spain South Africa Hungary Poland Category 1995 2016 1995 2016 2012 2016 1995 2015 2002 2015 1. Broad-based governance and institutional reforms 21 19 22 20 44 52 29 25 16 16 2. Basic health, education, and related human capital improvements 61 64 51 62 34 30 51 53 66 63 3. Provision of basic public services 6 4 9 8 9 8 6 7 9 9 4. Skill development, worker training, wage subsidies 4 6 6 5 1 1 5 2 3 2 5. Physical infrastructure for connectivity and to support local production 5 6 8 3 3 3 5 9 5 8 6. Subsidies and other incentives to capital 3 1 2 2 8 4 4 2 1 2 7. Growth poles, industrial districts, other location subsidiesa — — — — — — — — — — 8. Public sector industrialization and industrial location regulations 0.2 0.4 1.9 0.5 1.2 1.8 0.1 0.5 0.2 0.3 Source: Government Finance Statistics (GFS) database, International Monetary Fund. Note: — = not available. The period considered for each country uses the oldest and most recent years available in the IMF database. Expenditures on foreign aid are excluded from the analysis. a. None of the functions listed in the Classification of Functions of Government (COFOG) of the GFS system were classified in “7. Growth poles, industrial districts, other location subsidies.” This could be driven by the IMF database exclusion of public corporations, which underestimates potential market activities undertaken by comparator countries, such as growth poles and industrial districts. TABLE 4A.3  Expenditure distribution in the Middle East and North Africa is highly weighted toward spatially distortive policies Percentage of government expenditure Policy option Moroccoa Egypt, Arab Rep.b Jordanc Tunisiad 1. Broad-based governance and institutional reforms 0.2 8 8 15 2. Basic health, education, and related human capital improvements 5 15 46 17 3. Provision of basic public services 16 23 4 16 4. Skill development, worker training, wage subsidies 1 19 4 2 5. Physical infrastructure for connectivity and to support local production 15 1 12 44 6. Subsidies and other incentives to capital 34 8 21 3 7. Growth poles, industrial districts, other location subsidies 18 4 4 3 8. Public sector industrialization and industrial location regulations 10 24 1 0.1 Sources: Government data (as described in notes below). a. For Morocco, the information source is the 2017 Finance Law, specifically the “General Budget” and the “Public Establishments and Enterprises” segments. b. For Egypt, the Financial Statement 2018–19 was used for on-budget data, and the total expenses of the 48 economic authorities were retrieved from their 2016–17 financial statements from the Ministry of Finance website. c. For Jordan, the information source is “Capital and Current Expenses (2018),” classified by all line ministries, reported by the General Budget Department of the Ministry of Finance. d. The Tunisia analysis draws from the most recent five-year plan, the Development Plan 2016–20. distortive policies. This distribution is shown government functions published either by graphically in figure 4.7. ­ the International Monetary Fund or by each of the countries into the eight spatial categories. Annex 4B Disclaimers regarding Fu r ther, differences in govern ment the classification of government dissemination of expenditure data has expenditures meant that we do not have one standard- Table 4B.1 lists several disclaimers con- ized means for accessing expenditure cerning the classifications of each of the information. 1 5 0    C ONVERGEN C E TABLE 4B.1  Difficulties in classification of spatial categories of government expenditures Spatial category Difficulty 1. Broad-based governance and institutional reforms Regulatory reforms are not always explicitly spatially blind. Basic health, education, and related human capital 2.  Housing is classified in this category because this expenditure is usually oriented toward improvements addressing people’s basic needs. However, expenditures linked to “new communities,” as in Egypt, are considered spatially distortive, hence placed within category 8. 3. Provision of basic public services Some basic services can be provided to foster economic activity in a region, which is not clear from the description. 4. Skill development, worker training, wage subsidies Splitting worker training from standard education is not always straightforward. Physical infrastructure for connectivity and to support 5.  No difficulties noted. local production 6. Subsidies and other incentives to capital Spatial policies are difficult to identify specifically if they are undertaken by other 7. Growth poles, industrial districts, other location subsidies entities within the central government. They can be hidden from explicit government 8.  Public sector industrialization and industrial location functions. Moreover, subsidies can be implicit within projects, making it difficult to regulations disaggregate budget by specific activity. Annex 4C Classification of the IMF database TABLE 4C.1  Classification of subcategories of the IMF Government Finance Statistics (GFS) database into eight spatial categories Government functions in IMF database Spatial category, by number A.  Expenditure on general public services Expenditure on executive, legislative, fiscal, and external affairs 1. Broad-based governance and institutional reforms Expenditure on foreign economic aid Excluded from the analysis. Expenditure on general public services 1. Broad-based governance and institutional reforms Expenditure on basic research 1. Broad-based governance and institutional reforms Expenditure on general public services research and development 1. Broad-based governance and institutional reforms Expenditure on general public services not elsewhere classified 1. Broad-based governance and institutional reforms Expenditure on public debt transactions 1. Broad-based governance and institutional reforms Transfers between different levels of government 1. Broad-based governance and institutional reforms B.  Expenditure on defense  Expenditure on military defense 3. Provision of basic public services Expenditure on civil defense 3. Provision of basic public services Expenditure on foreign military aid Excluded from the analysis. Expenditure on defense research and development 1. Broad-based governance and institutional reforms Expenditure on defense not elsewhere classified 3. Provision of basic public services C.  Expenditure on public order and safety  Expenditure on police services 3. Provision of basic public services Expenditure on fire protection services 3. Provision of basic public services Expenditure on law courts 1. Broad-based governance and institutional reforms Expenditure on prisons 1. Broad-based governance and institutional reforms Expenditure on public order and safety research and development 1. Broad-based governance and institutional reforms Expenditure on public order and safety not elsewhere classified 3. Provision of basic public services D.  Expenditure on economic affairs  Expenditure on economic, commercial, and labor affairs 1. Broad-based governance and institutional reforms Expenditure on agriculture, fishing, forestry, and hunting 6. Subsidies and other incentives to capital Expenditure on fuel and energy 6. Subsidies and other incentives to capital Expenditure on mining, manufacturing, and construction 6. Subsidies and other incentives to capital Expenditure on transport 5. Physical infrastructure for connectivity and to support local production Expenditure on communication 3. Provision of basic public services table continues next page H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    151 TABLE 4C.1  Classification of subcategories of the IMF Government Finance Statistics (GFS) database into eight spatial categories (continued) Government functions in IMF database Spatial category, by number Expenditure on other industries 8. Public sector industrialization and industrial location regulations Expenditure on economic affairs research and development 1. Broad-based governance and institutional reforms Expenditure on economic affairs not elsewhere classified 6. Subsidies and other incentives to capital E.  Expenditure on environmental protection  Expenditure on waste management 3. Provision of basic public services Expenditure on wastewater management 3. Provision of basic public services Expenditure on pollution abatement 3. Provision of basic public services Expenditure on biodiversity and landscape protection 3. Provision of basic public services Expenditure on environmental protection research and development 1. Broad-based governance and institutional reforms Expenditure on environmental protection not elsewhere classified 3. Provision of basic public services F.  Expenditure on housing and community amenities  Expenditure on housing development 2. Basic health, education, and related human capital improvements Expenditure on community development 8. Public sector industrialization and industrial location regulations Expenditure on water supply 3. Provision of basic public services Expenditure on street lighting 3. Provision of basic public services Expenditure on housing and community amenities research and development 1. Broad-based governance and institutional reforms Expenditure on housing and community amenities not elsewhere classified 2. Basic health, education, and related human capital improvements G.  Expenditure on health  Expenditure on medical products, appliances, and equipment 2. Basic health, education, and related human capital improvements Expenditure on outpatient services 2. Basic health, education, and related human capital improvements Expenditure on hospital services 2. Basic health, education, and related human capital improvements Expenditure on public health services 2. Basic health, education, and related human capital improvements Expenditure on health research and development 1. Broad-based governance and institutional reforms Expenditure on health not elsewhere classified 2. Basic health, education, and related human capital improvements H.  Expenditure on recreation, culture, and religion  Expenditure on recreational and sporting services 2. Basic health, education, and related human capital improvements Expenditure on cultural services 2. Basic health, education, and related human capital improvements Expenditure on broadcasting and publishing 2. Basic health, education, and related human capital improvements Expenditure on religious and community services 2. Basic health, education, and related human capital improvements Expenditure on recreation, culture, and religion research and development 1. Broad-based governance and institutional reforms Expenditure on recreation, culture, and religion not elsewhere classified 1. Broad-based governance and institutional reforms I.  Expenditure on education  Expenditure on preprimary and primary education 2. Basic health, education, and related human capital improvements Expenditure on secondary education 2. Basic health, education, and related human capital improvements Expenditure on postsecondary nontertiary education 2. Basic health, education, and related human capital improvements Expenditure on tertiary education 2. Basic health, education, and related human capital improvements Expenditure on education not definable by level 4. Skill development, worker training, wage subsidies Expenditure on subsidiary services to education 2. Basic health, education, and related human capital improvements Expenditure on education research and development 1. Broad-based governance and institutional reforms Expenditure on education not elsewhere classified 2. Basic health, education, and related human capital improvements J.  Expenditure on social protection  Expenditure on sickness and disability 2. Basic health, education, and related human capital improvements Expenditure on old age 2. Basic health, education, and related human capital improvements Expenditure on survivors 2. Basic health, education, and related human capital improvements Expenditure on family and children 2. Basic health, education, and related human capital improvements Expenditure on unemployment 4. Skill development, worker training, wage subsidies Expenditure on housing 2. Basic health, education, and related human capital improvements Expenditure on social exclusion not elsewhere classified 4. Skill development, worker training, wage subsidies Expenditure on social protection research and development 1. Broad-based governance and institutional reforms Expenditure on social protection not elsewhere classified 1. Broad-based governance and institutional reforms 1 5 2    C ONVERGEN C E TABLE 4D.1  Estimated shares of public expenditures, by category, in Lebanon, the Republic of Yemen, and Iraq Percentage of public expenditure Policy option Lebanon Yemen, Rep. Iraq 1. Broad-based governance and institutional reforms 2 43 14 2. Basic health, education, and related human capital improvements 12 11 29 3. Provision of basic public services 38 21 20 4. Skill development, worker training, wage subsidies 1 17 2 5. Physical infrastructure for connectivity and to support local production 34 0.4 0 6. Subsidies and other incentives to capital 14 8 22 7. Growth poles, industrial districts, other location subsidies 1 0 1 8. Public sector industrialization and industrial location regulations 0.2 0.03 11 Sources: Harake and Kostopoulos 2018; National Budget 2013, Republic of Yemen; Citizens Budget 2018, Iraqi Ministry of Finance. Annex 4D Reasons for excluding Data on Iraq and the Republic of Yemen other Middle East and North are highly aggregated, so the classification Africa countries from the spatial shown in the table 4D.1 should be considered analysis of government with caution. For the Republic of Yemen, we expenditures used data classified by function in the National Budget 2013, which only has a three-digit Middle East and North Africa countries classification used by the IMF. For Iraq, we considered in the analysis of this chapter were used data published in the Citizen Budget included because of the availability of sec- 2018, albeit with few functional categories. toral data on government expenditures. For No detailed data on expenditures are countries such as Lebanon, Iraq, and the available for Algeria, Djibouti, Libya, West Republic of Yemen, publicly accessible data Bank and Gaza, and the GCC countries. were not available at a level of disaggregation sufficient to provide a robust analysis and compare with the rest of the region and com- Notes parator countries. We include information on 1. The Gulf Cooperation Council (GCC) coun- these three countries below. tries are Bahrain, Kuwait, Oman, Qatar, To the authors’ knowledge, Lebanon does Saudi Arabia, and the United Arab Emirates. not have a national budget. The team reviewed 2. This score includes the GCC countries, which a list of projects that Lebanon intends to invest average a GCI score of 4.75. Excluding the in, as reviewed in the World Bank’s assess- GCC subregion results in an average score ment of the country’s Capital Investment Plan of 3.93. (Harake and Kostopoulos 2018). Given the 3. The 2017–18 index considers 137 countries, limitations of this data, we did not include our with 66 countries scoring higher than 4.28 and analysis of it in the chapter; however, we 93 countries higher than 3.93 (Schwab 2017). reflect the distribution of projects based on 4. The Oslo Accords, a set of agreements signed by the government of Israel and the leader- our spatial classification criteria (table 4D.1). ship of the Palestine Liberation Organization Another source that could have been used is (PLO), were ratified in 1993 (Oslo I) and the lagging regions report produced by the 1995 (Oslo 2). Under the Accords, the PLO Lebanese Center for Policy Studies, which has agreed to formally recognize the state of data on the actual expenditure, by district, of Israel, and Israel in turn allowed for limited the Cou ncil for Development and self-governance in West Bank and Gaza. Reconstruction as of September 2017 (Atallah 5. “Registered factories” refers to those regis- et al. 2018). However, this expenditure repre- tered with Egypt’s Industrial Development sents a partial picture of the total expenditure Authority (IDA). executed in Lebanon, and using it would mis- 6. Exceptions are Israel and the Islamic Republic lead the reader. of Iran. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    153 7. See, for instance, the “China shock” debate in government’s expenditure in the Middle East the United States (Autor, Dorn, and Hanson and North Africa region. This means that allo- 2016). cation of expenses in comparator countries 8. For the analysis underlying the following may be underestimating shares in place-based estimates, see Ianchovichina and Ivanic interventions, while the expenses in the Middle (2014). East and North Africa will be considering 9. The only country from the Middle East them. It is also crucial to highlight that public and  North Africa in the IMF database is corporations in the region are focused not the  United Arab Emirates, but its data are only on place-based policies but also on peo- not disaggregated by specific function, mak- ple-based policies, diminishing the potential ing the spatial classification unmeaningful. bias toward place-based policies. Moreover, the only Latin America country in 17. With the exception of the United Arab the IMF database with disaggregated data Emirates, which is included in GFS but for is  El Salvador, which is not comparable which there are significant data gaps that ren- with  most Middle East and North Africa der developing an allocation profile unfeasi- countries. ble currently. Although the definition of the 10. Government units include all nonmarket expenditure profile of GFS countries using nonprofit institutions that are controlled and our territorial development spectrum is mainly financed by the government, including imperfect, doing so for Middle East and budgetary and extrabudgetary data. North Africa countries based on national 11. General government expenditure includes instruments for planning and budgeting is those performed by the central, state, and local also imperfect. We will make accessible the governments, including any social security files within which we assigned line items to fund in the administrative organization of the territorial intervention categories, and we country. Annex 4C includes the spatial classifi- will engage in consultations in Middle East cation of expenditure by each level of govern- and North Africa countries to ground-truth ment, including social security funds at the the categorizations. central level. There were no disaggregated data 18. The analysis of MENA countries is based on for Austria. one year of analysis and for the most recent 12. Extrabudgetary entities are those with indi- year for which complete data were available. vidual budgets not fully covered by the main Patterns observed may be the result of speci- (or general) budget (such as road transporta- ficities in spending in the given year (or five tion or nonmarket production of health and years in the case of Tunisia). education services). 19. Countries may differ on how to classify spe- 13. It is important to highlight that this classifi- cific types of subsidies or expenses. cation is subjective and that each of the 69 20. To classify functions of subsidies by spatial subcategories may contain more than one of dimension, only the four categories men- the 8 spatial categories; however, this exer- tioned above are considered. Moreover, cise offers a general sense of the degree of within these four categories, only subsidies spatial distortion introduced by government that accounted for more than 5 percent of expenditures. Additional disclaimers, by the total subsidies by category are included. spatial category, as well as specific percent- 21. Although subsidies for primary products and ages by country and category are shown in transport were established in the 1990s, the annex 4B. energy subsidy was introduced for the first time 14. The classification of each of the 69 subcate- in 2003 to promote the competitiveness of the gories is listed in annex 4C. private sector and support the purchasing 15. The period considered for each country uses power of the middle class because of increases the oldest and most recent years available in in energy prices in the international market. the IMF database. 22. The nine functional categories on on-budget 16. An important difference between the IMF data for Egypt are public services; public database and the analysis of the countries in safety; economic affairs; environmental pro- the Middle East and North Africa is that the tection; housing and social infrastructure; IMF database excludes public corporations, health; youth, culture, and religion; educa- which are highly relevant to understand the tion; and social protection. 1 5 4    C ONVERGEN C E References the Levant.” Policy Research Working Paper 7135, World Bank, Washington, DC. Atallah, Sami, Daniel Garrote Sánchez, Nizar IMF (International Monetary Fund). 2014. Hassan, Dima Mahdi, and Jana Mourad. Government Finance Statistics Manual 2014. 2018. “Introductory Chapter: Analytical Washington, DC: IMF. Framework for Improving Urban Resilience in Jouini, N., N. Lustig, A . Mou m m i, and Lebanon’s Districts Impacted by the Syrian A.  Shimeles. 2017. “Fiscal Incidence and Refugee Crisis.” Unpublished report, Lebanese Poverty Reduction: Evidence from Tunisia.” Center for Policy Studies, Beirut. Economics Working Paper 1710, Tulane Autor, D. H., D. Dorn, and G. H. Hanson. 2016. University, New Orleans, LA. “The China Shock: Learning from Labor- Lall, Somik V. 2009. “Territorial Development Market Adjustment to Large Changes in Policy: A Practitioner’s Guide.” Report No. Trade.” Annual Review of Economics 8 (1): 70398, World Bank, Washington, DC. 205–40. Lall, Somik, Ayah Mahgoub, Augustin Maria, Benhassine, Najy, Youssef Saadani Hassani, Anastasia Touati, and Jose Luis Acero. 2019. Philip E. Keefer, Andrew H. W. Stone, and “Leveraging Urbanization to Promote a New Sameh Naguib Wahba. 2009.  From Privilege Growth Model While Reducing Territorial to Competition: Unlocking Private-led Disparities in Morocco.” Urban and Regional Growth in the Middle East and North Africa. Development Policy Note, Report No. 137850, MENA Development Report. Washington, World Bank, Washington, DC. DC: World Bank. Malik, A., and B. Awadallah. 2013. “The Chahoud, T. 2011. “Syria’s Industrial Policy.” E conomics of the A rab Spring.”  World Working paper, German Development Institute Development 45: 296–313. (DIE), Bonn. MEP (Ministry of Economy and Planning, Collier, Paul. 2003. Breaking the Conflict Trap: Kingdom of Saudi Arabia). 2010. “Regional C i v i l Wa r a n d D e v e l o p m e n t Po l i c y . Development.” In “Ninth Development Plan Washington, DC: World Bank; New York: and (2010–2014),” 209–22. Ministry of Economy Oxford University Press. and Planning, Kingdom of Saudi Arabia, Deininger, Klaus. 2003. Land Policies for Growth Riyadh. and Poverty Reduction. World Bank Policy Mills, R., and F. Alhashemi. 2018. “Resource Research Report. Washington, DC: World Regionalism in the Middle East and North Bank; Oxford: Oxford University Press.  Africa: Rich Lands, Neglected People.” Diwan, I., P. Keefer, and M. Schiffbauer. Brookings Doha Center Analysis Paper 20, 2015.  “P y ra m id C apit a l ism : Pol it ic a l Brookings Institution, Washington, DC. Connections, Regulation, and Firm Productivity Nabli, M. K., J. Keller, C. Nassif, and C. Silva- in Egypt.” Policy Research Working Paper Jáuregui. 2006. “The Political Economy of 7354, World Bank, Washington, DC. Industrial Policy in the Middle East and North Gatti, R., M. Morgandi, S. Brodmann, D. Angel- Africa.” Working Paper, Report No. 116424. Urdinola, J. M . Moreno, D. M arot ta, World Bank, Washington, DC. M.  Schiff bauer, and E . Mata Lorenzo. Rijkers, B., C. Freund, and A. Nucifora. 2014. 2013.  Jobs for Shared Prosperity: Time for “All in the Family: State Capture in Tunisia.” Action in the Middle East and North Africa. Policy Research Working Paper 6810, World Washington, DC: World Bank. Bank, Washington, DC. Harake, Wissam, and Christos Kostopoulos. Schiffbauer, M., A. Sy, S. Hussain, H. Sahnoun, 2018. “Strategic Assessment: A Capital and P. Keefer. 2015. Jobs or Privileges: Investment Plan for Lebanon—Investment Unleashing the Employment Potential of the Opportunities and Reforms.” Working Paper, Middle East and North Africa. M ENA R e p o r t N o . 1 2 4 8 1 9, Wo r l d B a n k , Development Report. Washington, DC: Washington, DC. World Bank. Ianchovichina, Elena, and Maros Ivanic. S c hw a b , K l au s , e d . 2 017. T h e G l o b a l 2014. “Economic Effects of the Syrian War Competitiveness Report 2017–2018. Geneva: and the Spread of the Islamic State on World Economic Forum. H o w S t a t e s S h a p e Ma r k e t s t h r o u g h S p a t i a l a n d P r i v a t e S e c t o r D e v e l o p m e n t B e t s    155 Si ms , D. 2015. Eg y pt ’s De se r t D re a m s: Appraisal Document, Report No. 108094, Development or Disaster? Oxford: Oxford World Bank, Washington, DC. University Press. ———. 2017a. Doing Business Regional Profile UN-Habitat (United Nations Human Settlement 2017: M iddle E a st an d Nor th Afric a . Programme). 2012. The State of Arab Cities Washington, DC: World Bank. 2012: Challenges of Urban Transition . ———. 2017b. Doing Business 2017: Equal Nairobi, Kenya: UN-Habitat. Oppor tunit y for All. Washington, DC: WEF (World Economic Forum) and World Bank. World Bank. 2018. “The Arab World Competitiveness ——— . 2 0 18 a . D o i n g B u s i n e s s 2 0 18 : Repor t 2018.” I nsig ht Repor t , World Reforming to Create Jobs. Washington, DC: Economic Forum (WEF), Geneva. World Bank. World Bank. 2012. “Arab Republic of Egypt: ———. 2018b. “Jordan Housing Sector Review.” Reshaping Egypt’s Economic Geography: Technical Note, Report No. AUS0000232, Domestic Integration as a Development World Bank, Washington, DC. Platform.” Report No. 71289-EG, World ———. 2018c. “Policy note on How Tunisia Can Bank, Washington, DC. Leverage Its Regions in a Differentiated ———. 