DISCUSSION PAPER Report No. UDD-96 URBAN SUB-SAHARAN AFRICA IN MACROECONOMIC PERSPECTIVE: SELECTED ISSUES AND OPTIONS by Andrew Marshall Hamer April 1986 Water Supply and Urban Development Department Operations Policy Staff "he World Bank The views presented herein are those )f the author(s), and they 3hould not be interpreted as ref lecting ths,e of t-;e World Bank. Acknowledgment Among the many individuals who contributeti to improving the structure of the paper, several deserve particular credit: Michael Cohen (WUDOS); David de Ferranti and Carolyn Tager (WUDOD); George Gattoni (WAPUR); and Orville Grimes (EAPWU). Errors of omission or commission remain the responsibility of the author. Abstract Considerable attention has been directed in recent years toward the economic problems faced by the countries of Sub-Saharan Africa. Many bottlenecks have been identified and many solutions suggested. Yet, in spite of the fact that Sub-Saharan Africa's cities and towns house almost one-third of the region's population, and that over half of Gross Domestic Product is generated by the non-farm sector located in urban areas; little attention has been paid to the possible role of those centers in the process of structural adjustment and the renewal of sustained economic development. The paper argues that there are initiatives required at the subnational level without which macroeconomic reforms will remain incomplete. The urban agenda transcends physical planning and shelter. It lies at the heart of many of the intersectoral and geographic linkages that characterize both the economy and any successful investment programniing effort. Thus urban investment proposals should be judged by the contributions they make toward attaining critical development objectives, including those associated with agricultural develop- ment, industrialization, regional transportation, resource mobilization, investment planning, pricing reforms, and institution building. Table of Contents Page No. Summary ................. ..................................... v I. Introduction ................................................. A. A Statement of purpose ...................................1 B. The Challenge Created by Sub-SahLran Africa's Macroeconomic Constraints .....................2 II. Urban Centers in the African Economy ......................... 7 A. The Role of Cities and Towns in Economic Development ................................... 7 B. The Changing Level of Urbanization in Sub-Saharan Africa .................................. 8 C. An Overview of the Urban Productive Sector in Africa ...................................... 12 D. The Magnitude and Impact of Infrastructure Defi- ciencies on the Productivity of Cities ................... 18 E. Additional Constraints on Urban Business Activity: The Issue of Land Markets ............. ..... 26 F. Bottlenecks Created by the Failure of Capital Markets. ....................................... 28 G. Neglected Opportunities: Reinforcing Rural-Urban Links ..................................... 30 III. New Initiatives: Paths to Avoid ............................. 43 A. The Issue of Urban Bias ................................. 43 B. The Urban Sector After Structural Adjustment: A New Beginning ..........................46 C. "Excessive" Primacy as a Concern ........................50 IV. A Strategy for the Future .................................. 58 A. A Case for Strong Local Government ......................58 B. Evidence of Increased Local Self-Reliance ...............61 C. A Selective Agenda for Local Government .................. 63 D. The Need to Improve the Operation of Factor Markets ........................................ 69 E. The Urban Sector and Public Expenditure Reviews ............................................... 72 F. Next Steps .............................................. 74 -v - URBAN SUB-SAHARAN AFRICA IN MACROECONOMIC PERSPECTIVE: SELECTED ISSUES AND OPTIONS s',mmary and Conclusions Severe macroeconomic constraints are forcing the countries of Sub- Saharan Africa to re-examine many of their policies. As part of this reassessment and refocusing of effort, governments must deal with serious problems that exist at the subnational level, particularly in the cities and towns. One third of the population lives in those centers and they account for over half of the value-added generated in Africa. The efficiency with which the urban sector performs its tasks affects the success of strategies pursued in other areas, including agricultural development, industrialization, reaional transportation, public sector resource mobilization, and institution building. This report reviews selected issues uncovered through an examiniation of the urban productive sector and its links to the economy as a whole. In this summary section, the main recommendations for policy direction are discussed first, after which the principal issues that underlie these recommendations are outlined. A strategy for the future Over the last two decades central governments in the region have attempted to extend control over all aspects of economic life in an effort to enforce their vision of development. The result was a decline of institutional competence. One major victim of political centralization was municipal government. Local governments withered where once they were strong or failed to develop where local autonomy had not previously existed. Yet, today strong local government is necessary if the process of structural adjustment and reconstruction is to succeed. It is in the cities and towns that much national economic activity takes place; it is also at the local level that the main alternative to strong local government, i.e. the central government, has reached the limits of its competence. Municipal government must have the capacity and the opportunity to plan investment programs, implement them, and maintain the resultant infrastructure and services. The Center's role does not disappear in such a scenario; it is trimmed back and refocused by concentrating on oversiaht functions, and by administering loan facilities and incentive grants. In Sub-Saharan Africa onie model which has much to commend it is found in Zimbabwe, where local governments are strong and, not incidentally, municipal facilities and services are the best in the regionl. The gap between this model and the reality elsewhere is large. Central government and exte'rnal donor support will be needed for many years, as local institutions are rehabilitated and developed. But examples already exist, in many 3ank-funded projects, of increased municipal -ompetence, through greatly iniproved efforts at local resource mobilization. With this greater self-reliaice comes the possibility - vi - of fulfilling a key macroeconomic objective: the preservation and rehabi- litation of existing infrastructure and services, along with efforts to make more effective use of these existing assets. But the challenge to urban governments does not end there. The, must ease the adjustments to rapid growth through locally replicable solu-ions th.at increase new serviced land development and tap latent private initiative. At the rates of urban vrowth forecast over the next 15 years, urban Sub-Saharan Africa will grow by an average of 8 million to 10 million residents per year, or the equivalent of replicating Lagos plus Kinshasa annually. This will call for 80,000 to 100,000 serviced hectares per year for all uses. The costs of not responding to this challenge are evident across Africa, as both Lagos and Kinshasa demonstrate vividly. This task requires the coordinated introduction of affordable and unsubsidized infrastructure across thousands of cities and towns. In turn, this performance is only possible if certain conditions are met: a) the public sector must disengage from activities that subsidize the city dweller; b) new infrastructure standards are needed that are lower than those that typically prevailed during the construction of the colonial cities; c) there must be a serious effort to conserve public resources by encouraging private investments; d) a new system of intergovernmental relationls should focus on the creation of consolidated, self-financing infrastructure development funds which allow cities and towns to compete against one another for investment capital on equal terms; and e) institutional reforms are needed that will permit the development of minimally coherent city-wide development strategies. Success in these initiatives requires that attention be paid to factor markets, especially land. There is a serious need to develop land information systems that will not only make resource mobilization more effective, but will also allow entrepreneurs and households to freely trade land whose ownership is not in question. Land information systems can be devised which are progressively upgradeable. At its simplest, the responsible officials can record only the information needed for property taxation. More elaborate systems, that establish legal title, can be utilized, where feasible. As with land market issues, those involving loans to smaller-scale enterprises can be tackled if accumulated experience is used as a guide. Improvements in capital market access for smaller firms begin at the macroeconomic level. Like most forms of subsidization in poorer countries, subsidization of interest rates on loans to small business constrains access to funds by making such loans unremunerative to the supplier. Once the interest rate issue is resolved, there are many low costs ways of dealing with the problems cited. Referrals and formalities can be filtered through local extension services, community bank managers, and trade groups or other nongovernmental organizations. Collateral bottlenecks can be alleviated by the development of guarantee or insurance funds to cover such loans. Financial institutions can also use self-selected mutual security group loans, where funds are guaranteed against default by several potential borrowers. Other features of the loans can be adapted to encourage a good credit record. Problems with arrears can be dealt with by the use of substantial penalties and aggressive follow-up, leading to a restructuring, rescheduling or foreclosure of loans. The record of such carefully-crafted systems is generally very good, with high rates of return to income and with substantial employment effects. Because of the strategic importance nf urban Ereas in the national economies of the region, it is important thaL the sector be considered in major analyses of investments such as those performed by the Bank in its public expenditure reviews. Bank reports vividly mirror the serious boundary problem which obstructs the identification of intersectoral linkages both in the economy and in investment programming. Thus, for example, there are few attempts to coordinate urban investments by area with the expected geographic focus of rapid agricultural growth. Both in the countries themselves and within the donor agenicies more effort is needed to identify which cities and towns have key developmental potential in support of agriculture; which policies and institutions help or hinder the efficient symbiosis between town and country; and what bottlenecks exist that prevent the more efficient funtioning of these towns in support of rural areas, especially with regard to infrastructure and services. The same conclusion can be reached about linkages between urban areas and industrial, service, and artisan activities in general. With few exceptions, these ties are ignored in the preparation of macroeconomic adjustment packages, even when major urban bottlenecks to efficient operation of these subsectors are explicitly identified. In addition, the heavy volume of investments in interregional transportation facilities appears to be taking place in ways that are delinked from any urban development strategy. Urban centers in the African economy What then are the issues that prompt this set of recommendations? One should start with a re-examination of the conventional wisdom that views the urban economy as somehow once removed from the development process in Sub- Saharan Africa. Even when national development is based on an agriculture-led strategy, and the majority of the labor force is working on farms, most value- added will be generated by non-farm activities located in settlements of varying size. This feature will become more evident with continued economic development. The evolution of the macrozconomy cannot occur without a continuous restructuring of the spatial economy, favoring cities. The increased monetization and specialization of productive activity undergirds the growth of cities and towns in Africa. Using a standardized definition of urban settlements, one finds that the population of Africa's cities and towns grew from 28 million to 103 million between 1960 and 1980. Various forecasts suggest a further tripling of the urban population by 2000. At that time almost half of the population will live in the region's cities and towns. There is little to be gained by trying to stop this trend. There is much to lose in not seizing the opportunity to ensure that such growth contributes maximally to overall development goals. In Africa up to two-thirds of all non-farm employment is in rural areas, defined to include towns of up to 20,000 to 30,000 residents. In those areas roughly 40% of all workers have non-farm employment as their principal occupation. Given the undeveloped nature of rural infrastructure and the dispersed nature of the population at large, small cities and towns are critical hubs of local economic development. - viii - As one travels up the size hierarchy of cities and towns, enterprises are exposed to new potential benefits. The concentration of population tends to cut the costs of providing access to infrastructure and public services, making their installation more affordable and, therefore, more likely. Small manufacturing, repair, and service facilities begin to adapt to the availability of such services. In addition, many ventures begin to experience cost reductions because other activities, external to the enterprises, are in close proximity. Even the scope of tbhe market expands, as l-arger city are surrounded by larger dependent hinterland areas. Larger firms thus become more evident. Nevertheless, the importance of small firms continues to be felt, even in large cities. Such centers have an important percentage of their labor force working in small, usually unregistered enterprises. What's more the role of this so-called "informal" sector can only be enhanced by the structural adjustment process now underway in Africa. A declining proportion of new increments to the labor force can be expected to rely on new jobs provided by larger entet?rises, including those controlled by the public sector. By contrast, the small-scale firm already operates within a relative price framework that farors labor-intensive production. It is also the type of activity that most closely caters to the large and growing demand for customized and/or low-cost goods and services in these centers. Yet the smaller-scale enterprise cannot shoulder the full burden of supplying non-agricultural goods and services in these countries, and it is in the larger cities that one finds the bulk of medium-to-large firms. There, the opportunities for agglomeration economies are greatest. In addition, relatively easy access to the monetized economy favors larger cities. Furthermore, during the early stages of development, governments are likely to concentrate infrastructure in a few large places, so that at least these economic "city states" can service the widest possible range of enterprise demands. Only at a later stage of development will a portion of larger-scale productive activities tend to avoid operating in very large centers whose scale and associated costs are higher than justified by the clustering advantages they can reap. The issues facing the larger-scale firms will not be overtaken by the process of structural adjustment. Even under the most pessimistic assumptions, the bulk of existing firms will survive and be joined by additional businesses in at least somie productive sub-sectors. The magnitude and impact of infrastructure deficiencies on the productivity of cities One of the key problems facing the urban economy in Africa is the state of local infrastructure. This, in turn, is merely a symptom reflecting the neglect of local government in these countries, as well as the failure to mobilize resources at the local level. How bad are infrastructure conditions in Africa's cities? A review of the state of such facilities and services suggests conditions are often alarming. All the larjer centers have a core area, inherited at Independence, where infrastructure conditions are relatively good, in spite of those cities' need to accomodate much higher densities than originally foreseen. Elsewhere neglect has beerL the rule; the post-Independence peripheral neighborhoods are often devoid of more than token - ix - infrastructure coverage. Though this problem is best documented for large cities, it afflicts smaller centers as well. This general deficiency in the supply of infrastructure has several consequences for the productive sector. In differing degrees, enterprises depend on public infrastructure and services to permit land development (e.g. road access), to facilitate day-to-day operations (e.g. public safety), and to operate plant and equipment (e.g. power). The deficiencies can sometimes be compensated for, especially where larger firms are concerned; but the cost is considerable and usually exceeds the value of taxes and user charges needed to finance public provision. Often private substitutes are simply not competitive and the resulting mix of goods and services at the lccal level is either less diversified that it need be or the technology utilized is more primitive than it would otherwise be. In all the cases there are serious costs not only to the individual urban areas but to the economy as a whole. Additional constraints on urban business activity: land and capital markets A major problem facing the business community involves distortions in the markets for complementary inputs like land. Land markets tend to be disorganized, with conflicting, often unrecorded, ownership claims and unclear boundaries characteristic of most land units across Africa's cities. Without an adequate system for measuring and recording the boundary of land parcels, and without the registration of legal rights to such plots, it is impossible for a potential entrepreneur to go to any one or two sources in a city and find out what land is available for sale, what prices prevail in the market, and who the owners are. Property transfers are fraught with long-term risks, and transactions are time consuming, and may involve payments to several purported owners. The development of streamlined, effective procedures for transfering land rights and titles deserves the type of attentLon ordinarily accorded physical infrastructure alone. Small and medium-sized enterprises also face serious problems created by the fragmented nature of capital markets in urban areas. Typically, smaller businesses are restricted to their own resou::ces, borrowed funds from friends or relatives, or money raised at very high cost from money lenders. Good economic development projects are therefore either underfunded or not undertaken at all. Both project-specific jobs and employment indirectly stimulated by the project, may be