Africa’s Cities Opening Doors to the World Somik Vinay Lall J. Vernon Henderson Anthony J. Venables With Juliana Aguilar, Ana Aguilera, Sarah Antos, Paolo Avner, Olivia D’Aoust, Chyi-Yun Huang, Patricia Jones, Nancy Lozano Gracia, and Shohei Nakamura. 1 Africa’s Cities | Opening Doors to the World All queries on rights and licenses should be addressed to the Open to the World Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. © 2017 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. This work is a product of the staff of The World Bank with external Crowded contributions. Note that The World Bank does not necessarily own each component of the content included in the work. The World Bank therefore does not warrant that the use of the content contained in the work will not infringe on the rights of third parties. The risk of claims resulting from such infringement rests solely with you. 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 Disconnected 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 Costly This work is available under the Creative Commons Attribution 3.0 Unported license (CC BY 3.0) http://creativecommons.org/licenses/ by/3.0. 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: Lall, Somik Vinay, J. Vernon Henderson, and Anthony J. Venables. 2017. “Africa’s Cities: Opening Doors to the World.” World Bank, Washington, DC. License: Creative Commons Attribution CC BY 3.0 Closed for Business 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. 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; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Urban Planning ISBN (paper): 978-1-4648-1044-2 ISBN (electronic): 978-1-4648-1045-9 DOI: 10.1596/978-1-4648-1044-2 Design and production by Zephyr www.wearezephyr.com Infrastructure 2 Africa’s Cities | Opening Doors to the World Contents Acknowledgments ............................................................ 7 Overview Africa’s Cities: Part I Crowded and Opening Doors to the World Disconnected African Cities Chapter 1 Page 37 Crowded with people, not dense with capital The low development trap — Africa’s urban Crowded with people.....................................................38 economies are limited to nontradable goods and Slums: Workers’ only option when urban services ............................................................................ 12 economic density is low but highly concentrated ...................................................................38 Crowded, disconnected, and thus costly — Africa’s High population density at the city’s core, cities are limited to nontradables by urban form ..... 16 rapid tapering on the outskirts ................................ 41 Crowded cities ............................................................. 16 Disconnected cities .................................................... 19 Not dense with capital ................................................... 41 Costly cities ..................................................................22 Not dense with buildings ...........................................44 Not dense with amenities, not livable .....................45 Closed for business, out of service: Case studies: Access to amenities in Dar es The urgency of a new urban development Salaam, Durban, and elsewhere in Africa ...............50 path for Africa .................................................................26 Cities are “closed for business” ................................26 Low human capital .........................................................56 Cities are “out of service” ..........................................26 References ....................................................................... 61 Path dependence and interdependence.................27 Springing cities from the low development trap .......28 Formalize land markets, clarify property rights, and institute effective Chapter 2 Page 63 urban planning ............................................................28 Disconnected land, Make early and coordinated infrastructure people and jobs investments — allowing for interdependence among sites, structures, and basic services ...........29 Disconnected land .........................................................64 Opening the doors .........................................................30 Collections of small and fragmented neighborhoods............................................................64 Annex: African cities used in the analysis ..................32 Spatial fragmentation ................................................65 References .......................................................................34 People not connected to people: High fragmentation, low exposure, little potential for interaction .......................................69 People not connected to jobs .......................................71 Lack of transportation infrastructure .....................71 Lack of money for transportation ............................ 74 Inaccessible employment ..........................................79 References .......................................................................83 3 Africa’s Cities | Opening Doors to the World Part II Africa’s Low Urban Urban planning institutions and land use regulation ............................................................... 131 Development Trap Strengthening capacity and resources for urban planning ..........................................................132 Addressing coordination constraints across Chapter 3 Page 87 levels of government ................................................132 Costly for households, Urban regulations ........................................................133 costly for firms Streamlining administrative procedures ..............133 Relaxing planning standards ..................................133 High prices, low incomes...............................................88 To build cities that work, make land markets High wages, high costs of doing business ..................96 work — nothing less will do ........................................136 References .......................................................................99 References .....................................................................137 Chapter 4 Page 101 Chapter 6 Page 139 Africa’s urban Scaling up and coordinating development trap investments in physical structures and infrastructure Cities closed for business............................................102 The “nontradables trap”: Theory ............................102 Investing early in infrastructure to shape The “nontradables trap”: Evidence ........................105 urban structures ..........................................................140 Sunk costs, construction, and the Leveraging road investment .......................................143 expectations trap ......................................................... 112 Bus rapid transit: One option among many .........143 Integrated urban planning, regulation, References ..................................................................... 114 and transportation investments ............................145 Roads, densification, and land use change in four East African cities .............................................148 Part III Springing Africa from Its Citywide economic benefits of road Low Urban Development Trap improvements in Kampala ......................................150 Provide public goods and services for livability ...... 151 Chapter 5 Page 117 Finance for lumpy urban infrastructure investments...................................................................155 Clarifying property rights and strengthening urban planning References .....................................................................158 Why African cities fail to attract investment: An urban planner’s perspective ................................. 118 Capital misallocation ................................................ 118 About the contributors Institutional constraints .......................................... 118 Page 160 Ineffectual property rights ...................................... 121 The example of Nairobi ...........................................122 Clear land and property rights ...................................123 Land valuation and prices ...........................................129 Removing data and legal obstacles .......................129 Improving tax collection ..........................................130 4 Africa’s Cities | Opening Doors to the World Figures 1 The share of firms in internationally traded and nontradable 2.9 In Kigali, workers in the informal sector have sectors varies widely across developing country cities................13 shorter commutes ............................................................................75 2 In resource-exporting countries, urbanization is linked only 2.10 The typical household in many African cities cannot afford weakly to the development of manufacturing and services in public transport fares.......................................................................77 countries that export natural resources ........................................15 2.11 Scale economies in vehicle size should lead to the 3 Connections among people as a function of population near provision of collective transportation through large vehicles the city center: Nairobi, Kenya, is more fragmented and less toward the city center and smaller ones as distance to the well-connected than Pune, India ....................................................18 city center increases .........................................................................78 4 “Leapfrog development” undermines economies of scale 2.12 A very large share of trips in African cities are made by foot .....79 and agglomeration ...........................................................................20 3.1 African cities face high prices for their income levels ..................89 5 Paved roads occupy a smaller share of urban land in Africa 3.2 Consumers in Africa face high price premiums ............................91 than elsewhere — and usually drop off abruptly beyond the city center ..........................................................................................21 3.3 Households in African cities grapple with subsistence ................92 6 A fragmented urban form is associated with higher 3.4 Average time and cost to register property in Sub-Saharan urban costs ........................................................................................23 countries and international benchmarks ......................................93 7 Urban living costs in Sub-Saharan African countries in 2011 3.5 Average time and cost to deal with construction permits in exceeded costs elsewhere, relative to Africans’ lower per c Sub-Saharan countries and international benchmarks ...............94 apita GDP ...........................................................................................24 3.6 Infrastructure costs per capita decrease with density .................95 8 Urban households — especially poor ones — spent large shares 3.7 A fragmented urban form is associated with higher of their budgets on transportation (analysis from 2008) ............25 urban costs ........................................................................................96 1.1 Very high proportions of city dwellers live in slums in Africa .....40 3.8 Nominal manufacturing wages in African cities are higher 1.2 Population density in African cities is lower than in some than in other developing-country cities .........................................97 other regions .....................................................................................40 3.9 Sales revenue per worker in African and other developing- 1.3 The value of building stocks in African cities is low ......................44 country cities .....................................................................................98 1.4 Large shares of the population in Dar es Salaam lack access 4.1 Supply and demand for labor in a “local” city .............................103 to basic sanitation and water services ...........................................50 4.2 Supply and demand for labor in an internationally 1.5 Access to improved water and sanitation in Dar es Salaam competitive city ...............................................................................104 tends to rise with distance from the center ..................................51 4.3 The tradables sector is much larger in Asia than in Africa ........106 1.6 Access to improved sanitation and piped water at home is 4.4 Urbanization and manufacturing share of GDP in Africa and low across Dar es Salaam ................................................................52 outside Africa ..................................................................................108 1.7 Access to Services in Durban decreases with distance from 4.5 The share of land devoted to slums increases with distance the city center....................................................................................53 from the city center ........................................................................112 1.8 Moving to Durban improved migrants’ access to basic 5.1 Average distances to commercial and industrial areas services before but not after 2000 .................................................53 from formal and informal settlements in Dar es Salaam, 1.9 Within-city moves in Durban are less likely to yield better Tanzania, and Kigali, Rwanda ........................................................120 sanitation than they did in the past ...............................................54 5.2 Larger agglomerations have higher, and increasing, 1.10 Access to piped water, septic tanks, and electricity improved tensions over land prices ...............................................................124 in Maputo, Mozambique, between 1997 and 2013 ......................55 5.3 Key players in urban development in Tanzania and Uganda ...134 1.11 Access to piped water, septic tanks, and electricity improved 5.4 Average time and cost to register property in African in Kinshasa, Democratic Republic of Congo, between 2007 countries and international benchmarks in 2015 ......................135 and 2013 ............................................................................................55 5.5 Minimum lot sizes are large in Africa ...........................................136 1.12 Cognitive skills of most urban workers are limited, especially 6.1 Land values of de novo development projects are higher in Africa ..............................................................................................60 than values in other neighborhoods of Dar es Salaam, 2.1 Three aspects of being connected..................................................65 including rich ones .........................................................................141 2.2 The average “openness index” of cities in Sub-Saharan Africa 6.2 Differential impacts of de novo and upgrading projects in is not very different from the index of other regions ..................65 Dar es Salaam .................................................................................142 2.3 Land within five kilometers of the central business district is 6.3 Area around the Gangding station, in Guangzhou, often left unbuilt in Africa ................................................................66 China, before and after construction of the bus rapid 2.4 African cities are more fragmented in the center than transit system ..................................................................................143 comparably sized cities in India ......................................................67 6.4 Average daily bus rapid transit passenger counts 2.5 Urban people in Africa have less potential for interaction than are much higher in Lagos than in Johannesburg........................145 urban people in other regions ........................................................69 6.5a Integrated urban planning and regulation promote density ....147 2.6 African cities are becoming more fragmented .............................70 6.5b Integrated urban planning and regulation promote density ....148 2.7 Paved roads occupy a smaller share of urban land in Africa 6.6 Percentage of area devoted to paved and unpaved roads than elsewhere — and usually drop off abruptly beyond the in four East African cities, 2001 and 2013....................................149 city center ..........................................................................................71 2.8 Urban Africans spend a large share of their budgets on food, leaving little for transport ................................................................74 5 Africa’s Cities | Opening Doors to the World Boxes 6.7 Access to electricity by newly arrived migrants and other residents in the Democratic Republic of Congo, Ghana, Nigeria, and Sierra Leone ..............................................................154 6.8 Gap between capital needs and budget resources of city governments ............................................................................155 1 The promise of cities: Agglomeration economies and 6.9 Land-based financing instruments and city evolution ...............156 returns to scale .................................................................................14 2 Low capital investment in Sub-Saharan African cities during a period of rapid urban growth ................................17 Maps 3 Urban land and property rights: A need for clarification.............29 4 Leveraging land values to finance Africa’s urban infrastructure .........................................................................30 5 Building dense, connected, and efficient cities: A.1 Location of African cities used in the analysis ..............................33 Two models of success ....................................................................31 1.1 Housing conditions and shortages in Sub-Saharan Africa, 1.1 Life in Africa’s cities is often miserable: by subregion......................................................................................49 An urban migrant’s story .................................................................42 2.1 Change in land used by paved roads across four 1.2 The most valuable building structures are concentrated African cities ......................................................................................73 near the city center...........................................................................45 2.2 Most job opportunities in Nairobi are inaccessible to 1.3 The contribution of cognitive and noncognitive skills people without cars ..........................................................................80 to urban performance......................................................................59 3.1 Map of relative expensiveness of household consumption 2.1 Measuring fragmentation in urban form ......................................68 in Sub-Saharan Africa .......................................................................90 2.2 Is there a constant travel time budget? 5.1 Average built height in Nairobi, 2015 ...........................................123 The Zahavi conjecture ......................................................................75 5.2 Diversity of land rights in urban and peri-urban 4.1 Do natural resource exports explain why African Africa, 2009 ......................................................................................124 cities specialize in nontradables?..................................................110 5.3 Land tenure systems in Greater Kampala ...................................126 5.1 Inappropriate building regulations hamper affordability .........119 6.1 The Kampala’s Northern Bypass allows traffic to bypass 5.2 Customary land rights in Durban, South Africa ..........................125 Kampala’s city center......................................................................151 5.3 Welfare costs of stymied redevelopment in Nairobi..................127 5.4 Recent actions to improve land administration Tables and common knowledge in Africa ................................................128 5.5 Land sharing and readjustment: Two ways to include residents in urban redevelopment plans ....................................129 5.6 Leveraging land values to finance urban infrastructure............130 A.1 African cities included in the analysis, by population size ...........33 6.1 Bus rapid transit: Successful if handled with care ......................144 1.1 Slum population as percentage of total urban population 6.2 Alternate paths for improving urban mobility: Lessons in selected cities, historically and in 2014......................................38 from China and Japan ....................................................................146 1.2 Percentage of population with access to water and sanitation, 6.3 Providing water to the poor in African cities ...............................152 by region ............................................................................................46 1.3 Housing amenities in Sub-Saharan Africa, by type of area and expenditure quintile .................................................................47 1.4 Adult literacy rate, population 15+ years .......................................56 1.5 Youth literacy rate, population 15–24 years, both sexes .........................................................................................57 1.6 Skills assessed in the STEP Survey ..................................................58 2.1 Accessibility to formal jobs in Nairobi ............................................81 2.2 Estimated capital costs of building various types of transportation infrastructure ..........................................................82 4.1 Top 10 commodity exports from Asia and Africa, 2000–10 ......107 5.1 Percentage of land registered and number of days required to transfer property in selected countries and regions .................127 5.2 Ratio of registered planners to population in selected countries, circa 2011 ......................................................................132 6 Acknowledgments This volume is part of the African Regional Studies The report draws on a set of over twenty-five Program, an initiative of the Africa Region Vice- research papers produced as part of a research Presidency at the World Bank. This series of program on Urbanization and Spatial Development studies aims to combine high levels of analytical in Developing Countries, conducted by the World rigor and policy relevance, and to apply them Bank, Oxford University and the London School to various topics important for the social and of Economics. The research has been supported economic development of Sub-Saharan Africa. by generous financial contributions from UK- The Office of the Chief Economist for the Africa Aid through the Multi-donor Trust Fund on Region provides quality control and oversight. Sustainable Urbanization at the World Bank. A team led by Somik V. Lall, together with J. Vernon The team received valuable comments from Henderson and Antony J. Venables, prepared this Richard Damania, Marianne Fay, Indermit Gill, report. Members of the core team included Juliana William Maloney (peer reviewers), Souleymane Aguilar, Ana Aguilera, Sarah Antos, Paolo Avner, Coulibaly, Eric Lancelot, Mark Lundell, and Roland Olivia D’Aoust, Chyi-Yun Huang, Patricia Jones, Nancy White. The team thanks Francisco H. G. Ferreira Lozano Gracia, and Shohei Nakamura. Neeraj Baruah, for providing guidance to the team during his Louise Bernard, Julia Bird, John Felkner, Arti Grover tenure as Africa Chief Economist for the World Goswami, Rashmin Gunasekara, Rawaa Harati, Bank. The team benefited from discussions with and Dzhamilya Nigmatulina provided additional Abebaw Alemayehu, Andre Bald, Mapi M. Buitano, research inputs. Paul Collier, Indermit Gill and William Meskerem Brhane, Punam Chuhan-Pole, Narae Maloney were key advisors to the report team. Choi, Dean Cira, Sateh Chafic El-Arnaout, Sylvie Debomy, Roger Gorham, Andre Herzog, Sheila The main authors and contributors were: Kamanyori, Michel Matera, Megha Mukim, Shomik • The Overview was written by Somik V. Lall. Mehndiratta, Martin Onyach-Olaa, Dina Ranarifidy, Apurva Sanghi, Maria Angelica Sotomayor and Roland • Chapter 1 (Crowded with People, Not Dense White. The team appreciates the opportunity to with Capital) was written by Nancy Lozano discuss the policy framework and findings at various Gracia, J. Vernon Henderson, and Juliana Aguilar, forums including at the East and Central African with contributions from Ana Aguilera, Olivia Forum (Kampala), Habitat III (Quito), Johns Hopkins D’Aoust, Somik V. Lall and Tvisha Nevatia. University (Washington DC), UN Habitat (Nairobi), • Chapter 2 (Disconnected Land, People and Jobs) TDLC Seminar on Land Use Planning & Spatial was written by Paolo Avner, J. Vernon Henderson Development (Tokyo), UK DFID (London), and World and Somik V. Lall, with contributions from Neeraj Bank Sprig Meetings and workshops in Addis Ababa, Baruah, Louise Bernard, Julia Bird, Olivia D’Aoust, Dar es Salaam, Nairobi, Pretoria, and Washington. Somik V. Lall and Dzhamilya Nigmatulina. The report has been produced under the supervision • Chapter 3 (Costly for Households, Costly of Sameh Wahba, Director for Urban and Territorial for Firms) was written by Olivia D’Aoust, Development, and the overall direction of Albert Patricia Jones, and Shohei Nakamura, with Zeufack, Africa Chief Economist for the World Bank. contributions from Rawaa Harati. Nick Moschovakis and Bruce Ross-Larson, with a • Chapter 4 (Africa’s Urban Development Trap) was team at Communications Development, edited the written by Anthony J. Venables with contributions report. Zephyr designed and typeset the report. J. Vernon Henderson and Patricia Jones. • Chapter 5 (Clarifying Property Rights and Strengthening Urban Planning) was written by Chyi-Yun Huang, Olivia D’Aoust, and Somik V. Lall, with contributions from Juliana Aguilar and Julia Bird. • Chapter 6 (Scaling up and Coordinating Investments in Physical Structures and Infrastructure) was written by Olivia D’Aoust and Somik V. Lall, with contributions from Juliana Aguilar, John Felkner, J. Vernon Henderson, and Julia Bird. 7 Africa’s Cities | Opening Doors to the World 8 Overview Africa’s Cities: Opening Doors to the World The low development trap — Africa’s urban economies are limited to nontradable goods and services Crowded, disconnected, and thus costly — Africa’s cities are limited to nontradables by urban form Out of service, closed for business: The urgency of a new urban development path for Africa Springing cities from the low development trap Opening the doors 9 Africa’s Cities | Opening Doors to the World African cities are crowded, disconnected, and costly. Typical African cities share three features that constrain urban development and create daily challenges for residents: Crowded, not economically dense — investments in infrastructure, industrial and commercial structures have not kept pace with the concentration of people, nor have investments in affordable formal housing; congestion and its costs overwhelm the benefits of urban concentration. Disconnected — cities have developed as collections of small and fragmented neighborhoods, lacking reliable transportation and limiting workers’ job opportunities while preventing firms from reaping scale and agglomeration benefits. Costly for households and for firms — high nominal wages and transaction costs deter investors and trading partners, especially in regionally and internationally tradable sectors; workers’ high food, housing, and transport costs increase labor costs to firms and thus reduce expected returns on investment. 55% African households face higher costs relative to their per capita GDP than do households in other regions — much of it accounted for by housing, which costs them a full 55 percent more in this comparison 10 In eight representative African cities, roads occupy far lower shares of urban land than in other cities around the world. 20% African cities are 20 percent more fragmented than are Asian and Latin American ones. In Harare, Zimbabwe, and Maputo, Mozambique, more than 30 percent of land within 5 kilometers of the central business district remains unbuilt. 472 million Urban areas in Africa comprise 472 million people. That number will double over the next 25 years as more migrants are pushed to cities from the countryside. The largest cities grow as fast as 11 4 percent annually. Africa’s Cities | Opening Doors to the World Africa’s Cities: Opening Doors to the World Cities in Sub-Saharan Africa are experiencing rapid population growth. Yet their economic growth has not kept pace. Why? One factor might be low capital investment, due in part to Africa’s relative poverty: Other regions have reached similar stages of urbanization at higher per capita GDP. This study, however, identifies a deeper reason: African cities are closed to the world. Compared with other developing cities, cities in Africa produce few goods and services for trade on regional and international markets (figure 1). To grow economically as they are growing in size, dense — not merely crowded? How can they acquire Africa’s cities must open their doors to the world. efficient connections? And how can they draw firms They need to specialize in manufacturing, along with and skilled workers with a more affordable, livable other regionally and globally tradable goods and urban environment? services. And to attract global investment in tradables From a policy standpoint, the answer must be production, cities must develop scale economies, to address the structural problems affecting which are associated with successful urban economic African cities. Foremost among these problems development in other regions. are institutional and regulatory constraints that Such scale economies can arise in Africa, and they will misallocate land and labor, fragment physical — if city and country leaders make concerted efforts development, and limit productivity. As long as African to bring agglomeration effects to urban areas. Today, cities lack functioning land markets and regulations potential urban investors and entrepreneurs look and early, coordinated infrastructure investments, at Africa and see crowded, disconnected, and costly they will remain local cities: closed to regional and cities. Such cities inspire low expectations for the global markets, trapped into producing only locally scale of urban production and for returns on invested traded goods and services, and limited in their capital. How can these cities become economically economic growth. The low development trap — Africa’s urban economies are limited to nontradable goods and services How does the production of locally consumed, or Since the 1980s, much of the growth in developing nontradable, goods and services trap cities into low countries has depended on the expansion of exports economic growth? Put simply, producing for local through industrial production and higher technology. markets limits returns to scale. The consumer base of Unlike nontradables, tradable goods and services one city, however large, is much smaller than a regional face elastic global demand. They may also allow for or global market. Specializing in nontradables for local agglomeration economies, which increase returns to consumption leads to diminishing returns (both for employment (box 1). Rapidly growing cities require technological reasons, and because prices are set locally growth in employment — and the returns to expanding and decline as supply increases). In contrast, export employment are highest in tradable sectors. markets are key to a dynamic industrial sector. 12 Overview | The low development trap — Africa’s urban economies are limited to nontradable goods and services FIGURE 1 Share of firms in internationally traded and nontradable sectors, selected developing-country cities (latest post-2010 data) Luanda Gaborone Dar es Salaam Kampala Kigali Bamako Accra Nouakchott Africa Nairobi Dakar Addis Ababa Kinshasa Lagos Niger Mombasa Lusaka Harare Middle East North Africa Tunis Beirut and Amman Cairo La Paz Asunción Latin America Caribbean Cordoba and the Medellin Bogota Buenos Aires Lima Yangon Pacific & South Asia East Asia and the Bangkok Zhengzhou City Dhaka Shenzhen City Delhi Chittagong 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Nontradable Tradable Source: Calculation based on the World Note: The data is from the latest WBE surveys post-2010 (with more than 15,000 Bank Enterprise (WBE) surveys. firms in capital cities, or cities of at least one million inhabitants, and with at least 50 firms sampled). Only firms with five or more employees are interviewed. The sectoral specialization analyses used the UN International Standard Industrial Classification of All Economic Activities (3.1 revision). Manufacturing, wholesale and commission trade, and business services (such as travel agencies, transport, financial intermediation) are all tradable activities. By contrast, construction, local services, retail trade, health and social work, and other local activities are classified as nontradable. 13 Africa’s Cities | Opening Doors to the World Because of manufacturing’s importance in entering Why have African urban economies remained local? regional and global markets, one can look at the share Two reasons stand out. One, paradoxically, is natural of manufacturing in GDP to see whether an urbanizing resource development. Such development can create economy is opening its doors to the world — or a high demand for nontradable goods and services. closing them. For example, we compare the structures As growth in the natural resource sector raises factor of non-African and African economies during periods prices, this sector crowds out others — notably when the urbanized share of the population rises manufacturing (figure 2). Countries that depend to 60 percent. Based on a cross-section of African heavily on natural resource exports tend to sprout and non-African economies, the comparison urban economies dominated by nontradable services shows that Africa’s cities are indeed trapped in the (“consumption cities”). This syndrome is known as production of nontradables for local markets. As the Dutch Disease. African economies attain 60 percent urbanization, Another reason for Africa’s local urban economies their share of manufacturing in GDP stays flat (or is related to urban form: how cities are built and somewhat falling) at about 10 percent. In contrast, the spatially organized. The findings in this report draw manufacturing share of the non-African economies on spatial and economic analysis based on 64 cities rises from 10 percent to nearly 20 percent (falling back covering large, medium, and small cities across only when urbanization exceeds 60 percent). Africa and shows that cities are growing under a patchwork of constraints — inefficient land markets, BOX 1 The promise of cities: Agglomeration economies and returns to scale What is an urban agglomeration economy, and how has many benefits. Certain public goods — like does it arise from economic density? A simple case infrastructure and basic services — are cheaper to is the reduction of transport costs for goods: When provide when populations are large and densely suppliers are close to their customers, shipping costs packed together. Firms located near each other can decline. In the late nineteenth century, four fifths share suppliers, lowering input costs. Thick labor of Chicago’s jobs were compactly located within markets reduce search costs, giving firms a larger four miles of State and Madison Streets — near pool of workers to choose from. And spatial proximity residences and infrastructure (Grover and Lall 2015). makes it easier for workers to share information and And in the early 1900s, New York and London were learn from each other. International evidence shows manufacturing powerhouses because factories were that knowledge spillovers play a key role in boosting built there to access customers and transport services. the productivity of successful cities. Many agglomeration benefits increase with scale: Evidence from East Asia (China, the Republic of Each doubling of city size increases productivity by 5 Korea, Vietnam) points clearly to a close association percent, and the elasticity of income with respect to between episodes of rapid urbanization and economic city population is between 3 percent and 8 percent development. Unfortunately, these links appear (Rosenthal and Strange 2004). weak in Sub-Saharan Africa. Cities in Africa are not Productivity gains are closely linked to urbanization delivering agglomeration economies or reaping urban through their ties to structural transformation and productivity benefits; instead, they suffer from high industrialization. As countries urbanize, workers move costs for food, housing, and transport. These high from rural to urban areas in search of better paid and costs — rising from coordination failures, poorly more productive jobs. Similarly, entrepreneurs locate designed policies, weak property rights, and other their firms in cities where agglomeration economies factors that lower economic density — lock firms into will increase their productivity. Close spatial proximity producing nontradable goods and services. 14 Overview | The low development trap — Africa’s urban economies are limited to nontradable goods and services overlapping property-rights regimes, suboptimal with original research and analysis to explain how and ineffective zoning regulations — that hinder the the form of African cities is trapping them into local drive toward dense concentrations of structures. and nontradable production — and to point leaders More, the resulting scattered neighborhoods lack toward policies that can spring the trap. planned transport and infrastructure connections. To be sure, urban form is not the only constraint Without either high physical density or adequate on Africa’s international competitiveness. Other connective infrastructure, an urban area falls short of important factors include business regulation; the lack its potential: It cannot offer firms the cost efficiencies of access to finance (for residential and commercial and job matching advantages that open a city’s doors investments); the peculiarity of Africa’s demographic to regional and global trade. transition; the absence of agricultural productivity Even if the symptoms of Dutch Disease are mitigated gains; and, more generally, the macroeconomic by falling commodity prices, the typical African city context. These factors compound the risk that Africa’s will remain bound by constraints related to its form. cities will remain unwelcoming to investment — that These physical constraints deter regional and global their development will continue along paths that investment. And because they are likely to persist preclude their entry into higher-productivity tradable as the principal constraints on economic growth, goods sectors. And yet this threat of path dependency addressing them is one of Africa’s most urgent is itself closely, demonstrably related to the evolution challenges today. This report combines recent findings of cities’ physical form. FIGURE 2 In resource exporting countries, urbanization is linked only weakly to the development of manufacturing and services Nonresource exporters Resource exporters 100 Urbanization rate in 2010 (percent) Sub-Saharan Africa Other 80 60 40 20 20 40 60 80 100 20 40 60 80 100 Share of manufacturing and services in GDP in 2010 (percent) Source: Gollin, Jedwab, and Vollrath 2016. 15 Africa’s Cities | Opening Doors to the World Crowded, disconnected, and thus costly — Africa’s cities are limited to nontradables by their urban form Many Sub-Saharan African cities share three Workers should consume more diverse products and characteristics that constrain economic development services, pay less for what they consume, and enjoy and growth. Two appear directly in the cities’ physical easier commutes because of proximity to their jobs. structures and spatial form: They are crowded with Africa’s cities feel crowded precisely because they people and dwellings, and they are disconnected by are not dense with economic activity, infrastructure, a lack of transport and other infrastructure. Finally, or housing and commercial structures. Without and in Part because they are disconnected, cities are adequate formal housing in reach of jobs, and without also costly. Indeed, they are among the costliest in the transport systems to connect people living farther world, both for firms and for households — not least away, Africans forgo services and amenities to live in because of their inefficient spatial form. cramped quarters near their work. Often informal, these downtown districts are likely to lack adequate infrastructure and access to basic services. It is true that, within Africa as in other developing regions, population density is generally and strongly correlated with indicators of livability. For example, access to Crowded cities services is higher for African households in urban African cities are crowded in that they are packed with areas than in rural ones (Gollin, Kirchberger, and people who live in unplanned, informal downtown Lagakos 2016). But this relative advantage does not dwellings to be near jobs. Why? The immediate reason imply that cities are livable enough. Across Africa, 60 is that the urbanization of people is not accompanied percent of the urban population is packed into slums by an urbanization of capital (box 2). Housing, — much higher than the 34 percent seen elsewhere infrastructures, and other capital investments are (United Nations 2015a). lacking. Across the region, housing investment lags urbanization by nine years (Dasgupta, Lall, and Related to the predominance of informal housing near Lozano-Gracia 2014). African city centers is their relative lack of built-up area. For example, in both Harare, Zimbabwe and An underlying cause of this crowding is that African Maputo, Mozambique, more than 30 percent of land cities are not economically dense or efficient within five kilometers of the central business district enough to promote scale economies and attract remains unbuilt. This land near the core is not left capital investment. In principle, cities should benefit unbuilt by design in African cities, as it can be in well- businesses and people through increased economic developed downtowns such as Paris (which reserves density. Firms clustered in cities should be able to 14 percent of downtown land for green space, making access a wider market of inputs and buyers, with densely populated districts more livable). Instead, reduced production costs thanks to scale economies. outdated and poorly enforced city plans, along with dysfunctional property markets, create inefficient land use patterns that no one intended. The downtown lacks structures — despite being crowded. 3 Throughout Dar es Salaam, 28 percent of residents live at least three to a room This figure rises to 50 percent in Abidjan 16 Overview | Crowded, disconnected, and thus costly — Africa’s cities are limited to nontradables by their urban form BOX 2 Low capital investment in Sub-Saharan African cities during a period of rapid urban growth Africa’s cities are crowded because they lack formal, planned housing that is connected to jobs and services. Without sufficient formal development, informal settlements that are relatively central and thus close to jobs — such as Kibera in Nairobi, and Tandale in Dar es Salaam — are constantly growing in population. In Dar es Salaam, 28 percent of residents live at Housing investment in Africa has also lagged least three to a room; in Abidjan, 50 percent (World behind that in other low income and middle Bank 2015a, World Bank 2016). And in Lagos, income economies. Between 2001 and 2011, Nigeria, two out of three people dwell in slums African low income countries invested 4.9 percent (World Bank 2015b). of GDP in housing, compared with 5.5 percent elsewhere; and African middle income countries One factor in the crowding of Africa’s cities is invested 6.5 percent of GDP in housing, compared their lack of capital investment, which for the with 9 percent elsewhere (Dasgupta, Lall, and past four decades has remained relatively low Lozano-Gracia 2014). in the region, at around 20 percent of GDP. In contrast, urbanizing countries in East Asia — These figures underline the fact that Africa is China, Japan, the Republic of Korea — stepped up urbanizing while poor — indeed, strikingly poorer capital investment during their periods of rapid than other developing regions with similar urbanization. Between 1980 and 2011, China’s urbanization levels. In 1968, when countries in capital investment (infrastructure, housing, and the Middle East and North Africa region became office buildings) rose from 35 percent of GDP to 40 percent urban, their per capita GDP was 48 percent, while the urban share of its population $1,800 (2005 constant dollars). And in 1994, rose from 18 percent to 52 percent between when countries in the East Asia and Pacific region 1978 and 2012. In East Asia as a whole, capital surpassed the same threshold, their per capita investment remained above 40 percent of GDP at GDP was $3,600. By contrast, Africa, with 40 the end of this period. percent urbanization, today has a per capita GDP of just $1,000 (box figure 2.1). BOX FIGURE 2.1 Sub-Saharan Africa is urbanizing, but at lower levels of per capita GDP than other regions $3,617 41% 41% 37% 37% GDP $1,860 $1,806 Urbanization per capita rate (2005 dollars) $1,018 (percent) Latin America Middle East East Asia Sub-Saharan and the and North and the Africa Caribbean Africa Pacific (2013) (1950) (1968) (1994) Source: Estimations using United Note: Years in parentheses are those with available data in which the region was Nations 2014 and WDI 2014 for the closer to Sub-Saharan Africa’s present urban share of about 40 percent. In 1950 share of urban population, and WDI urbanization in Latin America and the Caribbean was 41 percent; in 1968 urbanization 2014 and Maddison Project to estimate in Middle East and North Africa was 41 percent; in 1994 urbanization in East Asia and GDP per capita. the Pacific was 37 percent; in 2013 urbanization in Sub-Saharan Africa was 37 percent. 17 Africa’s Cities | Opening Doors to the World FIGURE 3 Connections among people as a function of population near the city center: Nairobi, Kenya is more fragmented and less well-connected than Pune, India Nairobi (population 4.265 million) Pune (population 5.574 million) 0 - 364 365 - 103 1,033 - 22 2,295 - 40 4,042 - 67 6,735 - 97 9,777 - 13,984 13,985 - 19,534 19,535 - 27,747 27,748 - 39,955 Source: Henderson and Nigmatulina 2016. Note: The blue bars show the highest densities in the city. While these peaks 39,956 - 53,912 are concentrated in Pune, in Nairobi they are separated by lower densities. 53,913 - 71,973 18 Overview | Crowded, disconnected, and thus costly — Africa’s cities are limited to nontradables by their urban form Our analysis of imagery from satellites and geographic The lack of connections among neighborhoods information systems (GIS) confirms that in African means that African cities, compared with developed cities, capital investment not only appears low near and developing cities elsewhere, show both lower the urban core, but rapidly declines outside it. A stark exposure and higher fragmentation in connections contrast emerges between patterns of downtown among people living near the city center. population density — in which Africa largely resembles • Low exposure means that people are disconnected other regions — and of economic density (as reflected from each other. At a given distance (usually 10 in patterns visible from above that indicate capital kilometers), they cannot interact with as many investment). Africa’s generally low levels of urban people as in a city with higher exposure. capital investment also appear in the assessed worth of building stock. For example, the total economic • High fragmentation means that within a specified value of buildings in Dar es Salaam is estimated at area, population density varies widely: Its peaks around US$12 billion (Ishizawa and Gunasekera 2016), are scattered, not clustered in a way that could or just less than three times the city’s share of GDP. promote scale economies. Fragmentation increases Even lower are the estimated values for Nairobi, Kenya infrastructure costs, while it lengthens travel times ($9 billion) and Kigali, Rwanda ($2 billion). Compared among homes, job sites, and businesses. with cities in Central America, African cities have low According to a new study of 265 cities in 70 countries replacement values for their built-up area, built-floor that controls for total population and per capita GDP, area, and population. Thus, Nairobi has the highest average exposure near the center is 37 percent lower replacement value per square kilometer among the in African cities than in Asian and Latin American cities, four African cities studied, yet it is just 60 percent of while African cities are 23 percent more fragmented the value of Tegucigalpa, which has the lowest among (Henderson and Nigmatulina 2016). The contrast six Central American cities. between Nairobi, Kenya and Pune, India illustrates Although the capital investment shortfall that makes these differences (figure 3). African cities crowded appears across all building One pattern that explains the low exposure and high types, it is most severe in housing. In Nairobi, for fragmentation of African cities is their relative lack of example, commercial and industrial structures explain new development near the center. New construction 55 percent of the total value of building stock — even is not clustered to make capital more concentrated though these structures occupy just 4 percent of the and increase economic density. Instead, it tends to city’s area. Residential development is urgently lacking. push the boundaries of the city outward. In urban development language, this kind of building-out represents either expansion or leapfrog development; opposed to both is infill, which makes cities denser. • Expansion development enlarges a city’s footprint at Disconnected cities the edge of the consolidated urban area. While the lack of capital by itself might not always pose • Leapfrog development also enlarges the footprint, an obstacle to economic growth, African cities also but does so by establishing satellite areas — parcels are disconnected in that they are spatially dispersed. of newly built land that do not border on or overlap Structures are scattered in small neighborhoods. existing development. Without adequate roads or transport systems, commuting is slow and costly, denying workers access • Infill development is construction on unbuilt parcels to jobs throughout the larger urban area. People surrounded by existing developments. and firms are separated from each other and from Among the three types of new development, infill economic opportunity. And because urban form is is the best for economic exposure, or connections determined by long-lived structures that shape the city among people: It defragments the city and connects for decades — if not centuries — cities that assume a workers, jobs, and firms. Expansion and leapfrog disconnected form can easily become locked into it. development are the opposite: They are less likely to foster economic connections. Our analysis of GIS imagery for 21 African cities over 2000–2010 shows that, during this period, between 46 and 77 percent of new development occurred as expansion. The share of infill was typically much lower. 