2016. “Upper Egypt Local Development Manner for Shared Prosperity.” Project P rog ra m-for - Re su lt s P roje c t .” P roje c t 159072, World Bank, Washington, DC. Centralized Government: 5 Contributor to Economic Geography A s chapter 2 documented, many coun- enable job creation and provide public ser- tries in the Middle East and North vices, with some evidence that this preference Africa suffer large spatial disparities is especially pronounced in rural areas and in government service delivery, outcomes for low-income households. citizens, and citizen satisfaction. This chapter However, the policies that the region’s explains how excessive centralization in gov- governments deploy to meet these objectives ernment planning and provision of services are characterized by excessive centralization has contributed to spatial inequality, exacer- of resources, decision making, and service bating poor outcomes for citizens in the delivery mechanisms. This is true of policies region’s lagging areas. pursuing convergence in access to basic ser- In the Middle East and North Africa, vices (primarily through top-down allocation many governments have sought convergence and planning of public investment); conver- in living standards through the provision of gence in service quality (mainly through cen- jobs and public services. The development tralized service delivery mechanisms); and, as strategies and national plans of most of these discussed in chapter 4, convergence in con- countries emphasize enhancing welfare, sumption levels (typically through massive improving quality and efficiency of service public employment in lieu of poverty-targeted delivery, and increasing equality—as outlined social protection). in the Vision 2030 strategies recently released These policies closely reflect the “distinc- by Algeria, the Arab Republic of Egypt, Iraq, tively interventionist and redistributive” social Qatar, and the United Arab Emirates. contracts established in the region’s countries Moreover, recent national plans or govern- after independence, which have been marked ment programs for Jordan, Morocco, Saudi by (a) a preference for state planning over free- Arabia, and Tunisia specifically highlight the market outcomes; (b) the rise of a centralized, importance of reducing spatial imbalances in hierarchal bureaucracy or administration; and living standards, including through improved (c) a preference for redistribution, with the delivery of public services. This is in line with state perceived as responsible for providing citizen priorities: recent perception data con- welfare and social services including, to a firms that at multiple spatial scales, citizens great extent, employment (Yousef 2004). in the Middle East and North Africa believe This distinctively top-down approach to it to be the top priority for their states to welfare planning and delivery is reinforcing 157 1 5 8    C ONVERGEN C E spatial disparity. Drawing on empirical policies, par ticularly in ru ral areas. research and country-specific analytical work Subnational perception data from the 2016 within the Middle East and North Africa, Arab Barometer show that when asked to this chapter shows that these centralizing describe the essential characteristics of democ- policies are spatially biased. They encourage racy, respondents in the surveyed Middle East a vertically unbalanced flow of resources and and North Africa economies1 overwhelmingly decision making at the national level, com- value functions to improve their socioeco- bined with insufficient development of nomic well-being (narrowing the gap between human, financial, and technical capacity to rich and poor, providing basic necessities and respond to citizens’ needs for delivery of jobs for all, and providing quality public ser- qua l it y ser v ices at t he lo c a l level —­ vices) over characteristics reflecting political particularly in lagging areas. Effectively freedoms and fair institutions. meeting citizens’ demands for jobs and ser- On average, in the full sample of almost vices therefore requires moving from the 9,000 people in seven economies, 48 percent monumental to the incremental—away from of respondents agreed that government enable- top-heavy, state-centric models and toward ment of job opportunities for all is an essential more locally grounded, people-centric characteristic of democracy, compared with approaches to investment planning, service only 28 percent for government provision of delivery, and accountability. quality public services, 26 percent for free and fair elections, and 16 percent for freedom of speech (figure 5.1, panel a). This general rank- What do citizens expect of the ing of preferences holds in both urban and state in the Middle East and figure 5.1, panel b), and the prefer- rural areas (­ North Africa? ence for government enablement or provision What citizens expect of the state in the of both job opportunities and public services Middle East and North Africa varies spatially is particularly pronounced in the rural areas but with a clear preference for socioeconomic of most countries (figure 5.2). FIGURE 5.1  Respondents in surveyed Middle East and North Africa economies identified functions to improve socioeconomic well-being as the most essential characteristics of democracy Percentage of respondents selecting answers in multiple-choice format a. Responses, by economy b. All responses, urban versus rural Algeria Egypt, Arab Rep. Urban Jordan Lebanon Morocco Rural West Bank and Gaza Tunisia 0 20 40 60 80 0 10 20 30 40 50 Percent Percent Jobs Public services Fair elections Free speech Source: Arab Barometer Wave IV, 2016. Note: Percentages add up to more than 100 because they reflect respondents’ selection of one statement from each of four sets of answers to the question (Wave IV, no. 515): “If you have to choose only one from each of the four sets of statements that I am going to read, which one would you choose as the most essential characteristics of a democracy?” Categories shown summarize each choice described. C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    159 FIGURE 5.2  Rural respondents were more likely to cite government’s role in job creation and public service provision as essential characteristics of democracy Percentage of respondents selecting answers in multiple-choice format a. Importance of job creation b. Importance of public services Algeria Algeria Egypt, Arab Rep. Egypt, Arab Rep. Jordan Jordan Lebanon Lebanon Morocco Morocco West Bank and Gaza West Bank and Gaza Tunisia Tunisia 0 20 40 60 80 0 20 40 60 Percent Percent Urban Rural Source: Arab Barometer Wave IV, 2016. Note: Preferences reflect respondents’ selection of one statement from each of four sets of answers to the question (Wave IV, no. 515): “If you have to choose only one from each of the four sets of statements that I am going to read, which one would you choose as the most essential characteristics of a democracy?” Categories shown summarize each choice described. Between the two priorities, there is a clear objectives have been pursued in spatially preference for government’s role in enabling biased ways that limit achievement of the job creation over public service provision objective. at  the subnational level, which is poten- This section analyzes the policies that tially more pronounced in low-income house- Middle East and North Africa governments holds. In most governorates of Algeria, have used to uphold their end of the social Jordan, Lebanon, Morocco, and West Bank contract, focusing on efforts to encourage and Gaza,2 a greater proportion of respon- (a) convergence in access to basic services dents name government’s role in job creation (for which the dominant strategy has been over government provision of quality public top-down allocation and planning of public services as an essential characteristic of democ- investment), and (b) convergence in service racy (figure 5.3). For most countries, this pref- quality (mainly pursued through central- erence appears to be slightly more pronounced ized mechanisms of service delivery). The when considering only households in financial section presents evidence that these policies difficulty, although this finding should be are characterized by excessive centraliza- treated with caution given the low number of tion of capacity, resources, and decision- survey respondents in this category.3 m a k i ng pro ce sse s , wh ich u nder m i ne convergence objectives by reinforcing spa- tial disparities. Centralized government responses reinforce spatial bias, undermining instead of Centralized response 1: Top-down encouraging convergence allocation and planning of public capital investments results in As chapter 2 documented, authorities in fragmented or uncoordinated the region have struggled to uphold their end development of this redistributive social contract, suggest- ing that their policy menu needs to be revised. To assess the robustness of the Arab Barometer This chapter does not negate the importance data analyzed in this section, we analyzed World of redistribution but argues that convergence Values Survey (WVS) data for Middle East and FIGURE 5.3  Across Middle East and North Africa economies, subnational surveys also show a preference for governments’ role in job creation over public 160 service delivery Percentage of respondents selecting answers in multiple-choice format a. Egypt, Arab Rep. b. Jordan c. Lebanon d. Morocco Cairo Irbid Beirut Tanger-Tétouan Qalioubia Taza-AI Hoceima-Taounate Bani Swif Balqa Alexandria Oriental AI Bouhira Zarqa EI Nabatieh Fes-Boulemane Ismailia Tafilah Meknès-Tafilalet Gharbia Dakahlia Gharb-Chrarda-Beni-Hssen Amman Bekaa Sharqia Rabat-SalÈ-Zemmour-Zaer Damietta Aqaba Kafr AI Shiekh Grand Casablanca Giza Karak Chaouia-Ouardigha Mnoufia Mount Lebanon Tadla-Azilal Port Said Mafraq Suez Marrakech-Tensift-AI-Haouz AI Fayoum Jerash Doukkala-Abda Menia North Asyut Ajloun Souss-Massa-Draa Souhag Guelmim-Es-Semara Qena Madaba South Laayoune-Boujdour-Sakia El Hamra Aswan Luxor Ma’an Oued Ed-Dahab-Lagouira 0 20 40 60 80 100 0 20 40 60 0 20 40 60 80 0 20 40 60 80 Percent Percent Percent Percent e. Algeria f. Tunisia g. West Bank and Gaza Adrar Tunis Jenin Chlef Laghouat Ariana Oum EI Bouaghi Tobas Batna Ben Arous Bejia Taulkarm Beskra Manouba Béchar Blida Nabeul Qalqilia Bouira Zaghouan Tamanghasset Tbessa Bizerte Salfit Tlemcen Tiaret Beja Tizi Ouzou Nablus Algiers Jendouba Djelfa Jijel Kef Ramallah Setif Saida Seliana Skikda Jerusalem Sidi Bel Abbes Sousse Annaba Jericho Guelma Monastir Constantine Medea Mahdia Betlehem Mostaganem M’Sila Sfax Mascara Kairouan Hebron Ouargla Oran Kasserine EI Bayadh Jabalia Bordj Bou Arréridj Sidi bouzid BoumerdËs EI Taref Gaza City Gabes Tissemsilt EI Qued Mednine Khan Younis Khenchela Souk Ahras Tatouine Tipasa Deir al Balah Mila Gafsa Aïn Defla Na,ma Tozeur Rafah Aïn Témouchent Ghardaïa Kebili Relizane 0 20 40 60 0 20 40 60 80 0 20 40 60 80 Percent Percent Percent Jobs Public services Source: Arab Barometer Wave IV, 2016. Note: Preferences reflect respondents’ selection of one statement from each of four sets of answers to the question (Wave IV, no. 515): “If you have to choose only one from each of the four sets of statements that I am going to read, which one would you choose as the most essential characteristics of a democracy?” Categories shown summarize each choice described. C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    161 BOX 5.1  Comparing Arab Barometer and World Values Survey responses on government’s role The WVS Wave 6 (2010–14) ­ collected data about governments’ responsibility in enabling e ­ conomic what people consider “the most important aims of development (that is, jobs). Unfortunately, neither the country” (figure B5.1.1). It appears that people the subregional data nor the urban versus rural seg- consider economic growth, by far, to be the main mentation is available from the WVS for the Middle objective that their countries and governments East and North Africa countries, so we cannot com- should pursue. This corroborates the strong weight pare subnational results with the Arab Barometer. given by A rab Barometer respondents to their FIGURE B5.1.1  WVS respondents from most Middle East and North Africa economies identified economic growth as the country’s “most important” goal Percentage of respondents selecting answers in multiple-choice format Algeria Bahrain West Bank and Gaza Iraq Jordan Lebanon Libya Morocco Qatar Tunisia Egypt, Arab Rep. Yemen, Rep. 0 20 40 60 80 Percent Economic growth Security Freedom Prestige Sources: World Values Survey Wave 6 (2010–14 and 2015 datasets), World Values Survey Association 2015, http://www.worldvaluessurvey.org/WVSDocumentationWV6.jsp. Note: WVS = World Values Survey. Respondents selected from several choices in WVS Wave 6, question no. V60: “People sometimes talk about what the aims of this country should be for the next 10 years. . . . Would you please say which one of these you, yourself, consider the most important?” Categories shown summarize each aim described. North Africa economies. Although it does not basic services. Chapter 4 documents a func- include a question identical to the one analyzed tional imbalance between government spend- from the Arab Barometer, an analysis of some- ing on infrastructure (especially place based) what comparable questions and answers from the and spending on human capital (social ser- WVS reveal consistent perceptions (box 5.1). vices) in the region’s countries relative to The governments of the Middle East and comparator economies. North Africa have made sizable public capi- Indeed, sizable public infrastructure invest- tal investments in recent years as a key tool ments have been a hallmark of several of the for pursuing convergence in access to 1 6 2    C ONVERGEN C E region’s governments in recent years, including A related challenge is that the region’s gov- recent major infrastructure projects in Egypt ernments have tended to proliferate rather and megaprojects in Saudi Arabia. This focus than consolidate their subnational govern- on infrastructure is prevalent even in the social ment units, hindering their ability to take on services, with recent evidence showing that the a coherent spatial and regional development education budgets of many of these countries planning role. In Saudi Arabia, for example, heavily emphasize capital investments (such as the historic region of Hejaz is now split construction of new schools, rehabilitation between four governates. Iraq similarly took and expansion of existing facilities, or pro- a much simpler preindependence structure curement of school equipment) rather than and multiplied the number of governates. other learning inputs. Qatar, Kuwait, Lebanon is an extreme case, with more than Morocco, and Lebanon respectively allocate 1,100 municipalities (baladiyat) and some 53 24 percent, 21 percent, 13 percent, and 13 per- unions of municipalities, many of which have cent of public education spending to capital fewer than 10 member municipalities. The investment—much higher shares than the small size of these subnational units, their Organisation for Economic Co-operation and lack of local identity, and their upward-facing Development (OECD) average of 7.6 percent accountability pose significant obstacles to (World Bank 2018d). their viability as socioeconomic layers and Yet in most countries in the region, deci- removes what would otherwise be logical sions regarding the geographic allocation of c ou nt e r p a r t s for sp at i a l ly or i e nt e d investment expenditure are made by sectoral interventions. ministries within the central government, with little agency left to subnational govern- Centralized response 2: Service delivery ments. For example, in Egypt in 2016, local mechanisms favor primate cities, administration expenditures on nonfinancial especially political capitals assets for local development activities repre- sented only 3 percent of total local adminis- In the pursuit of convergence in service tration spending, reflecting that most local quality, most Middle East and North Africa investment expenditure in Egypt is decided countries have committed to decentralize and implemented by central government authority for service delivery to local bodies. agencies (Amin 2016). Likewise, in Jordan, Internationally, local authorities such as sectoral investments are generally planned at municipalities and city councils are com- the central level with limited attention to monly made responsible for delivering basic optimizing those investments across sectors services to citizens, and they are generally or applying a territorial planning framework viewed as the first and best contact points (World Bank 2018a). Even in Tunisia, which between the citizenry and the government has decentralized greater decision-making apparatus. authority to local administrations, five-year T he “I nter n at ion a l G u idel i ne s on municipal development plans make up the Decentralisation and Access to Basic sum of all municipal investment programs Services for All,” approved by the Governing but account for only 10 percent of total Council of the United Nations Human investment in urban infrastructure. The Settlement Programme (UN-Habitat) in remaining 90 percent is based on national 2007 and 2009, outline an international development plans financed and implemented consensus that the decentralization of directly by line ministries and state-owned responsibi l it ies , pol ic y ma nagement , service enterprises. This can lead to sectorally decision-­m aking authority, and sufficient fragmented or uncoordinated investment resources to local authorities are require- planning and funding decisions, without an ments for effective and sustainable service integrated view of regional and local priori- delivery (UN-Habitat 2009). Several Middle ties for socioeconomic development. East and North Africa countries have C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    163 recently made important advances in imple- visible in the region’s public administrations menting a decent rali zation agenda—­ today, particularly in the Mashreq subre- including Lebanon and Tunisia, which both gion. The public administration system is held long-awaited local elections in 2018. highly centralized in most of the region, Nonetheless, local government systems in with deconcentrated units of the central the region remain mostly deconcentrated government providing some services directly rather than decentralized,4 mirroring the (particularly health and education), while excessive concentration of the region’s urban basic services are provided by field offices of systems. Chapters 1–3 of the report docu- line ministries or governorates, districts, mented how territorial development in the and municipalities (UCLG 2010). Although Middle East and North Africa is character- local governments are usually run by elected ized by economic and demographic primacy councils, they typically have limited author- of the main city (spatial fragmentation ity over the services they are mandated to within the urban system). This is the legacy provide. Their role mostly consists in carry- of the political economy and administrative ing out service delivery decisions made by arrangements of the Ottoman Empire, the central government and performing lim- which concentrated public services and tax ited functions such as library and park ser- revenues in capitals and neglected lagging vices, street paving, street lighting, and regions (World Bank 2011). Subsequently, garbage collection (UCLG 2010). Figure 5.4 colonial powers developed metropolis- and summarizes the territorial governance struc- export-oriented economies that encouraged ture typical of the different Middle East and coastal agglomeration and hindered the North Africa subregions. The fiscal decen- development of effective administrative tralization indexes calculated by Ivanyna institutions (Brixi, Lust, and Woolcock and Shah (2012) suggest that the Mashreq 2015). This is especially the case in the subregion is significantly less decentralized Mashreq subregion, where the excessive than the Maghreb subregion, particularly in urban primacy and underdevelopment of the cases of Egypt, Iraq, Jordan, and the secondary cities seem to have been predomi- Syrian Arab Republic, as further shown in nantly the consequence of political bias chapter 3. (as further discussed in chapter 3).5 These decentralization patterns have High population growth, migration- resulted in a general mismatch between fiscal driven rapid urbanization, nondemocratic centralization and functional devolution of and centralized political government, and responsibility for delivering basic services. colonial history have all been identified as Two common metrics of decentralization in factors driving further urban concentration cross-country comparisons are the relation- (Faraji 2016). Politically centralized regimes ship of subnational revenues and expendi- in low- and middle-income countries tend to tures to national totals. Central transfers provide better services and safety in the cap- continue to be the main source of local ital city and give more attention to that local finance, and the share of local expenditures population. As such, there is evidence that in total government spending remains low migrants settling in the primate city come (figure 5.5). As a result, municipalities run not only from rural areas but also from perennial operational deficits. Where they small towns and medium-size cities (El-Din exist, the various local taxes, fees, and per- Haseeb 2012). As highlighted by Henderson mits that constitute the bulk of local revenues (2002), primate cities that are also political cover a small portion of the budget and go capitals are on average 25 percent bigger almost entirely to salaries and operations and than primate cities not concentrating politi- maintenance. Municipalities have little flexi- cal power. bility in adjusting the rates and mechanisms The legacy of this centralized structure of for levying fees for public services to reflect service provision and decision making is still the real costs of service provision. 1 6 4    C ONVERGEN C E FIGURE 5.4  The subregions of the Middle East and North Africa represent a spectrum in the degree of decentralization West Bank and Gaza Maghreb • Local governments are responsible for 27 functions • Countries are divided into including most public governorates headed by an services commonly appointed governor; delivered at the local level. municipalities have elected • The budget system is truly Mashreq decentralized. councils legally enabled to • Governorates focus mainly on approve plans, development public order and serve as the projects, regulatory controls, provincial seat through which and local budgets, with deconcentrated units of line varying degrees of political ministries plan and coordinate and administrative autonomy. investments or provide • Central governments control services. finances but the elected GCC • Central governments largely councils of municipalities Fiscal decentralization maintain control of financial can approve local • Saudi Arabia: Regional resources. budgets. councils headed by emirs Functional and fiscal (appointed by the king) oversee governorates, decentralization districts, and municipalities. Functional • Smaller GCC countries: municipalities are branches decentralization of the central government. • Quasi-totality of municipal Deconcentration funding is provided by the central government through fiscal transfers. Centralized territorial structure Source: World Bank review of decentralization literature, including fiscal decentralization indexes by Ivanyna and Shah (2012). Note: GCC = Gulf Cooperation Council, comprising Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. Maghreb includes Algeria, Libya, Morocco, and Tunisia. Mashreq includes the Arab Republic of Egypt, Iraq, Jordan, Lebanon, and the Syrian Arab Republic. Among the notable exceptions to these categorizations, the United Arab Emirates has a significant degree of fiscal decentralization. The reliance on central transfers to cover econometric analyses of spatial imbalances in local deficits and finance capital projects the Middle East and North Africa compared makes funding local initiatives and projects a with other world regions. World Bank (2011) pervasive problem. The main exception to presents statistical evidence that countries this in the region is West Bank and Gaza, with centralizing governance characteristics where 90 percent of property taxes (collected tend to follow a more spatially concentrated directly in Gaza or indirectly through development path. the Palestinian National Authority in the In related research, Kremer, Mijiyawa, West Bank) go to municipalities, which in and Whitmore (2012) find a strong negative turn depend on central transfers for only correlation between (a) a political rights percent of their total revenue (World Bank 15 ­ index, in which the Middle East and North 2017c). Maghreb countries, particularly Africa ranks lower than other world regions; Morocco, also tend to have somewhat greater and (b) a spatial agglomeration index (an financial autonomy at the local government indicator of the concentration of economic level—including some extent of commercial activities, which the authors use as a proxy subnational borrowing. measure for spatial imbalances), in which the This highly centralized administrative Middle East and North Africa ranks higher structure has led to uneven spatial develop- than other regions. This research concludes ment. The implication of this centralizing that an increase in political rights and a approach for spatial disparity is confirmed by greater sharing of political accountability C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    165 FIGURE 5.5  The fiscal decentralization of subnational governments in the Middle East and North Africa remains low compared with OECD countries 100 90 SNG spending (% of total government spending) 80 70 60 50 40 30 20 Morocco Egypt, Arab Rep. 10 Iraq West Bank and Gaza Tunisia Jordan Djibouti 0 10 20 30 40 50 60 70 80 90 100 SNG revenue from noncentral sources (% of total SNG revenue) Middle East and North Africa OECD countries Sources: OECD.Stat database 2015 (https://stats.oecd.org/) and 2014–18 World Bank country reports. Note: Figure shows each country’s (a) share of subnational expenditure in total national expenditure, in relation to (b) its share of total subnational revenue from sources other than central transfers. OECD = Organisation for Economic Co-operation and Development. SNG = subnational government. across a territory favors a reduction in spatial one hand, Oates (2005) suggests that this imbalances (World Bank 2011). will improve allocative efficiency by better Of course, not all spatial imbalances are matching public spending to differentiated negative. Indeed, one of the major intellec- needs across the territory and can also tual contributions of World Development increase service quality by fostering competi- R e po r t 2 0 0 9: R e s ha p i ng E c o n o m i c tion among local governments. This is the Geography was the argument that dispers- basis of the subsidiarity principle that gov- ing production more broadly does not foster erns European Union competences, whereby prosperity (World Bank 2009). However, decisions on social and political issues should this uneven spatial development is concern- be made at the most immediate (or local) ing because it contributes to spatial dispari- level that enables their resolution. The logic is ties in living standards, as the next section that lower tiers of government can often bet- discusses. ter identify needs of their citizens and firms and be more adept at designing, delivering, and coordinating than myriad central gov- Decentralization has complex ernment agencies. The theoretical argument implications for spatial disparity in favor of decentralization also centers on its Theoretically, the transfer of decision- potential to improve the quality, transpar- making authority and resources to lower lev- ency, and responsiveness of local governance els of gover n ment has a potent ia l ly by bringing government closer to the people ambiguous effect on spatial disparity. On the (Ivanya and Shah 2012). 