foregone. Opportunities are also diminished to mobilize personal and family savings through the availability of matching loan funds. Neglected opportunities: reinforcing rural-urban links The exploration of rural-urban links has become an object of increasing attention among students of development. Their studies stress the degree that rural households involved in agricultural production can draw on nearby towns, and the network of cities and towns connected to them. Input requirements are met through that network. Cities and towns act as marketing hubs that may reach out to regional, national, and international markets. This marketing function involves operations such as grading, storage, assembly for shipment, distribution to wholesales, and sales by retailers. The production and sale of agricultural commodities is not an end in itself; there must be "incentive" consumer goods and services, as well as savings and -x - investment opportunities. While not every option need come from area urban centers, those locations may be significantly more accessible than the rest of the urban system and thus play an important role in shaping rural household perceptions on available goods, services, and investment opportunities. Finally, public services provided by these urban centers are also accessible, at relatively low cost, to the region's rural community. Thus a substantial proportion of the urban population in Sub-Saharan Africa derives its livelihood primarily by serving the rural producer, directly or indirectly. One estimate suggests that up to 90 percent of the population in towns of less than 20,000 residents is so engaged; on a regional scale, 7 out of every 10 urban residents are linked, in some degree, to the primary sector. One final linkage bears mentioning. Across Sub-Saharan Africa there s evidence that rural-urban migration forms part of a rural household strategy to diversify sources of income. A review of existing data suggests that most migrants regularly reinit goods or money to rural areas and that these remittances represent between 10% to 20% of migrants' estimated household income. In addition, urban migrants often return to their rural homes, bringing experience and resources with them. This type of income stream is more than an insurance scheme; it can play a major role in promoting farm innovation. Such innovation is often foregone because self-financing out of current income is impossible and credit is difficult to obtain. Remittances help atenuate the perceived risks of innovation, while providing a source of financing. The same urban-based income helps to inject cash resources into the rural economy and to finance the bulk of non-farm entrepreneurial ventures. This wide variety of rural-urban linkages doesn't appear automatically. In fact, the region is full of examples of policies and institutions which short-circuit these interactions to the detriment of town and country. A prime culprit is the macroeconom:.c environment that has typified conditions across many of the region's couintries. But correct relative prices are not enough. Farmers react to a whole package of incentives, moct of which are linked to the adequacy of urban-based services. Sometimes the bottlenecks involved cannot be resolved at the local level. Ready access to urban marKets is a major factor in explaining the productivity of farmers. One can conceive of a network of urban centers whose demand impact on farmers in any locality is proportional to each city's size and inversely proportional to the difficulty of reaching it. At great distances from urban centers, isolation results in little more than subsistence farming. Around major urban centers one finds prosperous farm belts. In between, the state of local and interregional transport links will help to determine the degree to which farmers produce for the market. Both rural and urban development investment decisions are thus dependent on the strategies followed by regional and national transport planners. In other instances, the probleris restricting rural-urban links are susceptible to resolution at the regional or local level, but only after an appropriate reform in intergovernmental relations initiated at the national level. At present, most localities cannot tap local wealth and mobilize substantial resources at the local level to finance, directly or through - xi - loans, infrastructure programs that will provide electric power, piped water, improved schools, or better local roads. These communities cannot develop the infrastructure base necessary to create more investor confidence. As a result the local multiplier affect associated with local spending is weak. Additional bottlenecks are caused by deliberate government efforts to replace the private sector operating in cities and towns with highly centralized and vertically integrated parastatals that bypass urban areas. Most parastatals involved in agricultural development are based in the capital city and local offices, if located in secondary centers, are delinked from local activity. Area cities and towns thus have a moribund commercial and manufacturing sector, with little financial intermediation through local institutions. The international donor community has, in the past, been guilty of supporting these types of parastatals. In addition, there is evidence that even rural development projects financed with donor aid have ignored the urban network. Because these projects and institutions have proved inefficient, they are now under review. Agricultural parastatals face drastic restruc- turing; as a result the private sector, operating from cities and towns, could gain new importance. More generally, structural adjustment measures will lay the groundwork for an increase in commercial agricultural production, using more inputs and equipment than in the past. This suggests an enhanced role for cities and towns in any agriculture-led development strategy. The issue of urban bias and future urban development investments The term "urban bias" is meant to encompass all national policies that have been biased against agricultural development. Included are distorted relative prices, excessive directly-productive investment in cities, and inadequate cost recovery from infrastructure investments made in urban centers. There is ample evidence of all three types of policies in Africa. This having been said, one would still find disparities even in the absence of biases. The very process of economic development involves the emergence, and persistence over long periods, of output and income per capita differences between rural areas, between urban centers, between regions, and even between rural and urban sectors, taken globally. These desequilibria are apparent at all stages of development except -- at one end -- in a subsistence economy and -- at the other end -- in a fully developed economy. For that reason, it is meaningless to assume that investments per capita, be they in infrastructure or directly productive activity, should be roughly equivalent between town and country, or even between different sized centers or different regions. Getting rid of "urban bias" does not lead automatically to the adoption of misplaced egalitarianism. Furthermore, the discussion of "urban bias" in Sub-Saharan Africa needs to be updated to take into account of the massive changes underway today. Macroeconomic policies are being redirected and now increasingly support an agriculture-led strategy. In addition, financial constraints, now and the foreseable future, will make subsidization of urban infrastructure, urban public services, and urban-based parastatals, I:Qs and less attractive. The very foundations of "urban bias" are thus being destroyed. Wihat is the likely future of the urban sector in the emerging age of agriculture-led development strategies? With the restructuring of production, there will be a shift toward labor-intensive, agroprocessing activities, and a shift away from import-intensive and capital-iitensive outputs. This should encourage the growth of jobs in secondary urban centers, whose links with agriculture will increase. A redistribution of incomes, away from "formal" sector employees in and out of government, and toward farmers and the "informal" urban worker, should generate new demand for "basic" urban goods and services that rely relatively heavily on domestic inputs and labor- intensive technologies. Finally, the urban sector will share the benefits of a rise in domestic savings, due to the removal of subsidies and price controls. This will help boost economic expansion and that, inevitably, will increase the demand for urban products. It is thus possible to foresee an economic basis for demographic projections which forecast a continuing growth of the urban sector at rates that exceed natural population growth. The structural adjustment process lays the foundations for growth relatively free of the problems -enerated by "urban bias." That growth, however, can only be fully exploited if urban institutions are made to function more effectively. The high political visibility of the urban sector, especially the largest cities, provides an excellent arena for dramatic demonstration of the benefits of continued structural adjustment measures. Policies which have promoted inappropriate investments, inequitable subsidies, and unrealistic standards in cities can be modified, thus freeing up scarce resources for priority spending. The city can become a laboratory where more private initiative -- in business ventures, shelter construction or public services -- can be harnessed and public initiative can be redirected. To refuse this opportunity to tackle the urbaK agenda, out of fear of residual "urban bias," is irresponsible and or guarantees a continuation of practices which must be done away with, sooner or later. The issue of "excessive" primacy and future urban development investments Often fears about urbanization are mere proxies for concern about the size and growth of "`ery large" cities. Here the discussion dwells on three areas: the presumed high level of unemployment in large cities; the assumed upper limits to "optimal city size;" and the apparent need for a "balanced" system of cities, often reflecting the optimality of the so-called Rank Size Rule. None of these reasons has any merit. Without denying that the present structural adjustment crisis is creating unemployment problems in urban areas, most visibly in large cities, the fact remains that prolonged unemployment is not an option except for secondary household workers and school-leavers from relatively affluent households. Attempts to limit the size of relatively large cities by denying them investment funds are often cloaked with references to "optimal" city size. Sometimes optimal size is associated with attempts to minimize the per capita costs of public services and infrastructure. Among the many problems of this - xiii - rationale is its failure to consider the benefits associated with growing city size and its confusion between the costs oc affordable service standards and those of minimal infrastructure, at each city size. Furthermore choosing city size so as to maximize net benefits is not a viable alternative. "Net benefits" cannot be measured carefully enough to be made operational. In addition, whatever the net benefits associated with city size may be, they will vary markedly over time, in response to the changing location and size of other cities, the level of national development, the sectoral composition of output, and the functions performed by the city in question. Much the same can be said about the use of the so-called Rank Size Rule to identify cities that are too small or too large. The Rule states that each center's population should equal that of the largest city divided by the smaller center's size rank. As such it is one of an infinite number of possible size distributions; it has no prescriptive value whatsoever. A look at the facts themselves suggest little ground for alarm concerrning the largest cities. There are only 28 cities with more than 500,000 people in a subcontinent of 400 million people. Of these only nine have more than a million inhabitants. For those that worry about the spatial pattern of development, it is better to work for a set of efficient macroeconomic policies. If the policies in place allow producers to choose their production decisions so as to minimize the "scarcity" costs of producing their desired bundle of goods and servIces, then the resulting cities will approach their efficient size, whatever that might be. Next steps When the Bank's urban program began in 1972, a deliberate decision was made to follow an experimental approach, one of "learning by doing." This effort embraced the experience from mistakes as well as successes. Today much is known about preferred technical solutions to particular city infrastructure and public service problems. There is a firm basis in experience to deal with complex institutional and resource mobilization issues. At the same time, new issues have arisen that were not anticipated in the 1970s. The most critical new sectoral problem in Africa is how to integrate an urban perspective into the process of structural adjustment and the renewal of growth. While a spatial perspective is needed, the explicitly spatial instruments that were advocated in the past have proved banikrupt. Growth poles, migration controls, infant region promotion, and optimal city size controls -- all these have been tried and found wanting. Now what is needed is to test out impact analysis tools which seek to answer questions about how different urban initiatives might help to achieve objectives of rural development or transport planning; industrial development or export. promotion; resource mobilization or rehabilitation of infrastructure. By incorporating a broader perspective into the process, planning for rehabilitation and development in Africa would be more realistic and more rewarding. It would no longer be possible for "gatekeepers" to assign to the urban sector a percent of Bank or country investment funds based on past shares or on the relative prioritv granted narrow sectoral objectives treated in isolation. Instead, investment priorities should be determined first by - xiv - the contribution each proposal, regardless of sector of origin, makes toward attaining critical development objectives. In the future the urban sector should be allowed to compete for funds on the basis cf its demonstrable i.mpact on the success of investment programs in other, apparently unrelated sectors. I. Introduction A. A Statement of Purpose In recent years the economic difficulties facing Sub-Saharan African countries have received considerable attention. Much effort has been devoted to identifying bottlenecks and suggesting solutions. Yet, in spite of the fact that almost one-third of the area's population lives in cities, and that over half of Gross Domestic Product is generatea in urban centers, little attention has been focused on the contribution made by the urban econony to overall national development. This paper argues in favor of harnessing the human and financial resources of the African city as part of any strategy to ensure greater, and more widespread, economic development. What's more, it suggests that there are initiatives required at the subnational level without which measures being taken at the macroeconomic level will remain incomplete. The urban sector can no longer be considered merely as a social sector. The urban agenda goes well beyond physical planning and shelter and includes aspects of agricultural development, industrialization, regional transportation, resource mobili- zation, investment planning, pricing reforms, and institution building. In each area, the efficiency with which the urban sector performs a range of tasks will be vital to the success of policies elsewhere. This section provides an overview of the key macroeconomic constraints facing the region, and summarizes the principal objectives of emerging macroeconomic policy. Section II describes some of the key elements of the urban productive sector and underlines the types of constraints tnaE keep urban areas from making a more substantial contribution to achieving the cited macroeconomic goals. Sections III and IV are prescriptive in characcer, -2- and detail both useful and harmful measures that might be taken in relation to the urban economy. B. The Challenge Created by Sub-Saharan Africa's Macroeconomic Constraints The economies of Sub-Saharan Africa vary enormously in scope. l/ Of the 39 states considered to comprise the area, six have fewer than one million inhabitants and one, Nigeria, has over 93 million. Per capita incomes range from under $200 for six countries to over $1000 for three others. Gross Domestic Product (GDP) is less than one billion dollars in twelve states and over b4 billion dollars in Nigeria. Nevertheless, there are certain common- alities among these nations which, as a whole, have a population of 4UO million and generate GDP of 200 billion dollars. After a post-Independence decade of Lz.onomic promise, roughly 60% of the countries experienced either a decline in per capita output or virtual stagnation. 2/ Furthermore only ten countries managed to improve their GDP growth rate performance from 1973 to 1983 compared to the previous decade; with eight minor exceptions, all states also experienced increases in population growth rates over the two periods, compounding the problem. This deterioration of economic performance has affected not only agricultural and industrial output, but also the quality of infrastructure services and the financial and managerial viability of key development institutions. The causes of this decline and the advisaDility of different I/ Data drawn from the World bank's World Developmeiit Report 19b5 (Aew York, Oxford University Press, 1985), and Toward Sustained Development in Sub- Saharan Africa: A Joint Program for Action (Washington, 1964). 2/ This is based on 32 countries for which the necessary data breakdown is available in the cited World Development Report. Declines were experienced by 15 countries and marginal gains by four more. -3- reform measures have been discussed in various Bank reports and those conclusions are worth restating briefly, for they provide the context for some of the discussions that follow later in the text. 1/ The situation faced by each country in the region has its own individual features, but the Bank reports suggest that the recent difficulties often have similar exogenous and domestic causes. Among the exogenous factors are the oil price hikes of 1973 and 1979; the boom-bust cycle afrecting the prices of many primary exports; and the aggressive international lending practices of private banks, following the rapid growth of the Eurodollar market. These factors interacted with one another. The commodity price boom that followed the 1973 oil price increases encouraged area govern-e1its to sharply boost expenditures at home and borrowing abroad. Once commodity prices fell, there was additional recourse to external credit. Hany countries therefore worked themselves into a precarious economic corner when the 1979 oil price increases occurred. This event was followed oy a world recession, an increase in international interest rates, and widespread political instability and drought across mucn of Africa, all of which inflicted further damage. The Bank reports suggest that national adjustment policies were not comprenensive enough nor enacted rapidly enough to avoid a serious international debt servicing problem. This forced the adoption of restrictive import policies, compounding an already serious showdown in economic activity. 