19 Africa’s Cities | Opening Doors to the World FIGURE 4 Leapfrog development: Undermining scale and agglomeration economies in African cities 60 1990–2000 percentage of total new patches 50 2000–2010 New leapfrog patches a 40 30 20 10 0 a t k y Su is ta la a To o ou s no ey do ar Lu bi ka ak ra m i la M la Ba to o Ka al ot go kr oe ub Bu uj ay ak d pa ya ro ku u ou c ga ak g m sa ug Ka ch Ad Ab na Ac ap Ki La w m dh N ai ia ua D Co N N in W N O Source: Construction based on data in Baruah (2015). Note: Leapfrog patches as a share of all new development patches, by city, 1990–2000 and 2000–2010. Leapfrog patches are defined as continuous built-up area that do not border or intersect with existing development. An even greater concern than the preference for In Nairobi, the average journey-to-work time is one expansion over infill development is the increase of the longest for 15 global cities studied (IBM 2011). in leapfrog development, which is now appearing Part of the reason is that walking accounts for a large outside various cities. In Bamako and Maputo, such share of commuting — in Nairobi about 41 percent leapfrog patches account for more than 50 percent (UNEP and FIA Foundation 2013). But even if more of the change to the urban fabric over 2000–2010. In city dwellers could afford transport by car or minibus, many other cities this share approaches or exceeds 40 commutes would remain impractical for lack of roads. percent (figure 4). The patches often being small, their In eight representative African cities, roads occupy far isolation from existing development will undermine lower shares of urban land than in other cities around city governments’ efforts to provide the networked the world. services that require scale economies — and that The deficiency of urban road infrastructure is made undergird urban productivity. worse by its extreme concentration near the core The prevalence of expansion and especially leapfrog of African cities, leaving outer areas disconnected. development is just one pattern that makes urban Our GIS study shows that in well-developed cities commuting challenging in African cities; another is outside Africa, land allocated to roads declines only deficient transport infrastructure. Traffic congestion gradually as one looks out from the center toward the can hobble the economy with long commuting times. periphery: An example is Paris (figure 5). By contrast, 20 Overview | Crowded, disconnected, and thus costly — Africa’s cities are limited to nontradables by their urban form FIGURE 5 Paved roads occupy a smaller share of urban land in Africa than elsewhere — and usually drop off abruptly beyond the city center Built-up Paved roads Open space Barcelona London 100 100 Percentage of urban land Percentage of urban land 80 80 60 60 40 40 20 20 CBD CBD -25 -20 -15 -10 -5 0 5 10 15 20 25 -25 -20 -15 -10 -5 0 5 10 15 20 25 Paris Kigali 100 100 Percentage of urban land Percentage of urban land 80 80 60 60 40 40 20 20 CBD CBD -25 -20 -15 -10 -5 0 5 10 15 20 25 -25 -20 -15 -10 -5 0 5 10 15 20 25 Africa’s urban roads are disproportionately clustered density, Africa’s city centers remain dominated by a near the center. In Addis Ababa, Dar es Salaam, Kigali, retail industry that does not benefit from economies and Nairobi, paved roads drop off so abruptly outside of specialization: For example, in Kigali and Kampala the downtown area that they nearly disappear (Dakar many urban workers purvey food and beverages. being a notable exception to the African pattern). The spatial fragmentation of Africa’s cities prevents Households in African cities find it difficult to settle firms from reaping both scale and agglomeration outside central business districts, as the lack of benefits. It prevents scale economies by reducing paved roads makes commuting from the periphery workers’ access to jobs, constraining firm size: Africa’s impractical (Felkner, Lall, and Lee 2016). urban firms employ 20 percent fewer workers on average than comparable firms elsewhere (Iacovone, Considered as a whole, the average urban area Ramachandran, and Schmidt 2014). In addition, spatial in Africa is not strikingly less built-up than its fragmentation hinders agglomeration economies by counterparts in other regions (except in Asia, where preventing job market pooling and matching and the cities are more densely built; Angel and others transfer of skills and knowledge — a special concern in 2011). What is lacking is the economically dense light of African cities’ low human capital endowments. concentration of capital and infrastructure investment Urban agglomeration economies thrive on knowledge that enables households to live decently and spillovers, which presuppose a mix of specialized affordably near jobs. Because of this lack of economic 21 Africa’s Cities | Opening Doors to the World FIGURE 5 (cont.) Paved roads occupy a smaller share of urban land in Africa than elsewhere — and usually drop off abruptly beyond the city center Built-up Paved roads Open space Dar es Salaam Nairobi 100 100 Percentage of urban land Percentage of urban land 80 80 60 60 40 40 20 20 CBD CBD -25 -20 -15 -10 -5 0 5 10 15 20 25 -25 -20 -15 -10 -5 0 5 10 15 20 25 Addis Ababa Dakar 100 100 Percentage of urban land Percentage of urban land 80 80 60 60 40 40 20 20 CBD CBD -25 -20 -15 -10 -5 0 5 10 15 20 25 -4 -2 1 3 5 7 9 11 13 15 17 19 21 23 25 Source: Based on Antos, Lall, and Lozano- Note: CBD = Central Business District. Data for European cities are from the European Gracia 2016 and Felkner, Lall, and Lee 2016. Environment Agency’s Urban Atlas data layers. Data for African cities are from very high resolution (0.5 m) satellite images taken in 2013. cognitive skills in the labor market. African urban land use is fragmented. Its transport infrastructure workers are relatively poor in such skills, according is insufficient, and too much of its development to results from the first initiative to measure skills in occurs through expansion rather than infill. While the low-income and middle-income countries (the World underlying causes of these problems are regulatory Bank STEP Skills Measurement Program). If workers and institutional, the effects of spatial fragmentation are to sort by ability — as they should to generate are material: It limits urban economies. agglomeration economies — then Africa’s cities will need, among other things, to restructure their labor Costly cities market by attracting and growing more specialized talent. In sum, the ideal city can be viewed economically as an efficient labor market that matches employers and job seekers through connections (Bertaud 2014). Fragmented urban forms impose high living costs on The typical African city fails in this matchmaker role. workers and households, resulting in indirect costs A central reason for this failure — one that has not and other constraints for firms: In short, African yet been sufficiently recognized — is that the city’s cities are costly both to live in and to do business in. 22 Overview | Crowded, disconnected, and thus costly — Africa’s cities are limited to nontradables by their urban form FIGURE 6 A fragmented urban form is associated with higher urban costs 0.4 0.2 Price index 0 -0.2 -0.4 -1 -0.5 0 0.5 1 People within 10 kilometers of average worker Source: Estimations using Nakamura Note: The figure shows a residual-on-residual plot. The x-axis depicts the residuals from a et al. (2016) and Henderson regression of the Puga10 Index, log scale (based on Henderson and Nigmatulina, 2016) controlling and Nigmatulina (2016). for log GDP per capita, log population, a dummy for SSA, and percentage of urban population. The y-axis plots residuals of the adjusted price index, log scale (based on Nakamura et al., 2016), on the same controls. The lower the people within 10 km of the average worker, the higher the price index. According to the new research underlying this report, Africa’s higher urban living costs appear in rents, the higher cost of living in African cities is related to food prices, and prices for other goods and services. their lack of dense spatial form and infrastructure City dwellers pay around 35 percent more for food connections (figure 6). Higher spatial densities appear in Africa than in low-income and middle-income to reduce costs: For example, a 1 percent reduction in countries elsewhere: a premium that looms larger spatial fragmentation measured by the Puga Index is given the high share of African household incomes associated with a 12 percent reduction in urban costs, that goes to food. Even higher differentials apply controlling for income levels and city population. to urban housing (55 percent higher in urban areas of African countries, relative to their income levels) While higher living costs directly affect workers, they and transport (42 percent higher in African cities ultimately are borne by urban firms. Higher wages than cities elsewhere, including vehicle prices and mean lower returns — unless workers are more transport services). Overall, urban households pay 20 productive. And without the economic density that to 31 percent more for goods and services in African gives rise to efficiency, Africa’s cities do not seem countries than in other developing countries (figure 7). to increase worker productivity. The result is that investment expectations remain low for cities in the region. 23 Africa’s Cities | Opening Doors to the World FIGURE 7 Urban living costs in Sub-Saharan African countries in 2011 exceeded costs elsewhere, relative to Africans’ lower per capita GDP 210 190 Sub-Saharan African countries Other economies 170 Fitted values Adjusted price level index 150 Angola Gabon S. Africa 130 Namibia Central Equatorial African Congo Guinea Republic Mauritius Botswana Swaziland Congo. 110 Dem. Rep. Guinea- Chad Ghana Bissau Mozambique Sudan Malawi Lesotho Zambia Liberia Togo Senegal Nigeria Burkina Faso Benin 90 Niger Cameroon Mali Kenya Rwanda Guinea Sierra Leone Mauritania Burundi Uganda 70 Gambia Madagascar Tanzania Ethiopia 50 6 7 8 9 10 11 12 Log of GDP per capita (2011 PPP$) Source: Nakamura et al. 2016, based Note: The adjusted price level index (PLI) for household on data from the 2011 International consumption excluding housing rent is standardized Comparison Program (ICP) and WDI. so that the average PLI equals to 100. PLIs for 15 Asian countries are inflated by 10 percent. 24 Overview | Crowded, disconnected, and thus costly — Africa’s cities are limited to nontradables by their urban form Urban workers in Africa incur high commuting costs higher nominal wages than urban firms in other — or they cannot afford to commute by vehicle at all, countries at comparable development levels: unit and must walk to work. The informal, often colorful labor costs are three times higher in Djiboutiville, minibus systems that dominate collective motorized Djibouti, than in Mumbai, India and 20 percent higher transit in most African cities are far from cost-efficient: in Dar es Salaam, Tanzania than in Dhaka, Bangladesh. The buses’ low load factor (passenger capacity) Cities in Africa are costly for households, workers, and prevents them from realizing scale economies. For the businesses. Because food and building costs are high, poorest urban residents especially, the cost of vehicle families can hardly remain healthy or afford decent transport in some cities is prohibitive, as measured in housing. Because commuting by vehicle is not only a study from 2008 (figure 8). The need to walk to work slow but expensive, workers find it hard to take and limits these residents’ access to jobs. keep jobs that match their skills. And the need for The high cost of living affects not just households but higher wages to pay higher living costs makes firms also firms, which have to pay higher wages in cities less productive and competitive, keeping them out of where the cost of living is high. In addition, urban tradable sectors. As a result, African cities are avoided workers may need to be compensated for poorer by potential regional and global investors and trading living conditions in informal settlements with scarce partners. amenities. Manufacturing firms in African cities pay FIGURE 8 Share of urban household budgets spent or needed for transport in 11 Sub-Saharan African countries (analysis from 2008) 100 Percentage of household budget Percent of household budget spent on transportation 80 Percent of bottom quintile household budget needed for two trips/day 60 40 20 0 s am i n la ou bi a a a ar al go as al ab ja pa ro ak g ug ou id la sh Ki La Ab ai m D Ab Sa do D n N Ka s Ki ga es di Ad ua ar O D Source: Kumar and Barrett 2008. 25 Africa’s Cities | Opening Doors to the World Closed for business, out of service: The urgency of a new urban development path for Africa African cities are crowded as well as disconnected, The reason why a firm in the nontradable sector making them costly for firms and for residents (see can afford to pay higher wages — while a firm in the figure 6). Potential investors and trading partners tradable sector cannot — is that the nontradable quickly see evidence of the physical and economic producer can raise its prices citywide. By doing so, it dysfunction that constrains public service provision, passes its own cost increases on to consumers in the inhibits labor market pooling and matching, and urban market. But such price hikes make the cost of prevents firms from reaping scale and agglomeration living in a city even higher, contributing to the workers’ benefits. So these potential partners stay away, urban costs. This sequence can become a vicious cycle fearing lack of return on their investment. that keeps African cities out of the tradable sector and limits their economic growth. The problem is not a simple one of underinvestment leading to low infrastructure, but a more complex one Often, proposed solutions to Africa’s urban challenges involving the interdependence of many investment focus simply on increased investments in structures decisions. Business investment decisions depend or on reforming urban planning. These actions are on the presence of other businesses — a firm’s necessary and urgent — but, by themselves, they customers and its suppliers — and of workplaces that are unlikely to lift cities out of the nontradables trap. can be reached from residential areas. Investment Why? because coordination failures tend to inhibit the will flow into housing if demand rises, driven by rising formation of new clusters of economic activity, which worker incomes. Infrastructure finance depends on are necessary for efficient tradables production (see, revenues from a growing city. All these investments among others, Henderson and Venables 2009). are interrelated, and in all of them expectations are Given the dynamics described above, no firm wants crucial. Investors’ low expectations become self- to be the first to enter the tradables sector. Yet many fulfilling when one investment fails to take place, would become established if they could coordinate reducing the expected return to others. The resulting their entry. To enable coordination, a city needs a vicious circle locks cities into a low development trap. credible coordination agent: either a forward-looking (The underlying analytic framework describing such group of firms that can harmonize their plans and traps is presented in Chapter 4.) make a move together, or a large-scale land developer or municipal government that can realize its vision Cities are “closed for business” through major infrastructure investment (Henderson A firm’s business decision to produce internationally and Venables 2009). Without such coordination the tradable goods and services will depend on its input move into tradables will fail, leaving the city “closed for costs. Among these input costs are urban costs: the business.” added costs that workers face when living in a city. Urban costs include rent, commuting costs, and the Cities are “out of service” high price of many goods. To attract workers, firms More than 60 percent of African’s urban population must raise wages to offset (or partially offset) these lives in areas with some combination of overcrowding, costs. Yet even as nominal wages climb to reflect high low-quality housing, and inadequate access to clean or rapidly rising urban costs, real wages remain low water and sanitation (United Nations 2015a). Why (see chapter 4 for detailed discussion). have cities in the region remained so deficient in When urban costs drive nominal wages too high, firms housing and basic services? will not be able to compete in the tradable sector A fundamental reason is that Africa’s urban and will produce only nontradables. The nontradable dysfunction is self-perpetuating: It lowers sector includes certain goods (beer and cement are expectations, and low expectations deter the examples), the construction trade, the retail trade, investments needed for improvement. Housing and many service sector activities, including informal investment decisions shape urban form. Providing sector employment. Demand for these goods and housing in the formal sector means deciding to sink services comes from income generated within the city costs in long-lived structures. And such decisions and its hinterland — but also from income transferred depend critically on expectations for a city’s future from outside, such as resource rents, tax revenues, prospects. Cities that inspire high expectations will and foreign aid. 26 Overview | Closed for business, out of service: The urgency of a new urban development path for Africa attract greater investment in formal sector structures, Path dependence and interdependence including residential structures, which reduce urban When a city appears “closed for business” and “out costs and in turn attract more investment. In contrast, of service,” potential partners stay away, fearing low cities that seem likely to remain artisanal — based to no returns. At present this vicious cycle of low on low-value nontradables production — foster expectations appears likely to keep Africa’s urban low expectations for the growth of land rents over economies undercapitalized, making the region’s time. With little incentive for investment in formal development all the more challenging. structures, a lack of capital investment keeps cities disconnected and urban costs high, perpetuating the Compounding this problem of low urban expectations cycle. is the reality of path dependence –identified in recent work as a central concern for policymakers. Cities Alongside the general effect of low expected returns, that grow inefficiently, without any effective plans or specific features of the business and regulatory incentives to integrate their physical form, are likely environment in African cities create further barriers to be locked into the resulting disconnected forms. to capital investment. These features include property Urban structures share a “putty-clay” quality: Once law and land use regulations, along with the design built, they are difficult to modify and can stay in place and enforcement of urban plans. for more than 150 years (Hallegatte 2009). In addition, Systems of property law and land ownership in Africa infrastructure investment needs to be planned well are often the first and most cumbersome regulatory in advance; if a growing city lacks a comprehensive, burden weighing on urban development. For example, forward-thinking plan to provide basic infrastructure a majority of the land in Kampala, Uganda operates services — sewerage, drainage, electricity, clean under a complex land tenure regime that recognizes water, and connectivity — it will have to add them independent rights over land and structures — giving later. That means adding them inefficiently and at far rise to legal disputes and blocking investment (Muinde greater cost, and as afterthoughts and in response to 2013). The problem takes a different form in Nigeria, piecemeal demand from individuals (Collier 2016). where urban land transactions incur high costs, As important as path dependence is interdependence and inefficient regulations further bog down formal among urban structures, infrastructure, and services. development. In Lagos and Port Harcourt, titling Much of a structure’s value reflects complementarities expenses alone can reach 30 percent of construction with other structures in the neighborhood or city. For costs, while total transaction costs range from 12 to 36 example, this report documents the benefits of road percent of a property’s value (World Bank 2015b). As a investments for private investments in residential and result, land is developed informally: In Ibadan in 2000, commercial structures (chapter 6). All social returns researchers found that 83 percent of homes violated on public infrastructure depend on the proximity city zoning rules (Arimah and Adeagbo 2000). of housing and premises: Thus, a rapid transit Urban plans are largely ineffective in Africa. One system is more viable at higher densities. Policies reason is that they are divorced from reality: They need to leverage these complementarities, avoiding typically do not consider finances, market dynamics coordination failures and single-sector interventions and interests, social diversity, or differences among that get in the way of economic density. income groups. Another reason is that, when enacted, Cities that continue on inefficient development paths regulations lack built-in implementation mechanisms. are growing, but in a counterproductive direction. As a result, human capacity constraints and financial Their physical structures and infrastructure will resource constraints preclude effective enforcement. not keep up with their rising population. As they More generally, the intentions and outcomes of continue to amass sunk capital — while passing up urban plans are distorted by institutional failure and opportunities for complementary investments that will fragmentation (across sectors and levels); by political never come again — they will sink deeper into the low interference; and by lack of consideration of a city’s development trap. And they might not dig themselves political economy. out. They could remain “out of service” and “closed for Inappropriate or unrealistic regulations and opaque business” forever. guidelines, especially on land ownership, impede access to land and discourage the formal development of city centers. Political risk can make future rents even more unpredictable. As a result, the returns from construction in Africa’s cities are intolerably uncertain — and cities remain “out of service.” 27 Africa’s Cities | Opening Doors to the World Springing cities from the low development trap We now understand more about the low development (box 3). When these systems pose barriers to urban trap in which African cities find themselves. They are land access, they impede the consolidation of plots crowded rather than economically dense, and they are and the evolution of land use. Firms cannot readily physically disconnected; as a result, they are costly. buy downtown land to convert it from low-density High costs deter investors through low expected residential use into higher-density apartments, or returns — while the city’s unlivable appearance vividly to build clusters of new commercial structures. corroborates these low expectations. As a result, the Land transactions are long, costly, and complicated urbanization of capital in Africa is lagging far behind (World Bank 2015c). Such market constraints reduce the urbanization of people. Migrants crowd into slums, the collateral value of structures, giving developers simply to be near where the jobs are. little incentive to invest in residential height — while tempting all parties to enter informal arrangements How can Africa’s leaders and policymakers spring (Collier 2016). cities from this trap? Crucially, they must first realize that the problem does not begin with low capital Formalizing land markets is essential; so is making investment and the lack of physical structures, or them work. Constraints on formal land markets even with undersized infrastructure. To be sure, contribute to the typical African city’s spatial low investment in structures limits urban economic fragmentation and to the relatively low capital density; it exacerbates spatial fragmentation, and it investment near its core. Not only will efficient land precludes agglomeration economies. But the lack of markets notably increase economic efficiency, they investment results from low investor expectations, will also help African cities tap the potential of rising which result when cities are spatially dispersed and land values to finance infrastructure and other public disconnected. goods. (But such financing strategies bear risk; they presuppose stable property rights and predictable law When potential investors and trading partners look at enforcement.) African cities, they see spatial fragmentation and a lack of connections. They know that such fragmentation While urban land markets need to work more constrains public service provision, inhibits labor efficiently, cities also must strengthen their urban market pooling and matching, and prevents firms plans and land use regulations. African cities today from reaping scale and agglomeration benefits. use planning models and regulatory codes that So the key to freeing Africa’s cities from their low may be relics of colonial regimes, or that may be development trap is to set them on a path toward uncritically imported from developed countries physical and economic density, connecting them for (Goodfellow 2013). Urban planning documents do not higher efficiency and boosting expectations for the give credible accounts of finance, market dynamics, or future. distributional impacts. Guidelines are not sufficiently articulated, granular, or transparent to support The first priority is to reform land markets and land consistent and enforceable development planning. use planning — to promote the most efficient uses of Capacity and resource constraints undermine urban land, and to develop land at scale. implementation. City and country authorities will need to add urban planning capacity — and to make tough Formalize land markets,clarify property rights, political decisions informed by technical evidence and and institute effective urban planning assessments. Informal land markets are not good enough for African Land use regulations, such as zoning ordinances cities. Urban land is a vital economic asset, and asset and building codes, are necessary to make urban transactions are viable only where purchasers can rely plans into realities. Although planners may promote on enduring extra-legal documentation of ownership. spatial density as a public good, the cost of investing A formal market both offers purchasers the protection in housing and commercial structures is borne by of the state and — because transactions are readily households and firms. (The benefits of economic observable and recorded — generates the public good density and exposure are an externality.) Because of accurate valuation. private actors on their own will not prevent market Clear rights to urban land are a precondition for failures in the allocation and use of land, urban land formal land markets. African cities struggle with use regulations must be clear and their enforcement overlapping and sometimes contradictory property- predictable. rights systems — formal, customary, and informal 28 Overview | Springing cities from the low development trap BOX 3 Urban land and property rights: A need for clarification Unclear land rights are severely constraining urban fiscal obligations, so lenders cannot always use land land redevelopment throughout Africa, imposing high as collateral. In Sub-Saharan Africa, only 10 percent costs. Under the customary rules for land tenure that of total land is registered (Byamugisha 2013). In control much urban and peri-urban land, property West Africa, only 2–3 percent of land is held with a rights depend on the consent of local chiefs or family government-registered title (Toulmin 2005). elders. One example is Durban, South Africa. Other The good news is that African countries are taking examples are in Ghana, Lesotho, Mozambique, and steps to clarify land rights. Botswana took the Zambia. Such cities often struggle with overlapping bold step of regularizing customary lands in 2008, and conflicting tenure systems — formal, customary partly because the Land Boards faced challenges to and informal. administering tribal land (Malope and Phirinyane Even where formal titles or clear land rights exist, 2016). Zambia passed a new planning bill in 2015, basic mapping, geographic or ownership information extending planning controls across state and is often inaccurate or land records maintained poorly, customary land and designating all local authorities causing disputes. Applying for formal recognition as planning authorities (Wesseling 2016). Namibia can also be tedious and costly (Toulmin 2005). In recognizes traditional leaders as part of the formal Mozambique one can apply for concession to a land system; they are designated by the president land plot from the relevant municipal directorate or and their details published in the government gazette municipal cadaster services. But the application can (United Nations 2015b). involve as many as 103 administrative steps over Some countries and cities are developing hybrid several years (UN-Habitat 2008). The lack of a proper regimes to make formal and customary administration registration system prevents urban land markets from more compatible. For example, in Nigerian states with functioning well, and it creates obstacles to the raising largely Muslim populations, the emir’s representatives of capital for development and investment — and to subdivide and allocate land with the help of volunteer the raising of revenue by the local authority. professionals from government: An example is the Across Africa, opaque and inadequate land databases city of Rigasa, in the extreme west of Kaduna (Igabi, and information systems distort land prices and Local Government Area, Nigeria). Future Urban availability. Finally, land administration systems (such redevelopers in Africa may learn from the past as registries and cadaster records) are incomplete and successes of two approaches — land sharing and land underused for enforcing legal claims and landholders’ readjustment — in several Asian cities. The market pricing of land depends partly on other infrastructure projects carry high sunk costs: Like policies besides land use regulations. Taxes, charges, any large structures, they depreciate very slowly and subsidies can be used to complement regulations, over decades or even centuries (Philibert 2007). And creating financial incentives and disincentives. the costs of developing housing, infrastructure, and Revenues — land taxes, for example — can also be industrial premises depend on sequencing. Consider used to finance administration and infrastructure. the relation of new transport systems and industrial And implementation tools such as capital investment, zones. If not coordinated with one another, and with budget, and phasing plans can assist upstream land markets and land use regulations, these projects planning. can put cities on a counterproductive development path. Make early and coordinated infrastructure Such large investments, especially at scale, will require investments — allowing for interdependence financing through new systems of revenue. Public among sites, structures, and basic services infrastructure projects incur costs far in advance of their benefits to productivity and livability, and the Research conducted for this study supports the value large capital outlays required can appear daunting. of early investments in neighborhood infrastructure The central government transfers on which African and services (chapter 6). But coordination among cities often rely will not suffice. City leaders, country these investments is equally crucial, given that cities authorities, and the international aid community are both path-dependent and interdependent. Large 29 Africa’s Cities | Opening Doors to the World BOX 4 Leveraging land values to finance Africa’s urban infrastructure Making Africa’s cities well connected and economically on the economic activity of the owner — so, unlike in dense will entail huge infrastructure investments. production, no owner behavior exists to be distorted.) Urban public finance in the region has traditionally Higher revenues from land and real estate can come relied on revenues from intergovernmental transfers. through: Future investments should leverage the value of city assets — mainly land — to finance infrastructure and • Improved valuation of land and properties closer to provide public goods and services. their market value, deepening the tax base. Land-based infrastructure financing will bring the • Improved enforcement of land and property taxes biggest payoff where cities are growing rapidly. on a larger number of owners, broadening the tax Rapid growth drives swift increases in land prices base. and creates large revenue opportunities. Yet it also • Monetization of underused public land. magnifies infrastructure investment needs, requiring major sources of development finance. Land-based Devising systems of land and real estate taxation financing has funded large leaps in the scale of urban that promote economic density is not easy. Strong investment in France, Japan, and the United States. institutions are needed to clearly define property rights; to ensure standardized and objective methods Taxes on land can fund investments while also of land valuation; and to support and oversee land promoting more efficient land use — giving management, land sales, and tax collection. For pure landowners an incentive to develop the land to its real estate taxes, policymakers should realize that most profitable use given the market value of their property values generally respond more slowly than property. Valuable downtown land will become other taxable wealth to annual changes in economic more densely developed, attracting investment in activity — while “property areas” respond still more residential and commercial structures. And land taxes slowly. are nondistortionary. (Appreciated land values are economic rents for a scarce resource, not a return should therefore study various financing options. One strong signal to other potential investors. It has been is to leverage land values (box 4) — although many argued that “investments sunk historically, even small cities in Sub-Saharan Africa are not currently allowed ones that have now depreciated completely, might to raise revenues from land (World Bank 2015c), and serve as a mechanism to coordinate contemporary their weak fiscal cadaster records and capacities pose investment” (Bleakley 2012). a further challenge. Decisions about a city’s growth pattern, based on Unregulated markets are unlikely to solve the underlying transport investment choices, will strongly problems of coordination, path dependence, and influence future greenhouse gas emissions and interdependence. Public policy and planning are environmental sustainability. Scholars have proven needed to get urban structures “right.” This imperative the impact of urban form on driving behaviors, modal is especially challenging in Africa, where fragmented choices, transport-related energy consumption, and urban development may already be locking cities into carbon dioxide emissions (Newman and Kenworthy high-cost paths. And since the low expectations that 1989). African cities now enjoy a unique opportunity come with high costs are self-fulfilling — expectations to avoid carbon-intensive urban transportation affect investments, which in turn affect expectations trajectories. Getting these choices right the first — cities that lack durable capital today may have an time — while urbanization is still in its early stages even harder time financing its acquisition tomorrow. — is critical. Given the path dependence of urban settlements, polluting now and cleaning up later is not Even if developers expect an African city to grow, they an option. might not know where growth is likely to occur — a type of coordination failure. One mechanism for In coordinating land use policies with infrastructure overcoming such failures is a sunk investment made plans, it is finally important to consider risk from by the government or a group of investors. Sunk natural hazards. While 70 percent of high-income investments can have long-run effects, sending a countries integrate land use with the management of 30 Overview | Opening the doors natural-hazard risk, only about 15 percent of low- and water supply, is essential (World Bank 2012b). income countries do so (World Bank 2012a). Yet cities Swakopmund, Namibia, a city of 42,000 surrounded by in these low-income countries are more vulnerable environmentally sensitive areas, limits development to natural hazards, including the floods that are to zoned “townlands” and has protected watersheds now so destructive in many parts of the world. with integrated environmental, sector, and land use Coordinating land use planning with the management planning. of natural resources, including water resources Opening the doors That African cities are crowded is apparent from the housing, commercial structures, or connective ground — both in the growth of informal settlements, infrastructure. Such cities are not just difficult and and in the traffic that snarls urban roads. That the costly to live in, but costly to do business in — they same cities are disconnected can be seen from scatter firms, prevent labor market pooling, and limit satellite images showing land use. And that these specialization across settlements. The urban economy cities are costly appears in price and wage data, as is restricted to nontradable, as opposed to tradable, interpreted by economic analysis. activity. This report explains the high costs of living and So long as Africa’s cities are in evident disarray, with doing business in African cities as consequences fragmented forms and dysfunctional markets, they will of their inefficient urban form. Distortions in factor continue to signal low expectations and stay in this low and product markets leave cities without adequate development trap. At best, they will proceed farther BOX 5 Building dense, connected, and efficient cities: Two models of success One model of successful urbanization is the high-speed rail — the bullet trains that have shrunk Republic of Korea, where urban planning and Korea into a half-day travel zone. land management institutions evolved to meet A different sort of success story is that of Bangkok, challenges at each stage of urbanization. Land where less restricted land markets were able to adapt development programs were established first, to growing demographic and economic pressures and followed by a land use regulation system. Then came climbing costs. Over 1974-88, when growth was rapid comprehensive urban planning, with guidelines for and land and housing construction prices on the rise, mandatory 20-year visions, zoning decisions, and developers responded by increasing the density of planning facilities. Downtown development projects their housing projects. Average units per hectare rose systematically adhered to phased scenarios under the from 35 to 56. Multifamily housing increased from comprehensive plans. Later, in the 1990s and 2000s, less than 2 percent of new construction in 1986 to 43 Korea integrated separate laws regulating urban and percent in 1990. With these shifts, developers were nonurban areas, and in 2000 it instituted metropolitan able to profit through the construction of affordable city-regional planning (between the city and the housing (Dowall 1992). Over 1986-90, almost half the county or province). Meanwhile, the government growth in Bangkok housing stock was from private initiated large-scale apartment construction projects development, while informally produced housing that solved Korea’s most serious urban housing composed a mere 3 percent of the total. In other problems. Multiple transport modes were developed. cities with highly constrained land markets, informally Road projects over time have included urban produced housing composed 20-80 percent of the highways and pavement projects as well as a network total (Dowall 1998). of expressways. And the nation’s rail network includes urban subway lines alongside traditional railroads and 31 Africa’s Cities | Opening Doors to the World along the inefficient path of slow and inadequate land Annex: development and infrastructure investment. Fortunately, the need for more efficient cities is easy Coverage of African to see and impossible to ignore. Africa’s urban areas cities used in the are quickly gaining in population: Home to 472 million people now, they will be twice as large in 25 years. The analysis most populous cities are growing as fast as 4 percent annually. Productive jobs, affordable housing, and effective infrastructure will be urgently needed for residents and newcomers alike. In urgency lies opportunity. Leaders can still set their cities onto more efficient development paths if they Cities used in the analysis act swiftly — and if they can resist flashy projects, steadfastly pursuing two main goals in order of Small cities priority: (<800,000) • First, formalize land markets, clarify property rights, Country City and institute effective urban planning. Benin Abomey-Calavi Burundi Bujumbura • Second, make early and coordinated infrastructure CAR Bangui investments that allow for interdependence among Côte d’Ivoire Bouake sites, structures, and basic services. Namibia Windhoek Nigeria Maiduguri A third goal is to improve urban transport and Nigeria Nnewi additional services. But this must not come ahead of Somalia Hargeysa the two goals listed above — nor can it be achieved South Africa Soshanguve unless those are met first. Sudan Nyala Zimbabwe Bulawayo Models of success from other regions may offer illuminating analogies and contrasts with African cities, while exemplifying what leaders can achieve through Intermediate cities coordinated and sustained action (box 5). Of course, (800,000-2 million) political economy must be considered in designing Angola Huambo and implementing policies. Leaders need to foresee Congo Pointe-Noire policy impacts (opportunities, winners, and losers) and DRC Bukavu anticipate challenges to enforcement. DRC Kananga DRC Kisangani City growth will be central to development in Africa, Eritrea Asmara as it has been elsewhere. By starting with reforms to Guinea Conakry land markets and regulations, then making early and Kenya Mombasa coordinated infrastructure investments, governments Liberia Monrovia can take control of urbanization and build more Malawi Blantyre-Limbe connected and productive African cities: cities that Malawi Lilongwe Mauritania Nouakchott open their doors to the world. Mozambique Maputo Niger Niamey Nigeria Benin City Nigeria Ilorin Nigeria Jos Nigeria Kaduna Nigeria Uyo Rwanda Kigali Sierra Leone Freetown Tanzania Mwanza Togo Lomé Uganda Kampala Zimbabwe Harare 32 Overview | Opening the doors Spain Turkey Tunisia Morocco Syria Iraq Algeria Libya Egypt Saudi Arabia Western Sahara Cabo Mauritania Mali Niger Chad Sudan Verde Eritrea Yemen Senegal The Gambia Burkina Guinea- Faso Djibouti Bissau Guinea Benin Somalia Sierra Leone Côte Togo Nigeria D’Ivoire Ghana Central African South Sudan Ethiopia Liberia Republic Cameroon Equatorial Guinea Uganda Large cities Sao Tome and Principe Kenya Gabon Democratic (>2 million) Republic of Rwanda Angola Luanda Congo Congo Burundi Burkina Faso Ouagadougou Seychelles Cameroon Douala Cameroon Yaoundé Tanzania Côte d’Ivoire Abidjan DRC Lubumbashi DRC Mbuji-Mayi Ethiopia Addis Ababa Angola Comoros Gambia Sukuta Malawi Ghana Accra Ghana Kumasi Zambia Madagascar Kenya Nairobi Madagascar Antananarivo Mali Bamako Nigeria Abuja Zimbabwe Mauritius Nigeria Ibadan Namibia Mozambique Nigeria Kano Nigeria Lagos Botswana Nigeria Port Harcourt Senegal Dakar Senegal Touba Swaziland Somalia Muqdisho (Mogadishu) South Africa Cape Town Lesotho South Africa Durban South Africa Johannesburg South Africa South Africa Pretoria Tanzania Dar es Salaam Zambia Lusaka 33 Africa’s Cities | Opening Doors to the World References Angel, Shlomo, Jason Parent, Daniel L. Civco, Henderson, J. V., and A. J. Venables. 2009. UNEP (United Nations Environmental Program), and Alejandro M. Blei. 2011. Making Room for “The Dynamics of City Formation.” Review and FIA Foundation. 2013. Share the Road: a Planet of Cities. Policy Focus Report, Lincoln of Economic Dynamics 12 (2): 233–254. Design Guidelines for Non Motorised Transport in Institute of Land Policy, Cambridge, MA. Africa. United Nations Environmental Program. Henderson, J. V., T. Regan, and A. J. Antos, Sarah E., Nancy Lozano-Gracia, and Venables. 2016. “Building the City: Sunk UN-Habitat. 2008. Mozambique Urban Somik V. Lall. 2016. “The Morphology of African Capital, Sequencing, and Institutional Sector Profile. Nairobi: UN-Habitat. Cities.” Draft. World Bank, Washington, DC. Frictions.” Draft, March 23. Venables, A. J. 2016. “Breaking into Tradables: Arimah, C. B., and D. Adeagbo. 2000. Henderson, Vernon, and Dzhamilya Nigmatulina. Urban Form and Urban Function in a Developing “Compliance with Urban Development and 2016. “The Fabric of African Cities: How to City.” University of Oxford, United Kingdom. Planning Regulations in Ibadan, Nigeria.” Think about Density and Land Use.” Draft, WDI (World Development Indicators). Habitat International 24: 279–94. April 20, London School of Economics. 2015. http://data.worldbank.org/data- Baruah, Neeraj. 2015. “Splintered and Iacovone, L., V. Ramachandran, and M. Schmidt. catalog/world-development-indicators. Segmented? Fragmentation of African Cities’ 2014. “Stunted Growth: Why Don’t African Firms Wesseling, T. 2016. New Approaches to Footprints.” Presentation at the Spatial Create More Jobs?” Working Paper 353, Center Physical Planning in Zambia. Royal Haskoning Development of African Cities Workshop, World for Global Development, Washington, DC. DHV. http://www.royalhaskoningdhv.com/ Bank, Washington, DC, December 16–17. IBM. 2011. “Global Commuter Pain en-gb/innovation/world-cities-day/new- Bertaud, Alain. 2014. “Cities as Labor Survey: Traffic Congestion Down, Pain approaches-to-physical-planning-in-zambia. Markets.” Working Paper 2, Marron Institute Way Up.” http://www-03.ibm.com/press/ World Bank. 2012a. Inclusive Green Growth: on Cities and the Urban Environment, us/en/pressrelease/35359.wss. The Pathway to Sustainable Development. New York University, New York. Ishizawa, O., and R. Gunasekera. 2016. Washington, DC: World Bank. Bleakley, H., and J. Lin. 2012. “Portage “Economic Values of Buildings in Four African ———. 2012b. The Future of Water and Path Dependence.” Quarterly Journal Cities.” Background paper for this report. in African Cities: Why Waste Water? of Economics 127 (2): 587–644. Kumar, Ajay, and Fanny Barrett. 2008. Washington, DC: World Bank. Byamugisha, F. 2013. Securing Africa’s Land for “Stuck in Traffic: Urban Transport in ———. 2015a. Measuring Living Standards Shared Prosperity. World Bank, Washington, DC. Africa.” AICD Background Paper. within Cities. Households Surveys: Dar es Salaam Collier, Paul. 2016. African Urbanization: Malope P., and M. Phirinyane.2016. and Durban. Washington, DC: World Bank. An Analytic Policy Guide. Fourth Seminar in “Enhancing Property Rights through Land ———. 2015b. Nigeria Urbanization TICAD Seminar Series, “Land Use Planning Tenure Regularization in Botswana.” Review: From Oil to Cities: Nigeria’s Next and Spatial Development for Smart Growth Paper prepared for presentation at the Transformation. Washington DC: World Bank. in African Cities,” World Bank Tokyo. 2016 World Bank Conference on Land and Poverty. World Bank, Washington, DC. ———. 2015c. Stocktaking of the Housing Dasgupta, B., S. V. Lall, and N. Lozano- Sector in Sub-Saharan Africa: Challenges and Gracia. 2014. “Urbanization and Household Muinde, Damaris Kathini. 2013. “Assessing Opportunities. Washington, DC: World Bank. Investment.” Policy Research Working Paper the Effects of Land Tenure on Urban 7170, World Bank, Washington, DC. Developments in Kampala.” March. ———. 2016. Cote d’Ivoire Urbanization Review. Diversified Urbanization. Dowall, D. E. 1992. “A Second Look Nakamura, S., R. Harati, S. Lall, Y. Dikhanov, Washington, DC: World Bank. at the Bangkok Land and Housing N. Hamadeh, W. V. Oliver, M. O. Rissanen, Market.” Urban Studies 29 (1): 25–37. and M. Yamanaka. 2016. “Is Living in African Cities Expensive?” Policy Research Working ———. 1998. “Making Urban Land Markets Work: Paper 7641, World Bank, Washington, DC. Issues and Policy Options.” Prepared for seminar on Strategy on Urban Development and Local Newman, P. W., and J. R. Kenworthy. 1989. Governments, World Bank, Washington, DC. “Gasoline Consumption and Cities.” Journal of the American Planning Association 55 (1): 24–37. Felkner, John S., Somik V. Lall, and Hyun Lee. 2016. “Synchronizing Public and Philibert, Cédric. 2007. “Technology Private Investment in Cities: Evidence from Penetration and Capital Stock Turnover: Addis Ababa, Dar es Salaam, Kigali and Lessons from IEA Scenario Analysis.” Nairobi.” World Bank, Washington, DC. International Energy Agency, Paris. Gollin, Douglas, Remi Jedwab, and Rosenthal, Stuart S., and William C. Strange. Dietrich Vollrath. 2016. “Urbanization with 2004. “Chapter 49 Evidence on the Nature and without Industrialization.” Journal and Sources of Agglomeration Economies.” of Economic Growth 21 (1): 35–70. In Handbook of Regional and Urban Economics 4: 2119–71. Amsterdam: Elsevier. Gollin, Douglas, Martina Kirchberger, and David Lagakos. 2016. “Living Standards across Toulmin, C. 2005. Securing Land and Property Space: Evidence from Sub-Saharan Africa.” Rights in Sub-Saharan Africa: The Role of Local March 31. Available at https://collaboration. Institutions. Geneva: World Economic Forum. worldbank.org/docs/DOC-20505. United Nations. 2014. World Urban Prospects: Goodfellow, Tom. 2013. “Planning and The 2014 Revision. New York: United Nations. Development Regulation amid Rapid Urban ———. 2015a. Millenium Development Goals Growth: Explaining Divergent Trajectories Indicators. Indicator 7.10 Proportion of Urban in Africa.” Geoforum 48: 83–93. Population Living in Slums. http://mdgs.un.org/ Grover Goswami, A., and S. V. Lall. 2016. “Jobs unsd/mdg/seriesdetail.aspx?srid=710. and Land Use within Cities: A Survey of Theory, ———. 2015b. Thirteenth to Fifteenth Evidence, and Policy.” Policy Research Working International Convention on the Elimination Paper 7453, World Bank, Washington, DC. of All Forms of Racial Discrimination (ICERD) Hallegatte, Stephane. 2009. “Strategies to Periodic Report by Namibia. United Nations Adapt to an Uncertain Climate Change.” International Convention on the Elimination Global Environmental Change 19 (2): 240–47. of All Forms of Racial Discrimination. 