The The Urban TransitionRevised in Urban 35564 Transition Sub-Saharan Africa in Implications for Economic Growth Sub-Saharan and Poverty Reduction Africa Christine Kessides 1818 H Street, NW Washington, DC 20433 USA Tel: (202) 473-9233 Fax: (202) 522-3224 info@citiesalliance.org www.citiesalliance.org 35564 Revised The Urban Transition in Sub-Saharan Africa Implications for Economic Growth and Poverty Reduction Christine Kessides © The Cities Alliance 2006 1818 H Street, NW Washington D.C., 20433, U.S.A. http://www.citiesalliance.org All rights reserved First printing, August 2006 The material in this publication is copyrighted. Requests for permission to reproduce whole or portions of it should be directed to the Communications Unit of the Cities Alliance Secretariat at the address shown above. The Cities Alliance encourages active dissemination of its work. Permission to disseminate will normally be given promptly and, when the reproduction is for non- commercial purposes, without asking for a fee. Cover Photo: The fast­paced chaos of life in Lagos typifies both the challenges and opportunities urbanisation in Africa presents. Edgar Cleyne / Still Pictures Design: Circle Graphics, Inc. Contents F O R E W O R D vii A C K N O W L E D G E M E N T S ix A B B R E V I A T I O N S A N D A C R O N Y M S x E X E C U T I V E S U M M A R Y xiii 1. Introduction 1 2. The Context of Urban Development in Sub-Saharan Africa 5 The Demographic Setting for the Urban Challenge 5 The Urban Economy: Growing, Yes--But from Informal, Small-Scale and Vulnerable Enterprises 9 Urban Poverty: Already Significant, and Likely to Grow 17 The Institutional Setting: Urban Governments Becoming More Mature, from a Low Base 21 3. What Cities and Towns Can Do for a Country's Development Agenda--and How 25 Urban and Rural Development as a Virtuous Circle 25 The Role of Urban Centres in Fostering Productivity, Entrepreneurship, and Economic Modernisation 28 How Urban Development Can Be Good for Poverty Reduction 29 Urban Development as an Exercise in Local Governance and Institution-Building 34 4. Making African Cities More Effective in Promoting Economic Growth and Poverty Reduction 37 Listening to the Firms 37 What Cities Require to Be Productive 40 iii iv C O N T E N T S 5. The Opportunity Cost of Neglecting the Cities--What Is at Stake for National Development in Africa 45 6. Strategic and Operational Choices and Priorities: Attempting to Gain the Best from the Urban Transition 51 Issues of Spatial Focus 51 Issues of Sectoral Focus 54 Issues of Institutional Development 54 Priorities for Further Urban Research in Africa 55 7. Conclusions 57 A P P E N D I X A Review of Investment Climate Survey Results for Sub-Saharan Africa and Selected Asian Countries 61 A D D I T I O N A L S T A T I S T I C A L T A B L E S SA1. Population Growth 1982­2003: Total Population, Urban Population, Urbanisation, and Population in Largest City (Primacy) 68 SA2. Economic Sector Summary (Agriculture, Industry, Services) in GDP, 1990­2003 70 SA3. Sectoral Sources of Growth Among High Growth African Countries (1996­2003) 72 SA4. Poverty Rates, Urban and Rural, and Urban Share of Total Poor 73 E N D N O T E S 75 R E F E R E N C E S 81 C O N T E N T S v F I G U R E S 1.1. Real Per Capita Income Has Stagnated In Sub-Saharan Africa 2 2.1. Most Population Growth in the Next Thirty Years Will Be in Urban Areas 6 2.2. Sub-Saharan Africa Is Undergoing Urban Transition with a Relatively High Urban Population Growth Rate 7 2.3. Africa's Urban Population Distribution by Settlement Size is Similar to that of Other Regions (2000 estimates) 8 2.4a. Developing and Transition Economies: Comparing Levels of Per Capita Income and Urbanisation, 1990­2003 9 2.4b. Sub-Saharan Africa: Comparing Levels of Per Capita Income and Urbanisation, 1990­2003 10 2.5a. Sectoral Output as Share of GDP, 1990­2003 (average %) 11 2.5b. Average Annual Growth of Output by Major Sector, 1990­2003 (%) 12 2.5c. Contribution of Industry and Services to Total GDP Growth, 1990­2003 (%) 13 2.6. Industry and Services Were the Main Sources of GDP Growth among the Better African Performers from 1996­2003 14 2.7a. Rural and Urban Poverty Rates, latest year 18 2.7b. Urban Shares of Total Poverty, latest year and projected 2020 19 2.8. Urban Poverty Rates Are Not a Function of Urban Population Growth 20 4.1. Access to Land as a Constraint in Urban Centres 39 4.2. Crime, Theft, and Disorder as a Constraint in Urban Centres 40 5.1. Urban Poor Children Face Higher Disease Risks than their Rural Peers 46 5.2. The Urban Poor Do Not Always Have Effective Access to Healthcare 47 A1. Access to Land as a Constraint in Urban Centres 62 A2. Corruption as a Constraint in Urban Centres 64 A3. Crime, Theft, and Disorder as a Constraint in Urban Centres 65 T A B L E S 2.1. Relative Importance of the Informal Economy in Employment 15 2.2. Contribution of the Informal Sector to the GDP in Selected Developing Countries 15 4.1. Concerted Efforts Raised Municipal Financial Performance in Benin's Three Largest Cities, 1999­2004 42 5.1. Health Status of Children in Kenyan Slums 47 A1. Burdens of Obtaining Land 63 A2. Firm Responses on Costs Due to Security Problems 63 B O X E S 3.1. Urbanisation, Markets, and Rural Development 26 3.2. Rural-Urban Linkages: A Virtuous Circle 27 3.3. Urban Migrants Have Done Well in Uganda 32 3.4. Sustainable Growth Draws on Phenomena Requiring the Urban Transition and Well-Working Cities 35 5.1. Crime Issues and Responses in Three African Countries 48 Foreword A frica is urbanizing fast. In this short treatise, Christine Kessides provides a succinct outline of the challenges that rapid urbanization constitutes specifically in sub-Saharan Africa and some of the ways in which these can be met. She examines the context in which this rapid urbanization is taking place: (i) how the emerging cities and towns can be made to contribute positively to the development of the continent, reducing the pervasive poverty among both rural and urban residents; (ii) what requirements need to be met to make this happen; (iii) what is at stake or what are the opportunity costs to African countries if these requirements are not met; and (iv) what are the strategic and operational choices and priorities that would enable most African countries to gain the best in the process of transition to a largely urbanized society. These are all weighty questions to ask and even more daunting to attempt to answer. For the truth is that the urban transition in Africa is at the centre of two other major revolutions which are being played out with varying degrees of suc- cess in virtually all countries of the continent. These revolutions are, on the one hand, the increasing sway of democratic culture and values with its emphasis on decentralization and devolution of powers and resources to local and municipal governments; and, on the other hand, the displacement of state controlled economies by a very expansive free and globalizing market economy. These revolutions are why Kessides' concern is about the implications of the urban transition in sub-Saharan Africa for its economic growth and poverty reduction. The report stresses that the urban transition in sub-Saharan Africa is proceeding at a historically unprecedented rate averaging over 5 per cent per annum over the past two decades. Consequently, the critical issues that each country must face in dealing with the expected high increases to its urban popu- lation in the next decade relate to the nature of the institutional and other arrangements being made to receive these additional populations, provide them decent housing and ample employment opportunities, strive to reduce poverty among them, and ensure that they enjoy a tolerable quality of life and standard of living. Governance reforms and fiscal sustainability, based on participatory democ- racy and social inclusion, are thus one of four development priorities which vii viii F O R E W O R D Kessides posits will be important to achieve these This concise publication pulls together evidence goals. The other three are economic modernization and analysis on key themes that together inform to foster increased productivity through mobi- significantly the World Bank's approach to Africa. It lizing knowledge, skills, technology and in- takes on three key dimensions of African urban- novation; private sector development to nurture ization, setting the evidence squarely in an interna- entrepreneurship, develop centers of demand tional context. First, while African cities are growing (larger markets), information-sharing, easier rapidly, the overall share of the population living access to credit, and support risk-taking through in urban areas is on a familiar trend compared to various other income alternatives and through the experience of other countries. Second, urban- networks with other entrepreneurs; and finally, ization and economic growth typically go hand- in-hand in Africa as elsewhere in the world. Third, an agricultural transformation to enhance mar- urbanization is not the cause of poverty but can ket demand for agricultural goods and for make a significant contribution to its alleviation. non-farm activities in rural areas, stimulate agri- The recommendations here support the World cultural intensification and promote its evolution Bank's strategy that acknowledges urbanization as to higher value production, and generally trans- a positive factor to promote development in form the rural areas in keeping with enhanced Africa. The report calls for recognizing what cities living conditions in urban centres. can offer and suggests that urban policies and This publication contains numerous innova- investments need to be far more central to national tive ideas for fostering the development of urban development efforts. Taking a balanced approach, centres in all countries of sub-Saharan Africa and the report links the modern themes of urban pro- ensuring that they make major contributions to ductivity and investment climate to the traditional the economic growth and development of their sources of growth in Africa, such as rural develop- nations. ment and agriculture, showing how together they can create a virtuous circle. Institutionally, the Ibadan, June 1, 2006 focus is on the proactive role that city governments Akin L. Mabogunje and municipalities must play for Africa to realize the potential of urbanization. Both approaches can serve Africa and its development partners well. Washington, D.C., July 27, 2006 Katherine Sierra Vice President Sustainable Development The World Bank Acknowledgements T his report is a revised version of an earlier draft prepared for the Cities Alliance 2004 Annual Report. Cities Alliance and the World Bank Africa Region's Water and Urban Sector Units (AFTU1 and AFTU2), commissioned and funded the work, with additional support from the Urban Development Anchor (TUDUR) in the Bank's Sustainable Development Vice presidency. The author would like to thank in particular the peer reviewers, Vandana Chandra, Mari- anne Fay, Humberto Lopez, and Roland White, as well as others who provided input and detailed comments at various stages; including Franc Bosquet, Billy Cobbett, Catherine Farvacque-Vitkovic, Matt Glasser, Mark Hildebrand, John Hine, Nina Schuler, and Deepali Tewari. The work was completed under the man- agement review of Inger Anderson, Jaime Biderman, Eustache Ouayoro, and John Page. Mita Chakraborty (Consultant) conducted the economic growth and firm survey analyses. Matthew Ramsdell (Consultant) provided very able research assistance. Oscar Apodaca, Francis Muraya, and Kate Owens supplemented the content and aided in its presentation. Chii Akporji managed the editing and production of the report to its current form, supported by Erika Puspa. ix Abbreviations and Acronyms AFR Africa Region AFTU Africa Region Water and Urban Unit AIDS Acquired immune deficiency syndrome APHR African Population and Health Center CASSAD Centre for African Settlement Studies and Development CCRA Council of Cities and Regions of Africa CGIAR Consultative Group on International Agricultural Research CWIQ Core Welfare Indicators Questionnaire DFID United Kingdom Department for International Development DHS Demographic and Health Survey EAP East Asia and Pacific Region ECA Europe and Central Asia Region EPZ Export protection zone ERS Economy Recovery Strategy FAO Food and Agriculture Organization FIAS Foreign Investment Advisory Service GDP Gross domestic product GFS Government Finance Statistics GMR Global Monitoring Report HIV Human immunodeficiency virus ICA Investment Climate Assessments ILO International Labor Organization IMF International Monetary Fund LAC Latin America and Caribbean Region MDG Millennium Development Goals MNA Middle East and North Africa Region x A B B R E V I AT I O N S A N D A C R O N Y M S xi NGO Nongovernmental organization OECD Organisation for Economic Co-operation and Development ODPM United Kingdom Office of the Deputy Prime Minister PREM Poverty Reduction and Economic Management PRSP Poverty Reduction Strategy Paper RIC Rural investment climate SACN South African Cities Network SAS South Asia Region SSA Sub-Saharan Africa SSATP Sub-Saharan Africa Transport Policy Program TUDUR Urban Development Unit of World Bank UNDP United Nations Development Programme WBG World Bank Group WDR World Development Report WHO World Health Organization se Pictur /StillET SCHYT GENROJ Family life in a slum, Cotonu, Benin. Executive Summary To promote more sus- S ub-Saharan African countries (hereafter, Africa) face the formidable task of tainable growth and establishing sustained economic growth through increased productivity, the necessary basis for reducing poverty. Although these countries' situations are very poverty reduction, diverse, with a few well established on an upward path and several others fostering much more attention the conditions necessary for growth and good governance, many more countries (and a majority of the African population) remain mired in place. This report asks must be spent how the ongoing processes of urban development--demographic and economic transformation and the management challenges for cities and towns that result-- unblocking the can, as they should, be made to benefit all of the Sub-Saharan African countries, and productivity- and what conditions are needed to achieve these advantages. Urban transition represents an opportunity as well as a management challenge for Africa. Urban areas remain welfare-enhancing underutilised resources that concentrate much of the countries' physical, financial, potential of Africa's and intellectual capital. Therefore, it is critical to understand how they can better serve the national growth and poverty reduction agendas. urban areas. The processes of urbanisation and the growth of cities and towns favour national development by diversifying incomes, expanding options for more affordable service delivery, and widening horizons for innovation and skill acqui- sition. While ample evidence indicates that cities and towns do make such contributions to Africa's development goals, it is equally clear that the positive impacts are not as great as they could or should be. To promote more sustain- able growth and poverty reduction, much more attention must be spent unblock- ing the productivity- and welfare-enhancing potential of Africa's urban areas. The Context of Urban Development in Sub-Saharan Africa The demographic setting: Not off the norm, but very demanding Somewhat paradoxically, African countries on average are undergoing urbanisa- tion (raising the urban share of their populations) at about the same pace as other xiii xiv E X E C U T I V E S U M M A R Y regions, including the now-industrialised coun- in terms of facilities, finances, and local gov- tries in their time. African urbanisation, however, ernment capacity to meet the demands placed is taking place in a context of severe constraints on them. Thus the potential of both large and that other country groups in other periods did small urban areas in Africa remains greatly not face, among them full exposure to pressures compromised. of global competition; very limited outlets for ex- Migration from rural areas is not the primary ternal migration; and depredation of the produc- explanation for the growth of cities as urbanisa- tive workforce and of family security due to tion expands: Natural increase and the reclassifi- HIV/AIDS, which also drains the weak capacity cation of rural areas are also important factors. of local administrations. The evidence on internal migration in African What is historically unprecedented is the ab- countries, albeit much of it anecdotal, confirms solute rate of urban growth in Africa--averaging that the traditional view of one-way movement almost 5 percent per year,1 implying close to a mainly from rural to urban areas is by no means doubling of the urban population in 15 years. the whole story; in fact, that pattern is much less The urban growth rate reflects the continuing pronounced in overall population mobility than high natural growth of population. Yet the "take are circular and seasonal migrations. Moreover, off " in Africa's urban population growth is still to rural and urban boundaries are artificial dis- come. On average, the population of Africa is tinctions to households, which often distribute now one-third urbanised, a proportion higher members across different spatial and economic than South Asia's 28 percent. Africa is approach- activities to diversify income sources and reduce ing a demographic inflection point, with the risk. numbers of new urban residents projected to rise In sum, although the demographic picture in sharply by over 300 million between 2000 and Africa is one of rapid and dramatic change, the 2030--more than twice the rural population situation is neither anomalous for its level of increment. development nor wildly out of line relative to The spatial pattern of Africa's urban pop- other regions. The real surge in Africa's urbanisa- ulation is not very unusual by international tion is yet to appear. In the next thirty years the standards. Only about 15 percent of the urban urban population will become the majority. population resides in cities of over 1 million in- Population mobility will contribute to the habitants (about the same as in other regions), growth of cities, especially in the least urbanised while 52 percent (versus about 43 percent for countries, but many households will retain foot- other developing countries) live in urban areas holds in both the rural and the urban economies. with fewer than 200,000 people. Whereas cities Both the large cities (one million-plus residents) below this size threshold in middle and high and the many rapidly growing smaller cities will income countries typically have a reasonable pose major challenges for local government asset base composed of the infrastructure, exter- administrations with characteristically weak nal transport links, educated human capital, and capacity. other basic services and amenities needed to sup- port factories and universities, many of the small African cities remain bereft of such features The urban economy: Where most while facing rapid growth in the numbers of new economic growth is happening, residents. In countries with good urban manage- but from a vulnerable base ment, cities of one million or more inhabitants tend to be the most productive for a number of Africa's economic "growth tragedy" of the past reasons, including especially their ability to decades has disappointed hopes and expectations match workers with jobs. Yet most African cities in all sectors. But "urbanisation without growth" in this size range are completely unequipped does not, in fact, describe what has been happen- E X E C U T I V E S U M M A R Y xv ing in most of the African countries examined Transport costs, whether local, But "urbanisation here,2 which show a normal linear relationship regional, national, or interna- between urbanisation and economic growth, tional, for the typical African without growth" similar to that found in other regions. Moreover, country are at least twice those of does not, in fact, the economic growth that has taken place in the typical Asian country, a factor Africa in the 1990s derives overwhelmingly from that impedes the growth of a bal- describe what has the industrial (including construction and min- anced city system and the devel- been happening in ing) and services sectors, which are mainly opment of external markets. urban-based. With very few exceptions, these sec- Compared to the rest of the devel- most of the African ondary and tertiary activities accounted for at oping world average, tropical countries examined least 60 percent, and averaged almost 80 percent, Africa also has less of its popula- of GDP growth in the region between 1990 and tion within 100 km of a coast here, which show a 2003. This is even more true of countries that (25 versus 66 percent), greatly normal linear rela- have been relatively good performers during the reducing accessibility, and almost growth spurt of the late 1990s, as almost all of twice Asia's share of population tionship between them derived most of their boost in value-added living at low density, implying high urbanisation and from the nonagricultural sectors. To the extent costs for service delivery. Africa has that informal activity (estimated to account for a relatively high share of popula- economic growth, tion living in disadvantageous eco- 93 percent of all new jobs created and 61 percent zones (tropical or arid). All of similar to that of urban employment in Africa) is undercounted these spatial limitations could be statistically, the urban share of the economy is found in other lessened, and ultimately will be, by even greater. well-managed urbanisation. regions. But the growth record of these sectors has sim- Perhaps the cruellest factor ply not been sufficient to propel most of the coun- draining Africa's development impetus is the tries into the realm of per capita income increase scourge of HIV/ AIDS. The epidemic represents needed to overcome poverty sustainably. Just as a large part of the reason why Africa's hard-won most agricultural output in the Region has repre- investments in human capital and other assets, sented extensive rather than intensive (high pro- including administrative capacity, become un- ductivity) production, the industrial and services derutilised and degraded. sectors persist in approaches more traditional In brief, probing the sources of and potential for than modern. While the dominance of public economic growth in Africa is important because it enterprise has faded, a competitive and entrepre- is widely recognised that sustaining high income neurial private sector has yet to materialise. growth is a necessary condition for a significant Several structural features and disadvantages reduction in poverty, although it is not a sufficient in many of the African countries limit economic condition in countries with high inequality. It is transformation and the maturation of firms, estimated that reaching the millennium develop- despite urbanisation. Compared to the low in- ment goals (MDGs) would require that the real come Asian countries, in African industry total GDP of African countries grow by at least 6 percent factor productivity is much lower and the share per year or slightly less if inequality improves. Only of indirect costs is much higher, hampering four countries--Botswana, Mauritius, Mozam- export competitiveness. The growth taking place bique, and Uganda--sustained growth rates close in industry and services derives mainly from to this target in the past decade, thanks to strong small scale, informal enterprises using low skills industry and services. Importantly, these sectors and low capital endowments, operating mainly in tend to show a higher positive income elasticity of response to domestic demand and therefore, lim- demand than do traditional agricultural products, ited by it. especially over the medium term. xvi E X E C U T I V E S U M M A R Y Clearly, all the productive potential of the cash income for all goods and services, macro- African countries must be mobilised much more economic shocks tend to hit urban populations forcefully to ensure a durable decline in poverty. particularly hard. Some forms of deprivation, This report argues that the sectors located mainly such as the evidence of food insufficiency (and most efficiently) in urban areas should be reported in the Benin poverty assessment, result the particular focuses of national growth strate- from problems of affordability. Physical proximity gies because of their relatively strong historical to social and infrastructure services does not performance in Africa, as in other regions. This guarantee actual access or affordability for slum would be particularly true in African countries dwellers and other poor urban residents. Non- with less favourable or declining natural condi- monetary costs of obtaining water or using sani- tions for agriculture. tary facilities can be very significant in urban areas, despite reported physical proximity to sup- Urban poverty: Already significant ply points, because of the sheer numbers of peo- ple depending on them. and likely to grow Large, statistically significant gaps exist be- In virtually all of Africa, as in most other develop- tween the access of the urban poor versus that of ing countries, rural poverty rates exceed urban the urban non-poor to infrastructure and social poverty rates. This is expected, given that urban services, even though extending services from the areas provide a wider and deeper labour market, better-off to the less well-off neighbourhoods creating a higher capacity to pay for services, and nearby would cost much less than would reach- that the density of settlement and proximity to ing the same number of people in remote and centres of government allow many services to be scattered settlements. Informal settlements (often provided at lower per capita cost. It is therefore lacking secure tenure and environmentally pre- striking that urban poverty rates are quite close carious) are home to a majority of the inhabitants (within 20 percent) of the rural rate in Kenya, in most African cities. For all of Africa, UN- Malawi, Mozambique, and in two Habitat estimates that over 70 percent of the Informal settle- of the most populous countries, urban population suffers shelter deprivation in terms of inadequate housing, water supply, or ments are home to a Ethiopia and Nigeria. The magni- sanitation. The widespread lack of sanitation stan- tude of urban poverty amounts to majority of the dards adequate to urban density helps to explain a third or more of the urban popu- why the MDG target of reducing infant mortality inhabitants in most lation in the first four of these is projected to be met in urban areas in only one- countries, as well as in Madagascar, African cities. For quarter (6 of 24) of the countries studied (Sahn Niger, Senegal, and Zambia. With and Stifel 2002), even fewer than in rural areas. projected rates of urbanisation, all of Africa, UN- Correcting these intra-urban inequities would be and assuming no change in rural "good value" for countries: The urban slum and Habitat estimates and urban poverty incidence, in peri-urban residents form the core of the urban twenty years half or more of the that over 70 percent workforce, and it is cost-effective to ensure their poor would reside in urban centres effective access to basic services. of the urban popula- in Benin, Kenya, Mauritania, Mo- zambique, Nigeria, and Senegal. tion suffers shelter Urban poverty is not simply a deprivation in terms matter of queuing for jobs and The institutional setting: services in rapidly expanding cities. Urban governments becoming of inadequate hous- The phenomenon of urban poverty more mature, from a weak start ing, water supply, or reflects various economic and institutional factors. Since the As in other developing regions, many of the sanitation. urban population is dependent on African countries over the past decade have E X E C U T I V E S U M M A R Y xvii established newly democratic processes at the What Cities and Towns local government level, with elected mayors or Can Do for the Country's councils. But access to revenues and the capacity to deliver services in response to expectations Development Agenda-- have been slower to materialise. Cities' financial And How means, and therefore also their performance in service delivery, depend in the first instance on Urban and rural development the intergovernmental fiscal framework, which as a "virtuous circle" determines their authority to tax and access various forms of central revenues (directly or Some observers see the urban trend in Africa as a through transfers). In Africa, local revenue symbol of what has gone wrong with these amounts to 5 percent of GDP in Uganda and economies; they believe that Africa has shifted 3.5 percent in South Africa (IMF GFS 2004), but away from a "natural" historical sequence of the more common level is below 1 percent. As development leading from increased agricultural would be expected, African local governments productivity to the phased emergence of towns rely less on taxation and more on transfers than and cities. Yet in most of the countries, agricul- do more decentralised economies; however, in tural productivity has been stagnant, and some several of the countries studied the local revenues 250 million people on the continent (30 to 70 per- represented no more than 5 percent of central cent of the inhabitants of most of the countries) government receipts (and less than US$10 per live in "fragile" ecosystems with limited agricul- tural potential (WDR 2003). The preconditions capita). The tax and borrowing authority for the for a technology-based revolution in the sector, cities, especially large cities and localities facing as experienced by Asia, are not apparent in most rapid population growth, is typically not adapted of the countries. It is evident, however, that access sufficiently to their greater expenditure and serv- to urban-based activities as part of a household's ice delivery obligations. State spending on invest- "livelihood portfolio" can raise the level of the ment does not necessarily make up for the lack rural economy by providing knowledge and of local capital funding. In Cameroon, less than resources that can be invested in inputs or capital 1 percent of state revenues are spent on capital stock for agriculture or for nonfarm activities expenditure for 18 cities, largely for the admin- (Ellis and Harris 2004). Options for intensified istrative capital (Yaounde) and the economic agriculture and diversification into nonfarm centre (Douala). production prove to be complementary; they can In such circumstances, it is little wonder that in also be fostered by common conditions, especially many African cities firms and households subsist by effective access to major urban markets. by their own grit and that public services are The most constructive way of looking at the almost nonexistent outside the wealthy neigh- productive interlinkages among urban and rural bourhoods. Local administrations are barely up to areas may be as a virtuous circle, whereby access the task of maintaining services, let alone meet- to (urban) markets and services for nonfarm ing the demands of growth or recuperating production stimulates agricultural productivity the damages to the community incurred from and rural incomes, which in turn generate HIV/AIDS. Despite their weak fiscal and admin- demand and labour supply for more such goods istrative means, African local authorities are and services. The circle provides multiple entry beginning to perceive their potential and to look points, and opportunities should be seized where to each other for encouragement and to other they appear. Individuals, households, and com- regions for good practice through associations munities benefit wherever market access is eased that give a voice to local governments in the and diverse economic activities become available, Region. either through physical proximity (short dis- xviii E X E C U T I V E S U M M A R Y tances to an urban centre) or through individual provoking negative outcomes such as environ- mobility. It is not surprising that poverty levels mental degradation and social distress. are lower and high-value agriculture more com- International experience suggests that creating mon closer to urban centres and that mobility for a modern competitive economy requires risk- work starts at relatively short distances (for ex- taking entrepreneurship, innovation, exchange of ample, from a village to a local city). Even though information and knowledge (especially of tech- cheap imports of certain foods, such as rice, may nology), and skills. These are all elements that pervade urban markets, urban demand for fresh are most readily fostered in an urban setting, and higher-value foods will continue to grow and although agglomeration alone provides no guar- should provide a strong impetus to efficient local antee of their presence. Whether nurturing production and marketing. domestic firms and promoting the graduation of small-scale operators or seeking international investors, African countries will need to make Urban centres as platforms for their cities more viable and attractive platforms productivity, entrepreneurship, and ensure that local governments can become and economic modernisation trusted and competent hosts. The developmental role of cities, as understood from research and historical observation, is How urban development can founded on the scale and network economies be good for poverty reduction than emerge from the agglomeration of firms and The urban trend generally bodes well for national workers, leading to increased efficiency in flows poverty reduction. It is natural that the poor of goods and services, better matching of workers would be attracted to the greater opportunities to jobs, flourishing of markets, and spillovers of cities offer. The key question is whether poverty knowledge and information. Urban areas there- in cities is part of a healthy process of economic fore function not only in line with traditional transition and mobility for a country and its growth theories by aggregating larger pools of households or a perverse trap revealing dysfunc- labour, inputs, and capital, but, even more im- tional institutions. Greater attention must be portantly, they epitomise the process of endoge- given to identifying the particular deprivations nous growth, whereby productive facing the urban poor and to tackling these resources are used more produc- The key question is directly by removing causes of economic and tively and in new ways. social exclusion, such as insecure tenure, and whether poverty in These benefits do not arise from by providing effective access to affordable basic the mere physical concentration of cities is part of a services. people and firms but from the abil- healthy process of ity of cities--and, in particular, their governments--to create envi- Urbanisation supports the economic transition ronments in which economic demographic transition and mobility for a agents can easily interact, labour is mobile, urban land becomes avail- Africa has yet to fully experience the demo- country and its able for productive uses, and both graphic transition associated with urbanisation. Africa's dependency rate3 (89 percent) is not pro- households, or a citizens and firms trust that they can safely invest in the future. Fail- jected to fall below 50 percent until 2050. As this perverse trap reveal- ing to ensure these ingredients of transition (characterised by falling birthrates and swelling labour forces) takes place, it will initially ing dysfunctional effective cities leads to very high opportunity costs, both by