1/ Among the reports are two major Bank reports: Accelerated Development in Sub-Saharan Africa: An Agenda for Action (Washington, 1981) and Towara Sustained Development in Sub-Saharan Africa..., op.cit. See also K. Krumm, The External Debt of Sub-Saharan Africa: Origins, Magnitude, and Implications for Action, h4ashington, World Bank Staff working Paper Jo. 741, July 1985; and J. Nellis, "Public Enterprises in Sub-Saharan Africa," Washington, Public Sector Management Unit, Projects Policy Department, The World Bank, June 1985, processed. -4- The various studies point, as well, to endogenous factors. In many countries there was an overextension of the central government -- beyond the limits of its financial and managerial capacity -- as attempts were made to command and regulate large fractions of economic activity. Market signals, such as interest rates, foreign exchange rates, and relative prices, were often controlled. These tended to curtail incentives for private production, especially amonig primary producers. Many governments felt the necessity to expand their involvement in marketing and productive activities, partly through the use of parastatal enterprises. To finance these ventures and to build other facilities, such as infrastructure projects, large public investments were undertaken that appeared ill-advised in retrospect, partly because they often yielded rather limited returns. The Bank studies propose that two major macroecotnomic objectives dominate policy-making in the region for the foreseable future: increased production of output for both domestic and foreign markets; and the reduction in central government activities to more sustainable levels. This leads the reports to put forth four major sets of reforms for the region's economies. First, they recommend that market signals and private producers be given a grea,er role. They include, under this item, the dismantling of marketing controls exercised by parastatal monopolies involved in agriculture. A second proposal involves the parastatal sector more generally; suggestions are made that many be abolished or curtailed, and that those remaining receive assistance in such areas as financial management and manpower training. A third proposal calls for the slimmed-down public sector to give priority to expenditures for operations and maintenance of essential infrastructure and public services, with increased private sector participation, where possible. This would be accompanied by selective rehabilitation and removal -5- of bottlenecks, to allow fuller use of existing capacity. A final major concern of the Bank docuiments is that great care be given to the allocation of new investment funds, which are likely to be limited for a long time to come. While some may argue that this Bank vision is too ambitious to serve as a blueprint, one could point to at least one area where the reports don't go far enough. They largely ignore urban areas and the local institutions that must manage them. That is unfortunate because most problems and most solutions have important cross-sectoral and geographic implications, and these usually involve cities and towns. Among the historical antecedents of the present crisis that are ignored, for example, is the evolution of local government in Sub-Saharan Africa. The conventional wisdom in the revion was that centralized deeision- making was the key to rapid socio-economic change and to developnMent freed from parochial self-interest. Local autonomy became identified with the notion of competing centers of power and patronage, at a time when pluralism was increasingly unacceptable. Finally the concern with local administrative inefficiency was addressed not by an active defense of local institutions but by a strategy that assumed sucn shortcomings could be avoided by centralization. Instead the result was waste and ineffective governance, and a massive neglect of the local infrastructure networks and services wltrout which the other "sectors of the economy cannot operate effectively. To resolve Africa's economic problems, a new sensitivity will be needed, among donors and governments alike, concerning the role of cities and towns. The degree to which the urban environment is supportive of productive activities will affect significantly the attairnment of agricultural and industrial sector objectives. Efforts to increase resource mooilization, cut -6- subsidies, and redirect expenditures will be quite meaningless unless the urban agenda is addressed. Future Bank reports should therefore reflect a greater concern for the problems and potentials of cities in Sub-Saiaran Africa. -7- II. Urban Centers in the African Economy A. The Role of Cities and Towns in Elconoinic Developmenlt A considerable portion of economic activity takes place in cities and towns, even in lower-income countries. The reason for this is straightforward. All productive processes involve the acquisition of inputs, and the transformation of these inputs into outputs, which are then consumed or marketed. Even when national development is oased on an agriculture-led strategy, -and the majority of the labor force is working on farms, most value- added will be generated by non-farm activities taking place in sectlements of varying size. Furthermore, the reliance on cities and towns will increase markedly over time. The evolution of the macroeconomy cannot occur wichout a continuous restructuring of the spatial economty, favoring the cities; people and jobs cannot simply increase in situ, preserving the low settlement densities of an earlier age. The reason for this spatial transformation can be found by analyzing the emergence and increasing dominance of the market economy. In a "Robinson Crusoe" economy, all activity is for own household consumption. In such a subsistence setting, economic roles are largely undifferentiated and there are few, if any, incentives to live in cities and towns. Economic development implies the increasing specialization of economic activity under- taken by individuals and the growth of production for sale. This monetization and specialization of activity requires the growth of cities and towns; i.e., permanent groupings of specialized househlolds and firms concentratea in limited areas. From that moment forward, incentives will exist for both labor and capital to migrate between rural areas, between cities and ruraL cencers, and between cities of different sizes.l/ B. The Changing Level of Urbanization in Sub-Saharan Africa Most of the Sub-Saharan countries became independent in or around 1960. At that time urbanization rates were among the lowest in the world, reflecting, among other factors, colonial policies which restricted free uiovement of the non-white population. Using a consistenit definition of urban residence across all the relevant countries, one major study reveals that the population of African cities and towns rose from 28 million to 103 million between 19b0 and 1980. 2/ The rate of growth was twice as fast as that for the population as a whole, and the percentage residing in urban areas doubled to 29%. That same study forecasts that, by 2OUO, the urban population will have tripled when compared to 1980 totals. At that time 46% of the region's population would live in urban centers. United Nations forecasts, 3/ using 1/ This interpretation is now part of the conventional wisdom within the World Bank. See, for example, the World Development Report, 1982 (New York, Oxford University Press, 1982): "The transfer of labor from agriculture (where the amount of capital per worker is relatively low) to industry and services (where capital per worker and average productivity are relatively high) is the key to raising incomes and output. The higher average productivity of labor in the non-agricultural economy is reflected in income differences between it and agriculture.... This income difference provides the incentive for people to move out of agriculture into non-agricultural activities in urban centers. The agricultural shares of output and employment are roughly equalized only at a comparatively late stage of development." 2/ SCET International, SCET AGRI, SEDES, Une Image a Long Terme de l'Afrique au Sud du Sahara, Paris, Commission des Communautes Europeenes et Caisse des Depots et Consianations, January 1984. The cited data excludes the Republic of South Africa. The study identifies all settlements with 5000 or more residents as urban. 3/ Estimates and Projections of Urban, Regional, and City Populations, 195u- 2025: The 1962 Assessment, New YorK, Department of Lnternational Economic and Social Affairs, United Nations, 1985. -9- country-specific definitions of urban residence, yield different absolute totals. Yet, they agree that, between 1980 and 200U, the urban population will triple. Though many may bemoan the level of ur:,anization in the region, it is not at all clear that a more efficient macroeconomic or sectoral policy environment would decrease urbanization levels. In fact, when levels of urbanization today are regressed against explanatory factors such as per capita national income, Sub-Saharan Africa behaves in ways comparab,e to other regions. 1/ By and large, the countries the Bank identifies as having low incomes also have low levels of urbanization, and usually these fall below the 29% region-wide level for 1980 identified in the ILTA study. 2/ The exceptions, including Somalia, Zaire, Central African &epublic and Ghana, nave all experienced drastic declines in the rural economy. These particular cases may utidergo a rural renaissance that will slow the relative shift of population to the cities; for the rest, there is little reason to fault the level of urbanization at their level of development. Once again, the middle income oil importer countries, with an average rate of urbanization slightly aoove the ILTA reaional average, seem to experi- ence levels of urbanization not unlike comparator countries in other regions. If anything, there are several cases of countries with unusually low levels of urbanization. Zimbabwe, with only 24% of its population living in 1/ M. Montgomery and E. Brown, Miigration and Urbanization in SuD-Saharan Africa, Washington, Population, healtn and Liutrition Department, fhe World Bank, processed, April 1985. 2/ SCET International, et.al., Une Image a Long Terme de l'afrique au bud du Sahara, op.cit. - 10 - urban areas, is about 20 percentage points below its per capita income peers. 1/ This has as a procimate cause the colonial and neo-colonial government policy of restricting black African access to urban areas, and of bottling up the majority in so-called Tribal Trust or communal areas. 2/ In studies of Zimbabwe, as well as in other countries like Kenya, one finds concern that urbanization may not proceed fast enough to meet rural development objectives: If the annual rate of employment growth in the urban areas [over the next twenty yearsj were only 3.57., then effectively the [communal areas]... may still have to support over twice their present population in 20 years time implying per capita incomes falling to only 66% of their present level. 3/ Finally, among the middle income oil exporters, there appear to be no surprises. Nigeria's apparent low level of urbanization is an statistical artifact; using the uniform definition adopted by the ILTA study, it appears that Nigeria's urban population already accounts for 46% of the total popula- tion, a level matched by its other oil-export collegues. Some of the concern with "excess urbanization" appears to reduce itself to a complaint about "excess migration." Unce again, there seems to oe little in the African experience that is unusual, given its overall national population growth. Because, at earlier stages of development, migration I/ Other examaples include Swaziland a.d Lesotho, which appear to rely on South Africa's cities for the "missing" urban functions. 2/ These areas are characterized by the limited nature of land tenure rignts, as well as by the relatively low quality of the land. 3/ Zimbabwe Agricultural Sector Study, Washington, Easrern Africa Projects Department, Southern Agriculture Division, The World Bank, Report No. 4401-ZIN, December 1983. For Kenya, the World BanK report, Kenya: Growth and Structural Change, (Washington, 1983) notes "Even if nonagricultural employment were to quadruple from 1976 to 2UOU, agricultural employment would have to more than double... simply to prevent unemployment from growing." contributes halt or more of the aggregate growtn of urban areas, 1/ it is not surprising that the overall rate of African urbanization - 6.1% per annum -- bears a similar relationship to its overall population growth, which is averaging 3.3% per annum. 2/ Exceptions to this are easy to explain. Across Sub-Saharan Africa those countries with both above average urbanization rates and above average rates of migration number nine in total and include war- ravaged countries (Chad, Mozambique, Angola) where security in rural areas is poor; a few countries where the rural sector has collapsed (Zaire, Tanzania, Guinea, Niger); and two countries where rapid urbanization is linked to rapic economic growth (Ivory Coast, Cameroon). In any case, there is littie that can be done to contain migration. Draconian measures such as those used in Tanzania are politically unpalatable and very difficult to administer. 3/ Even rural development nas not proved an effective way to significantly reduce migration. 4/ What is needed is the political will to accomoaate the urban growtn that is inevitably coming to the region. Whether one relies on the United Nations' forecasts or on the assumptions of the ILTA study, population growth in Sub-Saharan Africa from 1980 to 2000 will add between 160 million (UN) and 1/ Migration, Population Growth, and Development, Baltimore, Population Information Program, Johns Hopkins University, Population L(eports Series U1, Number 7, September-October 1983. 2/ Towards Sustained Development in Sub-Saharan Africa..., op.cit. 3/ See, for example, the discussion of the 1983 Human Kesources Deployment Act in L. Shaidi, "Tanzania: The Human Resources Deployment Act 1983--A Desperate MLeasure to Contain a Desperate Situatin," Keview of African Political Economy, December 1984. 4/ Migration, Population Growth and Development, op.cit. - 12 - 200 million (ILTA) new residents to the region's cities and towns. -/ The urban population and its problems will not go away. C. An Overview of the Urban Productive Sector in Africa In discussions of African economic development, the concern for ,rehabilitating the agricultural sector sometimes leads analysts implicitly to slight the importance of the non-farm economy. Set the 2vY00 African cities and towns with population of 5000 or more play a major role in the region's development. 2/ At the lower end of the settlement spectrum, rural non-farm enterprises, most of them small in size, 3/ foster the arowing specialization of their area economy, as farm and village tiousehold production for own use gives way to market exchanges. In Africa up to two-thirds of all non-farm employment is in rural areas with towns of up to 20,000 to 30,000 residents 4/ in their midst. In such areas roughly 40% of all workers have non-farm employment as their principal occupation. Given the undeveloped nature of rural infrastructure (roads, power, water) and the dispersed nature of the population at large, small cities and towns are critical hubs of local economic development. It is also here that households with access to little 1/ Estimates and Projections of Urban, Rural and City Populations..., op.cit; and SCET International, et. al., Une Image a Long Terme..*, op.cit. 2/ J. M. Cour, Urbanisation, Croissance Economigue et Developpement; 'aris, SCET International, processed, 1985. 3/ In terms of employment, these are small enterprises have 0U or fewer employees, with most having fewer than 10. 4/ This and related data are presented in D. Anderson, &uril Enterprise and Non-Farm Employment, Washington, The World Bank, January 1978; E. Chuta and C. Liedholm, Rural Non-Farm Employment: A Review of the State of the Art, Department of Agricultural Economics, ˇlichigan State University, Rural Development Paper No. 4, 1979. - 13 - or nlo agricultural land find the closest opportunity ror additional income, thus adding an equity dimension to such rural development. The motive force behind growing non-farm rural employment is the high elasticity of demand for rurally-produced non-farm commodities among farmers. 1/ Apart from the marketing services, these town dwellers produce plows, irrigation pumps and motors; they repair equipment; and they provide "incentive" goods and services including clothing, leather goods, furniture, and metal working. It is important to emphasize that a large part of these uroan commodities are actually produced at the local level, since the markets for the products of the dominant suppliers, the small enterprises, are hignly localized. 2/ For both goods and services, marketing depends basically upon personal exchanges between the businessman and the final customer. As one travels up the size hierarchy of smaller cities and towns, enterprises are exposed to new potential benefits. The concentration or population tends to cut the costs of providing access to public utilities, education, health and social services, making their installation more affordable, and therefore, more likely. Small manufacturers, crop processors, and repair and service facilities begin to adapt to the availability of piped water and electrification. 3/ Many activities begin to experience cost 1/ See Chuta and Liedholm, &ural Non-iarm..., op.cit. on availabie evidence. 2/ J. Page, Jr., Small Enterprises in African Development: A Survey, Washington, The World Bank, Staff idorking Paper ao. Ji3, OctoDer 19/9. 3/ Anderson, Rural Enterprise..., op.cit. - 14 - reductions when other activies, external to the enterprise, are in close proximity 1/. It becomes easier to assemble the appropriate package of inputs, labor of varying skills, business services and credit. Even the scope of the market expands, as road and communication links are improved, and each city's hinterland widens. One now begins to find firms of more suostantiai size capturing a greater share of output. Nevertheless, the importance of smaller scale firms continues to be felt, even in large ci.ies. Such centers have an important percentage of their labor force working in small, usually unregistered enterprises. Survey results suggest that approximately 60 of the employment in Lagos can be classified as traditional, informal, or unregistered. 2/ For Abidjan, the total so classified equals roughly 55%.3/ In Kinshasa, virtually the same proportion applies for the unregistered sector. 