34 Part 1 Crowded and Disconnected African Cities Chapter 1 Crowded with people, not dense with capital Crowded with people Not dense with physical capital Not dense with human capital Chapter 2 Disconnected land, people and jobs Disconnected land People not connected to people: High fragmentation, low exposure, little potential for interaction People not connected to jobs 35 Africa’s Cities | Opening Doors to the World When well managed, city growth can spur economic Successful cities are economically dense and growth and productivity in two ways. First, it can boost connected. Africa’s cities are crowded but not incentives for investment through higher economic dense, and they are disconnected. Investments in density and proximity, which support clusters of firms infrastructure, industrial and commercial structures, and increase connectivity among workers. Second, and affordable formal housing have not kept pace it can make cities more livable for poor and middle- with the concentration of people. Congestion and its class residents, by providing affordable services, costs overwhelm the benefits of urban concentration. amenities, and housing. Through both channels, a Cities have developed as collections of small and successfully developed city offers firms the incentives fragmented neighborhoods. The lack of connectivity, of agglomeration and high returns on investment. particularly the absence of reliable transportation, limits job opportunities for workers and prevents Africa’s cities share neither of these attributes. They firms from reaping scale and agglomeration benefits. are predominantly local, lacking global or even Chapters 1 and 2 examine each of these themes, regional reach. They are unsuccessful because they describing the symptoms of Africa’s unsustainable have developed along a different trajectory — one that urban development path. imposes heavy costs on residents and firms and locks them in a low-growth poverty trap. Africa is far poorer than other developing regions were when they reached similar levels of urbanization. Urbanizing while poor means that cities lack the means to invest in physical capital, such as infrastructure. They have fewer resources for public services, such as schooling, and lack human capital, which hinders efficient administration and improvement of existing institutions. 36 Chapter 1 Crowded with people, not dense with capital Urbanization benefits people and businesses by increasing economic density. A worker in an economically dense area can commute more easily and consume a broader range of products. Firms clustered in cities can access a wider market of inputs and buyers. Scale economies reduce firms’ production costs — in turn benefiting consumers. 37 Africa’s Cities | Opening Doors to the World Population density is indeed strongly correlated with and other capital investments are lacking, especially indicators of livability — in Sub-Saharan Africa as outside the city center. Urban building stocks have low elsewhere (Gollin, Kirchberger, and Lagakos 2016). Yet replacement values. Across Africa, housing investment Africa’s cities are not economically dense or efficient. lags urbanization by nine years (Dasgupta, Lall, and They are crowded and unlivable. Most urban residents Lozano-Gracia 2014). are packed into low-rise, informal settlements without The population density of African cities is similar to adequate infrastructure or access to basic services. that of many cities elsewhere. What is holding these Two of every three people in Lagos, Nigeria dwell cities back is their low economic density — the lack in slums (World Bank 2015b). Thus, even though of thriving urban markets that depend on adequate households in densely populated areas of Africa are infrastructure and conveniently connected clusters better supplied with services than rural households, of residential and commercial structures. A dearth the mere fact of higher population density does not of capital and capital investment keeps Africa’s cities imply a livable environment. inefficient and less productive than they should be, Why do a majority of people in Africa’s cities live limiting firms and workers to the production of goods in slums? The immediate explanation is that the and services for small and local hinterland markets urbanization of people has not been accompanied by locking them out of much more lucrative regional and the urbanization of capital. Housing, infrastructure, international markets. Crowded with people Many of Africa’s urban workers live in crowded Slums: Workers’ only option when urban quarters near the city center. In Dar es Salaam, Tanzania, 28 percent of residents are living at least economic density is low but highly three to a room (World Bank 2015c); in Abidjan, Côte concentrated d’Ivoire, the figure is 50 percent (World Bank 2016). The crowdedness of African cities is most apparent in The reason for this crowding is that most people their slums. On average, 60 percent of Africa’s urban must live near the downtown district or industrial population is packed into slums — a far larger share zones if they hope to work. They cannot conveniently than the average 34 percent seen in other developing commute from outlying areas, because little or no countries (United Nations 2015). affordable transportation is available. During 1960–72, Sub-Saharan Africa and Latin Africa’s cities also suffer from a lack of adequate America had similar incidences of slums and squatter formal housing around the urban core. Consequently, settlements (table 1.1). More recently, slum levels in people settle in relatively central informal settlements Sub-Saharan Africa (62 percent) and Latin America that are densely populated, ill served by urban (24 percent) diverged. South Asia’s levels (35 percent) infrastructure, and, by many measures, unlivable. are similar to its historical levels, although some cities Paradoxically, Africa’s cities are sparsely built and laid have experienced large increases in slum proportions. out but feel crowded. Table 1.1 Slum population as percentage of total urban population in selected cities, historically and in 2014 Share of city population Share of national population Region/country City living in slums living in slums in 2014 Sub-Saharan Africa Cameroon Douala 80 (1970) 37.8 Yaoundé 90 (1970) Côte d’Ivoire Abidjan 60 (1964) 56.0 Ethiopia Addis Ababa 90 (1970) 73.9 Ghana Accra 53 (1968) 37.9 Kenya Nairobi 33 (1970) 56.0 Mombasa 66 (1970) 38 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital Table 1.1 Slum population as percentage of total urban population in selected cities, historically and in 2014 Share of city population Share of national population Region/country City living in slums living in slums in 2014 Liberia Monrovia 50 (1970) 65.7 Madagascar Tananarive 33 (1969) 77.2 Malawi Blantyre 56 (1966) 66.7 Nigeria Ibadan 75 (1971) 50.2 Senegal Dakar 60 (1971) 39.4 Somalia Mogadishu 77 (1967) 73.6 Sudan Port Sudan 55 (1971) 91.6 Tanzania Dar es Salaam 50 (1970) 50.7 Togo Lomé 75 (1970) 51.2 Burkina Faso Ouagadougou 70 (1966) 65.8 Zaire Kinshasa 60 (1969) 74.8 Zambia Lusaka 48 (1969) 54.0 Low-income Asia Afghanistan Kabul 21 (1971) 62.7 India Calcutta 33 (1971) 24.0 Bombay 25 (1971) Delhi 30 (1971) Madras 25 (1971) Baroda 19 (1971) Indonesia Jakarta 26 (1972) 21.8 Bandung 27 (1972) Makassar 33 (1972) Nepal Katmandu 22 (1961) 54.3 Pakistan Karachi 23 (1970) 45.5 Sri Lanka Colombo 43 (1968) 24.8 (1990) Latin America and the Caribbean Brazil Rio de Janeiro 30 (1970) 22.3 Belo Horizonte 14 (1970) Recife 50 (1970) Porto Alegre 13 (1970) Brasilia 41 (1970) Chile Santiago 25 (1970) 9.0 (2005) Colombia Bogotá 60 (1969) 13.1 Cali 30 (1969) Buenaventura 80 (1969) Ecuador Guayaquil 49 (1969) 36.0 Guatemala Guatemala City 30 (1971) 34.5 Honduras Tegucigalpa 25 (1970) 27.5 Mexico Mexico City 46 (1970) 11.1 Panama Panama City 17 (1970) 25.8 Peru Lima 40 (1970) 34.2 Arequipa 40 (1970) Chimbote 67 (1970) Venezuela Caracas 40 (1969) 32.0 (2005) Maracaíbo 50 (1969) Barquisimeto 41 (1969) Cíudad Guayana 40 (1969) Source: City data are from Linn 1979; Note: A slum household is defined as a group of individuals living under the national data are from the United Nations’ same roof in a dwelling that lacks one or more of the following conditions: Millennium Development Goals database. access to improved water, access to improved sanitation, sufficient living area, durability. Figures in parentheses are years of the statistic. 39 Africa’s Cities | Opening Doors to the World FIGURE 1.1 Very high proportions of city dwellers live in slums in Africa 100% 80% 60% 40% 20% 0% South Sudan Sudan Chad Guinea-Bissau Mozambique SSA (Av) Burundi Swaziland Non-SSA (Av) Zimbabwe South Africa CAR STP Mauritania Madagascar Sierra Leone DRC Ethiopia Somalia Niger Comoros Malawi Equatorial Guinea Burkina Faso Djibouti Mali Cote d’Ivoire Kenya Angola Zambia Uganda Rwanda Togo Lesotho Tanzania Nigeria Congo Guinea Senegal Ghana Cameroon Gabon Gambia Namibia Source: United Nations 2015. Note: Data are latest available for each country. FIGURE 1.2 Population density in African cities is lower than in some other regions 16,000 12,000 People per sq km 8,000 4,000 0 South Asia Middle East and Sub-Saharan East Asia Latin America Europe and North America North Africa Africa Pacific and the Central Asia Caribbean Source: Data from Demographia 2015. 40 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital High rates of slum living within urban areas are and economic densities are higher near the city characteristic of most African countries. Only two center and then fall steeply as one moves away. One countries, Zimbabwe and South Africa, fall below kilometer away from the city center, average ambient the non-African average (figure 1.1). The proportion population density falls 7 percent in African cities — a of Africans living in slums is not high because Africa much more rapid rate of decline than the 4 percent has higher urban population densities than other in cities elsewhere. Economic activity as measured by countries. The average population density of African nighttime light intensity falls at a rate of 15 percent cities tracks the global average; it ranks third among in African cities, compared with 11 percent in non- seven global regions (figure 1.2).1 African cities. In most African cities, the rate of decay of density gradients is above the sample median. High population density at the city’s core, Further, while most peak economic densities are below the median, most peak population densities are rapid tapering on the outskirts above the median. Africa does not stand out in a list of the world’s 50 densest cities (only 6 African cities make the list, most People are clustering in downtown locations not of them in the Democratic Republic of Congo). And because of the amenities or decent jobs they can although peak population density is higher in Africa access in central locations. These patterns reflect than elsewhere, the difference is not statistically broader dysfunctionalities in land markets as well significant. Using a sample of 265 cities in 3 regions, as limited investments in transport infrastructure, Henderson and Nigmatulina (2016) estimate that peak limiting the choices that people can make on where urban population densities are higher in Africa (57,600 to live and how to access jobs (see chapter 5). For people per square kilometer) than in Latin American example in Nairobi, the current patterns of informal (37,700) or Asia (50,900). In a sample of 39 cities in land use in centrally located settlements impose developing and developed countries, Henderson and an economic loss of over $1.8 billion (Henderson, Nigmatulina (2016) finds peak densities of 55,700 Venables, and Regan, 2016). And in Dar es Salaam, people per square kilometer in Africa and 50,800 poor residents face squalid living conditions in other regions. In most African cities, population alongside low earning potential in downtown informal settlements (box 1.1). Not dense with capital Capital investment in Africa over the past 40 Supporting rising population densities in African years has averaged about 20 percent of GDP. In cities will require investments in buildings, and contrast, urbanizing countries in East Asia — China, complementary physical infrastructure: roads, Japan, the Republic of Korea — stepped up capital drainage, street lighting, electricity, water, and investment during their periods of rapid urbanization. sewerage, together with policing, waste disposal, and Between 1980 and 2011, China’s capital investment health care. In the absence of higher levels of capital (infrastructure, housing, and office buildings) rose investment at around Asian levels, the potential from 35 percent of GDP to 48 percent; during roughly benefits of Africa’s cities are being overwhelmed by the same period (1978–2012), the urban share of its crime, disease, and squalor. population rose from 18 percent to 52 percent. In East Overcrowding increases exposure to communicable Asia as a whole, capital investment remained above 40 diseases. Inadequate drainage increases the risk percent of GDP at the end of this period, helping the of malaria, and lack of sanitation raises the risk of region become very dense economically. dengue (Sclar, Garau, and Carolini 2005). Lack of These contrasts underline the fact that Africa is access to clean water is a leading cause of diarrhea, urbanizing when poor — indeed, strikingly poorer than which is responsible for an estimated 21 percent other developing regions with similar urbanization of deaths among children under five in developing levels. In 1968, when countries in the Middle East and countries — 2.5 million deaths a year (J-PAL 2012). North Africa region became 40 percent urban, their Evidence from a randomized evaluation of a Mexican per capita GDP was $1,800 (in 2005 constant dollars). housing program shows how improving floor quality In 1994, when countries in the East Asia and Pacific in cities reduced the incidence of intestinal parasites, region hit the 40 percent urbanization mark, their per diarrhea, and anemia among young children while capita GDP averaged $3,600. By contrast, per capita improving their cognitive development and their GDP in Africa is just $1,000. 41 Africa’s Cities | Opening Doors to the World BOX 1.1 Life in Africa’s cities is often miserable: An urban migrant’s story In 1989, Fatma and her husband, Peter, sought to Why do they live in such conditions? Peter cannot escape poverty by moving from rural Mbeya to Dar es afford transportation, so he walks wherever he goes. Salaam. In many ways Dar es Salaam is better: Work is He leaves home every morning at 5 o’clock, walking to sometimes available, and services exist. But in some the city center, where he seeks day jobs as a mason. If ways life is more hazardous than it was in Mbeya. there is no work — and often there is not — his family goes hungry. But if the family were living farther from Despite savings, Fatma and Peter could not afford the downtown, Peter would struggle to find work at all. formal housing. To be close to work opportunities, they settled in a hazardous slum. Since then, rising Fatma stays at home taking care of the couple’s six housing costs or eviction have forced them to move children. She would like all of them to attend school, more than a dozen times. but only one does, because places for the others are lacking. Today they live in Tandale, one of the city’s largest slums. Built on a dumpsite just a few meters from Despite low wages and high unemployment the choked and contaminated river, their home (estimated at 22 percent of the active labor force in sinks into the garbage a few inches every year. They 2014), Dar es Salaam is a relatively expensive city. share an open toilet with their neighbors and live in extreme poverty. 42 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital A meal costs twice as much as in Bangkok — even though average income in Bangkok is 23 times higher. Most of Fatma and Peter’s income goes to food. Transportation, education, and health care are unaffordable. Like millions of Africans, Fatma and Peter have had to give up livability in their quest for opportunity. The only central location they can afford is full of dangers. When Fatma and Peter migrated to the city, Dar es Salaam was about one-third of its size today. It now has more than 4.3 million residents, and it is projected to reach 10 million or more by the early 2030s, making it a “megacity.” Many more families like Fatma’s may soon be living in Dar es Salaam. Source: World Bank 2015c. To avoid a future in which Fatma and Peter’s experience is the normal one, African governments must work with urban planners and development specialists to answer two key questions: • How can the city ensure housing and transportation for the urban poor, allowing them to work without living in danger? • How can high costs for people and businesses be reduced, making African cities kinder to their residents, enabling businesses to break into world markets, and lifting people like Fatma and her family out of poverty? 43 Africa’s Cities | Opening Doors to the World mothers’ mental health and happiness (Cattaneo access to water improved household welfare by and others 2009). The program was associated with freeing up the time usually spent to fetch water. a 20 percent reduction in the presence of intestinal Tensions over water-related issues (such as shortages) worms, a 13 percent reduction in the prevalence of decreased, and people reported spending more time diarrhea, and a 20 percent decline in the prevalence on leisure and social activities. The program increased of anemia. Children 12- to 30-months-old scored 30 reported happiness and improved social integration percent higher than controls on a communicative (Devoto and others 2012). development test. Improved access to electricity improves health Further, improved access to piped water in urban outcomes by reducing the inhalation of kerosene areas has been shown to decrease infant mortality. fumes and the risk of kerosene-related accidents, Galiani, Gertler, and Schardrogsky (2005) show that a frequent cause of burns. It also allows children to the privatization of water services in Greater Buenos study after dark (see Wu, Borghans and Dupuy 2010 Aires improved young children’s health outcomes by on Indonesia; Spalding-Fecher 2005 on South Africa; improving service quality and access to safe water, and UNDESA 2014). especially in the poorest areas, which benefited most from service expansion. Child mortality from Not dense with buildings infectious diseases dropped about 8 percent in areas Evidence of overcrowding and low levels of capital that privatized the services, with the largest decline investment appear in the value of building stocks. (26 percent) in the poorest areas. Gamper-Rabindran, Disaster risk profiles were developed for four African Khan, and Timmins (2010) show that infant mortality cities. This methodology estimates the economic value in urban areas in Brazil declines as living standards of building stock and its distribution across the city improve. at 1 square kilometer resolution. The total economic In urban Morocco, connecting houses to water had value of buildings in Dar es Salaam is estimated no sizable effect on health outcomes, but convenient at about $12 billion — about five times the city’s contribution to GDP. Even lower are the estimated FIGURE 1.3 The value of building stocks in African cities is low Replacement values per Replacement values per Replacement values km2 of land area ($m) km2 of built floor area ($) per person ($) 100 500 25,000 80 400 20,000 60 300 15,000 40 200 10,000 20 100 5,000 0 0 0 Kigali Addis Ababa Dar es Salaam Nairobi Tegucicalpa San Salvador Managua Guatemala City San Jose Panama City Nairobi Kigali Addis Ababa Dar es Salaam Tegucicalpa Managua San Salvador Guatemala City Panama City San Jose Kigali Addis Ababa Dar es Salaam Nairobi Tegucicalpa Managua San Salvador Guatemala City Panama City San Jose Source: Ishizawa and Gunasekera 2016. Sub-Saharan Africa Central America 44 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital values for Nairobi, Kenya ($9 billion); Addis Ababa, Not dense with amenities, not livable Ethiopia ($6 billion); and Kigali, Rwanda ($2 billion) In most of the developing world — from India and (Ishizawa and Gunasekera 2016). Vietnam to Brazil and Colombia — large cities usually The value of urban building stocks is much lower in have better living standards than smaller cities. In Africa than in Central America (figure 1.3). The value countries where governments are still struggling to per square kilometer in the four African cities studied bring water and sanitation to most of the population, ranges between $2.7 million (in Kigali) and $15.6 large cities usually fare better. In Brazil, where million (in Nairobi). In contrast, the range in Central access to piped water is converging toward universal America is $27.8–$90.4. The results are similar when coverage, smaller cities lag behind metropolitan areas, other measures or value are compares. As elsewhere with less than 90 percent coverage. In India, coverage in the world, residential and commercial buildings are levels reach 70 percent of the population in the largest concentrated downtown (box 1.2). cities reach but less than 50 percent in the smallest cities. BOX 1.2 The most valuable building structures are concentrated near the city center Building values are highly concentrated near the city center. In Nairobi, Kigali, and Dar es Salaam, more than 20 percent of the value of buildings in the city is concentrated in the square kilometer around the city center. Addis Ababa is less concentrated, although a small peak exists there, too (box figure 1.2.1). BOX FIGURE 1.2.1 Nairobi Kigali 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Dar es Salaam Addis Ababa 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 18 20 Source: Ishizawa and Gunasekera 2016. Non-residential Residential 45 Africa’s Cities | Opening Doors to the World Coverage in Africa is lower — and it is not, on average, At the national level, density and livability seem to be higher in larger cities. In Uganda, for example, correlated in Africa. But within cities, the relationship coverage hovers around 20 percent for large, medium, is not straightforward. Although households in denser and small cities (World Bank 2013). areas are better off than households in other areas of the country, living standards drop when crowding Two patterns of living standards (measured by supplants density. household durable assets, housing conditions, and infant survival rates) are evident in Africa (Gollin, Population densities are rising and cities expanding Kirchberger, and Lagakos 2016): rapidly in Africa: The largest cities are growing by as much 4 percent a year (United Nations 2014). • A wide variation of living standards is observed Infrastructure investments have remained low, within countries, with urban areas ranking however, making density look a lot like crowding. consistently higher across indicators. African cities have failed to absorb large increases • The links between density and living standards in population successfully because of lack of basic are strong in African countries, with a significant services, limited budgets in expensive cities, and and positive correlation found consistently across limited connective infrastructure. countries.7 During the mid-1970s, the share of the population Although population density and living standards are with access to water supply and sewage disposal correlated, higher population density does not imply a was similar in Saharan Africa and South Asia. Since livable environment. then, the gap between the two regions has widened dramatically (table 1.2). In urban areas in 2015, 88 Access to electricity increases as density rises in all percent of residents in South Asia but just 64 percent 20 of the Sub-Saharan African countries studied. The in Sub-Saharan Africa had access to improved water. cost to provide piped water in dense urban areas is The gap was even larger for improved sanitation, an estimated $0.70–$0.80 per cubic meter — far less which reached 71 percent of urban residents in South than the $2 per cubic meter in sparsely populated Asia but just 46 percent of urban residents in Sub- areas (Kariuki and Schwartz 2005). Data for Colombia Saharan Africa. suggest that the operating costs for solid waste disposal services in large and dense metropolitan areas can be one-third to one-half the cost in smaller, less dense cities (Trojanis and Lozano-Gracia 2016). Table 1.2 Percent of population with access to water and sanitation, by region Mid-1970s 2015 Access to sewage Access to Access to improved Access to water disposal improved water sanitation Region Urban Rural Urban Rural Urban Rural Urban Rural Sub-Saharan Africa 66 10 70 14 89 64 46 27 South Asia 66 17 67 3 93 88 71 55 East Asia and Pacific 58 10 67 15 96 87 85 69 Latin America and the Caribbean 78 35 80 25 97 89 88 77 Source: Data for 1970s are from Linn Note: Access to an improved water source refers to the percentage of the population using an 1979. Current data are from World improved drinking water source (including piped water on premises (piped household water Development Indicators 2016. connection located inside the user’s dwelling, plot or yard), and other improved drinking water sources (public taps or standpipes, tube wells or boreholes, protected dug wells, protected springs, and rainwater collection). Access to improved sanitation facilities refers to the percentage of the population using improved sanitation facilities, that is, likely to ensure hygienic separation of human excreta from human contact. They include flush/pour flush (to piped sewer system, septic tank, pit latrine), ventilated improved pit (VIP) latrine, pit latrine with slab, and composting toilet. 46 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital Analysis of low- and lower-middle-income countries shows how urban areas of Sub-Saharan Africa perform with respect to other regions. The analysis compares three measures: improved water, improved sanitation, and mortality caused by road traffic injury. The empirical estimation is specified as follows: WDI = θ11 Regdummy + θ2 log gdppc + θ3 log landarea + θ4 %urban + εi where WDI is the World Development Indicator mortality from traffic congestion. Europe and Central measure; Regdummy is a dummy taking a value of Asia and the Middle East and North Africa have much 1 if the worker lives in South Asia (the base case) better access to improved sanitation than South Asia. and different values for other regions; log gdppc Even controlling for poverty and urbanization levels, denotes the log value of GDP per capita in the country, Sub-Saharan Africa lags other developing regions in adjusted for purchasing power parity; log landarea these urban development indicators. gives the log value of the country’s land area in square Research based on household surveys for 20 kilometers; and % urban denotes the percentage of countries suggests that across Sub-Saharan Africa population residing in urban areas. the proportion of households with shelter and The results indicate that South Asia (the base case) infrastructure amenities varies not only by population has more than 15 times greater access to improved density but also by amenity type, urban area type, and sanitation than Sub-Saharan Africa and much lower household expenditure quintile (table 1.3). Table 1.3 Housing amenities in Sub-Saharan Africa, by type of area and expenditure quintile (percent of all households) Type of area Expenditure quintile Other Variable All areas Largest city Rural Bottom Top urban Shelter amenities Roof 16.1 18.6 19.6 13.7 15.4 18.0 Walls 65.8 90.3 80.0 53.0 58.7 72.1 Floor 52.8 91.4 77.0 35.2 40.7 65.8 Housing infrastructure Electricity 31.6 69.4 49.5 13.4 20.1 43.0 Toilet 27.6 48.8 38.3 17.6 19.6 35.5 Water 38.2 77.3 61.0 17.8 30.5 45.1 Other Home ownership 70.6 40.8 53.0 85.5 78.1 61.5 Lack at least one shelter amenity 92.5 85.3 87.6 96.6 94.9 89.2 Lack at least one shelter 89.9 70.1 82.4 98.0 89.2 83.7 infrastructure structure Overcrowded 17.6 17.8 16.0 18.4 30.4 7.4 Source: Lozano-Gracia Note: Figures are the means taken across all countries’ reported percentages of housing amenities. For example, and Young 2014. the percentage of households that have a permanent roof was calculated by country (and broken down by urban areas and quintiles); the figures shown are the mean percentages of households with a roof, where each observation is one country. This method prevents biasing the percentages toward countries with larger samples. 47 Africa’s Cities | Opening Doors to the World Amenities are poor across all areas and income quintiles. Only 16 percent of urban households in Africa have permanent roofs — and access is even lower in rural areas. Between 70 percent (in the largest cities) and 98 percent (in rural areas) of all Africans lack access to electricity, a toilet, or water. The differences between the proportions of households that have access to housing amenities also varies significantly across expenditure quintiles. They are statistically significant at the 1 percent level for all amenities and infrastructure, indicating the stark difference between housing for the richest and the poorest households in Sub-Saharan Africa. Housing conditions in Sub-Saharan Africa vary by subregion, country, and type of urban area. Shortages are less acute in Middle Africa than in Eastern and Western Africa (map 1.1). Eastern Africa has the largest share of households in overcrowded conditions, the smallest share of households with permanent floors and access to toilets, and the largest percentage of households with electricity connections. Middle Africa has the smallest share of households with electricity connections but the smallest share of people facing housing shortages. Western Africa has the greatest housing deficiencies and the largest proportion of households facing housing shortages. 16 % of urban households in Africa have permanent roofs Between 70 percent (in the largest cities) and 98 percent (in rural areas) of all Africans lack access to electricity, a toilet, or water. 48 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital MAP 1.1 Housing conditions and shortages in Sub-Saharan Africa Western Africa • Smallest share of households with electricity, water, and roof • Largest share of households facing infrastructure shortages • Largest share of homeowners Middle Africa • Smallest share of households facing shelter and infrastructure shortages; lowest share of households with electricity • Largest share of households with permanent walls, floor, toilet, and water Eastern Africa • Smallest share of households with nondirt floors and toilets • Largest share of households with electricity Note: Subregional classifications are • Largest share of households adopted from the United Nations. living in overcrowded housing Source: Lozano-Gracia and Young 2014. 49 Africa’s Cities | Opening Doors to the World Case studies Access to amenities in Dar es Salaam, Durban, and elsewhere in Africa Living standards also vary widely within cities, as recent surveys of Dar es Salaam and Durban reveal. Dar es Salaam, Tanzania About two-thirds of households in Dar es Salaam Access to water and sanitation is very limited share their toilet facilities, and there is some indication throughout the city. It is slightly higher in peripheral that access rates are relapsing: Of the households areas than in central areas (figure 1.5). Households in in the survey that do not currently have access to the city’s core rely mainly on improved pit latrines or sanitation, 15 percent said that they had access in pour flush to pit as their main toilets. The distribution a house they previously lived in, in Dar es Salaam. of households by type of toilet remains similar across The most common form of improved sanitation is city areas, except in shanty areas, where a larger share an improved pit latrine; other forms of improved of the population uses unimproved sanitation. Flush sanitation are rare. Pit latrines are the most common toilets connected to a septic tank are more common form of unimproved sanitation (figure 1.4). on the periphery. FIGURE 1.4 Large shares of the population in Dar es Salaam lack access to basic sanitation and water services Sanitation Water 1% Unprotected well, natural source, Bucket latrine 5% or rainwater 9% 11% or no toilet Protected well or spring Unimposed pit 15% 8% Water vendor, latrine (slab, open, or flush to pit) tanker, or bottled water Improved latrine 49% (slab and vip) Piped water from 39% 17% neighbouring or composting household Flush toilet to 43% a septic tank Standpipe/tap (piped water to yard/plot) Flush toilet to a — public or private piped sewer system Piped water inside dwelling Source: World Bank 2015c. Note: Improved latrine combines 33 percent slab latrine and 6 percent VIP latrine; unimproved sanitation includes 26 percent unimproved, 16 percent open slab, and 7 percent pour flush to pit. 50 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital Access to piped water is also very limited throughout While more than 60 percent of households in the city the city, with only 17 percent of households in the center and 80 percent in shanty areas report sharing city center and 14 percent on the periphery having their toilet facilities, only 36 percent of households piped water. In the city as a whole, only 11 percent of on the periphery do so. When asked about the main households report piped water as their main source constraint of not having piped water in their homes, of drinking water. The shares are much lower for more than 20 percent of households in all areas households in shanty areas in the center (9.5 percent) identified connection cost as the main hurdle. Half and in the consolidated city (7.7 percent). of all households indicated that they lacked access because their home was rented. Access to electricity is much better toward the city center, where more than 70 percent of households Even people with access to piped water in their (including households in shanty areas) have access. homes face challenges related to service quality. On Access decreases with distance from the city center, average, households reported having fewer than 5 with less than 45 percent of households on the days and 14 hours a day of service in the previous 7 periphery connected to the grid. Despite the high days. Low quality forces more affluent households connection rate, the majority of households use and businesses to rely on private boreholes, leading to charcoal as their main fuel for cooking. saltwater intrusion (Mtoni and Walraevens 2010; Jones and others 2016). Residents of Dar es Salaam seem to be benefiting to some extent from being closer to one another: More than 40 percent of households report using piped water from a neighboring household as their main source of water (figure 1.6). The share is much higher in shanty areas (51 percent of households). FIGURE 1.5 Access to improved water and sanitation in Dar es Salaam tends to rise with distance from the center Percentage of households with access to Percentage of households with access to improved sanitation improved water Core Consolidate Periphery Core Consolidate Periphery Shanty Shanty 100% 100% 90% 90% 80% 80% Share of households 70% Lowess 70% smooth 60% 60% 50% 50% Lowess smooth 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 13.4 26.7 13.4 26.7 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40 Distance to CBD (Kilometers) Distance to CBD (Kilometers) Source: World Bank 2015c. Note: Smooth lines are based on locally weighted regressions. 51 Africa’s Cities | Opening Doors to the World FIGURE 1.6 Access to improved sanitation and piped water at home is low across Dar es Salaam Sanitation type, by area Water source, by area 100 100 90 90 80 80 Share of households (%) Share of households (%) 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 Center Center Consolidated Periphery Center Center Consolidated Periphery (nonshanty) (Shanty) city (nonshanty) (Shanty) city Flush toilet to a piped sewer system Piped water inside dwelling Flush toilet to a septic tank Standpipe/tap (piped water to yard/plot) — public or private Improved latrine (slab and vip) or composting Piped water from neighbouring household Unimproved pit latrine (slab, open, or flush to pit) Water vendor, tanker, or bottled water Bucket latrine or no toilet Protected well or spring Unprotected well, natural source, or rainwater Source: World Bank 2015c. Durban, South Africa The story is quite different in Durban, where more Durban is also far ahead of Dar es Salaam on than two-thirds of households have access to piped sanitation. Variations in access are wide, however: water (against 14–17 percent in Dar es Salaam). Almost 92 percent of households in the urban core Within-city variations in access are wide, however. are connected to the sewerage system, but only 19 Almost 87 percent of households in the urban core percent in the outskirts of the city and 45 in shanty have piped water in their dwelling, but less than 20 areas are. In shanty areas, which in Durban are not percent on the outskirts of the city and less than very far from the city’s core, 44 percent of households 15 percent in shanty areas do. In shanty areas, still rely on pit latrines as their main form of sanitation, households rely less on neighbors than in Dar es against 15 percent in peri-urban formal areas and 76 Salaam: Less than 5 percent of households in shanty percent in rural areas on the outskirts of the city. areas and less than 8 percent of households in rural Access to piped water and a flush toilet in the house areas cite their neighbors’ dwelling as their main fall with distance from the city center (figure 1.7). source for drinking water. Almost 42 percent of The average increase beyond 35 kilometers from the households in shanty areas and 52 in rural areas rely city center should be interpreted with care, however, on yard taps. Street standpipes are also an important because it is driven by the greater access in peri-urban source in shanty areas, with 36 percent of households areas farther from the city center. (Further exploration relying on them for drinking water. using a local polynomial estimation suggests some over-smoothing is happening that hides these low access values in rural areas.) 52 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital FIGURE 1.7 Durban: Some challenges are met, but many remain Piped water in house Flush toilet in house 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 0 10 20 30 40 50 0 10 20 30 40 50 Distance from CBD (kilometers) Distance from CBD (kilometers) Improved water Piped water Improved sanitation Flush toilet Source: World Bank 2015c. Note: Smooth lines are based on locally weighted regressions. Migrants who arrived in Durban between 1960 and 2000 enjoyed improvements in access to basic services relative to their previous residence. Migrants today face a deterioration in living standards (figure 1.8). They sacrifice living standards in search of the opportunities the city offers. FIGURE 1.8 Moving to Durban improved migrants’ access to basic services before but not after 2000 Households with Improved water Households with Improved sanitation 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 60 70 80 90 00 10 1 5 0 0 0 0 0 0 5 19 19 19 19 20 20 20 96 97 98 99 00 01 01 re 0– 0– 0– 0 – 0– 0– re1 0–1 0–1 0–1 0–2 0–2 0–2 efo 196 197 198 199 200 201 efo 196 1 97 198 1 99 200 2 01 B B Decade of migration Decade of migration Before After Before After Source: World Bank 2015c. 53 Africa’s Cities | Opening Doors to the World Household moves within the city reflect a similar trend. Before 1980, about 31 percent of within-city moves were from a house that did not have access to improved sanitation to one that did. These figures fell to 18.8 percent for 1980–2000 and 10.6 percent for 2000–15 (figure 1.9). FIGURE 1.9 Within-city moves in Durban are less likely to yield better sanitation than they did in the past 100% 90% 80% 70% 60% Always improved From unimproved to improved 50% Always unimproved 40% 30% 20% 10% 0% Before 1980 1980–2000 2000–2015 Source: World Bank 2015c. Elsewhere in Africa a city with more than half a million people, peri-urban neighborhoods are growing without access to water Infrastructure and public services are concentrated and electricity (formal neighborhoods in the center in central areas and formal neighborhoods in Africa’s also have inadequate infrastructure) (World Bank cities, where access to water and electricity expanded 2016). In the communes (municipalities) of Abidjan between 1997 and 2013. Access to services declines where urban audits were conducted (particularly rapidly once one leaves the center, however; peri- Yopougon, the largest), more than half the residential urban areas and informal neighborhoods have limited neighborhoods have limited access to public services access to basic services. and infrastructure. The pressing challenges observed in Dar es Salaam In Ghana, access to piped water, waste disposal, and and Durban are not unique to those cities. In Côte toilet facilities in Accra and Kumasi decreases rapidly d’Ivoire, urban audits in 2013 showed that although as distance to the city center increases (World Bank infrastructure and public services are adequate in 2015d). In Maputo, Mozambique, access to basic central areas and formal neighborhoods, they are services is strongly concentrated in the city center, lacking in peri-urban and informal neighborhoods, dropping off rapidly outside the center (figure 1.10). where basic infrastructure has failed to keep up In Kinshasa, the Democratic Republic of Congo, access with population growth. In San Pédro, for example, to piped water and septic tanks remains relatively a city of about a quarter million people, informal low and declines with distance from the center (figure neighborhoods with limited access to basic services 1.11). cover about a third of the urban landscape. In Bouaké, 54 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital FIGURE 1.10 FIGURE 1.11 Access to piped water, septic Access to piped water, septic tanks, and electricity improved tanks, and electricity improved in in Maputo, Mozambique Kinshasa, Democratic Republic of between 1997 and 2013 Congo between 2007 and 2013 Piped water Piped water 100 100 100 100 100 % households with access % households with access % households with access % households with access % households with access 80 80 80 80 80 60 60 60 60 60 40 40 40 40 40 20 20 20 20 20 0 0 0 0 0 0 5 10 15 0 5 0 10 5 15 10 0 15 5 0 10 5 15 Distance from CBD in Maputo (km) Distance Distance from CBD in Maputo to downtown Kinshasa (km) (km) Distance Distance from CBD in Maputo to down (km) 2013 2007 1997 2013 Septic tank 2013 2007 1997 2013 2007 Septic 1997 tank 2013 2007 100 100 100 100 100 % households with access % households with access % households with access % households with access % households with access 80 80 80 80 80 60 60 60 60 60 40 40 40 40 40 20 20 20 20 20 0 0 0 0 0 10 15 0 5 0 10 5 15 10 0 15 5 0 10 5 15 10 15 0 5 in Maputo (km) Distance Distance from CBD in Maputo to downtown Kinshasa (km) (km) from CBD in Maputo Distance Distance to downtown Kinshasa (km) (km) Distance to down 07 1997 2013 2007 1997 2013 2007 2013 2007 1997 2013 2007 2013 Electricity Electricity 100 100 100 100 % households with access % households with access % households with access 80 80 80 80 60 60 60 60 40 40 40 40 20 20 20 20 0 0 0 0 0 10 5 15 10 0 15 5 0 10 5 15 10 15 0 5 10 15 Distance D in Maputo to downtown Kinshasa (km) (km) Distance Distance from CBD in Maputo to downtown Kinshasa (km) (km) Distance to downtown Kinshasa (km) 007 1997 2013 2007 2013 2007 1997 2013 2007 2013 2007 Source: Nakamura 2016. Source: Data from Demographic and Health surveys conducted in 2007 and 2013/2014. Note: Smooth lines are based on locally weighted regressions. Note: Distance is measured from Gombe. Smooth lines are based on locally weighted regression. 55 Africa’s Cities | Opening Doors to the World In Nigeria, only Lagos and Abuja have sewerage The lack of basic services puts further pressures on systems, and even in those cities the majority of household budgets. Alternative sources of water households are not connected: Just 6 percent of urban are almost always far more expensive than piped households have a pour/flush toilet connected to a water. In Lagos, the cost of buying informal water and piped sewerage system (World Bank 2005). In Ethiopia, garbage pickup is 1.3–3 times greater than the tariffs only Addis Ababa has a sewerage system — and the charged by the state (World Bank 2015b). In Accra, system serves only 10 percent of the population. the cost of water from small-scale water sources, In 2012, only 44 percent of the population of Addis such as bottles and water “sachets,” is typically five Ababa has access to safe water; full coverage of 100 to seven times higher than piped water (World Bank liters per person per day would require a two- to 2015d). In Bouaké, Côte d’Ivoire, most taps are broken, threefold increase in the water supply (World Bank forcing residents to buy potable water from informal 2015a). In other urban areas, access to water is even water vendors, who can price hikes and sell unsafe or lower and even waste collection is limited: For Sub- illegally sourced water (World Bank 2016). Saharan Africa as a whole, only 45–82 percent of locally generated waste in urban areas is collected. Low human capital For urban areas to be productive better human capital have worse health than people in other regions, is needed. Workers bring to the workplace a bundle of adversely influencing the overall quality of the urban skills that affect their productivity. Cognitive skills have labor supply. long been recognized as an important determinant of In attainment of education and literacy, Sub-Saharan productivity and wages. A growing body of evidence Africa lags behind other developing regions of the documents that, at least in the developed world, larger world and has done so historically. South Asia is cities reward educated workers more than smaller one exception which had poorer literacy than Sub- cities do—one reason larger cities tend to attract Saharan Africa in the 1990s, but over time overtook higher-skilled workers. the latter. Sub-Saharan Africa, which has surging Labor force participation rates in Sub-Saharan Africa youth population levels, has the lowest levels of youth do not lag other developing areas. But, on average, literacy today (tables 1.4 and 1.5). people in Sub-Saharan Africa are less educated and Table 1.4 Adult literacy rate, population 15+ years Region 1990 2000 2010 East Asia and Pacific 79.7 90.8 94.5 Latin America and Caribbean 85.2 88.6 91.5 Middle East and North Africa 56 68 77.9 South Asia 46.5 58 66.7 Sub-Saharan Africa 52.5 57 60.3 Source: UNESCO Institute for Statistics. Note: Percentage of the population age 15 and above who can, with understanding, read and write a short, simple statement on their everyday life. Generally, ‘literacy’ also encompasses ‘numeracy’, the ability to make simple arithmetic calculations. This indicator is calculated by dividing the number of literates aged 15 years and over by the corresponding age group population and multiplying the result by 100. Source: UNESCO Institute for Statistics. 56 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital Table 1.5 Youth literacy rate, population 15–24 years, both sexes (%) Region 1990 2000 2010 East Asia and Pacific 93.6 97.9 98.8 Latin America and Caribbean 92.9 96.0 97.5 Middle East and North Africa 76.2 86.1 91.8 South Asia 59.1 72.6 83.4 Sub-Saharan Africa 64.7 68.2 70.8 Source: UNESCO Institute for Statistics. Note: Number of people age 15 to 24 years who can both read and write with understanding a short simple statement on their everyday life, divided by the population in that age group. Generally, ‘literacy’ also encompasses ‘numeracy’, the ability to make simple arithmetic calculations. Divide the number of people aged 15 to 24 years who are literate by the total population in the same age group and multiply the result by 100. Source: UNESCO Institute for Statistics. Larger cities also pay a wage premium to workers with better social skills (Bacolod, Blum, and Strange 2009). This finding is unsurprising given that cities are often modeled as “interactive systems” in which the urban structure reflects the net benefits of interactions (that is, the value of interactions net of transportation costs) (see Fugita and Thisse 2002, among others). In France, human capital externalities increase earnings via greater communication in the workplace. This effect is larger in bigger and more educated cities (Charlot and Duranton 2004). The World Bank’s STEP (Skills towards Employment and Productivity) Skills Measurement Program is the first initiative to develop a consistent measure of skills — cognitive and social — in low- and middle-income countries. It integrates unique modules in household surveys to directly gauge the soft skills of respondents. The STEP survey incudes questions that directly assess reading, writing, and computer skills (cognitive skills) as well as questions assessing decision making, personality, and behavior (social skills). It also asks questions on whether physical skills are used at work (table 1.6). The STEP survey was used to analyze skills premiums in the capital relative to other urban areas in two African countries (Ghana and Kenya) and five comparators (Armenia, Colombia, Georgia, and Vietnam). The skill indexes were constructed for each respondent using factor analysis. 57 Africa’s Cities | Opening Doors to the World Table 1.6 Skills assessed in the STEP survey Index Measure Cognitive skill index • Read (forms, bills, financial statements, newspapers or magazines, (at and outside work) instruction manuals, books, reports, and so forth); length of readings • Write at work; length of writings • Show numeracy (measure of size, weight, distances, price calculation, more advanced mathematics, and so forth) • Use a computer; type of software used • Show thinking and show learning Social skill index • Interact with others (clients, public, colleagues) • Make a formal presentation • Supervise the work of others Responses to questions about personality and behavior (answers are almost always, most of the time, some of the time, and almost never). • Are you talkative? • Do you like to keep your opinions to yourself? Do you prefer to keep quiet when you have an opinion? • Are you outgoing and sociable — for example, do you make friends very easily? • Do you like to share your thoughts and opinions with other people, even if you don’t know them very well? • Do you ask for help when you don’t understand something? • Do you think about how the things you do will affect others? Physical skill index • Operate or work with any heavy machines or industrial equipment • Regularly have to lift or pull anything weighing at least 50 pounds • Have a physically demanding job Figure 1.12 (overleaf) shows the standardized distribution of skills between the capital and other urban areas in each country. Across countries, the left-tail density is higher than the right-tail density, indicating that most urban workers are poorly endowed with these skills. In most countries, the distribution of cognitive skills in the capital city is denser, because workers with higher cognitive skills prefer to work in larger cities. This pattern is not as pronounced in Ghana and Kenya, however, suggesting that such sorting is not as important in Africa as it is elsewhere. 