lower- appear in urban areas and create a "dividend" for institutions. ing returns to urban assets and by the economy. Urban areas provide an environ- E X E C U T I V E S U M M A R Y xix ment that tempers demographic dynamics, and it options for such income multiply Access to income is the cities that will need to serve the large youth in proximity to urban markets. and working-age cohorts over the next decades. Migrants within the urban popu- from nonfarm and While cities offer some natural advantages for lation do not simply swell the ranks urban-based activ- doing so, necessary conditions will include sound of the poor, as is often charged, and ities is associated management of the urban economy and ade- much evidence indicates that they quate urban public investment. do quite well. Demographic and with reduced rural health survey data reveal that, con- trary to common assumptions, poverty, and options Providing social and infrastructure recent migrants (whether from for such income services and meeting the MDGs-- rural or urban origins) were only more readily achieved in slightly more likely to be defined as multiply in prox- urban settings poor or to suffer less access to serv- imity to urban ices, and the differences were often On the supply side, input, output, and outcome not statistically significant (Mont- markets. measures of provision should be easier to achieve gomery et al. 2003). A poverty in cities than in less densely settled areas. The per study in Burkina Faso found no evidence that capita costs of many forms of infrastructure and immigrants to urban areas suffered more unem- social services are generally lower, as many more ployment than did other residents (Grimm and people can be reached. Relative to more remote Gunther 2004). The observation that migrants in locations, it is also easier to attract and retain per- urban areas are not significantly poorer or less sonnel, maintain operational quality, and provide well situated than incumbent residents reflects competition among alternative options. At the both self-selection among the individuals who same time, organising collective action among migrate--they tend to be better educated than heterogeneous populations is very challenging, their peers at home--and their ability to apply and densely built-up cities require more sophisti- cated construction and coordination of net- their human capital and initiative in the receiving worked infrastructure. economy. Household surveys in Benin reveal that compared to nonmigrants with similar char- Demand side factors. Demand side factors in acteristics (for example, education level), bread- urban areas can also facilitate reaching the coun- winners who migrate for employment achieve tries' MDGs. Individuals can more readily obtain higher consumption levels for their household-- information and witness the implications of cer- 14 percent higher, for those who migrate to urban tain behaviours--for example, by observing the areas and 7 percent higher in the case of migration opportunity costs of inadequate education and to another rural area (World Bank 2003c). the absence of family planning and the health Remittances represent such an important sup- impacts of poor sanitation. Urban settings do plement to some household incomes that they entail some specific health risks, including expo- clearly qualify as a tool (underutilised) in the sure to HIV/ AIDS, which is usually more preva- fight against poverty. While the magnitude of lent in large cities and informal settlements. But remittances, especially from domestic locations, here, too, such risks can be clearly perceived and is not easily compared with other financial flows, knowledge disseminated, supporting demands it is likely that they well exceed other special for action. resources available to the receiving households, Migration and remittances provide economic such as microcredit or development assistance. A options and mitigate economic risks. Access to recent poverty study of Ghana found that a major income from nonfarm and urban-based activities contributor to poverty reduction through the is associated with reduced rural poverty, and 1990s has been migration from slower-growing xx E X E C U T I V E S U M M A R Y to faster-growing regions. The biggest reduction ance of public services under their control, would in poverty was identified among the rural forest strengthen bonds with the governed in a very con- residents and was attributed to their receipt of crete way. The health of the city economies should remittances (McKay and Aryeetey 2004). not be of concern only to urban specialists or The people of Africa have fewer options for mayors. It is highly dependent on intergovern- migration abroad (outside the continent) than did mental, sectoral, and macroeconomic policies at poor populations in earlier periods. It is therefore the national level. especially critical to the welfare of rural residents (and especially those on fragile lands) that domes- tic migration remains unrestricted, so that indi- Making African Cities viduals can make their own location decisions. More Effective in Migration may add to income inequality in cases Promoting Economic Growth where the migrants and the households receiving remittances are among the better off in their com- and Poverty Reduction munities. The loss of the more productive mem- As against the potential, the urban reality in bers could also weaken the economy of origin. much of Africa is a physical concentration of Inequality should not be a dominant concern, people and activities that do not benefit from the however, in the face of overwhelming evidence of key "ingredients" we have come to expect from the benefits to individuals and households from more effective cities: (i) basic flexibility of the fac- income diversification and risk mitigation (Ellis tor markets--because of barriers to workers' job and Harris 2004). Migration (or mobility) is clearly and residential mobility and rigidities in land use; a favourable element in income growth and (ii) efficient local public services, both those that poverty reduction, for both urban and rural popu- can be produced through public-private partner- lations. Policies should enable labour mobility as an ships and those that are pure local public goods; element of general welfare and poverty reduction and (iii) trust and confidence in government. These strategies. At the same time, migration can pose a weaknesses deter the private sector and house- major challenge to the receiving areas by adding to holds from investing and partnering for the near-term demands for services, raising the stakes future and shift burdens to them that should be for good urban management. borne or reduced by the public sector. Urban local govern- ments can form the Urban development as Listening to the firms vanguard of good an exercise in local Enterprise surveys, such as investment climate assessments, indicate that firms in Africa often governance and governance and face hostile and unstable environments. But few institution-building sustainable public of the surveys have enquired into firms' percep- Urban local governments can form tions of the city in which they are located and of finance efforts, as the vanguard of good governance local government performance. For most of the their performance and sustainable public finance African respondents, institutional issues and efforts, as their performance is infrastructure pose equal degrees of concern. The is highly visible highly visible and they are the front firms' most pervasive complaints relate to inade- and they are the line of public administration for quate network infrastructure (power, telecoms, citizens. Institutional reforms, and transportation), corruption, and crime. front line of public including improving the collection Broader investigation reveals the extent of the administration for of local revenues and making burdens firms face, for example in land transac- municipal governments more di- tions. Anecdotal evidence from several countries citizens. rectly accountable for the perform- indicates that reported land acquisition delays are E X E C U T I V E S U M M A R Y xxi very long in Ethiopia and in Zambia. In Mozam- What cities require to be productive bique, firms pay on average $18,000 in processing fees for land, and in Nigeria, they must re-register Individuals and firms will continue to build for land to use it as collateral, a process that can take themselves and to provide their own services, up to two years and cost 15 percent of the prop- even in the absence of a functioning local govern- erty's value in official fees, before bribes (World ment. But clearly, much greater benefits could be Bank 2004d). An inquiry into investors' views in mobilised for the country and for urban inhabi- Senegal found that improvements in tax adminis- tants by providing complementary local public tration, business registration, and customs were goods in a safe and secure environment. This overwhelmed by the lengthening of the time implies focusing on basic investment and on effi- required to obtain land (World Bank 2003e). cient functioning of the essential core of land and The surveys reveal that some of the basic housing, environmental services, public trans- assumed advantages of urban agglomeration-- port, and local public finance. These, along with access to inputs and services affording economies the network utilities, are typically the missing of scale and connectivity to other producers and ingredients and weak links of the urban econ- low transactions costs--are not being realised in omies in Africa. Africa to the extent they should be, even in the largest or capital cities. Therefore, a firm located Land and housing. Land and housing (and in an African city gains a much lower produc- related construction activity) account for large tivity advantage than would be expected when shares of total investment, consumption services, comparing it to its non-urban counterpart, savings, and employment in most countries,4 and sometimes lower than that of firms in other they become no less important to the economy as regions facing otherwise similar conditions, such incomes grow. In Africa, the shelter sector is over- as market size or labour supply. At the same time, whelmingly informal, but with the right policies market access is so critical to firms that in Lagos and institutions it could become a powerful only 11 percent of large companies reported that engine for growth in jobs and for deepening the they would move out of the metro area, despite financial system. Inefficiencies in access to land serious problems with infrastructure and services and business premises have not often been (Tewari and Banerjee, 2005). Informal and small- studied in Africa, but where the issues have come scale firms are particularly vulnerable to the fail- to the fore, for example, in Addis Ababa, it is clear ures and high costs of local public services and that welfare costs can be high to both firms and are less able to protect or insure themselves households. An outcome is the burgeoning slum against local corruption and other failures of the and peri-urban settlements--insalubrious and city administration. illegal--where half to three-quarters of the Overall, the picture that emerges of urban population in most African cities find homes. enterprises in Africa shows them to be hamstrung African governments dominate the ownership by problems rooted in both national economic and use of urban land yet fail to protect rights of and city management. The advantages that enter- way or to prevent sensitive areas from being set- prises gain from an urban location can be swept tled. Misregulation leads to low density sprawl, away by poor economic and institutional frame- further raising the costs of services. Upgrading works at the national level. In Senegal, for ex- existing slums, while recognising the rights of ample, which has seen relatively good urban occupation (de facto tenure security), is far bet- management, cronyism and lack of competition ter than ignoring or relocating residents. But the continue to discourage private-sector entre- most cost-effective solution in rapidly growing preneurs (World Bank 2003e). To promote a cities (even compared to in-situ upgrading) is to dynamic and competitive private sector, both block out areas in advance for new settlement, national and city leaders need to be focused on identifying roads and drainage routes and secur- this common goal. ing rights of way, so that the needed infrastruc- xxii E X E C U T I V E S U M M A R Y ture can be provided cost-effectively as the zones commit to making government more responsive fill in. to citizens, municipal performance has improved The health and well-being especially of the remarkably. poor but also of the middle class, are put at risk In Benin, for example, the recently completed by from the dearth (and very poor quality) Decentralised City Management project enabled of urban environmental public goods and practices that more than doubled municipal rev- services--notably adequate water and sanitation, enues in the three largest cities. Similarly, in solid waste disposal, drainage, and green space. Senegal, municipal adjustment and investment International experience attests that an educated programmes achieved increases in municipal workforce and international investors are at- savings and rationalised the allocation of current tracted to "liveable" cities. For many urban revenues to much needed maintenance. All Africans, amenities are still defined as a hope of 67 municipalities under the Senegal programme getting regular garbage collection, freedom from have signed municipal contracts that commit street flooding, and an indoor toilet. both the national and the local governments to explicit standards of improved performance. Ease of moving people and goods. Ease of move- Introducing simple systems of street addressing ment of people and goods can be thought intrinsic in two cities--Thies and Kaolack--permitted to an urban setting. But failures in urban transport municipal tax billing to increase by about 50 per- policy seriously compromise the movement of indi- cent, with 90 percent collected (Farvacque- viduals as well as the circulation of goods, again Vitkovic et al. 2005). shuttering the urban marketplace in Africa. In many cities, even where a working bus system exists, the poor are simply priced out of using it. In Addis The Opportunity Cost of Ababa (a city of 3.6 million people), 70 percent of trips are by foot, and the average distance walked Neglecting the Cities--What per journey is 5 kms. In Nairobi, Is at Stake for National only 3 of 22 signalised intersections The major signal of were working at the time of one Development in Africa a well-functioning study, and in Addis, only 52 of 359 The failure to satisfy the basic conditions for traffic policemen had motorcycles city, and its chief effective cities will, in simplest terms, dilute these (SSATP 2002). Road traffic is barely benefits that could be gained for the country. In determinant, is the managed, and roads are highly unsafe for cars or pedestrians. some African countries, the issue presents itself quality of gover- more starkly: What is the difference between a patchwork of shantytowns, refugee camps, indus- nance and financial Governance and financial man- agement. The major signal of a trial zones, and gated residential communities-- management. It is well-functioning city, and its chief and a functioning city or town? determinant, is the quality of gov- Neglecting the cities also makes both firms important that as ernance and financial manage- and households more vulnerable to the dis- cities grow the pro- ment. It is important that as cities economies of urban agglomeration--high costs grow the professionalism and of land, congestion and inadequate mobility fessionalism and accountability of their local gov- within the city, a polluted environment, threats accountability of ernments mature as well. This to the social order and to public health, and obviously does not happen auto- crime. These risks, while never entirely avoidable their local govern- matically. But where systematic with population concentration, become greater ments mature as support has been given to improv- and are prematurely imposed by very inadequate ing practices and procedures and urban management. To put it bluntly, it is not well. to creating environments in which credible to argue that diseconomies are out- local and national authorities weighing the positive benefits of African cities E X E C U T I V E S U M M A R Y xxiii when these cities have virtually no working effective access or affordability of health care for public transport or safe waste disposal, much the urban poor. In Ghana, the 2003 Core Welfare of the land is held in public control with little Indicators Questionnaire (CWIQ) found a wors- available to meet market demand or the require- ening of health indicators, including underweight ments of public good purposes, and infant mor- status, among the urban poor relative to the rural tality is rising due to poor public sanitation in poor, even though the urban children had higher the neighbourhoods in which most of the work participation in health programmes--indicating force lives. that other risks, such as poor sanitation or food insecurity, may be at play. The burgeoning cohort of children and youth in African cities under- Higher costs and reduced scores the urgency of addressing urban public competitiveness of firms health issues. Crime and violence increasingly accompany Unreliable infrastructure and high transactions costs undermine firms domestically, but they deprivation in African cities. Next to Latin Amer- can be fatal to potential exporters. Africa, whose ica and the Caribbean, Africa is the most unequal share of world trade remains miniscule and region, and wide income disparities are associ- declining, cannot afford a comparative disad- ated with crime rates in cross-country studies vantage from the failure to nurture the business (Bourguignon 1999). The general problems of environment of cities and the waste of their pro- poverty and social exclusion within urban areas, ductive resources. The almost complete reliance the extreme weakness of national police and jus- of some large communities in major cities (such tice systems, and the absence of trust between as Karu in Abuja) on informal self-production communities and local governments compound and trade within a narrow radius is a sign of an the issue. Crime is now recognised as a significant urban economy insufficiently integrated with risk factor in African cities, especially those of itself, let alone with the regional or global South Africa and in Nairobi and Lagos, where economies. local and national governments and their police forces are working with the civil societies to find solutions. Worsening risks to public health and security One good overall reading of a population's wel- Strategic and Operational fare comes from its health status, as it stems both Choices and Priorities: from issues with specific services (for example, health care or the water supply) and from envi- Attempting to Gain the Best ronmental or social risk factors related to loca- from the Urban Transition tion. Although most studies report an urban health advantage in health outcomes (relative to There need be no debate about the first priority rural residents), poor urban children are also for healthy cities--a generally sound macro- found to be much less healthy than non-poor economic and fiscal framework and investment urban children and to face much greater disease climate at the national level. But the spatial focus and mortality risks. Particularly in Africa, urban of urban policy is becoming a source of con- poor children are found in some surveys to be tention, for example, in Ghana. Often, political less healthy than their rural counterparts (Mont- preference is given to reducing the relative domi- gomery et al. 2003). Children in the slums of nance of the largest (primate) city and to pro- Nairobi, for example, face enormously higher moting distribution of urban population and risks than do their peers living elsewhere in the economic activity across a number of geograph- same cities or in rural settlements. Apparent ically dispersed, smaller cities. In Africa, where proximity to health centres does not always imply the urban system as noted is not unusual, the xxiv E X E C U T I V E S U M M A R Y se Pictur /StillET SCHYT GENROJ Children at work and play in a slum in Accra, Ghana. argument is heard that smaller cities would link earlier, than a function of their sizes. As African better to the rural hinterland; against this view is countries urbanise and incomes grow, more cities stated the concern that investment should focus will emerge and expand naturally and gain in on very few cities to maximise economic importance. Decentralising political their role as trade centres. control and easing regulatory controls on the pri- A network of cities African countries need not vate sector will tend to promote this process nat- well-connected adopt a very deliberate or direc- urally, by reducing the incentive for firms to stay tive approach to reducing their physically close to centres of officialdom. At the within the country primacy rate and dispersing urban same time, global trade competition and fiscal or across its borders population among many smaller decentralisation tend to perpetuate spatial cities. In fact, it could be ineffective inequalities, and these differences can prove to can be a key ingre- and risky to do so. Large cities tend be very difficult to reduce through deliberate pol- dient in the growth to be the most productive and the icy. most attractive to innovative and A network of cities well-connected within the of interregional information-intensive economic country or across its borders can be a key ingredi- trading groups, activities. The observation that ent in the growth of interregional trading groups, large African cities pose major dis- an increasing priority of African governments. A an increasing economies even before all the factor that may be slowing the impacts of cur- priority of African potential gains of size have been rent efforts to create regional trading arrange- realised is more a testament to the ments in Africa is the relative absence of governments. neglect of urban policy, as noted city-regional corridors, reflecting the lack of E X E C U T I V E S U M M A R Y xxv large-scale industrial investors and of inter-city or other targeted investment zones Creating efficient transport linkages. may be justified in some cases, but Although national governments should not this should not preclude fuller and responsive local aim to dislodge the primate city from its dom- consideration of how to make the governments is key inant position in the economy, more basic city more hospitable to domestic support should be given to rapidly growing sec- investment and to all types of to creating effective ondary cities and their local governments to firms. cities. National improve their governance and service delivery Creating efficient and respon- capacity. Rather than attempting to "pick win- sive local governments is key to authorities and their ners" among emerging secondary cities or to creating effective cities. National external advisors create new cities as growth poles, national urban authorities and their external advi- policy should establish conditions and incentives sors (donors) need to put munici- (donors) need to put that help existing local governments to mobilise pal development at the centre of municipal develop- revenues and to respond to the evolving demands urban policy. Mature municipal in their jurisdictions for effective public services. management requires that local ment at the centre This is the approach of incentive-based munic- investment be on-budget and part of urban policy. ipal assistance projects, which fund secondary of an expenditure plan, rather than city investments on the basis of fiscal and other undertaken through ad-hoc assis- performance standards. tance arrangements. Accountability to the local Ultimately, the debate about urban policy and population and a voice for citizens should be the system of cities in Africa hinges on concerns encouraged as elements of mainstream perform- about unbalanced growth among regions. Rather ance by elected local governments and instilled through formal, transparent budgeting and mon- than attempting to manipulate urban develop- itoring processes. Sustainable urban development ment, the best approach to mitigating spatial is inseparable from local government capacity- inequality is the integration of rural areas within building and cannot be achieved simply through large urban markets, the mobilisation of labour investment in urban infrastructure--or by rely- and remittances, and strong fiscal mobilisation in ing only on practices introduced for less complex cities supported by intergovernmental transfer institutional settings (such as social investment policies. funds and community-driven development). Interest is growing in Africa in providing spe- Urban research and better urban data are cial industrial or export promotion zones (EPZs) needed to inform all of the issues discussed here. as a way of targeting favourable conditions (both in A priority is supporting the collection and use infrastructure and institutions) for new invest- of city-level data for local policy making and ment, especially in export sectors. The experience for benchmarking--as illustrated by the South of many export processing zones should give African Cities Network--so that variations in city pause, however, because the successful cases have performance can be better understood. A partic- been relatively few, and many have been costly ular interest would be to further document the failures, including some in Africa. To establish differential needs of capital or primate cities ver- linkages in the local economy, footloose investors sus the more numerous and faster growing sec- require grounding in strong local resources (such ondary cities. Existing data sources should also be as a capable and reliable workforce), institutional explored, notably through wider analysis of the relationships, and amenities. Rather than creating urban dimensions of household surveys. Not a few favourable investment enclaves, a more sus- least--more study is needed to gauge the finan- tainable and scaled-up approach would be to cial resources available to local governments and make the cities work better as support centres for to advocate for more adequate revenues so that business. A strategy favouring export promotion these authorities, especially those responsible for xxvi E X E C U T I V E S U M M A R Y Good urban man- large and growing urban areas, can Myths about the urban economies. Africa can- meet the challenges they face. not simply be characterised as "urbanisation agement feeds into without growth," and the term does not even fit the entire national many of the countries. The economic growth that Conclusions has taken place in the past decade derives mainly growth and welfare The development community, from urban-based sectors (industry and serv- agenda, stimulating including African policy makers ices), and this is especially true of the better- and donors, needs to move beyond performing economies. But cities have clearly not agricultural intens- debates that either criticise the lived up to their productive potential because of ification and divers- process of urban growth in the widespread neglect and bad management. Region or apologise for it. Instead, ification of rural they should focus attention and Myths about cities and poverty. Urban poverty income, poverty resources on the need to make the is not mainly a function of urban expansion, nor urban areas work more effectively is it a sign of the failure of urban economies in reduction, good for the benefit of the entire country. Africa. Evidence shows that much of the depriva- governance, and Africans need their cities if their tion in cities and the emerging urban public economies are to transform, but health problems relates to institutional failures fiscal resource they also need to transform their that perpetuate social exclusion and inequalities mobilisation. cities. between the urban poor and the urban non- poor. F I R S T -- the development com- S E C O N D -- the development community must munity must recognise and move beyond the recognise what cities can contribute to the na- myths that cloud much of the discourse on tional development agenda--and what this African urban development: requires in turn. Much of the development dia- logue over the past thirty years has been, and Myths about the demographics. remains, obsessed with the view that attention to African countries are not urbanising faster cities represents "urban bias." Yet cities suffer the than did other countries, and the distribu- effects of genuinely bad urban policy and finan- tion of urban population among large and cial neglect, misguided incentives that distort the very small cities is not unusual given these use of land and other investments, and hostile countries' level of development. That said, treatment of much of the population on which the absolute rate of urban growth creates a the city depends. Urban advocates and urban major challenge for urban management, critics need to take a hard look both at what the particularly in the secondary cities, which urban phenomena can offer national develop- tend to be the most under-serviced. ment across numerous channels and at what sup- Internal migration is not the only, nor even port cities and local governments require in turn main, source of urban growth in most of to achieve these results. countries--nor is it responsible for urban Good urban management feeds into the entire poverty. Although data are limited, evi- national growth and welfare agenda,5 stimulat- dence indicates that migration has been ing agricultural intensification and diversification favourable, on balance, for both sending of rural income, poverty reduction, good gover- and receiving areas in Africa. But popula- nance, and fiscal resource mobilisation. African tion mobility is much more fluid than is cities have the potential to be strong platforms recognised in the rural-to-urban model, and laboratories for most of the economic and and households wisely diversify their social behaviours needed for transformative activities across both areas. growth and productivity--including creative in- E X E C U T I V E S U M M A R Y xxvii novation, technological application, entrepre- neurship, open attitudes toward change, and risk-taking. Firms and individuals find more opportunity in cities and towns despite the many problems there, and the forces of agglomeration and migration appear as strong in Africa as they have proven to be elsewhere. Releasing the poten- tial of Africa's cities by addressing basic weak- nesses in land markets, public transport, and the provision of urban services could reduce an effec- tive "binding constraint" to future growth in Africa. At the very least, such a strategy would create a more hospitable environment for the investors and workers who will increasingly and invariably congregate in urban markets. T H I R D -- African governments that are respon- sive to their citizens and eager to achieve a sus- tainable growth path are increasingly looking to their cities and local authorities to play a greater role in the national development agenda. Local governments can become the strongest advocates for cities, linking up with each other (as the metro cities have done in South Africa) to share lessons and information to make their case. National governments and external donors should welcome the voice of local authorities and should include them in dialogues on the wide range of policies and actions affecting cities and towns. Supporting these efforts, the many good practices and lessons from external assis- tance programmes--including those of Alliance members--should be applied with renewed com- mitment to unblocking the development contri- bution of cities. Bank ldro W/k nemar CartruC se Pictur /StillET SCHYT GENROJ For women in Africa the informal economy is estimated to represent 92 percent of all job opportunities. Scene from an Accra market. 1 Introduction Urbanisation is an t has long been recognised that the growth of cities and towns and the shift- opportunity as well as I ing of population into these urban areas from rural settlements are inevitable a challenge for Africa, aspects of the long-term process of economic development. Yet in Sub-Saharan Africa (hereafter, Africa), the real contributions of urban development to eco- and urban areas are nomic growth and poverty reduction may not be adequately assessed and appre- underutilised ciated. More importantly, these contributions are actually less than they could or should be, due to weaknesses in policy or institutions, or to other factors. resources that con- Much has been written in recent years about Africa's disappointing record of eco- centrate much of a nomic growth (Sachs et al. 2004; The Commission for Africa 2005; The Millennium Project 2005a; World Bank 2005a). Over most of the last two decades, total factor country's physical, productivity, labour productivity, and agricultural productivity per worker have financial, and intellec- been declining or stagnant (World Bank 2002a; ILO 2004). Real GDP growth has been barely positive in per capita terms (Figure 1.1). Without the impetus of income tual capital, and will growth, poverty has remained deeply entrenched and in many cases has worsened. This average picture does not tell the full story, of course, as the continent is inevitably continue to highly diverse and becoming more so. A few notable success stories tell of income do so. gains coupled with declining poverty over the past decade or more (for example, in Botswana, Mauritius, and Uganda), and numerous other promising cases of relatively rapid growth coupled with good governance have emerged recently (including Benin, Burkina Faso, Ghana, Senegal, Mauritania, Mozambique, and Tanzania). But most countries, and the majority of the African people, are mired in place. Virtually all of the countries in the Region must find more viable growth strategies based on improving productivity and sharing its benefits more widely. In the face of this development challenge, every available asset and every phenom- enon experienced in the African countries should be scrutinised to determine how effectively it contributes to ensuring greater welfare and a more positive future for the population. Given this context, this paper asks how urban development--the process of demographic and economic transformation and the resulting management requirements of the affected cities and towns--can and should benefit the coun- tries and what conditions are needed to achieve these benefits. Africa has been 1 2 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A experiencing a historically high rate of urban growth, implying on aver- FIGURE 1.1. Real Per Capita Income Has Stagnated in age a doubling of the urban popu- Sub-Saharan Africa lation in the next 15 years. This GDP per capita across regions, 1990­2003 (PPP constant 1995 int'l $$$) paper views urbanisation as an 8000 opportunity as well as a challenge for Africa and considers urban areas to 7000 Europe & Central Asia be underutilised resources that con- Latin America & Caribbean centrate much of the country's 6000 physical, financial, and intellectual capital and will inevitably continue 5000 to do so. Therefore, it is critical to capita understand better how urban areas per Middle East & North Africa can serve the national growth and 4000 GDP poverty reduction agendas. The countries cannot afford to continue 3000 neglecting their cities; they must, East Asia & Pacific rather, identify cost-effective means South Asia 2000 Sub-Saharan Africa of supporting them. 1629 1580 1519 1496 1502 1501 1515 1585 1533 1547 1557 1574 1593 1618 The next section describes cer- 1000 tain key features of Africa's demo- 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 graphic, economic, and institutional Year setting that have important implica- Source: GDF & WDI (August 2004) , SIMA, 2004 tions for urban development. Here and throughout this report, com- Bank ldro Vitale/W Ami I N T R O D U C T I O N 3 parative data and analysis from other regions are cited where these help to elucidate the issues and the circumstances prevailing in Africa. Of neces- sity, given data shortages and limited comprehen- sive research, anecdotal evidence is sometimes cited to illustrate points. Section III summarises some of the findings from research on Africa and elsewhere indicating the actual and potential gains that urban development can offer to the national agenda of economic growth and poverty reduction. Section IV outlines the requirements for achieving these benefits, while section V sug- gests what is at stake for a country failing to ensure these conditions. Section VI lays out strategic and operational implications and choices posed by the above analysis. A final section draws conclusions from these various perspectives. se Pictur Still WARDS/ ED MARK Abidjan, Cote d'Ivoire. Bank ldro W/k nemar CartruC The urban population growth rate in Africa averages almost 5 percent per year over the past two decades. 