4/ For Bamako, fully bbXY of 1/ For a review of the literature on agglomeration effects see i. iloomaw, "Productivity and City Size: A Critique of the Evidence," Quarterly Journal of Economics, V.96, loo. 4; November 1961; J. V. aenderson, "General Equilibrium Modelling of Systems of Cities," in P. Nijkamp and E. Mills, eds., dandbooks in Regional and Urban Economics, Vol. ii, Amsterdam, North Holland, forthcoming; W. Wheaton and H. Shishido, "Urban Concentration, Agglomeration Economies and the Level of Economic Development," Economic Development and Cultural Change, October 1981; &. MHera, "The Changing Pattern of Population Distribution in Japan and its Implications for Developing Countries," Habitat International, 2, 1977; and K. Mera and d. Shishido, "A Cross-Sectional Analysis of Urbanization and Socio-Economic World," Washington, Discussion Paper No. 40, Water Supply and Urban Development Department, The World Bank. 2/ Wilbur Smith Associates, et.al., Master Plan..., op.cit. 3 Groupement SCET International, Perspective Decennales de Developpemeit d'Abidjan, Vol. 4: Annexes due Rapport de Premiere Phase, February 197b; somewhat lower estimates were arrived for earlier, incomplete surveys, see H. Josh;i, U. Lubell, J. Nouly, Abidjan: Urban Development and Employment in the IT-ary Coast, Geneva, International Labor Organization, 197b. 4/ M.K. Gozo, "Le Secteur Non Structure de Kinshasa," Addis Ababa, International Labor Organization, 1965. - 15 - all workers are excluded from the "modern" economy. 1/ In Dakar, the informal sector captures 56% of employment. 2/ In Nairobi, up to 44% of the labor force is employed in unregistered, small scale activities. 3/ And the proportion in Lusaka approaches 40%. 4/ Policymakers in Sub-Saharan Africa should concern themselves with the informal sector for reasons other than the relative share of the labor force involved. 5/ The alternative formal or "modern" sector has, in the past, been protected and subsidized by the central government. With the introduction of the structural adjustment process, this type of assistance is bound to decline in scope. Fewer and fewer of the new increments to the labor force will be able to rely on formal sector jobs. The informal sector, on the other hand, is already accustomed to working within a framework of more appropriate price signals, and is thus geared toward labor-intensive production methods. The informal sector is al3o capable of providing an efficient mechanism for "on- the-job" training, through its apprenticeship system. While expanding the stock of human capital in the cities, the informal sector helps attract savings that by-pass financial institutions, thus increasing the supply of physical capital as well. 1/ Direction du Projet Urbain de Mali, Etude du Developpement Urbain de Bamako: Rapport Socio-Economique, Bamako, March 1984. 2/ Senegal: Urban Sector Memorandum, Washington, Urban Projects Division, West Africa Region, The World Bank, Report No. 4537-SE, June 1, 1983. 3/ W. House, "Nairobi's Informal Sector: An Exploratory Study," in T. Killick, Papers on the Kenyan Economy: Performance, Problems, and Policies, London, Heinemann, 1981. 4/ R. Sandbrook, The Politics of Basic Needs: Urban Aspects of Assaulting Poverty in Africa, London, Heinemann, 1982. 5/ This section draws upon C. Becker, C. de Bodisco, and A. Morrison, "Urban Africa in Mlacroeconomic and Microeconomic Perspective: Issues and Options," Washington, The World Bank, processed, December 1985. 16 - Finally, the informal sector deserves attention because it satisfies a large and growing demand for customized and/or low-cost goods and services by an urban population that, on average, spends half of its income on food alone. Especially among low-income households, the outputs of small-scale firms are likely to be more appropriate, in quality and in price, than the production of most larger firms. Yet the smaller-scale enterprise cannot shoulder the full burden of supplying non-agricultural goods and services in these countries. And it is in the larger cities that one finds the bulk of medium-to-large businesses. The reasons for this are many. The opportunities for agalomeration economies are largest in such cities. In addition, during the early stages of economic development, the ease of access tc the national and international monetized economy will vary drastically from one center to another, favoring large cities priviledged. by reason of history and/or geography (e.g., possession of an excellent transshipment point such as a deepwater port). Furthermore, during the long period when investment funds are extraordinarily scarce given demand, governments are likely to concentrate infrastructure -- including power, water, roads, telecommunications, and education and health facilities -- in a few large places, so that at least those economic "city- states" can service the widest possible range of enterprise demands. The differential availability of infrastructure will act to restrict the range of settlements available for the location of certain economic activity, compared with what would be available at a more advanced level of development. Onlly as development proceeds -- and once variations in settlement size and density are no longer associated with major differences in public service quality and ease of access to markets -- will a portion of formal sector productive activities tend to avoid operating in very large centers - 17 - whose scale and associated cost are higher than justified by the clustering advantages they can reap. Until then the formal sector's health will be tied to large cities. One can get a aood grasp of the magnitude of these Dig city economies by examining their overall scale. Careful studies 1/ of the metropolitan economies suggest, for example, that without Nairobi, Kenya's GDP would be cut by one-third; without Abidjan, half of the Ivory Coast's GDP would not exist, without Bamako, Mali's economy would shrink by one-third; without Lagos, fully one-fifth of Nigeriats output would vanish; without Kinshasa, one-third of Zaire's output would disappear; and without Greater Dakar, Senegal's output would fall by one-half. Much of this output is produced by the regulated sector, despite the significant proportion of labor involved in the unregulated economy. 2/ And the regulated sector contributes heavily to making such large centers the most diversified local economies in their coun- tries. The issues facing the modern sector will not be overtaken Dy the process of structural adjustment that many of these countries are already involved in. Even under the most pessimistic assumptions, the bulk of existing industrial firms, and the specialized business services they foster, 1/ For Lagos, see Wilbur Smith Associates, et.al., Master Plan..., op.cit.; for Nairobi, see Nairobi Urban Study Group, Nairobi: Mletropolitan Growth Strategy, Volume 2: Technical Appendices, n.d.; for Abidjan, see Groupement SCET International, Perspectives..., op.cit.; for Bamako, see Direction du Projet Urbain de Mali, Etude..., op.cit.; for Kinshasa, author's own estimates, based on conversations with the Bureau d'Etudes d'Amenagements Urbains (Kinshasa); for Dakar, see Ministere du Plan et de la Cooperation, EPEVRY, Secteur Urbain, Preparation d'un Credit d'Engenierie, Rapport Intermediare, Dakar, ilay 1983. 2/ In Lagos for example, the unregulated sector contains 57% of the employment but only 147 of the value-added generated (wilour Smitfl Associates, et.al., Master Plan..., op.cit). - 18 - will survive the process and be joined by additional firms in an increasing number of industrial sectors. 1/ As one regional review notes: A number of the larger countries in Sub-Saharan Africa have built up eniough industry and acquired enough industrial experience so that they are already into the middle phases of industrial development. Their further success depends on building up large-scale, low-cost intermediate goods industries (in addition to cement and other building materials already being made), and increasing the competitiveness of a wide range of manufacturing industries including those making engineering products of various kinds, while at the same time expanding the overall economy despite foreign exchange, agricultural, and other constraints. Countries already into these middle phases include... Nigeria and Zimbabwe... [the] Ivory Coast and Kenya followed in some ways by Zambia. Nearing this stage are... Ghana, Tanzania, Ethiopia, Senegal... Cameroon, Zaire, Sudan, [and] Madagascar. 2/ Beyond these countries, most of the remainder -- smaller and poorer - face a challenge in building up even the simplest manufacturing industries. Taken as a group, however, the Sub-Saharan countries are expected to increase industrial output by at least 50% by 2000, or as much as 250% under more optimistic scenarios. 3/ The more sophisticated medium and large scale firms will obviously participate in this expansion, and their requirements should be met. D. The Magnitude and Impact of Infrastructure Deficiencies on the Productivity of Cities What type of problems thwart the development of the non-farm sector in Sub-Saharan Africa? In essenze, they are problems requiring solutions at a 1/ See, among others, D. Keesing, "Industrial Policy Issues in Eastern Africa," World Bank, processed, MIay 1984; and "Bank Operations for Industrial Development in Sub-Saharan Africa," Washington, Industry Department, The World Bank, processed, July 1985. 2/ D. Keesing, "Industrial Policy..., op.cit. 3/ Ibid. - 19 - subnational or local level, which complement measures being taken increasing at the macroeconomic level. Policy changes are needed to remove impediments and to take advantage of missed opportunities. These changes require recasting the role of the public sector, encouraging it to do things that only it can do, and to cease doing many things that promote inefficiency. The end result would be a pared-down public sector, with a lesser central government role, and a beefed-up role for muncipalities. It would not niecessarily invol.,e any increase in central expenditures devoted to urban problems. Some issues facing cities require few extra resources; others involve more effective use of existing resources; yet others necessitate more self- financing of urban activities. In all cases, resources would be airected increasingly to initiatives showing very high rates of return. One of major problems facing the urban economy in Africa is the state of local infrastructure. This, in turn, is merely a symptom reflecting the neglect of local government in these countries, as well as the failure to mobilize resources at the local level. How bad are infrastructure conditions in Africa's cities? As a rule of thumb, conditions are likely to be best in the very largest cities, such as the capital and/or major port. I)etailed city studies of smaller centers are rare, but one can assume that poorly functioning large cities are matched by deficiencies down the urban hierarchy. Thus, for example, in Nigeria, major secondary centers such as Ibadan, Kano, Kaduna, Benin City, and Port Hartcourt suffer serious service and infrastructure deficiencies similar to those identified in Lagos. in Ghana, the problems faced by Accra are also found in Kumasi. Like Kinshasa, virtually all the small urban centers in Zaire, including Kisangani and Bukavu, are in a state of disrepair that reflects 20 years of neglect. - 20 - Conversely, when infrastructure conditions are good in the largest city (Harare in Zimbabwe), they are good elsewhere as well. This is true not only in Bulawayo, with 450,000 inhabitants, but also in Kwekwe and Marondera each with fewer than 50,000 residents. A review of the state of urban services and infrastructure in major African cities is therefore revealing. All these centers have a core area, inherited at Independence, where infrastructure conditions are tolerable to good; this is the case in spite of the cities' need to accomodate much higher densities than originally foreseen. Elsewhere, as a rule, the situation is more alarming. For the very largest cities, such as Kinshasa and Lagos, conditions are unusually serious only because of the scale of the problem. They provide a vivid illustration of what many of tomorrow's African cities will look like if the politics of neglect continue to prevail. Kinshasa, 1/ with a population of 3 million, has increased in size from 6,000 hectares in 19bU to 25,000 hectares today. Yet the paved road network has remained virtually unchanged over that period. Paralleling these deficiencies is the absence of storm drains o'.' the obstruction of existing drains with uncollected solid wastes. These problems contribute to severe flooding in the rainy season, and accelerate the deterioration of existing streets. As a result, the potential coverage provided by public transit facilities is severely limited. Workers living in districts developed after 1960 are forced to walk long distances to reach a bus or to arrive directly at their work places. Then, there is the case of Lagos (population: 4.5 million). The problem can be detailed subsector by subsector: 2/ 1/ For Kinshasa, see Bureau d'Etudes d'Amenagements Urbains (Kinshasa), "Programme de Recherche Croissance Urbaine et Gestion de Villes, Rapport d'Analyze General, Kinshasa," processed, January 1984. - 21 - a) Water: "Il addition to bottlenecks in the transmission and distribution systems, much of the city lacks adequate tertiary networks; those that do exist are in poor condition." b) Storm Drainage: "...much of Lagos is inadequately drained or not drained at all. Most drains are not maintained and have, for lacK of maintenance, become dysfunctional. Consequently about 8,400 hectares of the city are subject to frequent and often severe flooding... c) Roads: "large land areas are often serviced by a single arterial or collector road and there is a lack of ad-equate arterial and collector facilities for cross town trips... Many of the major routes suffer considerable pavement distress. d) Power: "...vast areas of the metropolis still have inadequate supplies and outages are frequent." e) Communications: "The communications system in Lagos is characterized... by few telephones, much frustration and many delays in placing calls" Accra (population: 1.4 million) provides an example of a city that has lost its once proud status of a well-equipped city. Half the population now lives in areas where little infrastructure is available. The state of the remaining inifrastructure in nlow poor: Due to lack of maintenance in recent years the condition of the [road] network, particularly the heavily traveled paved roads, has deteriorated and suffers from various degrees of pavement distress, moderate to extensive, over much of the paved network.... The remaining pavement life is, on average, five years. This deterioration... and missing network links causes extensive delay in the movement of goods and services (particularly to and from the ports), damage to vehicles and equipment, [and] very high vehicle operating costs.... Due to the lack of routine maintenance, the 2/ Nigeria: Lagos Urban Sector Review, Urban Projects Division, West Africa Region, Washington, The World Bank, Report No. 4479-UNI, Sept 26, 1983. drainage system has also deteriorated to the point of ceasing to function in many places, and flooding during the rainy season causes extended road blockages and further deterioration of the road network. 1/ Many centers of even relatively moderate size suffer from similar problems. Bamako (population: 675,000) is described in one Bank report as follows: The oraanized city has only a minimum of facilities and over exploits a deteriorating infrastructure while nearly all the new areas expand without even makeshift equipment and infrastructure. 2/ Even more dramatic are the descriptions of conditions in Douala (population: 740,000), Cameroon's main port and industrial center: The lack of drainage facilities in most of the newly urbanized areas and the severe deterioration or even complete destruction of existing networks in older districts cause severe flooding during most of the year. Mud flows and accumulations, indiscriminate garbage disposal, and squatter settlements in natural streams have combined to worsen already minimal environmental conditions.... The road network has not kept pace with the expansion of residential areas, and an inefficient transport system is responsible for considerable loss of productivity and it increasingly affects the functioning of vital sectors such as the port and the railways system. 3/ Some cities have managed to expand infrastructure to accomodate some or most of the population growth. Among these most face problems of preserving and consolidating such gains, while tackling future growth. I/ Ghana: Accra District Rehabilitation Project, Urban Projects uivision, West Africa Projects Department, Washington, The World Bank, September 1984. 2/ "Mali's Urban Sector: A Financial Investment Strategy for Mali's Capital City, Bamako - 1985-95," (Urban Projects Division, West Africa Projects Department), Washington, The World Bank, processed, January 1985. 3/ "Cameroon: Douala Infrastructure Development Project: Initiating Project Brief," (Urban Projects Division, West Africa Projects Department), Washington, The World Bank, processed, February 1984. - 23 - Examples of this type can be found in Nairobi (population: 1.2 million), Abidjan (population: 2 million), and Dakar (population: 1.6 million). Nairobi's infrastructure network is relatively complete. There is an extensive network of paved roads with associated drainage facilities, power and piped water are readily available, and public transport facilities are reasonably good. However, the recent period of national austerity has led to policies of deferred road and drainage maintenance which threaten to create important new investment requirements simply to repair the results oE neglect. Solid waste collection is also emerging as a serious problem because collection and disposal equipment maintenance has been deferred for so long. Abidjan is characterized by generally good road infrastructure which, combined with careful attention to traffic management, ailows public transport to perform effectively in spite of the growing separation of job centers from areas of high resident growth. Water supplies are plentiful, and telephone service is satisfactory. The weaker services are electricity supplies and provision of an adequate storm drainage system. In addition, Abidjan has a problem in preserving the gains it has made since Independence, because, like Nairobi, maintenance functions have been neglected. Dakar has a weaker infrastructure base. Electricity supplies are adequate, but much other infrastructure is deficient. Out of 700 kilometers of roads, only about 200 are paved, and most are located in the ex-colonial core. This network is even less effective than it might be because little attention has been paid to traffic management (street lights, one-way streets, bus lanes). Like Abidjan, Dakar is experiencing ever longer average commuting time, as jobs remain tightly clustered in older core locations. 'Water supplies are constrained. Telephone service is poor. Solid waste disposal - 24 - procedures are deficient, which leads to the dumping of garbage into an already poorly-maintained drainage system. The major cities of Zimbabwe, whose low-income population growtLl was controlled until 1980, 1/ have the best conditions. As the Bank's Zimbabwe Urban Sector Review notes. 2/ ... primary infrastructure is already in place at a high standard. lWater and sewarage reticulation is complete, even in low-income areas. Major surplus capacity exists in most of the municipal streets.... Zimbabwe is in the unusual situation... of requiring infrastructure investment only for the increment of population, starting from a solid base. This situation, however, is exceptional. The earlier examples suggest the norm. Three characteristics of this generaL deficiency in the supply of infrastructure and services should be underlined. 