1 6 6    C ONVERGEN C E On the other hand, to the extent that effects on spatial disparity for the higher- more prosperous and productive regions are income countries in the region, particularly better equipped to plan and use resources, considering that the public sector in the placing this authority at a lower level could Middle East and North Africa is large by u ndermine the redistributive role of the ­ international standards and has significant state and exacerbate territorial imbalances redistributive capacity. This also implies, (Prud’homme 1995). Decentralization may however, that the positive effects of decen- thus allow inequalities to develop or become tralization on spatial inequalities will be reinforced in the absence of proper account- weaker in the region’s low-income countries ability frameworks and incentives. and when redistributive mechanisms are Internationally, empirical studies suggest poorly designed. The implications also differ that at middle-to-high income levels and in for countries struggling with separatist con- the presence of a sufficiently strong public flict and profound governance challenges, sector, decentralization tends to reduce such as Iraq, Libya, and the Republic of territorial disparities. In a rich review of the ­ Yemen, where administrative decentraliza- relevant literature, Martinez-Vazquez, Lago- tion has potential to reduce conflict and Peñas, and Sacchi (2017) conclude that improve service delivery but should be han- despite the mixed evidence of any direct dled with care to avoid reinforcing separatist impact of decentralization on poverty or tendencies (box 5.2). income inequality, “the evidence is quite With these nuances in mind, this chapter robust” regarding its impact on geographical does not make a specific policy recommenda- and interregional disparities, with decentral- tion regarding administrative or fiscal decen- ization generally contributing to regional tralization. It focuses instead on analyzing convergence and more equal access to ser- the extent to which the region’s top-down vices. Econometric analysis by Rodríguez- decision-making structures and centralized Pose and Ezcurra (2010) indicates that the service delivery mechanisms have contributed net effects depend on economic and political to the spatial disparities documented in chap- circumstances. In particular, the decentral- ters 1–3 of the report. ization of resources and decision making to subnational governments is most likely to Spatial bias 1: Budget processes may reduce territorial inequalities in middle- to result in unpredictable and regressive high-income countries and when the state has subnational resource transfers greater capacity to redistribute financial resources across regions (that is, a larger pub- The budget processes governing sectoral lic sector). However, in low-income countries and subnational resource allocations for pub- where the state has a lower redistributive lic investment in the Middle East and North capacity (a smaller public sector), decentral- Africa commonly lack transparency and pre- ization is found to increase regional inequali- dictability, potentially skewing resource flows ties. Econometric analysis by Lessmann toward areas with greater bargaining power. (2012) also finds that the positive effects of Several econometric studies from the OECD decentralization are conditional on a coun- find that political bias in the territorial alloca- try’s level of development, with a positive tion of central government grants across relationship at higher country income levels. regions is stronger as equalization systems The implications of these findings for the become weaker and less transparent (Pitlik, Middle East and North Africa vary by level Schneider, and Strotmann 2006; Simon- of country income and degree of fragility. Cosano, Lago-Peñas, and Vaquero 2013). Although the Rodríguez-Pose and Ezcurra To safeguard against political bias and (2010) study includes none of the region’s allow local governments greater predictabil- countries in its sample, its findings suggest ity in resource flows, guidance from the that decentralization may have beneficial OECD (2013), the World Bank (Boadway C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    167 BOX 5.2  Handling decentralization in fragile environments In an empirical analysis of the relationship between and state-building process to rebuild state legiti- decentralization and secessionism across 30 countries, macy, help integrate communities, and allow for Brancati (2006) finds that decentralization is gener- elite ­a ccommodations across ethnic groups. This ally beneficial in reducing conflict, by bringing ser- is in line with empirical findings from Alesina and vices closer to the people, but that it can also increase Spolaore (2003), who find there is a potential role for conflict indirectly by reinforcing ethnic and regional decentralization in alleviating secessionist pressures ­ identities. (Martinez-Vazquez, Lago-Peñas, and Sacchi 2017). On one hand, stronger subnational governments On the other hand, poorly implemented decentral- can improve service delivery if the state is absent or ization reform can establish the foundation for seces- otherwise constrained. For example, in West Bank sion. A recent report from the Brookings Institution and Gaza, local government units (LGUs) have suggests that this may be the case in Iraq, where the gained paramount importance in local service provi- recent dramatic push for decentralization of functions sion in the presence of a greatly constrained central to the local level is combined with low local capac- authority, and their role has grown with the increas- ity and poorly designed resource revenue-sharing ing political and geographical fragmentation over arrangements that give too much legitimacy to local the past two decades (World Bank 2016b). In the claims from resource-rich provinces. The authors Republic of Yemen, the success of Marib province explain that decentralization “is only as effective has reignited discussion on the potential of develop- as the political and administrative power-sharing ing a functional decentralized governance system arrangements between the central and subnational based on building local institutions’ capacity through governments in a given country” and should there- a local-first approach (Baron 2018; World Bank fore be combined with other policies targeting spa- 2017a). Commentators on Libya and Syria have also tial inequality at a national level, such as investment suggested the usefulness of shifting power from the in human capital and improving governance and the center to local authorities during the reconstruction ­ business climate (Mills and Alhashemi 2018, 19). and Shah 2007), and UN-Habitat (2009) rec- transfers from the central government to ommend that intergovernmental grant sys- local administrations are commonplace in tems should be timely, transparent, and many high-income and low- to middle- predictable. A lack of clear allocation crite- income countries and have been used ria, on the other hand, can lead to subopti- ­ p rimarily for equalization purposes, as mal spatial and economic outcomes. The performance, or as pass- incentives for good ­ Egyptian system illustrates how weaknesses through payments for a centrally mandated in the subnational budget process can rein- expenditure. International experience and force spatial disparity on several fronts empirical research show that such fiscal (box 5.3) because of “the phenomenal frag- equalization schemes can indeed “deliver mentation in planning and budgeting” that smaller spatial economic disparities across produces “a lack of accountability for the regions” (Henkel, Seidel, and Suedekum outcomes of public spending” and implicitly 2018, 2). favors geographic areas with higher bargain- However, such formulas can also have an ing power (World Bank 2012, 2). uneven spatial incidence that reinforces terri- Even in countries that apply formulas for torial imbalances. For example, in Iraq, cen- the subnational allocation of central govern- tral government budget allocations to ment resources, these are commonly ori- governorates are based primarily upon past ented by default toward leading areas spending, population, and a petrodollar for- (based on an area’s population or natural mula. The resulting allocations are a disin- resource revenues). Subnational resource centive to economizing and skew resources 1 6 8    C ONVERGEN C E BOX 5.3  Spatial bias in Egypt’s subnational fiscal architecture Lack of transparency and predictability in Egypt’s New efforts are under way to reform the country’s territorial budget allocation process reinforces spatial subnational fiscal architecture as part of an ongoing disparity in the following ways: engagement between the World Bank and the govern- ment of Egypt. The local investment planning pro- • The system uses no official formula or consistently cess is undergoing a revision to introduce multiyear applied set of criteria for making resource allocation planning, clarify functional assignments between the decisions, so it serves no clear redistributive function. governorate and district levels, and introduce formula- • All horizontal and vertical resource allocations based allocations by the Ministries of Planning and are determined through individual negotiations Local Development to allocate funds to and between between the Ministry of Finance and each of the governorates and districts in a more fair, transpar- 324 budget entities, a meaning that subnational ent, and predictable manner. With support from the entities with little bargaining power or negotiat- World Bank, the government is piloting these changes ing skills—such as potentially those in lagging even before codifying them through its integrated areas—may suffer lower resource allocations. Economic Development in Lagging Regions Program, • Budgeting takes place throughout the fiscal year which the World Bank is supporting through the under ad hoc circumstances, which prevents the Upper Egypt Local Development Program-for-Results parliament from recognizing and discussing the Project. geographical allocation of budgeting. • Deconcentrated implementing agencies can change a. These budget entities include line ministries and service delivery authorities at the spatial location of any assigned project with the central government level and governorate-level sectoral departments within only the approval of the minister of planning. each governorate. away from lagging governorates that tend to inequalities at the national level rather than be less populated: the correlation between a through redistributive mechanisms that governorate’s share of investment budget and explicitly reward resource-rich regions and poverty rate is −0.125, and the correlation perpetuate resource reliance (Mills and between a governorate’s poverty rate and Alhashemi 2018). its population is −0.15. The use of the petro- The excessively centralized, and some- dollar formula also rewards wealthier times spatially blind, allocation of central governorates—such as Basra and Karkouk, ­ government resources for capital investment which have poverty rates of 14 percent and can directly enhance spatial disparity. The 9 percent, respectively (below the governorate case of Lebanon is illustrative. Overall, public average of 21 percent in 2012 based on the investment in infrastructure in Lebanon has national poverty line)—resulting in signifi- been low and skewed more toward the most cant horizontal disparities in resource alloca- developed districts, even after controlling for tions (World Bank 2016a). population. In per capita terms for 2017, the Indeed, a recent Brookings Institution districts of Batroun, Beirut, Chouf, Jbeil, and report finds that the use of natural resource Marjaayoun received the highest investments endowments as a basis for revenue sharing is made by the Council for Development and widespread in the Middle East and North Reconstruction (a national agency, report- Africa’s fiscal transfer mechanisms and tends ing directly to the prime minister, that is to reinforce resource regionalism and fuel responsible for planning and coordinating separatist tensions (whereby resource-rich donor funding and financing for rebuilding regions demand to retain further benefits infrastructure at the national and local level). from locally generated revenues). The report Among those districts, only Marjaayoun can recommends that governments target spatial be considered a disadvantaged region, and its C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    169 relatively higher investment per capita is pri- keep existing capacity gaps from becoming marily driven by one big sewerage and waste- entrenched. These experiences are contribut- water project of a relatively short duration. ing to improved financial sustainability and On the other hand, there was limited management practices in local governments investment in more peripheral districts such and illustrate how a country’s subnational fis- as Bint Jbeil, Jezzine, and Nabatiyeh in the cal architecture can be designed to become a South; Baalbeck and Rachaya in the Bekaa; powerful tool for improving access to basic and Akkar in the North. The spatial distribu- services territorially. tion of public investment per capita leads to similar levels of disparity. Public investment Spatial bias 2: Human resource capacity therefore does not have a redistributive gaps create uneven quality of nature, and the infrastructure gap between investment planning and execution leading and lagging districts could actually be increasing (Atallah et al. 2018). Several countries display significant On an encouraging note, several of the regional imbalances in terms of local gov- region’s countries have reformed their subna- ernment staffing and human resource capac- tional fiscal transfer frameworks toward the ity that disproportionately affect lagging use of needs- and performance-based formu- areas. This is exemplified by the case of las that also consider the economic and fiscal Tunisia. The explanatory role played by spa- capabilities of individual regions. For exam- tial bias in human resource allocation is ple, Jordan’s system of grants to municipali- confirmed by a 2018 study on public ties is reported to be clear, simple, and employment in the country’s local govern- predictable: Article 23 of the Law on ments, which finds regional imbalances in Municipalities provides guidelines for the dis- staffing and human resource capacity tribution of total government revenue that between northern and coastal municipalities include elements that tend to correct spatial and between interior and rural ones (World disparities, including distance from the center Bank 2018c). Approximately 67 percent of of the municipality to Amman, nominal local governments lacked a qualified engi- gross domestic product (GDP) index, and neer, 75 percent lacked an architect on their municipal spending needs. payrolls, and 24 municipalities lacked a In Morocco, financial transfers to munici- qualified administrative or financial director palities are carried out through a value added within their administration. In total, the tax (VAT) designated account6 with explicit analysis estimated that more than 1,100 equalization objectives that are generally ful- staff would need to be hired to ensure that filled (World Bank 2017b). The impact of each of these key managerial and technical such transfers can be further enhanced by functions is filled by at least one staff mem- making greater use of incentive structures, so ber in all local governments. To address that municipalities are incentivized to build these gaps, the government of Tunisia has their technical capacity, financial sustainabil- begun piloting the use of financial incentives ity, and service delivery performance to to encourage staffing mobility to local gov- access the funds. ernments in lagging areas. Such conditional transfers based on local These capacity gaps translate into uneven government performance have been success- planning and execution among local govern- fully implemented by the Urban Development ments. Indeed, in many countries, the system- and Local Government Program (UDLGP) atic mismatch of planned budget and executed in Tunisia and the Municipal Development budget “tends to nullify the usefulness of bud- Program in West Bank and Gaza. These pro- get planning and the prioritization of expendi- grams are typically paired with a capacity- tures” (World Bank 2016b, 104). In Jordan, building element that helps underperforming few local authorities seem to prepare and reg- local authorities get up to speed and hence ularly update local development and 1 7 0    C O N V E R G E N C E investment plans, with the result that capital (Akkar and Minnieh-Dannieh) find that, expenditure allocations are often made on an compared with the national average, these ad hoc basis without a coherent strategic districts suffer from a small tax base (because vision and only limited citizen engagement or of small populations in each municipality) public participation (World Bank 2018a). In and low revenues compared with the national Morocco, the limited capacity of municipali- average (Atallah et al. 2018). Because of the ties to execute their investment budgets repre- weakness of their independent direct reve- sents a major constraint to local service nues, the municipalities have very limited delivery, with municipalities often executing autonomy from the central government, less than 50 percent of their investment plans because they depend on highly volatile (World Bank 2018b). transfers. Execution capacity also tends to be lower Geographic gaps in municipal revenue in lagging areas: a 2017 internal analysis of a mobilization potential can thereby exacer- survey of local government infrastructure bate spatial disparity in the quality of ser- quality in Tunisia finds that the execution vices. In the case of Lebanon, below-average rate of infrastructure projects was advancing municipal revenues in the lagging Akkar and faster in coastal (leading) areas, with 50 per- Minnieh-Dannieh regions “severely hinders cent of projects from communes’ 2016 annual the capacity of municipalities to provide nec- investment plans completed and another 25 essary public services for which they are percent under way, compared with 40 per- responsible” (Atallah et al. 2018). This is cent and 33 percent respectively in interior consistent with econometric findings from (lagging) areas (World Bank 2017d). multiple low- to middle-income, emerging, Execution of the annual investment plans and high-income economies showing that also tended to be higher in intermediary cities decentralization improves the efficiency of (10,000 –50,000 hectares) than in small public service delivery but only given suffi- towns. cient expenditure decentralization accompa- nied by sufficient revenue decentralization (Sow and Razafimahefa 2015). Spatial bias 3: Weak financial capacity In the Middle East and North Africa, spa- and fiscal autonomy of local tial inequalities in access to basic services governments drive service delivery appear to be particularly high in some of the disparities region’s most fiscally centralized countries, Additionally, the revenue-raising capa- particularly when considering multiple depri- bilities of municipalities are unevenly dis- vations and service quality. A 2016 World tributed, with weaker fiscal autonomy in Bank report on access to opportunities in the smaller municipalities and lagging areas. In region concludes that when opportunity is Jordan, the ability of municipalities to raise defined as access to a bundle of basic services, own-source revenues is directly related to or when the quality of services is taken into their size. The 2 municipalities of Greater account, location of residence explains more Irbid and Greater Zarqa mobilize almost than half the variation in access to opportu- three-quarters (74 percent) of their reve- nity, “with spatial advantages accruing to nues from own sources, whereas the next 8 capital cities and large metropolitan areas municipalities in population size split their that coincide with poles of economic activ- reliance about 50-50, and the remaining ity” (Krishnan et al. 2016, 49). That these 86 municipalities mobilize only 20 percent, differences are particularly high in the heavily relying on intergovernmental trans- Mashreq subregion supports the idea that fers for the rest of their funding (Timofeev excessive centralization may lead to spatially and Wallace 2017). biased outcomes. Likewise, in Lebanon, deep-dive analyses Improving the fiscal autonomy of local of municipal finances in two lagging regions governments can therefore be a powerful C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    171 tool to correct spatial disparities in service by reducing the agency and accountabil- delivery. In West Bank and Gaza, a recent ity of local service providers, particularly household survey shows that, even when in the social services. Specifically, in educa- ruling out differences due to size and loca- tion, decentralization of decision making to tion, “fiscal capacity can be identified as lower administrative units such as governor- one, if not the strongest driver of local gov- ates, school districts, or individual schools ernment performance” in access, quality, has been the topic of reform efforts interna- and reliability of local service delivery tionally. Evidence from the Programme for (World Bank 2017c). Economic models International Student Assessment (PISA) sug- using proxy indicators of fiscal sustainabil- gests that with proper support and account- ity (satisfactory collection efficiency and ability mechanisms, enhancing district own revenue sources as well as a basic sur- and school autonomy in core managerial plus in both operational and enterprise bud- responsibilities can improve student per- get s) y ield posit ive a nd st at ist ic a l ly formance, notably by aligning teacher and significant coefficients. Municipalities with parent incentives. Additionally, schools with satisfactory collection efficiency and own more autonomy over teaching content, stu- revenue sources average a 5.6 higher perfor- dent assessment, and resource allocation tend mance score, and municipalities with a sur- to perform better than those with less auton- plus in operational and enterprise budgets omy (OECD 2011). By contrast, the Middle perform on average 5.4 points higher. East and North Africa’s education systems Per capita revenues are strongly associated are overly centralized. with higher LGU performance in West Bank In the region’s countries that participated in and Gaza, and an even stronger relationship the 2015 PISA, teachers were found to have exists between per capita expenditures and far less decision-making responsibility than LGU performance. Furthermore, the geo- those in OECD countries, and most decisions graphic variation in LGU ­ performance scores are made at the central level, often resulting in measures of “laggi- is closely linked to various ­ a mismatch between school needs and service ness”—household wealth, remoteness from delivery (World Bank 2018d). In some cases, governorate, and poverty—suggesting that such as Egypt, attempts to decentralize deci- improving local capacity to mobilize and man- sion making over education have not been age resources, particularly in lagging areas, accompanied by the necessary resources to can have a significant impact on reducing spa- implement decisions. Elsewhere, efforts have tial disparities, assuming this is feasible given shifted administrative responsibilities to lower socioeconomic conditions in these territories. levels without any agency to effect change in In Saudi Arabia, increasing the own- choosing teachers, tailoring the curriculum, or source revenues of municipalities has become learning approaches to different needs—for a key performance indicator of the recently example, half-day school days in rural areas adopted National Transformation Program or different languages of instruction for refu- (KSA 2016a), reflecting the importance of gee communities, which can be particularly this issue in achieving the kingdom’s vision of relevant for enhancing learning in lagging “providing world-class government services areas or neighborhoods. which effectively and efficiently meet the An exception to this pattern is the United needs of our citizens” (KSA 2016b, 7). Arab Emirates, where there are clearly demarcated roles between the central minis- try and different emirates in delivering educa- Spatial bias 4: Excessive central tional services, with reportedly better government decision-making dilutes outcomes (World Bank 2018d). This is in line local accountability with empirical research analyzing the impact Highly centralized decision making on of decentralization on educational outcomes. service provision impedes service quality Decentralization also has a positive impact 1 7 2    C ONVERGEN C E on health outcomes, although the relation- accountability, can therefore help improve ship is weaker than with education (Martinez- the quality and reliability of service delivery Vazquez, Lago-Peñas, and Sacchi 2017). in lagging areas. International analysis shows The limited autonomy of local govern- that robust mechanisms of social account- ments over service delivery decisions also ability, transparency, and feedback have the undermines local accountability. This exter- potential to improve the quality and equity of nalization of decision making (concentration service delivery as well as to engage citizens of authority in the central government) can in behavioral change that can be critical in distance citizens from local government offi- improving outcomes (Schott 2014). cials, who do not hold the answer to their Similarly, comparative analyses of local problems. As expressed by a local council successes in service delivery in the Middle member interviewed in the Republic of East and North Africa highlight the impor- Yemen, “I exhaust all efforts in serving and tance of autonomy, accountability relation- helping the citizens, but I am helpless when ships, and local participation (Brixi, Lust, there is no response [from the state]” (Brixi, and Woolcock 2015). This is backed up Lust, and Woolcock 2015, 218). empirically by data from a 2017 World Bank To the extent that local governments in Local Government Performance Assessment lagging areas are particularly removed from (LGPA) in the West Bank and Gaza, which the decision-making apparatus of the central found that regardless of an LGU’s size and government, inhabitants of these regions may location or the institutional arrangement for feel less heard. According to a 2012 dataset service provision, the quality and reliability on perceptions of service delivery and gov- of services was higher for local governments ernment performance among urban youth in that performed better on responsiveness and Tunisia, the largest proportions of youth accountability metrics (World Bank 2017c). reporting dissatisfaction with municipal offi- Local governments with the highest shares of cers came from the (lagging) Center West households reporting that their LGU was and the North West subregions, where only “very responsive” to citizen concerns and 26 percent of respondents believed that complaints scored 16.4 points higher on aver- municipalities listen to residents of their dis- age than other local governments on an LGU tricts (World Bank 2014). Percentages were Performance Index (100-point scale) measur- higher in the coastal (leading) subregions ing the quality, reliability, and access to local such as the North East (47 percent). services. This distancing can be detrimental in sev- eral ways: it can lead citizens in less well-off areas to disengage from the state or lower Efforts to move from state-centric their expectations (potentially worsening to citizen-centric approaches vary spatial disparity because of reduced bottom- across the region up pressure from lagging areas) and ulti- This chapter has documented spatial mately to lose trust in the state. On the other biases arising from the excessive centraliza- hand, greater input from local officials— tion of government planning and service regarding both local priority setting and ­ p rovision in the Middle East and North local administration—has the potential to Africa. As they fine-tune the decentralization result in a higher-performing public sector, agendas currently taking shape across the where local administrative units spend their region (box 5.4), the region’s governments resources more efficiently and effectively should carefully consider the spatial implica- and where local officials are better posi- tions of alternative policy options. tioned to deliver responsive public services The discussion in this chapter would to citizens. suggest focusing on introducing more bal- Investing in citizen-centric modes of ser- ance into the vertical