2 The Context of Urban Development in Sub-Saharan Africa What is even more The Demographic Setting important to recognise for the Urban Challenge is that that economic Africa's urban population growth has been very rapid--but the"take off"is yet to come. The urban population growth rate of Africa is historically unprece- growth that has taken dented, averaging almost 5 percent per year over the past two decades (although place in recent years in official estimates are quite tentative in some cases).6 On average, the population of the Africa Region is now one-third urbanised,7 a proportion higher than South Africa is on the whole Asia's 28 percent. As portrayed in Figure 2.1, Africa is approaching a demographic mainly urban-based. inflection point as the numbers of new urban residents are projected to rise sharply by over 300 million between 2000 and 2030--more than twice the rural popula- tion increment. This implies that much of the new demand for services and for jobs, as well as the supply of human energy to meet the countries' future needs, will be appearing in urban areas. At the same time, Africa is not structurally over-urbanising relative to patterns in other regions. Figure 2.2 shows that the rate of urbanisation (the urban share of total population) over the past 25 years (1975­2000) has not been rising signif- icantly faster in the developing regions than was the case in currently industri- alised countries when they experienced a similar transition. That is, the width of the bars is about the same for each region (including Africa) over similar spans of time, both in the past quarter century and as projected for the next quarter cen- tury when urbanisation is expected to peak in Latin America and the Caribbean (LAC) and Europe and Central Asia (ECA). In other words, structurally the regions are transforming (in terms of the spatial shift of population) at a similar pace--Africa being no exception. However, it is also clear that Africa is under- going the transition from predominantly rural to predominantly urban while facing much faster absolute urban population growth rates. For Africa, the under- lying dynamic of urban growth is that of the total population, driven by persist- ently high fertility and slowly declining mortality. While there is no formula to suggest what an optimal level of urbanisation would be at any point in a country's development, clearly many countries are con- 5 6 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A FIGURE 2.1. Most Population Growth in the Next Thirty Years Will Be in Urban Areas Sub-Saharan Africa Urban & Rural Population Growth Trends (Millions) Rural population 600 increment 2000­2030 = 141 579 552 500 518 469 430 millions 400 411 in Rural population 345 300 308 Population 274 200 220 211 179 Urban population 135 100 84 Urban population 52 increment 2000­2030 = 367 31 0 1960 1970 1980 1990 2000 2010 2020 2030 Decades Source: UN-World Urbanization Prospects, 2003 fronting bigger demands to manage urban areas (United Nations 2004).8 In countries with good than they appear able to handle. African urbani- urban management, the larger cities can be the sation is also taking place in a context of severe most productive for a number of reasons, in- constraints that many other country groups in cluding especially their ability to match workers other periods did not face--notably, full expo- with jobs (Quigley 1998; Prud'homme 1994).9 sure to the pressures of global competition; more Especially in Africa, the largest city in each coun- limited outlets for external migration (WDR try (regardless of its absolute size) makes a 2003); and depredation of the productive work- disproportionate contribution to the national force and of family security due to HIV/AIDS, economy--Addis Ababa, for example, with which also drains the weak capacity of local 2.6 million residents representing only 4 percent administrations. of the total population, accounts for almost one- fifth of GDP (Cour 2003). In many developing Africa's urban landscape is not dominated by countries, especially in Africa, the quality of very large cities. Much public attention about urban management is the first issue in realizing urban growth focuses on the proliferation of the productive potential of cities, regardless of "mega" cities (defined by the UN as those with population size. more than 10 million residents) or very large cities (over 5 million residents). In 1970, Sub- Africa's city size distribution is also not un- Saharan Africa had no cities in the latter group; in usual. The share of the urban population in the 2000, there were two (Kinshasa and Lagos), and largest city (the urban primacy rate) is also not the UN's latest population projections do not out of line, on average, for Africa's level of devel- assume that this number will increase by 2015 opment, compared to that in the Middle East­ T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 7 North Africa (MNA) and Latin America and the The urban population in Africa is widely dis- Caribbean (LAC) regions.10 Excessive urban pri- persed across mainly small settlements, but not macy matters because it can entail a significant unusually so compared to other regions. As seen cost in economic efficiency. Urban primacy tends in Figure 2.3, Africa's city size distribution is quite to be positively related to low per capital income, comparable to that of other regions at all in- low trade integration, limited transport net- come levels, although a slightly higher share of works, non-democratic governance, and concen- urban Africans live in the settlement class below trated political power--characteristics shared by 200,000 inhabitants--about 52 percent versus many African countries--and inversely related 42 percent for all developing countries. 11 to geographic size (Ades and Glaeser 1995; Hen- About 12 percent of urban Africans live in set- derson 1999, 2000). The political and economic tlements of 1 to 5 million people, almost the same reforms underway in many of the African coun- proportion as in other developed and develop- tries will tend to gradually redress most of these ing regions. This city size category, which has the negative variables thereby reducing urban greatest potential for urban productivity, poses primacy naturally, even without active policy major management challenges, especially for attempts to affect the relative weight of the largest countries with very low incomes and weak city. administrative capacity. At the same time, sec- FIGURE 2.2. Sub-Saharan Africa Is Undergoing Urban Transition with a Relatively High Urban Population Growth Rate 6 1975­2000 2000­2025 5 Sub-Saharan Africa 4 Middle East & North Africa East Asia & Pacific (percent) South Asia Region Sub-Saharan Africa rate All developing countries 3 South Asia Region Latin America & Caribbean growth Developed countries (1900­1925) Middle East & North Africa All developing countries urban 2 East Asia Pacific Average Europe & Central Asia Latin America & Caribbean 1 Europe & Central Asia 0 0 10 20 30 40 50 60 70 80 90 100 Percent urban population share (25-year increment) Note: Dotted line shows 1975-2000 period (actual); solid line 2000-25 period (projected). Lines indicate increase in share of urban population between end-point years of these 25-year increments. All averages weighted by population. Source: Reproduced from WDR 2003, p. 113. 8 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A these two factors combined ac- counts for less than half (about FIGURE 2.3. Africa's Urban Population Distribution by Settlement Size is Similar to that of Other Regions (2000 estimates) 40 percent) of urban growth in most developing countries.12 Internal 60 migration rates tend to rise in peri- ods of economic growth and to fall Sub-Saharan Africa during economic downturns.13 50 Low income Middle income In general, rural-to-urban migra- High income tion can be explained by two forces: 40 the attraction of economic opportu- nity in cities exerts a "pull," while the cent 30 limitations of opportunity in rural erP areas create a "push." The latter may be stronger in some countries of 20 Africa where agriculture has been stagnant or declining or where local 10 conflict has devastated the country- side, even in the face of economic 0 problems in the receiving areas. But <200K 200­300K 300­400K 400­500K 500K­1M 1M­5M 5M­10M >10M migration research in recent decades No. of inhabitants (2000) has found that the traditional view of one-way movements mainly from Source: Vernon Henderson, World City Populations, 2002 rural to urban areas is by no means the whole story and is much less important in overall population mobility than are circular and seasonal migra- ondary and tertiary urban settlements tend to be tions (Ellis and Harris 2004). Moreover, rural and the most rapid growing, and managing expan- urban boundaries are artificial distinctions to sion alone is demanding. A further concern is households, which often distribute members that many of the secondary and tertiary urban across different spatial and economic activities areas have rather limited economic relations with to diversify income sources and reduce risk. each other and do not constitute an effective Options for migration, or, more accurately, pop- urban network--or an effective urban-and-rural ulation mobility, are crucial to ensuring sustain- circuit of exchange--because of the sparse able livelihoods, especially for households facing domestic transport linkages. constant uncertainty in climates and markets, as in Sub-Saharan Africa. Such mobility is essential Migration from rural areas is not the primary as well to enable individuals to gain new experi- explanation for urban population growth. A ences and income they can use wherever and common misconception exists that the vast however they perceive maximum utility. majority of new inhabitants in urban areas are migrants from the countryside. Urban dem- In sum, the demographic picture in Africa is ographic growth has three sources: natural one of rapid and dramatic change--and yet not increase among existing urban residents, reclassi- a situation that is anomalous or wildly out of fication of formerly rural areas as urban, and line for the region's level of development or rel- internal rural-urban migration. While it is diffi- ative to other regions. The real surge in Africa's cult statistically to separate reclassification from urbanisation is yet to come; it will occur in the migration on the basis of census results, estimates next thirty years, when the urban population is consistently indicate that the median value of projected to nearly triple and become the major- T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 9 ity. Most of this growth will occur due to the nomic development, but the annual growth rates transformation of rural settlements at the urban of each are not closely linked since urbanisation periphery, as they become more densely popu- often proceeds apace even in periods of economic lated and less dependent on agriculture, and to stagnation or decline (Fay and Opal 2000). This natural population increase in cities. Population becomes evident when comparing the levels of mobility will contribute to the growth of cities, urbanisation and of constant GDP per capita over especially in the least urbanised countries, but the last decade (between endpoints 1990 to 2003), many households will retain footholds in both the for a large sample of developing and transition rural and urban economies. Both the large cities countries (Figure 2.4a). A positive linear relation- (over 1 million residents) and the many rapidly ship is overall quite clear, although some countries growing smaller cities pose major challenges for show increased urbanisation with no increase in local government administrations with character- income (near vertical lines) and even with declin- istically weak capacity. ing income (backward bending lines).14 Looking more closely at Africa (Figure 2.4b), The Urban Economy: the linear relationship also emerges. Only 9 of the Growing, Yes--But from 24 countries shown15 are cases of "disconnect," showing urbanisation rising in the face of negligi- Informal, Small-Scale and ble or negative economic growth. As might be Vulnerable Enterprises expected, some of these have experienced civil unrest (Rwanda) or major transition (South Urbanisation with urban economic growth. Africa), although others reflect more complex Across all countries over time, the level of urbani- issues of performance (Cameroon, Cote d'Ivoire, sation is strongly correlated to the level of eco- Kenya, Madagascar, Niger, Togo, and Zambia). FIGURE 2.4a. Developing and Transition Economies: Comparing Levels of Per Capita Income and Urbanisation,1990­2003 90 Africa 80 East Asia & Pacific Latin America & Caribbean 70 Europe & Central Asia Middle East & North Africa 60 South Asia tion 50 popula 40 urban % 30 20 10 0 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 Ln gdp pc, ppp, 2000 intl dollars 10 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A FIGURE 2.4b. Sub-Saharan Africa: Comparing Levels of Per Capita Income and Urbanisation, 1990­2003 70 60 50 tion 40 popula 30 urban % 20 10 0 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 Ln gdp pc, ppp, 2000 intl dollars Benin Botswana Burkina Faso Cameroon Cote d'Ivoire Ethiopia Ghana Guinea Kenya Lesotho Madagascar Malawi Mauritania Mauritius Mozambique Niger Nigeria Rwanda Senegal South Africa Tanzania Togo Uganda Zambia While between 1990 and 2003 some countries in income can only be assessed indirectly as very few Africa experienced increasing "urbanisation with- developing countries, and none in Africa, pro- out growth" (Fay and Opal 2000), the urban trend vide spatially disaggregated national accounts. A is no more responsible for this disconnect or for "local domestic product" is estimated for some explaining the weak growth performance of the major cities in South Africa--for example, the economy than is any other phenomenon during economies of Johannesburg, Cape Town, and this period, such as structural adjustment or eThekwini (Durban) together make up some reforms in governance. In other words, Africa's 50 percent of the country's GDP but represent "growth tragedy" in the 1990s has disappointed only 20 percent of the national population hopes and expectations of all sectors. Yet what is (SACN 2004).16 even more important to recognise, as illustrated A very rough approximation of the contribu- below, is that the economic growth that has taken tion of the urban-based activities to the national place in recent years in Africa is on the whole economy can be made from the secondary and mainly urban-based. It has just not been sufficient tertiary sectors (industry and services, respec- to propel the countries into the rate of per capita tively), which in general take place in cities and income increase they need to overcome poverty towns, especially in terms of formal enterprise. sustainably. Agriculture, the primary sector, is of course Breaking down economic growth by sector and mainly rural-based, although urban agriculture source. The "urban" contribution to national is quite significant in developing countries.17 T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 11 Recognising that some margin of error may lead Combining the effects of sectoral share and to both overestimation and underestimation of sectoral growth rates gives the percentage contri- the value-added produced in urban areas, Table bution that industry and services have made to SA2 in the Appendix summarises the sectoral total GDP growth over the past decade (Figure shares of GDP, the annual growth rates of each 2.5c): from 79 percent in Africa to about 100 per- sector, and the combined contribution of indus- cent in ECA. Thus, despite the rather sluggish try and services (the putative "urban share") to performance of industry in many cases, it is rare overall GDP growth, over more than the past that the secondary and tertiary sectors combined decade (1990 to 2003). fail to provide the main power behind overall Surprisingly, African economies are no more reliant on agriculture (averaging 19 percent of GDP) than are those in the East Asia and Pacific FIGURE 2.5a. Sectoral Output as Share of GDP, 1990­2003 (average %) region (EAP), and less so than those in South Asia (Figure 2.5a). The Agricultural large share of services in Africa as Sub-Saharan Africa Industry more than half of total GDP is sur- Services prising as this is normally a feature East Asia & Pacific of more developed countries, but in Africa the outcome reflects the rela- tively underdeveloped state of the South Asia region's industry. The composition of services in Africa also represents Middle East & lower-value activities (more basic North Africa trading, less information and bank- Europe & ing intensive enterprises) than exist Central Asia in the other regions. Comparing the annual growth Latin America & rates of each sector over the period Caribbean reveals quite a mixed picture (Figure 2.5b). In most regions and income Lower middle income groupings, the services sector has grown more rapidly than has either agriculture or industry, while in- Low income dustry has tended not to lead and has sometimes trailed agricultural Middle income growth. In Africa, industry's growth fell behind that of overall GDP over most of the 1990s (although it Upper middle income picked up in the second half of the period, as discussed below). Agricul- High income tural output in Africa has grown at about the average rate for low in- 0 10 20 30 40 50 60 70 80 come countries. Only East Asia and Percent of GDP the Pacific (EAP), and to a lesser extent South Asia (SA), show a very Source: GDF & WDI Central, April 2005 vibrant growth rate for industry. 12 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A countries averaged a 3.4 percent FIGURE 2.5b. Average Annual Growth of Output by Major Sector, annual increase in GDP, almost 1990­2003 (%) triple the rate of the previous five years. Figure 2.6 shows that all of the relatively strong performers (coun- tries averaging at least 1 percent per Sub-Saharan Africa capita growth during the 1996 to 2003 period), with the exception of East Asia & Malawi, had positive growth in the Pacific nonagricultural sectors. This pattern is actually quite robust for other South Asia countries as well. Lopez (2005) finds, for a cross-region sample of Middle East & 14 countries in the 1990s and for a North Africa much larger global sample from the 1970s to the present, that "(eco- Europe & Central Asia nomic) growth has been high where the non-agricultural sector has Latin America & enjoyed high growth rates and it Caribbean has been low where that sector has shown low growth (i.e., growth is Lower middle income mainly driven by the nonagricul- tural sector)" (Lopez 2005, 10). Despite the relatively stronger Low income performance of the urban-based sectors, for most of Africa the growth of total GDP in per capita Middle income terms has been insufficient to achieve a turnaround in poverty. Upper middle income Agricultural Clearly, countries need to mobilise Industry more effectively all of their produc- Services tive capacities--including not only High income Gross Domestic Product agriculture, especially higher value- added crops, but also services ­3 ­1 1 3 5 7 9 11 and manufactured goods, which Percent growth tend to sustain a high positive income elasticity of demand over Source: GDF & WDI Central, April 2005 the medium term. Informality is the main game in GDP growth. In Africa, industry and services fall town. The dominant story of production and below half of GDP growth only in Cameroon, growth in Africa, as in many other low income Cote d'Ivoire, Malawi, Niger, and Rwanda, due developing and transition regions, is in the infor- to stronger agricultural performance in those mal economy. The informal economy workforce countries during the period.18 is estimated to account for 78 percent of nonagri- It is also useful to look at which sectors have cultural employment in Africa, 93 percent of all been the main sources of growth in the more new jobs created, and 61 percent of urban recent period (1996 to 2003) when the African employment. Similar figures for Latin America T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 13 and for Asia are also significant but lower (Table 2.1). For women in FIGURE 2.5c. Contribution of Industry and Services to Total GDP Africa, the informal economy is esti- Growth, 1990­2003 (%) mated to represent 92 percent of all job opportunities outside of agri- culture, overwhelmingly as self- Sub-Saharan Africa employment or own-account work (ILO 2002). While the nonagricul- tural informal economy cuts across East Asia & Pacific all sectors and locations, it flour- ishes especially in urban areas be- cause its activities are essentially South Asia demand-driven and therefore re- sponsive to population concentra- Middle East & tions (Cour 2004). North Africa The contribution of the informal economy to nonagricultural GDP Europe & tends to decline with level of devel- Central Asia opment. In Africa, this share is cur- rently estimated to average about Latin America & Caribbean 40 percent and even amounts to 31 percent in Asia (Table 2.2). The contribution of the informal "non- Lower middle income primary" (nonagricultural) activi- ties has been estimated to be as Low income much as one-third of even total GDP in Cameroon (Club du Sahel and PDM, 2004, 14). Thus, failures Middle income to adequately account for this activ- ity could lead to a statistical under- appreciation of economic output, Upper middle income especially of urban areas in Africa. While it was once commonly believed that the informal economy High income is entirely marginal and unproduc- tive, growing evidence in all de- 0 20 40 60 80 100 120 Percent veloping countries illustrates its contributions in output and its Source: Appendix Table SA2. importance both to workers and to consumers (for example, see Ma- loney 1999 on Mexico). Such find- ings make a case for enabling policies. In South facturing sector in Nigeria found, however, that Africa, surveys reveal that the openly unemployed most such enterprises appear profitable for their are less well-off than are persons in informal owners but generate little employment through employment and that the latter activities may firm growth. They often remain very small and have significant entry barriers, such as required undercapitalised and have minimal transactions experience and social networks (Kingdon and with larger, formal sector companies (CBN/ Knight 2003). A survey of the informal manu- NISER/FOS 2003). A parallel study looked at 14 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A FIGURE 2.6. Industry and Services Were the Main Sources of GDP Growth among the Better African Performers from 1996­2003 rate) 7.0 Mozambique* growth 5.0 (% Mauritius* Uganda* Botswana* Ethiopia* 3.0 Tanzania* Senegal* Burkina Lesotho* Mauritania* Faso* Guinea* Zambia Benin* Rwanda* added/capita Togo* 1.0 Ghana Nigeria* South Africa Cameroon* Niger value Madagascar ­7.0 ­5.0 ­3.0 ­1.0 1.0 3.0 Malawi* 5.0 7.0 Kenya ­1.0 Cote d'Ivoire services ­3.0 and ­5.0 Industry ­7.0 Agriculture value added/capita (% growth rate) *Denotes countries with an average GDP/capita growth rate of more than 1 percent over the period 1996-2003 Source: Appendix Table SA3. informal distributive trading, which has grown to structures, they must rely on their interpersonal dominate the informal economy in Nigeria networks. This keeps them trapped in local circles because real production, such as manufacturing, of exchange and discourages relations with out- has been discouraged by infrastructure bot- side partners who could enlarge their horizons tlenecks and by macro instabilities. Such cir- and markets (Collier and Gunning 1999; Murphy cumstances have discouraged 2002). Even formal sector firms depend heavily For women in risk-taking entrepreneurship and on ethnic ties, a dependence that slows entry and have favoured short-term, rent- may discourage the business community from Africa, the informal seeking activities, such as mer- pressing collectively for reform (Eifert, Gelb, and economy is esti- chandising. The studies found Ramachandran 2004). little exchange between the infor- mated to represent mal manufactures and traders, Structural constraints limit the economic 92 percent of all job however, and neither group re- returns to urbanisation--as to other develop- ported using formal banking chan- ments. In much of Africa where informal, non- opportunities out- nels (CBN/NISER 2003). wage employment is the norm, the concern to side of agriculture, Informal firms in Africa are also policy makers should not be that this activity is more vulnerable to failures in sys- "unhealthy" or unproductive, but rather that it overwhelmingly as tems of legal protection. When has not been accompanied by a robust growth of self-employment or they cannot count on fair and formal enterprises in the urban areas that impartial enforcement of laws and should be able to foster them. Even though eco- own-account work regulations or on business support nomic growth has predominantly depended on T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 15 secondary and tertiary activities, as indicated above, neither of these sectors is very deep or TABLE 2.1. Relative Importance of the Informal Economy buoyant. The decline over the past decade (1992 in Employment to 2002) in industry's share of GDP in Africa has Informal Workforce, as share of: Africa the Caribbean Asia been termed an "effective de-industrialisation" from what was already a shallow base (Fox et al. Nonagricultural employment 78% 57% 45­85% 2004). This decline resulted from the retrench- Urban employment 61% 40% 40­60% ment of public sector enterprise due to adjust- New jobs 93% 83% NA ment programmes, while formal sector wage Source: Charmes, J. Estimations and Survey Methods for the Informal employment has failed to pick up the slack and Sector, University of Versailles, 2002, as cited in Becker 2004. expand. What industrial sector growth took place has in some cases, for example, in Ghana, represented activity in mining rather than in greatly reducing accessibility, and it has almost manufacturing. Not surprisingly, the share of twice Asia's share of population living at low manufactures in total merchandise exports density, which implies high costs for delivering averaged less for Africa in 2000 to 2002 (35 per- services (Sachs et al. 2004). cent) than for any region except Middle Third, Africa has a relatively high share of East and North Africa, where fuels dominate population living in disadvantageous ecozones (WDI 2004). (tropical or arid). The Region is also depleting Despite urbanisation, several structural fea- tures and disadvantages that limit economic transformation and the maturation of firms con- front many of the African countries. One of the TABLE 2.2. Contribution of the main reasons for the soft performance of African Informal Sector to the GDP in industry is that, compared to low income Asian Selected Developing Countriesa countries, total factor productivity is much lower Country (year) Informal Sector GDP and the share of indirect costs is much higher. as Percentage of Nonagriculture GDP This makes African companies that would attempt to export noncompetitive and discour- Northern Africa 27 Sub-Saharan Africa 41 ages domestic investment (Eifert, Gelb, and Benin (1993) 43 Ramachandran 2004). The growth taking place in Cameroon (1995­96) 42 manufacturing and services therefore derives Kenya (1999) 25 mainly from small scale enterprises using low Mozambique (1994) 39 skills and low capital endowments and operating Tanzania (1991) 43 Latin America 29 mainly in response to domestic demand and Colombia 25 therefore limited by it. Mexico (1998) 13 A second and related problem is that transport Peru (1979) 49 costs, whether local, regional, national, or interna- Asia 31 tional, for the typical African country are at least India (1990­91) 45 twice those of its Asian counterparts (Starkey et Indonesia (1998) 31 Philippines (1995) 17 al. 2002). Transport costs are one of the main fac- tors favouring agglomeration, but inordinately Source: ILO, Women and Men in the Informal high costs burden even urban-based production Economy--A Statistical Picture (Geneva, 2002). a by impeding the growth of a balanced city system Note that estimates in this table are based on informal enterprises, not all informal workers. and the development of external markets. Com- Some estimates of informal value-added are pared to the rest of the developing world aver- included in official national accounts, but not age, tropical Africa also has less of its population necessarily to the extent indicated in this table. within 100 km of a coast (25 versus 66 percent), 16 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A Bank ldro W/k nemar CartruC Straw craft industry in Lagos, Nigeria. its natural resource base at a rapid rate, as indi- these costs back to households through reduced cated by land per capita endowments that have benefits and to government, especially at the local fallen by almost half in several countries between level, where municipalities face increased de- 1960­69 and 1990­90 (Sachs et al. 2004; WDR mands on services and reduced repayment capa- 2003). This trend reflects population growth as bilities (Rosen and Simon 2003). The epidemic well as reductions in the quality of land and water is much of the reason why Africa's hard-won available for effective cultivation. African farming investments in human capital and other assets, has continued to depend on extensive production including administrative capacity, become un- rather than shifting to intensive (higher produc- derutilised and degraded. tivity) cultivation, which would be key to rural transformation (World Bank 2002a). Conclusions on economic growth. Probing the sources and potentials of economic growth in The cruellest factor draining Africa's devel- Africa is important because it is widely recog- opment impetus is, perhaps, the scourge of nised that sustaining high income growth is a HIV/AIDS. In countries with high prevalence, necessary condition for a significant reduction including the region's largest economy, South in poverty, although it is not a sufficient condi- Africa, the economic cost of the epidemic may tion in countries with high inequality (Dollar and be far worse than the often-cited estimate of an Kraay 2002; Kraay 2003). It is estimated that, annual loss of one percent of GDP (Bell, Devara- assuming continued population growth of 2 per- jan, and Gersbach, 2003). Studies have confirmed cent per year, cutting poverty rates from 47 per- the impact to the private sector, especially of high cent in 2001 to 22 percent by 2015, in line with absenteeism, and the tendency for firms to shift the Millennium Development Goals, would re- T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 17 quire that the real GDP of African countries grow percent of the rural rate for Ethiopia, Kenya, by at least 6 percent per year, or slightly less if Malawi, Mozambique, and Nigeria. The magni- inequality improves (Iradian 2005). Only four tude of urban poverty is also significant by any countries--Botswana, Mauritius, Mozambique, standard, amounting to a third or more of the and Uganda--sustained growth rates close to urban population, in Ethiopia, Gambia, Kenya, this target in the past decade, thanks to strong Madagascar, Malawi, Mozambique, Niger, Sene- industry and services (see Appendix Table SA2). gal, and Zambia. Incidence of this order in other Clearly, all of the productive potential of the regions is seen in Mongolia, Bangladesh, Bolivia, countries must be mobilised much more force- Ecuador, Honduras, and Nicaragua. fully if a durable decline in poverty is to be With projected urbanisation, the relative rural ensured. It is argued here that the productive and urban poverty rates, if unchanged over the potential of these sectors located mainly (and next twenty years, would imply a growing share more favourably) in urban areas should be a of total poverty occurring in urban areas21-- particular focus of national growth strategies approaching or exceeding the majority in Benin, because of their comparatively strong perform- Kenya, Mauritania, Mozambique, Nigeria, and ance in Africa as in other regions. This would be Senegal (Figure 2.7b). particularly true in the African countries with relatively unfavourable or declining natural con- Urban poverty rates are not mainly a function ditions for agriculture. of pressures from urban population growth. The correlation between these two phenomena is weak, with a coefficient of 0.41 for the cross- Urban Poverty: regional sample, 0.22 for the Africa subsample Already Significant, alone (Figure 2.8). This indicates that urban and Likely to Grow poverty is not simply a matter of queuing for jobs and services. The phenomenon of urban poverty An almost universal finding in developing coun- can reflect various economic and institutional tries is that whether measured in terms of factors, varying across cities in the same country. income, consumption, or expenditure, rural As a general rule, access to essential services is poverty rates exceed urban poverty rates, often by better, and income poverty rates are lower, in a very large margin, and rural poverty is deeper larger versus smaller cities, although it is espe- (further from the estimated poverty line).19 This cially difficult to generalise for Africa.22 result would be expected, given that urban areas There are many methodological issues with provide a wider and deeper labour market, per- the measurement of poverty in any context, rural mitting higher incomes and capacity to pay for or urban. One type of problem (acknowledged services, and that density of settlement and prox- in the Burkina Faso poverty assessment, World imity to centres of government should allow pro- Bank 2003d) is that poverty surveys may fail to vision of many services at lower per capita cost. differentiate estimated poverty lines sufficiently Location in non-urban areas and sometimes even between rural and urban areas based on respec- distance from cities are found to be