3/ First, different forms of infrastrucL,ire and services have differing impacts in the efficiency of economic activity. Some types of infrastructure are prerequisites for land development (road access; drainage). Other public services affect the day-to- day operation of businesses (public transport; telecommunications; solid waste 1/ Prior to Independence, the legal movement of blacks to towns was tied to proof of formal employment. During that period, attempts at squatting" on vacant urban land was rigorously and effectively opposed. Zambia, Malawi, Kenya, and Tanzania all had similar influx controls prior to Independence. 9/ Water Supply and Urban Development Division, Eastern Africa Projects Department, Washington, The World Bank, Report #4171-ZIM, February 1, 1985. 3/ This section draws on the work of F. Kahnert, "Improving Urbani Employment and Labor Productivity," Washington, Water Supply and Urban Development Department, The Uorld Bank, processed, February, 1985. - 25 - removal; public safety). Yet others act as direct inputs into production process (power; water). Second, enterprises will differ in their dependence on particular services. Large and medium-sized enterprises are likely to see all basic infrastructure as necessary; while very small business may operace without power or telecommunications facilities. But these differences are less im- portant than they may seem at first because smaller enterprises do depend on the efficiency of larger enterprises. Large-scale enterprises often import or produce goods utilized by small businesses (e.g. factory-made yarns and dyes; metal goods; plastics; rubber; leather). 1/ In addition, many of the small business workers learn their trade and/or accumulate their initial capital, while working for larger firms. 2/ Finally, the wage incomes generated by larger firms helps create demand for commodities marketed by small scale firms. A final characteristic of these deficiencies is that they can be compensated for, but at considerable cost and that cost usually exceeds the value of the taxes and user charges needed to finance public provision of such services. Businesses can develop private infrastructure networks, investing in generators, bore hole wells, water purification equipment, private employee transport, security guards, walkie-talkies, and otherwise unnecessarily large inventories. If this financial burden is impossible to assume, there are only two remaining alternatives: to do without the service (e.g. operate with manual tools instead of powered machinery) or, more likely for anly sophisticated operation, to forego the investment opportunity. In each case, there are serious costs to the economy as a whole. 1/ A. O'Connor, The African City, New York, Africana Publishing Company, 1983. 2/ R. Sandbrook, op.cit. - 26 - E. Additional Constraints on Urban Business Activity: The Issue of Land Mar'cets A major problem facing the business community involves distortions in the markets for complementary inlputs like land., Land markets tend to be disorganized, wic:h conflicting, often unrecorded, ownership claims and unclear boundaries characteristic of most land units across Africa's cities. Without an adequate system for measuring and recording the boundaries of land parcels, and without the registration of legal rights to such plots; it is impossible for a potential entrepreneur to go to any one or two sources in a city and find out what land is available for sale, what prices prevail in the market, and who the owners are. Property transfers are fraught with long-term risks, and transactioins are time consuming, and may involve payments to several purported owners. These types of obstacles clearly impede publicly-sponsored development projects, as well. Examples of these problems are found in a number of Bank and non-Bank reports. A study of Monrovia, Liberia, notes that although most land is privately owned, there is often a proliferation of conflicting deeds for the same or overlapping properties involving government, tribes, organized community groups, institutions, families and individuals... The lack of clear ownership of urban land has created a number of problems.... It has prevented develop- ment of any effective system of mortuages, inhibited land sales and building construction, discouraged improvement and maintenance of existing structures, and undermined the establishment of an effective real estate taxation. Property in Monrovia is rarely bought or sold and when it is, prices fluctuate dramatically. I/ I/ Staff Appraisal Report: Liberia: M+Ionrovia Urban Developmenit Project, West Africa Projects Department, Wiashington, The World Bank, Report No. 3384-LBR, November 30, 1981. - 27 - In Kinshasa, Zaire, only 10% of the parcels are officially registered, while the rest have been acquired informally. Such plots are sold by customary chiefs and "validated" by local government commissioners, who grant "livrets de logeur." These permits are based on a colonial system which has since been abolished; they, therefore, have no legal standing. 1/ A report on urban conditions in Cameroon suggests that the new titles issued each year represent less than 10% of the demand. Those who wish to obtain a title must wait two to seven years, depending on the "persuasiveness" of the applicant. 2/ In Kampala, Uganda "abouc 2U percent of transactions are officially recorded and fully documented while about 80 percent take the form of verbal agreements, all technically illegal, which initially save time and money but which often lead later to disputes." 3/ In Accra, Ghana the preva- lence of customary land law is blamed for impeding the expansion of the central business district and for discouraging construction investments across the city. 4/ Finally, in Dakar, Senegal, where the land tenure status of about 60% of urban dwellings is unclear, "the authorities are unable to inform a private individual within a reasonable time of the whereabouts of sites approved for building or whether he even is eligible to obtain title to the land." 5/ 1/ Remy Prud'homme, "Le Financement des Services Publics Urbains a Kinshasa: Institutions et Mechanismes," draft report, Washington, The World Bank, November 1985. 2/ Cameroon: Urban Sector Review and Sector Assistance Strategy, West Africa Projects Department, Urban Division, Washington, The World bank, Report No. 4540-CM. 3/ A. O'Connor, The African City, op.cit. 4/ Ibid. 5/ Senegal: Urban Sector Memorandum, Part Two, West Africa Region, Urban Projects Division, Wastiington, The World Bank, Report No. 4537-SE, June 1983. - 28 - Land information systems facilitate private and public real estate transactions, increase the opportunity for using land as loan collateral, and undergird any system of local real property taxation. 1/ What is needed, above all, is the political will to develop a continously updated system, and to institute streamlined procedures for transferring land rights and titles. That, in turn, depends on the recognition that such systems are "as much a part of basic infrastructure as roads or electric grids: an essential element to expedite virtually all other forms of development." 2/ F. Bottlenecks Created by the Failure of Capital Markets Small and medium-size enterprises also face serious problems created by the fragmented nature of capital markets in urban areas. 3/ Typically "formal" market funds are available to large establistnments froin commercial banks or specialized development finance institutions. Uften these funds are lent at subsidized rates. Smaller enterprises, exceptions being made for token "small business" efforts, are restricted to their own resources, borrowed funds from friends or relatives, or money raised, at very niah costs, 1/ For a review of Bank experience in this area, see J. Bernstein, "Review of World Bank Lending for Land Registration and Property Tax Improvements in Urban Projects, FY 1972-82," Washington, Water Supply and Urban Development Department, The World Bank, processed, September 1985. A more detailed exposition of land issues can be found in d. Dunkerley, ed., Urban Land Policy: Issues and Opportunities, New York, Oxford University Press, 1983. 2/ W. Doebele, "Why Cadastral Systems are Important for Less Developed Countries," Wiashington, Urban Development Department, The World Bank, draft, July 31, 1983. 3/ See, among others, P. dugon, ed., Secteur Informel et Petite Production Marchande dans les Villes du Tiers vlonde, Revue Tiers-Monde, Volume XXI, No. 82, April-June 1980; S. Sethuraman, ed., The Urban Informal Sector in Developing Countries: Employment, Poverty and Environment, Geneva, International Labor Office, 1981. - 29 - from money lenders. Good economic development projects are theretore either underfunded or not undertaken ac -11. Both project-specific jobs and employment indirectly stimulaLed by the project, may be foregone. Opportunities are diminished to mobilize personal and family savings through the availability of matching loan funds. Studies on the informal sector conducted across African cities suggest that the latent demand for loans is very high. 1/ The cited studies also suggest that the type of loan demanded (fixed capital, working capital) differs according to type of informal activity. iylanufacturing activities, for example, generally require more start-up capital than commerce and services; often land may have to be purchased, structures built and tools and machinery bought. One survey of activities in Abidjan (Ivory Coast) places the fixed capital requirements of manufacturing enterprises are at least ten times those for commercial or service firms. 2/ This finding is confirmed in work on Freetown, Lagos, and Nairobi. 1/ Evidence can be found in, among others, Di. Van Dijk, "La Reussite des Petits Entrepreneuirs dans le Secteur Informel de Ouagadougou (Haute- Volta)," in P. Hugon, ed., Secteur Informel..., op.cit.; D. Fowler, "The Informal Sector in Freetown: Opportunities for Self-Employment," in S. Sethuraman, ed., The Urban Informal Sector..., op.cit.; 0. J. Fapohunda, "Human Resources and the Lagos Informal Sector," in S. Sethuraman, ed., The Urban Informal Sector..., op.cit.; A. Mabogunje and M. Filani, "Tle Informal Sector in a Small City: The Case of Kano (Nigeria)," in S. Sethuraman, ed., The Urban Informal Sector..., op.cit.; G. Aryee, "The Informal Manufacturing Sector in Kumasi," in S. Sethuraman, ed., The Urban Informal Sector..., op.cit. For Eastern Africa, see, among others Basic Needs in Danger: A Basic Needs Oriented Development Strategy for Tanzania, Addis Ababa, Jobs and Skills Program for Africa, International Labor Office, 1982; D. Ghai, M. Godfrey, and F. Lisk, Planning for Basic Needs in Kenya: Performance, Policies, and Prospects, Geneva, International Labor Office, 1979; and I. Livingstone, Rural Devlopment, Eployment and Incomes in Kenya, Addis Ababa, Jobs and Skills Program for Africa, International Labor Office, 1981. 2/ C. de Miras, "Le Secteur de Subsistence dans les Branches de Production a Abidjan" in P. Hugon, ed., Secteur Informel..., op.cit. - 30 - Whatever the type of loan desired, the same sot, ces confirm that formal financial institutions play a negligible role in promoting informal sector business development. G. Neglected Opportunities: Reinforcing Rural-Urban Links The exploration of rural-urban links has become an object of increasing attention among students of development. 1/ A brief review of the subject cannot delve into the degree of specificity that agriculture and rural development specialists might prefer, given the enormous number of commodities involved, each with its own characteristic production, processing, and distri- bution features. Nevertheless it is worth underlining the spatial aspects of the production and marketing cycle which is shared by most commodities. Broadly speaking, each rural household involved in agricultural production can draw on nearby towns and the network of cities and towns connected to them for a variety of reasons. Input requirements are met through the urban network, including market information, improved seeds, fertilizer, machinery and tools, spare parts, repair services, and credit. Once produced, outputs must be marketed. Nearby towns act as transportation and communication hubs that can reach out to regional, national, and international markets. This marketing function involves various operations, 1/ Among the pioneers, one can cite E.A.J. Johnson, The Organization of Space in Developing Countries, Cambridge, Harvard University Press, 1970; B.F. Johnston and P. Kilby, Agriculture and Structural Transformation: Economic Strategies in Late-Developing Countries, New York, Oxford University Press, 1975; and D. Rondinelli and K. Ruddle, Urbaniza.ion and Rural Development: A Spatial Policy for Equitable Growth, New York, Praeger, 1978. Within the Bank, an early attempt at tackling these issues is M. Cohen et.al., Urban Growth and Economic Development ini the Sahel, Washington, World Bank Staff Working Paper No. 315, January 1979. Among the bilateral donors, the earliest work was done under the auspices of USAID. For a quick review of this work see E.S. Belsky, "Review of Selected Documents on Rural-Urban Dynamics," Cooperative Agreement on Settlements and Resource System Analysis, Worcester, Clark University, May 1985. - 31 - including grading, storage, assembly for shi.pment, distribution to whole- salers, and sales by retailers. These same to;rns are often the centers of daily or periodic markets directed at local regional consumers. The production and sale of agricultural commodities is not an end in itself; there must be "incentive" consumer goods and services, as well as savings and investment opportunities. While not every option need come from area urban centers, those locations may be significantly more accessible than the rest of the urban system and thus play an important role in shaping rural household perceptions on available goods, services, and investment opportunities. The public services provided by these urban centers are also accessible, at relatively low cost, to the region's rural comaunity. Urban- based schools and health facilities can service a large number of farms and villages. Local municipal authorities can help to maintain area roads feeding into the urban centers, providing a form of asset maintenance otherwise hard to come by. The availability of power and piped water in the region's towns opens new options for agrobusinesses, especially those involved in processing activities. These, in turn, can provide farmers with a service and a potential outlet for savings. The provision of residentiary, transport, and communications infrastructure in urban centers can also create an appropriate local base for the skilled workers who must be attracted to man -specialized services, such as extension work or credit institutions. A substantial proportion of the urban population in Sub-Saharan Africa derives its livelihood primarily by serving the rural producer, directly or indirectly. One estimate 1/ suggests that up to 9U percent of the population in towns of 5,000 to 20,UUO is so engaged; wnile, on a regional 1/ J. Cour "Les Implications Macro-economiques de la Croissance Urbaine: Interactions Entre les Villes et leur Hinterland," Washington, The World Bank, draft September 1985. - 32 - basis, 7 out of every 10 urban residents are linked, in some degree, to ttie primary sector. Tla amounts to a total of 72 million people. One final Lnkage bears mentioning. Across Sub-Saharan Africa there is evidence that rural-urban migration forms part of a rural household strategy to diversify sources of income. 1/ A review of existing data suggests that most migrants regularly remit goods or money to rural areas and that these remittances represent between IU'o to ZUY of migrants' estimated household income. 2/ The relative weight of the remittances in rural household income also varies; in some cases, notably the communal farming areas of Zimbabwe or the villages of southeast Senegal, remittances can literally double annual household income. 3/ This income stream is more than an insurance scheme; it can play a major role in promoting farm innovation and the purchase of off-farm inputs. Collier and Lal, 4/ reviewing evidence fromn three provinces in Kenya, suggest that rural factor markets don't work very well. As a result, the adoption of innovations (including the associated capital and operating expenses) is often 1/ For a sketch of the evidence, see A. O'Connor, The African City, op.cit. 2/ Among others, see J. Caldwell, African Rural-Urban Migration: The Movement to Ghana's Towns, London, C. Hurst, 19b9; V. Moller, Urban Commitment and Involvement Among Black Rhodesians, Durban, University of Natal, 1978; G. Johnson and W. Whitelaw "Urban-Rural Income Transfers in Kenya," Economic Development and Cultural Change, Vol. 22, 1974; A. Adepoju, "Between Urban Migrants and Their Home Communities in Nigeria," Africa, Vol. 44, 1974; R. Sabot, Economic Development and Urban Migration: Tanzania, 1900-1971, Oxford, Clarendom Press, 1979. 3/ Zimbabwe Agricultural Sector Study, op.cit. For Senegal, see J. Venard, "Senegal: Urban-Rural Linkages Case Study," WXashington, DC, The World Bank, processed July 1985. 4/ P. Collier and D. Lal, "Why Poor People Get Rich: Kenya 1960-1979," World Development, Vol. 12, No. 10, 1984. See also 0. Stark, "On the Role of Urban-to-Rural Remittances in Rural Development," The Journal of Develop- ment Studies, Vol. 16, No. 3, April 1980. - 33 - foregone because self-financing out of current income is impossible and credit is difficult to obtain. The remittances associated with town and city employment help attenuate the perceived risk of innovation, while providing a means of servicing loans used to switch production to cash crops, imnproved livestock, hybrid maize, and a high level of purchased inputs. A proxy for measuring the impact of remittances on rural development can be found by examining the behavior of return rural-urban migrants, especially in areas that are relatively accessible to urban markets. Return migrants in Zambia, for example, tend to own larger farms than non-migrants, and to make greater use of improved agricultural practices and technology. 1/ Capital from migration allows farmers who are not creditworthy to invest in farm improvements, while giving others a chance to qualify for formal sector loans. Return migrants also help to inject cash resources into the rural economy, and to develop the bulk of non-farm rural entrepreneurial ventures. In each case, the earnings and savings of the return migrant constitute the major source of funds for such activities. That the Zambia example is not unusual if confirmed by Adepoju citing the work of such researchers as Mabogunje: Return migrants are known to serve as agents of cultural, social, and economic diffusion. They often influence the home community, on a macro level, by their personal example. By introducing new crops or improved varieties of existing ones, by employing new or improved techniques of production, and by encouraging education and organization, they serve as agents of change, opinion leaders and innovators. The ripple effect of their expenditures and investments in housing construction, trade, commerce, large-scale farming, and the introduction of new varieties of high-yielding seeds may directly promote rural development.... 