58 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital BOX 1.3 The contribution of cognitive and noncognitive skills to urban performance The empirical estimation is specified as follows log wi = θ1cci + S’i θ2 + (cci × S’i)θ3 + + X’i β + εi where wi is the nominal wage of worker i; cci is a Box table 1.3.1 shows the skills premium and dummy taking a value of 1 if the worker lives in the interaction effects capturing the effect of the capital city; S’i denotes a worker’s cognitive, agglomeration associated with more cognitive, social, and physical skills; and X’i is a vector social, and physical skills. In all countries, workers comprising the worker characteristics, including with more cognitive skills earn more. On average, educational attainment, the size of the firm, the a one standard deviation increase in the cognitive number of hours worked, and the number of tasks skills index is associated with a 20 percent assigned. Occupation and industry fixed effects increase in the nominal wage.1 Social skills also are accounted for. εi is the error term. tend to be positively associated with wages, albeit to a lesser extent. Agglomeration does not seem to reinforce the positive effect of soft skills on wages in these countries studied. Box table 1.3.1 Ordinary least squares estimates of the effect of agglomeration and soft skills African countries Comparator countries Item Kenya Ghana Vietnam Colombia Armenia Georgia Capital city 0.018 0.022 0.133** 0.018 0.225*** 0.316*** (0.052) (0.116) (0.052) (0.060) (0.063) (0.062) Cognitive skills index 0.224*** 0.193 0.180*** 0.216*** 0.133*** 0.198*** (0.039) (0.125) (0.044) (0.042) (0.047) (0.050) Cognitive skills × capital 0.043 0.040 0.021 –0.068 –0.072 –0.025 (0.061) (0.114) (0.050) (0.060) (0.053) (0.078) Social skills index 0.096*** –0.008 0.098*** 0.052 0.080* 0.039 (0.031) (0.069) (0.031) (0.036) (0.046) (0.046) Social skills × capital 0.044 0.142 –0.059 0.034 –0.134* –0.015 (0.055) (0.122) (0.046) (0.057) (0.068) (0.061) Physical skills index –0.032 0.036 –0.043 –0.052 0.008 –0.082 (0.031) (0.078) (0.032) (0.039) (0.041) (0.054) Physical skills × capital 0.020 0.070 0.050 –0.023 –0.076 0.060 (0.045) (0.102) (0.039) (0.063) (0.065) (0.073) Years of education 0.021*** 0.012 0.045*** 0.014 0.005 0.027** (0.007) (0.024) (0.008) (0.010) (0.013) (0.012) Constant 6.789*** 0.221 3.851*** 10.316*** 0.880* 0.064 (0.697) (0.673) (0.495) (0.348) (0.453) (0.711) Occupation dummies Yes Yes Yes Yes Yes Yes Industry dummies Yes Yes Yes Yes Yes Yes Observations 2108 763 2102 1635 895 634 R2 0.44 0.36 0.34 0.43 0.37 0.61 Source: Based on the World Bank Note: The dependent variable is the log of weekly earnings. Standard errors are clustered STEP Skills Measurement Program. by worker. All regressions control for gender, age and its square, marital status, firm size, and the number of tasks performed. Robust standard errors are in parentheses. *** statistically significant at the 1 percent level, ** statistically significant 5 percent, * statistically significant at the 10 percent level. 59 Africa’s Cities | Opening Doors to the World FIGURE 1.12 10 10 Cognitive 10 10 10 skills of most urban 10 10 10 workers are limited, especially 10 in Africa density kernal density density kernal density density kernal density density density kernal density density 8 8 8 8 8 8 Gaussian kernal density density density Vietnam 8 Colombia 8 Kenya 8 6 6 6 6 6 6 kernal kernal 10 10 10 10 10 10 6 6 6 kernal kernal 4 4 4 4 4 4 Gaussian Gaussian Gaussian Gaussian Gaussian Gaussian density density 8 8 8 8 8 8 4 4 4 Gaussian Gaussian 2 2 2 2 2 2 6 6 6 6 6 6 Gaussian kernal Gaussian kernal Gaussian kernal Gaussian kernal Gaussian kernal Gaussian kernal 2 2 2 0 0 0 0 0 0 4 -1 4 -1 00 01 12 23 34 4 4 -1 4 0-10 01 12 23 34 4 4 -1 04 -10 01 12 23 34 4 -1 0 Skills Cognitive Cognitive 1 (St.) Skills Index 2 (St.) 3 Index 4 -1 Cognitive Cognitive 0Skills Index 1 Skills Index (St.)2 (St.) 3 4 -1 0 Skills 1 Cognitive Cognitive Index 2 Skills(St.) Index 3 (St.) 4 2 2 2 2 2 2 Cognitive Skills Index (St.) Cognitive Skills Index (St.) Cognitive Skills Index (St.) CapitalCapital Other urban Other urban CapitalCapital Other urban Other urban CapitalCapital Other urban Other urban 0 0 Capital Other urban 0 0 Capital Other urban 0 0 Capital Other urban -1 -10 01 12 23 34 4 -1 -10 01 12 23 34 4 -1 -10 01 12 23 34 4 Cognitive Cognitive Skills(St.) Skills Index Index (St.) Cognitive Skills(St.) Skills Index Cognitive Index (St.) Cognitive Cognitive Skills(St.) Skills Index Index (St.) Ghana Armenia Georgia Other urban CapitalCapital Other urban CapitalCapital Other urban Other urban CapitalCapital Other urban Other urban 10 10 10 10 10 10 10 10 10 density kernal density density kernal density density kernal density density density kernal density density 8 8 8 8 8 8 Gaussian kernal density density density 8 8 8 6 6 6 6 6 6 kernal kernal 10 10 6 10 10 6 10 10 6 kernal kernal 4 4 4 4 4 4 Gaussian Gaussian Gaussian Gaussian Gaussian Gaussian density density 8 8 4 8 8 4 8 48 Gaussian Gaussian 2 2 2 2 2 2 6 6 2 6 6 2 6 26 Gaussian kernal Gaussian kernal Gaussian kernal Gaussian kernal Gaussian kernal Gaussian kernal 0 0 0 0 0 0 4 -1 0 4 -1 0 04 0 01 12 23 34 4 4 -2 4 -1 -2 -1 0 0 1 1 2 3 2 3 4 4 4 -2 -1 -2 0 -1 0 1 2 1 3 2 4 3 4 -1 0 1 2 3 4 -2 -1 0 1 2 3 4 -2 -1 0 1 2 3 4 Cognitive Cognitive Skills(St.) Skills Index Index (St.) Cognitive Cognitive Skills Index Skills(St.) Index (St.) Cognitive Cognitive Skills(St.) Skills Index Index (St.) 2 2 Cognitive Skills Index (St.) 2 2 Cognitive Skills Index (St.) 2 2 Cognitive Skills Index (St.) CapitalCapital Other urban Other urban CapitalCapital Other urban Other urban Other urban CapitalCapital Other urban Capital Other urban Capital Other urban Capital Other urban 0 0 0 0 0 0 -1 -10 Source: Based0 1 on 12 the World 23 Bank 34 4 -2 Note: -2 -1 The -1 0 x-axis shows1 0 1 2 2 3 a standardized 3 4 4 cognitive -2 skills -2 -1 index. -1 0 A Gaussian 0 1 kernel 1 2 3 2 4 3 4 Cognitive Cognitive STEP Skills Skills Skills Index Measurement Index (St.) (St.) Program. density Cognitive Skills Index Cognitive curve captures Skills(St.) Index (St.) the density of the index distribution Cognitive Cognitive on the Skills Index y-axis. Skills(St.) Index (St.) Other urban CapitalCapital Other urban CapitalCapital Other urban Other urban Other urban CapitalCapital Other urban Africa’s cities are crowded, and they lack the structures need to make them livable. They are not more densely populated than cities elsewhere, but more of their people live in slums. Concentrated urban populations should enjoy Also important are the relatively high concentrations of economic benefits from firm clustering, agglomeration, economic density within African cities. Urban workers and productivity. Instead, in Africa they endure crowd into informal settlements near the city center in substandard living conditions, because cities are not order to access job opportunities not found elsewhere. dense with economic activity and structures and capital The lack of economic density outside downtown levels are low (low building stock values, poor shelter districts can be seen, literally, in the low intensity of and infrastructure amenities, and low worker skills). nighttime lights. One reason Africa’s city perform poorly is that Africa is Poverty itself is not the problem, however: It locks poorer than other regions were at comparable stages cities into low development traps only if cities are of urbanization. As a result, housing investment lags spatially dispersed and disconnected, unable to attract such investment in low- and middle-income economies investment with efficiencies from agglomeration. Even elsewhere, and cities are relatively unbuilt. poor cities can become livable when they are developed in a more integrated manner, as chapter 2 shows. 60 Part 1 | Crowded and Disconnected African Cities Chapter 1: Crowded with people, not dense with capital References Bacolod, M., B.S. Blum, and W.C. Strange. 2009. “Skills in Linn, Justin. 1979. “Policies for efficient and equitable the City.” Journal of Urban Economics 65 (2): 136–153. growth of cities in developing countries.” Staff Working Paper 342, World Bank, Washington, DC. Bertaud, Alain. 2001. Metropolis: A Measure of the Spatial Organization of 7 Large Cities. Available: http://alainbertaud.com/ Lozano-Gracia, Nancy, and Cheryl Young. 2014. “Housing wp-content/uploads/2013/06/AB_Metropolis_Spatial_Organization.pdf Consumption and Urbanization.” Policy Research Working Paper 7112, World Bank, Washington, DC. Cattaneo, M. D., S. Galiani, P.J. Gertler, S. Martinez, and R. Titiunik. 2009. “Housing, health, and happiness.” American Mtoni, Yohana, and Kristine Walraevens. 2010. “Saltwater Economic Journal: Economic Policy 1 (1): 75–105. intrusion in the quaternary aquifer of the Dar es Salaam region, Tanzania.” Geologica Belgica, 15(1-2), 16-25. Charlot, S., and G. Duranton. 2006. “Cities and workplace communication: Some quantitative French Nakamura, Shohei. 2016. “Spatial Analysis of Greater evidence.” Urban Studies 43 (8): 1365–1394. Maputo: Housing, Infrastructure, and Poverty.” Background paper for Greater Maputo Urban Poverty and Dasgupta, B., S. V. Lall, and N. Lozano-Gracia. 2014. Inclusive Growth. World Bank, Washington, DC. “Urbanization and Household Investment.” Policy Research Working Paper 7170, World Bank, Washington, DC. Sclar, E. D., P. Garau, and G. Carolini. 2005. “The 21st Century Health Challenge of Slums and Cities.” Lancet 365 (9462): 901–03. Demographia. 2015. Demographia World Urban Areas. 11th Annual Edition ed. St. Louis: Demographia. Available Spalding-Fecher, R. 2005. “Health Benefits of Electrification at: http://www.demographia.com/db-worldua.pdf. in Developing Countries: A Quantitative Assessment in South Africa.” Energy for Sustainable Development 9 (1): 53–62. Devoto, F., E. Duflo, P. Dupas, W. Parienté, and V. Pons. 2012. “Happiness on Tap: Piped Water Adoption in Urban Morocco.” Thomson, H., M. Petticrew, and D. Morrison. 2001. “Health Effects American Economic Journal: Economic Policy 4 (4): 68–99. of Housing Improvement: Systematic Review of Intervention Studies.” British Medical Journal 323 (7306): 187–90. Fujita, M., and J. Thisse. 2002. The Economics of Agglomeration. Cambridge University Press, Cambridge. Trojanis, Z., and N. Lozano-Gracia. 2016. “Policy Note: Intermediate Cities in Peru.” World Bank, Washington, DC. Galiani, S., P. Gertler, and E. Schargrodsky. 2005. “Water for Life: The Impact of the Privatization of Water Services on Child UNDESA (United Nations Department of Economic and Mortality.” Journal of Political Economy 113 (1): 83–120. Social Affairs). 2014. Electricity and Education: The Benefits, Barriers, and Recommendations for Achieving the Electrification Gamper-Rabindran, S., S. Khan, and C. Timmins. 2010. “The Impact of of Primary and Secondary Schools. New York: UNDESA. Piped Water Provision on Infant Mortality in Brazil: A Quantile Panel Data Approach.” Journal of Development Economics 92 (2): 188–200. United Nations Department of Economic and Social Affairs, Population Division. 2014. World Urbanization Gollin, Douglas, Martina Kirchberger, and David Lagakos. 2016. “Living Prospects: The 2014 Revision, CD-ROM Edition. Standards across Space: Evidence from Sub-Saharan Africa.” March 31. Available at: https://collaboration.worldbank.org/docs/DOC-20505. United Nations. 2015. Millenium Development Goals Indicators. Indicator 7.10 Proportion of Urban Population Living in Slums. Gulyani, Sumila, Debabrata Talukdar, and Darby Jack. 2010. http://mdgs.un.org/unsd/mdg/seriesdetail.aspx?srid=710. “Poverty, Living Conditions, and Infrastructure Access: A Comparison of Slums in Dakar, Johannesburg, and Nairobi.” Policy World Bank. 2005. “Affordability of Public Transport in Developing Research Working Paper 5388, World Bank, Washington, DC. Countries. Transport Paper No. 3, Washington, DC: World Bank. Henderson, Vernon, and Dzhamilya Nigmatulina. 2016. “The ———. 2013. Planning, Connecting, and Financing Cities Now: Fabric of African Cities: How to Think about Density and Priorities for City Leaders. Washington, DC: World Bank. Land Use. Draft, April 20. London School of Economics. ———. 2015a. Ethiopia Urbanization Review: Urban Institutions Iacovone, L., V. Ramachandran, and M. Schmidt. 2014. “Stunted for a Middle-Income Ethiopia. Washington, DC: World Bank. Growth: Why Don’t African Firms Create More Jobs?” Working ———. 2015b. From Oil to Cities. Nigeria’s Next Transformation. Paper 353, Center for Global Development, Washington, DC. Nigeria Urbanization Review. Washington, DC: World Bank. Ishizawa, O., and R. Gunasekera. 2016. “Economic Values of ———. 2015c. Measuring Living Standards within Cities. Households Buildings in Four African Cities.” Background paper for this report. Surveys: Dar es Salaam and Durban. Washington, DC: World Bank. J-PAL. 2012. “J-PAL Urban Services Review Paper.” ———. 2015d. Rising through Cities in Ghana: Ghana Urbanization Cambridge, MA: Abdul Latif Jameel Poverty Action Lab. Review Overview Report Washington, DC: World Bank. Jones, Patricia, Julia Bird, J.M. Kironde, Lussuga, and Anne ———. 2016. Côte d’Ivoire Urbanization Review. Diversified Laski. 2016. Dar es Salaam: City Narrative. Available: https:// Urbanization. Washington, DC: World Bank. collaboration.worldbank.org/docs/DOC-20828. Wu, T., L. Borghans, and A. Dupuy. 2010. “Energy Kariuki, Mukami, and Jordan Schwartz. “Small-Scale Private Service Infrastructure and Left Behind Villages in Indonesia: Providers of Water Supply and Electricity: A Review of Incidence, Effects of Electric Energy on Children’s Educational Structure, Pricing and Operating Characteristics.” Policy Research Outcomes.” Maastricht University, the Netherlands. Working Paper 3727, World Bank, Washington, DC. 2015. Kolenikov, Stanislav, and Gustavo Angeles. 2009. “Socio- economic status measurement with discrete proxy variables: Is principal component analysis a reliable answer?” Review of Income and Wealth 55 (1): 128–165. 61 Africa’s Cities | Opening Doors to the World 62 Chapter 2 Disconnected land, people and jobs Even as Africa’s cities are crowded with inhabitants — but not dense with capital — they are physically fragmented and dispersed. They develop as collections of small, scattered neighborhoods. Without adequate roads or transportation systems, commuting is slow and costly, so workers lack access to jobs in the larger urban area. Many people and firms are disconnected — from one another and from economic opportunity. 63 Africa’s Cities | Opening Doors to the World The lack of connections among neighborhoods means development undermines city governments’ efforts to that — compared with developed and developing provide the networked services that benefit from scale cities in other regions — African cities show both lower economies — and that undergird urban productivity. exposure and higher fragmentation in the intensity of The prevalence of expansion, particularly leapfrog, how land is used near the city center. development is just one pattern that makes urban • Low exposure means that people are disconnected commuting challenging in African cities. Another from one another. At a set distance (usually 10 is deficient transportation infrastructure. Traffic kilometers), they cannot interact with as many congestion cripples the economy in cities such as people as in a city with higher exposure. Nairobi, where the average journey-to-work time is one of the longest among 15 cities studied (IBM 2011). • High fragmentation means that within a specified area, population varies widely, with scattered Economically, the ideal city can be viewed as an peaks rather than the clustered ones that can efficient labor market, matching employers and job enhance scale economies. Fragmentation increases seekers through connections (Bertaud 2014). With infrastructure costs and lengthens travel times matching, cities benefit because, by increasing the size between homes, jobs, and businesses. of their labor force and its diversity, employers and job seekers are more likely to find an appropriate match There are several aspects to African cities’ low that makes the best of workers’ skills and aspirations. exposure and high fragmentation. The first is the The typical African city fails in this matchmaker role. Its paucity of new building cover and volume near land use is fragmented and its transport infrastructure city centers. Although many African cities have tall insufficient — so its residents lack access to jobs. The buildings, there are patches of either undeveloped separation of formal housing areas from commercial land or land developed with low building volume. The and industrial areas, which makes commuting slow result is a lack of concentration of capital near the and costly, is made worse by emphasis on expansion center, or infill on unbuilt parcels. Development tends development. The lack of connections within the city to push the boundaries of the city outward (a type of stymies agglomeration economies, keeping costs development known as expansion). high and closing the doors of African cities to regional African cities also suffer from a large number of and global trade and investment. In short, because “leapfrog patches” — small parcels of newly built African cities are fragmented (disconnected) and their land that do not border on or overlap existing infrastructure inadequate, people are disconnected development. Their isolation from existing from people, and people are disconnected from jobs. Disconnected land Collections of small and fragmented With cities physically dispersed, Africans in urban areas are disconnected from one another. Urban neighborhoods expansion has increasingly occurred as leapfrog Nearly all African cities look like collections of small patches that do not border or intersect with existing neighborhoods. Outside the high-density city core urban built-up areas, leading to high transportation of a stylized African city, infrastructure is poor, costs and lower access to markets and other people specialization is low, and land has mixed uses — in the city. Within the urban core, population densities often unregulated and informal. New satellite towns vary widely across locations, reducing the chances for develop in peri-urban areas, where land is cheaper. large sections of the population to interact with other But because these new towns have few economic groups in the city. opportunities; lack social amenities (schools, markets); and are not well connected to the urban core, they do Cities exist because they reduce economic distance by not become viable, supporting centers. Instead, they concentrating, in a limited area, workforce, employers, emerge as secondary, disconnected neighborhoods. capital, costly infrastructure, ideas, and buyers and Large and mushrooming informal settlements emerge sellers. Connected people and firms enjoy labor along the few axes of connectivity, usually major market pooling, savings in the transport of inputs, arterial roads or rivers and canals. On the outskirts, and technological and information spillovers (Cervero the city’s rural areas see construction in pockets of 2001). Disconnectedness reduces the opportunity to private investments (Huang 2016). benefit from agglomeration, costing opportunities to learn from others and match needs. 64 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs Disconnectedness can stem from a combination of few density of people in the city, distance depends on how people to interact with, long physical distances between close buildings are to one another, and time depends people, and long travelling times to reach other people. on infrastructure availability (figure 2.1). African cities The three factors depend on the dispersion of people lag in all three areas. and capital. The scale of interactions depends on the FIGURE 2.1 Three aspects of being connected Scale Distance Time Potential for interaction Contiguity of built-up area Infrastructure to support interactions Spatial fragmentation 2000 (Angel and others 2011), show that most African cities were on average similar to other cities of the On measures such as population density gradients or world. The average openness index—the average the share of open space surrounding built-up areas share of open space in the walking distance circle inside urban areas, African cities do not appear to be around each built-up pixel in the city — is 39 percent more spatially fragmented than cities in other regions for African cities, similar to the average in land-rich (see chapter 1). The data, however, are unreliable: developed countries, Europe and Japan, and Southeast There is no comprehensive, up-to-date dataset on Asia (figure 2.2). built-up fragmentation in cities. The latest data, for FIGURE 2.2 The average “openness index” of cities in Sub-Saharan Africa is not very different from the index of other regions 60% 50% 40% 30% 20% 10% 0% Latin America Land-rich Europe Sub- Southeast Northern Eastern Western South and and the developed and Japan Saharan Asia Africa Asia and Asia Central Asia Caribbean Africa the Pacific Source: Angel and others 2011. Note: The openness index measures the average share of open space in the walking distance circle around each built-up pixel in a city. 65 Africa’s Cities | Opening Doors to the World The average city footprint ratio — the ratio of the city land is underutilized in the center, the city loses the footprint to the built-up area in the city — is 1.8 in opportunity to use some of its most productive land. African cities — similar to Latin America, North Africa, Not all land in the vicinity of the center needs to be and Europe and Japan. Sub-Saharan Africa’s cities built up; land can also be left empty for green spaces. thus do not appear to be more fragmented because In Paris, for example, 14 percent of the city center they have more urbanized open space than cities is unbuilt; in Singapore, about 20 percent of the elsewhere. land in the center is intentionally left undeveloped. African cities may be more fragmented than the global In contrast, in cities like Antananarivo, Madagascar; norm close to the city center. City centers have higher Brazzaville, the Republic of the Congo; and Harare, economic dynamism and are thus important spots Zimbabwe, noncontiguous built-up areas are scattered for exchanging goods, services, and ideas. They also throughout the center, with more than 30 percent of usually have the most valuable land in the city. When land within 5 kilometers of the city center left unbuilt (figure 2.3). FIGURE 2.3 Land within five kilometers of the central business district is often left unbuilt in Africa 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% o le e to da s bi sa ka r n cra dis kry no ka riv vil rar pu an La go iro ha Da tow Ac Ad na Ka sa na za Ha Ma Lu Na s e Co Lu tan a Braz Kin Fre An Residential Commercial/Industrial Unbuilt Paris (built-up) Source: Data on Africa are from Antos, Lall, and Lozano-Gracia 2016; data on Paris are from Duranton and Puga 2015. African cities are also characterized by lower exposure Lower exposure and higher fragmentation create and higher fragmentation in the intensity of land scattered density peaks instead of clustered densities occupation around the city center during the day. that could enhance scale economies. Henderson and Controlling for differences in city population and Nigmatulina (2016) use Landscan (see box 2.1) to country GDP per capita, African cities are about 33 measure these and other aspects of the urban form percent less exposed than Asian cities and somewhat in 265 cities in 70 countries in Africa, Asia, and Latin less exposed than Latin American cities, as measured America. They find, for example, that Nairobi is more by the de la Rocha and Puga index described in box fragmented in the city center than Pune, India, with 2.1.12 They are also about 25 percent more fragmented lower clustering of high-ambient densities (as evident near the city center than either Latin American or Asian from the smaller number of blue peaks in figure 2.4). cities (Henderson and Nigmatulina 2016).13 Kampala, Uganda is more fragmented than Surat, India. 66 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs FIGURE 2.4 African cities are more fragmented in the center than comparably sized cities in India Nairobi (population 4.265 million) Pune (population 5.574 million) Kampala (population 2.423 million) Surat (population 3.373 million) 0–364 2,295–4,041 9,777–13,984 27,748–39,955 365–1,032 4,042–6,734 13,985–19,534 39,956–53,912 1,033–2,294 6,735–9,776 19,535–27,747 53,913–71,973 Source: Henderson and Nigmatulina 2016. 67 Africa’s Cities | Opening Doors to the World BOX 2.1 Measuring fragmentation in urban form • Exposure: The de la Roca and Puga (2016) index • Fragmentation: The coefficient of variation of population exposure measures potential of density of the ambient population near interactions at the city level. It measures the the city center is the standard deviation of all exposure (“interaction”) of an average person to cells near the center normalized by the mean other people working and living in a city within a density of these cells. The index reported uses certain radius. The Landscan data focus on the the 8 percent of urban area cells closest to the ambient population: where people live and work city center. Using Landscan data from 2012, in a city over a 24-hour time period. The index Henderson and Nigmatulina (2016) calculated is calculated by taking the sum of the ambient this coefficient. population within 5 or 10 kilometers from each 1 • Fragmented built-up area: The fragmentation square kilometer cell within the city. The citywide of built-up area is measured by the number measure is the sum of the exposure measures and area of leapfrog patches, controlling for for every cell in the city, weighted by its share of initial city built cover and population, population the population. Using Landscan data from 2012, growth, and other factors. The analysis draws on Henderson and Nigmatulina (2016) calculated Baruah (2015). this index for 263 cities in Africa, Asia, and Latin America. Why are some cities more fragmented than others? higher, about 50 percent more leapfrog patches, and When population grows rapidly, cities struggle to about 40 percent more area dedicated to leapfrog provide infrastructure to new settlements. When development. land management and urban planning are weak The findings suggest that the different development and decentralized, new households tend to settle on paths followed by anglophone and francophone cities the outskirts (where land is cheaper), in a scattered, are linked with the colonial legacy. Three hypotheses fragmented manner. The intertwined effects of on why this might be need to be studied: population size, rates of growth, affordability and availability of land, and land management and urban • Legal systems and governance structures have planning lead to different levels of fragmentation. persisted and are more integrated in former French colonies than elsewhere, enabling governments The persistence of institutions from colonial times may to implement urban plans and enforce land use explain the development patterns of many African regulations. The French tended to reject customary cities today. Former British colonial cities appear land management, establishing a state monopoly more fragmented than former French colonies; their on land markets and management systems. In development has been patchier, more scattered contrast, the British established a dual mandate with and strung out, with more open space (Baruah, a dual local government structure: Different laws Henderson, and Peng 2016). governed the colonists’ plantations (or townships) Baruah and Henderson (2016) regress fragmentation and land held under indigenous or customary on a dummy variable for anglophone cities and titles (Baruah, Henderson, and Peng 2016). If control for time-invariant geographic features, francophone cities remained more centralized, their population growth, and time-varying country and city procedures for allocating land use (for example) characteristics. Their results consistently show that might be more standard across the city. In contrast, anglophone cities are more fragmented than other decentralization in anglophone cities may have cities in Africa. All other things equal, compared to allowed local governments to assign land under French cities, they have a sprawl index 25 percent various local rules. 68 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs • In addition to being more centralized, legal systems cities may have been more haphazard, driven by the in francophone countries were more prescriptive. interests of local governments. Procedures to assign land as well as zoning • The training of francophone urban planners — were well defined, possibly resulting in standard through scholarships and education in French procedures for permitting new construction and polytechnic schools — may have induced greater stricter zoning regulations. Land use in anglophone conformity and led to compact, centralized plans. People not connected to people: High fragmentation, low exposure, little potential for interaction Urban Africans have little connectivity in their African cities are more fragmented in the intensity neighborhoods, as shown by the low exposure and of land use near the city center. Development in the high fragmentation in the intensity of land occupation. city center is often fragmented, with low exposure They also have less potential for interaction, as punctured by random towers (Huang 2016). The the dense population near the city center quickly coefficient of variation of densities within the 8 diminishes with distance. People living outside the percent of pixels closest to the city center captures center have very few neighbors to interact with: The this pattern. Small and medium-size cities, such as Puga index at 10 kilometers is much lower in African Addis Ababa, Ethiopia; Antananarivo, Madagascar; cities than in Asian cities of similar size (see box 2.1). and Bujumbura, Burundi are more fragmented than African cities in the sample are mostly small and comparable cities in Asia and Latin America (figure 2.5, medium-size, with fewer than 3 million inhabitants. right-hand panel). Johannesburg’s exposure is about one-fourth that of similar-size cities in Latin America and Asia (figure 2.5, left-hand panel). FIGURE 2.5 Urban people in Africa have less potential for interaction than urban people in other regions Fragmented development: Potential for interaction (exposure): Coefficient of variation Puga index in 10 kilometers 1.8 400,000 25 cities Lagos 1.6 350,000 25 cities 6 cities Jo’burg 1.4 300,000 Lagos 6 cities 1.2 250,000 1 25 cities 200,000 0.8 Jo’burg 25 cities 150,000 0.6 0.4 100,000 0.2 50,000 0 0 LAC Asia SSA LAC Asia SSA Asia SSA LAC Asia SSA LAC Asia SSA LAC LAC SSA Asia LAC SSA Asia LAC SSA Asia SSA Asia LAC LAC SSA Asia 0.5-1 1-3 3-5 7-9 9-13 0.5-1 1-3 3-5 7-9 9-13 City size (million people) City size (million people) Source: Data from Henderson and Nigmatulina 2016. Note: LAC = Latin America and the Caribbean; SSA = Sub-Saharan Africa. 69 Africa’s Cities | Opening Doors to the World A physically fragmented city, with much empty space between the 1990s and 2000s. The number of between built-up areas, raises the costs of providing leapfrog patches per square kilometer of footprint fell services such as road networks, sewerage systems, in only four cities: Windhoek; Addis Ababa; Lusaka, and schools (Squires 2002). It increases air pollution, Zambia; and Ouagadougou, Burkina Faso. Cities such creates losses of natural land, and fragments the as Maputo, Mozambique; Nyala, Sudan; and Nairobi ecosystem (Johnson 2001). Conversely, the clustering are becoming increasingly fragmented. In Maputo, for associated with nonfragmented development makes instance, 51 percent of the patches created between cities potentially more productive and welfare 2000 and 2010 were leapfrogged. In 2010, there were enhancing. 95 new patches per square kilometer of land in the city (figure 2.6). Harari (2014) analyzes 400 Indian cities. She finds that compact (nonfragmented) cities are associated with higher population density, higher welfare, and higher concentration of productive establishments. A one standard deviation deterioration in city form is associated with 0.9 standard deviation decline in population density and a welfare loss equivalent to a 5 percent decrease in income. Leapfrog development increased in many African cities between 2000 and 2010, sowing the seeds for fragmentation. Regardless of their size, these patches increase the costs of providing network infrastructure. Among 21 African cities studied, only in Windhoek, Namibia did the share of total new fragments fall FIGURE 2.6 African cities are becoming more fragmented Share of new leapfrog fragments in total new fragments (percent) 60% 50% 40% 30% 20% 10% 0% ek mey akry ddis kuta buja ayo uba akar gou gos ano robi aka ccra hott igali pala yala uto ako ho a n A Su A law To D dou La K i s A akc K am N p m nd Ni Co Na Lu Ma Ba Wi Bu ag a No u K O u 1990s 2000s Source: Baruah 2015. Note: Share of new leapfrog fragments of new development with the alternatives being infill and extension. 70 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs People not connected to jobs Lack of transportation infrastructure population densities and the share of roads decline more gradually. African cities are allocating little land to roads. In a sample of 30 cities around the world, the 8 African Kigali, Rwanda and Nairobi dedicate a large share cities rank in the bottom 12 spots for road density, of land in the city center to roads, but the share of including Kigali (19th), Addis Ababa (24th), and Nairobi built-up area falls steeply as the share of roads almost (27th). African cities devote less than 16 percent of disappears, as it does in all four cities in map 2.1: the their land to roads (figure 2.7); cities in developed share of land for built-up areas (in orange) goes hand countries usually allocate more than 20 percent. in hand with the share of roads (in dark blue). Lower infrastructure provision, coupled with low African cities are usually serviced by radial arterial affordability of motorized transportation (see chapter roads emanating from the center, sometimes with 3), makes it difficult to access locations across the city. concentric ring roads. From a spatial planning Low affordability increases the number of people who perspective, a radial-ring structure is less efficient must walk; low road provision increases congestion than a grid structure, which allows for scalability and and pollution. easy orientation and provides alternative routes and Roads tend to be concentrated in the city center, thus enhanced travel efficiency. Increasingly, African providing little accessibility for people on the cities are adopting the grid structure, especially for periphery. Cities with well-developed transportation expansion areas or new satellite development, as in, infrastructure, like Paris, devote a larger share of for example, Bahir Dar, Ethiopia and Ouagadougou, land to roads throughout the city, providing the both of which have adopted a scalable grid modular connectivity required for compact development. unit at the neighborhood level, with a clear hierarchy These cities still follow a monocentric model, but their for roads, amenities, and services. FIGURE 2.7 Paved roads occupy a smaller share of urban land in Africa than elsewhere — and usually drop off abruptly beyond the city center Built-up Paved roads Open space Barcelona London 100 100 Percentage of urban land Percentage of urban land 80 80 60 60 40 40 20 20 CBD CBD -25 -20 -15 -10 -5 0 5 10 15 20 25 -25 -20 -15 -10 -5 0 5 10 15 20 25 71 Africa’s Cities | Opening Doors to the World FIGURE 2.7 (cont.) Paved roads occupy a smaller share of urban land in Africa than elsewhere — and usually drop off abruptly beyond the city center Built-up Paved roads Open space Paris Kigali 100 100 Percentage of urban land Percentage of urban land 80 80 60 60 40 40 20 20 CBD CBD -25 -20 -15 -10 -5 0 5 10 15 20 25 -25 -20 -15 -10 -5 0 5 10 15 20 25 Dar es Salaam Nairobi 100 100 Percentage of urban land Percentage of urban land 80 80 60 60 40 40 20 20 CBD CBD -25 -20 -15 -10 -5 0 5 10 15 20 25 -25 -20 -15 -10 -5 0 5 10 15 20 25 Addis Ababa Dakar 100 100 Percentage of urban land Percentage of urban land 80 80 60 60 40 40 20 20 CBD CBD -25 -20 -15 -10 -5 0 5 10 15 20 25 -4 -2 1 3 5 7 9 11 13 15 17 19 21 23 25 Source: Data from Antos, Lall, Note: CBD = Central Business District. The data for African cities come from very high-resolution (< 1m) imagery using a and Lozano-Gracia 2016 and semiautomated supervised classification approach (leveraging both textural and spectral data). The images are circa 2012 Felkner, Lall, and Lee 2016. for Nairobi and Kigali, circa 2013 for Dar es Salam, circa 2014 for Dakar, and circa 2011 for Addis Ababa. The European data come from the Urban Atlas, published by the European Environment Agency (EEA) (http://www.eea.europa.eu/data- and-maps/data/urban-atlas). The layers in this atlas were created in 2010, based on 2005–07 imagery. The central business district was identified using the location of the oldest building as a proxy (or a government building if necessary). 72 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs Investments in infrastructure shape how land is used. Roads and transit systems affect commuting costs and times differently across the city, incentivizing densification in some corridors and allowing households to locate farther from the city center. In the four cities in map 2.1, investments in road infrastructure increased population density by 17–37 percent within a 1 square kilometer of the road (Felkner, Lall, and Lee 2016). MAP 2.1 Change in land used by paved roads across four African cities Addis Ababa Kigali -1.0% — 0% 0.01% — 1.506% 1.506% — 1.748% 1.748% — 3.548% -0.886% — 0% 3.548% — 10,259.025% 0.01% — 0.891% 0.891% — 0.975% 0.975% — 1.449% 1.449% — 4,136.814% Dar es Salaam Nairobi -1.0% — 0% 0.01% — 0.084% 0.084% — 0.2% 0.2% — 0.905% 0.905% — 20,896.588% -1.0% — 0% 0.01% — 0.357% 0.357% — 0.977% 0.977% — 1.828% 1.828% — 36,640.027% Source: Felkner, Lall, and Lee 2016. 73 Africa’s Cities | Opening Doors to the World Lack of money for transportation With lower incomes, urban residents in Sub-Saharan Africa’s cities spend a large share of their budgets Sub-Saharan Africa is urbanizing rapidly, but it is on food, leaving little for transportation, housing, doing so at low per capita incomes, limiting the funds or other basic items. Food accounts for 60 percent available for transportation. In 2013, it was roughly 37 of total expenditures for the bottom 20 percent percent urbanized, and income per capita was a little of Sub-Saharan Africa’s urban households and 35 over $1,000 in 2005 dollars. Latin America and the percent even for the wealthiest quintile (figure 2.8). Caribbean reached 40 percent urbanization in 1950, Transportation expenditures are very low, at 3–8 with a per capita income of $1,860. The Middle East percent of total spending (Lozano-Gracia and Young and North Africa reached the threshold in 1968, when 2014). The small monetary share reflects the fact that per capita incomes was $1,800; East Asia and Pacific most commuting involves walking, trips for which reached it in 1994, when per capita income was $3,620 people in Asia and Latin American cities would use (WDI 2015). cars, scooters, or public transit. FIGURE 2.8 Urban Africans spend a large share of their budgets on food 60 Percentage of total expenditures 50 40 Food 30 Transportation Housing (total) 20 Housing (utilities) 10 0 1 2 3 4 5 (Poorest) (Richest) Source: Lozano-Gracia and Young 2014. Note: Data are from 20 countries. High out-of-pocket commuting costs to be part time, and usually better paying) within four disproportionately affect the disadvantaged (poor, months (Franklin 2015). Within the control group, young, unskilled) by creating a spatial mismatch. the employed were more likely to have part-time, Physical segregation of unskilled workers from job informal, and local jobs requiring less commuting — a opportunities leads to high commuting and job search pattern repeated in Kigali (figure 2.9). costs, which partially explain higher unemployment Travel budgets in Europe and the United States rates and lower average wages. In Addis Ababa, display a remarkable regularity (about 10–15 percent for example, a randomized experiment to provide of personal expenditures) (box 2.2). Monetary travel young residents in peripheral neighborhoods with budgets in Japan are lower and closer to African a nonfungible transit subsidy that could be used to averages (about 7 percent), because many Japanese search for employment in the city center increased use public transit rather than cars (Schafer and Victor employment (26 percent versus a mean of 19 percent) 2000). and led to better-quality jobs (more formal, less likely 74 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs FIGURE 2.9 In Kigali, workers in the informal sector have shorter commutes All 14.8% 23.4% 22.8% 39% Less than 10 min 10–20 min Informal 37.8% 20.8% 18.8% 22.7% 21–30 min More than 30 min Formal 11.5% 23.7% 23.5% 41.4% 0% 20% 40% 60% 80% 100% Source: NISR and MPSL 2013. BOX 2.2 Is there a constant travel time budget? The Zahavi conjecture Zahavi (1973, 1974; later with Talvitie) observed that, combined with average travel speeds in an that daily urban travel times in the United urban area, constrains the distances that can be States and Europe were similar, despite wide travelled and hence the space of accessibility. differences in transportation infrastructure; It is this space of accessibility that urban geography; and the incomes, cultural norms, dwellers seek to maximize given transportation and habits of residents of different cities. Zahavi constraints. This reasoning appears to be found that these average travel times are nearly compatible and may even explain urban areas’ constant for individual cities over time and growth through sprawl and increased travelled similar over space for different cities, at about distances. one hour a day. He concluded that travel times Looking at a large variety of settlements (cities are therefore spatially transposable, leading to and countries), Schafer and Victor (2000) find the idea of a predictable travel time budget. He some empirical evidence to support these also observed the regularity of a “travel monetary regularities, in particular for the travel time budget.” budget, which they estimate at about 1.1 hours This observation led to the formulation of an a day (box figure 2.2.1). The travel monetary assumption, known as the Zahavi conjecture, that budget displays slightly less regularity; it depends postulates that decreases in travel time caused on the motorization rate but tends to converge by increases in transportation speeds (switching beyond a certain rate. Schafer (2000) shows that from slower to faster transportation modes) travel monetary budgets were virtually constant, are entirely reinvested in more trips and longer oscillating between 10 and 15 percent, for a travelled distances. If the conjecture is accurate, panel of six developed countries from 1970 to it means that there is a natural travel time budget 2004. (continued overleaf). 75 Africa’s Cities | Opening Doors to the World BOX 2.2 (cont.) Is there a constant travel time 4.0 budget? The Zahavi conjecture 3.5 BOX FIGURE 2.2.1 Average 3.0 per capita daily travel times in selected cities 2.5 Travel time budget, hour per capita per day 2.0 1.5 1.0 0.5 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 GDP per capital, US$ (2000) Cities Countries African villages in Ghana and Tanzania Source: Schafer and Victor 2000; Schäfer and others 2009. Zahavi himself did not believe that there was a modeling found that threshold to be about 10 fundamental law that would lead to constant kilometers an hour, which would correspond to travel times over time, irrespective of local a speed at which a large share of travelers no conditions. When developing the unified longer walk for all trips. Before that threshold mechanism of travel model (Zahavi 1981), he is reached, travel time budgets can exceed one linked travel time budgets to the motorization hour and decrease strongly with any increase in rate (equivalent to an average speed in the urban travel speeds. This subtlety of Zahavi’s modeling, area). He did find that there is a speed beyond which is rarely appreciated, may explain some which travel time budgets converge and become findings for African cities in which travel time nearly insensitive to future speed increases. His budgets appear higher than elsewhere. 76 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs The cost of collective motorized urban transportation, These potential budget shares show wide variation. In which dominated in Africa by informal minibuses, is Dakar, a round-trip commute costs only 3.1 percent of high relative to household budgets in Sub-Saharan the average household’s total expenditures. For most Africa’s major cities, rendering it largely unaffordable of the other cities, it costs 5.1–27.5 percent. For the on a daily basis, especially for the poorest (Kumar bottom quintile, the situation is even worse. Excluding and Barrett 2008). In 8 out of 11 cities studied, the Dakar, the poorest households in these cities would average household could not afford one round-trip need to spend an average of about 19 percent of their a day using the minibus network. Figure 2.10 shows budget to afford a round-trip motorized commute. For actual average budget shares on transport and then the poorest quintile, the figure is 53 percent in Dar es what these shares would be for first the average Salaam and more than 100 percent in Lagos. households and then a poor household if they made two trips a day using the minibus network. FIGURE 2.10 The typical household in many African cities cannot afford public transport fares 120 % of household 100 budget spent on transport 80 % of average household budget needed 60 for 2 trips/day 40 % of first quintile household budget needed 20 for 2 trips/day 0 go s am ali an ala go u bi as a ua la ab a ka r La ala Kig idj mp ou iro sh Ab Da S Ab Ka d Na Kin Do is r Es aga dd Da Ou A Source: Kumar and Barrett 2008. The World Bank (2015b) conducted comprehensive of the population in African cities; starting from the travel demand surveys in Nigeria in Abuja (2013), Kano central business district and traveling outward, travel (2012), and Lagos (2009, 2012). They revealed that low- per passenger-kilometer can be provided at lower cost income households in these cities spend 49, 40, and 33 by bus than minibus and by minibus than car percent, respectively, of their household budgets on (figure 2.11). public transport; the average household spent 31, 32, The cost per passenger kilometer increases with and 24 percent, respectively; and middle- and upper- distance from the central business district because income households spent 18, 19, and 10 percent, density declines (Eskeland and Lall 2015). The increase respectively. These high costs relative to household is steeper (more convex) for buses than for minibuses budgets lead to travel patterns dominated by walking, because they have more seats to fill. They have to greatly limiting access to economic opportunities. spend more than minibuses to be sufficiently frequent The prevalence of minibuses rather than larger vehicles and close to demand. Cars have the flattest curve, increases the monetary costs of transportation, because they carry only a few passengers. contributing to its unaffordability for a large segment 77 Africa’s Cities | Opening Doors to the World FIGURE 2.11 Scale economies in vehicle size should lead to the provision of collective transportation through large vehicles toward the city center and smaller ones as distance to the city center increases Cbus Cminibus Cost per passenger km Ccar Central Walkers, Bus doughnut Minibus zone No service Business bikers (or only cars) District Distance; Declining density Source: Eskeland and Lall 2015. Note: The dotted lines indicate a change in the most cost effective means of transport. The ‘bus doughnut’ corresponds to the dense urban area served competitively by buses. Larger vehicles are more efficient if they can be kilometers. Such a catchment area would cover only filled; where population density falls, they may not about 7 percent of Nairobi’s core city, for example. be efficient. The size of the vehicle that can operate In Dar es Salaam, households travel only small cost effectively declines with the distance from the distances to get to work: Household heads’ average center. Travel costs therefore increase more than commuting distance is less than 6 kilometers, and proportionally with distance from the city center other adults travel no more than 4 kilometers on (Eskeland and Lall 2015). If the prices of transportation average (World Bank 2015c). These figures indicate reflect the costs, larger buses leveraging scale that Dar es Salaam, although a city of more than economies could significantly reduce transportation 5 million people, functions as a set of villages with costs for users in African cities. extremely local labor markets. The percentage of trips made by foot is very high in In Durban, South Africa, labor markets do not appear major African cities, largely because of the high cost to be as fragmented, with most people traveling 30–40 of bus trips. The figure is 30–45 percent in Nairobi, minutes from home to work; commuting modes are Lagos, and Addis Ababa and nearly 70 percent in Dar also more diverse than in Dar es Salaam. A third of es Salaam (figure 2.12). Durban’s population uses minibuses as the main Generously assuming that pedestrians can travel at mode of transportation, and almost 30 percent use an average of 4 kilometers an hour in a straight line, a private vehicles. A still considerable 21 percent walk to large share of city residents can access opportunities work. only within a 50 square kilometer area of where Another striking feature of African cities is that they live by walking for an hour — a problem in collective transportation is primarily informal. Matatus metropolises that often cover more than 1,000 square in Nairobi and Kampala, cars rapides (or ndiaga ndiaye) 78 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs FIGURE 2.12 A very large share of trips in African cities are made by foot Private car Abidjan Bicycle and walking Accra Mortorcycle Addis Ababa Public transport (bus) Private taxis Dar es Salaam Informal (minibus, collective taxis) Dakar Douala Johannesburg Lagos Nairobi Windhoek 0% 20% 40% 60% 80% 100% Source: International Association of Public Transport 2010. in Dakar, dala dalas in Dar es Salaam, tro in Accra, and human capital or facilitate the accumulation of the soft gbaka in Abidjan are all informal (and often colorful) skills they need to succeed. minibuses. Collectively, they are the main form of Mobility is essentially a derived demand, meaning motorized collective transportation in most African that it is desired not for itself but because it can cities. The service they provide is highly valued, as provide access to other goods and services. Cities’ demonstrated by their large share of trips. However, matchmaking function is to provide employers because public or collective transportation is informal, with candidates who seek their needs and for job the networks are mostly reactive: They accompany the seekers to find employment opportunities that growth of the city rather than structure it. In contrast, suit their talents and aspirations; it is an important in many developed country cities (Paris, London, reason why cities are more productive than the rest Barcelona), transportation helps determine the of an economy (World Bank 2015a). Employment structure of the city (Brooks and Lutz 2013). accessibility can be commonly defined as the number or share of jobs that can be reached within a given Inaccessible employment time. For cities to act as integrated labor markets and match jobs seekers and employers, they need to make Avner and Lall (2016) study access to formal jobs in employment accessible. African cities are failing to Nairobi. They find that accessibility is much higher for do so. Activities are scattered instead of clustered, car users than for users of matatus (privately owned because capital is misallocated. Fragmentation minibuses). Indeed, map 2.2 suggests that Nairobi is prevents labor market pooling — leaving workers a city built for car owners, who can reach about 90 with access to few jobs, keeping productivity low, percent of jobs within an hour, But car use accounts and locking cities into the nontradable sector. Such a for only 13 percent of trips for all purposes; 28 percent climate does not reward workers for investing in their of trips are made with matatus and 41 percent are 79 Africa’s Cities | Opening Doors to the World MAP 2.2 Most job opportunities in Nairobi are inaccessible to people without cars Shares of job opportunities accessible within an hour by car 0–10 10–20 20–30 30–40 40–50 50–60 60–70 70–80 80–90 90–100 Shares of job opportunities accessible within an hour by matatu (minibus) 0 5 10 Source: Avner and Lall 2016. 