markers for tive price indices, which should be even higher poverty in country studies. for large cities than for smaller ones, and thereby Figure 2.7a shows the poverty incidence can overestimate urban purchasing power. A (headcount rates) for a sample of countries in related issue is that the consumption basket used Africa and other regions based on official data for to estimate an urban poverty threshold may take both urban and rural aggregates.20 Given the insufficient account of the nonfood expenditures expected real economic advantages of urban that urban households must make to meet their location, what is striking is not where the figures basic needs, given that virtually all consumption show wide gaps but where they are actually quite requires a cash outlay. For example, the official close--the urban poverty rate being within 20 poverty line as a multiple of the costs of a "mini- 18 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A mum food basket" ranges from 1.0 FIGURE 2.7a. Rural and Urban Poverty Rates, latest year in Burkina Faso (nonfood expendi- tures are assumed to be equivalent Bolivia to food expenditure) to 2.0 in Chad, Brazil Ecuador while in the United States, a coeffi- El Salvador cient of 3.0 is used (Satterthwaite Guatemala Honduras 2004a). Mexico Estimates of actual expenditure Nicaragua on nonfood items do not reflect Albania whether households' nonfood needs Armenia Poverty headcount rural are being met--for example, when Azerbaijan (% of rural population) Bulgaria they walk long distances rather than Georgia Poverty headcount urban take public transportation to work Kazakhstan (% of urban population) or report consuming limited quanti- Kyrgyz Republic Moldova ties of water.23 Non-monetary costs Romania of obtaining water can be very sig- Tajikistan nificant in urban areas, despite Turkmenistan Uzbekistan reported physical proximity to sup- ply points, because of the sheer Egypt, Arab Republic Morocco numbers of people depending on Tunisia them. It has been shown that in 10 Yemen, Rep. urban sites in East Africa, for exam- Bangladesh ple, for households lacking in-house India connections, the average queuing Nepal Pakistan time for water rose from 28 minutes Sri Lanka per day in 1967 to 92 minutes in 1997 (Thompson et al. 2000). Cambodia China Indonesia Assessing the nature and extent of Mongolia urban poverty requires looking Philippines Vietnam beyond monetary measures. Some deprivations result from issues of Benin Burkina Faso affordability, which may explain, for Cameroon example, the food insufficiency Cote d'Ivoire reported in the Benin poverty Ethiopia Gambia assessment as being equally preva- Ghana lent in urban as in rural areas Guinea (World Bank 2003c). But low Kenya Madagascar income is not always the barrier. Malawi Problems in obtaining adequate Mauritania infrastructure and social services, Mozambique Niger health and education status, and Nigeria personal and communal security Rwanda Senegal can reflect underlying institutional Tanzania obstacles and social exclusion that Zambia even middle-income households 0 20 40 60 80 100 cannot fully surmount, as further described below. Source: Appendix Table SA4. T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 19 Disaggregating urban access data reveals the extent of intra-urban FIGURE 2.7b. Urban Shares of Total Poverty, latest year and inequalities. As would be expected projected 2020 for reasons listed earlier, the Demo- graphic and Health Surveys (DHS) Bolivia confirm that in all major developing Brazil El Salvador regions rural residents almost Guatemala invariably have much lower rates of Honduras Mexico access to infrastructure services Nicaragua (either piped water on premises, Albania flush toilet, electricity, or all three) Armenia than even the urban poor. At the Azerbaijan same time, there are large, statisti- Bulgaria Georgia cally significant gaps between the Kazakhstan access of the urban poor versus the Kyrgyz Republic urban non-poor, and these gaps are Moldova Romania often greater than those between the Tajikistan urban poor and the rural residents. Turkmenistan This observation is a testament to Uzbekistan urban inequality and institutional Egypt, Arab Republic rigidities rather than to absolute Morocco Tunisia resource constraints in urban areas. Yemen, Rep. In many cases, extending services Bangladesh from the better-off to the less well- India off neighbourhoods nearby would Nepal cost much less than would extend- Pakistan 2002 Urban Poverty Sri Lanka as % of Total Poverty ing those same services to the more remote and scattered rural popula- 2020 Urban Poverty Cambodia as % of Total Poverty tions.24 China Neighbourhoods of extreme Indonesia Mongolia deprivation in terms of basic serv- Philippines ices and local public goods--typi- Vietnam cally informal settlements, which Benin often lack secure tenure and are Burkina Faso located in environmentally precari- Cameroon Cote d'Ivoire ous sites--are home to a majority of Ethiopia the population in most African Gambia cities (for example, 72 percent in Ghana Guinea Douala and 62 percent in Yaoundé, Kenya Cameroon) (World Bank 2004a). Madagascar For all of Africa, over 70 percent of Malawi Mauritania the urban population is estimated to Mozambique suffer shelter deprivation in terms of Niger inadequate housing, water supply, Nigeria Rwanda or sanitation (UN-Habitat 2003).25 Senegal In many developing countries, the Tanzania informal settlements or slums house Zambia a mix of socioeconomic classes, 0 20 40 60 80 100 Source: Appendix Table SA4. 20 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A FIGURE 2.8. Urban Poverty Rates Are Not a Function of Urban Population Growth 8 Tanzania Mozambique 1990­2003 6 Niger Cambodia Gambia Kenya %), NepalMauritania Madagascar Nigeria Honduras ual Benin Burkina Faso Indonesia Cameroon Yemen, Rep.Ethiopia Malawi Senegal (ann Philippines Guinea Bangladesh 4 China El Salvador Rwanda Cote d'Ivoire owthrg Vietnam Pakistan Nicaragua Bolivia Morocco Ghana Guatemala Ecuador Tunisia India Zambia tion Turkmenistan Brazil 2 Sri Lanka Mexico Egypt, Arab Republic Albania Sub-Saharan Africa popula Mongolia Uzbekistan Azerbaijan Middle East & North Africa Kyrgyz Republic East Asia & Pacific urban Tajikistan 0 Romania South Asia Georgia Kazakhstan Europe & Central Asia Bulgaria Moldova Latin America & Caribbean Average Armenia Correlation coefficient: 0.41 ­2 10 20 30 40 50 60 70 Urban poverty rate, base year (% of urban population) including middle-income residents. Real differ- the exposure that residents of some areas face in ences in welfare, such as certain health outcomes, terms of the threat of crime and violence, which can be seen between residents and nonresidents are often very spatially concentrated within cities. of such settlements. These differences would be The urban population is dependent on cash expected to appear especially for aspects of wel- income to purchase all of their necessary goods fare that cannot be privately purchased through and services, including food. Therefore, macro- the market, so that even having disposable economic shocks, including cuts in government income is no guarantee of access. Such dimen- expenditures with fiscal adjustment, tend to hit sions of welfare would derive especially from urban areas particularly hard. Such effects were local externalities and public goods, which deter- seen in Zimbabwe after the combined fiscal mine many of the vulnerabilities that urban retrenchment and drought in the early 1990s households experience--those relating, for (Ersado 2003). In Burkina Faso and other West example, to an unsanitary environment.26 The African countries, the devaluation of the CFA widespread lack of sanitation standards adequate regional currency and related structural reforms to urban density helps to explain why the MDG led to a proximate rise in urban poverty (Grimm target of reducing infant mortality is projected and Gunther 2004).27 These impacts reflect the to be met in urban areas in only one-quarter (6 of vulnerability of the urban population to food 24) of the countries studied by Sahn and Stifel price increases and to declines in public sector (2002), even less than in rural areas where 10 of employment, which then have second-order 24 countries are projected to be on-target. So- effects on the urban poor as overall demand for called "neighbourhood effects" can also be seen in their labour and services declines. T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 21 In comparing changes in asset poverty and in The Institutional Setting: seven education and health indicators between Urban Governments DHS periods in African countries, the Sahn and Stifel (2003) dataset reveals that only in 8 out of Becoming More Mature, 24 countries have the majority of these indicators from a Low Base shown greater improvement in urban areas than in rural areas (or less deterioration in urban than As in other developing regions, large numbers of in rural areas). The urban areas showed relatively the African countries over the past decade have better improvement than rural areas in terms of established newly democratic processes at the asset poverty, neonatal care, and contraceptive local government level, with elected mayors or use--indicators in which urban areas might be councils. These countries have also increasingly expected to show advantages reflecting the sup- decentralised fiscal authority to these local posed ease of infrastructure coverage and service governments--at least formally, although not access. In all the other indicators studied, how- always in fact. ever, including school enrolments, infant mor- Cities' financial performance, and therefore tality rates, child stunting, and female adult also their performance in service delivery, depends malnutrition, urban populations improved less in the first instance on the intergovernmental fiscal (or worsened more) than did the rural popula- framework that determines their authority to tax tions.28 These indicators could reflect multiple and their access to various forms of central rev- inadequacies, including deteriorating service enues (directly or through transfers). Local gov- quality and social and institutional barriers to ernment revenue (including taxes and grants) and effective access to services.29 expenditure as a share of GDP varies widely across all regions, partly as a reflection of In brief. Urban poverty is clearly a major chal- decentralisation policy. In the Euro- The growth of lenge, both in income and non-income measures, pean Community, for example, in Africa as elsewhere. The growth of urban pop- local public expenditure averages urban populations ulations and increasing urbanisation will enlarge 11 percent of GDP (spanning 31 and increasing urban- the numbers of urban poor and will most likely percent in Denmark to 2.8 percent raise their share among the country's total poor, in Greece).30 Local revenue and ex- isation will enlarge but this trend will also tend to dampen the over- penditure each represent 1.4 per- the numbers of all incidence of poverty, since urban areas usu- cent of GDP for Mexico and range ally offer lower risks of poverty. In explaining from 5 to14 percent of GDP for a urban poor and will urban poverty, it would appear that factors other sample of transition countries.31 most likely raise their than the pressures of urban population growth In Africa, local revenue amounts are at play and that institutional failures represent to 5 percent of GDP in Uganda and share among the an important explanatory factor. The fact that 3.5 percent in South Africa (IMF country's total poor, most welfare indicators are better for urban resi- GFS 2004), but the more common dents than for rural residents on average--often level is 1 percent or less, as in but this trend will very much so--underscores the inherent advan- Benin, Burkina Faso, Cameroon, also tend to dampen tages that urban areas confer, in access both to Cote d'Ivoire, Ghana, Madagascar, incomes and to other assets and benefits. It can and Senegal (Chambas and Duret the overall incidence also be easier to combat poverty in urban areas 2000). African local governments of poverty, since than in rural areas by providing opportunities for rely less on taxation and more on work and for services. The disparities that persist transfers than do more highly urban areas usually within urban areas have little economic justifica- decentralised economies; how- offer lower risks of tion, however, and clearly reflect political and ever, in these last seven countries, institutional inequities. local revenues represent no more poverty. 22 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A than 5 percent of central government receipts. An State spending on investment does not neces- analysis of municipalities in Ghana, Madagascar, sarily make up for the lack of local capital fund- and Senegal found that their annual per capita ing. In Cameroon, less than 1 percent of state revenues amounted to less than US$10 (Cham- revenues is spent on capital expenditure for bas and Duret 2000). 18 cities, most going to the administrative capi- A common problem, for example in Mada- tal (Yaoundé) and the economic centre (Douala) gascar, is that the tax and borrowing authority (World Bank 2004a). In Niger, only 8 percent of for the cities, especially the large cities and the the national investment budget is devoted to localities facing rapid population growth, has urban areas (across all sectors), versus 80 percent not adapted sufficiently to their greater expen- to rural areas. The local governments in Niger, for diture and service delivery obligations (World their part, spend funds equivalent to 7 percent of Bank 2001). A wide dispersion exists in the the state's urban expenditure, and the capital and extent to which local governments use property maintenance budgets of Niamey are about two taxation, which can be a buoyant source to three times those of the regional capitals and for cities. South Africa is unique in relying on it smaller urban communities combined (World very extensively, for 72 percent of all local gov- Bank 2004c). In such circumstances, it is little ernment tax revenues (IMF GFS 2004). Some wonder that in many African cities firms and countries have tried to eliminate local sur- households subsist by their own grit and that charges and fees as "nuisance" taxes, but without public services are almost nonexistent outside the permitting the municipalities to replace them, wealthy neighbourhoods. thereby putting local government functions at Although African municipalities continue to risk. have very weak fiscal and administrative means, se Pictur TE/Still SCHYT GENROJ Checking out the book vendor on the main street in Cotonou, Benin. T H E C O N T E X T O F U R B A N D E V E L O P M E N T I N S U B - S A H A R A N A F R I C A 23 elected local authorities are beginning to perceive This chapter has reviewed some of the main their new potential and to look to each other for features and trends in Africa with respect to encouragement and good practice. Illustrations of urban and local government development and this trend include a region-wide conference, "Afri- the urban economies. The challenge is to harness cities" held every three years by the Municipal the strong demographic momentum and energy Development Partnership since 1998, the forma- of cities to lead the national economies forward tion of the South African Cities Network among with more sustainable income growth and the nine largest metropolitan cities in that coun- poverty reduction. The next chapter discusses try, and the creation in 2005 of a Council of Cities how urban areas can serve as resources and and Regions of Africa (CCRA) from the merger of sources of strength to the advantage of the coun- three other subgroups on the continent with the try's development agenda, followed by an out- aim of providing "a unified voice for sustainable line of the conditions and ingredients needed to local government development in Africa."32 realise these benefits. City putoa M Aerial view of Maputo Bay, Mozambique. What Cities and Towns 3 Can Do for a Country's Development Agenda-- and How Good urban manage- Urban and Rural Development ment poses numerous as a Virtuous Circle synergies with the Urban economic theory and most urban history hold that urbanisation emerges from the transformation of agriculture, with the resultant release of labour and prevailing national growth in demand for more diversified goods and services. Yet in most of Sub- development goals of Saharan Africa agricultural productivity has not "taken-off " over the decades since independence (ILO 2004).33 Many parts of these countries still face major sustainable growth, obstacles due to constraints of geography and transport and natural resource poverty reduction, and limitations in terms of declining soil fertility and the economic feasibility of exploiting water for crops (Bryceson 2004). The World Bank estimated in its 2003 good governance. World Development Report that 250 million people in Africa, or 30 to70 percent of the inhabitants of most of the countries, live in "fragile" ecosystems, which, by definition, have very limited agricultural potential.34 The physical and institu- tional conditions for a technology-based revolution in agriculture, such as occurred in Asia, are not yet apparent in most of the African continent (Ellis and Harris 2004). This means that the urban economies lack a buoyant source of domestic demand from the hinterland, although they have access to a wealth of basically unskilled labour with low purchasing power. The availability of urban-based activities as part of a household's "employment portfolio" can raise the level of the rural economy by providing knowledge and resources that can be invested in inputs or capital stock for agriculture or for non- farm activities. Ellis and Harris (2004) find from studies in four countries of East and Southern Africa that agricultural land productivity, household per capita incomes, and the proportion of total household income obtained from nonfarm sources are each closely correlated. This means that farm and nonfarm income opportunities are highly complementary and that, where farming has low poten- tial a purely, a localised nonagricultural economy will be unlikely to emerge. If intensified agriculture and diversification into nonfarm production are comple- mentary, they can also be fostered by common conditions--these conditions being the strength of available markets, which means effective access to an urban population. (See Box 3.1.) 25 26 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A BOX 3.1. Urbanisation, Markets, and Rural Development One of the many factors that has discouraged rural development in Sub-Saharan Africa is the presence of thin, inefficient, and isolated markets. Urbanisation and focused urban policies can be critically important in promoting rural develop- ment through improving the efficiency of the markets for transport services and agricultural produce. With low densities of demand and disconnected small markets, it is easy for cartels to develop that can control access to the market and hence control prices. Numerous studies have pointed to very much higher rural transport tariffs (with ratios from 2.5 to 6 times) in Africa as compared to Asia. Overall, it appears that isolated, inefficient, and monopolistic markets are the main explanation. Iso- lated location inevitably encourages inefficiency in the use of vehicles, and it is much more difficult to match demand with supply. As a result, more frequent empty running occurs and a poorer quality of service is provided, with longer time delays as vehicles wait to fill up. Formal and informal cartels operate to control the supply of vehicles at lorry parks all over Africa. The more isolated the location, the easier it is to control access to the market. The bigger the town, especially if rival transport centres emerge, the easier for competition to become established and for efficient operators to negotiate for work directly and achieve higher utilisation. In contrast to Africa, Pakistan, with its very low transport costs, has no centralised lorry parks. A typical town will have ten to twenty freight agents widely spread on the main roads, all competing with each other for business. Agricultural produce markets in Africa are also inefficient compared with those in Asia. Afeikhena and Ogunkola (2000) find that regional prices in food grains can differ by multiples of up to two to three times in Africa--a much greater variation than the ranges observed in the large markets of Asia, and one not explained by transport prices. Large price differences result from poor information, small markets (where one lorry load of produce can dramatically change prices), and the presence of marketing cartels. It is no coincidence that the poorest countries in Africa are among the least urbanised. In order to develop agricul- ture, a strong secure market is necessary. A good, reliable agricultural price will help to provide the financing needed to invest in agriculture. Many of Ethiopia's development problems, for example, result directly from the small size of its urban population (58 million rural dwellers will not get rich trying to compete to sell food to 11 million urban dwellers). Despite the general shortages of food in Ethiopia, it has not been economic in the past to develop the country's substan- tial water resources for irrigation, because urban markets are distant and relatively small. Source: John L. Hine, Transport and Urban Development, World Bank. References Afeikhena, Jerome, and Olawale Ogunkola. 2000. Characteristics and behaviour of African commodity/product markets and market insti- tutions and their consequences for economic growth. Center for International Development (CID) Working Paper No. 35, Harvard University. Hine, J. 1993. Transport and marketing priorities to improve food security in Ghana and the rest of Africa. International Symposium, Regional Food Security and Rural Infrastructure, Giessen, Germany. Hine, J., J.D.N. Riverson, and E.A. Kwakye. 1983. Accessibility, transport costs and food marketing in the Ashanti Region of Ghana. Transport and Road Research Report (TRRL) Report SR 809. Rizet, C., and J. Hine. 1993. A comparison of the costs and productivity of road freight transport in Africa and Pakistan. Transport Reviews 13 (2). Ersado (2003) cites research showing that in Poverty Assessment finds that agricultural spe- Zimbabwe's more remote areas nonfarm income cialisation and intensification, as well as off-farm sources increase income inequality because only activities, are most common in the regions the better off and well-connected farmers around urban centres. By the same token, often can diversify, while in areas better connected to the fastest-growing urban centres occur in the major urban markets nonfarm income regions where agricultural productivity is also decreases income inequality because such op- increasing very rapidly (Montgomery et al. portunities are more widely available. The Benin 2003). W H AT C I T I E S A N D T O W N S C A N D O F O R T H E C O U N T R Y ' S D E V E L O P M E N T A G E N D A -- A N D H O W 27 The most constructive way of looking at the mutual rural and urban gains. Much of Africa is productive interlinkages among urban and These investments also add to the rural areas may be as a virtuous circle. Such vir- effective mobility of the popula- entering the stage in tuous circles consist of access to (urban) markets tion and give individuals the abil- which increased and services for nonfarm production, which con- ity to determine the location of tributes to agricultural productivity and rural their livelihoods. productivity of incomes, which in turn generate demand and One cause for concern in recent towns is needed to labour supply for more such goods and services years is that food imports, fed by (Box 3.2). Most African countries have not man- agricultural subsidies in OECD improve urban aged to break into a cycle of sustainable growth countries and by dramatic im- incomes, thereby by raising agricultural productivity, often consid- provements in efficiency in export- ered the crux of a successful fight against rural ing countries, may break the flow providing alternative poverty. It is therefore important to recognise of the domestic rural-urban link- that multiple entry points into a circle are possi- age in food production and mar- work for the rural ble, and opportunities should be seized wherever keting. The one-third of Africans labour and stimulat- they appear. Individuals, households, and com- who are urbanites are estimated to munities benefit wherever there is ease of access account for about 60 percent of all ing agricultural to markets and diverse economic activities, either marketed food--a share that could investment in inten- through physical proximity (short distance to an increase to three-quarters by 2030, urban centre) or through individual mobility. It is given urbanisation trends.35 While sification through not surprising that poverty is lower and high- imports account for only about value agriculture more common closer to urban growing demand. 7 percent of total food consump- centres or that mobility for work (migration) tion in Africa, foods favoured by starts over relatively short distances (for example, urban dwellers, including dairy, from village to a local urban area.) meat, fish, oils, and wheat, are more likely to be imported, as is often rice because of low prices in National development strategies should, there- global markets (FAOSTAT 2004). Yet fresh fruits fore, build on factors that promote the virtu- and vegetables and many other foods are mainly ous circle of rural and urban development-- supplied domestically--often from very proxi- rather than hinging on hopes for a linear process mate sources in the cities' own periphery. In fact, in which cities will merit investment once agri- urban agriculture accounts for a large share of the culture is safely launched. Tiffen (2003) argues food produced in urban areas--40 percent in that in Africa the domestic (urban) market is becoming the main stimulus to agriculture but that the process is hindered by low productivity BOX 3.2. Rural-Urban Linkages: A Virtuous Circle of the urban sector due to inadequate urban investment and bad policies. Much of Africa is The agricultural chapter of the 2003 WDR argues, "The condi- entering the stage in which increased productiv- tions...to bring the rural poor out of poverty are overwhelm- ity of towns is needed to improve urban incomes, ingly associated with increasing rural-urban interactions and thereby providing alternative work for the rural more intensified use of existing agricultural land. In more labour and stimulating agricultural investment in dense rural areas with towns, credit markets are more apt to intensification through growing demand (Cour exist, and land more likely to qualify for collateral. Higher farm 2003). According to Tiffen (2003), public invest- gate prices associated with better roads and proximity to urban markets, and more opportunities for spreading risk, encourage ment in infrastructure to increase the productiv- higher-input agriculture. This in turn leads to greater value ity and purchasing power of urban centres can produced per unit area and generally to more off-farm jobs. be more favourable to agriculture than are export-oriented policies. Roads connecting rural Source: WDR 2003, Chapter 5 (« Transforming Institutions on Agricul- producers with urban markets and education are tural Land »), p. 88. both highly beneficial investments to foster 28 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A Kampala, for example--and is an important ical proximity has become no less important to source of nutrition especially for the poor.36 Over- firms' productivity. In fact, the growing role of all, despite the competition to domestic agricul- information technologies and of knowledge- ture from cheaper imports, there will remain an intensive production across many sectors seems inexorable and growing demand from the cities to be favouring greater concentration of workers that could be a strong stimulus to domestic farm- and activities using such skills, as opposed to ers who produce and market efficiently. diminishing the value of face-to-face interactions that cities offer (Gasper and Glaeser 1998; Glaeser 1998). At least based on industrial country data, The Role of Urban Centres innovation is associated with the circulation of in Fostering Productivity, ideas and knowledge, particularly in technologi- cally advanced and information-intensive activi- Entrepreneurship, and ties that are favoured by agglomeration (Glaeser Economic Modernisation et al. 1992; Duranton and Puga 2002). Evidence from both theory (including the New Economic A long tradition of scholarship explains the role Geography) and research is highly robust in of cities (or more generally, of urban areas) as revealing that agglomeration is an increasingly contributors to economic growth and social potent phenomenon of "self-reinforcing develop- transformation. The basic theoretical explanation ment" that encourages firms to co-locate in close is rooted in the concept of agglomeration econ- proximity to cities. Firms often remain where omies, which has an even older academic history they can have ready access to markets even in the in industrial location theory. Agglomeration face of diseconomies such as congestion and high economies represent the productivity advantages land and labour prices (Venables 2005). that firms and industries gain by locating in prox- Cities nurture entrepreneurs by providing imity to each other and to large centres of demand (larger markets), information markets. The increasing returns Thus, urban areas and technology sharing, easier access to credit, can be internal to the firm, such as and support for risk-taking (through income function not only in by permitting economies of scale alternatives and through networks of other entre- in production. Externalities may line with traditional preneurs). Most small- to medium-scale enter- also arise in the markets for goods prises, and the informal economy in general, are growth theories by and services, enabling wider access disproportionately located in urban areas for to services, infrastructure, workers, such reasons (Becker 2004). Cities are also key to aggregating larger and buyers, as well as "knowledge economic modernisation and diversification to pools of labour, spillovers", leading to more effi- higher value forms of production. Successful cient exchange of information and competition in global markets requires mo- inputs, and capital: lower transactions costs. Thus, bilising skills, information and technology, and Even more impor- urban areas function not only in capabilities for rapid response in production, line with traditional growth theo- marketing, services, transport, and logistics. tantly, they epito- ries by aggregating larger pools of Enabling such productivity requires attracting mise the process of labour, inputs, and capital: Even and retaining the knowledge resources--highly more importantly, they epitomise skilled and educated workers and an environ- endogenous growth the process of endogenous growth ment rich in amenities that will fosters their whereby productive whereby productive resources are interaction (OPDM 2003). Recent institutional used more efficiently and in new research focusing on the importance of dense resources are used ways. networks of personal exchange and relationships more efficiently and Recent research has suggested that foster innovation confirms the value for that despite the decline in trans- entrepreneurs and educated workers of remain- in new ways. port costs in recent decades, phys- ing spatially connected and the continued attrac- W H AT C I T I E S A N D T O W N S C A N D O F O R T H E C O U N T R Y ' S D E V E L O P M E N T A G E N D A -- A N D H O W 29 tion for them of large cities (Storper and Venables claims that urban areas offer eco- The advantages of 2004). nomic promise and an improved quality of life. Yet it is normal to cities and towns as The advantages of cities and towns as market- expect that there would be poor places for goods, services, and ideas do not arise marketplaces for people living in cities because of entirely laissez-faire, independent of public the attraction of greater job goods, services, and policy. Nothing in the academic theory of ag- opportunity and the availability of glomeration economies guarantees that the mere ideas do not arise services. The key question is physical concentration of people and of firms will whether poverty in cities is part of entirely laissez-faire, result in a well-functioning city, let alone a live- a healthy process of economic able one. As identified by Rosenthal and Strange independent of transition and mobility for the (2001) and summarised in Freire and Polese country and for households--or, public policy. (2003), the microeconomic foundations of rather, a perverse trap that reveals agglomeration economies are characterised by dysfunctional institutions and exclusion prevent- three elements: knowledge spillovers, labour ing individuals from moving forward. market pooling, and input sharing. Each of these The growth of cities favours poverty reduction depends on the ability of cities--and notably, of in several ways. First, it creates a wider and deeper their governments--to create an environment in labour market and facilitates the provision of which: (i) economic agents can meet and com- services. Urban areas both experience concen- municate easily, (ii) workers can move among trated demand and offer the widest options of jobs throughout the area, and (iii) services of supply of jobs and services. Just as importantly, public infrastructure are adequately provided and but less tangibly, the heterogeneity of urban soci- productive inputs can move efficiently. Each of ety contributes to widening horizons and loosen- these conditions assumes, therefore, that a mini- ing traditional attitudes, raising the aspirations of mally effective institutional framework operates disadvantaged groups and challenging the status in the urban area. quo (WDR 2003). Failures of these ingredients undermine the economic performance of cities, especially in Urbanisation supports the demographic transi- Africa. Poorly functioning cities create a drag on tion. Africa has yet to fully experience the demo- where they should give a boost to firms. This is graphic transition that follows urbanisation. true of enterprises attempting only to thrive in Africa's dependency rate37 (89 percent) is well the domestic marketplace, but the failure to realise the benefits of agglomeration is particu- larly damning to firms that would like to export to global markets. There is no reason to believe that cities in Africa cannot contribute as much to the economy as do cities in other regions. But Africa's cities require a commitment to ensuring that their urban assets are better managed to realise stronger gains. se How Urban Development Pictur Can Be Good for TE/Still Poverty Reduction SCHYT To some observers, the evidence of urban poverty GENROJ and deprivation, often most visible in slums, belie A local market in Maputo, Mozambique. 