flow of resources and vice delivery, with a greater emphasis on local decision making at the national level, C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    173 BOX 5.4  Recent advances in implementing decentralization agendas across the Middle East and North Africa, by region Maghreb subregion Mashreq subregion In Tunisia, 2018 was a particularly eventful year, In Lebanon, the first municipal elections since 2010 witnessing the first local elections since the 2010–11 were held in May 2016. The country has a dual system Jasmine Revolution and the long-awaited adoption of governance consisting of deconcentrated authori- of a local government organic code. These develop- ties at the regional level and decentralized authori- ments fit into the broader decentralization agenda set ties at the local level, represented by more than 1,100 into motion by the Constitution of 2014: the involve- municipalities. ment of different ministries (in committees) to define In Jordan, local political elections were held for key communal competences; the creation of a new the first time in August 2017 for provincial coun- Ministry of Local Affairs in 2016; and the restructur- cils, municipal councils, local councils, mayoralties, ing of the local government fund allocation system, and the Amman Secretariat, which governs the cap- including the introduction of an allocation formula- ital. A Fiscal Decentralization Unit was established based performance transfer. in the Ministry of Finance, with a detailed, phased In Morocco, based on the reforms introduced by fiscal decentralization plan developed for 2017–19. the Constitution of 2011 and subsequent organic These measures include providing revenue equal- laws, the govern ment has lau nched an ambi- ization grants through fiscal transfers from cen- tious reform program aiming to consolidate both tral government fuel tax collections and allocating (a) decentralization through the “advanced region- development budgets to governorates, in line with alization” project (which includes, among other Objective 2 of the Jordan Economic Growth Plan things, the consolidation of the country’s territory 2018–22, which sets forth the government’s vision into fewer regions with expanded competencies to “achieve developmental balance, reduce dispar- and financial resources); and (b) deconcentration ity and distribute development revenues fairly” through a Deconcentration Charter. (EPC 2018, 26). In Algeria, the government is currently develop- In Iraq, although responsibility for basic service ing a national “Vision 2035” policy document, which delivery has historically been almost exclusively in affirms a willingness to deepen the country’s decen- central government ministries, a strong push by the tralization process and the active role of regions and prime minister throughout 2015 has moved devo- territories in the country’s economic development. lution forward, building on the provisions of the Additionally, a local government organic law was Constitution of 2005 and Law 19 (2013), which iden- adopted in 2012 that suggests significant changes tified eight sector ministries for which some functions in administrative structures and public governance, and responsibilities were expected to be transferred although it has since suffered from severe implemen- to the governorates within a two-year time frame. tation challenges. However, this legal and policy framework is yet to be In Egypt, a country with a long history of strong followed through with efficient implementation. central government, a proposed local administra- In Syria, even before the civil war, the country’s tion bill calls for greater fiscal and administra- governance structure was highly centralized, with tive decentralization, seeking the establishment of decision-making power over security, political, bud- separate budgets at the different local administra- getary, and judicial issues in the hands of the gov- tive levels and the expansion of local government ernment. Although decentralization has often been responsibilities for revenue collection and invest- proposed as a solution to the conflict in the ongoing ment planning. The local administration bill seeks peace negotiations, no agreement has been reached. to establish a central government committee to set criteria for central transfers and design a frame- Other economies work for program-based budgeting across different In the Republic of Yemen, an ambitious Local Authority government levels. Law was adopted in 2000 to establish a framework for a box continues next page 1 7 4    C ONVERGEN C E BOX 5.4  Recent advances in implementing decentralization agendas across the Middle East and North Africa, by region (continued) decentralized local government system based on 22 gov- through a 2017 finance law, and creation in 2016 of a ernorates and 333 districts, each with directly elected ministry in charge of implementing a national decen- local councils. Although local councils were estab- tralization action plan. lished to devolve power and encourage local participa- In West Bank and Gaza, as a result of increas- tion in service delivery, the form of decentralization in i n g   p o l i t i c a l a n d g e o g r ap h i c a l f r a g m e n t a - the Republic of Yemen is a mix of deconcentration and tion, LGUs  have gained paramount importance devolution, with national ministries continuing to play a in ­ p roviding services to the local population, major role, and with the strong involvement of nonstate ­ p articularly in areas where the relatively central actors. These hybrid service delivery arrangements are ­ government is politically, geographically, and fiscally complicated by the ongoing civil war. constrained. Even in Djibouti, a small state, a decentralization In the GCC subregion, particularly in Saudi process was initiated in 2006 and has achieved some Arabia, there is recent evidence of government progress, including the holding of regular local elec- priorities moving toward placing greater responsibil- ­ tions, the establishment of local councils, the devolu- ity and resources at subnational levels, in line with tion of some tax collection authority to the local level national objectives of correcting spatial imbalances. combined with greater development of trust in government and in other political human, financial, and technical capacity to institutions, placing administrative functions respond to citizens’ needs and deliver qual- at lower levels of government does not neces- ity services at the local level (particularly in sarily go hand in hand with greater political lagging areas). W hether or not these autonomy or downward accountability at the reforms are couched within a decentraliza- local level. Shifting decision making and tion framework, the focus should be on accountability to subnational government (a) enhancing the state’s ability to better might facilitate greater allocative and produc- connect with the demands of the citizenry tive efficiency, but resources allocated to sub- by delivering services in an accountable national governments should come with and responsive manner, and (b) correcting agreements to ensure that local initiatives sources of spatial bias. improve national welfare along with local wel- fare. Likewise, feedback mechanisms and accountability loops between citizens and Concluding remarks local service providers need to be established Effectively meeting citizens’ demands for so that the service providers are responsive to quality public services requires shifting from the needs of lagging areas and tailored to local the monumental to the incremental: away realities. from top-heavy, state-centric planning and More attention must be paid to support- delivery of services toward placing greater ing effective governance arrangements that agency, capacity, and resources at the local engage citizens at all stages of development level. Enabling subnational governments to planning, investment prioritization, and provide services at the local level is a key part feedback on service delivery performance. of this agenda. Specific recommendations on developing Although Ligthart and van Oudheusden citizen-centric modes of service delivery are (2015) find that decentralization increases discussed in chapter 6. C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    175 Notes Baron, Adam. 2018. “The Marib Paradox: How One Province Succeeds in the Midst of Yemen’s 1. The Arab Barometer public opinion surveys in War.” Policy Brief, European Council on Wave IV (2016–17) were conducted in the fol- Foreign Relations, Berlin. lowing seven Arab economies: Algeria, Egypt, Boadway, Robin, and Anwar Shah, eds. 2007. Jordan, Lebanon, Morocco, Tunisia, and West Intergovernmental Fiscal Transfers: Principles Bank and Gaza. For more information, see the and Practice. Public Sector and Accountability website: https://www.arabbarometer.org/waves​ Series. Washington, DC: World Bank. /arab-barometer-wave-iv/. Brancati, Dawn. 2006. “Decentralization: Fueling 2. The Arab Barometer refers to West Bank and the Fire or Dampening the Flames of Ethnic Gaza as “Palestine.” Conflict and Secessionism?” International 3. Arab Barometer Wave IV (2016–17), Question Organization 60 (Summer 2006): 651–85. 1016, asks respondents to select which of sev- Brixi, Hana, Ellen Lust, and Michael Woolcock. eral statements best describes their household 2015. Trust, Voice, and Incentives: Learning income in terms of how well (or with how from Local Success Stories in Service Delivery much difficulty) their income covers their in the Middle E ast and Nor th Afric a . expenses. Washington, DC: World Bank. 4. “Decentralization” refers to the transfer of El-Din Haseeb, K. 2012. The Future of the Arab political, fiscal, and administrative powers Nation: Challenges and Options. Abingdon, from the central government to lower levels of U.K.: Routledge. government (through downwardly account- EPC (Economic Policy Council, Jordan). 2018. able actors, such as elected local authorities), “Jordan Economic Growth Plan 2018–2022.” whereas “deconcentration” refers to the trans- EPC, Hashemite Kingdom of Jordan, Amman. fer of administrative responsibility for specific Faraji, S. 2016. “Urban Primacy in Urban System functions to lower levels within the central of D e velopi n g C ou nt r ie s: I t s C au s e s government bureaucracy (which are part of and Consequences.” Human Research in the central government and upwardly account- Rehabilitation 6 (1): 34–45. able to it). Henderson, V. 2002. “Urbanisation in Developing 5. The Mashreq subregion comprises Egypt, Countries.” World Bank Research Observer Iraq, Jordan, Lebanon, and the Syrian Arab 17 (1): 89–112. Republic. The Maghreb subregion comprises Hen kel, M arcel, Tobias S eidel, and Jens Algeria, Libya, Morocco, and Tunisia. Suedekum. 2018. “Fiscal Transfers in the 6. The CAS TVA (Compte d’Affectation Spatial Economy.” CESifo Working Paper Spéciale – Part des Collectivités locales dans Series 7012, CESifo Group, Munich, Germany. la taxe sur la Valeur ajoutée) includes 30 per- Ivanyna, M., and A. Shah. 2012. “How Close is cent of national VAT revenue as well as con- Your Government to Its People? Worldwide tributions from line ministries and public Indicators on Localization and Decentralization.” foundations. Policy Research Working Paper 6138, World Bank, Washington, DC. Kremer, Alexander, Abdoul’ganiou Mijiyawa, and References Loic W hitmore. 2012 . “Does M ENA’s Alesina, A., and E. Spolaore. 2003. The Size of Governance Lead to Spatial Agglomeration?” Nations. Cambridge, MA: MIT Press. Middle East Development Journal 4 (2): 1–23. Amin, Khaled. 2016. Unpublished technical paper Krishnan, Nandini, Gabriel Lara Ibarra, Ambar for the Upper Egypt Local Development N a r a y a n , S a i l e s h T i w a r i , a n d Ta r a Program-for-Results Project, World Bank, Vishwanath. 2016. Uneven Odds, Unequal Washington, DC. Outcomes: Inequality of Opportunity in the Atallah, Sami, Daniel Garrote Sánchez, Nizar Middle East and North Africa. Directions in Hassan, Dima Mahdi, and Jana Mourad. Development S eries. Wash i ng ton, DC: 2018. “Introductory Chapter: Analytical World Bank. Framework for Improving Urban Resilience in K SA (K ingdom of Saudi A rabia). 2016a. Lebanon’s Districts Impacted by the Syrian “National Transformation Program: Delivery Refugee Crisis.” Unpublished report, Lebanese Plan, 2018–2020.” Planning document, KSA, Center for Policy Studies, Beirut. Riyadh. 1 7 6    C ONVERGEN C E ———. 2016b. “Vision 2030 Kingdom of Saudi Decentralized Countries: The Case of Spain.” A r a b i a .” C o u n c i l o f E c o n o m i c a n d Publius: The Journal of Federalism 44 (1): Development Affairs of Saudi Arabia, KSA, 135–56. Riyadh. Sow, Moussé, and Ivohasina F. Razafimahefa. Lessmann, C. 2012. “Regional Inequality and 2015. “Fiscal Decentralization and the Decentralization: An Empirical Analysis.” Efficiency of Public Service Delivery.” Working E nv iron m e nt an d Pl anning 4 4 (A): Paper WP/15/59, International Monetary 1363–88. Fund, Washington, DC. Ligthart, J. E., and P. van Oudheusden. 2015. “In Timofeev, Andrey, and Sally Wallace. 2017. Government We Trust: The Role of Fiscal “An  Assessment of the Subnational Fiscal Decentralization.” European Journal of Fra m e work i n Jord a n .” D ra f t Pap er, Political Economy 37: 116–28. International Center for Public Policy, Andrew Martinez-Vazquez, Jorge, Santiago Lago-Peñas, Young School of Policy Studies at Georgia and Agnese Sacchi. 2017. “The Impact of State University, Atlanta, GA. Fiscal Decentralization: A Survey.” Journal of UCLG (United Cities and Local Governments). Economic Surveys 31 (4): 1095–129. 2010. Local Government Finance: The Mills, Robin, and Fatema Alhashemi. 2018. Challenges of the 21st Century. Second Global “Resource Regionalism in the Middle East and Repor t on Decentralization and L ocal North Africa: Rich Lands, Neglected People.” Democracy (GOLD II). Cheltenham, U.K.: Brookings Doha Center Analysis Paper No. Edward Elgar Publishing. 20, Brookings Institution, Washington, DC. UN-Habitat (United Nations Human Settlements Oates, W. 2005. “Toward a Second-Generation Programme). 2009. International Guidelines Theory of Fiscal Federalism.” International on Decentralisation and Access to Basic Tax and Public Finance 12 (4): 349–73. Services for All. Nairobi, Kenya: UN–Habitat. OECD (Organisation for Economic Co-operation World Bank. 2009. World Development Report and Development). 2011. “School Autonomy 20 09: Reshaping Economic Geography. and Accountability: Are They Related to Washington, DC: World Bank. Student Performance?” PISA in Focus Note ———. 2011. Poor Places, Thriving People: 2011/9, OECD Publishing, Paris. How the Middle East and North Africa Can ———. 2013. Fiscal Federalism 2014: Making Rise Above Spatial Disparities . M ENA D ec e n t ra li s a t io n Wo rk . Pa r i s: OE C D Development Report. Washington, DC: World Publishing. Bank. Pitlik, Hans, Friedrich Schneider, and Harald ———. 2012. “Arab Republic of Egypt: Reshaping Strotmann. 2006. “Legislative Malapportion- Egypt’s Economic Geography: Domestic ment and the Politicization of Germany’s Inter- Integration as a Development Platform.” governmental Transfer System.” Public Finance 2  vols. Report No. 71249, World Bank, Review 34 (6): 637–62. Washington, DC.  Pr ud’hom me, R . 1995. “T he Dangers of ———. 2014. “Tunisia Urbanization Review: Decentralization.” World Bank Research Reclaiming the Glory of Carthage.” Report Observer 10 (2): 201–20. of the Pa r t nersh ip for Sustainable Rodríguez-Pose, Andrés, and Roberto Ezcurra. Urbanization, Collaboration for 2010. “Does Decentralization Matter for Development (C4D) platform, World Bank, Regional Disparities? A Cross- Country Washington, DC. Analysis.”  Journal of Economic Geography ———. 2016a. “Republic of Iraq: Decentralization 10 (5): 619–44. and Subnational Service Delivery in Iraq: Status Schott, Berenike. 2014. “Institutionalizing and Way Forward. Working Paper, Report No. Citizen Engagement in Local Governance: AUS17063, World Bank, Washington, DC. Tunisia.” Citizen Engagement in MENA ———. 2016b. “West Bank and Gaza: Public Case Note S eries No. 1, World B a n k , Expenditure Review of the Palestinian Washington, DC. Authority: Towards Enhanced Public Finance Simon-Cosano, Pablo, Santiago Lago-Peñas, and Management and I mproved Fiscal Alberto Vaquero. 2013. “On the Political Sustainability.” Report No. ACS18454, World Determinants of Intergovernmental Grants in Bank, Washington, DC. C e n t r a l i z e d G o v e r n m e n t : C o n t r i b u t o r t o Ec o n o m i c G e o g r a p h y    177 ———. 2017a. “Inclusive Service Delivery: Yemen l’urbanisation.”  [Morocco: Give cities the Policy Note 4.” Working Paper, Report No. financial means to meet the challenges 120534, World Bank, Washington, DC.  of urbanization]. Thematic Note, Report No. ———. 2017b. “Making Urbanization Work for 131221, World Bank, Washington, DC. Growth and Shared Prosperity in Morocco.” ———. 2018c. “Tunisia—Urban Development Urban and Regional Development Policy Note, and Local Governance Program Project: World Bank, Washington, DC. A d d it io n a l F i n a n c i n g .” P ro g r a m - for - ———. 2017c. “The Performance of Palestinian Resu lt s  I n for mat ion Do c u ment (PI D), Local Governments: An Assessment of Service R e p o r t   N o . 1 2 5 7 2 9, Wo r l d B a n k , Delivery Outcomes and Performance Drivers Washington, DC. in the West Bank and Gaza.” Report No. ———. 2018d. “Un leash i ng t he Potential ACS22456, World Bank, Washington, DC. of E duc at ion i n t he M idd le E a s t a nd ———. 2018a. “Jordan—Urban and Municipal North Africa.” MENA education flagship Program for Balanced and Inclusive Growth: draft for decision meeting, World Bank, Concept State Program Information Document Washington, DC. (PID).” Concept Note, Report No. PIDC151796, Yousef, Tarik M. 2004. “Development, Growth World Bank, Washington, DC. and Policy Reform in the Middle East and ———. 2018b. “Maroc: Donner aux villes les Nor th A frica since 1950.” Jour n al of moyens financiers de relever les défis de Economic Perspectives 18 (3): 91–116. Five Steps for Enabling Growth 6 through Thriving Cities and Towns in the Middle East and North Africa T he preceding chapters have shown that also potential to make important progress in the Middle East and North Africa, t h roug h more feasible “sc a f fold i ng” most governments have been oriented reforms—policies that help address barriers toward equalizing endowments across places to wholesale reforms by offering a transi- using hard investment, infrastructure, and tional path to better outcomes. This scaf- investment subsidies. Development strategies, folding may appear to be a second-best particularly targeting lagging areas, empha- approach when the range of binding con- size hard infrastructure. But efforts to equal- straints are not taken into account, but it ize endowments in the absence of reforms may constitute a first-best option when that relax institutional constraints to growth ­ b inding constraints and more complex have resulted in high-cost, low-return devel- ­ policy goals are considered. As certain con- opment strategies. Efforts to enhance endow- straints are relaxed during the transition, ments could enable greater growth and the returns to other reforms may also produce the resources needed to finance more change, meaning that careful thought as to spatially inclusive development, but for infra- the sequencing and coordination of reforms structure and other hard investments to pay is important. off, complementary interventions will be nec- This chapter therefore puts forward essary to address the institutional constraints options for a transitional path to unlock the to growth. potential of cities, people, and regions in the Designing and operationalizing institu- Middle East and North Africa. The region’s tional reforms to this end in the Middle East cities have underperformed as engines of and North Africa is difficult. Political econ- national growth, job creation, and social omy, institutional, social, and financial con- inclusion, but reforms within cities—as well straints make many bold, ideal reforms as connections between cities and the global unfeasible in many contexts in the region; markets able to fuel their growth—can raise pursuing these alone may risk more harmful their performance. isomorphic mimicry1 or simply the continua- The transitional path must also unlock tion of today’s ineffective approaches. people’s potential by delivering globally However, while governments strive ­ competitive skills that grant access to better toward ideal policies and outcomes, there is jobs and places and by directly addressing 179 1 8 0    C ONVERGEN C E welfare disparities. A transition that develops Adopt a framework for effective policies the potential of all subregions requires a more in lagging areas appropriate tailoring of policy to the needs Three dimensions are key to understand- and potentials of each place, supported by ing the success and failure of regional devel- local governments with the capacity to plan opment policies across the Middle East and and deliver reforms in a manner sensitive to North Africa: density, distance, and divi- these local conditions. Taking these agendas sions (World Bank 2009). One of the most together, a key insight is that economic important factors for productive firms and growth, social inclusion, and regional equity workers is access to large, diverse markets. are not mutually exclusive but can be sup- This market access is affected by the combi- ported by the same set of reforms. nation of Transitional Step 1: Adopt new, •  Density, referring to the concentration of evidence-based criteria to guide economic activity and people in the loca- spatial interventions tion (such as a large, dense city); •  Distance, referring to access to external Earlier chapters have highlighted some- markets (either domestic or abroad), such thing of a paradox in spatial development as through transport networks and prox- across the Middle East and North Africa: imity; and There have been heavy public investments, •  D i v i si o n s , refer r i ng to add it ion a l incentives, and regulations to direct firms ­ d ivisions—such as language barriers, to lagging, less-developed regions instead conflicts, or ethnic divisions—that may of leading cities. However, as also docu- fragment markets despite physical con- mented earlier, these have often failed to nectivity or density and hence require an deliver better outcomes for those regions, additional, tailored response. with leading cities retaining stronger pro- ductivity, firm density, public services, This market access cannot be equalized business environments, and wages relative across space. Density arises because agglom- to workers’ skills. eration economies cluster economic activity This is partially explained by the eco- in space—implying, in turn, that other places nomic advantages of leading cities, such as have relatively low density and that the over- agglomeration economies and any natural all economic landscape is “lumpy.” What’s advantages. However, uneven accountability more, the value of proximity increases the across places may also play a role: capital cit- returns to the scale of dense places, creating a ies tend to host the government and main certain momentum to density and conse- power base, encouraging greater sensitivity quently a challenge to reverse it and spread to local needs, while lagging regions often jobs and production more evenly across face weak structures for representation and space. Reflecting this dynamic, as shown in accountability as well as lower local govern- chapter 1, spatial concentration in the Middle ment capacity. As a result, investments in lag- East and North Africa strongly reflects his- ging regions are often highly visible but torical agglomerations. tokenistic, failing to address the key local As for distance, natural advantages (such bottlenecks or opportunities. as coasts and borders) and the costs of This transitional step presents a frame- extending connective infrastructure also work for more effective strategies to support ensure inequality in connectivity to external lagging regions, while Transitional Step 2 markets. And divisions within populations concerns higher-level policies to improve the can result in marginalization of certain accountability and performance of local groups; such groups may require further government. interventions to overcome division. F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    181 Tailor policies according to the local mix of pursuing spatial equity in living standards. density and distance Recent national plans or government pro- Territorial policies should try to raise mar- grams for Jordan, Morocco, Saudi Arabia, ket access in appropriate ways while also and Tunisia specifically highlight the impor- being sensitive to each place’s degree of mar- tance of reducing such imbalances. The ket access and attendant economic and social development strategies and national plans potential. Policies with high returns in a of most of the region’s countries echo the dense and well-connected place may have no, priorities of enhancing welfare, improving or negative, impacts in an isolated, sparse quality and efficiency of service delivery, location. Meanwhile, locally dense but poorly and increasing equality across regions. The connected locations may have different objectives and pillars of the Vision 2030 needs. This basic idea can inform a frame- strategies recently released by Algeria, the work for more evidence-based policy Arab Republic of Egypt, Iraq, Qatar, and making. the United Arab Emirates all emphasize these aspects. Lagging regions with low market access However, sparse, isolated, and otherwise In regions with low market access — lagging areas would benefit from strength- including low local density and weak exter- ened attention and tailored approaches to nal connectivity—policy should prioritize deliver equity in basic services and institu- addressing institutional bottlenecks and the tions. The provision of more basic services quality of basic services. Sparse, isolated and infrastructure (such as sanitation, edu- locations have both weak local markets and cation, and access to rural markets) has weak connectivity to other markets, limiting outsize impacts on quality of life and eco- their productive efficiency and attractiveness nomic dynamism in places where their ini- to investors. Investments in heavy productive tial provision was low—although, even or connective infrastructure often have low here, it must be tailored to demand and returns given low demand from either the affordability (Agénor, Nabli, and Yousef small local population or from external 2005). investors. In addition, although service delivery is a The provision of basic services and core priority everywhere, its challenges and appro- institutional functions are priorities, how- priate solutions manifest differently in differ- ever, being the foundations for both quality ent locations. 