1/ 1/ A. Chilivumbo, Migration and Uneven Rural Development in Africa: The Case of Zambia, New York, University Press of America, 1965. - 34 - This wide variety of rural-urban linkages doesn't appear automatically. In fact, Sub-Saharan Africa is full of examples of policies and institutions which short-circuit these interactions to the detriment of farm and town. A prime culprit is the macroeconiomic environment that has typified conditions across many of the regiotn's countries. Inadequate price incentives have been a common complaint among African farmers. With exceptions of countries where farm organizations have a strong organized lobby (Ivory Coast, Zimbabwe, Kenya), farmers have faced onerous price controls often backed by attempts to suppress unofficial private trading or, in the case of food goods, by efforts to flood the market with subsidized imports. 2/ Often these policies are accompanied by a protectionist trade regime and an overvalued currency, whose combined impact is to raise the prices of consumer goods to levels well above what they would otherwise be. In spite of the World Bank's reputation for recommending "get the prices right" as the principal means of solving sectoral problems, bank and non Bank sources alike provide repeated evidence that price incentives are not enough. A recent study of the impact of price and exchange policies on 1/ A. Adepoju, "Issues in the Study of Migration and Urbanization in Africa South of the Sahara" in P. Morrison, ed., Population Miovements: Their Forms and Functions in Urbanization and Development, Liege, Belgium, Ordina Press, 1983. 2/ One study on Zaire noted: "The large cities find it simpler to import foreign food products paid with the earnings from copper exports, than to take advantage of the national potential for agricultural production. (own translation) G. Gaveau and C. Monnier, Les Villes Secondaires en Afrigue: Leur Role et Leurs Fonctions dans le Developpement National et Regional: Le Zaire, Paris, Agence Francaise pour L'Amenagement et le Developpement a l'Etranger, 1984. - 35 - agriculture in the region 1/ concludes that the price and exchange rate policies followed since Independence have had an impact on agriculture production, and that impact has usually been negative. However, the analysis suggests that these policies are not the most important reason for the poor agricultural performance registered by so many of the region's ecotnomies. Instead, farmers react to a whole package of incentives, most of which are linked to the adequacy of urban-based activities. 2/ The Bank's agricultural sector report on Kenya notes, for example: Without additional measures to improve input delivery and incentives, it is unlikely that the Government's ambitious agricultural development programs will be effective, as input availability may well be as important to smallholders as price incentives in influencing production decisions. 3/ A study of Tanzanian agriculture seconds this, reporting that improved production cannot be achieved without the policies, prices and productive and service infrastructure that are provided by the urban dwellers and which boost or cripple peasants' efforts: The urban-provisioned rural infrastructure would have to involve an improvement of transport as well as increased productivity in urban manufacturing and 1/ K. Cleaver, The Impact of Price and Exchange Rate Policies on Agriculture in Sub-Saharan Africa, Washington, World bank Staff Working Paper No. 72d, 1985. 2/ Among others, this point is made forcibly in Tanzania Agricultural Sector Report, lWashington, Eastern Africa Projects Department, Southern Agriculture Division, The World Bank, Report No. 4052-TA, August, 1963; Zaire Econiomic Memorandum: Recent Economic and Sectoral Developments and Current Issues, Washington, Eastern Africa, Country Programs Department !I, The World Bank, Report No. 4077-ZR, December 1982; Zimbabwe Agricultural Sector Study, op.cit.; Ghana Agricultural Sector Review, Washington, Western Africa Region, The World Bank, Report No. 5366-GH, August 198); Nigeria Agricultural Sector Memorandum, Washington, Western Africa Projects Department, Agriculture Division B, The World Bank, June 1984. 3/ Kenya Agricultural Sector Report, Washington, Eastern Africa Projects, Central Agricultural Division, The World Bank, processed, January 1984. - 36 - services, thereby lowering the costs of consumer and producer goods and social services for the rural population. 1/ Sometimes there are bottlenecks restraining urban-rural links which cannot be resolved at the local level. Cour has pointed out 2/ that ready access to urban markets is a major factor in explaining the productivity of the farm community. In fact, one can conceive of a network of urban centers whose demand impact on farmers in any locality is proportional to each city's size and inversely proportional to the difficulty of reaching it. Around major urban centers one usually finds a prosperous farn belt speciaiizing in such high value-to-weight produce as vegetables and fruits, as well as raising chickens and producing dairy products. At great distances from urban centers, isolation results in little more than subsistence farming. In between, the degree to which farmers produce for sale will vary with the state of local and interregional transport links. This suggests that both rural and urban development investment decisions are best undertaken in a coordinated fashion rwith transportation plans for each region. In Zambia, for example, a review of rural development found that a strong correlation existed between provincial access to the network of urban centers and the level of agricultural development. 3/ The Southerni province has the best access and the most developed farm system. The Eastern province is far less accessible to urban centers and transport facilities and is 1/ D. Bryceson, Urbanization and Agrarian Development in Tanzania with Special Reference to Secondary Cities, London, International Institute for Environment and Development, February 1984. 2/ "Les Implications Macroeconomiques," op.cit. 3/ A. Chilivumbo, Migration and Uenven Rural Development in Africa..., op.cit. - 37 - correspondingly less developed. The Western and Luapula provinces are isolated and have thus participated minimally in cash crop production. Isolation from national urban markets also explains the fate of Matam (population: 12,000) in southeast Senegal. That area and its farmers are cut off from the bulk of potential buyers due to the absence of all weather inter- regional roads. As a results, Matam can do little but wait. A bilateral donor who financed an abattoir in Matam has erected little more thani a monument to excessive optimism. To a lesser degree, the urban centers of the Great Lakes region of Central Africa (Kivu province, Zaire; burundi; Rwanda) face demand constraints that can only be fully removed through more investment in interregional or international tranisport links. The potential rate of return in new transport facilities in these areas can be gauged by the fact that farmers in Kivu province export produce to Kinshasa by air, while cement is flown in the opposite direction. In other cases, the problems restricting urban-rural links are susceptible to resolution at the regional or local level, but only after an appropriate reform in intergovernmental relations initiated at the national level. In Zaire, or the Ivory Coast, for example, any infrastructural invest- ment of more than very modest dimensions is decided by the central government. Localities cannot, on their own, mobilize substantial resources at the local level to finance, directly or through loans, infrastructure programs that will provide electric power, piped water, improved schools, and better local area roads. As a result, cities like Oume (population: 2b,UUO) in the Ivory Coast or Beni (population: 40,000) in Zaire's Kivu province, cannot parlay their earnings from coffee production into additional infrascructure that will allow the urban economic base to expand. Conversely, a city like Marondera (population: 20,000) in Zimbabwe, has developed within - 38 - an environment that has allowed it to invest in its own development. As a result, there one finds a full spectrum of public services and infrastructure, mostly geared to satisfying the requirements of its agricultural hinterland. Whatever the proximate cause, there are widespread reports across Africa of communities that cannot develop the infrastructure base necessary to create more investor confidence. As a result, the money generated by commercial farming transactions and by remittances does not attract the volume of local urban economic activity that it might. The local multiplier effect is thus weakened.l/ As Cleaver's review 2/ of the factors behind poor agricultural performance points out, some of the bottlenecks appears to be caused by deliberate government attempts to repLace the private sector operating in cities and towns with highly centralized and vertically integrated parastatals. As a result the urban system fails to support agriculture, and farmers are forced to rely on the parastatals which are, as a rule, very poor substitutes. As an excellent example of this can be found in Senegal, by examining SAED, 3/ which is charged with the development of irrigated rice areas in the northeast. Studies prepared for the Commission of the European Communities underline the total detachment of the area's rice production system from the network of cities and towns: 1/ This phenomenon is extensively documented in D. Prindle, Zaire: &ural- Urban Profile, Washington, Office of Housing and Urban Programs, USALD, April 1984. See also, M. IcNulty, "Ghana/Nigeria: Kural-Uroan Linkage Case Study," preliminary report, September 1985. 2/ K. Cleaver, The Impact of Price and Exchange Rate Policies.., op.cit. 3/ Societe d'amenagement et d'exploration des terres du delta, du fleuve Senegal et des vallees du fleuve Senegal et de la Faleme. - 39 - The transporters, the artisans, che merchants, the local banks -- none are called upon, because all credit or inputs, all marketing activities follow the integrated circuit of the parastatal. 1/ (own translation) The parastatal's functions are developed in a featureless, plain characterized only by conditions linked to soil quality, to climate, and to water management. At the limit, one would think that the immense zone controlled by SAED is only populated by canals and a "work force". 2/ (own translation) As part of the structural adjustment process, SAED is to be largely dismantled. A review of urban functions in the Senegal River delta suggests that one center, Richard Toll (population ZU,000) has the potential to grow to 100,000 inhabitants in 20 years. This would come about largely because of SAED's retrenchment. Already private firms are emerging to cultivate, process and market rice using Richard Toll as a base. Input suppliers, like the Societe Senegalaise des Produits Chimiques, that previously serviced parastatals from Dakar, are begining to open up facilities in Richard Toll to reach the small-scale purchaser.3/ It is clear that, to a greater or lesser extent, lessons such as this one can be documented across Sub-Saharan Africa. The synthesis report of country studies prepared for the Commission of the European Communities, puts the case as eloquently as any. It reports that most parastatals are based in 1/ J. Champand, j. Lombard, and M. Sivignon, Les Villes Secondaires en Afrique, Leur Role et Leurs Fonctions Dans le Developpement National et Regional, Phase 2, Etude de Cas Regional: Senegal, Paris, Agence Francaise Pour l'Amenagement et le Developpement a l'Etranger, December 1984. 2/ J. Bissiliat and M. Fenet-Rieutord, Les Villes Secondaires en Afrique: Leur Role et Leurs Fonctions dans le Developpement National et Regional, Phase 1, Senegal, Paris, Agence Francaise pour l'Amenagement et le Developpement a 1'Etranger, March 1984. 3/ J. Venard, op.cit. - 40 - the capital, with local offices being sited away from regional townls or, if located in secondary ceniters, they are mere satellites delinked from all local activity: M4odern equipment, fertilizers and pesticides are supplied by the agencies themselves not by the town, stocks of inputs being transported direct from the center; heavy equipment, including tractors, is owned by the agencies or by central enterprises. ... the banking function, which is so important for a town, has to a large extent been supplanted by a closed system, of centralized origin, which sujpplies inputs Eor free against a proimise that the produce will be marketed. Funding comes directly from the ckinter, is changed into inputs and exchanged for produce, which represents a very low velocity of circulation. 1/ As a result, the network of towns have a moribund commercial and manufacturing sector, there is little financial intermediation through local institutions, and the centers have few resources to develop themselves. The international donor community has, in the past, been guilty of supporting these types of parastatals. 2/ In addition, there is evidence that even rural development projects financed with donor aid have ignored the urban network: Usually, the foreign and local management personnel take up temporary residence on site or in the nearest town, bring in their own vehicles, have their oWln stock of inputs and independent means of equipping the project. It is rare for the secondary towIn to be explicitly included in the operation and for measures 1/ M. Wertheimer, B. Becq, G. Gaveau, Secondary Towns in Africa, Their Role and Functions in National and Regional Development, Paris, Agence Francaise pour l'Amenagement et le Developpement a l'Etranger, 1984. 2/ "From the Government's point of view," reports the Bank's latest Tanzania Agricultural Sector Report (op.cit ), "the substantial donor support for parastatals has probably been seen as supporting evidence of their suitability as a form of agricultural organization" - 41 - to be taken, within the context of the project, to stimulate its supply of services and its demand for products, or, in other words, to organize mutual and reciprocal enrichment. 1/ Because the institutional structure prevalent until, now has beet, inefficient and damaging to marketed agricultural output, it is under review in many countries. The structural adjustment process often has as one of tts objectives the abolition or drastic restructuring of the agricultural parasta- tals. As a result, the private sector, operating from cities and towns, could gain new importance. As noted, SAED in Senegal is to be stripped of most of its present functions. Ghana's Cocoa Board is now committed to reducing its staff by 20,000, while being reorganized as a for-profit, commercial operation; the streamlined corporation will be far more reliant on the private sector than in the past. In Zambia, the National Agricultural Marketing Board has been reduced co interprovincial trade of maize and fertilizer and the holding of national grain reserves. In Uganda, there is broad agreement that the role of the main export crop marketing boards should be limited to the purchase of the processed crop and the transport to port for export. These types of actions are likely to be repeated across the coLntinent as the structural adjustment process intensifes. As institutional changes and reforms in pricing and trade regimes promise to restructure incentives, the question arises about the future course of the agricultural sector. There is agreement on two major points: (i) the sector must contribute much more commercial production than it has aI the past; and (ii) such increases will be made possible by a change in technology, which will be more input and equipment-intensive than is true, on average today. Both counts suggest that the relative importance of the subsistence 1/ M. Wertheimer, et.al., Secondary Towns in Africa..., p.-cit. - 42 - farmer -- who has little use for a town -- will diminisht and that commercial operations will become more significant. 1/ Thus, the fact that future economic growth will occur within the context of an agriculture-led development strategy in no way diminishes the role of cities and towns in the future; the contrary will be the case. 1/ These conclusions are sometimes overshadowed by the lively controversy over what units of what average size will play this vanguard role. Whatever size categories take the lead, the results should be similar. - 43 - III. New Initiatives: Paths to Avoid A. The Issue of Urban Bias Many analysts have called into question the role of the urban sector in Sub-Saharan Africa, citing the issue of "urban bias." 1/ According to this school of thought, national policies have been biased against agricultural development on three counts. First, macroeconomic decisions involving trade tariff structure, exchange rates, and price controls led to distorted economic signals; relative factor and product prices do not reflect true scarcity values. In particular, the-e policies, as a group, tend to raise non- agricultural unit value add-i -mo,je what would prevail at international prices. 2/ Second, the public sectcr deliberately invested resources in urban areas to improve infrastructure and public services without regard for high rate-of-return opportunities elsewhere in the economy and without recovering costs from the beneficiaries. Third, the central government compounded the effects of the first two transgressions by augmenting public employment and public directly-productive investment in cities to a degree not justified by any conceivable efficiency criteria. 1/ The case has been made for a number of regions in M. Lipton, Why Poor People Stay Poor: Urban Bias in World Development, London, Temple Smith, 1977. For Africa, see R. bates, MIarkets and States in 'Tropical Africa: The Political Basis of Agricultural Policies, Berkeley, University of California Press, 1981. For a different view, emphasizing the degree of depr-ivation found in urban Africa, and challenging Lipton's assumptions about the political foundations of "urban bias," see R. Sandbrook, 'he Politics of Basic Needs..., op.cit. 2/ That would apply to internationally traded commodities. For "non- tradeable" one would expect to find as symptoms of "urban bias," otherwise unexplained departures from prices of similar goods and services in agriculture. - 44 - There is ample evidence of all three types of "bureaucratic failure" both in Africa and elsewhere. For example, a survey by R. Agarwala, drawing on the experience of 31 countries, suggests that price distortions in foreign exchange, and factor and product pricing are widespread. 