30 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A above that of any other region, and its share of settings. As noted earlier, urban and rural aver- children is extremely high (44 percent). For age data on access to services hide the consider- Africa, the dependency rate is not projected to fall able disparities within the urban population. below 50 percent until 2050 (WDR 2003). Unmet urban demand remains for essential serv- Urban areas witness earlier and steeper ices, especially for the poor and for residents of declines in mortality and fertility than do the rest informal settlements (slums and peri-urban of a country, for many reasons--including higher areas). On the supply side, input, output, and effective demand for smaller families, more adop- outcome measures of provision are easier to tion of family planning, and generally better achieve in cities than in less densely settled areas. access to health services.38 Urban age structures The per capita costs of many forms of infrastruc- are typically more concentrated in the working- ture and social service are generally lower, and age and reproductive ages than are those of the many more people can be reached in urban areas total population. This implies a one-time demo- even if the technical level of services must be graphic (and economic) bonus, largely an urban higher than in rural areas (as in the case of sanita- phenomenon. (Montgomery et al. 2003). tion). It is also easier to attract and retain per- The implications of the projected demograph- sonnel,39 obtain spare parts and materials, and ics for the urban economy are clear. The large con- thus realise the desired operating efficiency and tingent of youth creates pressure for schooling quality of service, relative to circumstances in and jobs and can also be socially destabilizing by more remote locations. Competition among contributing to crime and violence, while the alternative service options can be more readily dearth of middle-aged adults (reflecting in part provided, as a force for innovation and efficiency, the impacts of HIV/AIDS) implies loss of experi- when the "market" for services is larger. ence in the workforce and in families. Yet the experience of Asian and other countries shows Demand-side factors in urban areas also facili- that the decline in dependency rates and increase tate the provision of services and attainment of in the working-age group permits more invest- goals such as the MDGs. In the urban setting, it ment in education per child and higher savings is easier to share information and to witness rates (Montgomery et al. 2003). examples of changed behaviours. Individuals can The per capita costs The demographic bonus can be appreciate the opportunity cost of inadequate used to particular advantage in education, lack of family planning, and the health of many forms of urban areas, where education is impacts of poor sanitation in their neighbour- easier to provide and investments hoods. Literacy can be reinforced with ready infrastructure and can produce diverse employment availability of written material, and publicity social service are opportunities for the new workers. campaigns have an immediate audience. All this But this requires the urban econ- is not to say that the urban context is without generally lower, and omies to absorb the growing complications and negative pressures. Organising many more people workforce productively, without collective action among a highly heterogeneous suffering increased unemployment population can be especially challenging. Densely can be reached in and depressing wages. So while built-up cities require more sophisticated con- urban areas even if urban areas generally favour the struction and coordination of networked infra- demographic transition, nothing structure to ensure that services function well the technical level of guarantees they will deliver such and are environmentally sound. Inequalities of services must be advantages in the absence of sound opportunity become highly visible and poten- management and adequate public tially destabilising. higher than in rural investment in the urban economy. Urban settings also entail some of their own areas (as in the case health risks. One of the most insidious threats to Providing services, and meeting health and well-being, HIV/AIDS, is more preva- of sanitation). the MDGs, can be easier in urban lent in urban than in rural areas and is generally W H AT C I T I E S A N D T O W N S C A N D O F O R T H E C O U N T R Y ' S D E V E L O P M E N T A G E N D A -- A N D H O W 31 highest in large cities, as well as in informal settle- cally significant.41 No greater disadvantage was ments. Risk factors include both the extent and found when the migrants from rural areas were nature of social contacts, including density and singled out.42 The observation that migrants in high mobility in urban settings, as well as the urban areas are not significantly poorer or less presence and proliferation of vulnerable groups, well situated than incumbent residents reflects such as street children and sex workers (Boerma, both self-selection among the individuals who Nunn, and Whitworth1999; Kelly 2003). migrate--they tend to be better educated than their peers at home--and their ability to apply Migration and remittances provide economic their human capital and initiative in the receiving options and mitigate vulnerability. Evidence economy, which is precisely the attraction of from many household surveys shows that urban areas.43 dependence on agriculture as the sole or major Other studies and data for Africa suggest sim- source of livelihood is associated with poverty ilar positive results. A poverty study of Burkina (World Bank 2005d). One of the main contribu- Faso found no evidence that recent immigrants tions made by the urban economy is the provi- into urban areas are unemployed in greater pro- sion of alternative sources of income, including portions than are other residents (Grimm and through migration and remittances, even though Gunther 2004). Household surveys in Benin re- the poorest rural households may be the least veal that breadwinners who migrate for employ- mobile and least able to benefit directly. ment achieve higher consumption levels for their Much of the discourse about development households than do others with similar charac- either ignores or discourages internal migration, teristics (for example, education level) who do assuming that it has largely negative impacts to not migrate. This was especially true for bread- the individual migrants or to the receiving or winners who migrated to urban areas; their sending areas.40 The DHS data indicate that, con- households' consumption increased by 14 per- trary to common assumptions, recent migrant cent versus 7 percent in the case of migrants to respondents were only slightly more likely to be another rural area. The Benin Poverty Assess- defined as poor or to suffer less access to serv- ment observes that "it pays to take risks and ices, and the differences were often not statisti- migrate when looking for work" and that migra- Bank ldro fman/W Huf windE Children collecting water from a community water pump, Lusaka, Zambia. 32 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A BOX 3.3. Urban Migrants Have Done Well in Uganda Preliminary analysis of the 2002­2003 household survey results for Uganda (see table) indicates that the majority of urban household heads are migrants, especially among the richer households. Only 14.3 percent of the top urban expenditure quintile represents non-migrants, versus 35.8 percent for the poorest urban quintile. The rural sample contains a larger share of non-migrants in all quintiles, as would be expected, but again, the richest households are less likely to be non-migrants (40.5 per- cent) than are the poorest ones (64.2 percent). Whether rural migrants came from other rural areas or are urban returnees is not specified. The evidence is thus pretty strong that migration is associated with urban residence but also, more importantly, with higher expenditure status. Uganda: Urban and Rural Comparisons of Migration, by expenditure quintile Percent Q1 Q2 Q3 Q4 Q5 All Household head Non-Migrant, Rural 64.2 61.0 57.9 56.1 40.5 56.1 Household head Non-Migrant, Urban 35.8 31.0 26.0 21.1 14.3 18.1 Number of Times Migrated: Rural ­ Never 64.2 61.0 57.9 56.1 40.5 56.1 ­ Once 28.2 27.1 31.1 31.3 42.5 31.9 ­ Two or more times 7.6 11.8 11.1 12.6 17.0 11.9 Urban ­ Never 35.8 31.0 26.0 21.1 14.3 18.1 ­ Once 40.4 50.9 51.8 56.2 52.9 53.0 ­ Two or more times 23.8 18.1 22.2 22.7 32.8 28.9 Reasons for Migration: Rural ­ Looking for work 8.6 9.5 16.7 20.2 34.7 19.4 ­ Other economic reasons 30.1 34.4 32.1 31.7 27.8 30.5 ­ Education 1.3 1.7 2.0 1.1 2.4 1.8 ­ Marriage 31.8 26.8 23.0 21.4 11.2 21.8 ­ Escape insecurity 5.8 5.9 4.8 4.2 2.7 4.5 ­ Combination reasons 22.3 24.6 21.3 21.4 21.3 22.1 Urban ­ Looking for work 11.4 35.0 35.6 43.2 35.9 36.7 ­ Other economic reasons 26.1 31.1 21.6 18.0 21.7 21.5 ­ Education 1.1 1.6 6.1 4.5 4.4 4.3 ­ Marriage 18.2 9.1 13.1 8.8 6.6 7.9 ­ Escape insecurity 7.8 3.7 3.8 3.4 0.9 1.9 ­ Combination reasons 35.3 19.5 19.9 22.0 30.3 27.6 Source: Q1 = poorest, Q5 = richest. Uganda National Household Survey, 2002­2003, Courtesy: Louise Fox, Africa Region, World Bank. The same table shows that the richer households migrate more often than do the poorer ones, both in the rural and the urban samples. Also, in both areas, the richer households more often report "looking for work" as their reason than do the poorer ones; and across the urban quintiles, all urban migrants report this as a more important reason than do the rural respondents. The rural migrants more often list other economic motivations, including escaping insecurity, and social aims (marriage). In short, for Uganda at least, urban residents appear to be drawn to migrant status for employ- ment and other economic interests, while rural residents may be responding more to duress and limited social options at home. These data suggest in general why policy makers who are determined to curb rural-to-urban migration have a hard time achieving this result. Urban migrants want urban employment and appear on the whole to achieve relative economic success, while the rural popula- tion seeks migration for both economic and social values. W H AT C I T I E S A N D T O W N S C A N D O F O R T H E C O U N T R Y ' S D E V E L O P M E N T A G E N D A -- A N D H O W 33 tion has proven to be one of the most effective fifth of rural household incomes in While the magni- coping strategies both as a short-term (for exam- Zimbabwe in 1990 and 1995, with ple, dry season for farmers) and as a permanent only slight reduction after an eco- tude of remittances, solution to poverty for many households (World nomic crisis (Ersado 2003). While especially from Bank 2003c). In Niger, it is estimated that each the magnitude of remittances, year half of the male household heads between especially from domestic locations, domestic locations, the ages of 20 and 29 seek work in urban areas, is not easily compared with other is not easily com- at least temporarily, after the harvest (World financial flows, it is likely that these Bank 2004c). shares of income well exceed other pared with other special resources, such as micro- financial flows, it is Remittances by migrants are often found to be credit or development assistance, very important both to senders and to receivers, available to households. likely that these although data are typically lacking to identify A recent poverty study of shares of income remittance flows of domestic urban origin.44 Ghana found that a major con- A survey of the informal sector firms in Karu tributor to poverty reduction well exceed other (a large settlement on the periphery of Abuja in throughout the 1990s has been special resources, Nigeria), where 80 percent of the inhabitants are migration from slower to faster migrants from other regions, found them rela- growing regions. The biggest such as microcredit tively prosperous as compared to the rest of the reduction in poverty was identified country. Those with a longer length of stay in among the rural forest residents, or development Karu had higher incomes, and new migrants reg- attributed to their receipt of remit- assistance, available istered more complaints about their life there-- tances. Nationally, remittances indicating, not surprisingly, that migrants start (both of domestic and interna- to households. out less well-off than the established residents but tional origin) have increased from manage to improve their situation over time. 14 percent to over 20 percent of household Interestingly, no correlation was found between incomes between 1992 and 1998 (McKay and length of stay and remittances home, with most Aryeetey 2004). remittances (61 percent) made by households in The people of Africa have fewer options for residence from one to five years. Households migration abroad (outside the continent) than engaged in informal sector activities sent more did poor populations in earlier periods.46 It is money home than did those working in the for- therefore especially critical to the welfare of the mal sector. This study confirmed that the direct populations in Africa's rural areas (and especially economic contributions of these migrants those on fragile lands) that domestic migration reached 24 of the 36 states of the country (CAS- remain unrestricted, so that individuals can make SAD 2002). their own location decisions. This means avoid- Another study of several African countries ing the kinds of policies effected in the past found that nonfarm income, including remit- by socialist countries such as China and Viet- tances from urban-based family members, was nam, where restrictions on internal migration especially significant as a share of income of the have created an underclass of illegal migrants better-off rural households and helped fund or "floating workers" denied housing and serv- investments back on the farm (Tiffen 2003). In ices. In Africa, Ethiopia has also practiced such Lesotho, which may be considered an extreme restrictions, and Nigeria most recently has case due to its traditional reliance on migrant espoused official aims of reducing rural-to-urban labour to South Africa, cash remittances in the migration. year 1994­1995 amounted to about one-quarter Migration may add to urban-rural inequality of total incomes across the board, only slightly in cases where the migrants and the households less for the very poor than the non-poor.45 Remit- receiving remittances are among the better-off in tances and other transfers comprised about one- their communities. The loss of the more produc- 34 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A It is therefore espe- tive members could also weaken Urban Development the economy of origin. Inequality, cially critical to the as an Exercise in however, should not be a domi- welfare of the popu- nant concern in the face of over- Local Governance whelming evidence of benefits to lations in Africa's and Institution-Building individuals and households from rural areas (and income diversification and risk It may be no coincidence that many of the devel- mitigation (Ellis and Harris 2004). oping countries facing rapid urbanisation, and especially those on Migration or mobility is clearly a those with already large urban populations, have fragile lands) that favourable element in income experienced pressures to grant more political growth and poverty reduction, autonomy to local governments and give greater domestic migration both for urban and rural popula- voice to citizens. Regardless of whether a coun- try has officially decentralised, running a city remain unrestricted, tions. Policies should enable labour mobility as part of general wel- showcases local governance because the link- so that individuals fare and poverty reduction strate- ages--for better or worse--among public expen- gies. At the same time, migration diture, local public goods, quality of services, and can make their own can pose a challenge to the re- quality of life are hard to miss on the street. location decisions. ceiving areas by adding to de- The growth of urban economies generates a mands for services in the near buoyant revenue base that can fund many of the term, raising the stakes for good city's own needs and (in the case of largest cities) urban management. provide large net revenue gains for the nation's Bank ldro W/k nemar CartruC Community construction efforts in Dakar, Senegal. W H AT C I T I E S A N D T O W N S C A N D O F O R T H E C O U N T R Y ' S D E V E L O P M E N T A G E N D A -- A N D H O W 35 coffers. But the collection and management of BOX 3.4. Sustainable Growth Draws on local taxation is sorely neglected in many coun- Phenomena Requiring the Urban Transition tries, especially in Africa. Supporting local rev- and Well-Working Cities enue strength, on the basis of sound fiscal transfer arrangements and encouraging local creditworthiness as far as a city's economy allows The report, Economic Growth in the 1990s (World Bank 2005a) describes the conditions for sustain- should be primary concerns of African govern- able growth as requiring "more than efficient use ments as the urban transition accelerates. of resources. Growth entails structural transfor- mation, diversification of production, change, risk Conclusion: Good urban management poses taking by producers, correction of both govern- numerous synergies with the prevailing na- ment and market failures, and changes in policies tional development goals of sustainable growth, and institutions. It is also a process of social transformation: people will change activities and poverty reduction, and good governance. Yet, it live in different places. Social relations will is relatively rare for urban development to figure change, and the informal networks of rural life prominently and explicitly in national strategies, will be lost as other more formal networks and official or donor. A notable exception proving the organisations are established. Entrepreneurs will rule is China's Tenth Five-Year Plan in 2001, invest . . . to produce new products and adopt new organisational forms. Farmers will adopt new which advocates increased urbanisation to stimu- farming methods and change their product mix. late rural and national economic development The economy will produce and demand different (New Star Publishers 2001). In Africa, it is time to goods and services. . . . Any growth strategy needs acknowledge the potential and the problems of to include actions, both on the policy and institu- the cities as one of the central elements of the tional front, that address and support this process national development agenda. Neglect of the of change." (Overview, 11). While growth of towns and cities is an cities may represent one of the least-considered inevitable accompaniment to this process, ensur- "binding constraints" in African economies. The ing that the urban areas are well-managed to next chapter details some of the key ingredients maximise benefits for the economy and society is to make better-functioning urban areas in the not inevitable and requires concerted attention. region. TA Mbau/UN-HABIT Susan Unblocking the economic potential of cities. Making African Cities 4 More Effective in Promoting Economic Growth and Poverty Reduction Neglect of African he evidence reviewed here confirms the ability of the urban economy to nur- cities does not remove T ture activities with strong growth prospects and to improve incomes and the impetus for their quality of life in the process. But what is increasingly seen in African cities is a physical concentration of people and activities that do not benefit from the key expansion, because "ingredients" that both theory and experience have led us reasonably to expect they remain the most of more effective cities. The key ingredients that are weak or relatively absent include: (i) basic flexibility of the factor markets--because of barriers to workers' efficient setting in job and residential mobility and rigidities in land use; (ii) efficient local public services, both those that can be produced through public-private partnerships and which entrepreneurs those that are pure local public goods; and (iii) trust and confidence in govern- can operate, relative to ment--the lack of which deters the private sector and households from investing and partnering for the future and shifts burdens to them that should be borne or the alternatives. reduced by the public sector. If these are not the prevailing conditions in many African cities, the concentra- tion of people and economic activity is less productive than it should be. Rather than devoting more attention to documenting or defending the urban contribu- tion to development, real energy needs to be spent unblocking it. Listening to the Firms Surveys conducted in recent years, mostly on formal (registered) manufacturing enterprises located in urban areas, attempt to gain the perspectives of firms on the ease or difficulty of working in different settings. Such investment climate (IC) surveys were reviewed for this study in six countries in Africa (Ethiopia, Kenya, Mozambique, Tanzania, Uganda, and Zambia) and for several other developing countries in East and South Asia (Cambodia, China, Bangladesh, and Pakistan) as comparators. The factors chosen for analysis are those that could shed light on the business environment in the various cities and, to the extent possible, policy implications for the local government. Overall, the main observations from this IC review for African countries can be briefly summarised as follows (see the Appendix for more detail). Within infra- 37 38 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A structure category, lack of reliable electricity is by responsibility of national authorities and there- far the biggest obstacle cited for all of the African fore are not "urban" or "local government" issues. firms. Many respondents in Nairobi and Dar es Nevertheless, they are significant in revealing that Salaam compensate with their own generators, some of the basic assumed advantages of urban yet 10 percent of output is reported lost due to agglomeration--access to inputs and services power outages in both cities. Undeveloped affording economies of scale and connectivity to telecommunications was reported to be a much other producers--are simply not being realised in less serious obstacle than electricity. Yet, firms African cities, even in those that might be con- in all urban areas in Kenya and those outside sidered relatively favoured, such as the capitals. Kampala in Uganda reported losing telephone Therefore, a firm locating in such a city gains a connections for more than 30 hours at a time, on much lower productivity advantage than would average. Poor transportation was reported to be its counterpart elsewhere, all else being constant an obstacle greater than, or at least on par with, (such as city size, natural resource base, labour telecommunications. Whereas problems in elec- supply, and so on). tricity and telecommunications were felt more The last infrastructure variable surveyed, in the capital or primate cities, transportation access to land, was least cited as a moderate/ issues were more pervasive and actually rated major constraint in the African sample, although much worse in the secondary city samples in it is a big issue in Ethiopia, with a 65 to 70 percent Mozambique, Tanzania, and Zambia. Problems rating (Figure 4.1). Land acquisition may be rated with access to skilled labour were rated about less badly in such surveys because, unlike other equal to transport by the African firms. factors and services, it is not a frequent expendi- These findings concern the infrastructure sec- ture. Reported land acquisition delays are very tors and the labour market that are largely the long in Ethiopia and in Zambia, where produc- ers in the capital cities have complained of wait- ing as long as 120 days to acquire plots. Steep processing fees are required in Mozambique and to use land as collateral in Nigeria. A separate sur- vey of investors' views in Senegal heard com- plaints that improvements in tax administration, business registration, and customs were over- whelmed by the lengthening of time required to obtain land. Among the institutional constraints in the IC surveys, corruption dominates the other issues as a major concern for all the African countries (except Ethiopia), as well as for the Asian (except China). Generally, corruption is felt to be a bigger problem in the capital or primate cities, indi- cating that it may pertain especially to national se government officials. Tax administration is the second reported constraint for most of the coun- Pictur tries, with relatively little variation across them or across the two types of cities. The lack of varia- WARDS/Still ED tion between the city groups may signify that national taxation is the main issue. There is a very MARK Congestion of people and services in Abidjan, Cote d'Ivoire. large inter-country variation in the reported con- M A K I N G A F R I C A N C I T I E S M O R E E F F E C T I V E I N P R O M O T I N G E C O N O M I C G R O W T H A N D P O V E R T Y R E D U C T I O N 39 cern with crime, theft, and disorder. Surprisingly, firms in capital or pri- FIGURE 4.1. Access to Land as a Constraint in Urban Centres mate cities do not claim to suffer this problem more than do those in Bangladesh secondary cities. Business licensing Dhaka received the fewest negative ratings. Other cities In sum, for most of the African Pakistan Islamabad & Karachi countries institutional and infra- Other cities structure issues are about equal as Cambodia areas of concern. Phnom Penh Surveys specifically of informal Other cities sector firms in South Africa (World China Bank 2002c) and in Tanzania (Tan- Beijing & Shanghai zania Planning Commission 1991) Other cities produced some differences in views Ethiopia Addis Ababa and priorities relative to the formal Other cities enterprises. In both countries, the Kenya two biggest constraints to informal Nairobi business were reported to be, first, a Other cities lack of credit or capital and, second, Mozambique lack of demand (effective market for Maputo their output). Problems of access to Other cities equipment, high costs of public Tanzania Dar es Salaam transport and infrastructure services Other cities (utilities), and difficulties in finding Uganda business premises also arose among South West* the top constraints.47 The informal Other cities firms might, therefore, be considered Zambia Moderate somewhat more sensitive to issues Lusaka Major rooted in the local economy or local Other cities government functions than are the 0 10 20 30 40 50 60 70 80 90 100 formal companies. Percent of firms Neglect of the African cities does not remove the impetus for their * Includes Kampala Source: Chakraborty (2005). expansion, because they remain the most efficient setting in which entrepreneurs can operate, relative to the alternatives. Large firms in Lagos report Despite these diseconomies rooted in poor urban that the main advantages of this location are the management, only 11 percent of the respondents much better availability of infrastructure than said they would prefer to relocate elsewhere in elsewhere, proximity to markets for inputs and Nigeria (Tewari and Banerjee 2005). outputs, and availability of space for expansion. Overall, the picture that emerges of urban The same firms reported their main constraints enterprises in Africa suggests that they are ham- as the decaying state of this infrastructure and the strung by problems rooted in both national eco- costs of obtaining premises for expansion, as well nomic management and in city management. as general congestion and security concerns. The advantages that enterprises gain from an 40 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A What Cities Require FIGURE 4.2. Crime, Theft, and Disorder as a Constraint in to Be Productive Urban Centres Individuals and firms will continue Bangladesh to build for themselves and provide Dhaka their own services even in the Other cities absence of a functioning local gov- Pakistan ernment. In fact, the vast majority of Islamabad & Karachi constructed capital in African cities Other cities represents the self-provision of shel- Cambodia Phnom Penh ter and shops. But clearly much Other cities greater benefits could be mobilised China for the country and for the urban Beijing & Shanghai inhabitants by providing comple- Other cities mentary local public goods in a safe Ethiopia and secure environment for the Addis Ababa urban economy. This implies focus- Other cities ing on basic investment and efficient Kenya functioning of the essential core of Nairobi land and housing, environmental Other cities services, public transport, and local Mozambique Maputo public finance--typically the miss- Other cities ing ingredients and weak links of the Tanzania urban economies in Africa. Dar es Salaam Other cities Land and housing. One of the most Uganda fundamental characteristics and South West* Moderate advantages of an urban economy is Other cities Major that it is a large and diverse market- Zambia place. But problems in the availabil- Lusaka ity (and affordability) of land for Other cities firms and for housing, and con- 0 10 20 30 40 50 60 70 80 90 100 straints in transport that reduce the Percent of firms effective mobility of goods and workers, can fracture the city into * Includes Kampala Source: Chakraborty (2005). disconnected subzones that become dead ends, especially for the poor. Investment in housing typically accounts for 2 to 8 percent of GDP, and the flow of housing services accounts for another 7 to urban location can be swept away by poor 18 percent of GDP in most countries. In the national economic/institutional frameworks. By United States, the housing construction sector the same token, it may be expected that a city alone accounts for an estimated 5 percent of the attractive for business is likely to be located in a economy and over 10 percent of economic similarly hospitable country. To enhance private growth. It has been estimated that a better func- sector enterprise and firm competitiveness, both tioning housing market could add 1 to 2 percent national and city leaders must focus on this as a to India's growth rate (McKinsey Global Insti- common goal. tute 2001). In Africa, housing is produced over- M A K I N G A F R I C A N C I T I E S M O R E E F F E C T I V E I N P R O M O T I N G E C O N O M I C G R O W T H A N D P O V E R T Y R E D U C T I O N 41 whelmingly by the informal economy but is no Experience in Africa and many The most efficient less important as a source of employment and other regions confirms that up- household income (through rental). Demand for grading existing slums by pro- solution for rapidly housing and business real estate is likely to viding basic infrastructure and growing urban remain income-elastic, and so ensuring a healthy communal services, with confir- housing and land market (and the related con- mation of existing tenure security areas, however, even struction sector) should be a key element of any (but not necessarily formal titles), cheaper on a per- development and growth strategy in the region. is much less disruptive or expen- Not much detailed study has been carried out sive per household than trying household basis on the economic costs of land use patterns in to relocate residents or directing than in-situ upgrad- African countries. Suggestive, however, is an them into public housing. The analysis of certain zoning regulations in Banga- most efficient solution for rapidly ing, is to block out lore, India (the floor-to-area ratio, a restriction growing urban areas, however, areas for new settle- on building heights that deters intensive develop- even cheaper on a per-household ment in central city locations), that found that basis than in-situ upgrading, is to ment in advance the welfare costs could amount to 3 to 6 percent block out areas for new settlement and to install very of household consumption, due to the impact on in advance and to install very basic increased housing prices and increased commut- sanitation and street layouts, en- basic sanitation ing costs (Bertaud and Brueckner 2004). Ethiopia abling progressive construction as is possibly an extreme case of land and housing the neighbourhood grows (The and street layouts, market dysfunction in Africa that illustrates the Millennium Project 2005b). enabling progressive impacts of socialist land use policies. South The lack of mortgage financing African cities reflect the legacy of another type and of microcredit for progressive construction as the of extreme, apartheid (Bertaud 2001; Bertaud housing development, and disin- neighbourhood and Renaud 1995). Although land in Ethiopia is centives for rental supply, make no longer restricted to private ownership, the housing expensive, even for the grows current policy of preventing it from being traded middle class. Ghana's Housing independently excludes most of the population, Finance Company is one case in and currently only the top 20 percent of the Africa where the beginnings of a mortgage sys- income distribution can afford plots. Much of the tem have been established, with initial World foreign remittances (which in total amount to Bank support. Many problems remain, however, almost 7 percent of GDP) are spent on real estate, rooted to a large degree in the supply constraints so distortions in the land market lead to wasteful of the land market. uses of these resources (Bertaud et al. 2004). Elsewhere in Africa, as in many other devel- Public transport. Along with limited effective oping countries, the common reality is for the choices in location, urban firms and residents public sector to dominate the ownership and use do not have good options for mobility. Failures of urban land. At the same time, governments in urban transport policy seriously compromise inadequately exercise their necessary role in pro- the movement of individuals as well as the cir- tecting rights of way and preventing sensitive culation of goods, again shuttering the urban areas from being settled. Authorities overly marketplace. restrict on-plot development and generally An ongoing analysis of the affordability of under-provide the infrastructure that would val- public transport fares in a global sample of orise land. Misregulation contributes to very low mainly large, non-African cities indicates that the density sprawl, as new settlement is only afford- poor are often simply priced out of public trans- able at the periphery--further raising the costs of port, which could cost a third of their income for infrastructure networks and complicating public normal use. A transport and poverty impact bus services. assessment for Lagos found that the average bus 42 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A fare for normal use would require over 50 percent Habitat Global Urban Observatory, 1998 data). of the household income of the poor, although The largest cities should be able to afford better drivers may charge less or more than posted fares. management of roads and public transport sys- These figures compare to an affordability bench- tems than can other urban areas, even with mark of 15 percent (Carruthers et al.2004). higher demand. Such figures testify to the policy A study of three African cities reports that in neglect that leaves the upper middle class to Addis Ababa (a city of 3.6 million people), untrammelled use of motorised vehicles and 70 percent of trips are made by walking, and pub- everyone else to walking. lic transport is estimated to cost 3 to 37 percent of household income. The average distance walked Sustainable municipal finances. Urbanisation per journey is 5 kms. In Nairobi, the average is tends to raise the profile of local government 4 kms (where 48 percent of trips are by non- functions. It is therefore critical that, as cities motorised transport, including walking). In Dar es grow, the professionalism and accountability of Salaam, the average figures are 2.2 kms and local government mature as well. This obviously 45 percent, respectively. Roads are poorly designed does not happen automatically. But with system- for walkers and extremely unsafe--including for atic support to improved practices