80 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs made on foot (JICA 2013, 2014). A central resident can These figures do not account for the distribution of access only 20 percent of all jobs within an hour using residents throughout the urban areas. As people tend the matatu network. to settle in denser patterns in zones that benefit from greater employment accessibility, residents’ location In contrast, in 2013, a resident of London could reach affects the outcome (table 2.1).10 2.5 million jobs (54 percent of all jobs in Greater London) within 45 minutes of the city center using the public transit network (TfL 2014). From any location in the urban area, the average number of jobs accessible within 45 minutes using transit in Greater London in 2013 was a little under a million (21.6 percent of all jobs). By comparison, in Nairobi, matatu users’ had access to only 5.8 percent of all jobs within 45 minutes. Table 2.1 Accessibility to formal jobs in Nairobi Cars Matatus + walking Walking only Maximum travel time Uncongested Congested Congested (minutes) 30 57% 31% 4% 3% 45 85% 58% 10% 7% 60 96% 77% 20% 11% Source: Avner and Lall 2016. Heavy congestion, high rates of walking, informal The early deployment of collective transportation collective transportation, and the spatial distribution systems preserves an option value (Henry 1974) of jobs and residents lead to low employment by limiting urban-form irreversibility. Early accessibility in Nairobi and the misallocation of labor. deployment also reduces the need for transportation Matatu users on average can access only 4 percent of infrastructure. Rode and others (2014) show that jobs within 30 minutes, 10 percent within 45 minutes, although the upfront costs of collective transportation and 20 percent within 60 minutes. These figures are infrastructure are on the same order of magnitude very low: In metropolitan Buenos Aires, equivalent as — or even higher than — the costs of high-capacity accessibility figures using public transportation are 7 highways, collective transportation has the edge when percent, 18 percent, and 34 percent for the same time cost estimates take carrying capacity into account thresholds (Peralta Quiros 2015). In Ugandan cities, (table 2.2). Compact cities (associated with collective 70 percent of work trips are on foot (Uganda Bureau transportation) also mitigate the impacts of higher oil of Statistics 2010), with the average share of jobs prices. reachable within one hour standing at just 19 percent (Bernard 2016).11 81 Africa’s Cities | Opening Doors to the World Table 2.2 Estimated capital costs of building various types of transportation infrastructure Capacity Capital costs (dollars Capital costs/ Type of infrastructure (pers/h/d) per kilometer) capacity (dollars) Dual-lane highway 2,000 10–20 million 5,000–10,000 Urban street (car use only) 800 2–5 million 2,500–7,000 Bike path (2m) 3,500 100,000 30 Pedestrian walkway/ pavement (2m) 4,500 100,000 20 Commuter rail 20,000–40,000 40–80 million 2,000 Metro rail 20,000–70,000 40–350 million 2,000–5,000 Light rail 10,000–30,000 10–25 million 800–1,000 Bus rapid transit 5,000–40,000 1–10 million 200–250 Bus lane 10,000 1–5 million 300–500 Source: Rode and others 2014, based on Rode and Gipp 2011, Litman 2009, Wright 2002, and Brilon 1994. High transportation costs, crippling congestion, and slow commuting speeds have prevented African cities from acting as matchmakers and fostering agglomeration economies through firm clustering. Such mixed land-use patterns — with jobs and people dispersed throughout urban areas — penalize large-scale transit systems, which need high ridership to function efficiently. African cities are locked into low-level equilibria land use and transportation infrastructure. Whatever that are unkind to their residents and unproductive shape a city assumes — and various urban forms are for firms. Coordination mechanisms are needed to consistent with more productive and inclusive cities harness the forces of urbanization through kinder, — leaders must think very carefully about the long- more efficient development trajectories (chapter 6). term consequences of their decisions. Given the high inertia and path dependencies of urban areas, these These coordination mechanisms can take many decision makers are building their cities for decades forms. All, however, include synchronized action on and possibly centuries to come. 82 Part 1 | Crowded and Disconnected African Cities Chapter 2: Disconnected land, people and jobs References Alonso, W. 1964. Location and Land Use. Cambridge, Gobillon, Laurent, Harris Selod, and Yves Zenou. 2007. “The MA: Harvard University Press. Mechanisms of Spatial Mismatch.” Urban Studies 44 (12): 2401–28. Angel, Shlomo, Jason Parent, Daniel L. Civco, and Alejandro M. Harari, M. 2014. “Cities in Bad Shape: Urban Geometry in Blei. 2011. “Making Room for a Planet of Cities.” Policy Focus India.” Economics Department, MIT, Cambridge, MA. Report, Lincoln Institute of Land Policy, Cambridge, MA. Henderson, Vernon and Nigmatulina, Dzhamilya. 2016. “The Antos, Sarah E., Nancy Lozano-Gracia, and Somik V. Lall. 2016. “The Fabric of African Cities: How to Think about Density and Morphology of African Cities.” Draft, World Bank, Washington, DC. Land Use.” Draft, April 20, London School of Economics. Arthur, W. B. 1986. Industry Location Patterns and the Henry, Claude. 1974. “Investment Decisions under Uncertainty: The Importance of History. Center for Economic Policy ‘Irreversibility Effect.’” American Economic Review 64 (6): 1006–12. Research, Stanford University, Stanford, CA. Huang, Chyi-Yun. 2016. “Enabling Structure: Role of Public Avner, Paolo, and Somik V. Lall. 2016. “Matchmaking in Sector and Institutions.” Background paper for this report. Nairobi: The Role of Land Use.” Policy Research Working Iacovone, L., V. Ramachandran, and M. Schmidt. 2014. “Stunted Paper 7904, World Bank, Washington, DC. Growth: Why Don’t African Firms Create More Jobs?” Working Baruah, Neeraj. 2015. “Splintered and Segmented? Paper 353, Center for Global Development, Washington, DC. Fragmentation of African Cities’ Footprints.” Paper presented International Association of Public Transport (UITP). at the Spatial Development of African Cities Workshop, 2010. Major Trends and Case Studies. Brussels, Belgium: World Bank, Washington DC, December 16–17. International Association of Public Transport. Baruah, Neeraj, and Vernon Henderson, with Cong Peng. 2016. IBM. 2011. “IBM Global Commuter Pain Survey: Traffic “Fragmentation of African Footprints. Colonial Legacies in Shaping Congestion Down, Pain Way Up.” Available at: http://www- African Cities.” Paper presented at the Annual Bank Conference 03.ibm.com/press/us/en/pressrelease/35359.wss. on Africa 2016: Managing the Challenges and Opportunities of Urbanization in Africa, Oxford University, June 14. JICA (Japan International Cooperation Agency). 2013. JICA Nairobi Personal Travel Survey. Tokyo. Baum-Snow, N. 2007. “Did Highways Cause Suburbanization?” Quarterly Journal of Economics 122 (2): 775–805. ———. 2014. The Project on Integrated Urban Development Master Plan for the City of Nairobi in the Republic of Kenya. Draft. Tokyo. Baum-Snow, N., L. Brandt, J. V. Henderson, M. A. Turner, and Q. Zhang. 2016. “Roads, Railroads and Decentralization of Chinese MCLAU (Ministère de la Construction, du Logement, de Cities.” LSE Working Paper, London School of Economic, London. l’Assainissement et de l’Urbanisme), SDUGA, and JICA (Japan International Cooperation Agency). 2014. Le Projet de Développement Bernard, Louise. 2016. “Job Access in Kampala.” University of Oxford. du Schema Directeur d’Urbanisme du Grand Abidjan (SDUGA). Bertaud, Alain. 2014. “Cities as Labor Markets.” Rapport Interimaire, Volume I, Resume. Oriental Consultants Co. Working Paper 2, Marron Institute on Cities and the Ltd., Japan Development Institute, JICA, and Asia Air Survey Co. Ltd. Urban Environment, New York University. Johnson, M. 2001. “Environmental Impacts of Urban Brooks, Leah, and Byron Lutz. 2013. “Vestiges of Transit: Urban Sprawl: A Survey of the Literature and Proposed Research Persistence at a Micro Scale.” Trachtenberg School of Public Policy and Agenda.” Environment and Planning A 33: 717–35. Public Administration, George Washington University, Washington, DC. Krugman, P. R. 1991. Geography and Trade. Cambridge, MA: MIT Press. Cervero, Robert. 2001. “Efficient Urbanisation: Economic Performance Kumar, Ajay, and Fanny Barrett. 2008. “Stuck in Traffic: Urban and the Shape of the Metropolis.” Urban Studies 38 (10): 1651–71. Transport in Africa.” Africa Infrastructure Country Diagnostic de la Roca and Puga. 2016. “Learning by Working Background Paper. World Bank, Washington, DC. in Big Cities.” Review of Economic Studies. Lall, S. V. 2009. “Shrinking Distance: Identifying Priorities for Duranton, Gilles, and Diego Puga. 2004. “Micro-Foundations of Territorial Integration.” Draft, World Bank, Washington, DC. http:// Urban Agglomeration Economies.” In Handbook of Regional and ec.europa.eu/dgs/policy_advisers/docs/15_paper_lall.pdf. Urban Economics, vol. 4, 2063–117. Amsterdam: Elsevier. Lall, S. V, E. Schroeder, and E. Schmidt. 2009. “Identifying ———. 2015. “Urban Land Use.” In Handbook of Regional and Spatial Efficiency-Equity Tradeoffs in Territorial Development Urban Economics, vol. 5, 467–60. Amsterdam: Elsevier. Policies: Evidence from Uganda.” Policy Research Working Paper 4966, World Bank, Washington, DC. Ellison, G., and E. L. Glaeser. 1997. Geographic Concentration in U.S. Manufacturing Industries: A Dartboard Approach. Lecocq, Franck, and Zmarak Shalizi. 2014. “The Economics of Targeted Journal of Political Economy 105 (5): 889–927. Mitigation in Infrastructure.” Climate Policy 14 (2): 187–208. Eskeland, Gunnar S., and Somik V. Lall. 2015. “A Crowded Levinson, David. 2013. Access across America. CTS 13-20, Center for City: Agglomeration and Mobility in Urban Development.” Transportation Studies, University of Minnesota, Minneapolis. Draft, August 20, World Bank, Washington, DC. Lozano-Gracia, Nancy, and Cheryl Young. 2014. “Housing Felkner, John S., Somik V. Lall, and Hyun Lee. 2016. Consumption and Urbanization.” Policy Research “Synchronizing Public and Private Investment in Cities: Working Paper 7112, World Bank, Washington, DC. Evidence from Addis Ababa, Dar es Salaam, Kigali and Lucas, Robert E., and Esteban Rossi-Hansberg. 2002. “On the Nairobi.” April 22, World Bank, Washington, DC. Internal Structure of Cities.” Econometrica 70 (4): 1445–76. Franklin, Simon. 2015. “Location, Search Costs and Youth Marshall, Alfred. 1920. Principles of Economics. London: Macmillan. Unemployment: The Impact of a Randomized Transport Subsidy in Urban Ethiopia.” CSAE Working Paper WPS/2015-11, Oxford.. Muth, R. F. 1969. Cities and Housing: The Spatial Pattern of Urban Residential Land Use. Chicago: University of Chicago Press. Fujita, Masahisa, and Ogawa, Hideaki, 1982. “Multiple Equilibria and Structural Transition of Non-Monocentric Urban Configurations.” Mills, E. S. 1972. Studies in the Structure of the Urban Regional Science and Urban Economics 12 (2): 161–96. Economy. Baltimore, MD: Johns Hopkins Press. 83 Africa’s Cities | Opening Doors to the World NISR (National Institute of Statistics), and MPSL (Minister for Public Service and Labour). 2013. Rwanda National Manpower Survey 2011–2012. National Institute of Statistics of Rwanda, Kigali. Owen, Andrew, and David Levinson. 2014. Access across America: Transit. CTS 14-11, Center for Transportation Studies, University of Minnesota, Minneapolis. Peralta Quiros, Tatiana. 2015. “Mobility for All: Getting the Right Urban Indicator.” Note 25, Connections Series, World Bank, Washington, DC. http://www.worldbank. org/en/topic/transport/brief/connections-note-25. Rode, Philipp, Graham Floater, Nikolas Thomopoulos, James Docherty, Peter Schwinger, Anjali Mahendra, and Wanli Fang. 2014. “Accessibility in Cities: Transport and Urban Form.” New Climate Economy Cities, Paper 03. LSE Cities, London School of Economics and Political Science, London, UK. Rosenthal, Stuart S., and William C. Strange. 2004. “Chapter 49: Evidence on the Nature and Sources of Agglomeration Economies.” In Handbook of Regional and Urban Economics 4: 2119–71. Amsterdam: Elsevier. Schäfer, Andreas. 2000. “Regularities in Travel Demand: An International Perspective.” Journal of Transportation and Statistics 3 (3): 1–31. Schäfer A., J. B. Heywood, H. D. Jacoby, and I. A. Waitz. 2009. Transportation in a Climate-Constrained World. Cambridge, MA: MIT Press. Schäfer, Andreas, and David G. Victor. 2000. “The Future Mobility of the World Population.” Transportation Research Part A: Policy and Practice 34 (3): 171–205. Squires, G. 2002. Urban Sprawl: Causes, Consequences, & Policy Responses. Urban Institute, Washington, DC. Syverson, C. 2004. “Market Structure and Productivity: A Concrete Example.” Journal of Political Economy 112 (6): 1181–222. TfL (Transport for London). 2014. Travel in London. Report 7, Transport for London. Uganda Bureau of Statistics. 2010. Uganda National Household Survey 2009–2010. Kampala. WDI (World Development Indicators). 2015. Washington, DC. http://data.worldbank.org/data-catalog/world-development- indicators.World Bank. 2009. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. World Bank. 2009. World Development Report 2009: Reshaping Economic Geography. Washington, DC: World Bank. ———. 2015a. Competitive Cities for Jobs and Growth. Washington, DC: World Bank. ———. 2015b. From Oil to Cities. Nigeria’s Next Transformation. Nigeria Urbanization Review. Washington, DC: World Bank. ——— 2015c. Measuring Living Standards within Cities. Households Surveys: Dar es Salaam and Durban. Washington, DC: World Bank. ———. 2015d. Rising through Cities in Ghana: Ghana Urbanization Review Overview Report Washington, DC: World Bank. Zahavi, Yacov. 1973. “The TT-relationship: A unified approach to transportation planning.” Traffic engineering and control, 205–212. ———. 1974. “Travel Time Budgets and Mobility in Urban Areas.” Report prepared for the U.S Department of Transportation, Washington, D.C. and Ministry of Transport, Federal Republic of Germany, Bonn. ———. 1981. The UMOT Project. Washington, DC: US Government Printing Office. Zahavi, Yacov, and Antti Talvitie. 1980. Regularities in Travel Time and Money Expenditures. Transportation Research Record 750: 13–19. 84 Part 2 Africa’s Low Urban Development Trap Chapter 3 Costly for households, costly for firms High prices, low incomes High wages, high costs of doing business References Chapter 4 Africa’s urban development trap Cities closed for business: The “nontradables trap”: Theory The “nontradables trap”: Evidence Sunk costs, construction, and the expectations trap References 85 Africa’s Cities | Opening Doors to the World African firms face multiple obstacles to becoming Fragmented cities thus have higher business costs. competitive on global markets. Among the usual They are less attractive sites for investment in suspects are corruption; low access to finance; limited infrastructure and in commercial, industrial, and domestic goods markets; and unreliable infrastructure residential structures. And they are less able to realize for water, transport, and electricity (Malikane 2015 scale economies. As a result, African firms find it reviews the main findings). harder to break into global markets. But this is not the end of the story. This difficulty retards income A less obvious constraint is the fragmented spatial growth in African cities and lowers expectations about form of the typical African city. This type of urban future income growth. As a consequence, residential form burdens firms with high costs. By hindering investment decisions get delayed, resulting in a lower agglomeration economies and necessitating high quality housing stock. This contributes further to the wages, fragmented cities lock African firms into fragmented and high cost urban form of African cities, producing nontradables — a low urban development creating a low level development trap in which many trap. African cities appear to be trapped. When urban forms are fragmented, workers and The next two chapters present evidence and analysis households face higher living costs — and firms face in support of this view of African cities. Research for constraints and indirect costs as a result. For example, this report confirms that Africa’s urban living costs commuting by vehicle is unaffordable for most are relatively high and that this contributes to high workers, who thus have difficulty finding jobs that labor costs. Wages are about 15 percent higher in match their skills. High food and building costs directly African cities than in non-African ones (at official challenge families who are trying to remain healthy exchange rates, conditional on national per capita and secure decent housing. But when firms pay a high GDP), while overall labor costs are estimated at up to nominal wage to cover workers’ high living costs (or 50 percent higher. Chapter 3 documents these facts compensate them for living in poor conditions), that about cost and wages. Chapter 4 shows how these higher wage — driven by high urban costs instead features, together with those described in chapters 1 of high labor productivity — makes the firms less and 2 can combine to create an urban development competitive. They can survive selling nontradable trap, in which African cities become locked into poor goods in local markets, but are not able to compete in urban form and unable to break in to production of tradable goods production or break into international internationally tradable goods and services. markets. Reference Malikane, Christopher. 2015. “The Theory of the Firm in the African Context.” In The Oxford Handbook of Africa and Economics: Volume 1: Context and Concepts. Edited by Célestin Monga and Justin Yifu Lin. Oxford, United Kingdom: Oxford University Press. 86 Chapter 3 Costly for households, costly for firms African households face higher costs relative to their per capita GDP than households in other regions. According to a new study of price level indices at the urban level, based on data collected by the International Comparison Program, urban households in 39 Sub-Saharan African countries pay 20–31 percent more overall than urban households in other countries at similar income levels (Nakamura and others 2016). A similar comparison of urban prices — based on data from 125 cities, including nine in Sub-Saharan Africa (the Economist Intelligence Unit’s Worldwide Cost of Living Survey) — finds African cities to be about 31 percent more costly for households than cities in low- and middle-income countries elsewhere (Nakamura and others 2016). 87 Africa’s Cities | Opening Doors to the World Housing and transport are especially costly in urban In deciding where to live, households choose the best Africa. Relative to their income levels, urban residents home they can afford, with the amenities they value pay 55 percent more for housing in Africa than they do most, and make tradeoffs in allocating their budgets in other regions. over time. Some choose to live in the center of a city, where rents are higher, sacrificing size for access Urban transport, which includes prices of vehicles to amenities. When public amenities are limited, and transport services, is about 42 percent more basic services lacking, and connective infrastructure expensive in African cities than in cities elsewhere. deficient, households may underconsume housing Urban workers in Sub-Saharan Africa incur high and make suboptimal decisions on housing quality commuting costs — or they simply cannot afford to in order to access these services. For example, when commute by vehicle, leaving them no option but to transport systems are unavailable, households may walk (or possibly bike). The informal, often colorful choose to live in close-in slums so that they can minibus systems that dominate collective motorized access jobs, schools, or health care. Recent work transport in most African cities are far from cost- suggests that low investments in formal housing may efficient: the buses’ small size and low load factors be pushing households into searching for informal (passenger capacity) prevent them from realizing solutions. scale economies. For the poorest urban residents especially, the cost of vehicle transport in some cities The high costs and lack of amenities faced by is prohibitive. The need to walk to work limits these households also matter for firms. Workers need to residents’ access to jobs. The price premium for food be compensated for the high costs of living that they is also large (about 35 percent). face, which translates into higher wage costs for firms. Wages are generally at least 15 percent higher than wages in comparable countries. High prices, low incomes Price levels are generally higher in high-income The figure is based on data from the 2011 round countries, and they are higher in urban areas than of the International Comparison Program. It covers rural ones (Nakamura and others 2016). Goods 62 countries (including 39 in Sub-Saharan Africa), and services in lower-income African countries are with price level index data collected mainly in urban generally less expensive than in higher-income areas. Where the price information is not entirely countries. For example, the price level of food and urban based, adjustment was made using within- nonalcoholic beverages in Ethiopia is almost half that country data for urban–rural differentials.20 Price in the United States. Within countries, urban areas level indices are calculated by dividing the purchasing generally have higher prices than rural, partly because power parities by the nominal exchange rate for each the costs of commuting, land, rent, and some goods country. are also high. Econometric analysis of these data show that, These relationships hold quite widely, but the data controlling for income levels, price levels for reveal that African cities have particularly high prices household expenditures (excluding housing rent) are relative to their level of development. Figure 3.1 plots on average 31 percent higher in Sub-Saharan African a measure of the cost of living in cities in various countries than in other countries. A group of relatively countries relative to their GDP per capita (adjusted expensive countries includes Angola, Mozambique, for purchasing power parity). It confirms that richer Malawi, Niger, Chad, and the Central African Republic. countries generally have higher price levels, but it also By contrast, The Gambia, Mauritania, Madagascar, shows that African countries face price levels that are and Tanzania have relatively low price levels. Map 3.1 higher than expected given their low income levels (overleaf) illustrates the spatial pattern in Sub-Saharan (Nakamura and others 2016). Africa by showing the residuals from the regression. 88 Part 2 | Africa’s Low Urban Development Trap Chapter 3: Costly for households, costly for firms FIGURE 3.1 African cities face high prices for their income levels 210 190 Sub-Saharan African countries Other economies 170 Fitted values Adjusted price level index 150 Angola Gabon S. Africa 130 Namibia Central Equitorial African Congo Guinea Republic Mauritius Botswana Swaziland Congo. 110 Dem. Rep. Guinea- Chad Ghana Bissau Mozambique Sudan Malawi Lesotho Zambia Liberia Senegal Togo Nigeria Burkina Faso Benin 90 Niger Cameroon Mali Kenya Rwanda Guinea Sierra Leone Mauritania Burundi Uganda 70 Gambia Madagascar Tanzania Ethiopia 50 6 7 8 9 10 11 12 Log of GDP per capita (2011 PPP$) Source: Nakamura and others (2016), based on 2011 International Comparison Program data. 89 Africa’s Cities | Opening Doors to the World Mauritania Mali Cabo Niger Verde Sudan Eritrea Chad The Senegal Gambia Burkina Guinea- Faso Djibouti Bissau Guinea Benin Nigeria Somalia Sierra Leone Côte Togo Ethiopia D’Ivoire South Central African Ghana Sudan Liberia Republic Cameroon Equatorial Guinea Uganda Kenya Sao Tome and Principe Gabon Democratic Rwanda Republic of Congo Congo Burundi Seychelles -0.440 to -0.299 Tanzania -0.299 to -0.087 -0.087 to 0.043 0.043 to 0.172 Comoros Angola 0.172 to 0.384 Malawi no data Madagascar Zambia Zimbabwe Namibia Mozambique Botswana Mauritius MAP 3.1 Swaziland Relative expensiveness of Lesotho household consumption in South Africa Sub-Saharan Africa Source: Nakamura and others 2016. 90 Part 2 | Africa’s Low Urban Development Trap Chapter 3: Costly for households, costly for firms Similar findings are obtained using data collected What goods and services are most important in by the Economist Intelligence Unit. These data are driving this price premium? The Africa price premium compiled for quite different purposes — the cost varies depending on the groups of goods and of living of expatriates traveling from developed services. It is particularly large on essentially urban countries for business. Its price survey therefore commodities, such as housing (55 percent), followed collects data on items typically consumed by by communication (46 percent), and urban transport expatriates. Using these data, Nakamura and others (42 percent). But food and nonalcoholic beverages are (2016) find a highly significant positive Africa effect, also relatively expensive in African cities (35 percent indicating that (controlling for income levels), African premium); particularly expensive among food items cities are about 30 percent more expensive than are fresh or chilled vegetables, eggs and egg-based comparable cities elsewhere. products, and fresh milk (figure 3.2). FIGURE 3.2 Consumers in Africa face high price premiums Household consumption 31% except for housing Household consumption 26% except for housing and food Food and nonalcoholic beverages 35% Housing rent 55% Transport 42% Communication 46% Restaurants and hotels 41% Source: Nakamura and others 2016. Given the expenditure patterns of urban households 2003 and 2010, spending on food accounts for 39–59 in Africa, higher prices of food deepen livability percent of monthly expenditures by urban households challenges for households and impose a severe (figure 3.3). The poorest households (households in constraint on the choices they have on where to live the bottom expenditure quintile) spend an even larger or work. The budgets of the poor are spent mainly on share on food, ranging from 44 percent in Uganda food, reducing their opportunities to spend on health, to 68 percent in Zambia. Rising incomes allow for education, and housing. According to household consumption of nonfood items, such as housing, surveys collected in several African countries between recreation, and so forth (Dasgupta and others 2014). 91 Africa’s Cities | Opening Doors to the World FIGURE 3.3 Households in African cities grapple with subsistence 80% 70% 68% 66% 66% 66% 59% 60% 61% 60% 60% Share of food expenditure 54% 54% 51% 50% 44% 44% 40% 30% 20% 10% 0% da on ire r a ia da ue on a i ia l ge an iop ny law mb ga an ero ’Ivo Ni Gh an biq Le Ke Ma ne Ug m d Eth Rw m ra Za Se Ca ôte za ier C Mo S Average Poorest Source: Data from the Survey-based Harmonized Indicator Program (SHIP) of the World Bank’s Poverty Reduction and Economic Management Statistics Practice Group, Africa Region. Two other observation about the cost of urban living of property value). Obtaining construction permits in in Africa are important. The first is that building Sub-Saharan Africa takes on average 171 days and formally is expensive. Registering property in Sub- costs 737 percent of income per capita. This average Saharan Africa is generally more time consuming and time is comparable to other regions; it is lower than costly (relative to property value) than in other regions in South Asia, Latin America and the Caribbean, and of the world (figure 3.4); dealing with construction Europe and Central Asia. However, the average cost permits is fairly quick but still costly (relative to income is second only to South Asia and, at 84 percent of per per capita) (figure 3.5). On average in Sub-Saharan capita income, nearly nine times as expensive relative Africa, it takes 59 days and 9 percent of property value to incomes as in the high-income OECD countries to register property — more than twice as long and (World Bank 2015). The high costs of obtaining three times as much as in Europe and Central Asia property registration and construction permits in (26.5 days and 2.8 percent of property value) and Sub-Saharan cities contributes to further growth of high-income OECD countries (24 days and 4.4 percent informal settlements. 92 0 50 100 150 200 250 300 An go B la Bu Bots enin rki wa FIGURE 3.4 na na Fa Time (days) Ce Ca Buru so ntr al b n Afr Ca o Ve di ica me rde n R ro ep on ub Co l ng C Ch ic o, om ad De or Co m. R os Cô ngo, ep. te Re d’I p. vo Er ire Eth itrea io Ga G pia mb abo ia, n T Gh he Gu G a ine ui na a-B nea iss Part 2 | Africa’s Low Urban Development Trap K au Time and international benchmarks Le enya so Source: World Bank 2015, based on data from Doing Business. Ma Lib tho da eri ga a s Ma car la Chapter 3: Costly for households, costly for firms Ma M wi ur ali it Cost Mo Mau ania za riti mb us Na ique mi Sã b oT N ia om Ni iger g é a R eri nd wa a Pr nda in S cip Se ene e Sie ych gal r e So ra Le lles So uth A one uth fr Su ica da S Sw ud n az an Ta ilan nz d an i T a Ug ogo an Z d Zim am a ba bia La Eu East ti r o A b we Mi n Am pe & sia & dd er Ce Pa le ic n Ea a & tra ific c Average time and cost to register property in Sub-Saharan countries st l OE & N Carib Asia CD or be hig th A an Su h fr b-S So inco ica ah ut me ara h A n A sia fric a 93 0 4 8 12 16 20 24 Cost (% of property value) 94 0 100 200 300 400 500 An go Bo Ben la Bu ts rki wa in na na FIGURE 3.5 Ce F Time (days) nt Ca Bur aso ra lA b o V und fri Cam er i ca n R ero de ep on Co ub ng C C lic o, om had D o Co em. ros Cô ngo Rep te , Re . d’I p. v Eth oire i Ga G opia mb ab ia, on Time GhThe Gu a ine Gu na a-B ine iss a K au Le enya Africa’s Cities | Opening Doors to the World so Ma Lib tho da er Cost ga ia Source: World Bank 2015, based on data from Doing Business. Mascar law Ma i ur Mal i i Mo Mau tani za rit a mb ius Na iqu countries and international benchmarks mi e Sã b oT N ia om Ni iger é a R ger nd wa ia Pr nda in S ci Se ene pe Sie ych ga r e l So ra L lles e So uth one ut Afr h S ic ud a a Sw Sud n a an Ta zilan nz d an T ia Ug ogo a Zim Za nda mb La Eu Eas ti b ab ia ro t A we Mi n Am pe si dd & a& le eric Ce Pa Ea a & ntr ci s a fi OE t & N Cari l As c CD or bb ia hig th A ean Su h i fri b- n c Sa Sou coma ha th e ra Average time and cost to deal with construction permits in Sub-Saharan n A Asia fri ca 0 8 16 24 32 40 Cost (% of warehouse value) Part 2 | Africa’s Low Urban Development Trap Chapter 3: Costly for households, costly for firms The second observation is that there appears to higher “exposure” for the largest city in a country be a positive association between urban costs and is associated with a lower urban price index. When the extent to which a city fails to provide density the urban form is fragmented, economies of scale or is fragmented. Estimates of the cost of proving in service delivery are sacrificed, opportunities for urban infrastructure indicate that doubling urban agglomeration economies are lost, and transportation density reduces the per capita cost of a package of is more expensive, because people are dispersed infrastructure improvements by about 25 percent and more kilometers of road network are needed. (Foster and Briceno-Garmendia 2010). The decrease Ordinary least squares regressions show that a 1 is particularly large for infrastructure associated with percent increase in the Puga index is associated with high capital cost per capita, as shown in figure 3.6. urban costs that are lower by 12 percent, controlling for income levels and city population. Alongside Figure 3.7 also shows that a fragmented urban form is high urban costs, lack of urban amenities and high associated with higher costs. Using the Puga measure congestion reduces household wellbeing in Africa’s of urban fragmentation (discussed in chapter 2), cities. FIGURE 3.6 Infrastructure costs per capita decrease with density 1,400 Secondary City 1,200 1,000 $ per capita 800 600 400 200 0 C 0 5,000 10,000 15,000 20,000 25,000 30,000 Density (people/km2) Private tap High quality roads Power grid Source: Foster and Briceno-Garmendia 2010. Source: Foster and Briceno-Garmendia 2010. 95 Africa’s Cities | Opening Doors to the World FIGURE 3.7 A fragmented urban form is associated with higher urban costs 4 2 Price index 0 -2 -4 -1.0 -0.5 0.0 0.5 1.0 People within 10km of average worker Source: Henderson and Note: Figure shows a residual-on-residual plot. The x-axis plots the residuals from a regression of the Nigmatulina 2016 and Puga10 Index, log scale (based on Henderson and Nigmatulina 2016), controlling for log GDP per capita, Nakamura and others 2016. log population, a dummy for Sub-Saharan Africa, and the percent of urban population. The y-axis plots the residuals of the adjusted price index, log scale (based on Nakamura and others 2016) on the same controls. The lower the number of people within 10 kilometers of the average worker, the higher the price index. High wages, high costs of doing business Chapters 1 and 2 point to many aspects of African percent (in nominal terms) over equivalent firms in cities that impose costs on firms. An important other developing country cities (figure 3.8). Slightly additional factor is that firms need to pay high larger estimates can be found when comparing nominal wages to compensate workers for their high industrial labor costs across countries with the same cost of living as well as for their poor living conditions. data: African firms paid 50 percent more in labor African firms pay significantly higher nominal wages costs than equivalent firms elsewhere (Gelb, Meyer, than firms in other regions at comparable levels of and Ramachandran 2013). Higher wages may be one real income. reason why Africa’s manufacturing sector is so small and has been declining (as a share of global output) Data from World Bank Enterprise Surveys indicate since the 1980s (UNIDO 2009). Today, the average firm that urban wages in manufacturing are higher in in Africa hires about 20 percent fewer employees than African cities than in other cities at comparable levels equivalent firms elsewhere (Iacovone, Ramachandran, of economic development. Manufacturing firms and Schmidt 2014). in African cities pay a wage premium of about 15 96 Part 2 | Africa’s Low Urban Development Trap Chapter 3: Costly for households, costly for firms FIGURE 3.8 Nominal manufacturing wages in African cities are higher than in other developing country cities 10 Africa Other developing countries Buenos Aires Fitted values Fitted values Santiago Beirut Johannesburg Sao Paulo 9 Belgrade Montevideo Minsk Tunis Beijing Djibouti San Bogota Asuncion Lima Jose Mexico Amman Kingston S Domingo Harare Skopje La Paz Guat Log nominal wage, 2010 US$ Tirana Panama city 8 Dakar Casablanca Tegucigalpa Sofia Juba Ulan Batar Blantyre Kigali Nairobi Managua Kiev Mumbai Tbilisi Bishkek Cairo Noukchott Dar Jakarta Kabul Yerevan Addis Dhaka Accra 7 Ababa Lusaka Vientienne Ouaga Abidjan Kampala Kinshasa Bamako 6 Conakry 5 5 6 7 8 9 10 Log GDP per Capita, 2010 US$ Log GDP per capita, 2010 US$ Source: Data from World Bank Enterprise Surveys. 97 Africa’s Cities | Opening Doors to the World FIGURE 3.9 Sales revenue per worker in African and other developing-country cities 11 Johannesburg Africa Buenos Aires Sao Paulo Other developing countries Santiago Montevideo Beirut Fitted values Belgrade Fitted values Tunis Beijing Lima San Bogota Jose Mexico Djibouti Amman Kinshasa Asuncion Kingston Panama city S Domingo Harare Casablanca Nairobi Skopje 10 La Paz Log sales pwer worker, 2010 US$ Dakar Guat Kigali Minsk Tegucigalpa Managua Noukchott Ulan Batar Yerevan Ouaga Cairo Addis Bamako Ababa Kabul Bishkek Tirana Mumbai 9 Dar Tbilisi Kiev Juba Kampala Accra Lusaka Blantyre Dhaka Jakarta Vientienne Conakry Abidjan 8 5 6 7 8 9 10 Log GDP per capita, 2010 US$ Source: Data from World Bank Enterprise Surveys. 98 Part 2 | Africa’s Low Urban Development Trap Chapter 3: Costly for households, costly for firms African firms could afford to pay high wages only if revenue per worker is high. This turns out to be the References case, by about 25 percent (figure 3.9). It is therefore Dasgupta, P. 2014. “Measuring the important to ask: “Do these higher revenues reflect Wealth of Nations.” Annual Review of higher productivity (output per worker) or do they Resource Economics 6 (1): 17–31. simply reflect higher prices?” Gelb, A., C. Meyer, and V. Ramachandran. 2013. “Does Poor Mean Cheap? A The implications of high costs is that urban wages Comparative Look at Africa’s Industrial in manufacturing are higher in African cities than Labor Costs.” Working Paper 325, Centre for Global Development, Washington, DC. in other cities at comparable levels of economic development. When a city’s urban wage is higher Foster, V., and C. Briceno-Garmendia. 2010. Africa’s Infrastructure: than the international wage (for tradables), it makes A Time for Transformation. it harder for the city to break into global markets. In Washington, DC: World Bank. contrast, in the nontradable sector, higher nominal Henderson, Vernon, and Dzhamilya wages may exist in the absence of productivity gains, Nigmatulina. 2016. “The Fabric of because, absent competition, firms can pass labor African Cities: How to Think about Density and Land Use.” Draft, April costs on to local consumers (Venables 2016). As 20, London School of Economics. labor costs continue to rise in China and other Asian Iacovone, L., V. Ramachandran, and countries, international firms will be searching for new M. Schmidt. 2014. “Stunted Growth: cities in which to invest and set up industrial plants. Why Don’t African Firms Create More Jobs?” Working Paper 353, Center for Global Development, Washington, DC. Nakamura, S., R. Harati, S. Lall, Y. Dikhanov, N. Hamadeh, W. V. Oliver, M. O. Rissanen, and M. Yamanaka. 2016. “Is Living in African Cities Expensive?” Policy Research Working Paper 7641, World Bank, Washington, DC. UNIDO (United Nations Industrial Development Organization). 2009. Breaking In and Moving Up: New Industrial Challenges for the Bottom Billion and the Middle-Income Countries. Industrial Development Report 2009. Vienna: UNIDO. Venables, A. J. 2016. “Breaking into Tradables: Urban Form and Urban Function in a Developing City.” University of Oxford, United Kingdom World Bank. 2015. Stocktaking of the Housing Sector in Sub-Saharan Africa: Challenges and Opportunities. Washington, DC: World Bank. 99 Africa’s Cities | Opening Doors to the World 100 Chapter 4 Africa’s urban development trap An urban area that is difficult for commuters and firms because of a fragmented plan, lack of affordable transportation, and unexpectedly high labor costs might as well post a “Closed for business” sign. One that looks unlivable, without decent housing and amenities, could set up an “Out of service” sign as well. 101 Africa’s Cities | Opening Doors to the World A city that is a candidate for posting these signs has customers and their suppliers and on workplaces already entered a low development trap. Potential being reachable from residential areas. Investment investors and trading partners quickly see the in housing will occur if there is increasing demand, evidence of the physical and economic dysfunction driven by rising incomes of workers, and if revenues that constrains public service provision, inhibits labor from a growing city are available to finance market pooling and matching, and prevents firms infrastructure. These decisions are interrelated — from reaping scale and agglomeration benefits. They and in all of them expectations are crucial. Investors’ stay away, fearing lack of return on their investment. low expectations become self-fulfilling, as failure to This dynamic will keep Africa’s urban economies implement one project reduces the return to others, undercapitalized, making their development even locking cities in a low development trap. more challenging than it otherwise would be. This chapter looks at the form and function of City and country governments should recognize African cities to spotlight key inefficiencies and their the problem for what it is: not simply one of immediate effects — the signs that warn business underinvestment leading to low infrastructure, but away, limiting Africa’s appeal to the business one of the interdependency of many investment world and consigning its cities to producing mostly decisions. Business investment decisions depend nontradables in the informal sector. on the presence of other businesses that are their Cities closed for business Chapters 1–3 laid out key features of developing one outcome or the other. In one outcome, a city country cities and documented the low levels of produces only nontradable goods, with low land capital, lack of connectivity, and high costs facing values, little investment in buildings, and consequent households and firms in African cities. There are failure to achieve the scale and density required for strong interlinkages between all these elements. modern tradable production. The other outcome Urban form (the built fabric or physical environment) has the city producing tradable goods and services shapes urban costs, such as commuting expenses alongside nontradables. Land values are higher, as and rents. Costs determine the urban function are expectations of rents, incentivizing developers to (the mix of activities undertaken in the city), in turn invest more in building a taller and denser city. In both shaping productivity, labor demand, wages, and rents. cases, expectations become self-fulfilling: If developers There are feedback loops, as expectations about anticipate increased tradable production, their the city’s performance — employment, productivity, expectations of wages and land prices will be higher, wages, rents — determine investment decisions and leading them to invest more in taller buildings. “Taller” hence urban form. These interactions can lead to cities have greater economic density than “flatter” virtuous cycles of rapid urban development — or to a cities, increasing the likelihood of agglomeration development trap exhibiting the features described in economies and the future growth potential of the city chapters 1–3. and decreasing urban costs. This chapter pulls the key elements together. It first looks at the ways in which business investment The “nontradables trap”: Theory decisions in different sectors depend on the way the Africa’s failure to industrialize, and to create the jobs city functions, in particular its urban form. It then that come with industrialization, are a major cause of turns to residential investment decisions, in particular concern. Many factors lie behind this failure. The focus the ways in which they depend on expectations about here is on the role that poorly functioning cities play, future city growth. drawing on the framework developed by Venables (2016), which reflects the interactions that shape the Much of the chapter is analytical, pulling together performance of a developing country city. findings from previous research and setting the stage for the chapters on housing and land tenure The key ingredients are supply and demand for (chapter 5) and infrastructure (chapter 6). The labor in the city. The supply curve gives the wage analytic framework demonstrates how the different that the city has to offer to attract population from elements of the city fit together and highlights the neighboring areas (figure 4.1). It slopes upward fact that there can be several equilibrium outcomes. because cities impose “urban costs” — the additional Expectations can be self-fulfilling, leading a city to costs workers face when living in the city because of 102 Part 2 | Africa’s Low Urban Development Trap Chapter 4: Africa’s urban development traps high rent, commuting costs, and the price of many rents, tax revenues, and foreign aid. The curve is goods (as shown in chapter 3). These costs increase downward sloping because of diminishing returns; the with city size. Urban costs — and hence the height and more workers employed produce nontradables, the slope of the supply curve — also increase with urban greater will be the supply of the goods they produce inefficiencies in delivering housing, transportation, and hence the lower the price of the goods. Given and public services. Labor supply will be forthcoming local demand, the more workers are crammed into only if wages rise to offset the costs of living in a city. this sector, the lower will be their earnings. The demand curve is drawn with two distinct sections, In contrast, the price of tradable goods is set on the one representing labor demand in the nontradable world market. A city’s export activities do not run into sector, the other in the tradable sector. Nontradables a constraint set by the size of the local or regional are goods and services that are sold within the city market. The price of imports (and import-competing and its hinterland (perhaps extending to national or products) is set largely by the possibility of supply even regional sales). They include beer and cement, from the rest of the world. The demand for labor construction, and many services, including retailing, curve is therefore relatively flat. It is drawn in figure as well informal sector activities. Demand for these 4.1 as upward sloping because these sectors are likely goods and services comes from income generated to experience agglomeration economies, creating within the city and its hinterland and from income increasing returns to scale. transferred from outside sources, such as resource FIGURE 4.1 Supply and demand for labor in a “local” city Labor demand Labor supply WN (high urban costs) Labor demand WT N $ per worker w=WN w* wo CBD Population (= area x density) Source: Venables 2016. Note: CBD = Central Business District. W0 is outside wage. W* is the international wage, or the threshold for which the city is competitive in tradable production. WN is the wage in the local city, WT is the wage in the global city. 103 Africa’s Cities | Opening Doors to the World FIGURE 4.2 Supply and demand for labor in an internationally competitive city Labor demand Labor supply WN (high urban costs) T w=WN Labor demand WT $ per worker N w* wo CBD Population (= area x density) Source: Venables 2016. Note: CBD = Central Business District. W0 is outside wage. W* is the international wage, or the threshold for which the city is competitive in tradable production. WN is the wage in the local city, WT is the wage in the global city. What is the outcome of the interplay between supply demand at point T, where both the nontradable and and demand? If urban costs are high or increase tradable sectors are active. The larger the city, the sharply, then the situation is as in figure 4.2. Labor flatter the labor supply curve, although nominal wages supply equals labor at point N. Four observations can may be high. Firms in the tradable sector can afford be made. First, nominal wages are high. Second, real to pay these wages because employment in these wages remain low (they are set by the supply of labor sectors is high; agglomeration economies increase from outside the city; high nominal wages merely productivity. compensate for high urban costs). Third, nominal The argument so far shows how urban form — urban wages are too high to attract any tradable sector costs and hence the shape of the labor supply curve production, so the city produces only nontradables. — determine the size of the city and its production Fourth, unlike tradable sector firms, firms in the structure. Why is production of only nontradables nontradable sector are able to afford to pay these high considered a “trap”? In figure 4.2, there are three wages by passing them on to consumers, contributing points at which labor supply equals labor demand. further to the high cost of living in the city. The middle one is unstable and can be ignored, but Figure 4.2 illustrates an outcome in which urban costs points N and T are both sustainable outcomes. At N increase less rapidly with city size, so the labor supply the city produces only nontradables; at T both sectors curve is flatter. In this case, labor supply equals labor are active. 