1/ Among the ten best performers, three are African (Malawi, Cameroon, Kenya). In the interme- diate group, three of the region's states are again represented (Ethiopia, Ivory Coast, and Senegal). While among the eleven worst performers, one finds Tanzania, Nigeria, and Ghana. These results underline two obvious poilnts: a) the macroeconomic policies of the region's countries are not uniform, and it is unfair to lump Kenya with Tanzania, or the Ivory Coast with Ghana; and b) the presence of at least this type of "urban bias" has not, in the past, kept the World Bank and other donors from working in the urban sectors of Brazil, Mexico, Indonesia, India, and Pakistan -- none of which are star performers on Agarwala's list. As for the second type of bias, that involving public services, the Bank's urbani sector work has documented, for Africa and elsewhere, the problems of heavy subsidization of public services, inadequate attention to affordable standards, and the consequent restricted coverage of most public services. Most households in most African cities have limited access to public services. The World Bank has takeni the lead in convincing African governments to scale down the quality of additional infrastructure and public services, and to employ user charges and local taxes to recover a large 1/ Price Distortions and Growth in Developing Countries, Wasnington, Woria Bank Staff Working Paper No. 575, 1983. - 45 - fraction of capital and recurrenit expenses in urban areas. I/ This should free up resources, over time, for tnew initiatives in rural areas. It is not immediately apparent, however, what types and levels of public services would flow to rural areas where low incomes and low densities may militate against rapid expansion of infrastructure. Finally the present process of structural adjustment is helping to document widespread overmanning of public sector enterprises and to catalog inappropriate productive investments in parastatal ventures. Since many of these are sited in urban areas, they have promoted particular cities to the detriment of other centers and of rural areas. This having been said, several caveats are in order. In the absence of these biases, one would still find disparities. The very process of economic development involves the emergence, and persistenee over lona periods, of output and income per capita differences between rural areas, between urban centers, between regions, and even between rural and urban sectors, taken globally. 2/ These disequilibria are apparent at all stages of development except -- at one end -- in a subsistence economy and -- at the other end -- in a fully developed economy. Thus, under most circumstances, there will be incentives to move between regions, between occupations, and between economic subsectors. 1/ This experience is reviewed in A Review of Project Development and LooK Ahead, Washington, Urban Projects Division, 4estern Africa, The world Bank, 1980; M. Cohen, The Political Economy of Urbani Reform in Africa, Toronto, Canadian Association of Africaa Studies, May 1982; A. Lemer, Resource Mobilization, Cost Recovery, and Replicability, Easter-n Africa Urban Projects, Washington, Water Supply and Urban Development Division, Eastern and Southern Africa Projects Department, The World Bank, processed, July 1985. 2/ These references are to real income or output per capita differences. Nominal measures would vary even in long-run equilibrium because different areas or regions have different price structures. - 46 - For that same reason, it is meaningless to assume that investments per capita, be they in infrastructure or directly productive activity, should be roughly equivalent between city and country, or even between differenrt sized centers or different regions. Departures from equal allocations per capita are the norm even in an efficient model economy. Getting rid of "urban bias" does not lead automatically to the adoption of misplaced egalitarianism, nor should it. B. The Urban Sector After Structural Adjustment: A New Beginning The discussion of "urban bias" in Sub-Saharan Africa needs to be updated to take into account of the massive changes underway today. 1/ Macro- economic policies are being redirected and now increasingly support an agriculture-led development strategy. In addition, financial constraints, now and over the foreseable future, will make subsidization of urban infrastructure, urban public services, and urban-based parastatals, less and less attractive. The very foundations of "urban bias" are thus being destroyed. Evidence of this is available in report after report. Take, for example, the issue of urban wages and salaries. There is mounting evidence that both in absolute terms, and relative to agriculture, urban wages have suffered precipitous declines: a) Sudan: "[Civil serviceJ salaries had already decreased by 50% itn real terms during the earlier 1970's; and lost another 30 percentage points since 1978" 2/ 1/ Sandbrook, in The Politics of Basic Needs..., op.cit., argues that thie compression of income differentials between rural and urban areas and within urban areas has a longer history. 2/ Sudan: Prospects for Rehabilitation of the Sudanese Economy, Washington, Division EA2NE, Eastern and Southern Africa Regional Office, The World Bank, Report No. 5496-SU, June 1985. - 47 - b) Nigeria: "Unskilled wages in 1973 were about 80X higher in nonagricultural occupations than in agricultural ones. Over 1973-81 real agricultural wages rose 50% and in other sectors fell by 15%. As a result by the early 1980's earnings differentials were mucth less pronounced so that labor should have been willing to relocate into agriculture in the event of an alteration in employment opportunities. This indeed appears to have been what happened" 1/ c) Zaire: "One of the stark consequences of Zaire's economic and financial crisis has been the continuous decline in the real purchasing power of income for both public and private sector employees. Although in 1983 public service salaries were seven times higher in nominal terms than in 1975..., in real terms, they ...represented less than one-fifth of the 1975 level. A similar erosion in the level of income has affected employees of the private sector... As of June 1984, the implementation of the Lon going administration] reform had led to the firing of around 5000 civil servants, the retirement of about 17,000, and an annual reduction of roughly 25,000 staff since 1980. Such actions have reduced total civil employment ...to about 200,000 employees in 1983. 2/ d) Ivory Coast: Between 1980 and 1984 there has been a 25% decline in per capita real income. The drop in per capita income mainly affected the urban population while agricultural producer price increases helped maintain rural real incomes. The lowest 40% of the urban population is now worse off than its rural counterparts. 3/ e) Tanzania: Real wages of urban workers have declined by 50% since the mid-1970's. 4/ 1/ P. Collier, "An Analysis of the Nigerian Labor Melarket," Wasnington, Wesc Africa Programs Department, The World Bank, processed, July 1985. 2/ Zaire Economic Memorandum: Economic Change and External Assistance, Washington, Country Programs Department II, Eastern and Southern Africa, The World Bank, Report No. 5417-ZR, March 1985. 3/ Ivory Coast: Third Structural Adjustment Loan, Washington, The World Bank, June 1985. 4/ Tanzania Agricultural Sector Report, op.cit. - 48 - f) Ghana: In 1970, the average urban unskilled wage rate in the formal sector was twice the rate for casual farm labor. By 1984, the rural wage rate was three times higher than the official minimum wage. Today, most civil service salaries are very low in real terms; top civil servant wages are "scarcely sufficient for a family to survive for a week". 1/ g) Sierra Leone: Most civil servants experienced a 30-50% drop in real wages during the 1970's. 2/ As noted, the restructuring of relative wages is part of a larger reform, affecting relative prices. An excellent example of the impact of tnis wider effort has been documented for Zaire, 3/ once a major offender in terms of inappropriate macroeconomic policies. Between 1972 and 1982, real national output stagnated as modest gains in agricultural output were offset by declines in the secondary and tertiary sectors. The 20 percent increase in real agricultural output was accompanied by a 59 fold increase in agricultural prices, a rate which far outdistanced the 16 fold increase in industrial prices, and the 24 fold increase in service prices. As a result, the ratio of current nonagricultural to agricultural value added fell from 4.8 to 1.8. Since the non-agricultural population also rose much more rapidly than the agricultural population, the cited ratio, when adjusted to reflect per capita value-added, fell from 17.7 in 1972 to 3.5 in 1982. This provides a graphic illustration that future urban economic growth will be based on very different price signals than in the past. 1/ Ghana: Priorities for Public Expenditures, 198o-1986, Wasnington, Country Programs I, Western Africa, The World Bank, Report No. 5824-GH, August 1985. 2/ Sierra Leone: Prospects for Growth and Equity, Washington, Western Africa Region, The World Bank, Report No. 3375-SL, July 1981. 3/ M. Arnaud, "Esquisse de Schema d'Amenagement de la Region du Bas-Zair: Contexte Spatial et Evolutions a Long Terme," Kinshasa, Departement des Travaux Publics et d'Amenagement du Territoire, Government of Zaire, October 1985. - 49 - What is the likely future of the urban sector in the emerging age of agriculture-led development strategies? With the restructuring of production, there will be a shift toward labor-intensive, urban agroprocessing activities, and a shift away from import-intensive and capital-intensive outputs. This should encourage the growth of jobs in secondary urban centers, whose links with agriculture will increase. A redistribution of incomes, away from "formal" sector employees in and out of government, and toward farmers and the "informal" urban worker, should generate new demand for "basic" urban goods and services that rely relatively heavily on domestic inputs and laoor- intensive technologies. Finally, the urban sector will share the benefits of a rise in domestic savings, due to the removal of subsidies and price controls. This will help boost economic expansion and that, inevitably, will increase the demand for urban products. It is thus possible to foresee an economic basis for demographic projections which forecast a continuing growth of the urban sector at rates that exceed natural population growth. The structural adjustment process lays the foundations for growth relatively free of the problems generated by "urban bias." That growth, however, can only be fully exploited if urban institutions are made to function more effectively. The high political visibility of the urban sector, especially the largest cirit2s, provides an excellent arena for dramatic demonstrations of the beneeits of continued structural adjustment measures. Policies which have promoted inappropriate investments, inequitable subsidies, and unrealistic standards in cities can be modified, thus freeing up scarce resources for priority spending. The city can become a laboratory where more private initiative --in business ventures, shelter construction or public services-- can bei harnessed and puolic - 50 - initiative can be redirected. To refuse this opportunity to tackle the urban agenda, out of fear of residual "urban bias," is irresponsible and only guarantees a continuation of practices which must be done away with, sooner or later. C. "Excessive" Primacy as a Concern 1/ Often fears about urbanization or urban bias are mere proxies for concern about the size and growth of "very large" cities. Here the discussion dwells on three areas: the presumed high level of unemployment in large cities; the assumed upper limits to "optimal city size;" and the apparent need for a "balanced" system of cities, often reflecting the optimality of the so- called Rank Size Rule. 2/ None of these reasons has any merit. Without denying that the present structural adjustment crisis is creating unemployment problems in urban areas, most visibly in large cities, the fact remains that prolonged unemployment is not an option except for secondary household workers and school-leavers froin relatively affluent households. 3/ Higrant household 1/ Primacy refers gpnerally to the size of distribution of urban centers, particularly the relative weight of the largest cities in the overall urban population. 2/ The Rank Size Rule applies where the size of each center bears a fixed relationship to both larger and smaller centers. Each center's population equals that of the largest city divided by the smaller city's size rank. Thus the second largest city is roughly one-half the size of the largest; the third largest city is one-third the size of the largest, etc... 3/ Kahnert, Improving Urban Employment and Labor Productivity, op.cit.; P. Collier and D. Lal, Poverty and Growth irL Kenya, Washington, WorlT Bank Staff Working Paper No. 389, May 1980; P. Collier, "An Analysis of the Nigerian Labor MIarket," op.cit., P. Fallon, "Tne Labor Market in Kenya: Recent Evidence, Washington, Labor Market Division, Development Research Department, The World Bank, August 1985; Etude du Developpement Urbain de Bamako, op.cit. - 51 - heads must either work or move, thus being unlikely to create a "welfare" problem for cities, large or small. What may, and does, happen is that a substantial proportion of the labor force becomes employed in so-called "informal" activities which do not take place iTn large, capital-intensive offices and factories. "Informal" activities may thus offend some sensitivities, but that is hardly grounds for elevating the issue to the level of a muajor public policy concern. A parallel, and more serious problem is the alleged overma'aning of public and parapublic enterprises, which tend to locate in larger centers. But the solution to that issue is to reform the individual entreprises or institutions, not to impose city arowth controls. A very popular misconception which often appears in discussions of excess primacy" is the notion of optimal city size. The less sophisticated version of this hypothesis involves a belief that as cities grow, infrastruc- ture and public service costs first fall then bottom out and rise. The "inflection" point is thus associated with the optimal city size. Usually the size is listed as being between 100,000 and a million inhabitants. 1/ The fatal problem with this "optimal" size argument is twofold. First, it ignores the benefits associated with working and living in cities of different sizes. Unit cost minimization is no substitute for maximizing net benetits (i.e. benefits-costs). This same complaint can be lodged against those who want to control pollution or congestion by controlling city size; cLearly olne can control pollution and congestion through self-financed new investments, and the result may be efficiently to encourage further growth. Secondi, thie argument over costs is itself flawed. The observed structure of infrastructure costs across cities reflects both unit costs and the income 1/ See, for exam-ple, the discussion on urban areas in W4orld Development Report, 1984, New York, Oxford UTniversity Press, 1984. - 52 - levels which help dictate the quality or standards of investments. From the demand side, if households are willing to pay for a given level of service, "costly" though it may be, there is little to argue about. In addition, the unit cost curves of different services (one for each presumed level of quality) will begin increasing at different hypothetical city sizes. How tihen does one decide how to weigh all the various "optima" (one Lor bus teatworks, one for standpipes, one for paved roads, one for power, etc...)? Finally unit costs of different services can be very sensitive to city-specific factors. Population densities will affect the costs of the distribution networks ot water pipes or bus services. Geographic features will also play a role. Even the incremental nature or the infrastructure investment affects its unit costs; costs may differ as between adding services in a piecemeal fashion or all at once. Costs may also vary depending whether completely new production facilities (water, power) must be tapped. The unit cost issue then is hardly simple; and the optimizing rule is hardly adequate. 1/ Instead, as one observer has noted, Service costs and per capita expenditures imply nothing as to whether infrastructure service levels are excessive or deficient in particular places. That can be done only by careful benefit-cost analysis. The more fundamental question is what system of infrastructure financing and decision- making leads to incentives that induce appropriate infrastructure investments and hence appropriate levels of urbanization. 2/ 1/ For a good discussion of the problems involved, see J. Linn, "The Costs ot Urbanization in Developing Countries," Economic Development and Cultural Change, Vol. 30, No. 3, April 1982. 2/ E. Mills, "The Concept of Urbanization," NagarloK, Vol. XV, No. 2, April- June 1983. - 53 - The more sophisticated hypothesis is that, with each city size, there are associated economic benefits and costs of doing business. At some size -- and given a static environment where prices, technology, demand and other such -actors are frozen -- undoubtedly net benefits will be maximized. 1/ Identi- fying such a maximum and doing something about it is, of course, a ditterent (and unsurmountable) problem: First, data on all the relevant parameters are not available and statistical tools of measuremente are sufficiently crude that we can only confidently specify wide estimated ranges of any particular parameter value that we have the data to estimate. Second, in even the simplest numerical examples, the calculated efficient size can be very sensitive to alterations in the hypothetical values of parameters. 2/ In fact, a small alteration in the assumed parameters can alter "efticient city size" by a factor of 2 or 3. Finally, whatever the optimal size of a city may be, it is dependent upon several factors that change over time: the location and size of other cities, the level of national development, and the functions performed by that center. This leads one to conclude, with H. Richardson, that "It might be possible to dilute the concept of a unique optimum into the weaker proposition of an optimum size for an individual city at a specific location with specific functions at a particular phase of its development. Needless to say, this 1/ J. V. denderson, Economic Theory and the Cities, New York, Academic Press, 1977; H. Richardson, City Size and National Spatial Strategies in Developing Countries, Washington, The World Bank Staff Working Paper .lo. 252, 1977. 2/ J. V. Henderson, A Framework for International Comparisons of Systems of Cities, Washington, Urban and Regional Economics Division, Development Economics Department, The World Bank, Urban and Kegional Report No. 80-3, March 1980. - 54 - dilution generates so many optima that it can provide no guidelines to policymakers." 1/ A final argument on "excess" primacy makes reference to the cited Rank Size Rule. The Rule describes but one of an infinite number of possible size distributions, i.e. the relationships of diffErent sized cities to one another. It is not uncommon for analysts to pass judgements on the city size distribution according to its deviation from the Rank Size Rule: The Rank/Size distribution for Nigeria ... shows ...there are still gaps in the hierarchy of settlemenits in many areas... 2/ T'ne population of each city multiplied by its rank should be equal to the population of the largest city in the s,sttem... [Excess] primacy is present if there are fe:!r cities of intermediate size than would be expected from the rank-size rule. [Aside from Mombasa] the other intermediate towns [in Keniya] have a much smaller population com- pared to Nairobi, than would be expected in normal rank-size distribution... 3/ The Rule however has no more prescriptive use than the other standards considered earlier. For urban Africa, a recent study of 24 countries (including South Africa) found one country -- Burkina-Faso -- whose urban system is approximated by the Rank Size Rule. Seven others had less 1/ City Size and National Spatial Strategies..., op.cit. 2/ P. Sada and A. Onokerhoraye, The Emergenice of Secondary Cities and National Development in Nigeria, Dakar, Ninth (USAlD-Sponsored) Conference on Housing in Africa, April 1984. 