and proce- cars, especially given how they are managed. In dures, and in an environment where local and Nairobi, only 3 of 22 signalised intersections were national authorities commit to making local gov- working at the time of the study, and in Addis, only ernment more responsive to citizens, municipal 52 of 359 traffic policemen had motorcycles performance has improved remarkably. (SSATP 2002). In Benin, for example, the recently completed Most comparative city data on travel-to-work Decentralised City Management project fostered times, regardless of country income level, indi- dramatic increases in municipal resource mobil- cate that about 30 minutes is the mode and aver- isation in the three largest cities (Table 4.1). Sim- age. UN-Habitat city data, however, reveal that ilarly, in Senegal municipal adjustment and in Africa the capital or primate cities have much investment programmes introduced by the Bank- higher travel times--in the range of 45 to 60 min- supported first Urban Development and Decen- utes, sometimes more than double the figures tralisation Programme (closed December 2004) for secondary cities in the same country (UN- achieved increases in municipal savings and TABLE 4.1. Concerted Efforts Raised Municipal Financial Performance in Benin's Three Largest Cities, 1999­2004 Cotonou Porto-Novo Parakou Total municipal revenue growth over period 82% 148% 131% Share of city's recurrent cost budget spent from 65 to 81% from 39 to 66% from 52 to 72% on local public services--increase over period Share of city's recurrent cost budget spent from 22 to 23 % from 9 to 23 % from 1 to 43% on maintenance of roads and drainage-- increase over period Collection rate for billed local taxes 67% (target 75%) 48% (target 40%) 72% (target 65%) (property and professional taxes)-- by 2004 Source: Republic of Benin, Decentralised City Management Project, implementation completion data. M A K I N G A F R I C A N C I T I E S M O R E E F F E C T I V E I N P R O M O T I N G E C O N O M I C G R O W T H A N D P O V E R T Y R E D U C T I O N 43 rationalised the allocation of current revenues to mental relationships) and from Many of the con- much needed maintenance. All 67 municipalities below (through accountability to under the Senegal programme have signed the local population), leadership straints facing cities municipal contracts that commit both the groups in cities can identify the suggest corrective national and local governments to improved obstacles to better performance-- performance within a simple framework. This for example, through a city devel- actions that the city framework aims to help the local governments opment strategy process or a leaders can under- prioritise and plan service delivery; formulate, management contract agreed to implement, and monitor budgets; mobilise fiscal with central government. Cur- take, provided they and nonfiscal revenues; and become accountable rently, a few instances of such have a commitment to communities. municipal activism are taking In eleven African countries, a system of street place in Africa, and these experi- to installing capable addressing (or addressage) has been established in ences must be shared with other professionals in the the major municipalities. Addressing represents a local governments. practical and low-cost information system for local government local governments and utilities to identify land and to involving uses and residences, providing data that can also Yet, many issues for cities are help improve property taxation and service pro- rooted in national and even inter- stakeholders among vision. Introducing simple systems of street national policies and institutions. addressing in two cities in Senegal--Thies and Senegal, for example, has a history local businesses and Kaolack--increased municipal tax billing by of relatively good urban man- the community. about 50 percent, with 90 percent collection rates. agement, yet the entrepreneurial In Burkina Faso, Togo, and Mauritania, street sector remains handicapped by ha- addressing has aided in inventorying the local tax bitual practices that discourage competition and base and in implementing residence taxation. favour rent-seeking, despite efforts of macro- Guinea is using the system to organise solid waste economic reform (World Bank 2003e). Even in collection and secondary street maintenance these cases, an empowered and informed city (Farvacque-Vitkovic et al. 2005). leadership can advocate appropriate actions by Many of the constraints facing cities suggest national governments and network with other corrective actions that the city leaders can under- cities to collaborate in pursuing common con- take, provided they have a commitment to cerns, for the benefit of their residents and of the installing capable professionals in the local gov- country at large. A very notable example is the ernment and to involving stakeholders among South African Cities Network, which is engaging local businesses and the community. Empowered in national dialogues on policies affecting the from above (through supportive intergovern- member cities and the society. City putoa M The opportunity costs of neglecting infrastructure in Maputo, Mozambique. The Opportunity Cost of 5 Neglecting the Cities What Is at Stake for National Development in Africa Neglecting the cities he failure to satisfy the basic conditions for effective cities will, in simplest makes both firms and T terms, dilute the benefits that could be gained for the country. In some households more African countries, the issue presents itself more starkly: What is the difference between a functioning city or town--and a refugee camp or an enclave such as vulnerable to the dis- an industrial zone or gated residential community? economies of urban Neglecting the cities also makes both firms and households more vulnerable to the diseconomies of urban agglomeration--high costs of land, congestion and agglomeration--high inadequate mobility within the city, a polluted environment, threats to the social costs of land, conges- order and to public health, and crime. These risks, while never entirely avoidable with population concentration, become greater and prematurely imposed by tion and inadequate very inadequate urban management. The projected growth of urban popula- mobility within the tions over the next decades, and the burgeoning cohort of children and youth in particular, underscore the challenges. city, a polluted envi- Higher costs and reduced competitiveness of firms. The impact of inadequate ronment, threats to infrastructure and high transactions costs on reducing the competitiveness of the social order and to firms has been widely documented in many countries (WDR 2004). Africa, whose share of world trade remains miniscule and is declining (from 1.9 percent public health, and of all merchandise exports in 1990 to 1.5 percent in 2003), cannot afford a com- crime. parative disadvantage from failure to nurture the business environment of cities and from the waste of their productive resources. Lagos, for example, like many other major cities of the region, hosts half of Nigeria's invested manufacturing capital and 60 percent of its non-oil economy. While the small scale and infor- mal enterprise sector often appears quite dynamic in Nigeria and other coun- tries of the region, the predominance of such enterprises is a concern where few signs show that they are maturing in size or interacting with larger or formal firms companies.48 The cities should nurture not just the proliferation but also the maturing of these firms. The almost complete reliance of some large communi- ties in major cities (such as Karu in Abuja) on informal self-production and trade within a narrow radius is a sign of an urban economy that is insufficiently inte- grated with domestic and regional markets, let alone globally. 45 46 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A infant and child Worsening risks to public health mortality risks. Particularly in Africa, urban poor and security. Health status pro- children are found in some surveys to be less mortality and other vides a good overall reading of healthy than their rural counterparts.49 A com- health indicators risks to welfare, both stemming parison of DHS results in 22 African countries, from issues with specific services for example, found that diarrhoeal point preva- such as child growth (health care or water supply) and lence was higher on average for the poorest quin- measures have from environmental or social risk tile in urban areas than for the poorest quintile factors related to location. The in the rural sample (28 to 24 percent), and this remained signifi- DHS and other studies that disag- relationship appeared for half of the countries cantly better on gregate by urban and rural areas (Doumani 2002). generally confirm an urban advan- Comparisons across intra-urban localities average in urban tage in health outcomes. That is, often present a stark picture. Table 5.1 documents infant and child mortality and large disparities in basic health as measured by than in rural areas, other health indicators such as infant and child mortality and diarrhoea inci- for a number of child growth measures have dence in Kenya. The data reveal that children in remained significantly better on the slums of Nairobi face enormously higher reasons including average in urban than in rural health risks than do their peers living elsewhere better public infra- areas, for a number of reasons in the same cities or in rural settlements. including better public infrastruc- The 2001 to 2002 Zambia DHS found that structure, higher ture, higher maternal education, slum children in Lusaka were more likely to have maternal education, and better access to health care in suffered diarrhoea or a serious cough in the pre- urban areas in general. Poor urban vious two weeks than either non-slum urban or and better access to children are found to be much less rural children, although action had been taken health care in urban healthy than non-poor urban chil- (healthcare sought) as often or more often for the dren (Figure 5.1), however, and to slum children than for these other cohorts areas in general. face much greater disease and (Mboup 2004). Apparent proximity to health centres does not always imply effec- tive access or affordability of health care for the urban poor, however FIGURE 5.1. Urban Poor Children Face Higher Disease Risks than (Figure 5.2). In Ghana, the 2003 their Rural Peers Core Welfare Indicators Question- Morbidity inequalities naire (CWIQ) found a worsening of Acute respiratory infections among children health indicators, including under- 25 weight status, among the urban Urban richest poor relative to the rural poor, even ne Urban poorest 20 Rural average though the urban children had drilhcfo higher participation in health pro- 15 grammes--indicating that other egatnecreP risks such as poor sanitation or food 10 insecurity may be at play. 5 Negative health externalities often affect spatially concentrated 0 areas of disadvantage, such as slums. Cameroon Guinea Gabon Niger Ghana Kenya These are urban health penalties, in Source: Mboup (2003) the sense of risks more highly asso- ciated with urban than with rural T H E O P P O R T U N I T Y C O S T O F N E G L E C T I N G T H E C I T I E S 47 living. Such risks go beyond those of infectious diseases that can be reduced by public infrastruc- TABLE 5.1. Health Status of Children in Kenyan Slums ture and include stress-related diseases and oth- Infant and Under-Five Mortality Rates and Diarrhoea Prevalence in Kenya ers, as well as certain kinds of accident and disaster risks. The incidence of HIV/AIDS is gen- Infant Under-5 erally highest in large cities, as well as in infor- Mortality Mortality Prevalence mal settlements, for multiple reasons.50 Rate Rate of Diarrhoea* Location (per 1000) (per 100) (%) Crime and violence, noted earlier as a concern Kenya (national) 74 112 3.0 to enterprises, increasingly accompanies depri- Rural 76 113 3.1 vation in African cities. Research in Latin Amer- Nairobi 39 62 3.4 ica and globally has confirmed that income Other urban centres 57 84 1.7 Informal settlements in 91 151 11.3 inequality is an explanatory factor in national Nairobi, of which: crime rates (Bourguignon 1999), and Africa Kibera 106 187 9.8 shares the distinction with LAC of being the two Embakasi 164 254 9.1 most unequal regions (Global Monitoring Report 2005). The U.N. Office on Drugs and Crime has *Incidence of diarrhoea with blood in children under 3 in the two weeks prior to interview. concluded that a major cause of crime in Africa is the continent's wide income disparity.51 The gen- Source: African Population and Health Research Centre (2002). Table from Satterthwaite (2004b). eral problems of poverty and social exclusion within urban areas, extreme weaknesses of national police and justice systems, and absence of trust between communities and local govern- where the governments, the private sector, and ments compound the issue. Crime is now recog- the citizens have been developing various strate- nised as a significant risk factor especially in the gies to confront the problems (Box 5.1). cities of South Africa and in Nairobi and Lagos, Negative health FIGURE 5.2. The Urban Poor Do Not Always Have Effective Access to externalities often Healthcare affect spatially con- rotcod 80 centrated areas of 70 Urban richest hti Urban poorest disadvantage, such 60 w Rural average erac 50 as slums. These are al 40 urban health penal- nate 30 ant ties, in the sense of g 20 intteg 10 risks more highly % 0 associated with Ghana Kenya Guinea Cameroon Gabon Benin Niger urban than with Source: Mboup (2003) rural living. 48 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A BOX 5.1. Crime Issues and Responses in Three African Countries Kenya. The security situation in Kenya has deteriorated in tion of the crime has increased. A comprehensive National recent years, particularly in the country's urban centres, Crime Prevention Strategy was adopted in 1996 with the with an upsurge of car-jacking, robberies, and murders. aim of expanding from a focus on crime control to crime The law enforcement agencies are seriously below prevention. Components of the NCPS include coordination strength and lack the necessary resources and capacity to and integration of criminal justice functions, special tar- combat the perpetrators effectively. With the cooperation geting of high crime areas, research, advocacy, and facili- of the business community, the police in Nairobi have con- tation of crime prevention programmes. All levels of structed information centres within the central business government are involved in the programme. The interna- district to respond to urgent calls. Unfortunately, a legacy tionalisation and high economic stakes of crime in South of distrust still exists between the police and the public, Africa calls for a well-trained professional police service. emanating from past police practices. The police force The appointment of specialised crime-fighting units and realises this problem and, together with NGOs, is attempt- cooperation between government and the private sector ing to raise public awareness about the need for more businesses have begun to show effects against crime. effective policing and adequate resources and duties within the framework of human rights. The Safer Cities Initiative, through UNDP funding and Nigeria. In Nigeria, law enforcement has long been seen as implemented by UN-Habitat, has conducted a survey on not doing enough to protect life and property both in the crime in Nairobi, supported capacity building of local rural and the urban areas. The country has developed its partners, and assisted in the development of appropriate own versions of community-based crime prevention, with coordination mechanisms at the national level. As a result, many vigilante groups forming in response to requests a three-year citywide crime prevention strategy and action from Nigeria's various ethnic groups. Throughout the plan for Nairobi is being developed to respond to growing years, "neighbourhood watch" has grown from an "extra concerns about local level urban safety issues. Considering eyes and ears" approach to crime prevention to a much the recent political developments in Kenya, developing more proactive, community-oriented endeavour. Neigh- and implementing a participatory crime prevention strat- bourhood Watch groups are now incorporating activities egy in Nairobi is viewed as a way to improve urban gov- that not only address crime prevention issues, but also ernance and to augment the government's Economic restore pride and unity to a neighbourhood. It is not Recovery Strategy (ERS). uncommon to see Neighbourhood Watch groups partici- pating in neighbourhood cleanups and other activities that impact the quality of life for community residents. South Africa. South Africa has one of the highest national and urban crime levels in the world--it ranked fourth of Sources: Compiled by Francis Muraya, from Seventh United about 60 countries reporting rates of burglary, murder, Nations Survey of Crime Trends and Operations of Criminal Justice and robberies. According to the Johannesburg-based Cen- Systems, covering the period 1998 to 2000 (United Nations Office tre for the Study of Violence and Reconciliation (CSVR), on Drugs and Crime, Centre for International Crime Prevention); South Africa's current high rate of violent crime is just as United Nations Office on Drugs and Crime, Centre for Interna- related to economic and social marginalisation as it was tional Crime Prevention website; Centre for the Study of Violence during the 1980s, although the diversity and sophistica- and Reconciliation (South Africa) website. T H E O P P O R T U N I T Y C O S T O F N E G L E C T I N G T H E C I T I E S 49 Bank ldro W/k nemar CartruC Children in Kibera slum, Nairobi, Kenya. Bank ldro W A market along rail tracks in Accra, Ghana. Strategic and Operational 6 Choices and Priorities Attempting to Gain the Best from the Urban Transition There need be no ational policy makers (and their external advisors) who wish to take debate about the first N advantage of what urban development can offer the national agenda face priority for healthy many choices, given resource constraints. There need be no debate about the first priority for healthy cities--a generally sound macroeconomic and fiscal environ- cities--a generally ment and investment climate at the national level. Beyond this, alternatives and sound macroeconomic trade-offs may be faced in the spatial focus of urban policy and in the sectoral focus. In supporting cities, getting the right institutional balance is a further challenge. and fiscal environment and investment Issues of Spatial Focus climate at the Large city versus small cities--or, urban concentration versus dispersal. One national level. frequent debate regarding urban policy in Africa and other developing regions concerns the balance of support between one or very few relatively large cities, versus the distribution of urban population and economic activity across a wider number of geographically dispersed, smaller cities. Although, as noted in chap- ter Two, the urban system of most African countries is not unusual in compara- tive terms, often the prevailing political preference is to reduce the relative dominance of the largest (primate) city. The frequent belief is that the primate city has become overgrown and inefficient (and sometimes politically inconven- ient). It is thought that smaller, more spread-out cities would create greater link- ages to the rural hinterland and better sharing of welfare throughout the country. In contrast, a concern with enhancing global competitiveness and the very limited investment funding available may argue in favour of focusing resources in few locations: the major cities. The weight of both theory and empirical evidence suggests, however, that African countries need not adopt a very deliberate or directive approach to reduce their primacy rate and disperse urban population among many smaller cities. In fact, it could be ineffective and risky to do so. In the first place, the relatively large cities of any country are generally the most productive for all the reasons of agglomeration economies discussed earlier. Multicentric, metropolitan areas tend to be the centres of the most diversified production, of innovative and 51 52 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A information-intensive industries and services, ingredient in the growth of interregional trading and of entrepreneurship because they offer the groups, an increasing priority of African govern- deepest labour market and more opportunities ments. A factor that may be slowing the impacts for exchange of ideas (Henderson, Shalizi, and of current efforts to create regional trading Venables, 2001).52 There is no reason to believe arrangements in Africa is the relative absence of that the large cities in African countries do not city-regional corridors--extended conurbations exert this functional advantage, despite the many comprised of the physical integration of neigh- developmental constraints unrelated to the cities bouring settlements and industrial areas. City- that limit producers' efforts. The observation that regions often develop following the location of large African cities pose major diseconomies even industry along transport corridors or occur in before all the potential gains of size have been greenfields that start as enclaves but blend even- realised is a testament to the neglect of urban pol- tually, forming a more or less continuous urban icy. The blame for serious congestion, for exam- region. Such corridors are apparent in East Asia ple, should fall first on the virtual absence of and parts of Latin America, but in Africa only the public transport services and the lack of traffic Gauteng region around Johannesburg-Pretoria management, not on the size of the city itself. is an obvious example.55 Many African cities do Similarly, the neglect of urban transport and of feature physical "sprawl" at very low density, with solid waste management (most African cities do wide swaths of peri-urban informal settlement not collect more than half of the waste produced, having minimal or nonexistent services--but this and few have sanitary waste disposal) and the low periphery is much less rooted in a web of formal coverage of household sanitation explain high sector investment (industrial or infrastructural) levels of pollution. Excessive land prices are than is seen, for example, in the East Asian city inevitable when the land supply is heavily regu- regions. This notable difference in the nature of lated or publicly owned. urban expansion reflects the relative absence of A deliberate urban policy to promote decon- inter-urban transport infrastructure in Africa, the centration of urban populations in Africa is very weak integration of peri-urban settlements with difficult to implement effectively and can be very the rest of the city, and the lack of large-scale costly if exercised through public fiscal transfers, industrial manufacturing investment that forms as evidenced from the disappointing regional the core of such urbanised spread elsewhere development policies of many other countries, (Webster 2004; Ellis and Harris 2004). such as Brazil.53 As African countries urbanise Although national governments should not aim to "dislodge" the primate city from its dom- and incomes grow, more cities will emerge and inant position in the economy, more basic expand naturally, gaining relative economic support should be given to rapidly growing sec- importance. Decentralising politi- ondary cities and their local governments to As African cal control and easing regulatory improve governance and service delivery capac- controls on the private sector will countries urbanise ity. The intergovernmental fiscal framework, and tend to promote this process natu- in particular the design of transfers, should and incomes grow, rally, by reducing the incentive for encourage local governments to mobilise rev- firms to stay physically close to more cities will enues to the extent possible from their emerging the major city.54 Providing more local economies, while recognising that smaller emerge and expand and better infrastructure in and cities will likely remain more reliant on fiscal between secondary cities would, of transfers than the larger cities.56 Rather than naturally, gaining course, facilitate this process. attempting to "pick winners" among emerging relative economic A network of cities that is well- secondary cities or to "create" new cities as connected within a country or growth poles, national urban policy should estab- importance. across its borders can be a key lish conditions and incentives that help existing S T R AT E G I C A N D O P E R AT I O N A L C H O I C E S A N D P R I O R I T I E S 53 local governments respond to the evolving wide range of developing coun- The best response to demands of the population and of producers in tries in the past, including in their jurisdiction for good public services. The Africa, should impose a major mitigate inequality World Bank's municipal development projects in note of caution on the current is to encourage the Africa, which provide structured assistance to debate. The successful cases have improve local government performance in finan- been relatively few, while costs in integration of rural cial management and make investments propor- public expenditure and foregone areas with large tional to such performance, illustrate an tax revenues can be considerable.57 incentive-based approach that supports second- Much of the globally-oriented urban markets, ary cities based on their effective demand. manufacturing in Africa that has mobility of labour Ultimately, the debate about urban policy and emerged with foreign investment the system of cities in Africa hinges on concerns (garments in Lesotho, export and remittances, about inequality among regions and specifically zones in Madagascar) has had only between the more urbanised and less urbanised a superficial impact on the rest of strong fiscal mobili- areas of the country. The main goal of African the economy in terms of backward sation in cities, and governments is to sustain economic growth to linkages or building fungible skills reduce poverty, and inequality is a worry as it and entrepreneurship, although a appropriate inter- weakens the transmission from growth to pov- few programmes, such as those in governmental trans- erty. Much of urban policy and investment in the Mauritius, have proven more suc- past couple of decades has been judged as a con- cessful than others. The solution to fer policies--rather tributor to inequality (through "urban bias"), creating more employment in such than simply to avoid neglecting its role in growth and poverty reduc- zones, however, is not simply to tion. This report argues that urban policies can be locate them more widely through- urban development made more effective in ensuring cities' contribu- out the country, as some have tions to growth and poverty reduction and that argued (GMR 2005, Box 4.2), out of fear of the African countries cannot afford to neglect because modern exporting firms inequality. these channels. Some regional inequalities are generally need access to a major inevitable and may increase during the transition. city and its amenities. The best response to mitigate inequality is to For many types of manufacturing and serv- encourage the integration of rural areas with ices, foreign investors are notoriously footloose, large urban markets, mobility of labour and and retaining such investment requires ground- remittances, strong fiscal mobilisation in cities, ing in strong local resources (such as a capable and appropriate intergovernmental transfer and reliable workforce), institutional relation- policies--rather than simply to avoid urban ships, and amenities. Rather than creating a few development out of fear of inequality. favourable investment enclaves, a more sustain- able and scaled-up approach would be to make Promoting industrial or export zones--or the cities work better as support centres for busi- cities? There is also growing interest in Africa in ness. Increasingly, even developed countries are providing special industrial or export promotion gearing their industrial policies to build on the zones as a way of targeting favourable conditions localised strengths of a city or region, such as its (infrastructural and institutional) for new invest- entrepreneurs and skilled workforce, partner- ment. The argument is that many other countries ships across public and private sectors, linkages to have pursued this strategy to stimulate export markets, and even its history, to attain sustainable sectors especially and that African countries can economic development. The basic preconditions at best afford to provide the desired conditions would be national policies that favour private in a few locations. investment and fiscal policies that give local gov- As summarised by Madani (1999), the results ernments (especially of the larger cities) the of many export processing zones (EPZs) in a means to mobilise revenues and to deliver decent 54 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A Creating efficient services to the entire city. An Issues of Institutional industrial policy favouring export Development and responsive local promotion or other targeted in- vestment zones may be justified in Creating efficient and responsive local govern- governments is key cases, but it should not precede ments is key to creating effective cities. This chal- fuller consideration of how to lenge requires that national authorities and to creating effective make the city more hospitable to external advisors (donors) recognise the need for domestic investment and to ex- further conditions and choices--in particular, cities. isting firms. Greater economic putting municipal development at the centre of growth for African countries may urban policy. require them to satisfy the exigen- First, central governments need to commit to cies of foreign export markets, but without meet- appropriate fiscal frameworks that give local gov- ing the standards for a healthy domestic urban ernments incentives for fiscal effort and responsi- market an export-oriented strategy has a soft footing. ble spending but also reasonable certainty in planning tax and transfer revenue. African municipalities have developed responsible bor- Issues of Sectoral Focus rowing within such frameworks, as demonstrated in many World Bank-supported and other urban African countries will seek national growth projects. Use of credit is indispensable for financ- strategies to suit their particular conditions, and ing local infrastructure as cities grow. each of these is likely to have urban develop- Second, mature municipal management ment requirements. Resource-scarce coastal requires that local investment be on-budget and countries (such as Ghana), for example, may part of an expenditure plan, rather than made be most inclined to consider an export-led through ad-hoc assistance arrangements or spe- approach, for manufacturing or services. Exter- cial extra budgetary funds. nal competitiveness begins with an efficient Third, accountability to the local population domestic marketplace, but rigid urban land use and a voice for citizens should be encouraged as policies can distort the location or relocation of part of the mainstream performance of elected factories. Land-locked resource-scarce countries local governments and instilled through open, (for example, Uganda and Rwanda) may choose transparent budgeting and monitoring processes. to foster more regional integration to facilitate Community involvement in neighbourhood their trade and develop modernised services sec- investments should be encouraged as part of tors with low transport costs--but these sectors municipal planning, but it cannot be the solution will also require that their major cities become to creating sustainable and well-integrated in- more attractive to educated workers and frastructure and service systems in the urban investors. The majority of the African countries context. may increasingly aim to develop more intensive In brief, sustainable urban development is agriculture and agro-processing--for which inseparable from local government capacity- the consumer demand and support services in building and cannot be achieved simply through urban areas will be an essential asset. In short, investment in urban infrastructure--or by relying virtually all the "non-urban" growth strategies only on practices derived for less institutionally- for Africa will have major urban development complex settings (such as social investment funds implications. This will be especially true for and community-driven development). Many countries that emphasise efforts to rapidly trans- urban issues, in fact, require institutional cooper- form and diversify their economic structure and ation across municipal jurisdictions to address to stimulate globally competitive industries and larger spatial externalities, such as solid waste dis- services. posal and public transport management. S T R AT E G I C A N D O P E R AT I O N A L C H O I C E S A N D P R I O R I T I E S 55 Priorities for Further data collection, would be a major contribution to advancing knowledge of the links among Urban Research in Africa urban development, economic growth, and Collecting city-level data. On virtually every poverty reduction. topic mentioned in this paper, the state of knowl- edge is very limited with respect to Africa. A par- Priorities for use of existing and new data. ticularly notable gap is the lack of quantified city While better data are a prerequisite for better studies because of the dearth of reliable data at knowledge, data sources that are available could the city level that would permit tracking key indi- be mined to further illuminate many of the issues cators of performance over time and among mentioned here. A first priority would be to cities. Developing such data is critical for research examine a larger set of household surveys in but also for better governance and decision- Africa to explore multiple dimensions of urban making by and for the cities themselves. poverty, which is only illustrated very partially The aim to document how well the city is serv- here. A second priority might be to look more ing its residents and businesses, what it is produc- systematically at comparisons across cities in ing and contributing to the country, and where some countries to identify and target responses it is falling short in performance is central to the to the varying needs of large/primate/capital concept of a city development strategy. The cities versus those of smaller/secondary cities. recent publication by the South African Cities Third (but not least), focusing on the financial Network, State of the Cities Report 2004, demon- resources available to cities--contrasting the strates the rich texture that can be created to actual revenues collected with a reasonable fiscal describe the issues facing cities, using census and harvest that might be gleaned from the observed other data available within the country and in the economic activity--would be a revealing and cities themselves. A commitment by the interna- possibly energising research.58 Each of these top- tional community and by city leaders to promote ics could inform the urban and national policy the use of policy-relevant indicators in other dialogue and help define better-tailored urban countries, by supporting a process of sustainable interventions. Bank ldro W/k nemar CartruC Dakar, Senegal. City putoa M Boarding the Matatu to the city in Maputo, Mozambique. 