2 Remote rural areas may of life and private sector development. require more novel “last mile” solutions to Strengthening the local business environment service delivery (box 6.1), whereas certain and related institutions can remove barriers dense urban areas require highly adaptive to whatever private sector job creation poten- systems to accommodate rapid population tial is present. National institutional func- changes. It is important to disaggregate sta- tions like progressive taxation, transfers, and tistics about delivery and outcomes across social services are also critical to redistribute places to identify pockets of higher depriva- the gains of economic concentration else- tion or stress, set an agenda for regional con- where to these less advantaged areas and vergence, and identify the tailored local ensure universal basic service provision— policies required for convergence. This should which not only contributes to human welfare consider that the biggest challenge for service directly but also equips people with better delivery across the Middle East and North human capital to pursue income opportuni- Africa concerns quality rather than access to ties across space, addressing the challenges of services: frequent blackouts, shortage of med- “stuck” people discussed in chapter 2. ical supplies, ineffective pedagogy, wide- Governments across the Middle East and spread absenteeism of doctors and teachers, North Africa have committed themselves to and so on. 1 8 2    C ONVERGEN C E BOX 6.1  Spatially sensitive “last mile” education provision Sparse rural areas of the Middle East and North coordinated by a common management body, including Africa face challenges of small scale and high trans- a cluster principal. This system helps overcome the high port costs, challenging their ability to deliver basic costs of specialized teachers in rural areas (World Bank services to all citizens, such as quality basic education. 2016a). In the Middle East and North Africa, school Spatially tailored solutions may be required to achieve clustering could address the serious shortage of male sci- universal targets in service delivery, as demonstrated ence and math teachers in rural areas. in international examples of solutions to deliver basic In other isolated areas (such as mountainous areas education in such locations. of Bhutan), boarding becomes increasingly common School clustering leverages economies of scale at higher levels of education, at which schools require between small schools. In the Catalonia region of Spain, a certain scale to teach specialized subjects, and stu- clustered schools share teachers who travel between dents are mature enough to thrive while living away the different rural schools in the cluster. Each cluster is from home. Lagging but dense regions with higher also exposing the location to greater competi- market access tion from outside (box 6.2). Broadly speak- Compared with a sparser area, a lagging ing, sectors that were already competing with region with a large, dense population has imports from other locations, and failing to relatively strong local market access, raising export elsewhere, tend to struggle under this its potential productivity. Given this, it is greater competition and may decline, while important to diagnose the remaining bottle- sectors that were already “exporting” to the necks to stronger growth and job creation. connected locations tend to benefit from this These may include institutional weaknesses improved access and hence grow. A lagging and human capital, as in sparse lagging areas. area with few competitive exporting sectors However, these locations also often lag compared with import-competing sectors can because of their dislocation from external decline absolutely following a connective markets, such as leading domestic cities or investment (Duranton and Venables 2018). global markets that might otherwise offer strong opportunities for trade and growth. If Avoid reliance on high-stakes, place-based so, investments in connectivity to mitigate interventions the effects of distance from markets can be Certain policies place particularly high bets transformative. on the productive potential of a targeted loca- To avoid adverse consequences, however, tion, investing a large amount in fixed immo- connective investments must be tailored to bile capital or tax incentives. These policies local demand and density. It is important to include industrial parks, special economic avoid overinvestment not matched by zones (SEZs), generous regional tax incentives, demand, which not only has low returns but or industrial location directives and land-use also creates stranded assets whose high recur- policies forcing sectors to the target location. rent operations and maintenance costs often In the Middle East and North Africa, govern- fall on the targeted area. Diagnostics must ments have sometimes invested heavily in such also assess whether connectivity is indeed the capital-intensive place-based policies in lag- binding constraint to local productivity. This ging areas that lack the underlying conditions is essential because connectivity operates in for success. This may be motivated by the two directions—improving the target loca- urgency of local need such as persistent local tion’s access to markets to sell products but challenges of unemployment or poverty, F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    183 BOX 6.2  Do cheap land and labor create opportunities for lagging regions? Strategies to promote industry in lagging regions The price of factors that cannot move across space often assume that firms can be attracted by cheap will adjust according to demand. This includes costs local land, labor, and so on. The assumption is that of land, preexisting built structures like buildings local prices will adjust to compensate for weaknesses and infrastructure, and any less-mobile labor (such as in the local economy. However, local prices do not workers who are less educated or who face divisions simply adjust to meet demand, because many factors preventing migration such as language barriers, lack of of importance to firms can move from the lagging social networks across space, cultural norms, and so area, in search of higher returns elsewhere. Low local on). Disadvantaged areas will likely struggle to grow wages, for example, attract firms but also repel work- industries that rely heavily on mobile assets like skilled ers, who may relocate to leading areas where returns workers, but they are more likely to have a competi- to their skills are higher. Finance and machinery can tive edge in sectors that are intensive in more immobile also typically relocate in search of higher prices. Thus, factors—such as agriculture and low-skill activities. rather than factor prices simply adjusting to competi- Relatedly, as documented in World Development tive levels in lagging regions, the supply of factors may Report 2009, as land and unskilled labor prices climb also fall, allowing absolute decline in the local econ- in leading cities, the opportunities for secondary or omy (Duranton and Venables 2018). lagging areas to benefit from spillover growth often Opportunities for lagging and disadvantaged increase as cheap land and unskilled labor become rela- areas remain, however, thanks to immobile factors. tively more important to firms (World Bank 2009). divisions that prevent people from accessing In general, outcomes would be improved opportunities elsewhere (stuck people), by reduced reliance on these high-stakes, regional risks to national stability or security, place-based policies. Such policies can instead and so on. It may also stem from more politi- be mobilized as a final nudge once the tar- cal motivations—such as a squeaky wheel geted location’s underlying competitiveness is ­ syndrome whereby a region of largely histori- secured through more foundational institu- cal importance retains an outsize voice in the tional reforms, local agglomeration econo- national government today. mies, and external connectivity (figure 6.1). Sometimes strategies to spread jobs and The components of figure 6.1 should be growth to lagging regions are criticized as understood as building blocks . Even in a inefficient because of their relatively small place with a high agglomeration of firms and direct impacts. However, appraisals should people, capital-intensive place-based indus- not overlook the indirect effects of such poli- trial policies (as described above) are likely to cies on outcomes of concern (such as unem- fail without the foundations of strong institu- ployment, poverty, stability, and security). tions, adequate human capital, and strong However, even considering indirect connectivity to markets for their products. impacts, these investments have rarely met Likewise, if connectivity to a large market is their objectives in locations that lack strong delivered to a place that lacks quality institu- market access in the form of both major local tions and human capital, then, as discussed agglomerations (proximity to dynamic cities) above, the outcome may be increased impor- and strong connectivity to international tation of goods and services without parallel export markets. Similarly, impacts have been exports and hence the gradual loss of people poor when hard place-based investments are and firms to the larger market. Effective poli- not complemented by the building blocks cies for the development of lagging areas are mentioned earlier: institutions, basic infra- bot h mu lt id i mensiona l a nd properly structure and services, and human capital. sequenced. 1 8 4    C ONVERGEN C E FIGURE 6.1  Framework for effective spatial policy in the Middle East and North Africa, from foundations to final steps • More spatially distortive Place-based • Needs more industrial market access • Industrial parks and SEZs, policy industrial location regulations, • Final steps place-based subsidies and incentives investments Connective • Physical infrastructure for connectivity (ports, roads and road improvements, rail, logistics centers, and so on) People-based • Skill development, worker training, and wage subsidies policy • Provision of basic public services (water, sanitation) • Less spatially distortive • Basic human capital improvements (broad-based education, health) • Needs less based policy market access Institution- • Policies to reduce frictions to factor mobility across places • Foundations • Broad-based governance and institutional reforms (trade policy, competition, factor markets, Doing Business indicators, and so on) Source: Adapted from World Bank 2009. Note: SEZs = special economic zones. Improve the outcomes of place-based investment despite heavy public investments in policies some dimensions of competitiveness. Chapter 4 discussed this problem, documenting The above framework recommends a particularly low returns to infrastructure ­ sequential approach to support lagging areas, investments in the Middle East and North differentiated by the market access and pro- Africa, likely because infrastructure is planned ductive potential of each place. Three principles without important complements that address can help to identify strategies that avoid the the remaining weakest links. common pitfalls of place-based policies and The importance of complementarities and successfully tap the available local potential: the risk of weakest links often leads policy (a) account for complements, (b) coordinate makers to consider big-push efforts for lag- across places, and (c) focus on market failures. ging areas, investing heavily across multiple areas to ensure that every possible comple- Account for complements ment is accounted for. However, successful When investment is not flowing into a big pushes were not indiscriminate, place- region, it is typically the result of not one but blind programs to spark investment by fiat. several disadvantages. The need for multiple Instead, they were often oriented around a factors to be in place in a location can create a big local opportunity and addressed pre- weakest-link problem, where the absence of cise bottlenecks preventing its realization. just one critical factor (such as security, access For instance, arguably the most celebrated to external markets, or the ability to obtain big push—the Tennessee Valley Authority land or a business license) prevents private (TVA) programs in the United States after the F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    185 Great Depression—focused on exploiting the 10 percent of local income for several years, potential for cheap local hydropower due to while the greatest poverty reduction in the powerful local rivers and booming national region was in fact attributable to a more industrialization as well as on managing atten- natural process of urbanization (Duranton dant risks and constraints (flooding, barriers and Venables 2018). The case of the Tanger to navigation of the rivers, and environmental Mediterranean Port (Tanger-Med) in Morocco degradation of farmland). The wider pack- tells a similar story, where a big push strategi- age of complements came in behind this more cally targeted a big opportunity—modernizing strategic, focused push to heighten impacts. the port to connect Tangier to international Big pushes can also demand huge spend- markets—and complementary investments ing levels to see returns; even granted posi- addressed the remaining bottlenecks revealed tive effects, their cost-effectiveness is often by consultations with the local business com- questionable. The TVA cost approximately munity (box 6.3). BOX 6.3  Big bottleneck or big opportunity: Targeted place-based policies in Afghanistan and Morocco One big bottleneck in Afghanistan: Security local and regional economy (including strengthening Afghanistan has considered constructing special tourism), and increased production in manufacturing-­ economic zones (SEZs) in former military air bases. intensive sectors such as the automotive industry This fragile and conflict-affected state enjoys good (Ducruet, Mohamed-Chérif, and Cherfaoui 2011). access to power and a well-connected location, but of Given the increase of employment, a new city between course, security and gaps in infrastructure have been Tangier and Tétouan is being developed, with invest- major concerns. The military bases had been particu- ment committed for housing developments. larly well serviced with infrastructure and could also Such success is rare among regional development provide a more secure environment for firms. In this efforts, so what made the difference in Tangier? way, the SEZ construction targets a precise remaining • The project centered on a big untapped oppor- major bottleneck. tunity: access to international markets. • Closely informed by the private sector, the project One big opportunity in Morocco: Tanger-Med Port was effectively coordinated between the relevant Nestled on the Strait of Gibraltar, Tangier occupies a public and private stakeholders. A strong public-­ strategic location for international trade and valuable private partnership known as the Tanger-Med shipping routes, but it had been neglected for many Special Agency (TMSA) instilled collaboration years and was relatively peripheral to the Moroccan across private and public actors throughout the economy by the late 1990s. development of the strategy and coordinated the Aiming to support stronger development in northern actions of various government agencies. Further, Morocco, the government undertook a range of national regional government actors built their own com- reforms and local development efforts to capitalize on munication and feedback loops between private Tangier’s potential, centered on opening Tangier to inter- and public actors throughout the life cycle of national trade while developing local productive capaci- implementation. ties. Central to this was the development of the Tanger • The resulting strategy was multidimensional , Mediterranean Port (Tanger-Med) in 2002 at a cost of addressing not only the port but also the essential €1 billion (Ducruet, Mohamed-Chérif, and Cherfaoui complements for its success. These included national 2011), complemented by national reforms including reforms promoting liberalized trade, investment, and free trade agreements, open skies agreements for airline migration; national transport infrastructure for roads travel, and relaxed investment and visa regimes. and rail; local free trade zones for industry and logis- Between 2005 and 2012, Tangier created new jobs tics; training and education of the local workforce; three times as fast as Morocco as a whole. The manu- city regeneration through activities such as improved facturing sector enjoyed 28 percent employment growth water supply and waste management, protection from 2002 to 2004 and an annual increase of invest- of green spaces, cultural heritage, and beaches; and ment of 13.2 percent. The strategy has also supported a modernization of the old port for cruise ships. deepening of backward linkages, diversification of the Sources: Bismil 2018; Kulenovic 2015. 1 8 6    C ONVERGEN C E The examples of the TVA and Tanger- policy. For a local leader, a policy is effective Med highlight the positive side of the weakest-​ if it brings jobs and investment to that lead- link problem: in a lagging area with major er’s own locality at reasonable costs to the latent advantages, an intervention that diag- local government. However, despite these noses and closely targets the remaining bot- positive local impacts, a policy or investment tlenecks can unlock disproportionately large may worsen outcomes for another competing benefits. Overall, the importance of comple- place. Consider, for instance, offering tax mentarities often entails high discontinuities incentives that attract firms from one region in the impacts of place-based policies, with to another: this benefits the implementing the chance of no returns to heavy invest- locality but harms the city from which firms ments when weakest links are not addressed and people leave. And if the implementing but also the opportunity for transformation locality is a less efficient place for those firms with the right intervention targeted at to produce, it also harms overall job creation the  remaining big bot tleneck and and growth nationally. opportunity. Without coordination across local govern- One low-risk, high-return approach to the ments, their individual efforts at local eco- delivery of complements is to better coordi- nomic development can become a race to the nate existing activities. Different line minis- bottom—a costly fiscal competition in spend- tries and national, regional, and local ing and taxes to attract firms to their locations governments should coordinate their activi- that entails high waste on a national level. An ties locally to better exploit complementari- uncoordinated strategy where each locality ties between investments (box 6.4). For tries to attract the most and best firms can also example, if an industrial zone is planned, it impede the development of agglomerations by should be easily accessible from good high- scattering firms’ incentives across space. Thus, ways and well serviced by power by the time overall, when place-based policies are com- of completion, and so industrial zone, power, bined, they often lead to a national framework and highway plans should be coordinated. with considerable wastage and inconsistency Addressing local human capital deficits may as well as missed opportunities to work require coordination on school reform, hous- together to develop scale and specialization ing and social support, internet connectivity, across places. There is a need to reconcile agen- and so on. das across subnational units, and between sub- national and national governments, so that all Ensure horizontal coordination pull in the same direction in the country’s best Coordination across localities is also interests. Transitional Step 2 discusses practical essential to optimize returns to place-based options for such coordination. BOX 6.4 Industrial zones in Egypt: Suffering a lack of density and complements Egypt’s industrial zones have often provided sub- surrounding Greater Cairo were also inadequately sidized land to industry, but many lack the comple- planned, meaning that routes from Cairo to the towns ments needed for productivity. Several of Egypt’s are seriously congested. The industrial zones have also industrial zones were not located on major roads often lacked adequate infrastructure to complement the to connect them to local markets or ports, which cheap land, because of coordination failures between implies that zones will either remain poorly accessible agencies. Finally, investment facilitation services— or entail high costs to retroactively connect them to standard in industrial zones worldwide and important markets. to their success—are limited in Egypt’s zones. Some of the roads between Cairo and indus- trial zones in the “new towns” built in the deserts Sources: Sims 2015; World Bank 2016b. F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    187 Address market failures •  Coordination problems impeding the pri- Strategies for local economic development vate provision of shared infrastructure should focus on addressing market failures, •  Negative externalities such as congestion simply because private firms will indepen- and pollution in major cities dently pursue a productive opportunity •  Incomplete information markets, pre- unless market failures are present. Many venting investor or household knowledge types of market failures can reduce local about local potential. investment, including the following: The first mover problem implies that, to •  “First mover” problems preventing the attract firms to lagging areas, governments creation of new agglomerations have a role in coordinating expectations •  Gaps in factor markets such as problems about where new firm clusters will develop in property transfer, credit, and business (box 6.5). The importance of agglomeration registration economies means that many firms would not BOX 6.5  How the dynamics of large investors can justify government intervention Large investors in economic theory largest apparel company. The government worked closely Private investments that are large relative to the local with PVH to identify areas of mutual interest and deliv- economy can drive up local factor prices (such as the costs ered them to create an attractive environment for PVH in of land and immobile labor). This benefits local owners the new Hawassa Industrial Park (HIP). They included of those factors (local landowners and immobile work- • Demonstrating government commitment by ers who enjoy higher wages), but it reduces the investor’s raising capital, building facilities fast, and deliv- own profits. Because some benefits of the investment are ering world-class factories with high environ- social and do not benefit the investor, some such invest- mental standards as agreed; ments can be forgone despite being socially beneficial. • Delivering strong access to international mar- In these circumstances, final-straw improvements kets (particularly important for the fashion that nudge the large investor just over the private industry), including through the enterprise bar- profit threshold can have large social impacts: they gaining agreement (EBA), the 10-year exten- unlock the previously forgone net social benefit as sion of the African Growth and Opportunity well as the newly available private marginal benefit to Act (AGOA), streamlined customs procedures, the investor. These impacts are demonstrated econo- extension of the new Djibouti–Addis Ababa rail metrically by Duranton and Venables (2018). line to Hawassa,a and improvement of the Addis This potential creates a role for government to Ababa–Kenya highway on which Hawassa lies; attract marginal large investors by capturing and • Exploiting abundant hydropower to undercut returning a portion of the social benefits to the regional competitors in electricity price and reli- investor. An obvious case is a subsidy or tax incen- ability (electricity being a major cost in the gar- tive, though other mechanisms could involve local ment industry); infrastructure or skills training relevant to the firm. • Complementing the large pool of local labor with However, correctly calibrating these mechanisms is employee selection and training programs run challenging, and governments must always be wary of jointly by PVH, the government, and donors; and the risks of simply subsidizing uncompetitive, unsus- • Offering tax holidays, described as “icing on the tainable industries, or as above, engaging in a destruc- cake” rather than a key element. tive race to the bottom across localities. Thanks to PVH’s investment, 18 foreign and 5 Large investors in practice: Hawassa Industrial domestic supplier companies have already committed Park, Ethiopia to follow it to HIP, and more are expected. The park was inaugurated in July 2016 and is planned to create The example of Hawassa Industrial Park (HIP) in Ethiopia 60,000 direct jobs on US$1 billion in export sales. illustrates how a government might overcome first mover problems by attracting a catalytic large investor—in this Source: Adapted from Duranton and Venables 2018. case, Phillips-van Heusen (PVH), the world’s second a. Hawassa is also referred to as Awasa or Awassa. 