104 Part 2 | Africa’s Low Urban Development Trap Chapter 4: Africa’s urban development traps Both outcomes are sustainable because The “nontradables trap”: Evidence agglomeration economies make it hard to start Chapter 3 established that the high-cost/high-wage producing in new sectors. At point N the wage is configuration is a feature of African cities. There is also too high to trigger tradable production, because plenty of evidence that African cities — and countries productivity in the tradable sector is low — because — have not broken in to supplying internationally there is no tradable production, there are no tradable manufactured goods or services. The agglomeration economies. contrast between Asia and Africa is illustrated in This situation stems from coordination failure, a figure 4.3, which shows the shares of firms in tradable standard problem in developing new clusters of and nontradable sectors in selected cities. It shows economic activity (see, among others, Henderson and that the share of tradables in Asian cities is about 70 Venables 2009). It arises when a group of firms (or percent, 20 percentage points higher than in African individuals) has the resources to achieve a desirable cities (about 50 percent). outcome but fail to do so because they cannot (or do Africa’s failure to industrialize is a cause for concern not) coordinate their decision making. The chicken and because much of the growth in developing countries egg problem is that no firm wants to be the first to set since the 1980s has been linked to the expansion up but many would become established if they could of industrial production and higher-technology coordinate their entry. exports (Nallari and others 2012). Rapidly growing To solve the problems, cities need either a forward- countries, like China, have switched from exporting looking group of firms to harmonize their plans and mainly resource and agro-based products to make a move together or a large-scale land developer increasingly participating in global production chains, or city government with strong expectations that in particular high-tech products like optical devices, will establish its credibility by making irreversible transportation equipment, and domestic appliances investments in the new location (Henderson and as well as related assembly services. As countries Venables 2009). Absent coordination, agglomeration are increasingly specializing in tasks rather than economies will not be realized, leaving all firms stuck products, trade statistics are now moving from “naive” in the suboptimal equilibrium. Cities will develop at a export valuation to domestic value added measures. smaller scale, with fragmented neighborhoods, as they The switch is particularly important in countries have in Africa. like China, whose share of processing exports is above 50 percent. In high-tech sectors, the share of Three messages emerge from this analysis. First, it is foreign value added reaches 80 percent, resulting possible for real incomes to be low but nominal prices in an overestimated trade value for these products and wages high. In this scenario, the city produces (Koopman and others 2008). only nontradables, with high prices passed on to the local market; the city is too high cost for tradable The big winners in China were “exports of electronic sectors to be able to operate. and telecommunications products and office equipment, the shares of which grew from 5.4 percent Second, even if urban costs are lower, it may be hard in 1985 to more than one-third in 2006” (Nallari to break out of the nontradables trap. Coordination and others 2012). Many other countries in East and failure means that the city is stuck at point N in figure Southeast Asia experienced similar transitions in their 4.2, and it is not profitable for any firm to start up export mix during the first decade of this century tradable production. (table 4.1). By contrast, the majority of African exports Third, there is a dichotomy of urban types. Some remain resource- and agro-based. cities — including many in Asia — produce tradable goods at scale, bringing high productivity and high real incomes. Others — such as most in Africa — have not yet broken into this sector. They remain stuck at point N. 105 Africa’s Cities | Opening Doors to the World FIGURE 4.3 The tradables sector is much larger in Asia than in Africa Luanda Gaborone Dar es Salaam Kampala Kigali Bamako Accra Nouakchott Africa Nairobi Dakar Addis Ababa Kinshasa Lagos Niger Mombasa Lusaka Harare Middle East North Africa Tunis Beirut and Amman Cairo La Paz Asunción Latin America Caribbean Cordoba and the Medellin Bogota Buenos Aires Lima Yangon Pacific & South Asia East Asia and the Bangkok Zhengzhou City Dhaka Shenzhen City Delhi Chittagong 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Nontradable Tradable Source: Data from World Note: The data is from the latest WBE surveys post-2010 (with more than 15,000 firms in capital Bank Enterprise surveys cities, or cities of at least one million inhabitants, and with at least 50 firms sampled). Only firms conducted since 2010. with five or more employees are interviewed. The sectoral specialization analyses used the UN International Standard Industrial Classification of All Economic Activities (3.1 revision). Manufacturing, wholesale and commission trade, and business services (such as travel agencies, transport, financial intermediation) are all tradable activities. By contrast, construction, local services, retail trade, health and social work, and other local activities are classified as nontradable. 106 Part 2 | Africa’s Low Urban Development Trap Chapter 4: Africa’s urban development traps Table 4.1 Top 10 commodity exports from Asia and Africa, 2000–10 Commodity Trade value (billions of dollars) Asia (East Asia, South Asia, and Oceania) Electrical, electronic equipment 7,409.6 Nuclear reactors, boilers, machinery, etc. 5,044.7 Vehicles other than railway, tramway 2,175.4 Mineral fuels, oils, distillation products, etc. 1,926.4 Optical, photo, technical, medical apparatus, etc. 1,085.4 Plastics 903.9 Articles of apparel, accessories, knit or crochet 798.2 Articles of apparel, accessories, not knit or crochet 774.1 Iron and steel 744.0 Pearls, precious stones, metals, coins, etc. 722.5 Africa Mineral fuels, oils, distillation products, etc. 1,224.4 Pearls, precious stones, metals, coins, etc. 169.7 Iron and steel 74.3 Ores, slag, and ash 64.0 Articles of apparel, accessories, not knit or crochet 62.2 Vehicles other than railway, tramway 59.3 Electrical, electronic equipment 57.4 Nuclear reactors, boilers, machinery, etc. 54.3 Cocoa and cocoa preparations 43.5 Inorganic chemicals, precious metal compound, isotopes 35.5 Source: Data from UN Comtrade Database. Globally, urbanization is strongly correlated with the Sub-Saharan Africa has not developed in this way: its expansion of manufacturing (figure 4.4). For most pattern of growth has been described as “urbanization countries, manufacturing as a share of GDP rises with without industrialization” (Fay and Opal 2000; Jedwab urban shares until about 60 percent of the population 2013; Gollin, Jedwab, and Vollrath 2016). This theory lives in cities and manufacturing accounts for about 15 argues that Africa’s urbanization growth has been percent of GDP. driven by natural resource exports, a third explanation for relatively high demand for nontradable services produced in cities (box 4.1). 107 Africa’s Cities | Opening Doors to the World FIGURE 4.4 Urbanization and manufacturing share of GDP in Africa and outside Africa Outside Africa 40 Manufacturing, value added (% of GDP) Fitted values CHN KOR 30 THA CZE BLR Manufacturing share of GDP HUN SVN MYS DEU PHL DNK SLV IDN IRL 20 LKA GTM LTU CHE JOR POL SGP ZAR MEX IND HND AUT VNM MAR TUN FIN KHM EGY TUR BGD NIC EST DOM SWE KGZ MDA ECU ARG LBN URY ITA RUS PAK GEO HRV BOL ESP SVK BEL AFG UZB PRT UKR COL BIH FJI CHL PRY FRA BRA KNA KAZ MKD LVA IRN NLD QAT 10 GRL SAU LAO ARM GBR JAM MNG OMN TON BTN AUS NPL GUY NOR ALB KWT MDV VCT MNE CYP LUX GRD BHS VUT DMA AZE LCA ATG MHL PLW HKG FSM 0 0 20 40 60 80 100 Urbanization rate (%) Source: Data from World Development Indicators. 108 Part 2 | Africa’s Low Urban Development Trap Chapter 4: Africa’s urban development traps Africa 40 Manufacturing, value added (% of GDP) SWZ Fitted values 30 Manufacturing share of GDP 20 MUS BEN CMR ZAF ZWE SEN MWI CIV KEN NAM 10 BDI MOZ LSO NGA UGA MRT COM SDN SYC NER GIN CAF BWA BFA GHA TZA COG ETH TGO TCD RWA STP GAB SLE 0 0 20 40 60 80 100 Urbanization rate (%) Source: Data from World Development Indicators. 109 Africa’s Cities | Opening Doors to the World BOX 4.1 Do natural resource exports explain why African cities specialize in nontradables? Countries can follow one of two paths to achieve endowments. Positive productivity shocks to the higher urbanization rates as they structurally resource sector shift workers into the sector and out transform. The first path moves workers out of of the food and tradable sectors. Surplus income agriculture and into manufacturing (box figure 4.1.1). generated from natural resource productivity shocks This type of structural change — the path taken by causes a disproportionate rise in the demand for most countries in Europe, Latin America, and Asia — urban goods and services (relative to food). This creates “production cities,” which produce tradable added demand is met largely through imports goods for domestic and international markets. (except for urban services, which are produced locally). Urbanization is driven by consumption, not The second urbanization path reflects the experience production, creating “consumption cities” (Gollin, of African countries with large natural resource Jedwab, and Vollrath 2016). BOX FIGURE 4.1.1 Urbanization and industrialization take different paths in resource exporters and nonresource exporter Nonresource exporters Resource exporters 100% 100% 90% SSA 90% Other 80% 80% Urban share of population in 2010 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0% 0% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Share of Manufacturing and Services in GDP in 2010 (%) Source: Gollin, Jedwab, and Vollrath 2016. 110 Part 2 | Africa’s Low Urban Development Trap Chapter 4: Africa’s urban development traps If this analysis is correct, urbanization in Sub-Saharan Disproportionate demand for nontradables can offset Africa may not generate the same productivity effects efforts to reduce urban costs — one of the factors that as urbanization in other developing regions — in keeps cities locked into local markets. part because of rapid growth in the natural resource In box figure 4.1.2, heavier demand for nontradable sector, which increases urban employment mostly in goods and services shifts the labor demand curve to the nontradable sector. The different growth paths the right. Starting from a city with two equilibria, an taken by countries that do and do not rely on natural increase in demand for nontradables leads to the resource exports appear in box figure 4.1.1. single equilibrium that specializes in nontradables. BOX FIGURE 4.1.2 Large natural resource rents lead to an equilibrium dominated by nontradables (urban “Dutch disease”) Labor demand WN Labor supply T $ per worker Labor demand WT N1 N2 w* wo CBD Population (= area x density) Source: Venables 2016. Note: N is a city producing nontradable goods for the domestic market. T is a city producing tradable goods for both local and international markets. CBD is central business district. W0 is outside wage. W* is the international wage, or the threshold for which the city is competitive in tradable production. 111 Africa’s Cities | Opening Doors to the World Sunk costs, construction, and the expectations trap Urban form depends on private investment decisions, than can be a lengthy and expensive (see chapter 5). including decisions made in the residential sector. Imperfect land markets make it difficult and expensive For formal sector housing, these decisions involve to assemble parcels of land of size sufficient to justify sinking costs in constructing long-lived structures. infrastructure investments and other large-scale Such decisions depend critically on expectations about expenditures. the future prospects of the city. If the city is expected The future returns from building a structure are to be “artisanal,” based on low-value production of extremely uncertain in many developing cities. nontradables, then land rents will be expected to Lack of clarity in land rights and political risk create remain relatively low and it will not be worth investing uncertainty about future rental incomes. Low in formal structures. The lack of incentive to invest expectations about the future growth of the city mean perpetuates the disconnectedness and high urban low expectations about the growth of rents. Even if costs that are one of the obstacles to investment. developers expect the city to grow, they might not Expectations are therefore self-fulfilling. In contrast, know where in the city that growth is likely to occur. more optimistic expectations increase investment An investor may be confident that growth will occur in formal sector structures, including residential somewhere in the city, but if it could be in any one structures, bringing down urban costs and making of many possible locations, the ensuing uncertainty the city more attractive as a place for local investment means that no one will invest. These issues reflect (flattening the labor supply curve). coordination failure. The costs involved in building durable formal sector There are mechanisms for overcoming such failures. structures include several components. Construction Sunk investments—made by the government or a costs are high, particularly in building tall. There may group of investors—are one. They provide a strong also be high conversion costs in going from informal signal to other potential investors and can have long- to formal. They include the costs of road layout run effects. It has been argued that “investments and the provision of water, sewerage, and other sunk historically, even small ones that have now infrastructure. They also include legal and institutional depreciated completely, might serve as a mechanism costs. Building durable structures is unlikely to take to coordinate contemporary investment” (Bleakley place until ownership of land is made clear, a process FIGURE 4.5 The share of land devoted to slums increases with distance from the city center 0.25 Share of land area 0.2 0.15 0.1 0.05 0 0 2 4 6 8 10 Distance from Center (kilometers) Source: Henderson, Regan, and Venables 2016. 112 Part 2 | Africa’s Low Urban Development Trap Chapter 4: Africa’s urban development traps 2012). This argument explains why places that have An inefficient level of housing investment has seen large sunk investments in the past (like towns implications that go beyond the housing sector. along rail lines) continue to attract investment today The supply curve of urban labor discussed earlier (Jebwab and Moradi 2016; Jedwab, Kerby, and Moradi depends on, among other things, the stock of housing 2015). available to workers. Inefficiency and undersupply shift this curve to the left, raising costs and restricting Given these high sunk costs and low or uncertain the supply of workers. This shift is one of the factors expected returns, it is not surprising that investment that makes it more likely the city will be stuck in in formal structures has been low. A further the nontradables trap. The coordination failure is consequence is the patchwork of building types therefore much wider than just between firms or observed in many African cities, with informal between property developers. It intersects both sets of settlements still present in central areas, adjacent to activities. There is an expectations trap, as developers modern development. will not construct housing unless they expect the city Henderson, Regan, and Venables (2016) argue that to grow (and other market failures in the housing informal settlement is likely to be part of an “efficient” market to be ironed out), and undersupply of housing urban structure in a growing low-income city but is one of the factors undermining the profitability of that the coexistence of formal and informal sector tradable sector production, which in turn retards the settlement near the center is inefficient. Their detailed growth of the city. study of Nairobi indicates that although there are no Countries in Sub-Saharan Africa are urbanizing remaining slum areas within 2 kilometers of the city rapidly. Some 472 million people live in urban areas center, beyond that 10–20 percent of land is occupied across Africa, and this number is expected to double by slums (figure 4.5). Slums are less tall than formal in the next 25 years (United Nations 2014). By 2030, areas but have a higher proportion of land area Sub-Saharan African cities need to create 160 million covered by buildings (so less space is used for road additional jobs. and other amenities). The building volume per unit land is similar in slum and formal areas. However, the Africa’s pattern of urbanization, however, is different land rent differential between uses is large, indicating from that underway in other developing regions, inefficient use of land. They estimate that the capital where increased urbanization has been accompanied value forgone by not developing Kibera, Nairobi’s by a rise in manufacturing activities. As two-thirds of largest slum, is on the order of $1 billion. Africa’s urbanization still needs to happen, now is the moment to make African cities more productive. Theory and evidence combine to paint a challenging picture of the current state of African cities. The urban landscape of African cities has generated economic inefficiencies that have increased urban costs. Low expectations have resulted in lack of investment in residential, commercial, and industrial buildings and in infrastructure. Such shortfalls in structures makes cities costly and forestall economic agglomeration. 113 Africa’s Cities | Opening Doors to the World References Bleakley, H., and J. Lin. 2012. “Portage and Path Dependence.” Quarterly Journal of Economics 127 (2): 587–644. Fay, M., and C. Opal. 2000. “Urbanization without Growth: A Not-So-Uncommon Phenomenon.” Policy Research Working Paper 2412, World Bank, Washington, DC. Gollin, D., R. Jedwab, and D. Vollrath. 2016. “Urbanization with and without Industrialization.” Journal of Economic Growth 21 (1): 35–70. Henderson, J. V., T. Regan, and A. J. Venables. 2016. “Building the City: Sunk Capital, Sequencing, and Institutional Frictions.” CEPR Discussion Papers 11211, Center for Economic and Policy Research, Washington, DC. Henderson, J. V., and A. J. Venables. 2009. “The Dynamics of City Formation.” Review of Economic Dynamics 12 (2): 233–254. Jedwab, R. 2013. Urbanization without Structural Transformation: Evidence from Consumption Cities in Africa. Jedwab, R., E. Kerby, and A. Moradi 2015. “History, Path Dependence and Development: Evidence from Colonial Railroads, Settlers and Cities in Kenya.” Economic Journal. Accepted Author Manuscript. Jedwab, R., and A. Moradi. 2016. “The Permanent Effects of Transportation Revolutions in Poor Countries: Evidence from Africa.” Review of Economics and Statistics 98 (2): 268–284. Koopman, R., Z. Wang, and S.-J. Wei. 2008. How Much of China’s Exports is Really Made in China? U.S. International Trade Commission. Office of Economics Working Paper, Washington, DC. Linden, G., K. Kraemer, and J. Dedrick 2007. “Who Captures Value in a Global Innovation System? The Case of Apple’s iPod.” Personal Computing Industry Center, Irvine, CA. Nallari, R., B. Griffith, and S. Yusuf. 2012. Geography of Growth: Spatial Economics and Competitiveness. Washington, DC: World Bank. United Nations. 2014. World Urban Prospects, the 2014 Revision. New York: United Nations. Venables, A. J. 2016. “Breaking into Tradables: Urban Form and Urban Function in a Developing City.” University of Oxford, United Kingdom. 114 Part 3 Springing Africa from Its Low Urban Development Trap Chapter 5 Clarifying property rights and strengthening urban planning Why African cities fail to attract investment: An urban planner’s perspective Urgent need to clarify land and property rights Strengthen land valuation Strengthen urban planning institutions Streamline urban regulations To build cities that work, make land markets work — nothing less will do References Chapter 6 Scaling up and coordinating investments in physical structures and infrastructure Invest in early infrastructure to shape urban structures Leverage road investment Provide public goods and services to enhance livability Finance for lumpy urban infrastructure investments References 115 Africa’s Cities | Opening Doors to the World To create an internationally competitive tradable 2. Coordinate and increase early infrastructure sector, African cities must cease to be crowded, investments. After taking firm and decisive steps disconnected, and costly, and instead become livable, to improve institutional structures, authorities can connected, and productive. How? build on those efforts to adapt physical structures and infrastructure — including housing, transport The answer lies in swift action by mayors and infrastructure (including roads), and basic services ministers to enhance the functioning of land markets — to a future of urban productivity. Leaders should (the factor market most urgently in need of reform) do this by: and to strengthen urban planning, regulation, and enforcement — followed by actions to coordinate • Making infrastructure investments early, and and scale up investment in cities’ physical structures coordinating them with land market intentions and infrastructure. In short, city leaders must make and with the plans and regulations that guide decisive and concerted efforts to: physical structures (ensuring that infrastructure investments will be integrated with the growth 1. Reform urban land markets and regulations. of neighborhoods and structures in predictable Leaders can act immediately to improve the ways). institutional and capacitive structures that govern land markets and land use — structures that • Intensifying these early, coordinated depend on human capital, and that will ultimately infrastructure investments to take full advantage determine a city’s ability to mobilize investment of scale economies in housing, transport, and capital. Leaders should do this by: services (avoiding inefficient and fragmented investments that diverge from market demand). • Simplifying and clarifying transfers of property rights among land market participants (freeing Both efforts should aim at structural improvements in these procedures from today’s unclear, the allocation of a city’s land, capital, and structures. overlapping property-rights regimes). Their aim should be to achieve urban development at scale and for scale, while fostering economic • Supporting the effective management of urban specialization. Chapters 5 and 6 provide findings from development through foresighted planning, emerging empirical studies on the key policy areas realistic regulation, and predictable enforcement. outlined above. 116 Chapter 5 Clarifying property rights and strengthening urban planning Over the next 20 years, the growth of Africa’s urban populations will propel new demand for infrastructure, housing and other physical structures, and amenities. To meet this new demand, city leaders and planners will need adaptable strategies. Plans and regulations should allow the best use of land — but they must also permit uses, and users, to change over time, as demand evolves further. Three key considerations are as follows: • How to handle land and property rights • How to manage land valuation and prices • How to strengthen land use and urban planning 117 Africa’s Cities | Opening Doors to the World Africa’s cities are not developing in a well-planned impractical (Felkner, Lall, and Lee 2016). Increased fashion. Instead, they grow informally and develop investments in roads could increase productivity, even informally. Public planning is ineffective; private while affecting commuting costs and times differently development is hobbled or repelled by opaque across the city — but only if such investments are well or inappropriate regulations. Informal dwellings thought through in advance. Similarly, African cities house not only poor but also middle-income have an urgent need for well-planned and forward- households, essentially because of constraints on thinking transportation systems. All transportation formal land markets. These land market constraints development plans are not equal.17 also do much to explain the typical African city’s The lack of physical and technological structures — spatial fragmentation and the relatively low capital housing, services, and transportation — in Africa’s investment near its core. cities points to the need for planning capacity. Without The crowded streets of African cities attest to a lack of proper local planning guidance, it is impossible to formative, integrated urban plans. Traffic congestion coordinate and implement infrastructure, public stems from limited road infrastructure, limited amenities, and other investments. No planning, or parking, and the lack of formal addresses. And the poor planning, is one of the fundamental reasons why informality and small scale of public and collective African cities are too crowded, too disconnected, and transport in African cities indicate that these networks too costly to attract regional and global investors and are mostly reactive — they emerge in response to the trading partners. city’s growth. They do not structure growth, as did the Reforming land markets is a necessity in any policy introduction of rail infrastructure or streetcars in well- effort to transform city growth patterns — to get developed cities like Paris and London (Brooks and African urbanization right. Success will bring large Lutz 2013). payoffs in economic efficiency. If land market reform Households in African cities find it difficult to locate enables Africa’s cities to tap the potential of rising land outside the central business district because the lack values, the ability to finance infrastructure and other of paved roads makes commuting from the periphery public goods will follow.18 Why African cities fail to attract investment: An urban planner’s perspective Beyond the poor connective infrastructure market failures, including inappropriate regulation discussed in chapter 2, three features of African and land titling or capital market imperfections. The cities directly explain their low appeal to investors, consequences of such imperfections are long-lasting. which undermines their prospects for economic development. The first feature is capital misallocation. Institutional constraints The second is institutional constraints. The third is Restrictive regulations are strangling development in ineffectual property rights. African cities, discouraging investment and limiting formal housing options for the poor. For example, Dar Capital misallocation es Salaam (box 5.1) requires lots to be at least 400 Central to the problem of “cities out of service” is square meters. Anyone who wants to buy a stand- capital allocation in urban structures, which is shaped alone house in the formal sector near the center must by sunk capital, sequencing, and institutional frictions. be able to afford a lot this size. The only ownership Because urban investments in structures are durable alternative is informal — and one-fourth of Dar es and long-lived, investment decisions are based on Salaam’s homeowners have no documentary proof of expectations about future land rents, as driven by ownership (World Bank 2015d). future incomes, populations, and policy. Another kind of regulatory burden can be imposed by Henderson, Regan, and Venables (2016) develop the system of land ownership itself. A majority of land a core analytic model to examine these issues. It in Kampala, Uganda operates under a complex land shows the importance of expectations in shaping tenure regime that recognizes independent rights over urban investment decisions: Low expectations of the land and structures, giving rise to legal disputes and city’s future development distort investment levels blocking development (Muinde 2013). below their efficient levels. Investment is deterred by 118 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning BOX 5.1 Inappropriate building regulations hamper affordability Land use regulations, zoning, and building regulations than one-tenth those allowed in other Asian cities. The are some of the most valuable tools for governments restriction increased housing prices by an estimated to guide development and promote livability. Yet 15–20 percent of income, according to Buckley and certain interventions in urban land markets can Kalarickal (2006). negatively affect affordability and access to serviced In Dar es Salaam, inappropriate size regulations make land if they are not benchmarked against what the local the majority of buildings de facto illegal, regardless population can afford to pay. Evidence from around of formal land title or the quality of the structure. the world indicates that inappropriate minimum Developments that are out of compliance (building standards actually increase informal development, areas below 375 square meters) are condemned to be even on formally titled land. unplanned and excluded from water and sanitation Many cities in India have imposed strict limits for services, making it extremely costly and difficult to building heights. In Bangalore, the policy resulted in redevelop the land later legally. A more effective horizontal low-density expansion of built-up area and approach would be to rationalize standards for increased housing costs by 3–6 percent of median development based on performance (for example, household income. In Mumbai, which is constrained structural integrity) and affordability by the local by the surrounding topography, the impact was even population. more pronounced. Building heights were limited to less BOX FIGURE 5.1.1 Distribution of buildings in Dar es Salaam by distance from center 80,000 75,000 Less than 375m2 Total buildings More than 375m2 50,000 25,000 0 0 10 20 30 Kilometers from city center Source: World Bank Staff calculations 119 Africa’s Cities | Opening Doors to the World In Nigeria, high costs and burdensome regulations In Dar es Salaam, for example, people live in Tandale have stymied the formal development of urban land. — the informal district — not for its services and Titling expenses alone can reach as high as 30 percent amenities but despite its lack of them. Its central of construction costs in Lagos and Port Harcourt, location puts people close to where most jobs and where total transaction costs range from 12 to 36 economic opportunities are. In 2010, Dar es Salaam’s percent of a property’s value (World Bank 2015c). informal housing areas were on average much closer Zoning can also push people out of the formal sector than formal ones to commercial and industrial areas and into unplanned development. In Ibadan in 2000, (figure 5.1). Similar evidence is found for Kigali. Across researchers found that 83 percent of homes violated cities, households are willing to compromise on living city zoning rules (Arimah and Adeagbo 2000). conditions. In Nairobi, most residents of informal settlements have jobs and comparatively high levels of As challenging as the stranglehold of overregulation education relative to those living in formal housing, yet is the lack of connective infrastructure. Given faster their living conditions remain basic (Gulyani, Talukdar, and more affordable transportation, more African and Jack 2010). This choice probably reflects the city dwellers might forgo a downtown location for a premium they place on accessibility. house with better amenities a few kilometers from the center. But long commutes are an obstacle for most residents (see chapter 2). Some people live on the outskirts of the city, but many others settle for more centrally located informal housing — the only affordable kind. FIGURE 5.1 Average distances to commercial and industrial areas from formal and informal settlements in Dar es Salaam, Tanzania and Kigali, Rwanda Dar es Salaam c. 2010 Average Distance from Residential Area to Commercial/Industrial Area Residential irregular Commercial/Industrial Residential regular to Commercial/Industrial 8.9 km Residential irregular to Commercial/Industrial 5.7 km 0 km 10 km 20 km Source: Antos, Lall, and Lozano-Gracia 2016. 120 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning Ineffectual property rights In West Africa, only 2–3 percent of land is held with a government-registered title (Toulmin 2005). Clear rights to urban land are a precondition for the emergence of a formal land market, whether The lack of a proper registration system prevents formal, customary, or informal. But many African urban land markets from functioning well. It creates cities struggle with overlapping and sometimes obstacles to the raising of capital for development and contradictory systems, severely constraining urban investment — and to the raising of revenue by the land redevelopment and imposing high costs. And local authority. even where formal titles or clear land rights exist, These dysfunctions distort the price and availability basic mapping, geographic, or ownership information of land for efficient urban development, prompting is often inaccurate or land records maintained poorly, recourse to informality in building. Henderson, Regan, causing disputes. and Venables (2016) distinguish between formal and Applying for formal recognition can also be a tedious informal sector construction. Formal buildings involve process. Land administration systems (such as sunk capital costs, can be built tall, and are hard to registries and cadaster records) are incomplete and modify once constructed. Because they are durable, underused for enforcing legal claims and landholders’ investment decisions are based on expectations fiscal obligations, so lenders cannot always use land about future land rents, as driven by future incomes as collateral. In Sub-Saharan Africa, only 10 percent of and populations. As the city grows, there will be rural land is registered (Byamugisha 2013). periodic demolition and redevelopment of formal Kigali c. 2010 Average Distance from Residential Area to Commercial/Industrial Area Residential irregular Commercial/Industrial Residential regular to Commercial/Industrial 5.8 km Residential irregular to Commercial/Industrial 4.6 km 0 km 10 km 20 km Source: Antos, Lall, and Lozano-Gracia 2016. 121 Africa’s Cities | Opening Doors to the World areas. The city may also contain informal or slum The example of Nairobi structures. Given the technology and materials used In Nairobi, informal building volume per unit of in construction, these building are not likely to be built land is lower than formal development (map 5.1). tall; they can be rebuilt and adjusted after their initial As land prices decrease with distance to the center, construction. The capital used in such structures is not the gains from conversion are larger closer to the sunk but remains malleable. central business district and usually exceed the cost The cost of converting informal to formal land use of conversion. For example, converting informal slums varies over time and even across properties in the to formal structure would increase building volumes city. The process can be hindered by institutional by up to 148 percent at 2 kilometers from the center issues. Barriers can include poor enforcement of and by 53 percent at 4 kilometers from the center land rights when multiple tenure systems coexist. (Henderson, Reagan, and Venables 2016). Most slum In slums, rules are often “split between a galaxy of structures are built with unimproved materials (57 private sector actors, landlords, chiefs, bureaucrats percent of slum dwellings are built with sheet metal and gangs” (Marx, Stocker, and Suri 2013, quoted and 15 percent of mud and wood) that are cheaper in Henderson, Regan, and Venables 2016). The core but not suitable for taller buildings. Ninety percent insight from the framework of Henderson, Reagan, of formal residences are built with stone, brick, or and Venables is that urban form (the size and shape of cement block. There has been redevelopment up to the city) is sensitive to the expected returns to durable about 2 kilometers from the center in Nairobi; beyond investments and to the costs of converting informal to that distance, development is much less than what formal sector usage. the model of Henderson, Reagan, and Venables (2016) predicts. More broadly, the lack of affordable formal housing in Africa’s cities stems partly from the difficulty of adding They argue that the 1,000 acres covered by the Kibera new stock onto the old with the building methods slum (3.5 kilometers from downtown Nairobi) is too now in use. Housing and other structures in Africa large and too complex for conversion costs to be employ cheap methods that do not allow substituting tractable. The land is owned by the government but capital for land (Bernard, Bird, and Venables 2016). managed by slumlords and political elites who control By contrast, in much of developing East Asia — as in the land and have no interest in redevelopment developed countries — building methods allow for (because they do not own the land), which would various floor areas and building heights. In Bangkok, take away their very profitable slum business. for example, the average height of the tallest buildings Rough calculations based on 1,000 acres of land at 4 is 62 floors; in Kuala Lumpur it is 84 floors. In contrast, kilometers from the center suggest that the gains of in Dar es Salaam it is just 15 floors. converting Kibera would reach $1 billion. The surplus generated by such a transformation could be used More than 60 percent of African’s urban population to help relocate tenants and potentially buy off the lives in areas with some combination of overcrowding, people blocking redevelopment, but transformation low-quality housing, and inadequate access to clean would require making deals with slumlords. water and sanitation (UN Habitat 2015). The result is urban dysfunction that, across the region, keeps expectations low and deters investment. 122 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning MAP 5.1 Average built height in Nairobi, 2015 Kibera slum CBD Slum Formal 21 49 18 39 15 32 13 25 11 20 9 16 7 12 5 9 3 6 Average height (m) Source: Modified from Henderson, Note: Map shows 150 × 150 meter cells. The boundary of the city spans about 22 kilometers east to west Regan, and Venables 2016. and 11 kilometers north to south. The map tilt may distort the appearance of distances. Background imagery copyright Airbus Defence and Space 2016 taken from the SPOT5 satellite, September 20, 2004. Clear land and property rights Clear rights to urban land are a precondition for the peri-urban and urban land, property rights depend emergence of a formal land market. Informal, illegal on the consent of local chiefs or family elders. Many markets can function in almost any conditions. But countries (mostly in Central Africa) have a wide range informality in land markets is distinctively limiting. of land rights systems in urban and suburban areas Informal asset transactions are viable only if the (traditional, customary, collective, religious, and purchaser can rely on some enduring, extra-legal “modern”). This diversity is problematic. means of having new ownership recognized (such Tenure insecurity, measured as the share of the as sanction by the local community). In contrast, a population with no formally recognized land tenure formal market does not merely offer purchasers the rights, increased in Africa from about 55 percent protection of the state, it also generates the public in 2009 to about 66 percent in 2012, with “strong good of accurate valuation through transactions that tension” more prevalent (and rising) in larger are readily observable and recorded. agglomerations than in average-sized towns (figure African cities struggle with overlapping and sometimes 5.2). Overlapping property rights are common in contradictory property rights systems — formal, Durban, South Africa (box 5.2) and in Ghana, Lesotho, customary, and informal (map 5.2). Under the Mozambique, and Zambia. customary rules for land tenure that control much 123 Africa’s Cities | Opening Doors to the World MAP 5.2 Diversity of land rights in urban and peri-urban Africa, 2009 Single Wide range No data Source: Data from Picarelli 2015. Note: The map plots an index of the diversity of land tenure rights systems in urban areas that ranges from 1 (wide range of systems) to 4 (single land system). No information was available for countries in grey. FIGURE 5.2 Larger agglomerations have higher, and increasing, tensions over land prices 100% 80% 60% 40% 20% 0% 2009 2012 2009 2012 Average sized towns Large agglomeration Strong tension Low tension Source: Data from Picarelli 2015. 124 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning BOX 5.2 Customary land rights in Durban, South Africa Customary property rights systems retain a strong The municipality and the board operate two entirely influence on land tenure in Africa, particularly in peri- different and separate administrative regimes, and urban and rural areas. For example, in Durban, areas their planning and implementation systems are not formally administered by the eThekwini Municipality aligned. The link to the formal administration occurs occupy just 36 percent of the eThekwini Metropolitan at the central level, where the board reports to the Area (EMA). Another 27 percent of the EMA falls under national Minister for Rural Development and Land non-scheme agriculture areas, administered jointly Reform. At the local level, the board in effect decides by local and provincial governments. The remaining how to govern and manage activities within its lands. 37 percent of the EMA is jointly administered by The board is required to inform and seek advice from the municipality and the traditional authority (the the municipality on land use decisions within the Ingonyama Trust Board), with primary responsibilities Trust Lands, but such coordination does not occur falling to the latter. consistently, a marker of how hard it is to overcome entrenched divisions. Source: Huang 2016. Kampala, Uganda also labors under multiple systems Mailo tenure is prominent close to the central (map 5.3), including the dual-ownership Mailo system business district and thus particularly bad for (instituted by the British government in 1900), which productivity. Productivity in local services would rise 8 helps makes Uganda’s regime among the world’s percent by switching entirely to leasehold from Mailo, most complicated (Muinde 2013). Under Mailo, the removing the strict regulation of freehold areas that tenant owns the structures on each plot while the may well impede small businesses in setting up easily. Mailo landowner — typically the Buganda Kingdom Turning to where people choose to live, customary but possibly one of a few other private landholders land is less attractive than leasehold land for high- — owns the land. The tenant pays below-market skilled workers. Mailo land is associated with high rent, determined by the Land Board at the Ministry of residential densities of low-skilled workers. Paired Lands. Developing the land requires permission from with the fact that Mailo land has a negative impact on the landowner and the structure owner. local services, this finding suggests that this land is Recent research led by Tony Venables (as part of this particularly residential and undersupplied with small project) examines the broader welfare consequences shops relative to other areas in the city. In Kampala of the land tenure system in Kampala. Bernard, Bird, City, the high end of the skilled distribution seems to and Venables (2016) developed a spatial computable value settling on Mailo land, suggesting that people general equilibrium model for the city. They find make a tradeoff between being close to the central that land tenure and geographic constraints explain business district and having a tenure system without up to 38 percent of the variation in productivity and incentives to invest in better housing (under which 48 percent of the variation in amenity value across mainly low-skill, low-income people informally reside). Greater Kampala. The negative effect is particularly strong for customary land tenure, which if converted to leasehold would increase productivity by 3 percent in manufacturing, 4 percent in business services, and 11 percent in local services. A parish’s share of Mailo land also has a detrimental effect on service sector productivity, especially local services. 125 Africa’s Cities | Opening Doors to the World MAP 5.3 Land tenure systems in Greater Kampala Freehold Owner has full rights. Leasehold Leaseholder has rights for 49 years. Needs approval of the Land Board for use, transfer, and development. Customary Residents gain rights by long- term presence, but rights are hard to convert to free- or leasehold. Cannot be used as collateral or redeveloped. Mailo Dual ownership: Tenants own the structures and Customary landowner the land. Tenants pay below-market rents Freehold (determined by the Land Board). Development needs Leasehold approval from both parties, and both are compensated. Mailo Source: Bernard, Bird, and Venables 2016. Across Africa, when barriers to urban land access Unclear land rights severely constrain urban land arise from an overly complex property rights regime, redevelopment throughout the continent, imposing they impede consolidation of plots and transfer of high costs. In Nairobi, the cost of misallocating 1,000 land among users and among uses. Firms cannot acres of land within 4 kilometers of the city center — readily buy downtown land to convert it from low- the edge of Kibera — is huge (box 5.3). Unclear land density residential use to higher-density apartments rights also impede the collateral value of buildings or build clusters of new commercial structures. Land constructed on the land, constraining investment in transactions are long, costly, and complicated (World residential structures. Bank 2015d). Such market constraints reduce the collateral value of structures, giving developers little incentive to invest in residential height while tempting all parties to make informal arrangements. 126 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning BOX 5.3 Welfare costs of stymied redevelopment in Nairobi In 1912, the British government set up Kibera as Failure to convert this land from informal to formal a settlement for Nubians who fought in the King’s uses imposes high welfare costs. Kibera is located on African Rifles regiment of the British Army. Nubians some of Nairobi’s most desirable land. The welfare were the only people with settlement permits for the cost of misallocating 1,000 acres within 4 kilometers land until independence, in 1969, when the Kenyan of the city center — the edge of Kibera — amounts to government revoked their claims. Other migrants, who $1 billion, or about $200 per person in Greater Nairobi never had permits, settled informally on the land not (about 70 percent of Kenya’s per capita GDP). directly occupied by the Nubians. Today Kibera is one of the largest slums in Africa, with 600,000 inhabitants. Source: Henderson, Regan, and Venables 2016. Table 5.1 Percentage of land registered and number of days required to transfer property in selected countries and regions Percentage of land Number of days to Country registered transfer property Rwanda 70–100 25 Kenya 35 73 Uganda 18 52 Tanzania 5 68 Sub-Saharan Africa 10 65 Organisation for Economic Co-operation and Development 70 30 Source: World Bank 2014. Low land registration (table 5.1) may result partly The good news is that African countries are taking from cumbersome, expensive registration and steps to clarify land rights and strengthening land transfer systems loaded with survey expenses and administration (box 5.4). Botswana took the bold fees, which make registration unaffordable for many step of regularizing customary lands in 2008, (Toulmin 2005). The principal obstacles to improved partly because the land boards faced challenges land governance in Africa include land grabs, poor to administering tribal land. Zambia passed a new documentation, inefficient land administration, a lack planning bill in 2015, extending planning controls of transparency, and low capacity and demand for across state and customary land, and designating professional land surveyors (Byamugisha 2013). all local authorities as planning authorities. Namibia recognizes traditional leaders as part of the formal land system; they are designated by the country’s president, with their details published in the government gazette (World Bank 2015d). 127 Africa’s Cities | Opening Doors to the World BOX 5.4 Recent actions to improve land administration and common knowledge in Africa Improving the reach of land registries • Computerizing land records and registration systems helped cut the number of days to transfer • Since 2005, Rwanda has implemented property in Ghana (from 169 to 34) and Uganda comprehensive land tenure reform that has shown (from 227 to 48) (Byamugisha 2013). early success. Between 2005 and 2012, it pursued its nationwide program to issue land titles based on photomapping technology, at a cost of less than $10 Improving tenure security among informally per parcel. Madagascar, Namibia, and Tanzania have settled populations undertaken similar efforts (Byamugisha 2013). • In 2012, Namibia passed the Flexible Land Tenure • Tanzania surveyed all its communal lands and Act, which allows communities to obtain blocks registered 60 percent of them, at a cost of $500 of multiple plots and a “starter title” that grants per village. Ghana and Mozambique have begun to perpetual occupancy and transfer rights. This act follow Tanzania’s example (Byamugisha 2013). is aimed at the 30 percent of Namibian residents who live in informal settlements (CAHF 2013). • Ethiopia issued certificates for 20 million parcels of Residents can also apply for full, mortgageable land land at less than $1 per parcel and mapped them titles. Upon receipt of title, the communities are onto a cadastral index map at less than $5 per responsible for upgrading the site infrastructure. parcel in 2003–05 (CAHF 2013). The legislation is regarded as innovative in its recognition of incremental tenure and building Streamlining registration methodology (Byamugisha 2013). • In 2009, Kenya adopted a new land policy that • In 2011, Senegal passed a new Land Tenure Act, strives to streamline land administration processes under which people with temporary occupancy by reducing the stamp duty, from 25 percent of permits in urban areas can convert the permits into the principal amount to 5 percent; providing value permanent title deeds at no cost. Improved tenure added tax exemptions for developments with more security further helps increase housing investment than 20 low cost units; and reducing the tax on and improvement, access to housing finance, and mortgages, from 0.2 percent to 0.1 percent (Johnson the activity of the formal land market. and Matela 2011). • Kenya, Lesotho, and Tanzania are using bulk • The introduction of Lesotho’s Land Administration surveying and land use planning approaches to Authority in 2012 improved land registration by regularize tenure in slums (Byamugisha 2013). reducing wait times and improving application turnaround. It gained general support from land- holding communities (Byamugisha 2013). Source: World Bank 2015d. Some countries and cities are developing hybrid households. It could be formalized and extended to regimes to make formal and customary administration land subdivision and site-infrastructure-reservation more compatible. For example, in Nigerian states with authority. A similar pattern is evident in Ghana, largely Muslim populations, the emir’s representatives where local authority professionals survey the land subdivide and allocate land with the help of volunteer for the chiefs, who allocate it, in a manner also seen professionals from the government (an example is in Malawi and Zambia (World Bank 2015b). Future the city of Rigasa, in the extreme west of Kaduna urban redevelopers in Africa may learn from the [Igabi Local Government Area]). In Ethiopia, local successes of two approaches — land sharing and land officials play a dual role for the local authority and for readjustment — in Asian cities (box 5.5). 