7 Conclusions Much better managed F I R S T -- recognise and move beyond the myths that cloud much of the dis- course on African urban development. urban development Myths about the demographics: could therefore play a African countries are not urbanising faster than other countries have, and the significant role in distribution of urban population among the large and very small cities of launching African Africa is not unusual for their level of development. That said, the absolute rate of urban growth is historically unprecedented and a challenge for urban economies onto a management, particularly in the secondary cities, which tend to be the most stronger, more underserviced. Internal migration is neither the only nor even the main source of urban sustained path of growth in most of countries--nor is it responsible for urban poverty. economic growth. Although data are limited, evidence indicates that migration has been favourable on balance for both sending and receiving areas in Africa. But population mobility is much more fluid than the rural-to-urban model would indicate, and households wisely diversify their activities across both areas. Myths about the urban economies. Africa cannot simply be characterised as "urbanisation without growth," and the term does not even fit many of the coun- tries. The economic growth that has taken place in the past decade derives mainly from urban-based sectors (industry and services), and this is especially true of the better-performing countries. But cities have clearly not lived up to their produc- tive potential because of widespread neglect and bad management. Myths about cities and poverty. Urban poverty is not mainly a function of urban growth, nor is it a sign of failure of the urban economies. Evidence indicates that much of the deprivation in cities, and the consequent emerging urban public health problems, relate to institutional failures that perpetuate social exclusion and inequalities between the urban poor and the urban non-poor. S E C O N D -- recognise what cities can offer the national development agenda-- and what this requires in turn. Much of the development dialogue over the past 57 58 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A What is needed thirty years has been, and remains, countries must foster productivity by mo- obsessed with the view that atten- bilising knowledge, skills, technology, and from both urban tion to cities represents "urban innovation--factors that have been widely advocates and urban bias." Yet cities suffer the effects of associated with an urban (and large city) genuinely bad urban policy and platform. Yet for many African countries, critics is an honest neglect (no financial security, pal- amenities are still defined for residents as a and hard look both try investment in local public hope of getting regular garbage collection, goods), misguided incentives that at what the urban drains to prevent flooding after the rain, distort the use of land and other and an indoor toilet. Africans need their phenomena can investments, and hostile treatment cities to allow the economy to transform, of much of the population on offer national de- but they also need to transform their cities. which the city depends. What is Governance and fiscal sustainability. The velopment across needed from both urban advocates major signal of a well-functioning city, and and urban critics is an honest and numerous channels, its major determinant, is the quality of gov- hard look both at what the urban ernance and financial management. African and what support phenomena can offer national cities cannot be ushered into the present development across numerous cities and local gov- century without systematic improvements channels, and what support cities in accountability and service delivery at the ernments require and local governments require in local level. With the right incentive frame- turn to achieve these results. in turn to achieve work, municipalities can demonstrate sig- Cities are essential contributors nificant gains in resource mobilisation these results. to national development priorities and responsible borrowing--indeed, many on many fronts. are doing so and providing citizens with ground-level examples of good government. The agricultural transition. By providing market demand for agricultural goods and Much better-managed urban development for the nonfarm activity of rural areas, cities could therefore play a significant role in launch- are an underutilised resource for stimulating agricultural intensification and evolution to ing the African economies onto a stronger, more higher value production. Cities are thus a sustained path of economic growth. While there key link in the "virtuous cycle" of urban and is much criticism of the negative externalities rural livelihoods. In many cases, addressing seen in many of the African cities--problems of bottlenecks in city performance will be an environmental degradation, congestion, and, effective entry point into this "virtuous increasingly, crime, these shadows should be cycle." recognised as signals of serious neglect. To put it Private sector development. Cities nurture bluntly, it is not credible to argue that disec- entrepreneurs by providing centres of onomies outweigh the positive benefits of African demand (larger markets), information shar- cities when they have virtually no working pub- ing, relatively easy access to credit, and sup- lic transport or safe waste disposal, much of the port for risk-taking through other income land is held in public control with little availabil- alternatives and through networks of other ity for market demand or public good purposes, entrepreneurs. But because of failures in and infant mortality is rising due to poor public urban institutions and services, many firms sanitation in the neighbourhoods where most of are not gaining the cost advantages of their the work force lives. urban location. Economic modernisation. Whether to gain African cities have the potential to be strong plat- access to export markets or for more forms and laboratories for most of the economic competitive domestic markets, African and social behaviours needed for transforma- C O N C L U S I O N S 59 tive growth and productivity. These favourable to their citizens and eager for a sustainable economic and social behaviours include creative growth path are doing just that. Local govern- innovation, technological application, entrepre- ments can become the strongest advocates for neurship, openness of attitudes to change, and cities, linking up with each other (as the metro risk-taking. Firms and individuals find more cities have done in South Africa) to share lessons opportunity in cities and towns despite the many and information to make their case. The national problems there, and the forces of agglomeration governments and external donors should wel- and migration appear as strong in this Region as come the voice of local authorities and include they have proven to be elsewhere. Releasing the them in dialogues on the wide range of policies potential of Africa's cities by addressing basic and actions that affect them. This would include weaknesses in land markets, public transport, and engaging the local officials in national develop- the provision of urban services could reduce ef- ment strategy, as well as in the design and fective "binding constraints" to future growth in interpretation of national poverty surveys and Africa. At the very least, such a strategy would investment climate analyses, the planning of create a more hospitable environment for the regional investments and large sectoral under- investors and workers who will increasingly and takings, and the formulation of major fiscal invariably congregate in urban markets. reforms. Supporting these efforts, the many good practices, as well as lessons from less successful T H I R D -- look to cities and municipalities to attempts, in past assistance programmes should play greater roles in national development agen- be applied with renewed commitment to un- das. Increasingly, African governments responsive blocking the development contribution of cities. se Pictur TE/Still SCHYT GENROJ New slum shacks in Nouakchott, Mauritania. Bank ldro W/k nemar CartruC Nairobi city park with the skyline in the background, Kenya. A Review of Investment Climate Survey Results for A P P E N D I X Sub-Saharan Africa and Selected Asian Countries 59 T he essential ingredients of any sound market economy have been described as the "investment climate" (WDR 2005). Several types of surveys have been conducted in recent years to gain the perspectives of firms on the ease or difficulty of working in different settings. Most of these surveys have been conducted with formal (registered) manufacturing enterprises located in urban areas. These firms' responses provide a view of the business environment in which they operate--which comprises conditions pertaining not only to the specific city location and its local government but also, of course, to the country and national government. As background for the present study, investment climate (IC) surveys com- pleted as of early 2004 were reviewed for six countries in Africa (Ethiopia, Kenya, Mozambique, Tanzania, Uganda, and Zambia) and for four developing coun- tries in East and South Asia (Cambodia, China, Bangladesh, and Pakistan), as comparators. This analysis reveals how the firms surveyed, which are all in manufacturing sectors and related services, rate the various constraints or bottlenecks they face and differentiates between firms located in the capital and/or primate city from those in other cities of the country. The factors chosen for analysis are those that could shed light on the business environment in the various cities and, to the extent possible, policy implications for the local government. In most cases, the questionnaires were not sufficiently detailed to distinguish the source of policies or practices as between the local government and the national government. Two categories of indicators were examined: (i) infrastructure, land and labour--including telecommunications, electricity, transportation, access to land (the only variable in this category largely under the responsibility of local government), and the skills and education of workers; and (ii) institutional constraints--including tax administration, customs and trade regulations, labour regulations, business licensing and operating permits, corruption, and crime, theft, and disorder. Of this second set of con- straints, customs and trade regulations would be purely national policy, and local governments would share responsibility with central governments for the others (except for business licensing and permitting, which is most likely a local government matter). Responses were recorded as the percentage of firms sampled 61 62 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A who rated a particular topic as posing a "moder- In the African sample, firms in capital or ate" or "major" constraint. primate cities cited worse problems, except Overall, the main observations from this IC in Ethiopia and Zambia. The reported review can be summarised as follows. objective indicators of electricity quality (days of connection, extent of outages, and Constraints within infrastructure: so on) in the same surveys do not show a Electricity is by far the biggest obstacle cited consistent comparison between the two for all of the African firms, 65 to 90 percent groups of cities, possibly because many of whom rated it as a moderate/major con- firms compensate with their own genera- straint, as did 90 percent of Bangladeshi tors, for example, 74 percent of firms in firms and about 60 percent of the Pakistani. Nairobi and 67 percent in Dar es Salaam. Still, 10 percent of output is reported lost due to power out- FIGURE A1. Access to Land as a Constraint in Urban Centres ages in both cities. Telecommunications was reported Bangladesh to be a much less serious obstacle Dhaka than was electricity (except in Other cities China). Yet, firms in all urban areas Pakistan in Kenya and those outside Kam- Islamabad & Karachi pala in Uganda reported losing Other cities telecom connections for more than Cambodia Phnom Penh 30 hours at a time, on average.60 Other cities Firms in the capital or primate China cities reported most concern with Beijing & Shanghai telecoms, possibly because it is Other cities more difficult to compensate by Ethiopia getting around for face-to-face Addis Ababa meetings. Other cities Transportation was reported to be Kenya Nairobi an obstacle greater than or at least Other cities on par with telecommunications Mozambique in both the African and the Asian Maputo samples. Whereas problems in Other cities electricity and telecoms were felt Tanzania more in capital or primate cities, Dar es Salaam transportation issues were more Other cities balanced and actually rated much Uganda South West* worse in the secondary city sam- Other cities ple in Mozambique, Tanzania, Zambia Moderate and Zambia. The surveys did not Lusaka Major differentiate among types of Other cities transportation, but it is likely that the firms in secondary cities are 0 10 20 30 40 50 60 70 80 90 100 particularly vulnerable to the Percent of firms shortcomings of intercity road or * Includes Kampala rail access. In Ethiopia, domestic transport costs are estimated as A R E V I E W O F I N V E S T M E N T C L I M AT E S U R V E Y R E S U LT S 63 eight times those of China, four-times those of South Africa, and twice Kenya's. Mada- TABLE A1. Burdens of Obtaining Land gascar has only one road leading to the port, Days from the date and its disruption can totally block exports of release to the (World Bank 2004d). In such circumstances Days for land use of the plot, Country City acquisition average (median) even port cities, which exist from locational advantage, may demonstrate much less eco- Ethiopia Addis Ababa 538.0 (180.) 115.0 (60.) nomic benefit in reality. Other urban 150.6 (120.0) 58.0 (60.0) Mozambique Maputo 38.3 (12.0) It is worth noting here that whereas these find- Other urban 11.7 (2.5) ings concern infrastructure sectors that are Zambia Lusaka 232.8 (180.0) largely the responsibility of national authorities Other urban 393.2 (240.0) and therefore are not "urban" or "local govern- ment" issues, they are of great interest in revealing Note: Results are from only about a dozen firms that answered this question in these surveys, so they cannot be taken as representative. that some of the basic assumed advantages of urban agglomeration--access to inputs and services affording some economies of scale and con- TABLE A2. Firm Responses on Costs Due to Security Problems nectivity to other producers--are simply not being realised in devel- Sales loss oping country cities, even in those Security Protection due to costs payments theft etc. that might be considered rela- Country City (% of total sales) (% of total sales) (% of total sales) tively favoured. Therefore, a firm locating in such a city gains a Ethiopia Addis Ababa 0.9 0.4 Other urban 0.9 1.3 much lower productivity advan- tage than its counterpart else- Kenya Nairobi 2.7 0.1 3.9 where possesses, all else remaining Other urban 2.9 0.0 4.7 constant (such as city size, natu- Mozam- Maputo ral resource base, labour supply, bique Other urban and so on). Tanzania Dar es Salaam 2.9 0.2 Other urban 6.5 0.2 Access to skilled labour is Uganda South West rated almost as much of a (incl. Kampala) 1.8 0.0 problem as transport by the Other urban 2.1 0.4 African firms, while labour is Zambia Lusaka 4.0 0.1 5.9 China's biggest reported Other urban 1.9 0.0 3.3 constraint. In about half the countries, firms in the capital Bangladesh Dhaka 0.6 Other urban 0.7 or primate cities rate labour as a bigger concern than do Pakistan Islamabad and firms in the other cities. Karachi 1.6 0.9 0.3 Other urban 0.9 0.3 0.1 Larger cities create higher demand for skilled labour, Cambodia Phnom Penh 16.8 3.4 2.9 although they also tend to Other urban 12.5 1.8 1.5 attract more skilled workers. China Beijing 0.7 0.9 0.4 Access to land, the last in- Other urban 0.9 0.6 0.2 frastructure variable, is the 64 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A and in Zambia (Table A1). In FIGURE A2. Corruption as a Constraint in Urban Centres Mozambique firms pay on aver- age $18,000 in processing fees for Bangladesh land, and in Nigeria they must re- Dhaka Other cities register land to use it as collateral, a process that can take up to two Pakistan Islamabad & Karachi years and cost 15 percent of the Other cities land value in official fees, before Cambodia bribes. Rental costs for private Phnom Penh firms in Ethiopia can be about Other cities 20 percent of total sales, while China government owned firms pay Beijing & Shanghai Other cities little for rental (Office of the Chief Economist 2004). A FIAS Ethiopia Addis Ababa inquiry into investors' views in Other cities Senegal elicited strong com- Kenya plaints about real estate as being Nairobi one of the most serious obstacles; Other cities improvements in tax administra- Mozambique tion, business registration, and Maputo customs were said to be over- Other cities whelmed by the lengthening of Tanzania Dar es Salaam time required to obtain land Other cities (World Bank 2003e). Uganda South West* Constraints within institutions: Other cities Corruption dominates the other Zambia Moderate Lusaka Major issues as a major concern for all Other cities the African countries (except Ethiopia) and for the Asian coun- 0 10 20 30 40 50 60 70 80 90 100 tries (except China). Generally, Percent of firms corruption is felt to be a bigger * Includes Kampala problem in the capital or primate cities, indicating that it may per- tain especially to national govern- ment officials (Figure A2). Tax administration is the second most fre- least cited as being a moderate/major con- quently reported constraint for most of the straint in the African sample, in Cambodia, countries, with relatively little variation and in China, although it is a big issue in across them or across the two types of cities. Dhaka and in Ethiopia (with a 65 to 70 per- The lack of variation between the city cent rating). (See Figure A1.) Land acquisi- groups (except in Mozambique, where tion may be rated less badly in IC surveys Maputo dominates the concern) may signify because, unlike other factors and services, it that national taxation is the main issue-- is not a frequent expenditure. Reported land local taxation being almost trivial in these acquisition delays are very long in Ethiopia countries anyway. A R E V I E W O F I N V E S T M E N T C L I M AT E S U R V E Y R E S U LT S 65 Crime, theft and disorder occa- sion a very large inter-country FIGURE A3. Crime, Theft, and Disorder as a Constraint in variation in reported levels of Urban Centres concern (Figure A3). This issue is cited by 80 to 90 percent of Bangladesh firms in Kenya and Maputo, Dhaka closely followed by other Zam- Other cities bian cities and by Phnom Pakistan Islamabad & Karachi Penh. Surprisingly, firms in Other cities capital or primate cities do not Cambodia claim to suffer this problem Phnom Penh more than do those in second- Other cities ary cities. Firms in other urban China areas in Kenya and Tanzania, Beijing & Shanghai and in Lusaka, reported incur- Other cities ring the highest levels of costs Ethiopia or sales losses due to security Addis Ababa Other cities issues, but the African cities did not rate worse in all cases than Kenya Nairobi did the Asian cities (Table A2). Other cities Business licensing had the least Mozambique negative ratings, although it is Maputo cited as a moderate/major con- Other cities straint by about half the firms Tanzania in Maputo and Dar es Salaam Dar es Salaam and by 60 percent in Dacca. Other cities The two types of cities did not Uganda show a large difference in re- South West* Moderate Other cities Major sponse on this point. Zambia Lusaka Summarising main Other cities conclusions by country: 0 10 20 30 40 50 60 70 80 90 100 In Ethiopia, land problems Percent of firms rival electricity as the major constraints, exceeded only by * Includes Kampala Source: Chakraborty (2005). tax administration. Kenya shows serious problems in most categories but corrup- tion and crime top the list. about equal areas of concern, while for In Mozambique, Maputo stands out as the Cambodia, China, and Pakistan institutional problematic location for business. constraints dominate. For most of the African countries, the insti- Bangladesh rivals the African countries in tutional and the infrastructural issues are the extent and seriousness of IC obstacles. Additional Statistical Tables Note: Countries within each Region shown in these tables are an indicative subset for which the World Bank has an active operational dialogue and data available; excluded are very small countries or those in extensive civil conflict. Regional averages include all countries in Africa according to World Development Indicators data. 67 68 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A TABLE SA1. Population Growth 1982-2003--Total Population, Urban Population, Urbanisation, and Population in Largest City (Primacy) Urban Primacy­population of Total population Total population population largest city as % of (in millions) (in millions) (% of total) urban population % growth p.a. % growth p.a. 1982 2003 1990-2003 1982 2003 1990-2003 1982 2003 2005 Sub Saharan Africa Benin 3.7 6.7 2.7 1.1 3.0 4.7 28.7 44.5 28* Botswana 1.0 1.7 2.3 0.2 0.9 3.6 22.5 50.3 22* Burkina Faso 7.3 12.1 2.4 0.7 2.1 4.4 9.6 17.6 37.0 Cameroon 9.3 16.1 2.5 3.1 8.2 4.3 33.1 51.2 22.0 Cote d'Ivoire 8.8 16.8 2.7 3.2 7.6 4.3 35.8 44.9 44.0 Ethiopia 39.9 68.6 2.3 4.3 11.4 4.4 10.9 16.6 23.0 Ghana 11.8 20.7 2.3 3.7 7.7 3.1 31.6 37.1 24.0 Guinea 4.6 7.9 2.4 0.9 2.3 4.1 20.1 28.9 60.0 Kenya 17.9 31.9 2.4 3.1 11.6 5.6 17.6 36.3 22.0 Lesotho 1.3 1.8 1.0 0.2 0.5 4.2 14.6 30.3 53* Madagascar 9.3 16.9 2.9 1.8 5.3 5.1 19.5 31.4 31.0 Malawi 6.6 11.0 2.0 0.6 1.7 4.4 9.6 15.9 30* Mauritania 1.7 2.8 2.6 0.5 1.8 5.2 31.0 61.7 34* Mauritius 1.0 1.2 1.1 0.4 0.5 1.4 42.0 42.3 27* Mozambique 12.7 18.8 2.2 1.9 6.7 6.2 14.6 35.6 18.0 Niger 5.9 11.8 3.3 0.8 2.6 5.8 13.3 22.1 35.0 Nigeria 75.8 136.5 2.7 21.5 63.6 4.9 28.4 46.6 16.0 Rwanda 5.5 8.4 1.5 0.3 0.6 3.1 4.8 6.6 45* Senegal 5.9 10.2 2.6 2.1 5.1 4.2 36.4 49.6 43.0 South Africa 29.0 45.8 2.0 14.0 27.1 3.5 48.2. 59.2 12.0 Tanzania 19.8 35.9 2.6 3.1 12.7 6.4 15.9 35.4 19.0 Togo 2.7 4.9 2.6 0.7 1.7 4.2 24.4 35.1 47* Uganda 13.5 25.3 2.9 1.2 3.9 5.3 9.2 15.3 31.0 Zambia 6.1 10.4 2.2 2.4 4.2 2.4 39.8 40.3 33.0 Total 406.3 704.5 2.5 88.6 237.6 4.6 21.8 36.0 24.0 East Asia and Pacific Cambodia 7.1 12.5 2.6 0.9 2.5 5.6 12.5 19.0 44* China 1008.6 1280.4 1.0 211.0 498.0 3.6 20.9 39.0 2.0 Indonesia 154.2 211.7 1.4 36.6 94.7 4.2 23.7 44.0 13.0 Mongolia 1.8 2.4 1.3 0.9 1.4 1.2 53.2 57.0 55* Philippines 50.4 79.9 2.2 20.0 49.7 3.9 39.6 61.0 63.0 Vietnam 55.7 80.4 1.6 10.8 20.7 3.3 19.4 25.0 23.0 Total 1402.4 1838.5 1.5 314.8 725.5 3.6 22.4 39.0 8.0 South Asia Region Bangladesh 89.9 135.7 1.7 14.3 37.0 4.1 15.9 27.0 31.0 India 718.4 1048.6 1.7 169.2 301.3 2.5 23.6 28.0 6.0 Nepal 15.2 24.1 2.4 1.1 3.2 5.2 7.0 13.0 20* Pakistan 87.4 144.9 2.4 25.0 50.6 3.3 28.6 34.0 22.0 Sri Lanka 15.0 19.0 1.3 3.2 4.6 2.1 21.5 24.0 16* Total 942.6 1401.5 1.8 215.4 403.5 2.8 22.9 28.0 10.0 A D D I T I O N A L S TA T I S T I C A L TA B L E S 69 TABLE SA1. Population Growth 1982-2003--Total Population, Urban Population, Urbanisation, and Population in Largest City (Primacy) (Continued) Urban Primacy­population of Total population Total population population largest city as % of (in millions) (in millions) (% of total) urban population % growth p.a. % growth p.a. 1982 2003 1990-2003 1982 2003 1990-2003 1982 2003 2005 Middle East & North Africa Egypt, Arab Rep. 43.0 67.6 1.9 18.9 28.9 1.8 43.9 43.0 37.0 Morocco 20.3 30.1 1.7 8.6 17.3 3.0 42.7 57.0 21.0 Tunisia 6.7 9.9 1.5 3.5 6.7 2.6 52.4 67.0 30.0 Yemen, Rep. 9.2 19.2 3.7 1.9 4.9 4.6 20.3 26.0 30.0 Total 185.1 311.6 2.1 91.2 181.8 2.8 49.2 59.0 27.0 Europe & Central Asia Albania 2.8 3.2 -0.3 0.9 1.4 1.3 34.1 44.0 27* Armenia 3.2 3.1 -1.1 2.1 2.1 -1.1 66.0 67.0 52.0 Azerbaijan 6.4 8.2 1.1 3.4 4.3 0.8 53.2 52.0 42.0 Bulgaria 8.9 7.8 -0.8 5.6 5.3 -0.7 62.5 68.0 20.0 Georgia 5.2 5.1 -0.5 2.7 2.9 -0.3 52.5 57.0 36.0 Kazakhstan 15.2 14.9 -0.7 8.3 8.3 -0.9 54.8 56.0 13.0 Kyrgyz Republic 3.8 5.1 1.0 1.4 1.7 0.3 38.3 34.0 47.0 Moldova 4.1 4.2 -0.2 1.7 1.8 -1.1 41.5 42.0 34* Romania 22.5 21.7 -0.5 11.2 12.1 -0.2 50.0 56.0 14.0 Tajikistan 4.2 6.3 1.3 1.4 1.7 0.3 33.9 28.0 25* Turkmenistan 3.0 4.9 2.2 1.4 2.2 2.2 46.6 45.0 45* Uzbekistan 16.8 25.6 1.7 6.8 9.4 1.0 40.7 37.0 73.0 Total 434.1 472.2 0.1 258.2 301.1 0.2 59.5 64.0 .. Latin America & Caribbean Bolivia 5.6 8.8 2.1 2.6 5.6 3.2 47.4 64.0 26.0 Brazil 127.1 176.6 1.4 87.0 146.2 2.1 68.4 83.0 12.0 Ecuador 8.4 13.0 1.8 4.1 8.4 3.0 48.6 64.0 27.0 EI Salvador 4.7 6.5 1.9 2.1 4.1 3.8 45.1 64.0 33.0 Guatemala 7.2 12.3 2.6 2.7 5.0 3.1 37.5 41.0 18.0 Honduras 3.8 7.0 2.9 1.4 3.9 4.9 36.0 56.0 25.0 Mexico 70.8 102.3 1.6 47.9 76.7 1.9 67.6 75.0 24.0 Nicaragua 3.1 5.5 2.8 1.6 3.1 3.4 50.8 57.0 35.0 Total 372.1 532.7 1.6 246.5 407.8 2.1 66.3 77.0 23.0 Low income 1637.3 2311.9 2.0 376.3 686.5 3.3 23.0 30.0 16.0 Lower middle income 1859.3 2655.5 1.1 673.7 1319.8 2.6 36.2 50.0 13.0 Upper middle income 246.1 333.1 1.3 164.8 251.1 1.7 67.0 75.0 26.0 High income 842.5 972.1 0.7 619.1 758.3 1.0 73.5 80.0 19.0 World 4585.2 6272.5 1.4 1833.9 3015.7 2.2 40.0 49.0 16.0 *Own Calculations, based on the World Urbanization Prospects (2003) Source: World Development Indicators (2005) 70 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A TABLE SA2. Economic Sector Summary (Agriculture, Industry, Services) in GDP, 1990­2003 Contribution of Average annual sector to GDP % GDP Average annual % growth % share of GDP growth, 1990­03 growth 1990­2003 Average GDP 1990­2003 (growth* share) explained by pc % growth industry and Agr Ind Svcs GDP 1990­2003 Agr Ind Svcs Agr Ind Svcs services** Sub-Saharan Africa Benin 5.3 5.2 4.3 4.8 1.9 36.0 13.9 50.1 1.9 0.7 2.1 60.2 Botswana ­0.3 4.2 7.2 5.3 2.8 3.5 48.9 47.6 0.0 2.0 3.4 100.2 Burkina Faso 3.7 2.4 4.9 4.0 1.6 31.7 17.8 50.4 1.2 0.4 2.5 71.2 Cameroon 4.7 0.5 ­0.1 1.6 ­0.9 37.8 22.7 39.5 1.8 0.1 0.0 3.3 Cote d'Ivoire 2.8 1.0 0.9 1.3 ­1.5 26.6 21.7 51.8 0.7 0.2 0.5 48.2 Ethiopia 1.4 2.3 4.9 3.1 0.7 51.9 10.0 38.1 0.7 0.2 1.9 73.7 Ghana 3.2 3.4 5.8 4.3 1.9 38.5 23.2 38.3 1.2 0.8 2.2 71.3 Guinea 4.3 4.0 3.1 3.8 1.3 23.5 34.4 42.1 1.0 1.4 1.3 72.3 Kenya 1.0 1.6 2.8 1.8 ­0.7 25.4 17.9 56.7 0.3 0.3 1.6 87.9 Lesotho 0.9 6.2 3.6 3.7 2.6 18.1 39.9 41.9 0.2 2.5 1.5 96.2 Madagascar 1.7 1.8 1.9 1.7 ­1.2 28.9 12.9 58.2 0.5 0.2 1.1 73.5 Malawi 7.3 0.9 2.2 3.2 1.1 37.0 20.8 42.2 2.7 0.2 0.9 29.4 Mauritania 2.6 2.0 5.7 3.8 1.2 24.9 30.3 44.7 0.7 0.6 2.5 82.8 Mauritius 1.0 5.3 6.2 5.2 4.0 9.3 32.0 58.7 0.1 1.7 3.6 98.3 Mozambique 4.4 10.5 4.8 6.1 3.9 31.8 22.4 45.8 1.4 2.3 2.2 76.2 Niger 3.3 2.0 1.8 2.3 ­1.0 39.4 17.3 43.3 1.3 0.3 0.8 47.1 Nigeria 3.6 2.8 4.7 3.6 0.8 30.6 46.3 23.1 1.1 1.3 1.1 68.7 Rwanda 4.7 2.8 2.9 3.3 0.5 40.8 20.0 39.2 1.9 0.6 1.1 47.2 Senegal 2.6 5.0 3.6 3.7 1.0 18.8 20.0 61.1 0.5 1.0 2.2 86.8 South Africa 0.7 0.9 2.5 1.8 ­0.2 4.0 33.9 62.1 0.0 0.3 1.5 98.5 Tanzania 3.5 4.5 3.7 4.1 1.3 46.2 15.7 38.1 1.6 0.7 1.4 56.3 Togo 2.9 2.9 2.3 2.3 ­0.3 37.5 20.4 42.1 1.1 0.6 1.0 58.3 Uganda 3.8 10.0 7.8 6.5 3.4 44.0 16.6 39.4 1.7 1.7 3.1 73.7 Zambia 3.8 0.2 2.5 1.5 ­0.8 21.5 35.4 43.1 0.8 0.1 1.1 58.4 Median 2.8 2.0 2.7 2.4 ­0.1 18.9 31.1 50.0 0.5 0.6 1.3 78.8 East Asia & Pacific Cambodia 3.8 15.2 5.1 6.6 4.0 43.6 19.9 36.5 1.7 3.0 1.8 74.4 China 3.9 12.2 8.1 9.3 8.2 19.4 48.1 32.5 0.8 5.9 2.6 91.9 Indonesia 2.1 5.6 5.0 4.6 3.1 17.6 42.7 39.7 0.4 2.4 2.0 92.3 Mongolia*** 2.2 ­0.7 0.3 ­1.3 ­2.6 33.5 21.7 44.9 0.7 ­0.2 0.1 ­1.5 Philippines 2.2 3.0 4.1 3.3 1.0 18.8 32.5 48.7 0.4 1.0 2.0 87.9 Vietnam 3.8 10.5 7.2 7.3 5.6 28.2 31.8 40.2 1.1 3.3 2.9 85.1 Median 3.0 10.1 7.0 7.7 6.5 18.7 45.3 36.0 0.6 4.6 2.5 92.8 South Asia Bangladesh 3.4 7.0 4.6 4.9 3.1 26.1 24.5 49.4 0.9 1.7 2.3 81.7 India 2.9 5.8 7.4 5.6 3.8 27.8 26.8 45.5 0.8 1.6 3.4 86.0 Nepal 2.9 6.0 5.2 4.5 2.1 42.6 21.2 36.2 1.2 1.3 1.9 71.9 Pakistan 3.6 4.4 4.5 3.9 1.4 25.7 24.0 50.3 0.9 1.1 2.3 78.5 Sri Lanka 2.0 5.7 5.6 4.8 3.5 22.6 26.5 51.0 0.5 1.5 2.9 90.6 Median 2.9 5.7 6.7 5.3 3.4 27.4 26.2 46.3 0.8 1.5 3.1 85.4 A D D I T I O N A L S TA T I S T I C A L TA B L E S 71 TABLE SA2. Economic Sector Summary (Agriculture, Industry, Services) in GDP, 1990­2003 (Continued) Contribution of Average annual sector to GDP % GDP Average annual % growth % share of GDP growth, 1990­03 growth 1990­2003 Average GDP 1990­2003 (growth* share) explained by pc % growth industry and Agr Ind Svcs GDP 1990­2003 Agr Ind Svcs Agr Ind Svcs services** Middle East & North Africa Egypt, Arab Rep. 3.1 5.4 3.8 4.2 2.2 17.1 32.3 50.5 0.5 1.7 1.9 87.2 Morocco 6.4 3.5 3.3 3.1 1.3 16.5 31.6 51.9 1.1 1.1 1.7 72.6 Tunisia 2.9 4.2 5.2 4.8 3.2 13.3 28.7 58.0 0.4 1.2 3.0 91.6 Yemen, Rep. 6.4 5.0 5.8 5.4 1.6 18.7 34.4 46.9 1.2 1.7 2.7 78.8 Median 3.3 3.6 3.9 3.8 1.5 12.5 39.4 48.2 0.4 1.4 1.9 88.9 Europe & Central Asia Albania 5.3 ­0.2 3.7 2.2 2.4 37.5 23.5 39.0 2.0 0.0 1.4 41.2 Azerbaijan 0.8 4.5 134.5 ­0.8 ­1.8 23.5 39.5 37.0 0.2 1.8 49.8 99.6 Bulgaria*** 2.2 ­3.1 ­2.0 ­0.8 0.1 15.5 34.7 49.8 0.3 ­1.1 ­1.0 119.4 Georgia ­0.2 6.1 7.5 ­4.8 ­4.4 34.1 24.2 41.6 ­0.1 1.5 3.1 101.3 Kazakhstan ­1.2 4.5 3.1 ­0.1 0.7 12.6 36.2 51.2 ­0.1 1.6 1.6 104.8 Moldova*** ­7.8 ­4.2 1.1 ­5.2 ­5.0 31.5 27.6 36.8 ­2.4 ­1.2 0.4 24.1 Romania*** 0.2 ­0.9 ­7.1 ­0.5 ­0.1 18.1 40.6 41.3 0.0 ­0.4 ­2.9 100.9 Tajikistan*** ­1.6 ­5.3 4.0 ­3.9 ­5.2 30.3 31.4 38.3 ­0.5 ­1.7 1.5 22.2 Turkmenistan 3.2 ­0.1 0.8 2.7 0.5 22.7 44.9 32.3 0.7 0.0 0.3 25.0 Uzbekistan 1.5 ­1.4 0.8 1.0 ­0.7 33.3 27.9 38.8 0.5 ­0.4 0.3 ­19.8 Median ­0.2 ­1.3 1.6 0.2 0.0 11.0 35.8 53.2 0.0 ­0.5 0.9 104.9 Latin America & Caribbean Bolivia 2.3 3.5 3.6 3.5 1.3 15.9 31.4 52.7 0.4 1.1 1.9 89.2 Brazil 3.8 1.1 2.0 1.8 0.4 7.6 31.2 61.1 0.3 0.3 1.2 84.6 Ecuador ­0.7 1.1 4.1 2.3 0.4 13.4 29.8 56.8 ­0.1 0.3 2.3 103.5 El Salvador 1.4 4.4 4.6 4.1 2.1 12.5 28.5 59.0 0.2 1.3 2.7 95.9 Guatemala 2.5 3.3 4.2 3.6 1.0 23.9 19.7 56.3 0.6 0.7 2.4 83.7 Honduras 2.8 3.3 3.2 2.9 0.1 19.3 30.1 50.6 0.5 1.0 1.6 82.8 Mexico 1.9 3.1 3.1 3.0 1.4 5.6 27.7 66.7 0.1 0.9 2.0 96.4 Nicaragua 2.6 3.0 2.4 2.9 0.1 22.4 23.8 53.8 0.6 0.7 1.3 77.7 Median 2.3 2.1 2.7 2.5 0.9 7.7 30.9 61.4 0.2 0.6 1.6 92.8 Lower middle income 2.3 4.6 3.8 3.6 2.5 14.3 38.5 47.3 0.3 1.8 1.8 91.5 Low income 2.9 4.8 5.6 4.5 2.4 29.2 26.6 44.2 0.9 1.3 2.5 81.4 Middle income 2.2 4.1 3.5 3.4 2.2 11.7 37.3 51.0 0.3 1.5 1.8 92.7 Upper middle income 1.7 3.0 3.0 2.8 1.5 6.4 34.9 58.7 0.1 1.0 1.8 96.3 High income 1.2 1.6 3.0 2.5 1.7 2.2 29.5 68.2 0.0 0.5 2.0 99.0 Source: Global Development Finance (GDF) & World Development Indicators (WDI) Central (April 2005), SIMA Database. * Calculated using the values from "contributions of sector to GDP growth" by the following formula: (Industry+Services)/(Agriculture+ Industry+Services). **For these countries, the numbers in last column denote the extent to which the negative growth rate can be explained by decline in the growth rates of the industry and the services sectors combined, given their shares. 72 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A TABLE SA3. Sectoral Sources of Growth Among High Growth African Countries (1996­2003), % per capita p.a. GDP Agriculture Non-agric. Country Name growth rate growth rate growth rate Mozambique* 5.40 3.30 6.14 Mauritius* 3.97 0.00 4.47 Botswana* 3.37 ­3.30 3.60 Uganda* 2.51 0.00 4.18 Guinea* 2.34 3.10 2.06 Tanzania* 2.25 0.90 2.96 Burkina Faso* 2.22 0.70 2.88 Lesotho* 2.20 2.00 2.52 Cameroon* 2.18 3.80 0.71 Senegal* 2.15 ­0.30 2.72 Rwanda* 2.13 3.00 1.42 Benin* 1.95 2.60 1.55 Ethiopia* 1.60 ­1.60 3.61 Mauritania* 1.42 ­1.30 2.30 Nigeria* 1.37 1.80 1.32 Togo* 1.11 0.70 1.36 Malawi* 1.10 4.50 0.01 Ghana 0.93 0.70 1.07 South Africa 0.62 1.00 0.65 Zambia 0.53 ­1.90 1.59 Niger ­0.23 ­0.30 ­0.16 Madagascar ­0.40 ­1.2 ­0.30 Kenya ­1.10 ­1.40 ­0.82 Cote d'Ivoire ­1.16 ­0.30 ­1.46 Median 1.78 0.68 1.57 Non-agriculture consists of all industry and services. Growth rate calculated between endpoint years. Countries marked * sustained minimum 1% GDP growth pa over period. Source: WDI 2005. A D D I T I O N A L S TA T I S T I C A L TA B L E S 73 TABLE SA4. Poverty Rates, Urban and Rural, and Urban Share of Total Poor (most recent year and 2020) Poverty Poverty Year Most recent year Urban poverty headcount, headcount, I-PRSP*** urban poverty as % of urban (% of rural (% of PRSP** as % of total total poverty, population) population) SIMA* poverty 20201 Sub-Saharan Africa Benin 23.2 33.0 1999** 35.3 47.1 Botswana .. .. .. .. .. Burkina Faso 19.9 52.3 2004** 7.4 11.9 Cameroon 22.1 49.9 2001* 31.0 41.5 Cote d'Ivoire 23.0 42.0 1998*** 30.5 39.1 Ethiopia 37.0 45.0 2000* 13.7 21.4 Gambia 32.5 51.1 1998** 22.9 33.1 Ghana 17.3 36.0 1998/99** 21.8 28.5 Guinea 25.0 52.5 2002** 17.8 30.5 Kenya 49.0 53.0 1997* 33.5 48.9 Lesotho .. .. .. .. Madagascar 52.1 76.7 1999* 23.2 33.7 Malawi 54.0 65.5 1998** 13.1 20.7 Mauritania 25.4 61.2 2000* 33.1 58.2 Mauritius .. .. .. .. .. Mozambique 62.0 71.3 2000** 31.3 48.6 Niger 52.0 66.0 1993** 17.8 27.4 Nigeria 30.4 36.4 1993* 41.3 53.9 Rwanda 22.6 67.9 2000** 2.2 3.7 Senegal 59.0 88.0 20012** 38.2 50.3 Tanzania 29.5 38.7 2001* 28.4 43.9 Togo .. .. .. .. Uganda .. .. .. .. Zambia 56.0 83.1 1998* 31.1 38.8 East Asia & Pacific Cambodia 13.9 40.1 1999* 7.0 12.7 China 2.0 4.6 1998* 20.8 33.2 Indonesia 16.3 34.1 3 26.5 40.2 Mongolia 39.4 32.6 1998* 62.8 66.0 Philippines 21.5 50.7 1997* 39.0 51.5 Vietnam 6.6 35.6 2002* 5.7 9.4 South Asia Bangladesh 36.6 53.0 2000* 19.6 29.4 India 24.7 30.2 2000* 24.2 30.3 Nepal 23.0 44.0 1996* 7.0 11.8 Pakistan 24.2 35.9 1999* 25.6 33.1 Sri Lanka 15.0 27.0 1996* 14.5 21.6 Middle East & North Africa Egypt, Arab Rep. 22.5 23.3 1996* 41.9 47.4 Morocco 12.0 27.2 1999* 36.7 47.0 Tunisia 3.6 13.9 1995* 34.2 44.0 Yemen, Rep. 30.8 45.0 1998* 18.9 26.4 (Continued) 74 T H E U R B A N T R A N S I T I O N I N S U B - S A H A R A N A F R I C A TABLE SA4. Poverty Rates, Urban and Rural, and Urban Share of Total Poor (most recent year and 2020) Poverty Poverty Year Most recent year Urban poverty headcount, headcount, I-PRSP*** urban poverty as % of urban (% of rural (% of PRSP** as % of total total poverty, population) population) SIMA* poverty 20201 Europe & Central Asia4 Albania 15.1 23.4 33.3 44.0 Armenia 22.4 16.6 73.6 77.1 Azerbaijan 23.9 15.6 62.3 66.1 Bulgaria 14.9 30.6 50.2 54.2 Georgia 21.0 19.0 59.2 66.2 Kazakhstan 14.8 26.2 41.7 46.1 Kyrgyz Republic 14.1 23.2 24.1 27.4 Moldova 21.9 19.0 45.3 51.3 Romania 11.4 30.2 32.0 37.6 Tajikistan 16.8 20.9 23.4 28.1 Turkmenistan 9.5 28.0 21.9 27.8 Uzbekistan 15.6 22.6 28.6 32.1 Latin America & Caribbean5 Bolivia 52.0 82.0 1999 50.0 61.1 Brazil 27.0 58.0 1998 65.0 80.2 Ecuador 60.0 .. 