1 8 8    C ONVERGEN C E move to an underdeveloped lagging area Transitional Step 2: Devolve without assurance that others would join greater functional authority and them. This creates a first mover problem, resources for local revenue which can cement existing agglomerations generation and service provision even where new agglomerations may be more to local governments efficient. Governments can help assure firms and households that others will move, such Chapter 5 argued that to improve out- as by coordinating investors, addressing local comes across the Middle East and North bottlenecks, and publicizing these efforts; Africa’s regions—particularly its lagging updating local plans and branding; attracting regions—there is a need to move “from the catalytic large private investors around whom monumental to the incremental—away from other firms may coordinate; and so on. top-heavy, state-centric models and toward Smaller shifts in agglomerations are easier more locally grounded, people-centric to support than large ones. For example, approaches” through the strengthening and Egypt’s new towns performed far better empowerment of local government. when close to Cairo than in brand new, iso- lated desert locations. Another example Strengthen the capacity and would be encouraging the consolidation of effectiveness of local governance an existing industrial cluster on better land within the same city or in a nearby satellite This speaks to the importance of regional town. development policy that is sensitive to local Tax incentives and other subsidies are needs. However, this local sensitivity is often often used to attract investment to priority in tension with two other essential qualities locations, citing first mover justifications of effective local governance: scale economies among others. However, Gaubert (2018) and capacity, on the one hand, and horizon- and others have shown that tax incentives tal coordination across locations on the other are typically low priorities for firms decid- (Bahl 2008). ing where to locate, compared with factors Sensitivity to local needs. Given the more fundamental to productivity like importance of sensitivity to local needs, the market access and trade facilitation; basic principle of subsidiarity asserts that govern- infrastructure; and reliable, strong institu- ment functions should be decentralized to the tions. This means, in turn, that the exemp- most local level available unless there is a tions and subsidies needed to attract firms particular argument for their allocation to are often very high. In addition, firms that higher levels. Whereas planning has often chose a location because of tax incentives, been disconnected from realities on the rather than its broader suitability, tend to ground in the Middle East and North Africa, be less complementary to the local econ- taking better advantage of local governments’ omy and create weak local spillovers proximity to the population can improve (Gaubert 2018). information about local needs and strengthen If used at all—in line with the earlier planning through better channels of account- framework—evidence-based spatial-policy ability and feedback. Local populations and tax incentives or subsidies should be a final circumstances are also generally more homo- nudge to bring investors to an already com- geneous, enabling more tailored policies. petitive location. Tax incentives also need to Smaller local governments may also avoid the be coordinated across localities given the risk diseconomies of scale associated with the of “beggar thy neighbor” tax competition larger and more complex bureaucracies of and footloose firms that drive down the many central governments, such as blocked margin of social benefits by shopping for the channels of communication, a lack of agility, best incentive package from each subnational onerous procedures, and slow decision mak- government. ing and responsiveness. These factors give F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    189 reason to move from highly centralized gov- Others are negative—such as pollution, con- ernance in many regional contexts. gestion, or a race to the bottom in local tax Scale and capacity. However, more-local competition. In either case, fragmented local authorities tend to suffer capacity and effi- authorities without dedicated coordination ciency challenges due to their small scale. frameworks often fail to control negative This is particularly so in the Middle East and spillovers or jointly develop shared assets and North Africa, where many local authorities may engage in wasteful “beggar thy neigh- have been subdivided (resulting in very low bor” competition. scale) and suffer from historic underresourc- As argued in chapter 5, in contrast to the ing and weak institutional frameworks. principle of subsidiarity, the default in many Small-scale authorities have less capacity to Middle East and North Africa governments hire skilled and specialized staff (as docu- (particularly in the Gulf Cooperation Council mented in chapter 5) or to invest in thorough [GCC] and Mashreq subregions) is the central- research and planning processes. In addition, ization of political, administrative, and finan- certain public functions (such as major hospi- cial powers and structures—leading to tals, universities, railways, or industrial planning that lacks adequate sensitivity to local zones) are only efficient when undertaken at needs and conditions. This has been problem- scale (given a need for specialized inputs or atic for efforts aimed at local economic devel- large capital investments and other fixed opment, for example, which requires sensitivity costs), making provision by small local units to local conditions to address the key local inefficient and infeasible. bottlenecks, as noted earlier. Coordination of external effects. Small Decentralization, however, needs to be sen- local authorities may also fail to properly sitive to how scale economies and coordina- manage the external effects of their actions, tion challenges affect local governments and to the detriment of wider regional and the functions they must perform. Effective national development. Investments, policies, decentralization is often therefore both incre- economic activities, and so on in one area mental and asymmetric, tailored to the vary- often have external impacts on other areas. ing capacities and needs of different localities. Some are positive—such as when a district Decentralization efforts also need to improve that hosts a vibrant city benefits surrounding scale and capacity, coordination, and local districts as a source of jobs and services. accountability structures (box 6.6). BOX 6.6 Instruments to improve scale and coordination among local governments Vertical coordination: France’s State-Region CPERs are coordination tools: they do not make Plan Contracts new resources available; rather, preexisting resources (from the region and the state) are aligned around the In France, binding contracts between regions and the priorities set. The central government delineates broad central government promote effective region-led plan- categories of projects that are eligible for CPERs, ning while aligning efforts of the local and central within which regional governments propose their government. State-Region Plan Contracts (CPERs) own priority projects, increasing the coherence and have been implemented in France since the mid-1980s. complementarity of regional plans. The national and These are five- to seven-year agreements between the regional governments legally engage jointly to deliver national government and the regions (the highest level the projects, agreeing on their financing, delivery of local government) to deliver the region’s highest- modalities and agencies, beneficiaries, selection crite- priority development projects. ria, and results indicators. CPERs can also coordinate box continues next page 1 9 0    C ONVERGEN C E BOX 6.6 Instruments to improve scale and coordination among local governments (continued) actors beyond the regional and central governments— irrigation, river-based trade facilitation, and economic such as sectoral line ministries, national transport development initiatives; and a final group may consist agencies, cross-regional bodies, and subregional bod- of municipalities within a city as well as a wider catch- ies (such as metropolitan areas or nature areas)—that ment area (such as the commuting span) not within the often cosign CPERs. city boundaries. Because spillovers and coordination Contracts between states and regions give the needs do not always align with administrative bound- regions considerable responsibility to deliver on the aries and vary with the geography of each function, agreed-upon goals: the regions must contribute a large local governments in the Middle East and North Africa part of plans’ financing, and if implementation of any may benefit from instruments for flexible and function- contracted project is not begun within 18 months, it is specific collaboration or coordination. cancelled automatically. An example is the French public institution of Overall, CPERs promote strong bottom-up plan- intercommunal cooperation (Établissement Public ning sensitive to local needs while aligning efforts and de Coopération Intercommunale, EPCI)—a flexible resources among national and regional actors. Other instrument through which municipalities can collabo- countries have since adopted similar systems inspired rate to perform either a single narrow function (such by this model, such as Colombia’s contratos plans. An as local water supply) or broader functions (such as innovation of the contratos plans is that several local city management or, for a more rural area, comple- authorities can cooperate to sign the same state-region mentary environmental, tourism, training, and social contract. and economic development activities). Finland’s municipalities play the major role in local Horizontal coordination for broad local administration in Finland but suffer from very low planning and management: England’s scale, which motivated a 2005 reform—the Project combined authorities to Restructure Local Government and Services In England, groups of local authorities can submit pro- (PARAS)—to raise the scale of local governance. For posals to form joint combined authorities. These exist example, municipalities are responsible for the design, above the level of the local authorities and are often provision, and oversight of all health and social care.a formed to improve the management of major cities and Yet in 2008, the average municipality had just 4,700 their surrounding metro and rural areas, which other- inhabitants. Demographic and migration pressures wise spill across administrative boundaries. Combined added to the pressures on municipal services and authorities conduct functions benefiting from shared finances, with several municipalities facing declin- planning across the region, such as regional economic ing working-age populations alongside an increased planning, overseeing regional transit services, and number of elderly requiring care. The PARAS reform investing in innovation hubs or industrial parks. Based encouraged municipalities to merge and cooperate to on their needs and performance, combined authorities achieve greater scale economies to meet these mount- negotiate for incrementally greater decentralized pow- ing challenges. Mergers were voluntary but encour- ers and responsibilities, through a sustained dialogue aged through large financial incentives. with the central government. Through these voluntary mergers, the number of municipalities fell from 444 in 2005 to 320 by 2015. Horizontal coordination for narrow functions: In addition, it is compulsory for municipalities to form France’s EPCI and Finland’s municipal joint bodies to finance and manage certain functions requiring major scale economies—including universi- coordination ties, central hospitals, disabled people’s services, and The different functions local governments must per- regional councils. Cooperation on other aspects of form often have geographies, not aligned with admin- service provision is optional. Overall, joint municipal istrative boundaries, such as those of a district, region, bodies have been popular, with each municipality a or city. For example, a particular group of scattered member of approximately 5–10 intermunicipal bodies. municipalities hosting heritage sites may benefit from collaboration for preservation and tourism promo- Source: World Bank 2019. tion; a different group along a river basin may benefit a. “The Kainuu Regional Experiment,” Ageing Studio (2010) Dossier, Helsinki from collaboration for environmental management, Design Lab archive, http://helsinkidesignlab.org/dossiers/ageing/kainuu.html. F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    191 Regarding capacity, local authorities must with the domestic political economy. The be properly resourced to carry out their link between local governments and their decentralized functions. This includes own- residents and businesses is not always func- source revenues, and central-to-local trans- tional. Many of the region’s governments fers. Relevant areas for fiscal decentralization have been pursuing relevant reforms, include increasing local governments’ control such  as the adoption of transparent and over fees to cover service costs and decentral- p erformance-based recruitment and pay ­ izing land and property tax regimes and systems. The process of building account- administration. Central-to-local government ability is complex, however, and transi- transfers also require greater transparency, tional measures may be required to improve predictability, and more equitable formulas, the quality of local service delivery while it particularly in the Mashreq subregion. As is developed. argued in chapter 5, transparency and pre- An alternative source of accountability is dictability are important to counteract spa- between residents and private sector organi- tial biases in spending and outcomes, while zations. In a competitive market, clients can in many cases, formulas also need revision to directly hold private providers accountable reduce the bias toward resource-rich loca- through their purchasing choices. Private tions and better target poverty in lagging providers may also address gaps in capacity regions. and financing. However, the private sector However, financial resources alone are requires proper government oversight and insufficient. Many local governments fail to regulation, as illustrated in two examples spend and execute the limited budgets they from the Republic of Yemen and Kenya possess. Any reform should clarify the divi- (box 6.7). sion of responsibilities for each level of gov- Community management and monitoring ernment and otherwise coordinate the local can also hold frontline delivery agents and national activities affecting each directly to account. For example, the School region. and District Development Program in Jordan Overall, expanding the scale of local gov- has supported community-managed schools ernance may be important in the Middle East through appointment of a domain leader who and North Africa, where authorities have uses input from stakeholders (parents, stu- been highly fragmented. This can be achieved dents, and teachers) to design the school through consolidation (reducing the number improvement plan and submit it to the local of local governments, as by incentivizing education council. The program has resulted local governments to merge) as well as in better learning outcomes (Brixi, Lust, and through coordination between authorities. Woolcock 2015), as have school monitoring The latter might involve incentivizing local programs in Egypt and Morocco (Beddies authorities to form joint bodies, either for et al. 2011). larger tasks such as metropolitan manage- Citizen participation can also improve ment and regional planning or for narrower local government performance through both service-delivery tasks (for example, Finland’s accountability and information channels. requirement that municipalities cooperate to Incorporating the opinions of residents and jointly deliver services subject to large-scale businesses may take time and require extra economies). resources up front but can help ensure that service provision has the desired impacts. Local businesses, for example, usually know Ensure local accountability for service very well what infrastructural bottlenecks delivery are hindering their daily operations, such as Service delivery challenges in the Middle lack of road connectivity or stable electricity East and North Africa’s local governments supply (Australian Embassy and Asia are not purely technical but intertwined Foundation 2016). 1 9 2    C ONVERGEN C E BOX 6.7  Government-regulated private sector service delivery in the Republic of Yemen and Kenya A partnership for water delivery in the economies, brings down costs, and serves as an entry Republic of Yemen point to more sustainable use of water resources while also developing more effective regulation of water In the Republic of Yemen, a considerable water supply quality (World Bank 2017a, 2017b). gap is being filled by private wells and water tanker trucks. Typically, urban or peri-urban wells sell water to tankers, which then deliver the water to urban A franchise model for clinics in Kenya households. However, this arrangement raises serious In Kenya, the Child and Family Wellness Clinic has questions about water quality, pricing, and sustain- implemented a franchise model since 1997. The model able water resource management. provides clinic and pharmacy franchises with business The World Bank has implemented a pilot pro- incentives to both comply with regulations and pro- gram to create a private partnership with water vide high-quality, affordable health services (World tankers in Sana’a. In this program, private tanker Bank 2017c). Private clinics or pharmacies in the net- trucks sign a contract with local water corporations work have access to a cost-effective drug and medical and deliver water from wellfields managed by local equipment supply chain as well as incentives to retain water corporations to customers at affordable prices. customers, while regulations require all participating This approach helps increase supply and raise scale clinics to follow standardized medical procedures. Transitional Step 3: Step away for education that confers those skills. from credentialist education and Raising demand for education may toward schooling that cultivates require policy makers to address barri- globally tradable skills ers to private sector job creation, such as by reducing the dominance of public Citizens of the Middle East and North employment and addressing the disparity Africa will benefit most from trade integration in wages and conditions between the pub- if they possess the skills to compete for and lic and private sectors. In addition, exam secure high-quality jobs in a global market- reform to ensure that test scores better place. Human capital is also critical for the reflect skills can make paper credentials a knowledge spillovers responsible for much of better signal of real skills for employers. the dynamism and productivity of cities as •  Discipline and inquir y: Curriculum well as to empower people in sparse and lag- reform and teacher training are needed ging areas to migrate to opportunities. to raise students’ engagement (and reduce However, as documented in chapter 2, the passivity) as well as to ensure that memo- region’s citizens underperform in the techni- rization of facts (dominant in many sys- cal and problem-solving skills required to be tems) is properly complemented by critical competitive in global labor markets. A World thinking and other skills. Bank flagship report on education in the •  Tradition and modernity: In certain Middle East has diagnosed four underlying Middle East and North Africa coun- tensions that need to be addressed to improve tries (such as Jordan and Kuwait), such outcomes in the region on the scale required curriculum reforms have received deep (El-Kogali and Krafft 2019): criticism—for reflecting imported val- ues and undermining national character •  Credentials and skills: Greater employer or religious identity. Reforms should not demand for skills can raise students’, par- import foreign systems but rather assess ents’, and other stakeholders’ demand the effectiveness of current teaching in F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    193 conferring the desired opportunities to to address the region’s low preprimary enroll- the next generation and incrementally ment (30 percent gross, on average). The adjust accordingly. United Arab Emirates has demonstrated one •  Control and autonomy: A greater decen- reform path, targeting universal enrollment tralization of authority to schools and in its UAE Vision 2021, and is now on track teachers can potentially increase local to achieve 95 percent enrollment by 2021, ownership while also addressing chal- complemented by investments in quality lenges of low innovation, adaptation, and improvements. motivation in schools. Global evidence F o c u s i ng o n fo u n d a t i o n a l sk i l l s . shows that assigning greater local author- Standardized tests for literacy (such as the ity to schools and teachers tends to raise Early Grade Reading Assessment, or EGRA) outcomes but must be complemented by and mathematics (such as the Trends in adequate local capacity, resources, and International Mathematics and Science accountability, reflecting the broader les- Study, or TIMSS) show that students in the son of Transitional Step 2. Middle East and North Africa tend to greatly underperform their global counterparts on Building out from these underlying ten- core skills. An urgent focus on these founda- sions, the report offers specific practical rec- tional skills is needed in the first three years ommendations: (a) closing the spatial and of school. demographic gaps in educational access and This agenda would be supported by align- outcomes; (b) targeting preprimary enroll- ing preprimary and primary education to ment; (c) focusing on foundational learning ensure that children arrive in primary school skills; (d) improving teacher performance; ready to learn (also done in the United Arab and (e) engaging the private sector to increase Emirates) and to support students who are work-relevant skills. transitioning from local languages and dia- Closing gaps in educational access and lects to Modern Standard Arabic. To sustain outcomes. Human capital outcomes—and learning at higher levels, where instruction of their returns on the labor market—are not certain subjects may be in English or French, equal across regions, neighborhoods, classes, many students will also need to develop a and genders. A focus is needed on diagnosing stronger foundation in these languages at the these disparities and tailoring policy to primary level. address them. Subnational variations in edu- I m p r o v i n g t e a c he r p e r fo r m a n c e . cational inputs and human capital outcomes Shortcomings in teacher skills and pedagogy should be systematically monitored and used can be addressed through large strategic to inform targeted investments for regional investments in professional development and convergence, thus helping to ensure that by introducing greater sensitivity to pedagog- regardless of where people are born, they can ical skills (over credentials) in teacher recruit- access productive opportunities and quality ment and promotion. Professional of life. development should raise teachers’ focus on The Middle East and North Africa also inquiry and skills over discipline and exami- displays above-average disparity in educa- nations (to complement curricula reforms) tional outcomes between richer and poorer and address teachers’ own skill gaps, such as students as well as high gender disparity. foreign language skills. (Girls far outperform boys in many of the In addition, teacher motivation and region’s countries, but men nonetheless enjoy accountability suffer from the centralization better returns to education in terms of access of hiring, firing, and promotion decisions; to job opportunities.)3 decentralizing autonomy over these decisions Targe ti ng p re p r i m ar y e n rollm e nt . to schools may result in better incentives for Preprimary education is key to prepare stu- performance and effort (through recognition dents to learn in primary school. A priority is and promotions) and better ability to manage 1 9 4    C ONVERGEN C E poor behavior such as absenteeism. This productivity-enhancing effect of cities, the devolution of responsibilities may benefit concentration of firms and people in cities is from complementary activities (such as train- not only beneficial for growth and inclusion ing and the provision of frameworks and but also inevitable. guidance) to raise schools’ and districts’ The Middle East and North Africa is no capacity to manage resources, staffing, curri- exception. A high concentration of the cula, and so on (Mansuri and Rao 2013). region’s population, and its poor, live in cit- Increasing work-relevant skills. Greater ies, where productivity and income opportu- interaction between the school system and nities are higher. However, chapters 1–4, the private sector can improve the relevance outlined how several of the region’s cities of schooling and the school-to-work transi- may be underperforming relative to their tion. Workplace training (such as apprentice- potential. ships, technical training, and internships for Modernist planning ideals have left cities university students) can be employed to fragmented internally and disconnected from increase students’ employability after the global economy. These modernist ideals ­ g raduation. In Tunisia, for example, intro- are observed in the government-led develop- ducing an entrepreneurship track that com- ment of new sites fragmented from the main, bines business training with personal “messy” urban fabric and in regulations out- coaching helped to reshape the business skills lawing market-appropriate housing and com- of university students (Valerio, Parton, and mercial developments. The disjunction Robb 2014). Building partnerships between between top-down visions and standards and schools and the private sector can also close bottom-up market demand and purchasing information gaps between the labor market power has been accompanied, on the one hand, and education by consulting on curriculums, by underused, expensive d ­ evelopments—such providing internships and vocational training as the largely uninhabited new towns of Egypt, curricula, hosting job fairs, or providing Morocco, Saudi Arabia, and the United Arab career counseling (Gatti et al. 2013). Emirates (for instance, Abu Dhabi’s Masdar City)—and, on the other hand, by the prolif- eration of informality as stifled markets try to Transitional Step 4: Renew the breathe. focus on nurturing urban These planning practices tend to raise the agglomerations by streamlining costs and reduce the connectivity of the land transfer procedures and Middle East and North Africa’s cities, limit relaxing zoning regulations in the exploitation of productive opportunities existing cities, lowering the at a cost to job creation, and prevent the regulatory barriers to their market from meeting households’ needs. As redevelopment a result, large numbers of residents unable to Much of the rapid growth and poverty meet the elevated costs are forced into infor- reduction of the past century has taken place mal homes and jobs, where a vicious cycle of in, and been shepherded by, cities. Well- poverty and social segregation follows. managed cities offer large, well-connected Weak connectivity, high costs of living and markets—for products, services, labor, jobs, doing business, and heavy restrictions on and ideas—to firms and residents. This con- productive activities prevent scarce urban nectivity can greatly reduce the costs of a land from being used most productively, huge range of transactions, raising efficiency reduce efficiency and job creation, and leave and productivity. Dense markets in turn the region’s cities less able to break into enable firms to produce at scale and with global tradables and stuck in a model of higher specialization, two dynamics that are “consumption cities.” fundamental to productivity, economic How can these patterns be reversed? An g row th, and job creation. Given the important strategy is to move away from F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    195 modernist planning ideals and work with pri- An ideal response may involve clarification vate sector demand for positive urban den- of land and property ownership as well as sity. This does not demand an acceptance of wholesale reforms to land and property regis- urban chaos; indeed, disorderly informal tration and transactions, building permitting, housing and jobs have been the result of plan- and business licensing. However, this will be ning that is out of sync with demand. Rather, a lengthy process for many countries, test- it means supporting a market-led urban den- ing institutional capacities, while in others, sity and affordability that creates space in the vested interests and broader political sensi- formal economy for lower-income residents tivities may pose insurmountable barriers to and firms and ensures their access to the meaningful reform and modernization.4 basic services required for their productivity Where the complete modernization of and well-being. The role of the government factor markets is unrealistic, important versus the private sector would be adjusted, gains can nonetheless be made through with governments identifying and getting incremental scaffolding reforms that lay the behind emerging density and demand to pro- foundations for later modernization or vide the complementary public goods and tackle the most harmful bottlenecks. For services that support productivity and avert example, land and property market modern- the downsides of density such as congestion, ization can be approached in bite-size pollution, contagion, and crime. Specifically, chunks, such as first digitizing the existing this reorientation would include the cadaster and updating records incremen- following: tally, beginning with priority locations such as major commercial and industrial areas or •  Reviewing city master plans to encour- important urban neighborhoods. Attaching age market-led density, connectivity, and more demanding land and property tax affordability obligations to land and property ownership •  Diagnosing and shedding unnecessary, claims can also encourage owners them- distortive, or infeasible planning codes selves to clarify ownership. and regulations (such as building stan- Where broader reforms are not attainable, dards, industrial location directives, and countries and regions have also used indus- zoning regulations) that may be stifling trial parks and SEZs to create pockets of affordability, densification, productivity, ­ better-functioning land and property mar- and adaptability of the city kets and wider business environment institu- •  Prioritizing in situ upgrading of basic tions. Others have forgone these benefits by neighborhood and trunk infrastructure applying the same or even more stringent in the urban core instead of focusing pre- regulations within industrial parks and hence dominantly on greenfield development— failing to create attractive pockets of well- ensuring that current neighborhoods functioning institutions. have proper infrastructure for sanitation, Although Transitional Step 4 focuses on floodproofing, street lighting, waste col- city-level reforms, they will not alone create lection, and so on, as well as diagnosing the growth and job creation needed in the the need for bus lanes, improved junc- Middle East and North Africa. Also key will tions, and wider traffic management be delivering more effective support to lag- •  Identif ying and planning the urban ging areas (Transitional Step 1); making city expansion area and addressing incen- management more responsive and effective tives that encourage sprawl (such as (Transitional Step 2); strengthening human more relaxed regulations in peripheral capital so workers are globally competitive versus central areas and low central area (Transitional Step 3); and opening cities to attractiveness) while undergoing basic the large global and regional markets that grid planning to ensure that any further c a n s u s t a i n a b l y f u e l t h e i r g r ow t h expansion is orderly. (Transitional Step 5). 