128 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning BOX 5.5 Land sharing and readjustment: Two ways to include residents in urban redevelopment plans Land sharing was used in Bangkok in the 1970s Land readjustment has been used in the Republic and 1980s, when rapid economic growth drove up of Korea, Japan, and Germany to assemble and plan urban land prices. Many slums were in desirable and privately owned land on the peri-urban fringe and accessible urban areas; the government brokered develop it with infrastructure and services. In the seven land-sharing deals with slum dwellers, Republic of Korea, 95 percent of urban land delivery accommodating commercial development without between 1962 and 1981 occurred through land displacing the residents. Existing development was readjustment. In Japan, 40 percent of the total annual to be densified, enabling verticalization of low-rise or supply of urban building plots from 1977 to 2000 was low-density residential uses and the opening up of secured in the same manner (Povey and Lloyd-Jones some of the land for new development (Rabe 2010). 2000). In land readjustment, the government pools privately owned parcels in an area and prepares a land The deals — struck in cases where land rights had use plan, designating spaces for public infrastructure long been disputed by landowners and 10,000 slum and services such as roads and open spaces. It then dwellers — allowed the building of high rises for implements the plan, providing trunk infrastructure, existing residents, releasing other portions of the and distributes lots to landowners, proportional to land for lucrative real estate development. In all the original parcels but smaller (for example, 50–60 seven cases, the slum dwellers paid for part of the percent). Because the new lot is serviced, it is worth construction through a loan program. Land sharing more than the landowner’s original parcel. The can work both for squatter households, who gain the government retains selected, strategic land parcels, right to remain on the site (though in new, multifamily, which it auctions or sells at market rates to recover medium- to high-rise housing), and landowners, who the cost of infrastructure and service delivery (Lozano- recover and benefit from part of their land (Rabe Gracia and others 2013). Although land readjustment 2010). is useful for urban regeneration where land ownership is divided among many private parties — and avoids the need for the government to buy land outright — it presupposes strong local institutions and a sound legislative framework. Source: Amirtahmasebi and others 2015. Land valuation and prices The pricing of land on the market depends partly Removing data and legal obstacles on policies, which must be designed with great care. Taxes, charges, and subsidies can be used to Land valuation is outdated or incomplete in many complement regulatory controls on land use, creating African countries. In Kenya, the valuation and rating financial incentives and disincentives. Revenues, such system has not been updated since colonial times, as those from land-based financing, can be used and property rolls are outdated: Mombasa’s was to finance administrative costs and infrastructure. last updated in 1992 and Nairobi’s in 1981 (World Implementation tools such as capital investment, Bank 2016a). In Ghana, property valuations have not budgets, and phasing plans can help with planning. changed in the past 15–20 years (World Bank 2015b). Some cities in Ethiopia do not even have such rolls (World Bank 2015a). In Malawi, only ratable areas are listed and valued for tax purposes. However, over time, some nonratable areas have become indistinguishable from ratable areas. 129 Africa’s Cities | Opening Doors to the World BOX 5.6 As a result, Lilongwe City Council’s property valuation Leveraging land roll is estimated to list about 45 percent of the values to finance properties in the city and Blantyre’s lists about a third (World Bank 2016b). urban infrastructure In Malawi, the Local Government Act prescribes that To be economically dense and well connected, the preparation and updating of valuation rolls can Africa’s cities will need huge investment in be done only by registered valuers (that is, valuers infrastructure. Although revenues from the registered under the Land Economy Surveyors, intergovernmental transfer system have been Valuers, Estate Agents and Auctioneers Act), but the mainstay of urban public finance, there is there are very few in the country, pushing up costs. a need to explore how cities can leverage the Property valuation methods are also inappropriate value of their assets — mainly land — to finance and cumbersome, relying on individual, rather than infrastructure and provide public goods and mass, valuations (World Bank 2016b). In Nigeria, the services. sales comparison approach is the preferred method of valuation, but most cities lack the information on Land-based infrastructure financing has the transactions needed (World Bank 2015c). In Kenya, greatest payoff where there is rapid urban the system relies on individual valuations that can growth. Rapid growth causes land prices to rise be subject to ratepayer objection before the roll is rapidly, creating an opportunity to generate finalized, leading to very outdated property rolls in the significant revenue. Yet rapid growth also more urban counties (World Bank 2016a). Some laws magnifies infrastructure investment needs, even prohibit (or severely limit) land fees and taxes requiring substantial sources of development (World Bank 2015d). finance. France, Japan, and the United States used land-based financing most heavily during Even if land revenue laws were sound, cities would periods of rapid urban growth, when there were still have little power to leverage land for revenue, leaps in the scale of urban investment. Taxes because fiscal cadaster records and capacities are on land can also improve the efficiency of land weak. Moreover, cities’ reliance on central government use, because property owners have an incentive transfers means that they have few incentives to make to develop the land to its most profitable use such efforts. Given the inadequacy of revenues from commensurate with the market value of the intergovernmental transfers, Africa’s cities should property. Valuable downtown locations with consider land and property taxes to finance urban higher land prices will experience densification infrastructure and public services (box 5.6). and investment in residential and commercial structures. Improving tax collection Several cities are making improvements in their Not only can such taxes incentivize dense urban tax collection systems. In Blantyre, Malawi, minor development, they are also nondistortionary, improvements, including direct payment to a because appreciation in land values is merely an commercial bank, have led to a sharp increase in economic rent for a scarce resource rather than collection rates, from less than 50 percent in 2011/12 a return on the economic activity of the owner. to almost 60 percent in 2013/14. The Mzuzu City There is thus no behavior by the owner to be Council in Malawi set a more complete reform that distorted. doubled tax collection in just one year, from MWK 120 Higher revenues from land and real estate can million to more than MWK 220 million. The reform of come from (a) improved valuation of land and its property tax administration system included a low- properties closer to their market value, which cost mass valuation method that used a Geographic deepens the tax base; (b) improved compliance, Information System (GIS) and follow-up in the field so that more property owners pay land and by local staff. It increased the number of assessed property taxes, which broadens the tax base; properties from 8,000 to almost 40,000 (World Bank and (c) monetization of underused public land. 2016b). However, setting in place land and real estate In 2004, Hargeisa, Somaliland began to create a land tax systems that support economic density is and property database and a method for classifying not straightforward. Strong institutions are and generating property tax invoices. Data were essential to define property rights clearly; stored in a GIS database for quick retrieval and ensure standardized and objective methods of mapping, allowing the local government to begin land valuation; and support and oversee the tax collections very quickly. The property survey, process of land management, land sales, and tax collection. 130 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning prepared over one year, cost only $48,500, or $0.82 administered own-source revenue system to a per property (excluding equipment such as personal modern local government revenue collection digital assistants, office computers, and software but information system integrated with a GIS platform. including satellite imagery). The new system enabled The new system allows the local government to use the local government to increase tax collections from satellite data to identify taxpayers and includes an $60,000 in 2008 to $282,725 in 2011. Between 2006, electronic invoicing system that notifies and tracks when the GIS became operational, and 2013, the payments. The city identified 102,904 buildings with share of taxed properties increased from 5 percent to this new method, a huge increase from the 23,000 in 45 percent; the number of properties on record also past databases. In the first 15 months after the switch climbed, from 15,850 to 59,000, in five districts (UN- to the new system, the number of eligible taxpayers Habitat 2013). more than trebled, from 31,160 to 104,629. Within one year, the city council boosted annual revenues by In 2014, the Arusha City Council became the first of 75 percent, from 2.6 billion shillings in 2012/13 to 4.6 seven Tanzanian cities to switch from a manually billion shillings in 2013/14 (World Bank 2015e). Uban planning institutions and land use regulation Land and property rights affect the transfer of land building the city depends on private rights over land between users; land prices determine the intensity and structures, such enforcement is fundamental of investments in structures. Both land transfers and to successful urbanization. For land registers and land prices are affected by land use regulation and mortgage collateral to perform their core functions of urban planning — the policies that determine how and supporting a market in parcels of land, and providing where land is used. finance for investment in structures, they need well- functioning on-the-ground enforcement. Urban planning and land use regulations are central for enhancing urban connectivity, productivity, and Beyond enforcing property rights, public authorities livability because of externalities and coordination are important for enforcing building regulations in two failures. Unregulated markets are unlikely to get urban areas: coordination and information. Costs are often densities and form right, because the productivity lower if design is standardized: Firms can coordinate of firms and the job-generating aspects of increased on a common publicly set standard. Some features of density are positive externalities accruing freely to a structure, such as its foundations, can be observed all, whereas the increased costs of construction, such only during construction. It is therefore useful for as buildings, roads, and network utilities necessary standards on foundations to be publicly enforced, for higher density to remain efficient, are not fully so that subsequent purchasers know what they internalized by firms and households. are buying. Standardization and information make valuation easier, which enhances the collateral value These market and coordination failures lead to less of structures. than optimal investment and ultimately to weaker productivity gains, less job creation, and lower wages. Across Africa today, urban plans and planning Further, well-functioning cities require that economies institutions appear ineffective. They are not of scope and complementarities be leveraged in the coordinating investment in structures or managing provision of physical infrastructure (roads, drainage, the spatial form of cities. One source of difficulty is street lighting, electricity, water, sewerage, and the inappropriate adoption of regulatory codes and waste disposal) alongside policing and health care. planning models inherited from colonial regimes Although each infrastructure sector and service can or imported from developed countries (Goodfellow be addressed by appropriate government policies, 2013). Another problem is that plans do not give addressing only one or two has little payoff if the credible accounts of finance, market dynamics, or others remain unresolved. With foresight and strong distributional impacts. Minimum lot sizes, for example, implementation, urban planning can help prevent may be intended as pro-poor land use regulations, these failures. but in practice they limit households’ investment choices (and in Brazil they have been associated with Strong urban planning institutions have a unique increased slum formation [Lall and others 2006]). Yet power of enforcement that is valuable in many another challenge arises from capacity and resource aspects of urbanization. Strong public authorities are constraints. essential to enforce private property rights. Because 131 Africa’s Cities | Opening Doors to the World Strengthening capacity and resources for Addressing coordination constraints across urban planning levels of government City and national authorities will have to make tough Urban planning has traditionally been largely the political decisions based on technical evidence function of the central government. Only gradually and assessments. They will need to increase since the mid- to late 1900s has decentralization urban planning capacity and resources, even given gained momentum. These shifts are usually competing funding priorities. In 2011, a survey of 12 formalized through some form of local government African countries found, on average, 0.89 planners for act that empowers local governments to undertake every 100,000 people (Africa Planning Association and their own urban planning activities. UN-Habitat 2014) — a far lower the ratio than in three However, such devolution of responsibilities is not high-income countries (table 5.2). It reported that entirely clear-cut or always accompanied by the recruiting skilled urban planners in a reasonably short necessary fiscal empowerment. At the local level, the time is a challenge in Africa, although institutional systems and institutions to support inclusive and processes may have been at fault, given that ministry- effective community participation may also be lacking. level staff have to be recruited centrally through the Ministry of Public Services. The lack of staff capacity Most African countries have vertical institutional constrains effective management. It is even more fragmentation and a complex relationship among crippling to enforcement, often the greatest challenge the multiple bodies, as exemplified by Tanzania even where all necessary structures and regulations and Uganda (figure 5.3). Both national and city-level are in place. agencies are often still involved in land and urban Table 5.2 Ratio of registered planners to population in selected countries, circa 2011 Population No. of accredited No. of planners Year of (million) 2011 planners per 100,000 estimate APA Countries Burkina Faso 16,970,000 14 0.08 2011 Ghana 24,970,000 150 0.60 2011 Nigeria 162,500,000 2,333 1.44 2011 Mali 15,840,000 50 0.32 2011 Kenya 41,610,000 194 0.47 2011 Uganda 34,510,000 90 0.26 2011 South Africa 50,800,000 1,690 3.33 2011 Malawi 15,300,000 30 0.20 2011 Mauritius 1,286,000 27 2.10 2011 Tanzania 46,200,000 158 0.34 2011 Zambia 13,400,000 60 0.45 2011 Zimbabwe 12,700,000 262 2.06 2011 Other countries United Kingdom 61,126,832 23,000 37.63 United States 304,059,724 38,830 12.77 2010 Australia 18,972,350 4,452 23.47 2009/10 Pakistan 173,593,383 755 0.43 2010 India 1,210,193,422 2,800 0.23 2011 * Countries that regulate the registration of planning at a national level. Source: Africa Planning Association Survey, Newman 2012. 132 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning management and in infrastructure and services authorities, and the Minister for Local Government provision, often with overlapping or unclear regulatory approves the urban authorities’ bylaws, budgets, scope and responsibilities. Such ambiguities muddy and proposals for generating own-source revenue. lines of accountability and complicate planning, (Most local government authorities are still largely implementation, monitoring, and enforcement. A dependent on central transfers.) Several central “silo” mentality across sectors and departments also government agencies or national parastatals are perpetuates institutional fragmentation, working crucial in areas under the jurisdiction of these against coordination. authorities, in service provision (roads, water, electricity, drainage), land use regulations (especially In Dar es Salaam, the central government retains land use planning and land allocation), and multiple controls over local authorities. For example, environmental management, for example (Kironde it appoints senior personnel to run the urban 2009). Urban regulations Urban regulations are key instruments to put urban processes, to lend them credibility. Although these plans into action, because they determine the documents may not be legally recognized, they are pattern of future land use. Yet many African cities’ considered socially legitimate and in some cases help urban regulations make it economically infeasible for reduce fraud and offer greater security to dwellers households and firms to acquire planned land, forcing in informal settlements. Informal landholding may them to seek alternative land sources and contributing also be more gender sensitive, privileging widows’ to extensive informal settlements. Typically, at least or orphans’ rights over extended family claims, as in two forms of urban regulation need to be satisfied: Malawi (UN-Habitat 2010). administrative procedures (steps that individuals have to take in order to apply for and acquire planned land) Relaxing planning standards and planning standards (such as minimum plot size) Excessively stringent planning standards help keep (Jones and others 2015). housing out of the formal sector. In particular, large minimum lots make formal land unaffordable, so Streamlining administrative procedures the poor often have no option but to illegally access Registering property in Africa is generally more time and subdivide land into very small parcels, creating consuming and costly than in other regions of the slums. A survey of regulations in five countries shows world. This contributes to the growth of informal that Kenya has the lowest minimum plot size, at 112 settlements. Formal property registration in Africa square meters (figure 5.5). Rwanda does not specify takes, on average 58.9 days and costs 9 percent of a minimum size nationally; minimums are likely to be the property value. Both figures are more than twice set by a local authority. the comparable figures in Europe and Central Asia Cities outside Africa cities have set lower minimum (26.5 days and 2.8 percent) and OECD high-income plot sizes, to allow access to formal tenure to the countries (24.1 days and 4.4 percent). Time and costs poorest population and promote formal, planned vary widely in Africa (figure 5.4, overleaf): Rwanda’s growth. When Philadelphia was settled, for instance, values, for example, are below even those in Europe the city authorities set a minimum plot size of about and Central Asia, while registration can take almost 30 square meters. 300 days in Togo and can cost more than 20 percent of the property value in the Republic of Congo. Continuous monitoring of on-the-ground conditions to determine the appropriateness of regulations drawn In response to such rigidities, alternative systems and up would better inform the parameters to use. mechanisms to recognize ownership and facilitate market transactions have developed informally. In the Kampala informal settlements of Kamwokya, Mbuya, and Busega, for example, most land transfers are accompanied by a “letter of agreement” that acknowledges the transaction between seller and buyer or an informal “certificate of title” to prove ownership (Nkurunziza 2007). In many places, recognized leaders in the community witness the 133 Africa’s Cities | Opening Doors to the World FIGURE 5.3 Key players in urban development in Tanzania and Uganda Tanzania Urban planning and Services provision Environmental development (water and sewage) management • Ministry of Lands, Housing • Ministry of Water • Vice President’s Office, and Human Settlements Environment Division • Energy, Water, Utilities Development Central Regulatory Authority • National Environment Management Council • National Water Board • Ministry of Natural Resources & Tourism • Ministry of Health Prime Minister’s Office Regional Administration and Local Government Dar es Salaam Regional • Basin Water Boards Dar es Salaam Regional Regional Commissioner and Sectretariat Commissioner and Sectretariat • Catchment Committees • Water User Associations or Groups • Dar City Council, • Dar es Salaam Water and • Dar City Council, Public Health Master Planning Sewerage Corporation • Temeke DLA, Public Health • Temeke DLA • Dar es Salaam Water and City • Ilala DLA, Public Health Sewerage Authority • Ilala DLA • Kinondoni DLA, Public Health • Dar City Council, Urban Planning, • Kinondoni DLA Utility Services and Environment Uganda Urban planning and Services provision Environmental development (water and sewage) management Central Ministry of Lands, Housing and National Wate & Sewerage • Ministry of Water & Urban Development Corporation Environment • National Environmental Management Authority • Wetlands Department • Kampala Capital City Authority City • Physical Planning Department, Public Health & Environment Department, Engineering Department, etc. Source: Huang 2016. 134 0 50 100 150 200 250 300 An g B ola Bu Bots enin rki wa FIGURE 3.4 na na Time (days) Ce nt ra C a Bu Faso lA bo run fri Cam Ver di ca n R ero de ep on Co ub ng C C lic o, om had D o Co em. ros Cô ngo Rep te , Re . d’I p. vo Er ire Eth itre i a Ga G opia mb ab ia, on T Gh he Gu a ine Gu na a-B ine iss a K au Time Le enya so Ma Li tho da ber Source: World Bank 2015d, based on data from Doing Business. ga ia Mascar law Ma i ur Mal i i Cost and international benchmarks in 2015 Mo Mau tani za rit a Part 3 | Springing Africa from Its Low Urban Development Trap mb ius Na iqu mi e Sã b oT N ia om Ni iger é a R ger nd wa ia Pr nd in a S ci Se ene pe Sie ych ga r e l Chapter 5: Clarifying property rights and strengthening urban planning So ra L lles e So uth one ut Afr h S ic ud a a Sw Sud n az an Ta an i l nz d an T ia Ug ogo an Za da Zim m La Eu Eas ba bia tin ro t A bw Mi Am pe sia dd & e le eric Ce & Pa Ea a & ntr ci s a fi Average time and cost to register property in Sub-Saharan countries OE t & N Cari l As c CD or bb ia t hig h A ean Su h i fri b- n c Sa Sou coma ha th ra e n A Asia fri ca 135 0 4 8 12 16 20 24 Cost (% of property value) Africa’s Cities | Opening Doors to the World FIGURE 5.5 Minimum lot sizes are large in Africa m2 350 300 250 200 150 100 50 0 Kenya Ethiopia Uganda Tanzania Philadelphia Source: Based on Huang 2016 for African cities; city visit for Philadelphia. Minimum plot sizes in Ethiopia vary across cities, ranging from 75 to 300 square meters. In other countries, they vary with type of construction. To build cities that work, make land markets work – nothing less will do African cities lack the institutions needed for of key factor and product markets. To build cities functional land markets, far-sighted planning, that work — cities that are livable, connected, and and effective regulation. As a result, their physical affordable and therefore economically dense — policy structures and infrastructure lag far behind the growth makers need to direct their attention toward the of the urban population. Not just housing (discussed deeper structural problems that misallocate land, in chapters 1–4) but other basic infrastructure and fragment development, and limit productivity. Above services are constantly struggling to catch up. all, city and national leaders must reform land markets and urban regulations, in order to enable investment Africa’s crowded, disconnected, and costly cities are and development, reward compliance, and ensure symptoms rather than causes of urban dysfunction. enforcement. Titles to property must be clear and They appear “closed for business” and “out of service” secure. Real property transactions must not be unduly because of underlying distortions in the functioning costly or burdened by bureaucracy. The key that will unlock the future for African cities is the establishment of high expectations and common knowledge through credible institutions to govern the transfer, valuation, and use of land. Such institutions are central for building economic density. Only policies that nurture them will enable Africa’s cities to support dense clusters of firms; respond nimbly to changing economic circumstances; and become more kind to their residents, whose future rests in policy makers’ hands. 136 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 5: Clarifying property rights and strengthening urban planning References Amirtahmasebi, Rana, Mariana Orloff, Sameh Wahba, and Andrew Lozano-Gracia, N., C. Young, S. V. Lall, and T. Vishwanath. Altman. 2015. Regenerating Urban Land. A Practitioner’s Guide 2013. “Leveraging Land to Enable Urban Transformation. to Leveraging Private Investment. World Bank, Washington, DC. Lessons from Global Experience.” Policy Research Available at: http://urban-regeneration.worldbank.org/node/31. Working Paper 6312, World Bank, Washington, DC. Africa Planning Association and UN-Habitat. 2014. The State of Muinde, Damaris Kathini. 2013. “Assessing the Effects of Land Planning in Africa. An Overview. Nairobi, Kenya: UN-Habitat. Tenure on Urban Developments in Kampala.” March. Available at: https://www.itc.nl/library/papers_2013/msc/upm/muinde.pdf. Antos, Sarah E., Nancy Lozano-Gracia, and Somik V. Lall. 2016. “The Morphology of African Cities.” Draft, World Bank, Washington, DC. Newman, K. 2012. “Benchmarking the Supply of Planning Graduates against International Practice: Its Implications for Arimah, B. C., and D. Adeagbo. 2000. “Compliance with the Re-curriculation Process at the Cape Peninsula University Urban Development and Planning Regulations in Ibadan, of Technology.” Paper presented at the State of Planning in Nigeria.” Habitat International 24 (3): 279–294. Africa Conference, 17–19 September, Durban, South Africa. Bernard, Louise, Julia Bird, and Anthony J. Venables. 2016. Nkurunziza, E. 2007. “Informal Mechanisms for Accessing “The Urban Land Market: A Computable Equilibrium Model and Securing Urban Land Rights: The Case of Kampala, Applied to Kampala.” Draft, July, University of Oxford. Uganda.” Environment and Urbanization 19 (2): 509–526. Brooks, Leah, and Byron Lutz. 2013. “Vestiges of Transit: Urban Picarreli, Nathalie 2015. Assessing Urban Land Tenure Persistence at a Micro Scale.” Trachtenberg School of Public Policy and Rights in Sub-Saharan Africa. November 20. Public Administration, George Washington University, Washington, DC. Povey, M., and Lloyd Jones, T. 2000. “Mixed Use Development: Buckley, R. M., and J. Kalarickal. 2006. “Land Market Issues: Mechanisms for Sustaining the Livelihoods and Social Capital The Mystery of Capital Revisited.” Thirty Years of World Bank of the Urban Poor in Core Areas.” Paper presented at the 17th Shelter Lending. Washington DC: World Bank: 27–47. Inter-Schools Conference on Sustainable Cities: Sustainable Byamugisha, F. 2013. Securing Africa’s Land for Shared Development. Oxford Brookes University, Oxford, United Kingdom. Prosperity. Washington, DC: World Bank. Rabe, Paul. 2010. Land Sharing in Phnom Penh and Bangkok: Lessons CAHF. 2013. Housing Finance in Africa. A review of some from Four Decades of Innovative Slum Redevelopment Projects in Two of Africa’s housing finance markets. Parkview, South Southeast Asian “Boom Towns. Washington, DC: World Bank. Africa: Center for Affordable Housing Finance. Toulmin, C. 2005. Securing Land and Property Rights in Sub-Saharan Felkner, John S., Somik V. Lall, and Hyun Lee. 2016. Africa: The Role of Local Institutions. Geneva: World Economic Forum. “Synchronizing Public and Private Investment in Cities: UN-Habitat. 2008. Mozambique Urban Sector Profile. Nairobi, Kenya: Evidence from Addis Ababa, Dar es Salaam, Kigali and United Nations Human Settlements Programme (UN-Habitat). Nairobi.” April 22, World Bank, Washington, DC. ———. 2013. Urban planning for City Leaders. Nairobi, Kenya: Goodfellow, Tom. 2013. “Planning and Development United Nations Human Settlements Programme (UN-Habitat). Regulation amid Rapid Urban Growth: Explaining Divergent Trajectories in Africa.” Geoforum 48: 83–93. ———. 2010. Malawi Urban Housing Sector Profile. Nairobi, Kenya: United Nations Human Settlements Programme (UN-Habitat). Gulyani, S., D. Talukdar and D. Jack. 2010. “Poverty, living conditions, and infrastructure access: A comparison of World Bank. 2014. Tanzania World Bank Country Economic slums in Dakar, Johannesburg, and Nairobi.” Policy Research Memorandum. Washington, DC: World Bank. Working Paper 5388, World Bank, Washington, DC. ———. 2015a. Ethiopia Urbanization Review: Urban Institutions Henderson, J. V., T. Regan, and A. J. Venables. 2016. “Building for a Middle-Income Ethiopia. Washington, DC: World Bank. the City: Sunk Capital, Sequencing, and Institutional ———. 2015b. Ghana Urbanization Review Overview: Rising Frictions.” CEPR Discussion Papers 11211, Center for through Cities in Ghana. Washington, DC: World Bank. Economic and Policy Research, Washington, DC. ———. 2015c. Nigeria Urbanization Review: From Oil to Cities: Huang, Chyi-Yun. 2016. “Enabling Structure: Role of Public Nigeria’s Next Transformation. Washington, DC: World Bank. Sector and Institutions.” Background paper for this report. ———. 2015d. Stocktaking of the Housing Sector in Sub-Saharan Johnson, S., and M. Matela. 2011. Reforming Land Administration Africa: Challenges and Opportunities. Washington, DC: World Bank. in Lesotho: Rebuilding the Institutions. Available at: http:// www.landadmin.co.uk/Documents/PN-32%20Reforming%20 ———. 2015e. The Tanzanian Strategic Cities Project. Land%20Administration%20in%20Lesotho.pdf. Improving Local Governments’ Own Source Revenues: The Arusha Experience. Washington, DC: World Bank. Jones, P., J., Bird, J. Lussuga Kironde, and A. Laski. 2015. Dar es Salaam: City Narrative. Available at: https:// ———. 2015f. Vietnam 2035 Report. Washington, DC: World Bank. collaboration.worldbank.org/docs/DOC-20828. ———. 2016a. Kenya Urbanization Review. Washington, DC: World Bank. Kironde, J. L. 2009. Improving Land Sector Governance in ———. 2016b. Malawi Urbanization Review: Strengthening Africa: The Case of Tanzania. In Workshop on Land Governance Rural-Urban Linkages. Washington, DC: World Bank. in support of the MDGs. World Bank, Washington DC. Lall, S. V., H. G. Wang, and D. da Mata. 2006. Do Urban Land Regulations Influence Slum Formation? Evidence from Brazilian Cities. World Bank, Washington, DC. Lavigne-Delville, P., and A. Durand-Lasserve. 2009. “Land Tenure and Development, Land Governance and Security of Tenure in Developing Countries, Summary.” AFD/French Ministry for Foreign and European Affairs, Paris. 137 Africa’s Cities | Opening Doors to the World 138 Chapter 6 Scaling up and coordinating investments in physical structures and infrastructure At the same time as they pursue the recommendations in chapter 5, African cities have an urgent need to improve two sets of urban structures — physical and infrastructural — given that they are chronically underserviced by them. Infrastructure coverage is actually declining on some metrics, across the region. In 2010, for example, when the urban population was 37 percent, only 34 percent of urban residents had access to piped water, down from 43 percent in 1990, when the urbanization rate was 30 percent. In 2006, the Africa Infrastructure Country Diagnostic (AICD) estimated that addressing the infrastructure backlog would require $68–$93 billion a year over the next three decades, a third of which would be for maintenance. 139 Africa’s Cities | Opening Doors to the World At the same time as they pursue the firms’ investments in accessible jobs. And any social recommendations in chapter 5, African cities have an return on public infrastructure depends on the urgent need to improve two sets of urban structures proximity of housing to premises. For example, a — physical and infrastructural — given that they are rapid transit system is more viable at higher densities. chronically underserviced by them (Banerjee and Policies need to leverage complementarities and Morella 2011). Infrastructure coverage is actually manage coordination failures that lead to single-sector declining on some metrics, across the region (Banerjee interventions that hinder economic density. and others 2009). In 2010, for example, when the Effective coordination will therefore be crucial to urban population was 37 percent, only 34 percent of African cities’ success in managing path dependence urban residents had access to piped water, down from and interdependence. A city’s ability to make early and 43 percent in 1990, when the urbanization rate was coordinated investments directly determines its later 30 percent (UNICEF/WHO 2012). In 2006, the Africa appeal to firms considering their own investments Infrastructure Country Diagnostic (AICD) estimated in the urban economy. Only efficient infrastructure that addressing the infrastructure backlog would — and service provision — will generate economic require $68–$93 billion a year over the next three density and improve livability, job market matching, decades, a third of which would be for maintenance and productivity. Inefficient structures can set back (Foster and Briceño-Garmendia 2010). urban development for decades. Pysical structures and infrastructure pose special Much of a structure’s value is determined by challenges. The first is path dependence: The complementarities with other structures in the costs of developing housing, infrastructure, neighborhood or city. The first structures built will and industrial premises depend on sequencing. dictate the options for further investments in the Making infrastructure investments first, followed vicinity: Path dependence implies that investors need by investments in housing and then in industrial to anticipate what other structures will be built nearby. premises, reduces the cost of all three, because These expectations are self-fulfilling — investments sewerage, drainage, electricity, clean water, and affect expectations, which in turn affect investments connectivity are cheaper to provide at scale than if (see chapter 4). The problems of path dependence and they are added to houses and factories individually interdependence are all the more pressing because of and at a later date (Collier 2016). Furthermore, this circularity. urban structures share a “putty-clay” quality: Once constructed, they are difficult to modify and can stay This chapter provides insights from recent research on in place for more than 150 years (Hallegatte 2009). African cities that examines how early infrastructure shapes urban structures; how road investment can A second challenge is interdependence among stimulate private investment in structures and achieve investments in physical structures and infrastructure. other citywide gains; how public goods can enhance For firms, the productivity of premises depends on the livability; and what African cities will need to do to proximity of infrastructure, workers, and customers. finance lumpy and huge infrastructure investments. For households, the utility of housing depends on Investing early in infrastructure to shape urban structures One reason for the early installation of infrastructure unable to provide new infrastructure. Now that people is that it is a coordinating device — an irreversible, have settled all urban spaces, local opposition to road and therefore credible, commitment that is highly construction perpetuates the severe lack of roads. visible and so generates common knowledge. Another A research project carried out for this study examines reason is that, if postponed until after population the longer-term benefits of “sites and services” settlement, it is far more expensive and difficult to projects that the World Bank used in the 1970s and install. It is more costly because all the services that 1980s to lay down infrastructure ahead of growth need to be located underground are easier to install at of urban settlements (Regan and others 2016). scale on clear sites rather than retrofitted piecemeal Many of these projects were undertaken with the beneath (or over or around) existing structures. It is idea of preventing slum formation or setting up harder because the inevitable disruption to private durable foundations for slum upgrading into formal homes generates indignant political protests. For neighborhoods. World Bank projects covered tens of example, Freetown in Sierra Leone grew rapidly during thousands of households in over 20 neighborhoods the civil war of 1991–2002, when the government was 140 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure FIGURE 6.1 Land values of de novo development projects are higher than values in other neighborhoods of Dar es Salaam, including rich ones 40 Median value per square meter (Tsh) 35 30 25 20 15 10 5 0 i ld) ni i i bi ma en Sin za me an ala un ya ch i (o as am im um jito n Mi ko he n ag o Ms an y Mz du g Ki c M an N ko Mi Mw Sites and Services Other areas Source: Regan and others 2016. in cities in Brazil, El Salvador, Jamaica, Peru, Senegal, In Dar es Salaam, sites with de novo development Tanzania, Thailand, and Zambia. Both types of projects have higher land values than land in other sites and services projects — building on empty, parts of city, including rich neighborhoods, partly unpopulated lands (known as de novo construction) because the sites and services areas have a higher and upgrading squatter settlements — included building footprint to plot area ratio (figure 6.1). infrastructure investment in roads, electricity, water, The research also shows that plots are bigger where and public buildings (schools, clinics, community investment was made ahead of settlement (as in Sinza, centers, etc.). The projects were discontinued during shown in the upper left of figure 6.2). These projects the late 1980s because their costs were high, despite have higher land value per square meter than projects anecdotal evidence that they had beneficial long-term in upgraded areas (such as Manzese, shown in the impacts. bottom right), where roads are disorganized, plots Researchers evaluated the long-run outcomes of both are small and irregular, and the cost-benefit ratio of types of projects on neighborhoods, including the valuing for tax collections would be prohibitive. The costs and benefits of each program. Benefits included sites and services plans drawn in the 1970s closely whether infrastructure investments increased the match the shape of today’s road network, showing value of certain areas and whether and how each type that investment in infrastructure is enduring, shapes of project shaped the urban landscape. In the long urban landscapes, and leads to higher land values, run, there is an expectation that the sites and services which are taxable and can finance future investments. programs would increase land values, which translates into a potential tax base. Slum areas have low land value and require recurrent investments in upgrading. 141 Africa’s Cities | Opening Doors to the World FIGURE 6.2 Differential impacts of de novo and upgrading projects in Dar es Salaam Source: Regan and others 2016. Note: The upper left of the photograph shows Sinza, a de novo project. The bottom right shows Manzese, an upgrading project. 142 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure Leveraging road investment Serious transportation problems are likely to transportation–land use relationship than cities in have major implications for the overall economic more developed countries, because of weaker land performance of a city. They reduce the connectivity use controls; the impacts on land use from road or of firms to workers, firms to other firms, and firms other transportation investments are therefore likely to consumers — and with it both the livability and to be more immediate. What is needed is the capacity the productivity of the city. Economic theory and to manage the land use shifts that are put into motion empirical evidence from around the world indicate by transportation infrastructure (Cervero 2013). that a reduction in transportation costs — which Ideally, such investments could be used to match could be brought about by road investments or other land development to feasible transportation capacity, improvements, such as bus rapid transit (discussed creating a more efficient balance between jobs and below) — can help increase connectivity between housing and giving workers access to a larger number business and residential areas, improving intracity of jobs (Srinivasan and Bhat 2005). mobility and reducing commuting costs (Fujita and Ogawa 1982; Lucas and Rossi-Hansberg 2002; Glaeser Bus rapid transit: One option among many and Kohlhase 2004; Srinivasan and Bhat 2005; Liu International evidence shows that bus rapid transit 2005; Owen and Phillips 1987). (BRT) systems can reduce commuting times. In Reductions in transportation costs and gains in Guangzhou (China) the introduction of the BRT in 2010 mobility foster land use changes and economic reduced travel times by 29 percent for bus riders and growth as the city moves to a new equilibrium of 20 percent for private car commuters. The change urban land use patterns (Grover Goswami and — readily apparent in figure 6.3 — has yielded total Lall 2016). Gakenheimer (1999) suggests that savings of 52 million hours a year, valued at RMB 158 cities in developing countries may have stronger million ($23 million) (Suzuki, Cervero, and Luchi 2010). FIGURE 6.3 Area around the Gangding station, in Guangzhou, China, before and after construction of the bus rapid transit system Source: Suzuki Cervero, and Luchi 2013. 143 Africa’s Cities | Opening Doors to the World In Bogota, the first BRT line reduced travel time by 15 implementation of the BRT, SO2 emissions declined minutes per passenger day (Hidalgo and Yepes 2004); 43 percent, NOx 18 percent, and particulate matter the first and second lines reduced travel time by or 12 percent (Turner and others 2012, cited in EMBARQ 12–14 minutes per passenger day (about 19 percent) 2013). Car crashes and injuries fell in two of the (Perdomo, Castañeda, and Mendieta 2010). BRT users system’s main corridors (Bocarejo and others 2012, in Istanbul can save 28 days’ worth of commuting cited in EMBARQ 2013). In Lagos, the BRT project a year by shifting to BRT (World Bank 2015). In reduced CO2 emissions by 13 percent and greenhouse Johannesburg, the BRT reduce travel times 13 minutes gas emissions by 20 percent (Peltier-Thiberge 2015). each way (Venter and Vaz 2011, cited by EMBARQ Having boomed since the early 2000s, especially in 2013). In Lagos commuting time fell by an average Latin America, BRTs are starting to grow in African of 25 minutes along a 22-kilometer corridor and wait cities. Four were recently implemented, in Lagos time was reduced from 45 minutes to 10 (Peltier- (2008), Johannesburg (2009), Cape Town (2011), and Thiberge, 2015). Dar es Salaam (2016). Together the four systems have Other benefits of BRT are reductions in pollution and 104 kilometers of exclusive bus lanes. South Africa’s improved road safety. Bogota’s BRT (TransMilenio) urban BRTs have underperformed (figure 6.4). BRT and new regulations on fuel quality reduce CO2 projects indeed have great potential — but they must emissions by an estimated 1 million tons a year. After be carefully planned and implemented (box 6.1). BOX 6.1 Bus rapid transit: Successful if handled with care Less often noticed have been the pitfalls and • Fares were defined by political authorities, shortcomings of these and other BRT projects. A World sometimes without a complete calculation of costs Bank review reveals key determinants of success along and revenues. with certain challenges: • The public was not adequately educated about • Bureaucracy was circumvented. In all 11 cities route changes. Communication failures occurred studied, planning and implementation teams were in Bogota; Santiago, Chile; Mexico City; and León, formed outside existing public structures, to avoid Mexico during expansion, leading to chaotic bureaucratic obstructions. conditions and, in some cases, public protests. • Political leadership was on board. Projects went • Existing transport operators protested when their forward rapidly in cities where the mayor or other interests were sidelined. For example, they were political leaders had a clear vision for BRT (Bogota, not involved in the BRT process through direct Curitiba, Jakarta, and Ecuador’s Guayaquil). Projects negotiations, or the bidding process failed to satisfy were stalled, sometimes for years, in cities where no them. such political commitment was present. • Fare collection systems were not integrated • Lead times were reduced to match political with public transit systems everywhere. Such timetables, but the quick schedule produced gaps disconnections occurred in Beijing, Bogota, and in planning. Steps toward project completion were Mexico City. Even BRT corridors were not always hastened so that elected officials could claim credit integrated with one another. Examples include before the end of their terms in office. As a result, Jakarta and Quito. crucial institutional, legal, and financial issues were • Bus scheduling sometimes led to overuse at peak sometimes neglected — though BRT planners gave hours and underuse at off-peak hours. Peak-hour thought to busway designs (median or curbside), crowding and off-peak inactivity are perhaps the platform types (high or low), fuel technologies (diesel most visible weaknesses in BRT operation. One or compressed natural gas), and fare collection or both can be seen in Beijing, Bogota, Curitiba, mechanisms (on board or prepaid). Guayaquil, Jakarta, León, Mexico City, and Quito. Source: Hidalgo, Custodio, and Graftieaux 2007; ITDP 2007. 144 144 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure FIGURE 6.4 Average daily bus rapid transit: Passenger counts are much higher in Lagos than in Johannesburg 12,000 10,000 8,000 6,000 4,000 2,000 0 rg a n pe ica nia ia s a bu ric ow ro Afr ea As go ric es me eT Eu La me n A p Oc A an ern Ca tin Joh rth La No Source: brtdata.org. Note: Values for Africa do not include Dar es Salaam. Different densities require different solutions. Mass transformation of rural land into urban land and transit needs high population densities, which make it development of large-scale infrastructure. In contrast, more likely that the system will be used with sufficient Japan has limited the amount of rural land conversion frequency and that high capital investments will be and linked its spatial hierarchy of subcenters by the justified. The appropriate transportation systems most developed subway network in the world. The for a city varies with population and job densities result has been a vibrant urban economy and social throughout the city. Improving logistics and designing integration. routes and bus stops might be enough to improve Integrating urban planning and regulation with connectivity in places with low densities; higher transport investments can help enhance ordered density can require a BRT, a light-rail train, or a and efficient transport development, as it has in subway. A study of the United States by Guerra and Curitiba, Brazil (figure 6.5b). Curitiba’s 1965 master Cervero (2011) finds that to be in the top quartile of plan first envisioned the city structure. Today the cost-effective investments, a BRT system with a cost city has created articulated densities along its BRT of about $50 million per mile would need about 18 corridors, with buildings strategically built along BRT jobs and residents per acre within a half mile of the corridors. As a result, the city has lower greenhouse stations, a light-rail train would need 50 jobs and gas emission levels, less traffic congestion, and more residents per acre, and a heavy-rail system would livable urban spaces than similar cities in Brazil. Public need about 60 per acre. transportation is also more widely used. For instance, although São Paulo has at least 10 times Curutiba’s Integrated urban planning, regulation, and population, in 2000 the number of annual public transportation investments transit trips per capita was higher in Curitiba (355 As African cities grow larger, policy makers need to versus 330 in São Paulo) (Suzuki, Cervero, and Luchi carefully plan the modes of transportation that will 2013). best enhance urban mobility. Chinese cities have seen massive sprawl and huge dependence on cars (box 6.2). They have grown through a mammouth 145 Africa’s Cities | Opening Doors to the World BOX 6.2 Alternate paths for improving urban mobility: Lessons from China and Japan China’s cautionary tale: Avoid getting locked global inefficiency, because it obliges sizing the whole system for peaks that could be better dissipated by into the “large-scale” mindset a variety of modal choices in a large “space of paths,” In the early stages of urbanization, the massive and fitting the distances to travel at a finer grain and accelerated conversion of rural land into urban creating a better-structured city organized around land brought in large amounts of capital, further a variety of scales. Peaks cannot be dissipated fuelling the process. The resulting economic growth into capillary networks, because the system lacks has led to large-scale industrialization and social capillarity. To avoid congestion, the large scales are transformation. China now suffers from severe urban overdimensioned and segregated from the smaller sprawl based on giant infrastructure and isolated scales, which prevents the emergence of intermediary buildings, a marked contrast to the much finer grain and small scales connected to the large scale. This of Chinese historical spatial forms. This new stage reinforces large scales against intermediary and small of development of Chinese cities has severed the scales and eventually locks cities into “large-scale traditional links between family generations and dependence.” between neighborhoods. Spatial zoning and large- scale separation of activities has greatly increased Japan’s successful organization, based on fine- mobility, imposed strict separations between economic classes, and increased the amount of time grain and local connectivity spent at work as opposed to leisure activities. The Japan increased its rate of urbanization from 25 diversity of Chinese cities, with their different climates percent in 1950 to 65 percent in 1980, while rising, in and cultures on a semi-continent, has been reduced 30 years of sustained growth from poverty after the to a uniform category of modern city, which consumes destructiveness of World War II, to become the world’s massive amounts of energy and will be locked into car third-largest economy, after the United States and dependency in the future. This urbanization model China. Japan developed the largest city in the world, is not only environmentally unsustainable, it also Tokyo, with 38 million inhabitants, in a highly efficient jeopardizes the future of China’s transition toward a way, through myriad micro-processes, creating a more mature society that is less dependent on low- highly complex, well-integrated, and well-connected cost labor and more dependent on innovation-based spatial hierarchy of subcenters, linked by the most economic growth. developed subway network in the world. Rather than relying on excessive rural to urban land conversion China has built its intra- and interurban networks with and then suffering from the detrimental urban sprawl a very large grain compared with other countries. The that resulted, Japan based its urban growth on internal result has been a decline in the number of possible intensification through well-balanced spatial planning links and paths between urban elements in China policies and a balance between micro-processes and compared with cities and city systems in Europe and larger-scale structuring interventions. the United States. This lack of medium- and small- scale street networks has a strong impact on the By preserving and reinforcing fine-grain and local management of traffic flow. The system is inefficient; connectivity, Japan has avoided the destruction among the resulting problems is congestion at the wrought by the collapse of the centuries-old social level of thoroughfares and subway lines. structure in China. Japanese society supports both social resilience and vibrant economic activity. By As a result, the deployment of a variety of providing an appropriate framework for investment transportation choices (walking, biking, buses with at the district, city, and national scales, Japan has short distances between stops, tramways, dense produced well-balanced cities with high degrees of subways, regional trains) is ruled out. People cannot connectivity at the district, city, and national scales. efficiently time their commuting schedules daily. Its numerous medium-scale projects and investments Everyone has only two choices: the large-scale mass have avoided the large-scale dependence that China transit system or the urban highway. The lack of suffers from and contributed to the emergence of short-range local choices and diversity creates strong livable, low-carbon, and successful cities. Source: World Bank 2016b. 146 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure FIGURE 6.5A Integrated urban planning and regulation promote density Bogota, Colombia TM trunk lines Urban perimeter FAR by block 0–1 1–2 Above 2 Source: Suzuki Cervero, and Luchi 2013. 