1998 n.a. .. El Salvador 16.0 53.0 1998 30.0 37.5 Guatemala 26.0 66.0 1998 20.0 32.2 Honduras 41.0 71.0 1998 37.0 40.4 Mexico 15.0 55.0 1998 43.0 52.5 Nicaragua 45.0 67.0 1998 46.0 56.1 Urban and Rural Headcount are for most recent year available. All figures cited are for basic poverty, not extreme poverty. Source: GDF & WDI Central (April 2005), SIMA database, except where noted. **PRSP data summarised in Diana Mitlin, 2003, "Understanding Urban Poverty: What the Poverty Reduction Strat- egy Papers Tell Us," Poverty Reduction in Urban Areas Series, Working Paper 13, Human Settlements Programme, International Institute for Environment and Development, London. For Burkina Faso, poverty data is from the PRSP, July 2004. For Guinea, poverty data is from the PRSP, January 2002. ***For Cote d'Ivoire, poverty data is from the Interim PRSP, January 2002. Notes: (1) Population data from United Nations (2004). Figures were estimated by taking base year urban and rural poverty headcount rates and applying them to projected 2020 urban and rural populations. (2) The PRSP indicates a range of 44­59% urban poverty and 72­88% rural poverty. The highest values have been used for this table. (3) Based on "iterative method"; see Menno Pradhan, Asep Suryahadi, Sumarto Sudarno, and Lant Pritchett2000, "Measurements of Poverty in Indonesia: 1996, 1999, and Beyond," Social Monitoring and Early Response Unit (SMERU) Working Paper, Jakarta, Indonesia. (4) Dimensions of Urban Poverty in ECA, Final Draft, March 11, 2004, Energy and Infrastructure (ECSIE), Europe and Central Asia Region (5) Fay (2005). For some countries, there are large disparities in poverty headcounts from different sources; for example, alterna- tive poverty headcounts from PRSPs, as cited by Mitlin: Urban Poverty Rural Poverty Cambodia 42.4(14.6) 56.1 Vietnam 7.8 19.7 Yemen 10.0 19.9 Honduras 37.0 58.0 Endnotes EXECUTIVE SUMMARY the mid-1990s. (The latest census for Benin is from 1992, for Cameroon 1987, for Ethiopia 1. Estimated rates of urban growth should be 1994, for Madagascar 1993, for Nigeria 1991 taken with caution because of incomplete [provisional], and for Senegal 1998.) Benin, census data in many countries. It should also Ethiopia, and Madagascar have had only one be noted that urban and rural distinctions census, Cameroon and Senegal only two are more an administrative artifice than real (www.citypopulation.de ). Hence, projec- boundaries, except at the extremes of den- tions for many countries remain tentative. It sity and settlement size. has been suggested that where there has been 2. Excluding the smallest and most of the most economic stagnation or decline in the past conflict-ridden countries. decade, urban growth (at least the share due 3. Defined as the ratio of non-working-age to internal migration) has probably been less population (under 15 and over 64 years old) than the official estimates in the absence of to the working age population. census data (Satterthwaite 2002). 4. Shelter can represent 5 to 10 percent of GDP, 7. Countries differ on how they define "urban." 15 to 20 percent of capital formation, more Criteria usually include a settlement size than half of national wealth, and three- threshold (which varies among countries quarters of the financial sector activity in from about 2500 to 10,000 residents) and developed markets. may include a minimum share of nonagri- 5. This is rarely acknowledged in explicit cultural economic activity and administra- national policy statements in any country. A tive conditions. Criteria may change between notable exception is China's 10th Five Year census periods or settlements shift from one Plan (2001), which advocates urbanisation category to another, leading to discontinuous to stimulate rural and national economic changes in the urban-to-total population development. ratio. 8. The 2001 revision of the same document had predicted that Addis Ababa, Luanda, and CHAPTER 2 Abidjan would exceed 5 million residents by 6. Urban population data for Sub-Saharan 2015. Africa should be taken with caution, since 9. As a rule, larger urban areas are the most some countries have not had a census since productive, since they allow for greater spe- 75 76 E N D N O T E S cialisation in labour use, better matching of backward-bending line on these graphs) skills with jobs, and a wider array of con- were Estonia and Egypt, while Kazakhstan sumption choices for workers and ancillary showed no change in GDP (WDI data). services to producers. As long as this greater 15. The sample excludes mainly the smallest and productivity from positive externalities out- most conflict-ridden countries of the weighs higher costs for land, labour, housing, Region. and other necessities, the city can grow and 16. Estimates have been made for some African thrive. Once diseconomies of urban scale countries of the spatial shares of GDP based (negative impacts such as traffic congestion, on analysis of the population distribution pollution, and crime) become too great, and its participation in primary sector, infor- however, the city may start to lose its edge in mal sector, and modern sector activities, creating jobs, attracting high quality labour, drawing on local surveys and comparing or improving welfare of residents. these to formal national accounts data. For 10. The urban primacy rate varies widely among example, such estimates indicate that for the African countries, and the variance Cameroon in 2002, which is 50 percent does not seem closely related to per capita urbanised, 70 percent of GDP derives from income or country size. See Table SA1 in the urban areas; the two largest cities, Douala Appendix. and Yaounde, with 10 and 8 percent, respec- 11. An alternative data source was also looked at tively, of total population, account for 25 and that breaks down the smallest settlement 16 percent, respectively of GDP. The analysis category to a minimum of 100,000 residents also finds that rural productivity (average (www.citypopulation.de). Possibly incomplete rural income) is growing due to the demand coverage, especially in the small-settlement pull from urban areas (Club du Sahel and category, makes further breakdown at this end PDM, 2004). The same methodology is used of the distribution unreliable, however. for the Addis estimate, cited earlier. 12. The range of estimates runs from about 7 to 17. Industry includes manufacturing, mining 70 percent (Montgomery et al. 2003, Table (which may be mainly rurally located), con- 3-4, 90; Lucas 1998). Both the rates of urban struction, and utilities. Services include natural increase and migrant shares of banking and insurance, trade, and all other growth tend to be high at the initial stages of services. Agriculture includes forestry and urbanisation and to decline as urbanisation fishing. rises (Montgomery et al. 2003, 151). 18. The thirteen-year period portrayed here is 13. Montgomery et al.2003, Figure 3-4, 91. This somewhat arbitrary, but the same analysis observation underscores the earlier point was done for the two halves of this period, that in the 1990s, in countries whose per and for the years 1992 to 2002 and the three capita economic growth was nil or negative, years 2000 to 2002. The results are basically the estimated urban population growth rates the same in all cases, with these two sectors in the absence of recent census data may be accounting for over 80 to 95 percent of aver- somewhat exaggerated. age GDP growth in every country grouping. 14. During 1990 to 2002, only 18 countries expe- 19. Household poverty studies vary in the wel- rienced declining urbanisation rates, of fare indicator they use to measure so-called which 3 were tiny island states, 13 transition "income poverty"--ranging from actual economies (all but Mongolia are in Europe income (relatively rare, because most diffi- and Central Asia, ECA), and two in MNA cult to capture), expenditure, or consump- (Egypt and Iraq). The only cases of declin- tion translated into monetary terms. ing urbanisation with positive per capita "Non-income poverty" reflecting quality of GDP growth (which would also appear as a life in terms of access to or quality of essen- E N D N O T E S 77 tial services, education and health status, sat- Gabon, versus 82 for the rest of the city (UN- isfaction, security, and so on, is indicated by Habitat, Global Urban Observatory, 1998 objective or qualitative measures pertaining data). to each characteristic. Poverty or welfare can 24. This fact can be seen especially in the case of also be described in terms of access to assets electricity, where the costs of extending the or to varied forms of capital (human, physi- network throughout an urban area are rela- cal, financial, social, intellectual, and natural tively small and therefore failure to do so is resources) or in terms of vulnerabilities and difficult to explain on purely financial risks faced. grounds. In Africa, whereas 52.2 percent of 20. Note that poverty incidence rates should not the urban non-poor have electricity, only be compared across countries because of dif- 19.7 percent of the urban poor do (a gap of ferences in estimated poverty thresholds. See almost 33 percentage points), twice the gap Table SA4 in the Appendix. between the urban poor and the 4.3 percent 21. Urban poverty incidence is actually higher of rural residents with the service. A similar than the rural average in Mongolia and in comparison holds on average for electricity Armenia, Azerbaijan, Georgia, and Moldova. across other developing regions and for Given the extent of urbanisation and relative shares of households lacking all three net- urban-to-rural poverty rates, the urban poor worked services (Montgomery et al., 2003, already represent half or more of the total Table 5-4, 175). poor in these countries, as well as in Bulgaria, 25. The UN-Habitat Global Urban Observatory Bolivia, Brazil, and several other LAC coun- estimates the number of slum dwellers in tries. This is also a pattern common to terms of five criteria of shelter deprivation upper-middle income and high income (nondurable housing structure, overcrowd- countries; for example, it is true of Hungary ing, lack of safe water, lack of sanitation, and and Russia, as documented in World Bank insecure tenure). The estimate is based on (2004b). the first four criteria, as there are currently 22. Montgomery et al. (2003, Table 5-3,173-174) few good estimates of tenure status. looked at results of Demographic and Health 26. For example, according to the Zambia Liv- Surveys (DHS) across the developing regions ing Conditions Monitoring Survey of 1998, and concluded that access to infrastructure only 6 to 8 percent of urban residents, across services fairly consistently rises with settle- all expenditure quintiles, have their house- ment size category, from under 100,000 resi- hold garbage collected. dents at the low end to over 5 million at the 27. Similar effects on urban poverty were seen high end. The contrasts are less often statis- in Indonesia due to the financial crisis of tically significant for Sub-Saharan Africa 1997 to 1998 (World Bank 2003b). than for the other regions, however. 28. In Ghana, for example, the recent Core Wel- 23. Household surveys report what households fare Indicators Questionnaire (CWIQ- say they spend on different goods and serv- national household survey) reveals that the ices, not what they would need to spend to nutrition indicators (percent of children meet their minimum needs. UN-Habitat's underweight, stunted, or wasted) worsened city indicators database reveals that per much more in Accra than for any other capita daily water consumption in informal region between 1997 and 2003 (WB memo settlements in Antananarivo, Madagascar, by Carlos Cavalcanti, 9/27/04). was only 20 litres (the level considered by 29. Sahn and Stifel (2003), tables 3 and 4. Note WHO to be the minimum health standard) that the authors focus on the findings of versus 40 for all settlements in the city; and absolute disadvantage of rural relative to 30 in informal settlements of Port-Gentil, urban areas and conclude that their evidence 78 E N D N O T E S confirms their hypothesis of persistent urban refer to migration cast it in pejorative terms, (anti-rural) bias. as a cause of poverty, disease, crime, or other 30. DEXIA 2003. www.dexia.com. problems for development. 31. Including Bulgaria, Georgia, Kazakhstan, 41. The DHS queries this detail of women Kyrgyz Republic, Moldova, Mongolia, and respondents of reproductive age, of whom Romania(International Monetary Fund one-fourth reported moving to her current [IMF] Government Finance Statistics [GFS], city or town within the past 5 years. 2004). 42. One exception to this positive result in the 32. From founding congress final declaration of DHS data was in the survival rates of infants CCRA, May 2005 (www.uclga.co.za). whose mothers had migrated from rural areas, which were found to be significantly worse than those for non-migrants, for rea- CHAPTER 3 sons unclear (Montgomery et al. 2003, 287). 33. According to ILO (2004), both labour pro- 43. A household living standards survey study ductivity and total factor productivity in SSA in Indonesia, the Philippines, and Vietnam agriculture have remained below those of likewise found that migrants were not worse every other region (including China) since off than were other respondents. A major 1980, although the trend has been on the shortcoming of the Vietnam household sur- upturn since the mid-1990s (figures 3.14 vey, however, was that it did not capture non- and3.15). registered migrants, who are officially 34. WDR 2003 (Tables 4.2, p. 61 and 4.3, p. 62). excluded from access to public benefits and 35. Estimate by Clare Romanik, Urban Institute services (World Bank 2003). (December 2004 draft, "An Urban-Rural 44. One recent study based on a nationally rep- Focus on Agricultural Markets"), extrapo- resentative household survey in Guatemala lated from Malawi and Mozambique house- distinguished between international and hold survey data. internal remittances, finding that both types 36. Urban Harvest. "Science for People and the reduce the level, depth, and severity of Planet." In Annual Report 2003. CGIAR Sys- poverty. The poverty headcount fell by equal tem-wide Initiative for Urban and Peri- amounts--0.8 and 0.9 percent--when inter- Urban Agriculture, cited in Romanik, 2004. nal and international remittances, respec- 37. Defined as the ratio of non-working-age tively, are included in household income. population (under 15 and over 64 years old) Since a large share of the receiving house- to the working age population. holds are in the lowest income decile, their 38. According to 90 Demographic and Health income status improved most dramatically Surveys (DHS) in 56 countries, urban fertility (Adams 2004). is on average 25 percent lower than rural fer- 45. These shares had dropped from about one- tility; in SSA the rates are 5.07 to 6.5, a 22 per- third of incomes in 1986 to 1987, due to cent difference (Montgomery et al., 2003). external factors (World Bank 2003f, p. 56). 39. There is widespread evidence of difficulties 46. While in the late nineteenth­to early twenti- in recruiting and retaining health and educa- eth centuries as much as a third of the labour tion personnel to rural areas, relative to force migrated abroad from depressed rural urban centres (Global Monitoring Report areas (for example, from Ireland and Scan- 2005). dinavia to the United States), international 40. Black et al. (2003) found that of 48 Poverty migration to the OECD area is more Reduction Strategy Papers (PRSPs) exam- restricted today. Cumulative migration to the ined, migration was not even mentioned in United States between 1970 and 2000 ac- 21 of them, and nearly all of those that did counted for less than 2 percent of the labour E N D N O T E S 79 force in Sub-Saharan Africa and less than across countries or has studied metropolitan 5 percent in LAC (the region with the highest areas in developed countries. migration ratio); moreover, while the labour inflows in the last century consisted mainly CHAPTER 6 of peasants, today more than half the migrants to the United States have higher 53. The main finding of Brazilian and other education or other skills (WDR 2003, studies of regional development policies is Box 4.4). that agglomeration economies and market access create strong increasing returns that tend to reinforce the spatial concentration CHAPTER 4 of economic activity in the dominant cities 47. Preliminary analysis of a survey of 2000 and regions of countries. This pattern is very informal businesses in Lagos also finds that difficult to change by fiscal or financial poli- improvement in infrastructure (electricity, cies. See the literature review in M. Fay, edi- water, good roads) is cited first as the contri- tor, "Brazil Economic Development at State bution most desired from the state or local Level--(Some) Lessons from Experience," government, closely followed by "help with World Bank, Latin American and Caribbean access to credit" and "safety and security on Region, May 2005 draft. Korea's attempts at the streets" (Tewari and Banerjee, 2005). industrial relocation in the 1970s an 1980s to new towns outside of Seoul succeeded as far as reducing the capital's share of manu- CHAPTER 5 facturing employment and raising it in other cities, with some efficiency costs. The urban 48. Tewari and Banerjee (2005) find that an population concentration in Seoul barely overwhelming share of both large manufac- changed, however. See Kyu Sik Lee and Sang- turing firms and small-medium-micro scale Chuel Choe, 1989, "Changing Location firms do not export even within Africa, and Patterns of Industries and Urban Decentral- there are relatively few reported trading link- isation Policies in Korea," in Jene K. Kwon, ages across firm sizes. ed., Korean Economic Development (West- 49. Evidence summarised in Montgomery et al. port, CT: Greenwood Press); and Henderson, (2003), Chapter 7, Table 7-4. Shalizi, and Venables 2001. 50. Risk factors include the extent and nature of 54. This was one of the lessons from the Korean social contacts, including density and high efforts toward industry de-concentration. mobility in urban settings, as well as the pres- 55. Rakodi (1999) observes that the only cross- ence and proliferation of vulnerable groups, border metropolitan region apparent in such as street children and sex workers Africa is that between Douala, Lagos, and (Boerma, Nunn, and Whitworth 1999; Kelly Abidjan, following the creation of ECOWAS. 2003). Measures are being taken to encourage an 51. Cited by Carsten Hyttel, Eastern Africa urban corridor between Johannesburg and regional representative for the U.N. Office on Maputo. Drugs and Crime, Centre for International 56. Rodriguez-Pose and Gill (2003) show for a Crime Prevention. global sample of countries that fiscal devo- 52. There is no rule as to the absolute size at lution that simply increases the fiscal auton- which a city becomes more productive or at omy of subnational governments can worsen what size negative externalities may impinge. regional inequalities and reinforce the Most of the research on city productivity has stronger regions, unless national govern- looked at the relative size (primacy rate) ments maintain equalisation transfers. 80 E N D N O T E S 57. In Tanzania, for example, industrial estates 58. This is a major focus of the ECOLOC have been found not to provide superior program research of the Municipal Devel- infrastructure services to the resident firms opment Program and OECD/Club du Sahel. (Vandana Chandra, Pooja Kacker, and Ying Li, 2005, "Identifying the key constraints to APPENDIX growth, export competitiveness and employ- ment in Tanzania's manufacturing sector," 59. The median number of hours lost was about World Bank, PREMED, draft Executive 10, so the range experienced was very wide. Summary). References Adams, Richard. 2004. Remittances and Poverty in Boerma, J.T., A.J. Nunn, and A.G. Whitworth. 1999. Guatemala. Presentation at World Bank. "Spread of HIV infection in a rural area of Tanzania." Ades, Alberto F., and Edward L. Glaeser. 1995. "Trade and AIDS 13: 1233­1240. Circuses: Explaining Urban Giants." The Quarterly Bourguignon, Francois. 1999. Crime, Violence and In- Journal of Economics (February):195­227. equitable Development." World Bank Annual Bank African Population and Health Research Centre Conference on Development Economics, Washington, (APHRC). 2002. Population and Health Dynamics in D.C. Nairobi's Informal Settlements. Nairobi. Bryceson, D.F. 2004. "Rural livelihoods and agrarian Becker, Kristina Flodman. 2004. The Informal Economy. change in Sub-Saharan Africa: processes and policies." SIDA Fact Finding Study. Sweden. In Rural Livelihoods and Poverty Reduction Policies, Bell, Clive, Shantayanan Devarajan, and Hans Gersbach. ed. F. Ellis and H.A. Freeman. London: Routledge. 2003. The Long-Run Economic Costs of AIDS: Theory Carruthers, Robin, with Malise Dick and Anuja Saurkar. and an Application to South Africa. World Bank: 2004. Affordability of Public Transport in Developing Policy Research Working Paper WPS3152. Countries. Transport and Urban Development Bertaud, Alain. 2001. The Costs of Utopia: Brasilia, Department, World Bank. Johannesburg and Moscow. Revised. http://alain Central Bank of Nigeria (CBN) and Nigerian Institute of bertaud.com/ (Accessed June 10, 2006) Social and Economic Research (NISER). 2003. An In- Bertaud, Alain, and Jan K. Brueckner. 2004. Analyzing Depth Study of Nigeria's Informal Distributive Trade Building Height Restrictions: Predicted Impacts, Wel- Sector. Draft. Abuja. fare Costs, and a Case Study of Bangalore, India. Central Bank of Nigeria (CBN), Nigerian Institute of World Bank, Transport and Urban Development Social and Economic Research (NISER), and Federal Department, Urban Unit. Policy Research Working Office of Statistics (FOS). 2003. Indepth Study of Paper 3290. Nigerian Informal Manufacturing Sector. Draft Final Bertaud, Alain, and Bertrand Renaud. 1995. Cities With- Report. Abuja. out Land Markets: Location and Land Use in the Socialist City. World Bank Policy Research Working Centre for African Settlement Studies and Development Paper 1477. (CASSAD). 2002. Economic and Employment Trends Bertaud, Alain, Robert Buckley, Sumila Gulyani, Kate in Karu. Ibadan, Nigeria. Owens, and Kevin Villani. 2004. Real Estate Reform Chakraborty, Mita. 2005. Acting Locally While Thinking in Addis Ababa: Making the City a Platform for Mar- Globally: Subnational Strategies to Promote Invest- ket Development. Executive Summary. ment Climate. TUDUR. Draft. Black, R., R. Sabates-Wheeler, R. Skeldon, C. Waddington, Chambas, Gerard, and Elsa Duret. 2000. La Mobilisation and A. Winters. 2003. Mapping Study of Migration des Ressources Locales au Niveau des Municipalities Issues. Sussex Centre for Migration Research. d'Afrique Subsaharienne. Ministere des Affaires Etran- Boerma JT, A. Nunn and J. Whitworth. "Mortality impact geres, Direction Generale de la Cooperation Interna- of the AIDS epidemic in less developed countries: Evi- tionale et du Développement. France. dence from community studies," AIDS (1998) 12(Sup- Club du Sahel and Programme de Développement plement 1) S4­S13. Municipal (PDM). 2004. Assessing Local Economies 81 82 R E F E R E N C E S and their Prospects. Technical Document 1.3, Ecoloc Potential of the Labour Force in Africa: Barriers and Handbook, vol. 1. Opportunities. Paper prepared for World Bank Collier, Paul, and J.W. Gunning. 1999."Explaining African Labour Markets Conference (November 18­19, 2004). Economic Performance." Journal of Economic Litera- Draft. ture 37 (1): 64­111. Freire, Mila, and Mario Polese. 2003. Connecting Cities The Commission for Africa. 2005. Our Common Interest: with Macroeconomic Concerns: The Missing Link. Mon- Action for a Strong and Prosperous Africa. London. treal: Institut National de la Recherché Scientifique Cour, Jean-Marie. 2004. Assessing the "benefits"' and and the World Bank. "costs" of urbanisation in Vietnam. Annex to "Urban- Gasper, J., and E. Glaeser. 1998. "Information Technol- isation and Sustainable Development: A Demo- ogy and the Future of Cities." Journal of Urban Economic Conceptual Framework and its Application Economics 43: 136­156. to Vietnam." Report to Fifth Franco-Vietnamese Eco- Glaeser, Edward L. 1998. "Are Cities Dying?" Journal of nomic and Financial Forum, Ha Long. Economic Perspectives 12 (2): 139­160. ______. 2003. Sustainable development and poverty Glaeser, E.L., H.D. Kallal, J.A. Scheinkman, and A. Shleifer. reduction strategy revisited: a demo-economic con- 1992. "Growth in Cities." Journal of Political Economy ceptual framework and its application to Ethiopia. 100 (6): 1126­1152. The Ethiopian Economic Association, 2003 Interna- Grimm, Michel, and Isabel Gunther. 2004. Operational- tional Conference on the Ethiopian Economy. Addis ising Pro-Poor Growth: A Country Case Study on Ababa. Draft. Burkina Faso. Joint initiative of AFD, BMZ (GTZ, Dollar, David, and Aart Kraay. 2002. "Growth is good for the poor." Journal of Economic Growth 7: 195­225. KfW Development Bank), DFID and the World Bank. Doumani, Fadi M. 2002. Environmental Health in Sub- Henderson, J. Vernon. 2000. How Urban Concentration Saharan Africa: A Road Map. World Bank, Africa Affects Economic Growth. World Bank Policy Res- Region. Annex VIII. Draft. earch Working Paper 2326, Washington, D.C. Duranton, G., and D. Puga. 2002. "Diversity and Special- ______. 1999."Inter-City and Inter-Regional Patterns of isation in Cities: Why, Where and When Does it Mat- Production with Economic Development. Brown ter," in P. McCann (ed.) Industrial Location Economics, University. Draft. ed. P. McCann. Cheltenham, U.K.: Edward Elgar. Henderson, J., Zmarak Shalizi, and Anthony J. Venables. Eifert, Benn, Alan Gelb, and Vijaya Ramachandran. 2004. 2001. "Geography and development." Journal of Business Environment and Comparative Advantage in Economic Geography 1 (1): 81­105. Africa: Evidence from the Investment Climate Data. International Labour Organisation (ILO). 2002. Decent World Bank. Draft. work and the informal economy. International Labour Ellis, Frank, and Nigel Harris. 2004. New Thinking about Conference, 90th session. Urban and Rural Development. Keynote paper for ______. 2004. World Employment Report 2004-05. DFID Sustainable Development Retreat, University of Geneva. Surrey. Draft. International Monetary Fund (IMF). 2004. Government Ersado, Lire. 2003. Income diversification in Zimbabwe: Finance Statistics. Washington D.C. Welfare Implications from Urban and Rural Areas, Iradian, Garbis. 2005. Inequality, Poverty and Growth: World Bank. Draft. Cross-Country Evidence. IMF Working Paper WP/ Farvacque-Vitkovic, Catherine, and Lucien Godin. 1998. 05/28. Washington, D.C. The Future of African Cities: Challenges and Priorities Kaufmann, Daniel, Frannie Leautier, and Massimo for Urban Development. Directions in Development Mastruzzi. Governance and the City: An Empirical Series. Washington, D.C.: The World Bank. Exploration into Global Determinants of Urban Farvacque-Vitkovic, Catherine, Lucien Godin, Hugues Performance. World Bank Institute. Draft. Leroux, Florence Verdet, and Roberto Chavez. 2005. Kelly, K. 2003. Supporting Local Government Responses Street Addressing and the Management of Cities. World to HIV/AIDS: Positions, Priorities, Possibilities. Cen- Bank Directions in Development Series. Washington, D.C.: The World Bank. tre for AIDS Development, Research and Evaluation, Fay, Marianne, ed. 2005. The Urban Poor in Latin America. Institute for Social and Economic Research, South Directions in Development Series. Washington, D.C.: Africa. Paper presented at 2003 Urban Research Sym- The World Bank. posium, World Bank, Washington, D.C. Fay, Marianne, and Charlotte Opal. 2000. Urbanisation Kingdon, Geeta Gahdhi, and John Knight. 2003. Unem- without Growth: A Not So Uncommon Phenomenon. ployment in South Africa: The Nature of the Beast. World Bank Policy Research Working Paper 2412, Oxford University, Centre for the Study of African Washington, D.C. Economies. Fox, Louise, Gordon Betcherman, Vandana Chandra, Kraay, Aart. 2003. "When is Growth Pro-Poor? Cross- Benn Eifert, and Arvil Van Adams. 2004. Realizing the Country Evidence." Journal of Development Economics. R E F E R E N C E S 83 Lopez, Humberto. 2005. Growth and Inequality. Paper Rakodi, C. 1999. "Trans-border urban regions -- are they presented at Pro-Poor Growth Seminar, World Bank, emerging in Africa?" In Development and Urban Washington, D.C. Draft. Africa, ed. C. Stein, 199­219. Barcelona: Centre d'Es- Lucas, Robert E. B. 1998. Internal Migration and Urbani- tudis Africans. sation: Recent Contributions and New Evidence. Rodriguez-Pose, Andres, and Nicholas Gill. 2003."Is there World Bank, Washington, D.C. a Global Link between Regional Disparities and Devo- Madani, Dorsati. 1999. A Review of the Role and Impact lution?" Environment and Planning A 36 (2004): of Export Processing Zones. The World Bank Devel- 2097­2117. opment Research Group. PRWP 2238. Washington, Rosen, Sydney, and Jonathon Simon. 2003. "Shifting the D.C. burden: the private sector's response to the AIDS epi- Maloney, W. 1999."Does Informality Imply Segmentation demic in Africa." Bulletin of the World Health Organi- in Urban Labour Markets? Evidence from Sectoral sation 81 (2): 131­137. Transitions in Mexico." World Bank Economic Review Rosenthal, S.R., and W.C. Strange. 2001. "The Determi- 13: 275­302. nants of Agglomeration." Journal of Urban Economics Mboup, Gora. 2004. Lusaka, Zambia: Intra-City Differen- 50: 191­229. tials. Internal report. Nairobi: UN-Habitat, Monitor- Sachs, J., J. McArthur, G. Schmidt-Traub, M. Kruk, Chan- ing Systems Branch. drika Bahadur, Michael Faye, and G. McCord. 2004. Mboup, Gora. 2003. Improving Health Conditions of Ending Africa's Poverty Trap. Paper prepared for the Slum Communities: Health Interventions versus Slum Millennium Project, Draft. New York: Columbia Uni- Upgrading. Urban Poverty and Health in Sub-Saharan versity. Africa. Nairobi Urban Health Equity Gauge Conference, Sahn, David E., and. David C. Stifel. 2003. "Urban-Rural APHRC, Nairobi. Inequality in Living Standards in Africa." Journal of McKay, Andrew, and Ernest Aryeetey. 2004. Operational- African Economies 12 (4): 564­597. Centre for the ising Pro-Poor Growth: A Country Case Study on Study of African Economies. Ghana. Joint initiative of AFD, BMZ (GTZ, KfW ______. 2002. "Progress Toward the Millennium Devel- Development Bank), DFID and the World Bank. opment Goals in Africa." World Development 31 ( 1): McKinsey Global Institute. 2001. "India: The 23­52. Growth Potential." http://www.mckinsey.com/mgi/ Satterthwaite, David. 2002. Coping with rapid urban publications/India.asp. growth. RICS Leading Edge Series, Royal Institute of The Millennium Project. 2005a. Investing in Development: Chartered Surveyors. A Practical Plan to Achieve the Millennium Develop- ______. 2004a. The under-estimation of urban poverty in ment Goals. New York: UN Office of Secretary General low and middle-income nations. Poverty Reduction in and UN Development Group. Urban Areas Series Working Paper 14. London: Inter- ______. 2005b. A Home in the City: Task Force on Improv- national Institute for Environment and Development. ing the Lives of Slumdwellers. New York: UN Office of ______. 2004b. Urban poverty and inequality: How their Secretary General and UN Development Group. scale is hidden by aggregate data. London: Interna- Montgomery, Mark, Richard Stren, Barney Cohen, and tional Institute of Environment and Development. Holly E. Reed, eds. 2003. Cities Transformed: Demo- Unpublished manuscript. graphic Change and its Implications in the Developing South African Cities Network (SACN). 2004. State of the World. Washington, D.C.: National Research Council Cities Report 2004. of the National Academies, National Academies Press. Starkey, P., S. Ellis, J. Hine, and A. Ternell. 2002. Improv- Murphy, J.T. 2002. "Networks, Trust and Innovation in ing Rural Mobility--Options for Developing Tanzania's Manufacturing Sector." World Development Motorised and Nonmotorised Transport in Rural 30 (4): 591­619. Areas. World Bank Technical Paper No. 525. Wash- New Star Publishers, The Tenth Five-Year Plan of China ington, DC: The World Bank. (Summary). Beijing: China, 2001. Storper, Michael, and Anthony J. Venables. 2004. "Buzz: Office of the Deputy Prime Minister (OPDM). 2003. The Economic Force of the City." Journal of Economic Productivity in the UK. London: HM Treasury. Geography. 4. Prud'homme, Remy. 1994. On the Economic Role of Sub-Saharan Africa Transport Policy Programme Cities. Paper prepared for conference on Cities and the (SSATP), World Bank and Economic Commission for New Global Economy, organised by Government of Africa. 2002. Scoping Study: Urban Mobility in Three Australia and the OECD, Melbourne. Cities. SSATP Working Paper No. 70. Quigley, John M. 1998. "Urban Diversity and Economic Tanzania Planning Commission and Ministry of Labour Growth." Journal of Economic Perspectives 12 (2): and Youth Development. 1991. Tanzania the Informal 127­138. Sector 1991. Dar-es-Salaam. 84 R E F E R E N C E S Tewari, Deepali, and Sudeshna Banerjee. 2005. Lagos ______. 2004a. Le Secteur Urbain au Cameroun: Con- Strategy for Economic Development and Poverty tribution de l'IDA a la Stratégie du Gouvernement. Reduction: Preliminary Findings from Four Firm Central and West African Urban and Water Sector Level Surveys. World Bank, AFTU2. Draft. Unit Draft. Thompson, J., I.T. Porras, E. Wood, J.K. Tumwine, M.R. ______. 2004b. Dimensions of Urban Poverty in ECA. Mujwahuzi, M. Katui-Katua, and N. Johnstone. 2000. Energy and Infrastructure Department, Europe and "Waiting at the tap: changes in urban water use in East Central Asia Region. Africa over three decades." Environment and Urbanisa- ______. 2004c. République du Niger: Stratégie d'Appui tion 12 (2): 37­52. au Secteur Urbain. Central and West African Urban Tiffen, Mary. 2003. "Transition in Sub-Saharan Africa: and Water Sector Unit (AFTU2). Draft. Agriculture, Urbanisation and Income Growth." World ______. 2004d. Competitiveness and Private Sector Development 31 ( 8): 1343­1366. Development in Africa: Lessons from Investment Cli- UN-Habitat. 2003. The Challenge of Slums: Global Report mate Assessments. Office of the Chief Economist, on Human Settlements 2003. Nairobi: Global Urban Africa Region. Draft. Observatory. ______. 2003a. World Development Report 2003: Sustain- able Development in a Dynamic World. World Bank United Nations. 2004. World Urbanisation Prospects: The and Oxford University Press. 2003 Revision. Data Tables and Highlights. Department ______. 2003b. Urban Poverty in East Asia: a Review of of Economic and Social Affairs, Population Division, Indonesia, the Philippines, and Vietnam. East Asia New York. Urban Working Paper Series No. 11. United Nations Millennium Project. 2005. "Africa's Spe- ______. 2003c. Benin Poverty Assessment. Africa Re- cial Needs." Chapter 10 in Investing in Development: A gion, Poverty Reduction and Economic Management Practical Plan to Achieve the Millennium Development Network. Goals. New York. ______. 2003d. Burkina Faso: Risk and Vulnerability Venables, Anthony J. 2005. "Spatial disparities in devel- Assessment. Africa Region, Human Development oping countries: cities, regions and international Unit. AFTH2. trade." Journal of Economic Geography 5: 3­21. ______. 2003e. Senegal: Policies and Strategies for Accel- Webster, Douglas. 2004. Summary of Peri-Urbanisation: erated Growth and Poverty Reduction--A Country The New Global Frontier. Enschede, Netherlands: Economic Memorandum. Africa Region, PREM 4. International Institute for Geo-Information Science ______. 2003f. Lesotho Country Economic Memoran- and Earth Observation. dum: Growth and Employment Options Study. World World Bank. 2005a. "Africa's Growth Tragedy: An Insti- Bank, Report 26130-LSO. Draft. tutional Perspective." Country Note G in Economic ______. 2002a. From Action to Impact: the Africa Growth in the 1990s: Learning from a Decade of Reform. Region's Rural Strategy. Rural Development Opera- Washington, D.C. tions, the Africa Region. ______. 2005b. Global Monitoring Report 2005--Millen- ______. 2002b. Madagascar Poverty and Socio Economic nium Development Goals: from Consensus to Momen- Developments: 1993-1999. Report No. 23366-MAG. tum. Washington, D.C. ______. 2002c. South Africa Constraints to Growth in ______. 2005c. World Development Report 2005: A Better Johannesburg's Black Informal Sector. Africa Region, Investment Climate for Everyone. The World Bank and PREM 1, Report No. 24449-ZA. Oxford University Press. ______. 2001. République de Madagascar: Développe- ______. 2005d. Pro-Poor Growth: Country Experiences in ment Urbain Note de Strategy. Central and West the 1990s. World Bank, PREM Network. African Urban and Water Sector Unit. Draft.