1 9 6    C ONVERGEN C E Transitional Step 5: Expand integration with external markets. Its start- market access for cities by ing point delineated three types of countries thinning the “thick borders” that based on their relationships with world mar- inhibit mobility across the Middle kets and large emerging markets (World East and North Africa for both Bank 2009): regional trade and migration •  Type 1 countries are close to the leading At present, the concentration of people world markets such as North America, and firms in the Middle East and North Europe, and East Asia. Most of the Africa’s leading cities is unmatched by those Middle East and North Africa countries cities’ integration into larger external mar- belong to this type, such as North African kets, because “thick borders” restrict critical countries with European and global input and output flows throughout the region access via seaports. (as examined in chapter 3). Yet regional and •  Type 2 countries have large emerging global integration—and the market access markets nearby but are far from world this brings—will be crucial to the economic markets. These include the countries dynamism of these cities. neighboring Brazil, China, India, the Globally, firm and industry value chains Russian Federation, and South Africa. are now increasingly spread across countries Middle East and North Africa countries because of the falling costs of transport, with relatively large neighboring econo- ­ c om mu n i c at ion , a nd d at a t r a n sf e r. mies can, with similar considerations, Manufacturing, but increasingly also ser- implement policies to tap into the growth vices, has become more about specialized potential of their dominant neighbor. This tasks within global or regional value chains may apply, for example, to the Islamic than about finished products. This interin- Republic of Iran, Iraq, and the Syrian dustry and intraindustry trade (rather than Arab Republic with respect to Turkey. trade in final goods) is what has increased tre- •  Type 3 countries are in regions far from mendously over the past half century, espe- world markets and also lack large local cially in East Asia, North America, and emerging markets. These regions include Western Europe. An example is the consumer Central Africa, Central Asia, East Africa, technology cluster in East Asia, where China West Africa, and the small Pacific islands. has done most of the assembly but high-tech The category also includes countries inputs come from Japan or the Republic of affected by conflicts or otherwise politi- Korea and some lower-tech ones from cally fragmented from world markets, as Southeast Asia (box 6.8). experienced in many Middle East and North Africa countries and regions. For A country typology offers a general instance, although Syria adjoins Turkey, framework for market integration its civil war has greatly reduced the chance to trade, and hence Syria is arguably tem- Despite the formidable barriers to similar porarily in this third type. Likewise, the regional integration in the Middle East and Republic of Yemen would fall into this North Africa, the stakes are too high to category because conflict has fragmented abandon efforts. A precondition for partici- it from trade opportunities. pation in these cross-national production net- works has been countries’ willingness to open their markets for inputs and outputs Type 1 countries: Align institutional and reduce barriers to trade and to capital reforms with international market and labor flows. standards World De velopme nt Re por t 20 0 9: The greatest opportunities come from the Reshaping Economic Geography outlined a largest global markets. Where these are framework to understand priorities for accessible (in Type 1 countries), the priority is F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n M ENA   197 BOX 6.8  How special economic zones supported China’s incremental integration into global markets In the 1970s, the Chinese government had resolved opened another 11 border cities and 6 ports along the to experiment with greater openness to global trade Yangtze River (map B6.8.1, right panel). and free market institutions. Rather than liberalizing The developments reflected a strategy of incremen- the entire economy—which was politically and insti- tal liberalization of ever-expanding pockets of the tutionally infeasible at the time—China approached economy, building on positive experiences, which may this process incrementally and experimentally. An offer lessons for challenging cases in the Middle East early step was the introduction of four special eco- and North Africa where full liberalization is infeasible. nomic zones (SEZs) in 1980, strategically located Several countries have tried to replicate China’s suc- along the southeastern coast (Shenzhen, Zhuhai, cess but misinterpreted its success factors and merely and Shantou in Guangdong Province and Xiamen placed cookie-cutter, infrastructure-heavy economic in Fujian Province) at China’s gateway to the global zones in targeted lagging regions. China’s strategy economy (map B6.8.1, left panel). succeeded because it closely targeted a specific and Following the early success of these SEZs, 14 cities major binding constraint in the local economy (being along the coast began to accept overseas investment in closed to foreign investment and featuring illiberal eco- 1984, and the island of Hainan was assigned SEZ status nomic institutions) and located the SEZs in places with in 1988 (map B6.8.1, middle panel). Around the same strong access to global markets. China’s accompany- time, the coastal belt around the Yangtze and Pearl ing economic boom and significant poverty reduction River Deltas and the Xiamen-Zhangzhou-Quanzhou have been intensely concentrated in these advantaged Triangle in south Fujian Province opened for business coastal locations, while sparsely populated inland with the world. In the early 1990s, the government places remain much less economically dynamic. MAP B6.8.1  China gradually increased special economic zones from 1980 through the 1990s Sources: IBRD 35899, January 2008; World Bank 2009. 1 9 8    C ONVERGEN C E to align domestic institutions and regulations Japan and later Korea linked up with China with the standards of the international and other Asian neighbors). Regional inte- market. gration is important to compensate for the Countries close to world markets have a relatively small size of individual countries’ geographical advantage because multina- economies and scaling-up capacity; to tional enterprises look to extend their supply improve specialization and scale across the chains to countries peripheral to the main region; and to strengthen the Middle East markets served. These investors look, above and North Africa’s basis to trade with the all, for effective and transparent institutions world. With this regional integration, the and regulations, good and stable governance, Middle East and North Africa can particu- a stable macroeconomy, and any relevant larly benefit from nearby Europe as an trade agreements. Therefore, such founda- important final market and source of capital tional institutional reforms are critical to and know-how. unlock potential in such countries. This approach can be seen in the Middle East and Type 3 countries: Take advantage of global North Africa, such as in select sectors in and regional incentive mechanisms Morocco and Tunisia that integrated into For countries far from major global global production chains via institutional markets and lacking a large, dynamic reforms such as improved governance and neighbor—the Type 3 countries—more a l ig n ment of qu a l it y st a nd a rd s a nd targeted incentives may be necessary. For certification. example, the European Union has a com- pensating mechanism to finance infra- Type 2 countries: Align infrastructure and structure and other development projects regulations with large neighboring markets i n l e s s - d e ve l o p e d a n d s l ow- g r ow t h Middle East and North Africa countries regions, spreading the gains from more- with a large neighboring market (Type 2) advanced regions. Other incentive mecha- may wish to target integration with that n i sm s m ig ht i nclude d i re c t a id a nd neighbor. The largest regional markets are preferential market access. relatively well placed to attract global invest- Although Middle East and North Africa ment and can form hubs for regional supply countries have relatively good proximity to chains, into which neighbors can be inte- large markets in Europe, current or historic grated through connective infrastructure, conflict and other divisions have nonetheless regulatory alignment, and broader trade and often isolated countries from opportunities diplomatic agreements. This approach may to trade (as seen recently in Iraq, Libya, be a priority for countries in the region with Syria, and the Republic of Yemen). More relatively low gross domestic product (GDP) broadly, not all countries that pursue and foreign investment but whose neighbors reg iona l i nteg rat ion benef it equa l ly, are well integrated in global supply chains, although practically all have raised their such as the countries bordering Turkey: the economic fortunes. Where the political will- Islamic Republic of Iran, Iraq, and the Syrian ingness exists to compensate those that fell Arab Republic. relatively behind, regional integration can However, regional integration among the generate the resources to do so. Thus, region’s countries is a complement, rather regional economies most strongly positioned than an alternative, to deeper integration to gain from global trade may benefit from with other world regions. The Middle East also supporting less-advantaged regional and North Africa lacks a single dominant, counterparts through investments in infra- reliable anchor economy in the region to structure, human capital development, and serve as the gateway to international mar- institutional strengthening that can both kets (in the way Western Europe served this further regional integration and raise those role for Eastern Europe after 1990 or as countries’ gains from trade. F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    199 Small steps toward trade integration power markets. Weak logistics sectors appear could bring immediate benefits to be a major institutional barrier to regional integration in the Middle East and North In the Middle East and North Africa, an Africa (box 6.9). ideal, comprehensive, and resilient regional In many contexts, the mechanisms to track trade agreement and infrastructure network is shipments are not available; customs proce- unlikely. However, smaller steps—which dures require many documents and check- neighboring countries, or groups of them, can points; customs clearance information is not implement—can provide immediate benefits shared between border agencies, even within and move countries along the transition path the same country; customs regulations are not to greater long-term integration. Informed by coordinated across borders; international this framework, governments could consider a standards of customs procedures are not pragmatic strategy of incremental improve- adopted; and bribery and other kinds of ments comprising three elements: ­ corruption are widespread in border crossings. •  Institutional reforms that enable economic There is also underutilization of ICT, which actors to take advantage of cross-border could complement these deeper reforms to economic opportunities would include effi- raise the efficiency of procedures (as seen in cient customs procedures, regulatory har- Tunisia’s and Morocco’s recent online goods monization, migration regimes, and broad clearance systems, Tunisia TradeNet [TTN] business environment reforms. and the BADR e-payment system, respectively) •  Incremental improvements to the physi- (Kulenovic 2015). Finally, state monopolies of cal infrastructure for trade would focus the logistics sector in most of the region’s on borders, transport and energy, and countries have undermined incentives to information and communication technol- reform. ogy (ICT). Regulatory harmonization. While liberal- •  Appropriate incentives take into con- ization of the logistics sector can improve sideration that not all participants in a efficiency, more incremental efficiency regional agreement will benefit equally improvements can also shepherd substantial from it. Certain forms of compensation or progress without overstretching institutions. offsetting benefits will often be necessary. Incremental, technocratic reforms to address these challenges—such as streamlining cus- toms procedures, harmonizing customs regu- Reform institutional functions to unlock lations across borders, and upgrading cross-border opportunities logistics services—have the potential to Physical trunk infrastructure to connect reduce delays, costs, and poor reliability markets, while helpful, is rarely the binding (Rouis and Tabor 2013). Morocco provides a constraint on regional trade in the Middle positive example, supporting the develop- East and North Africa. Rather, institutional ment of logistics zones that overcame broader reform is required to enable and facilitate the weaknesses in the sector (in Casablanca, exchange of goods, services, capital, people, Tangier, and elsewhere); removing regula- and ideas; address risks; and strengthen eco- tions deterring foreign direct investment nomic fundamentals to ensure that a coun- (FDI) in the logistics sector; and establishing try’s outputs are indeed internationally new customs regimes suitable for logistics competitive. activities (Rouis and Tabor 2013). Logistics sector improvements. Greater The case of China’s SEZs (discussed ear- competition in infrastructure service provi- lier in box 6.8) also showed how institutions sion may be important to increase goods can be strengthened in pockets of the econ- trade efficiency in areas such as port opera- omy as scaffolding when wholesale reform tion (to reduce container dwell times), trans- across the economy was unattainable. port services (to reduce hauling costs), and The  Democratic Republic of Congo and 2 0 0    C ONVERGEN C E BOX 6.9  Logistics, more than infrastructure, impedes trade in the Middle East and North Africa Middle East and North Africa countries tend to fall other countries in the region ranked in the middle behind less in infrastructure quality than in areas of or lower portions of the LPI: Three Maghreb states logistics quality, international shipping, and the qual- (Algeria, Morocco, and Tunisia) ranked between 105 ity of customs services. The poor quality of trade sup- and 117. Two Mashreq states (Syria and Iraq) as well port services—such as customs procedures, trucking, as the Republic of Yemen and Libya came in the low- warehousing, or freight forwarding—increases trade est in the region, ranking between 138 and 154. costs substantially. A mong the non- GCC countries, the Islamic The World Bank’s 2018 Logistics Performance Republic of Iran, Djibouti, and Egypt achieved the Index (LPI) summarized these elements for 160 coun- greatest improvements in their LPI scores between tries around the world. The Gulf Cooperation Council 2010 and 2018 (figure B6.9.1, panel a). On the cus- (GCC) countries and Israel ranked relatively high toms quality indicator, Egypt, the Islamic Republic of across these logistics indicators, followed by Egypt Iran, Israel, and Jordan raised their scores the most and the Islamic Republic of Iran (table B6.9.1).a Most (figure B6.9.1, panel b). TABLE B6.9.1  Logistics Performance Index (LPI) rankings of Middle East and North Africa countries, 2018 Overall LPI International Logistics quality Tracking and Country rank Customs Infrastructure shipments and competence tracing Timeliness United Arab Emirates 11 15 10 5 13 13 4 Qatar 30 38 27 9 31 30 36 Israel 37 29 28 75 34 32 48 Oman 43 44 39 36 49 66 29 Saudi Arabia 55 66 43 56 57 46 67 Bahrain 59 63 68 55 58 60 68 Kuwait 63 56 45 98 67 96 59 Iran, Islamic Rep. 64 71 63 79 62 85 60 Egypt, Arab Rep. 67 77 58 73 63 89 74 Lebanon 79 106 73 70 104 74 77 Jordan 84 88 70 119 93 84 76 Djibouti 90 113 60 118 135 72 85 Tunisia 105 107 133 115 123 71 70 Morocco 109 115 93 103 101 112 114 Algeria 117 138 96 122 113 103 124 Syrian Arab Republic 138 154 82 126 124 128 148 Yemen, Rep. 140 104 131 141 131 146 151 Iraq 147 153 140 131 159 144 129 Libya 154 149 115 159 153 160 123 Mid. East & N. Africa avg. 84 89 72 89 88 85 81 Source: World Bank Logistics Performance Index (LPI) 2018. Note: The 2018 LPI covered 160 countries globally. No data were available for West Bank and Gaza. Shading designates global quintiles, from light blue (highest) to dark orange (lowest). box continues next page F i v e S t e p s f o r E n a b l i n g G r o w t h t h r o u g h T h r i v i n g C i t i e s a n d T o w n s i n MEN A    201 BOX 6.9  Logistics, more than infrastructure, impedes trade in the Middle East and North Africa (continued) FIGURE B6.9.1  Among the Middle East and North Africa countries, about half improved their logistics performance between 2010 and 2018 a. Overall LPI score b. Customs score (sorted by change in score) (sorted by change in score) Qatar Qatar Oman Egypt, Arab Rep. United Arab Iran, Islamic Rep. Emirates Iran, Islamic Rep. Israel Djibouti Jordan Egypt, Arab Rep. Algeria United Algeria Arab Emirates Iraq Djibouti Jordan Tunisia Israel Yemen, Rep. Saudi Arabia Libya Libya Iraq Tunisia Saudi Arabia Yemen, Rep. Kuwait Kuwait Morocco Bahrain Bahrain Syrian Arab Republic Oman Morocco Syrian Arab Republic Lebanon Lebanon 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 2010 2018 Source: World Bank Logistics Performance Index (LPI) 2010 and 2018. Note: Countries are arranged in order of most to least improvement on the LPI, which covered 160 countries globally in 2018. No data were available for West Bank and Gaza. For Morocco, the 2010 and 2012 scores were used. a. The Gulf Cooperation Council (GCC) countries are Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. The Mashreq subregion comprises the Arab Republic of Egypt, Iraq, Jordan, Lebanon, the Syrian Arab Republic, and West Bank and Gaza. The Maghreb subregion comprises Algeria, Libya, Morocco, and Tunisia. 2 0 2    C O N V E R G E N C E Rwanda offer other examples of scaffolding: Migration-related reforms. Chapter 3 also they created pockets of liberalized trade and emphasized institutional barriers to services movement between border towns through trade and the movement of people, which a  novel visa and small traders’ regime limits not just services trade directly but also (box 6.10). the exchange of ideas and know-how and the Customs reforms are most effective when establishment of stronger social and business harmonized between neighbors (ECA 2015). networks across borders. Reforms in this Certain governments in the region have cre- area are particularly important in the Middle ated a national agency to promote coordina- East and North Africa, given the size of ser- tion of border control services within the vices sectors and remittances in the region’s country (for example, the Moroccan Agency economies. for Logistics Development). Similar institu- Reforms include aligning domestic regula- tions might usefully work with neighbors tions that license service providers across and wider regional partners to fast-track borders (such as mutual recognition agree- harmonization. ments for professional bodies of architects, BOX 6.10  Scaffolding for cross-border trade and migration in the Great Lakes Region The Rwanda–Democratic Republic of Congo border allowing nationals to simply scan their ID cards and cuts between the twin cities of Goma (Democratic pass through a turnstile to enter Goma. Republic of Congo) and Rubavu (Rwanda) in the In theory, the CEPGL is also complemented by the north, and between Bukavu (Democratic Republic Common Market for Eastern and Southern Africa’s of Congo) and Rusizi (Rwanda) in the south. These Simplified Trade Regime, allowing small traders cities sit at the edge of eastern Democratic Republic (identified by their membership in an association issu- of Congo’s troubled Kivu provinces, the epicenter of ing IDs) to trade goods up to US$2,000 in value from major conflicts and illicit activities for decades. a harmonized list of products (agreed to bilaterally) However, an innovative regime facilitates fluid, without any tariff or nontariff barriers. However, in small-scale, cross-border trade and migration between practice, because of poor communication about the the cities. For example, in 2018, an estimated 90,000 Simplified Trade Regime on both sides as well as par- people crossed the Goma–Rubavu border each day, ticularly weak institutions in the Democratic Republic predominantly to trade small volumes of items such of Congo, this provision is little known or accessed as fresh produce, meat, clothes, or water but also for by traders. For instance, traders may face checks and services trade like accessing a bank, barbers, or res- irregular taxes and fees from up to 17 agencies on taurant, and for social purposes. the Democratic Republic of Congo side (only 5 of the The key to fluid trade across this challenging bor- agencies are permitted to be present). der has been the Economic Community of the Great Studies have found, however, that despite such Lakes Countries (CEPGL) protocol. Through this hurdles, the intense border trade has created impor- initiative, residents of each border city can use their tant livelihood opportunities for low-income people voter or national IDs to obtain a jeton card, enabling in each border city, particularly women, who form free travel into the cross-border town (not beyond the the majority of traders. It also fills important gaps town’s limits) without any further documentation. in access to goods and services, particularly for the Additional inexpensive passes are available to travel Democratic Republic of Congo, such as trade of water farther into each country without a passport for a lim- during cuts, food supplies, and access to more modern ited number of days per trip and cumulative days per banking facilities. There is also evidence the scheme year. For purposes of small trade, the CEPGL is com- has improved social relations, with 40 percent of trad- plemented by a directive that simplifies procedures ers reporting new social invitations from across the and reduces tariffs (to 5 percent) for agricultural con- border since they began trading. signments under US$500 or 60 kilograms. Rwanda has recently also introduced electronic border gates, Sources: Mwanabiningo 2015; Vanguard Economics 2018. F i ve S t e p s for E n a b l i ng G ro w t h t h ro u g h T h r i v i ng C i t i es a n d To w ns i n M E N A   203 accountants, lawyers, and so on), reducing subsidies and setting tariffs high enough to barriers to FDI in services, reducing visa recover costs is important, yet politically and requirements, and removing barriers to the technically complicated, and reform has flow of information. These would be comple- often failed because of strong political pro- mented by developing more competitive tests. The Islamic Republic of Iran achieved human capital able to create and take advan- temporary success in removing a large energy tage of the opportunities beyond low-value- subsidy. The government introduced highly added services trade. publicized compensatory mechanisms, Broad business environment reforms. including mass cash transfers, and held fre- Unblocking trade between the Middle East quent dialogues and campaigns with citizens and North Africa’s countries will yield only to gain favor for the reforms (Salehi-Isfahani muted benefits if the underlying bottlenecks 2011). Although this suggests ingredients of to private sector development, competition, success, the reform also struggled over the and diversification are not also addressed. long term in the context of international The well-documented institutional bottle- sanctions and high inflation, and political necks include preferential treatment of challenges arose to focusing cash transfers favored firms, other regulations that con- only on the poorest to manage affordability. strain competition and entry, high shares of government employment, dominance of nat- Improve the regional infrastructure ural resource rents in exports, and firm subsi- for trade dies that impose a large fiscal burden without Infrastructure that connects places to benefiting the poor. For example, firm-level markets can shape the fortunes of cities and data indicate that the region’s exporting regions. Better-connected areas offer greater firms are no more productive than firms that access to markets, typically encouraging do not export, while there is a large and densification and elevated investment. ­ significant productivity premium for manu- However, as noted in Transitional Step 1, facturing firms that import (Francis and for infrastructure to benefit the local econ- Schweiger 2017). omy, it should be within the framework of a This situation is consistent with the exis- multidimensional strategy that addresses tence of barriers and privileged access affect- complementary constraints to competitive- ing imports and exports. For example, cities ness. These include the business environ- are unlikely to benefit from lower tariffs on ment and trade institutions, basic public intermediate inputs if favored importers con- services, and human capital to ensure that trol distribution and would simply capture corridors are attractive sites for investment, the savings (Hoekman 2016). Similarly, even not merely recipients of final imports or fly- with tariff and nontariff barriers, energy over towns. trade is unlikely to take off without address- A concrete vision for a regional network of ing energy subsidies and price controls. development corridors, and an assessment of Chapter 3 also noted that despite high its benefits, could help create greater support capital openness, regional capital flows are for closer integration in the Middle East and low, in part because of underlying institu- North Africa. Other regions have pursued tional weaknesses. In addition, it noted that integration through such strategic regional meaningful digital integration in the region is visions as the Trans-European Transport hampered by such underlying barriers as low Network (TEN-T) to improve connectivity human capital and restrictions on the across the continent, especially between exchange of information (for example, media Eastern and Western Europe (complemented controls). by the regulatory harmonization of the During the recent decline of oil prices, European Union)5 and China’s Belt and Road many of the region’s countries started to Initiative, connecting China not just within reform their energy sectors. Reducing Asia but as far as Europe and Africa. 