147 Africa’s Cities | Opening Doors to the World FIGURE 6.5B Integrated urban planning and regulation promote density Curitiba, Brazil Source: IPPUC 2008. Note: The red, green, and orange lines indicate the city boundaries. In contrast, transportation in many cities has been (Felkner, Lall, and Lee 2016). The study used data motivated by the unique objective of improving from very high-resolution satellite images to measure mobility. In these cases, the opportunity of road investment using remote-sensing classification pro¬moting sustainable patterns of urban growth methods at a very high spatial resolution. It then used in the long run is lost. For instance, in Bogota, land econometric techniques that measure roads and use regulation has not been coordinated with urban structure over time — including differences- TransMilenio corridors. The city maintains a low floor in-differences with propensity score matching — to area ratio (0–2) throughout the city, except within estimate the quantitative associations between road the central business district and a few other selected investment, land use changes, economic productivity, spots (see figure 6.5a) (Suzuki, Cervero, and Luchi and population density. 2013). These cities lose the opportunity to achieve less The results quantify the full extent of the road network fragmented, livable cities. in each city for five categories of roads: three-lane paved roads, two-lane paved roads, two-lane paved Roads, densification, and land use change in roads with paved service lanes, one-lane paved four East African cities roads, and unpaved roads. Also quantified is the full As part of this research project, a study of roads extent of road investment and road changes over in four East African cities — Addis Ababa, Dar es 2003–13 (figure 6.6). The spatial resolution of the Salaam, Kigali, and Nairobi — examined the extent data enabled the authors to identify precisely where to which the timing and spatial incidence of public road investment was made in each city for each type investment drives population densification and private of road and to spatially quantify the level of that investment in housing and industrial structures investment. 148 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure FIGURE 6.6 Percentage of area devoted to paved and unpaved roads in four East African cities, 2001 and 2013 Percentage of area (km2) devoted to paved Percentage of area (km2) devoted to unpaved roads, 0.5km2 gridcells roads, 0.5km2 gridcells 2003 2013 2003 2013 Addis Ababa Kigali Dar es Salaam Nairobi Total area: 0% 0.01%–1.053% 2.022%–2.457% 4.223%–7.512% 1.053%–2.022% 2.457%–4.223% 7.512%–364.486% Note: Figures show percent of each square kilometer devoted to roads. Each grid cell is half a square kilometer. Source: Felkner, Lall, and Lee 2016. 149 Africa’s Cities | Opening Doors to the World African cities generally have low levels of road • Road investment and formal residential land investment (chapter 2): In a representative sample of use. Results are inconclusive. When estimated for all 30 global cities on the proportion of land area devoted cities pooled, paved road investment is associated to roads (UN Habitat 2013), Kigali ranked 19th, Addis with a significant decrease in formal residential land Ababa 24th, and Nairobi 27th. Yet these four cities use growth. Results are negative for Dar es Salaam saw heavy road investment over 2003–13, measured and Nairobi but positive for Kigali. The results are in terms of total road length and total area devoted positive for Addis Ababa for unpaved roads. to roads. The area for roads increased 54 percent in Nairobi, doubled in Dar es Salaam, trebled in Kigali, Citywide economic benefits of road and almost quintupled in Addis Ababa. Total road length almost doubled in Dar es Salaam (increasing by improvements in Kampala 98 percent, from 1,771 kilometers to 3,498 kilometers) The research findings reported above show and rose by 78 percent in Addis Ababa. It grew the spatially localized benefits of urban road more modestly in Nairobi (23 percent) and Kigali (19 improvements. There are also broader citywide percent). economic benefits of transportation improvements. In recent research on road improvements in Kampala, The study finds that paved road investment was Bernard, Bird, and Venables (2016) examine the positively and significantly associated with population potential benefits of improving the northern bypass density growth; growth in economic activity, as around the city, aimed at improving connectivity proxied by night light radiance; and industrial land use between the west and east of the city (map 6.1) and growth. upgrading the existing road network to facilitate • Road investment and population density. Paved movements within the city. road investment is associated with a 37 percent The direct beneficiaries of these improvements increase in population density for all cities pooled. are workers using motorized transport, whose It is associated with a 17 percent increase in Kigali transportation costs decrease. Other residents also and a 34.4 percent increase in Nairobi. Most of the benefit, because lower transportation prices reduce estimated associations are strongest within 1–2 the cost of living, at least in the short term, because kilometers of the road investments. The bulk of the over time, people relocate: The better-off locate paved road investment impact appears to come farther from their jobs as commuting times decrease, from one-lane paved roads, less from two-lane reducing pressure on land close to the city center and paved roads with paved service lanes. allowing poorer people to settle close to their jobs. • Road investment and economic activity as Low-skilled informal workers may actually benefit proxied by night light radiance. Paved road more than high-income workers, depending on the investment is associated with an increase in strength of agglomeration effects. If these effects are economic activity (as proxied by night light radiance) strong enough in the informal sector, as they were of 24 percent in Kigali and 13 percent in Nairobi. in Colombia (Duranton 2016), the relocation and As with population density, the magnitude of the increased clustering engendered by the transportation estimated coefficients tends to be highest within 1–2 improvement will boost productivity in the kilometers of the locations of the investment. These nontradable sector (in which most low-skilled workers results held across multiple robustness checks. are employed). The resulting increase in wages in that (Results for Addis Ababa are questionable, because sector might surpass the wage effect for high-skilled of likely errors in the city’s night light data related to workers. the linear interannual calibration process.) The long-term effect of the bypass on urban welfare • Road investment and industrial land use. Paved is eight times it short-term impact; the long-term road investment is associated with a 31 percent effect of upgrading the road network is three times growth in industrial land use, a result that holds as great. Transportation investments not only across robustness checks. Two- and three-lane decrease the aggregate commuting time of users of paved road investments have stronger associations motorized and nonmotorized transport, they also with industrial land use growth than one-lane paved make easier for firms and households to relocate and roads. Paved road investment corresponded to a have agglomeration effects on the city. If households 74 percent increase in industrial land use in Addis and firms can relocate, both projects lower the cost Ababa, an 83 percent increase in Dar Es Salaam, of living by 19 percent for high-income households and a 325.4 percent in Kigali. The results for Nairobi and 6 percent for low-skilled workers. With stronger were negative and significant. scale economies in the informal sector and increasing returns to scale, the difference between low- and 150 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure MAP 6.1 The Kampala’s Northern Bypass allows traffic to bypass Kampala’s city center CBD Northern bypass Primary road Secondary road Other roads and tracks Source: Bernard, Bird, and Venables 2016. Note: CBD = Central Business District. high-skilled workers decreases, with a reduction in the This research shows that the benefits of changes cost of living of 3 percent and 9 percent, respectively. in the urban landscape, such as investment in These investments’ long-term effects on urban welfare transportation infrastructure, take time to emerge. are much higher than their short-term impacts: those Firms and households respond to these changes from the construction of the bypass eight times, those gradually. In the long term, planning policies from upgrading the road network, three times. should support relocation in response to change of connectivity for the full benefits of investments to be realized. Providing public goods and services for livability Many African cities provide low access to public for a dense urban population than for a dispersed services and amenities. In expanding public services rural population (box 6.3; see also chapter 1). Being such as schooling, health, water, and sanitation, less costly, the optimal level of provision will likely be planners must consider that these services are subject higher, so cities can provide better education, health, to economies of scale and specialization. It is often and other services than rural areas. less costly to provide a given level of per capita service 151 Africa’s Cities | Opening Doors to the World BOX 6.3 Providing water to the poor in African cities The experience of some African cities shows that credits the continued renewal of his contract by access to safe water can be widely provided to the the elected city council to eThekwini’s success in poorest households. A forthcoming World Bank report serving poor communities. Uganda’s National Water studies how five African cities — Dakar (Senegal), and Sewerage Commission noted that serving poor Durban (South Africa), Kampala (Uganda), Nyeri communities creates a powerful ally, saying “In (Kenya), and Ouagadougou (Burkina Faso) — have Kampala, the poor vote.” Its leadership reports that provided reliable service to poor people in challenging the “Water for All” campaign has won support from environments (World Bank 2016). It concludes that both poor communities and the government. these cities have improved access for the poor 3. Independent structures are helpful but not primarily through effectively managed utilities, which sufficient for effective service provision. Formal recover all operating costs and some capital costs structures, such as independent boards and while scoring well on other measures of efficiency and regulators, are not sufficient to ensure effective cost-effectiveness. service provision, because they are not immune to Five lessons emerge from the report: predation or capture. However, these structures can be useful for bolstering professional corporate 1. Successful reforms need local leadership. All cultures and coordinating supportive relationships reforms were started and implemented by local with external stakeholders. Affermage and related leaders. Outsiders assisted only with knowledge contracts in Senegal provide clear rules that are and financial support. The start of successful costly to change. Provided the utility keeps doing a reforms displayed three mutually reinforcing good job, the contracts support success. The Office conditions: (a) a catalytic event or space for reform, National de l’Eau et de l’Assainisement (ONEA) (b) a skilled technical leader motivated to improve in Burkina Faso is publicly owned and operated. service, and (c) a relatively stable political leader Its performance contract with the government who supported and protected the reform. These is supervised by a multistakeholder committee conditions are inherent to a particular context and comprising representatives of customers, NGOs, cannot be manufactured, created, or effectively and the donors who finance the sector. The driven by outsiders. committee monitors performance of both the 2. Utilities that serve the poor well involve utility and the government under the contract the community. In Durban, eThekwini Water on the basis of independently audited financial established a consultative committee with low- and technical reports. Such designs embed income communities. The committee became accountability to external stakeholders in formal engaged and supportive when the utility addressed structures that can also help mobilize support requests to increase the free basic water allowance. against predation. The utility leader who drove eThekwini’s turnaround Because the gains from enhanced urban service Recent research shows that access to basic services provision accrue to urban households, they offer an increases steadily with population density (Gollin, extra incentive for migration. As migration eventually Kirchberger, and Lagakos 2016; for differences in leads rural and urban living standards to converge, the access to particular services.30 Rural and urban are thus gain in well-being from urban public services will tend not binary but a continuum: Moving to more densely to moderate urban wages. Urban workers face some populated areas can improve one’s living standards. costs that are higher than for rural workers — such as Many obstacles could prevent a rural family from housing and commuting — and so need higher wages moving to a city; fewer would prevent households in a to compensate for them (see chapter 3). Better urban small town from moving to a larger one. services will counter these effects, helping the city break into markets for internationally traded goods. 152 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure 4. Efficiency needs to improve — it’s not just a matter of getting more outside financing. 280 These turnarounds required hundreds of millions of dollars of investment in networks 260 and production facilities. Betweeen 2006 260 and 2015, the utilities serving the five case study cities increased the number of water 240 238 connections by an average of 93 percent (box figure 6.3.1). They financed these improvements 225 partly by improving their efficiency, so that 220 Number of connections, index (2006=100) they could increase their operating cash flows to raise and repay loans for infrastructure 200 investment. Some borrow commercially, but most rely mainly on development bank lending, taking advantage of credit enhancement offered 180 by their national governments. 5. Different approaches are necessary to 160 make access affordable for the poorest 144 households. These cities implemented 140 increasing block tariff structures and used cross-subsidies between commercial/industrial 125 and residential customers to guarantee 120 affordable access by the poorest households. Where informal land tenure does not allow 100 service provision, or the pattern of settlement or topography makes conventional network 06 07 08 09 10 11 12 13 14 15 20 20 20 20 20 20 20 20 20 20 designs infeasible, they improvised new technical and institutional arrangements, including working with small providers to Dakar eThekwini Kampala deliver services where the utility is restricted by Nyeri Ougadougou capacity or mandate (as in Ouagadougou). This option is proving handy for a wider range of utilities. Source: World Bank 2016a. Why has such migration from smaller to larger urban Depending on the country and the service, migrants areas not been observed? It may be that too little can be better off or worse off than other residents information is information for a detailed study of (figure 6.7). Differing urban management policies migration in Africa; most data are on rural-to-urban might explain these differences in access to services. migration. Other possible explanations include the fact that denser areas have a limited absorptive capacity; government policies are working to prevent slum proliferation; and migration is limited by unclear land titling, risk aversion, and poor information (Gollin, Kirchberger, and Lagakos 2016; De Brauw, Mueller, and Lee 2014). 153 Africa’s Cities | Opening Doors to the World FIGURE 6.7 Access to electricity by newly arrived migrants and other residents in the Democratic Republic of Congo, Ghana, Nigeria, and Sierra Leone Ghana Democratic Republic of Congo 1.0 1.0 0.8 0.8 0.8 0.8 0.6 0.6 Electricity Electricity Electricity Electricity 0.6 0.6 0.4 0.4 0.4 0.4 0.2 0.2 0.2 0.2 0.0 0.0 44 66 88 10 10 22 44 66 88 10 10 Log Log ofpopulation of population density density Log Log ofpopulation of population density density Rural-Urban Rural-Urban Migrant Migrant All All Rural-UrbanMigrant Rural-Urban Migrant All All Sierra Leone Nigeria 0.6 0.6 1.0 1.0 0.8 0.8 0.4 0.4 Electricity Electricity Electricity Electricity 0.6 0.6 0.2 0.2 0.4 0.4 0.0 0.0 0.2 0.2 22 44 66 88 10 10 44 66 88 10 10 Log Log ofpopulation of population density density Log Log ofpopulation of population density density Rural-Urban Rural-UrbanMigrant Migrant All All Rural-UrbanMigrant Rural-Urban Migrant All All Source: Gollin, Kirchberger, and Lagakos 2016. 154 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure Financing lumpy urban infrastructure investments Public authorities should seize the opportunity Land-based financing can thus be an attractive option of residential density to build highly productive for infrastructure investment. It takes different forms infrastructure and public services. That effort will at different stages of a city’s evolution (figure 6.9). require building revenue systems to finance provision. Such financing has clear advantages in principle. The capital costs of public infrastructure must be First, if the infrastructure is worth providing, the incurred far in advance of the productivity and appreciation in land values that it generates must livability benefits. The large capital outlays required exceed the cost of provision, most likely by a wide can be daunting. They are likely to far exceed the margin. Second, as the appreciation in land values is budget of any city government (figure 6.8). an economic rent rather than a payment to a factor of production — that is, it does not depend on the Because cities with good connectivity generate large effort of the land or its owner — taxing that rent does gains in productivity for their inhabitants, they offer not distort productive behavior. Appreciation does many potential tax points. Transactions within the not require the tax point to be a land transaction, in city could be subject to a sales tax, and households which there is considerable scope for falsifying the (or firms) could be made subject to a local income (or transaction to evade taxes. An alternative is an annual business) tax. But the least distorting form of taxation tax based on a market estimate of the average value may be appreciation in urban land values. As the city of land in the area. becomes more productive, in order to benefit from the enhanced productivity that it enables, firms and households must locate to it. As this happens, the value of land parcels appreciates, usually with the level of connectivity. This appreciation capitalizes the additional productivity provided by locating on that parcel. FIGURE 6.8 Gap between capital needs and budget resources of city governments $ Capital Tentative budget allocation Capital costs of expenditure to city government for replacement and urban infrastructure refurbishment Cash outflow Operating and maintenance expenses Construction Operation phase phase T Source: KPMG 2011. 155 Africa’s Cities | Opening Doors to the World FIGURE 6.9 Land-based financing instruments and city evolution Progressive evolution of cities Increased building Focus on building Undeveloped Developed property height and floor area performance, green property with basic services rations space, recreation Negotiated Developer Impact fees; ‘one off’ Property tax, tax Tax Increment ‘in kind’ development payments for surcharge, etc. Finance (TIF) contribution charges infrastructure Land sale, land lease, sale of Betterment development tax/levy rights City operating account City capital City capital Dedicated investment account account account Basic Improved Advanced infrastructure Infrastructure Specific service levels; infrastructure; serving focused on infrastructure higher capacity mass transit; individual improved for identified infrastructures CBD upgrades; property quality of life properties systems parks, etc. developments Source: African Centre for Cities 2015. Note: CBD = Central Business District. 156 Part 3 | Springing Africa from Its Low Urban Development Trap Chapter 6: Scaling up and coordinating investments in physical structures and infrastructure African cities are generally poorly positioned to finance the most significant volume through the land-lease their infrastructure needs through typical channels system. Kenya and Zimbabwe have legislation in place such as grants, transfers, own-source revenue, or but are underperforming or ineffective in ring-fencing borrowing. Land-based financing holds great potential the proceeds for infrastructure. to contribute to infrastructure financing — as long as Where land-based financing instruments are in place some conditions are met: in Africa, the modality is largely limited to in-kind • Demand for property. contributions, such as installation by developers of the secondary infrastructure connections and, at times, • An effective supply of developable land, land rights, bulk infrastructure required for particular projects. property finance, and a working real estate sector. This infrastructure may not be optimally located or • A sound regulatory and policy framework, with good not well coordinated enough to form an integrated legal backing. infrastructure network. These land-based financing instruments also tend to benefit middle- and high- • The financial and technical capacity to implement income groups, who are the main consumers of and monitor land-based financing instruments. such development projects. In many cases, the local Most African cities do not yet meet these government subsidizes them in an attempt to boost preconditions; they should work toward meeting local economic development. them, with the goal of leveraging land values to fund A better form of land-based financing may be the much needed infrastructure projects. A study of development charge, a one-off payment made by Ethiopia, Kenya, and Zimbabwe and a scan of 28 large a developer when land use changes are approved property development projects in Africa finds only very or new developments begin. It is matched by a little land-based financing (African Centre for Cities complementary policy and governance framework, as 2015). Of the three countries, Ethiopia has generated well as by implementation capacity. If cities in Sub-Saharan Africa remain crowded, disconnected, and costly, they can be neither kind to their residents nor productive. These cities are still being built. Before it is too late, they can ensure that they are not locked into inefficient and unsustainable patterns of urbanization. Given the high sunk costs and enduring nature of infrastructure, any approach to urban development that lacks early planning and coordination will burden future generations with cleaning up the mess — a terribly inefficient strategy. To inspire higher expectations, cities in Africa need better institutions. It is up to local and national authorities to undertake the institutional reforms needed for effective planning and coordination that will increase urban economic density and productivity and spur Africa’s belated structural transformation. Institutional structures must lead, not lag, urban infrastructure. If they do, the region’s cities will become not only better connected and more efficient but also kinder to their inhabitants, whose skills will be critical to economic growth and development. Only when this happens will the doors of African cities stand open to the world. 157 Africa’s Cities | Opening Doors to the World References African Centre for Cities. 2015. Urban Infrastructure in Hidalgo, D., and T. Yepes. 2004. “Are Bus Rapid Transit Sub-Saharan Africa: Harnessing Land Values, Housing and Systems Effective in Poverty Reduction? Experience of Transport. University of Cape Town, South Africa. Bogota’s Transmilenio and Lessons for Other Cities.” Paper presented at the Transportation Research 84th Annual Banerjee, S., A. Diallo, V. Foster, and Q. Wodon. 2009. “Trends in Board Meeting, January 9–13, 2005, Washington, DC. Household Coverage of Modern Infrastructure Services in Africa.” Policy Research Working Paper 4880, World Bank, Washington, DC. IPPUC (Instituto de Pesquisa e Planejamento Urbano de Curitiba). 2008. Curitiba: Integrated Urban Planning. Urban Planning Banerjee, S. G., and E. Morella, 2011. Africa’s Water and Regeneration: Key to Tackling Climate Change. Presented and Sanitation Infrastructure: Access, Affordability, and in Dundee, March 18th, 2008. IPPUC: Curitiba, Brazil. Alternatives. Washington, DC: World Bank. ITDP 2007. Bus Rapid Transit Planning Guide. New-York: Bernard, L., J. Bird, and A. J. Venables, 2016. “Transport in a Institute for Transportation and Development Policy. Congested City: A Computable Equilibrium Model Applied to Kampala City.” Draft, March 29, University of Oxford. KPMG. 2011. Financing the growth of your city. Singapore, Singapore: KPMG. Cervero, R. 2013. “Linking Urban Transport and Land Use in Developing Countries.” Journal of Transport and Land Use 6 (1): 7–24. Liu, F. 2005. “Interrupted Development. Failure in Urban Development and Housing Markets.” Paper presented to the Collier, Paul. 2016. “African Urbanization: An Analytic Policy Federal Reserve Board–George Washington University Seminar Guide.” Paper presented at the Fourth Tokyo International in Real Estate and Urban Economics, Washington, DC. Conference of African Development, “Land Use Planning and Spatial Development for Smart Growth in African Cities,” Tokyo. Lucas, R. E., and E. Rossi-Hansberg. 2002. “On the Internal Structure of Cities.” Econometrica 70 (4): 1445–76. de Brauw, A., V. Mueller, and H. L. Lee. 2014. “The Role of Rural–Urban Migration in the Structural Transformation Owen, A. D., and G. D. A. Phillips. 1987. “The Characteristics of of Sub-Saharan Africa.” World Development 63, 33–42. Railway Passenger Demand: An Econometric Investigation.” Journal of Transport Economics and Policy 231–53. Duranton, G. 2016. “Agglomeration Effects in Colombia.” Journal of Regional Science 56 (2): 210–38 Peltier-Thiberge, N. 2015. “Lagos’ Bus Rapid Transit System: Decongesting and Depolluting Mega Cities.” World Bank EMBARQ. 2013. Social, Environmental and Economic Transport for Development Blog. December 8. Available at Impacts of BRT Systems. Bus Rapid Transit Case Studies http://blogs.worldbank.org/transport/lagos-bus-rapid-transit- from around the World. Washington, DC. system-decongesting-and-depolluting-mega-cities-0. Felkner, J. S., S. V. Lall, and H. Lee. 2016. “Synchronizing Public and Perdomo, J. A., H. Castañeda, and J. C. Mendieta. 2010. “Evaluación Private Investment in Cities: Evidence from Addis Ababa, Dar es de impacto de las Fases I y II del sistema de transporte masivo Salaam, Kigali, and Nairobi.” April 22, World Bank, Washington, DC. TransMilenio sobre el tiempo total de desplazamiento de los Foster, V., and C. Briceño-Garmendia, eds. 2010. Africa’s usuarios del transporte público tradicional en Bogotá.” Documentos Infrastructure. A Time for Transformation. Agence Française de CEDE 2010-11, Universidad de los Andes, Bogotá, Colombia. Développement and the World Bank, Paris and Washington, DC. Regan, T., Nigmatulina, D., Baruah, N., Rausch, F., and Fujita, M., and H. Ogawa. 1982. “Multiple Equilibria and Michaels, G. 2015. “Sites and Services and Slum Upgrading Structural Transition of Non-Monocentric Urban Configurations.” in Tanzania.” Paper presented at the Spatial Development of Regional Science and Urban Economics 12 (2): 161–96. African Cities Workshop, World Bank, December 16–17. Gakenheimer, R. 1999. “Urban mobility in the Developing Srinivasan, S., and C. R. Bhat. 2005. “Modeling Household World.” Transportation Research Part A 33 (7–8): 671–89. Interactions in Daily In-Home and Out-of-Home Maintenance Activity Participation.” Transportation 32 (5): 523–44. Glaeser, E. 2011. Triumph of the City: How Our Greatest Invention Makes Us Richer, Smarter, Greener, Healthier, and Happier. London: MacMillan. Suzuki, H., R. Cervero, and K. Luchi. 2013. Transforming Cities with Transit. Transit and Land-Use Integration for Glaeser, E. L., and J. E. Kohlhase. 2004. “Cities, Regions and the Decline Sustainable Urban Development. Urban Development of Transport Costs.” Papers in Regional Science 83 (1): 197–228. Series. Washington, DC: World Bank. Gollin, Douglas, Martina Kirchberger, and David Lagakos. 2016. “Living UN-Habitat. 2013. “The Relevance of Street Patterns and Standards across Space: Evidence from Sub-Saharan Africa.” March Public Space in Urban Areas.” UN Habitat Working Paper. 31. Available at https://collaboration.worldbank.org/docs/DOC-20505. UNICEF (United Nations Children’s Fund), and WHO Grover Goswami, A. G., and S. V. Lall. 2016. “Jobs in the City: (World Health Organization). 2012. Progress on Drinking Explaining Urban Spatial Structure in Kampala.” Policy Research Water and Sanitation. 2012 Update. New York. Working Paper 7655, World Bank, Washington, DC. World Bank. 2015. “Take the Bus: It’s Faster than Your Car.” Guerra, E., and R. Cervero. 2011. “Cost of a Ride: The Effects of December 2, Washington, DC. http://www.worldbank.org/en/ Densities on Fixed-Guideway Transit Ridership and Capital Costs.” news/feature/2015/12/02/take-the-bus-its-faster-than-your-car. Journal of the American Planning Association 77 (3): 267–90. ———. 2016a. Providing Water to Poor People in African Cities Hallegatte, S. 2009. “Strategies to Adapt to an Uncertain Climate Effectively: Lessons from Utility Reforms. Washington, DC: World Bank. Change.” Global Environmental Change 19 (2): 240–47. ———. 2016b. Vietnam 2035 : Toward Prosperity, Creativity, Hidalgo, D., P. Custodio, and P. Graftieaux. 2007. “A Critical Equity, and Democracy. Washington, DC: World Bank. Look at Major Bus Improvements in Latin America and Asia: Case Studies of Hitches, Hic-ups, and Areas for Improvement; Synthesis of Lessons Learned.” World Bank, Washington, DC. 158 159 Africa’s Cities | Opening Doors to the World About the contributors Somik Vinay Lall J. Vernon Somik Vinay Lall is the World Bank’s Global Lead on Henderson Territorial Development Solutions and its Lead Economist for Urban Development in Africa. He heads a World Bank global research program on urbanization and spatial development and founded the Bank’s urbanization reviews program. He is an expert on development policy related to urban and territorial competitiveness, J. Vernon Henderson is the School agglomeration and clusters, infrastructure, and impact Professor of Economic Geography at the evaluation, with more than 18 years’ experience in Asia, London School of Economics. His research Africa, and Latin America. He was a core member of the focuses on urbanization in developing team that wrote the World Development Report 2009: countries. He is a co-principal investigator Reshaping Economic Geography; a senior economic on a major ongoing research project on counselor to the Indian prime minister’s National urbanization policy in Africa and globally. Transport Development Policy Committee; and the He has published dozens of articles in lead author of the World Bank’s flagship report on peer-reviewed journals. He is the co-editor urbanization Planning, Connecting, and Financing Cities of the Journal of Urban Economics and the Now. His work focuses on “place-shaping policies” Handbook of Regional and Urban Economics around cities, clusters, and corridors and the functioning and serves on several editorial boards. He of factor and product markets. He has published dozens is the founder and a past president of the of articles in peer-reviewed journals. Urban Economics Association. Anthony J. Anthony J. Venables is Professor of Economics at Oxford University where he directs a programme of research on urbanisation in developing Venables countries and the Oxford Centre for the Analysis of Resource Rich Economies. He is a Fellow of the Econometric Society and of the Regional Science Association, and is a Fellow and Council member of the British Academy. Former positions include chief economist at the UK Department for International Development, professor at the London School of Economics, research manager of the trade group in the World Bank, and advisor to the UK Treasury. He has published extensively in the areas of international trade and spatial economics, including work on trade and imperfect competition, economic integration, multinational firms, economic geography, and natural resources. Publications include The spatial economy; cities, regions and international trade, with M. Fujita and P. Krugman (MIT press, 1999), and Multinationals in the World Economy with G. Barba Navaretti (Princeton 2004). 160 Juliana Aguilar Olivia D’Aoust Juliana is an economist, Aguilar is consultant an economist, consultant in the in theRural Social, Urban, is Olivia an Urban D’Aoust is an Urban Economist in the Economist Social, Urban, Social, in the Rural and Social, Urban, Rural and Resilience Global and Practice Resilience at the World Global Bank, at Practice Resilience Urban, Rural and Practice Global Resilienceat Global the World Practice Bank,at the she where the where sheBank, World workswhere on the link she between works on the urban linkform between works on World urban Bank, wheredevelopment she works in urbanparticularly onAfrica, development the urban form and urban development and urban development and on urban studies and on urbanin in drivers ofparticularly Africa, and impediments the drivers to cities’ of and impediments productivity and Mozambique studies and the Democratic in Mozambique Republic of and the Democratic Congo. of Republic livability. to She is involved cities’ productivity andin livability. projectsShe in the Democratic is involved in She holds Congo. She master’s a holds a master’s degree in economics degree from the in economics from projects Republicin Congo, ofthe Ethiopia, Democratic Republic South Africa, of Congo, Ethiopia, South Sudan, the Universidad Universidad de los de Andes los Andes (in Colombia). (in Colombia). and Uganda. South Africa, South Sudan, and Uganda. She holds a PhD in economics from the Université libre She holds a PhD in economics from the Université libre Ana Aguilera de Bruxelles and a master’s degree in demography de Bruxelles and a master’s degree in demography from the Université catholique de Louvain. Ana Aguilera works as an Urban Development from the Université catholique de Louvain. Specialist at the World Bank Group. Her work focuses on improving city management with an emphasis on Chyi-yun Huang Chyi-Yun Huang urban economics and spatial development. She has Chyi-Yun Huang is an urban specialist at the World is an urban specialist at the World Bank with more studied the relationship between urban morphology, Bank with more than 10 years’ experience in integrated than 10 years’ experience in integrated urban planning productivity and access to basic services in cities. urban planning and design, city management, and and design, city management, and sustainable Her work also comprises survey management and sustainable development. She is currently leading or development. She is currently leading or contributing design to measure living standards and socioeconomic contributing technical expertise to various operations technical expertise to various operations and analytical indicators. In 2014 she was awarded with the Youth and analytical studies in Ethiopia, South Africa, studies in Ethiopia, South Africa, Tanzania, and Uganda. Innovation Fund for her work using Big Data to Tanzania, and Uganda. Before joining the Bank, as a Before joining the Bank, as a Young Professional, she understand mobility patterns in cities. Ana graduated Young Professional, she was an executive planner with was an executive planner with Singapore’s national as an Economist from Universidad Católica Andrés Singapore’s national planning agency. She holds an planning agency. She holds an MSc in urban planning Bello in Caracas, and holds a MSc. in Public Policy from MSc in urban planning from Columbia University and from Columbia University and a BA in architecture The University of Chicago. a BA in architecture from the National University of from the National University of Singapore and is a LEED Singapore and is a LEED accredited professional. accredited professional. Sarah Antos Patricia (Tracy) Patricia (Tracy) Jones Jones Sarah Antos is an urban geographer at the World Patricia (Tracy) Jones is a project manager and Bank who has worked extensively with spatial data in Patricia (Tracy) Jones is a project manager and researcher for the Urbanization in Developing the field of international development. In particular researcher for the Urbanization in Developing Economies Project. Her research focuses on economic her work as focused on remote sensing and the Economies Project. Her research focuses on economic development in Sub-Saharan Africa, particularly the classification of high resolution imagery. Before joining development in Sub-Saharan Africa, particularly the interplay between history, institutions, and long-run the World Bank, she worked at the World Health interplay between history, institutions, and long-run growth. She holds a DPhil in economics from the Organization and US’s Office of Foreign Disaster growth. She holds a DPhil in economics from the University of Oxford. Assistance, where she used satellite imagery and University of Oxford. survey data to improve disaster recovery efforts. She holds a bachelor’s and master’s degree in Geography Nancy Nancy Lozano-Gracia Lozano-Gracia from The George Washington University. Nancy Lozano-Gracia is a senior economist in the Nancy Lozano-Gracia is a senior economist in the Urban, Rural and Social Global Practice at the World Urban, Rural and Social Global Practice at the World Paolo Avner Bank. She was one of the main authors of Planning, Bank. She was one of the main authors of Planning, Connecting, and Financing Cities—Now! She provides Paolo is Avnereconomist an urban is an urban ineconomist the Social,in the Social, Urban, Rural and Connecting, and Financing Cities—Now! She provides diagnostic tools and a policy framework for policy Urban, Resilience Global Rural Resilience and Practice of Global the World Bank. Practice of the His current diagnostic tools and a policy framework for policy makers to manage the challenges brought about by work focuses World Bank. His on current the impacts workof land use focuses the impacts onregulations makers to manage the challenges brought about by urbanization and has led work using innovative data and of transport land systems on use regulations and the spatial development transport systems on of urbanization and has led work using innovative data collection methods such as satellite imagery, new developing the country cities. spatial development of Prior developing to joining the Bank, country collection methods such as satellite imagery, new survey designs, and Big Data approaches to improve cities. Prior to in Paolo worked France joining the asBank, a research engineer Paolo worked in in survey designs, and Big Data approaches to improve the understanding of urban challenges. She holds a France LEPII inas a research order to integrate engineer the urban in LEPII in order to dimension into the understanding of urban challenges. She holds a PhD in applied economics from University of Illinois in integrate the worldwide energy the urban dimension prospective into the worldwide model POLES. He PhD in applied economics from University of Illinois in Urbana-Champaign. then joined energy the Center prospective model for International POLES. He then joined the Research in Urbana-Champaign. Environment Center and Development for International Research in Environment (Paris) and collaborated to the and development Development of anand (Paris) collaborated applied land use to - transport the Shohei Shohei Nakamura Nakamura development interaction model of an (NEDUM land His applied 2D). usework specifically - transport Shohei Nakamura is an economist in the Poverty and Shohei Nakamura is an economist in the Poverty and focused on model interaction of public the ability (NEDUM 2D).policies His work and investments specifically Equity Global Practice of the World Bank, where he Equity Global Practice of the World Bank, where he focused to curb greenhouse on the ability ofemissions gas public policiesfromand urban transport investments specializes in urban poverty and spatial inequality. specializes in urban poverty and spatial inequality. to while curblimiting gasof greenhouse the costs emissions these policiesfrom forurbanurbantransport Before joining the Bank, he was a visiting fellow at Before joining the Bank, he was a visiting fellow at while residents. limiting Paolothe has costs graduated of these from policies for urban La Sorbonne Harvard’s Graduate School of Design. He holds a PhD Harvard’s Graduate School of Design. He holds a PhD University Paolo residents. and from has University graduatedParis fromX La Sorbonne - Nanterre as an in urban and regional planning from the University of in urban and regional planning from the University of economistand University andfrom is currently finishing University Paris his PhD at École X - Nanterre as an Michigan. Michigan. des Hautesand economist is currently Etudes en Sciences finishing Sociales, Paris. his PhD at École des Hautes Etudes en Sciences Sociales, Paris. 161 Africa’s Cities | Opening Doors to the World Notes 1 This study uses urban footprints rather than 20 See Nakamura and others (2016) for details. administrative areas to define urban areas, because 21 The World Bank Enterprise Survey data are based on a countries define urban in different ways. standardized questionnaire, which makes them comparable 3 The six cities are Mogadishu (Somalia); Kananga, across countries. The sample is stratified by firm size Tshikapa, Kinshasa, and Bakavu (Democratic and geographic location, which means that the data are Republic of Congo); and Dakar (Senegal). representative at the city level for many countries. The analysis 7 National population density is computed based on is restricted to manufacturing firms based in the largest city data from WorldPop (http://www.worldpop.org. in a country. As the data include only firms with five or more uk). It is used to derive measures of local population employees interviewed, they are unlikely to be representative of density matching the living standards indicators. the average African firm, but they may be representative of firms that are likely to expand into export markets. More than 10,000 10 The analysis is based on nominal wages, as in firms in 67 cities, including 16 cities in Africa, were analyzed. Bacolod, Blum, and Strange (2009). They claim that Urban nominal wages were converted using 2010 exchange rates. firms do not care about the cost of living. 22 The definition of tradables used is broad, including, for example, 11 The answers to all questions are ordinal and therefore all manufacturing, transportation, and financial intermediation, violate the assumptions about the use of standard methods and the data include only firms with at least five employees. when performing factor analysis, which are based on Pearson’s correlation matrix. The polychoric correlation 23 For example, a naive export value for an IPod is $150, but its matrix was therefore used (Kolenikov and Angeles 2009). domestic value added in China is $4 (Linden and others 2007). 13 Lower exposure means that at a certain distance 24 Two contrasting paths for the development of urban (usually 10 kilometers), people can potentially interact mobility appear in the experiences of China and Japan with fewer people than they do in other cities. (World Bank 2015f). More details are provided in box 6.2. 14 Higher fragmentation near the city center shows that density 25 International experience suggests that strong institutions are a values vary widely in this area. Henderson and Nigmatulina (2016) precondition for successful implementation of all instruments argue that African cities’ terrain is highly rugged, suggesting for tapping land value increases to finance infrastructure. that there could be less usable land near the city center. They include institutions that assign and protect property rights, institutions that enable independent valuation and 15 This report conceptualizes better connectivity as a driver of public dissemination of land values across uses, and a strong more specialized allocation of land in either commercial or legal framework supported by a healthy judicial system to residential use, causing households and firms to cluster (see handle disputes and oversee the process. But as land-based Fujita and Ogawa 1982; Lucas and Rossi-Hansberg 2002). Other instruments are considered for financing infrastructure, their views exist on the effect of a decline in commuting costs. Better risks must be recognized. These risks include the cyclicality connectivity can incentivize workers, especially the rich, to move of real estate markets; incentives for corruption over land to the suburbs (Alonso 1964; Muth 1969; Mills 1972). Besides transactions; and the “transformation” of local authorities pushing workers out from the city center, improved connectivity into real estate developers, favoring profits not welfare. pushes manufacturing firms farther out, to the urban periphery (where land and labor are cheaper) while keeping services 26 A subsequent version of the paper is in (which require less space and benefit more from localized Henderson, Reagan, and Venables (2016). agglomeration spillovers) downtown. See Baum-Snow (2007) for 27 These indicators are based on the Institutional Profiles Databases the United States and Baum-Snow and others (2016) for China. (which include a land tenure component), produced by the 16 See Gobillon, Selod, and Zenou (2007) for a Agence française de développement (AFD) and the French detailed review of theoretical explanations for and Finance Ministry based on the set of indicators developed empirical evidence of spatial mismatch. by Lavigne-Delville and others 2009. The data used rely on Picarreli (2015), who conducted additional desk research to 17 An average that weighs employment accessibility throughout eliminate inconsistent observations that would bias the results. urban areas by the number of people residing in each location is The retained indicators are deemed sufficiently consistent to arguably a more accurate depiction of accessibility in cities. Such provide an accurate picture of urban land tenure security. figures were not used here because they were not available for London. Table 2.1 provides a more complete picture of Nairobi. 28 Other countries, including Mauritius, Nigeria, South Africa, and Zimbabwe, have higher ratios. 18 Conversely, car accessibility is very high in the United States— very few metropolitan areas have less than 100 percent average 29 BRT is a bus system that dedicates road lanes exclusively for accessibility to employment within an hour (Levinson 2013)—and buses. The system uses bus stations, allowing pas-sengers to pay public transit accessibility is much lower. The city in the United before boarding, similar to metros and light rail systems. BRTS States with the highest average employment accessibility via have proven to be cheaper and faster to build than other transit public transit is Salt Lake City, with only 25.4 percent of jobs systems. Building a heavy rail system can cost 10 times as much reachable within 60 minutes (Owen and Levinson 2014). In many as building a BRT system. Though less expensive than heavy rail, cities the average is far less: Only 2.2 percent of jobs in Riverside, light rail can still cost four times as much as BRT. In addition, California and 2.8 percent in Atlanta, Georgia are accessible installing a BRT can take just two years; building an underground by public transit within 60 minutes. These numbers reflect the metro can take a decade (Suzuki, Cervero, and Luchi 2013). shaping of American cities by cars and roads, which decentralize 30 The exception is what Edward Glaeser calls “demons both people and jobs. In such contexts, efforts to fully connect of density,” which eliminate the positive effect at the workers and firms through transit are unlikely to succeed. high end of the distribution (Glaeser 2011). 19 The nine African cities are Abidjan, Dakar, Douala, Harare, Johannesburg, Lagos, Lusaka, Nairobi, and Pretoria. 162