2 0 4    C O N V E R G E N C E Academic research has made considerable •  Better leveraging border free-trade zones, progress in evaluating the benefits of such ini- such as that between Tunisia and Libya, tiatives at both the national level and for for incremental integration. places within countries.6 A comprehensive Infrastructure does not only concern assessment of a regional integration strategy roads and railways. Infrastructure and could estimate the potential gains from over- complementary reforms affecting regional coming integration barriers as well as the energy trade may be particularly transfor- geographical and distributional impacts mative. National power generation capacity within each country. Some of the specific is generally small in the Middle East and transport infrastructure improvements that North Africa, while energy demand is could be evaluated include the following ­ projected to increase by 50–60 percent in (Rouis and Tabor 2013): the region between 2018 and 2048.7 Even •  Enabling road traffic from Egypt to Jordan with potentially large efficiency increases, and Saudi Arabia, which currently relies the energy supply will need to expand. on ferry service, by opening a 15-kilometer I nte g rat ion of energ y ne t work s a nd corridor through Israel (box 6.11) increased trade can make these supply •  Facilitating cross-border trade in the increases more affordable. Mashreq subregion (Iraq, Israel, Jordan, Larger systems improve capacity utiliza- Lebanon, and Syria), in part to encourage tion and supply reliability even while reduc- economies of scale in port operations and ing required reserve margins, because trends shipping services, which are now scat- in daily and seasonal supply and demand tered across seven medium-size ports tend to vary across countries and subnational •  Expanding rail connections from Saudi regions. This becomes even more important Arabia to Jordan and beyond as well as to as the share of intermittent renewable power other Gulf countries sources increases. For instance, high renew- •  Improving rail connections from the able energy shares in Europe would not be Islamic Republic of Iran to Turkey, possible without the ability to trade electric- including building the missing rail link ity among countries. All of these benefits arou nd L ake Van, which cu rrently result in greater reliability and lower prices requires a ferry to cross for end users. BOX 6.11  The gains to border regions in Egypt and Jordan from better regional integration Improving transport infrastructure and reducing • Partial integration: the opening of the road cor- trade barriers between Egypt and Jordan could bring ridor with a reduction by half of the border wait- the better-connected areas greater access to markets. ing time Using a quantitative spatial model similar to Redding • Full integration: the opening of the road corri- and Rossi-Hansberg (2016) that incorporates a geog- dor with no border waiting time. raphy of trade costs between districts, local wages, Simulations of the welfare gains for all scenarios and labor mobility, the benefits from opening a show that the welfare gains are high and are much 15-kilometer corridor through Israel and facilitating larger when both a 15-kilometer corridor through border crossing between Jordan and Egypt can be Israel is open and border crossing is facilitated evaluated. Three scenarios are considered: (table B6.11.1). Full integration almost doubles the • Infrastructure connection: the opening of the welfare gains for Egypt from the opening of the road road corridor only corridor and more than doubles the gains for Jordan. box continues next page F i ve S t e p s for E n a b l i ng G ro w t h t h ro u g h T h r i v i ng C i t i es a n d To w ns i n M E N A   205 BOX 6.11  The gains to border regions in Egypt and Jordan from better regional integration (continued) Trade facilitation reforms maximize the benefits from better connectivity toward new markets. In Egypt, cross-border road investments. the welfare gains range from 3 percent to 4 percent Regions at the borders between Jordan and Egypt growth in GDP per capita for the regions in the Sinai, will benefit the most from the opening of the cor- while they are around 2 percent for the rest of the ridor and partial trade facilitation at the border country. In Jordan, the gains for the southern regions (map B6.11.1). Previously isolated regions in the Sinai are higher than 17 percent, while they are around and in the south of Jordan will benefit the most from 15 percent for the rest of the country. TABLE B6.11.1  A road corridor connecting Egypt and Jordan would bring welfare gains from each scenario Increase in GDP per capita, percent Country Infrastructure connectiona Partial integrationb Full integrationc Egypt, Arab Rep. 1.45 2.2 2.9 Jordan 9.6 15.1 24.1 Source: World Bank computations from spatial model simulations as described in Redding and Rossi-Hansberg 2016. Note: Each scenario assumes the opening of a 15-kilometer road corridor through Israel from Egypt to Jordan. a. “Infrastructure connection” scenario involves only the opening of the road corridor. b. “Partial integration” scenario would add facilitation of border crossing that cuts waiting time in half. c. “Full integration” scenario would add facilitation of border crossing that eliminates waiting time. MAP B6.11.1  Spatial welfare gains from partial integration of an Egypt–Jordan road corridor Beirut Netanya Tel Aviv Amman Cairo Welfare gains Scenario: half border time 2%–2.2% 2.3%–2.7% 2.8%–3.4% Rabigh 3.5%–16.4% 16.5%–19.4% Jeddah Source: World Bank computations from spatial model simulations as described in Redding and Rossi-Hansberg 2016. Note: The simulation assumes the opening of a 15-kilometer road corridor through Israel from Egypt to Jordan plus facilitation of border crossing sufficient to cut waiting time in half. 2 0 6    C O N V E R G E N C E Several power grid interconnection worse off. 9 However, orchestrating such schemes already exist or are planned in the redistribution is not straightforward. Middle East and North Africa: Domestically, such compensation requires that governments capture some of the gains •  In the M aghreb subregion, Algeria, of trade and use them to benefit, for example, Morocco, and Tunisia are interconnected, people in isolated lagging areas who may see with Morocco also connected to the an out-migration of firms to newly connected pan-European transmission network via regions. Redistributive policies are far more Spain. complex at an international level. Only the •  In the Mashreq subregion, the intercon- European Union has extensive mechanisms nection involves eight countries (including for redistributing resources across countries Turkey). through several structural and investment •  In the GCC subregion, the interconnec- funds.10 Such mechanisms require far closer tion involves six countries. international agreements than seem feasible A vision for a pan-Arab electricity net- in the Middle East and North Africa in the work exists (World Bank 2013). It proposes a foreseeable future. parallel process of strengthening the infra- Instead, agreements more limited in geo- structure and expanding trade while also graphic and substantive scope would consti- building the institutional, legal, and policy tute important transitional steps for regional framework. The result would be (a) full inter- integration. For example, a larger, more connection between the three subregions and influential nation may be the most ready to any countries not yet included in one of these; benefit from openness, the most likely able to (b) additional connections to other countries make concessions in return, and hence the in Europe, Africa, and Asia; and (c) open most able to lead the financing of regional access to the transmission network with full infrastructure or offer the greatest conces- wholesale competition. Electricity trade sions in regional institutional reforms.11 would be overseen by a regional market oper- Alternatively, countries that benefit from ator that would also ensure functioning of increased economic activity around a mod- financial markets as required for effective ernized regional port might in turn offer their trading. Many steps will be necessary to neighbors greater access to their market to achieve this vision. Initial scaffolding actions sell agricultural goods or allow greater labor could begin today to strengthen existing net- inflows. Such narrow agreements have been works and gradually expand connections to successful even between countries and new participants and interconnections regions that otherwise have strained rela- between the networks as reforms in individ- tions; an example is the Indus Water Treaty ual countries allow. between India and Pakistan, which has held up throughout many decades of tense Design incentives to share the gains from relations. integration Integration will benefit the region overall, but particular people or places may gain less Concluding remarks or even be harmed. Regional integration can Rather than representing new develop- increase the concentration of economic activ- ment objectives, the transitional steps set ity, exacerbating inequality within or across forth in this chapter offer opportunities for countries despite the overall gains. Some past governments in the Middle East and North regional integration agreements have failed in Africa to pursue their same final objectives part because of the perception of unequal with greater efficacy. In concluding, we illus- gains.8 Gains should be great enough that trate how the above considerations may those who gain from regional integration can inform more effective strategies to meet a compensate those who do not or who become range of possible policy objectives. F i ve S t e p s for E n a b l i ng G ro w t h t h ro u g h T h r i v i ng C i t i es a n d To w ns i n M E N A   207 Create jobs for youth signaled demand by attracting households and firms. Such places may include the If the government’s aim is to create jobs urban expansion area of a main city, fast-­ for young people, the first lesson is that large growing satellite towns, or highly acces- cities with strong market access offer the sible secondary cities. Opportunities in greatest potential for doing so. How can cit- these alternative locations can be unlocked ies then deliver these jobs? This is the concern through the same urban reforms as well of Transitional Steps 1, 2, and 3: unlocking as better local governance (Transitional the potential for private job creation in cities Step 5) while observing the warnings and through principles of Transitional Step 2. •  Proper public investment in the funda- •  Address gaps in welfare and human capi- mentals for growth and reduced regula- tal in lagging areas. This can prevent tion that stifles private investment; residents from leaving prematurely before •  Connections between cities and world urban opportunities are available while markets as sources of growth, particularly also equipping residents with stronger by deepening regional integration; and human capital to tap into opportunities •  Preparation of young people, ensuring that do arise locally or in cities. they are equipped not only with paper credentials but also with globally tradable Revive a particular lagging area skills. If the aim is to revive a particular lagging Where social divisions affecting certain area, it is important to recognize the chal- youth (such as language or cultural barriers lenges of such efforts (given path dependency between their home and the leading city) pre- and any underlying disadvantages that vent migration to leading cities and thus explain the place’s underperformance) and to make local cities more relevant, Transitional build on whatever advantages and potential Steps 2 and 3 provide guidance on how to are already there. This means identifying better support local job creation in spite of existing centers of density and market access the challenges. (such as major regional cities, ports, transit corridors, and so on) that might serve as Decongest a leading city poles of growth for the area and ensuring connectivity to external markets (through If policy makers are concerned by what soft and hard reforms) as a minimum. This economist Edward Glaeser has termed the high accessibility is the most common source “downsides of density” and wish to decon- of “one big opportunity” for an area, but gest a leading city (for example, a capital complementary factors may include fixed city), consider three key complementary assets such as tourism heritage. approaches: Beyond building on the area’s existing •  Address the negative externalities in the potential, it is essential to diagnose the key city itself (which are unlikely to recede local bottlenecks to greater productivity, of their own accord because of strong including through engagement with the local demand for major cities) by investing business community and residents, and to in transport and other basic infrastruc- carefully target development strategies to ture and regulating negative externalities address these bottlenecks. Strategies that are like pollution and traffic, as discussed in sensitive in this way to local needs and oppor- Transitional Step 4. tunities and properly coordinated in the lag- •  Unlock any latent potential in alter- ging area can be particularly encouraged by native locations for firms and people strengthening the role—while building the that offer strong access to national and capacity and accountability— of local international markets and have already governments. 2 0 8    C O N V E R G E N C E Launch an effective SEZ successes in ways that often conflate form and function. This mimicry leads to a situa- Regional policies should avoid the mis- tion where “looks like” substitutes for take of assuming that SEZs are the solution “does”; that is, governments look capable to their problem. However, an SEZ may be after the mimicry but are not actually more justified, such as to serve as a second best capable. Isomorphic mimicry may explain pocket of liberal trade and investment in why countries fail to build real capability a more restrictive regime. If an SEZ is even after years of policy and reform engage- planned, how can its chances of success be ment and billions of dollars of capacity- building work—especially where public optimized? sectors have become closed to novelty but •  First, zones should be in locations with open to (and supportive of) agenda strong access to international markets conformity. and must offer the essential soft comple- 2. While remote, rural areas of the region have ments such as appropriate openness, ease the largest share of households lacking access to basic services, dense urban areas often host of doing business, trade facilitation, and the largest number of households suffering so on. This may require liberalization of poor service delivery, and certain areas face relevant trade services. particular demand shocks from internally •  The intention of benefiting the local mar- displaced persons and refugees. ket should be remembered; hence, zones 3. For example, girls outperform boys in math- should be placed close to existing popula- ematics, engineering, and computer informa- tion centers with relevant capacities and tion in many Middle East and North Africa should facilitate supply chain linkages and countries, but they are underrepresented in technology transfer. the booming technology sector (El-Alfi and •  Unhelpful competition between subna- Boutros 2018). tional authorities should be avoided, and 4. For example, the World Bank report, “Reshaping Egypt’s Economic Geography,” opportunities for authorities to coordi- notes that Egypt’s land and property system nate to develop a shared zone with greater would ideally undergo a complete overhaul agglomeration economies should be but that such a process has been attempted explored. several times and failed because of entrenched •  Weaknesses in local governance that systems and institutions as well as the sheer risk undermining the reliability of infra- scale of the challenge (World Bank 2012). structure and quality of business services 5. “Infrastructure and Investment,” Mobility should be addressed (or compensated for, and Transport, European Commission web- such as through private management of site: https://ec.europa.eu/transport/themes​ the zone). /infrastructure_en. •  Given large uncertainty about the poten- 6. For a review, see Roberts et al. (2018). 7. International Energy Outlook 2017 projects tial of the zone or wider region, feedback that energy use will increase during 2018–48 loops should be built in and frictions to by 56 percent in the Middle East and by the transformation of land, buildings, and 63 percent in Africa (EIA 2017). productive activities in the zone reduced 8. Examples are Central America in the 1960s to allow competition between firms and and East Africa in 1970s (World Bank sectors and to help zones adapt as compe- 2009). tencies and advantages evolve. 9. This principle is the so-called Kaldor-Hicks criterion, as discussed in Donaldson (2015). 10. “European Structural and Investment Funds,” EU Regional and Urban Development, Notes European Commission website: http:// 1. Andrews, Pritchett, and Woolcock (2017) ec.europa.eu/regional_policy/en/funding/. define “isomorphic mimicry” as the tendency 11. For a discussion, see chapter 9 in World Bank of governments to mimic other governments’ (2009). F i ve S t e p s for E n a b l i ng G ro w t h t h ro u g h T h r i v i ng C i t i es a n d To w ns i n M E N A   209 References ECA (United Nations Economic Commission for Africa). 2015. “International Transport and Agénor, P.-R., M. K. Nabli, and T. M. Yousef. Trade Fac i l it at ion i n N or t h A f r ic a .” 2005. “Public Infrastructure and Private Publication CEA-AN/PUB/14/1, Subregional Investment in the Middle East and North Office for Nor th A frica, EC A , Rabat, Africa.” Policy Research Working Paper 3661, Morocco. World Bank, Washington, DC. EIA (U.S. Energy Information Administration). Andrews, Matt, Lant Pritchett, and Michael 2017. “International Energy Outlook 2017.” Woolcock. 2017. “Looking Like a State: The Publication No. IEO2017, EIA, Washington, Seduction of Isomorphic M imicr y.” I n DC. Building State Capability: Evidence, Analysis, El-Alfi, Ahmed, and Iris Boutros. 2018. “What Action , edited by Matt A ndrews, Lant Tech Can Do for Arab Women.” The Cairo Pritchett, and Michael Woolcock, 29–52. Review of Global Affairs 31 (Fall 2018). Oxford: Oxford University Press. El-Kogali, Safaa El Tayeb, and Caroline Krafft, Australian Embassy and Asia Foundation. 2016. eds. 2019. Expectations and Aspirations: “Coordinating Roads and Infrastructure A  New Framework for Education in the Investments for Development (CR+ID): Working Middle East and North Africa. Washington, Towards More Strategic and Sound Local Public DC: World Bank. Investments.” Coalitions for Change project Francis, D. C., and H. Schweiger. 2017. “Not So note, Asia Foundation, San Francisco. Different from Non-Traders: Trade Premia in B a h l , R . 2 0 0 8 . “ T h e P i l l a r s of F i s c a l the Middle East and North Africa.” Economics Decentralization.” CA F Working Paper of Transition 25 (2): 185–238. 2008/07, CA F (Corporacion Andina de Gatti, R., M. Morgandi, R. Grun, S. Brodmann, Fomento)–Development Bank of Latin America, D. Angel-Urdinola, J. M. Moreno, D. Marotta, Caracas, Venezuela, RB. M. Schiffbauer, and E. Mata Lorenzo. 2013. Beddies, Sabine, Mariana Felicio, Gabriel Dedu, Jobs for Shared Prosperity: Time for Action in Fatou Fall, and Caroline Vagneron. 2011. t he M i d d l e E a s t a n d N o r t h A f r i c a . “MENA Local Service Delivery Initiative: Washington, DC: World Bank. Promoting Social Accountability and Demand G au b e r t , C . 2 018 . “ F i r m S o r t i n g a n d for Good Governance.” Arab World Brief Agglomeration.” N BER Working Paper No.  3, Report No. 69347, World Bank, 24 478 , Nat iona l Bu reau of E conom ic Washington, DC. Research, Cambridge, MA. Bismil, Farmanullah. 2018. “The Future of Hoekman, B. 2016. “Intraregional Trade: Special Economic Zones in Afghanistan: How Potential Catalyst for Growth in the Middle Does the Trump Administration’s New South East.” Policy Paper 2016-1, Middle East Asia Strategy Change the Plan for Establishing Institute, Washington, DC. SEZs in Afghanistan?” The Diplomat, June 6. Kulenovic, Z. Joe. 2015. “Tangier, Morocco: Brixi, H., E. Lust, and M. Woolcock. 2015. Trust, Success on the Strait of Gibraltar.” Private Voice, and Incentives: Learning from Local Sector Development Blog, November 25. Success Stories in Service Delivery in the ht tps: // blogs.worldbank.org /psd /tangier​ Middle East and North Africa. Washington, -morocco-success-strait-gibraltar. DC: World Bank. Mansuri, Ghazala, and Vijayendra Rao. 2013. Donaldson, D. 2015. “The Gains from Market Localizing Development: Does Participation Integration.” Annual Review of Economics Work? World Bank Policy Research Report. 7 (1): 619–47. Washington, DC: World Bank. Ducruet, César, Fatima Zohra Mohamed-Chérif, Mwanabiningo, Nene Morisho. 2015. Deriving and Najib Cherfaoui. 2011. “Maghreb Port Maximum Benefit from Small-Scale Cross- Cities in Transition: The Case of Tangier.” Border Trade between DRC and Rwanda. PORTUSplus 1 (1). L ondon: I nter national A ler t ; Nai robi: Duranton, Gilles, and Anthony Venables. 2018. TradeMark East Africa. “Place-Based Policies for Development.” Redding, Stephen J., and Esteban Rossi-Hansberg. NBER Working Paper 24562, National Bureau 2016. “Quantitative Spatial Economics.” of Economic Research, Cambridge, MA. NBER Working Paper 22655, National Bureau doi:10.3386/w24562. of Economic Research, Cambridge, MA. 2 1 0    C O N V E R G E N C E Roberts, M., M. Melecky, T. Bougna, and Y. Xu. R ep or t N o. 71 289 - E G , World B a n k , 2018. “Transport Corridors and Their Wider Washington, DC. Economic Benefits: A Critical Review of the ———. 2013. “Middle East and North Africa Literature.” Policy Research Working Paper Integration of Electricity Networks in the Arab 8302, World Bank, Washington, DC. World: Regional Market Structure and Rouis, Mustapha, and Steven R. Tabor. 2013. Design.” Energy study, Report No. ACS7124, Regional Economic Integration in the Middle World Bank, Washington, DC. East and North Africa: Beyond Trade Reform. ———. 2016a. “Hashemite Kingdom of Jordan: Directions in Development Series. Washington, Education Sector Public Expenditure Review.” DC: World Bank. R ep or t N o. AC S18935, World B a n k , Salehi-Isfahani, Djavad. 2011. “Iran: Subsidy Washington, DC. Reform amid Regional Turmoil.” Op-Ed, ———. 2016b. “Upper Egypt Local Development March 3, Brookings Institution, Washington, P rog ra m-for- Re su lt s P roje c t .” P roje c t DC. https://www.brookings.edu/opinions/ Appraisal Document, Report No. 108094, iran-subsidy-reform-amid-regional-turmoil/. World Bank, Washington, DC. Si ms , D. 2015. Eg y pt ’s De se r t D re a m s: ——— . 2017a. “Di re St ra it s: T he C r isis Development or Disaster? Oxford: Oxford Surrounding Poverty, Conflict, and Water in University Press. the Republic of Yemen.” WASH Poverty Valerio, Alexandria, Brent Parton, and Alicia Diagnostic Series, Report No. 11711, World Robb. 2014. Entrepreneurship Education and Bank, Washington, DC. Training Prog ram s aroun d the World: ———. 2017b. “Inclusive Service Delivery: Yemen Dimen sion s for S uccess . Directions in Policy Note 4.” Working Paper, Report No. Development Series. Washington, DC: World 120534, World Bank, Washington, DC. Bank. ———. 2017c. “Opportunities for Improving Vanguard E conomics. 2018. “Democratic Urban Service Delivery in South Sudan: A Tale Republic of Congo Borderlands Report.” of Two Cities. Part I: Service Delivery Status Report, Vanguard Economics, Kigali, Rwanda. Report.” World Bank, Washington, DC. World Bank. 2009. World Development Report doi:10.1596/29199. 20 09: Reshaping Economic Geography. ———. 2019. “Territorial Development in Support Washington, DC: World Bank. of N DS 2030.” Reimbursable Advisory ———. 2012. “Arab Republic of Egypt: Reshaping Services (RAS) paper for the Croatia 2030 Egypt’s Economic Geography: Domestic National Development Strategy (NDS), World Integration as a Development Platform.” Bank, Washington, DC. ECO-AUDIT Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage electronic publishing options and print-on-demand technology, which is located in regional hubs world- wide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)–certified paper, with nearly all containing 50–100 percent recycled content. The recycled fiber in our book paper is either unbleached or bleached using totally chlorine-free (TCF), processed chlorine–free (PCF), or enhanced elemental chlorine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility. Policy makers across the Middle East and North Africa have for many years articulated plans to integrate their people spatially and economically. Wishing to bring communities together and narrow economic gaps, governments have made large capital investments in transport corridors and “new cities.” Hoping to provide jobs in places with little economic activity, governments have designated new industrial zones supported by spatially targeted business incentives. Yet the results of these place-based initiatives in the region have largely underdelivered in terms of yielding more sustainable jobs and growth. Spatial inequality contributes to a 63 percent larger share of inequality in this region than in any other region. Sharp disparities remain between capital cities and lagging areas, as well as between richer and poorer sections within cities. Across much of the region, a fortunate few are connected to opportunities, while many more people are marginal to the formal economy—or live outside of it. Convergence: Five Critical Steps toward Integrating Lagging and Leading Areas in the Middle East and North Africa considers the numerous and varied challenges. It explains that many of these place-based policies are ineffective because they treat the spatial and physical symptoms of inequity rather than its root causes, which are economic and institutional. Convergence presents the five roots of spatial inequity in institutional inefficiencies across the Middle East and North Africa—urban regulatory frictions, credentialist education systems, centralized control over local public services, barriers to the spatial mobility of goods and people, and barriers to market entry and lopsided business environments—within cities, within countries, and across national borders. It proposes five transitional steps to enable convergence informed by economic geography: • Strengthen coordination and complementarities across initiatives. • Redistribute roles and responsibilities across tiers of government. • Enable mobility of people between lagging and leading areas. • Build dense and connected cities. • Enhance market access nationally and regionally. 978-1-4648-1450-1 90000 9 781464 814501 SKU 211450