3/ W. McKin, "Patterns of Spatial Interaction in Kenya," in R. Obudho and D. Taylor, eds., The Spatial Structure of Development: A Study of Kenya, Boulder, Westview Press, 1979. - 55 - primate distributions than suggested by the Rule, and the rest had more primate distributions. 1/ Alone, these findings are meaningless. The facts thiemselves suggest little ground for alarm concerning "primacy." Adopting a realistic cut-off point for defining a "large" city -- 500,000 residents or more 2/ -- one finds 28 such cities, 9 of which are in Nigeria alone. The percent of the urban population living in such centers is low by international standards: for the low-income semi-arid countries the proportion is zero; for other low-income countries in Sub-Saharan Africa it is approximately 32%. 3/ Middle-income oil importers weigh in at 35%, and the middle-income oil exporters at 25%. 4/ Using the share of these large cities in the total population as a guide one finds only eight cities whose share of total population exceeds 10%: Kinshasa, Accra, Lusaka, Luanda, Harare, Abidjan, Conakry, and Dakar. Only Dakar and Conakry have more than 15X of national population within their boundaries. 5/ Yet another way of looking at this issue is to review the individual size of the largest cities in each country. There are only nine "millionaire" cities, and of these only Lagos, Kinshasa, Ibadan, Abidjan, and Khartoum have 1/ >W. Mali, Size Distributions of Urban Areas in Africa, unpublished M. A. Thesis, Vanderbilt University, 1983. (Cited in C. Becker, C. de Bodisco, and A. Morrison, Urban Africa in Macroeconomic and Microeconomic Perspective: Issues and Options, op.cit. 2' That is the figure utilized in the Bank's World Development Reports. 3/ These figures are drawn from Towards Sustained Development in Sub-Saharan Africa..., op.cit, except for corrected urban population data drawn for the ILTA report for Benin and Guinea. The proportion listed in the Bank report for Nairobi's share of the urban population is off by 20 percentage points; it should read "37%." 4/ See footnotc ' for sources. Corrections were made, using IL'Ta data, for urban population totals in Senegal, Lesotho, Botswana, and, most significantly, for Nigeria. 5/ The percentages are 16% for Conakry and 22% for Dakar. - 56 - populations well over I million. By international standards, few cities in Africa would be considered very large. What can then be said on the subject ot primacy that is not in dispute? The existing literature on the evolution of the size distribution of cities over time suggests that, for any one country, there is a tendency to begin the process of development with a relatively dispersed urban structure anchored around poorly-linked, economically autonomous subsystems. Development tends to bring selective improvements in inter-regional transport and communications investment, often based on efforts to promote primary product exports. As noted, large centers tend to get access to urban infrastructure first, thus creating economic "city states" that can produce a varied mix of aoods and services with far less infrastructure than it would take if a large number of smaller centers were so equipped. These same centers provide, at this stage of development, easy access to monetized markets, and they attract complementary inputs liKe specialized labor and business services. Eventually, dispersion becomes apparent, as regional centers begin to grow more rapidly than the largest cities and a process of deconcentration gets underway. Many factors appear to be involved: improved access to all parts of the country, improved infrastructure in regionial centers, income growth in the "periphery" due to exploitation of local resource endowments, and the spinning off of activities originally lodged in large cities but later forced out by the relatively high costs of doing business there. I/ If the policies in place allow producers -- public and private -- to choose their (spatial) production decisions so as to minimize 1/ Documentation of the trends in concentration of urban population is available in J. Parr, "A Note on the Size Distribution of Cities Over Time," Journal of Urban Economics, Vol. 18, 1985. - 57 - the "scarcity" costs of producing their desired bundle of goods and services, then the resulting cities will be approaching their efficient size, whatever that may be. In this sense the best spatial policy is a set ot efficient macroeconomic policies. Among policies which could have an impact on the relative growth of different sized cities over time one can list four. First, as noted, the interregional transport and communications network strategy followed by a country during any given period has spatial consequences. These act to limit the dynamism exhibited by any set of smaller, emerging centers. Urban development efforts concentrated on promoting isolated urban enclaves makes little sense; as does neglecting cities in areas of rapidly improving interregional access. Second, development strategies that rely heavily on capital-intensive, import inputs-intensive industrialization are liKely to favor the growth of larger cities, especially ports. Third, policies that subsidize the costs of living or doing business in urban areas are most likely to attract business to the most favored, usually the largest cities. With subsidies and price controls, there is no feedback from the increasing cost of living in large cities to individual household and business decisions on where to live and do business. Yet it is this mechanism which has proved most effective in promoting urban decentralization among developed countries. Finally, the dearee of local government autonomy affects thte economic viability of smaller centers. Typically, when decision-making authority is concentrated in central government ministries, smaller centers cannot capitalize on their emerging economic strength by taxing themselves and borrowing funds to improve their infrastructure. Simultaneously, larger centers tend to get more infrastructure investments than they would choose if the costs were borne locally. - 58 - IV. A Strategy for the Future A. A Case for Strong Local Government Looking back over more than two decades of independent government across Sub-Saharan Africa, one generalization can be made which is widely applicable: central governments attempted to extend control over all aspects of economic life in an effort to enforce their vision of development against the "whims" of other sources of initiative. Instead of success, chere was a decline of institutional infrastructure matched by a growing disengagement of the population into parallel chainels of activity in which the state became increasingly irrelevant. 1/ One major victim of central government attempts to stifle independent initiative was municipal government. In this instance, withdrawal into parallel markets was not possible; local governments simply withered: "top government officials took authority away from local governments under tie illusion that they were thereby acquiring more power. They discovered instead that they had gained the negative power to prevent other political actors from taking independent initiatives, but not the positive power to implement policies that would improve social welfare. 2/ Where they were once relatively strong, local governments had their financial autonomy curtailed and their freedom of action severely circumscribed. In some cases, such as in Tanzania, local government was actually abolished during part of the last decade. in others, such as in 1/ Among interesting case studies, see N. Chazan, An Anatomy of Ghanaian Politics: Managing Political Recession, 1969-1982, Boulder, Colorado, Westview Press, 1983; and G. Hyden, No Shortcuts to Progress: African Development Management in Perspective, Berkeley, University of California Press, 1983. 2/ N. Kasfir, "Designs and Dilemmas: An Overview," in P. Aawhood, ed., Local Gov cnment in the Third World: The Experience of Tropical Africa, New York, John Wiley and Sons, 1983. - 59 - Nigeria, municipal governments saw many of their functions taken over by parastatals and state governments. Some municipalities, such as 'Mlonrovia (Liberia), were given few internally-generated revenues: Central Government has pre-empted most significant sources of local revenue including the property tax, typically the most common source of municipal revenue elsewhere. City income includes minor fees for market stalls, burials, trash collection, and city court fines totalling only about US $164,000... These are the only funds Lthe local governmentJ... has authority to spend directly. 1/ Across Francophone and Lusophone Africa, local government at Indepen- dence was embryonic to begin with and, after two decades, was still subject to the capriciousness of highly centralized decision-making. Even in Anglophone countries, like Kenya, which maintained a commitment to local authority, some traditional sources of revenue were curtailed even as new responsibilities were added. And yet, the process of structural adjustment and reconstruction -- with its commitment to a retrenchment of central government, greater resource mobilization by public bodies, encouragement of private initiative, and the preservation and rehabilitation of existing assets -- cannot succeed without the simultaneous strengthening of local government, especially in the cities and towns of the region. It is in these cities and towns that much national economic activity taKes place; and it is also at the local level that all alternatives to viable autonomous local government have reached the limits of their competence. These alternatives include not only the overextended central authority at the local level; it also encompasses private voluntary organizations which, though often praised for their selfiess actions, readily 1/ Staff Appraisal Report: Liberia, Monrovia..., op.cit. - 6u - admit that, given the ' tude of local governance requirements, much of what they do in this field is more symbolic than substantive. 1/ Municipal governments must have the ability to govern the cities and towns. They must plan investment programs, implement them, and maintain the resultant infrastructure and services. For this, municipalities must have autonomy as well as institutional competence. The Center's role does not disappear in such a scenario; it is trimmed back and refocused. There are, for example, the important oversight functions: there must be mechanisms in place to ensure minimum standards and administrative efficiency. The tasks involved include regular auditing of accounts; prescribed standards for staff qualifications; periodic inspections by central supervisors and approval and review of procurement procedures. In addition, the Center has a role to play in administering infrastructure development loan facilities, incentive grants, and other such funds. The Center and its donor counterparts will need to do one more thing: have patience. As one analyst concluded, after studying a number of local government experiments in Africa, The case studies show how easy it is for an autonomous body, left to itself, to run into trouble through lack of resources, external competition from public agencies and individual corruption, or inefficiency. It takes time for a new institution to find its feet, to be tested for its response to the needs of the situation, and to be intelligently modified. ˇiany of the cases... include the story of local government 1/ "...groups outside local government... are often successful in getting togetlher large quantities of money and administering it honestly on projects for the public benefit. They cannot provide a substitute for local government; the voluntary principle favors single one-off efforts, but works less well when it comes to year-by-year routine maintenaiice," (P. Mawhood, "Decentralization: The Concept and the Practice," in P. Mawhood, ed., Local Government.., op.cit. - 61 - structures planted and hastily uprooted because they were not producing instant results. [Yetj... both councillors and officials need time to build up their experience before they can begin to work effectively. The period needed will vary with circumstances and cultures, but if the first apprenticeship is passed in as little as ten years, that will be a cause for congratulation. 1/ B. Evidence of Increased Local Self-Reliance In Sub-Saharan Africa one model which has mutch to comamend it is found in Zimbabwe, where local governments are strong and, not coincidentally, municipal facilities and services are the best in the entire region: Government financial assistance to local authorities... is quite limited. There is no system of revenue stiaring, with only modest grants for general administrative purposes which are being gradually phased out. There are no subsidies for local public services. Government loans are made to local authorities for capital projects...Funds are provided on a project-by--project basis, with the Government appraising the projects financed. The Central Government... exercises a number of financial control over local authorities, including approval of budgets and all borrowing powers... Zimbabwe's cities have well-developed public services and are in sound financial conditions. They produce detailed budgets and accounts and maintain strict control over expenditure. The principal of cost recovery is stringently applied to services that can run commercially... and tariffs are raised as necessary to keep pace witn costs. Revenue collections operate well and the accounts do not show any substantial arrears. 2/ The gap between this model and the reality elsewhere is large. Sustained external support will be needed as local institutions are rehabilitated and developed. But examples exist, in many Bank-funded projects, 1/ P. Mawood, "Coniclusion: The Problem of Decentralization for Governmnents," in P. Mawood, ed., Local Government..-, op.cit. 2/ Staff Appraisal Report: Zimbabwe: Urban Development Project, op.cit. - 62 - of increased municipal self-reliance. 1/ Bamako's tax revenues tripled in US dollar terms between 1979 and 1985; and are expected to grow at 14 percent per annium over the foreseeable future. This has come about through better enforcement of existing tax sources and through new taxes introduced during the Bank's first urban project. In Bujumbura, local revenues rose four-fold in terms of US dollars between 1980 and 1985; by the latter year, 80 percent of revenues were generated by taxes approved during the first urban project. Accra, benefitting from a re-evaluation of its property tax base, expects to almost triple locally-generated revenues between 1984 and 1969, expressed in constant terms; this includes a projected increase in property taxes that by 1989 will be 4 1/2 times the 1984 level. Antananarivo, whose receipts fell 30 percent in real terms during the 1970's, is forecast to double revenues between 1983 and 1994, after inflation is taken into account; central government grants, held constant in value terms, will fall from about 1/3 of the total in the first year to less than 1/6 by the end of the period. And Lagos, as part of the Bank's solid waste and storm drainage project, has the potential to increase its property tax collections from US U15 million in 1984 to US $56 million. I/ Examples are drawn from "Mali: Second Urban Project: Yre-departure Meeting for Appraisal Mission," Washington, Western Africa Projects Department, Urban Division, The World Bank, May 14, 1985; "f5urundi: Proposed Secondary Towns Project Identification Mission: Back-to-Office and Full Report," Washington, Eastern africa Projects Department, Water Supply and Urban Development Division, The World Bank, August 23, 1985; Ghana: Accra District Rehabilitation Project, op.cit.; Staff Appraisal Report: Democratic Republic of 'Madagascar: Urban Development Project, Washington, Eastern Africa Projects Department, Water Supply and Urban Development Division, The World Bank, May 24, 1984; Staff Appraisal Report: Nigeria: Lagos Solid lWaste and Stormdrainage Project, Washington, Western Africa Projects Department, Urban Division, The World Bank, August 26, 1985. - 63 - C. A Selective Agenda for Local Governments With increased self-reliance comes the possibility of fulfilling a key macroeconomic objective: the preservation and rehabilitation of existing infrastructure and services, along with efforts to make more effective use of these existing assets. Two examples of this type of benefit can be found among Bank urban projects. In Accra a project was recently developed to support the Government's economic recovery program. It provides for selective new investments to remove infrastructural bottlenecks constraining the road movement of exports and imports, along with a maintenance program for the city's road and drainage networks. Economic rate of return ranging froi 45 percent to 155 percent were calculated for the various road sections being completed or rehabilitated. The rates of return for the periodic and rcutine maintenance work are estimated to vary between 30 percent to 46 percent. 1/ Similar results are likely in major Eastern Africa urban centers such as Dar es Salaam, Tanzania; Kampala, Uganda; and Kinshasa in Zaire; where civil strife and/or neglect has blighted urban areas over the recent past. There are also examples where better use of existing facilities has made planned new investments unnecessary. One of the best cases of this emerged in Abidjan during the implementation of the Bank's first urban project. Better use of roads and bridges--through the employment of traffic management measures and the introduction of preferential lane treatment for 1/ Ghana: Accra District Rehabilitation Project, op.cit. - 64 - buses--allowed the government to defer new investments totalling more than US $400 million. 1/ The challenge to urban government does not end there. They must ease the adjustments to rapid growth through locally replicable solutions that increase serviced land development and tap latent private initiative. At the rates of urban growth forecast over the next 15 years, urban Sub-Saharan Africa will grow by an average of 8 million to IU million residents per year, or the equivalent of replicating Lagos plus Xinshasa annually. This will call for 80,OOU to 100,UUU serviced hectares per year for all uses. 2/ The costs of not responding to this challenge are evident across Africa, as both Lagos and Kinshasa demonstrate vividly. This task requires the coordinated introduction of affordable and unsubsidized infrastructure across thousands of cities and towns. In turn, this performance is only possible if certain conditions are met: a) the public sector must disengage from activities that subsidize the city dweller; b) new infrastructure standards are needed that are lower than those that typically prevailed during the construction of the colonial cities; c) there mtust be a serious effort to conserve public resources by encouraging private investments; d) a new system of intergovernmental relations should focus on the creation of consolidated, self-financing infrastructure deve±opment funds which allow cities and towns to compete against one another for investment capital on equal terms; and e) institutional reforms are needed that will permit the development of minimally coherent city-wide development strategies. 1/ C. Tager, "Sub-Saharan Urban Africa in Macroeconomic Perspective: West Africa Project Evaluations," draft report, Washington, Water Supply and Urban Development Department, The World Bank, November 1965. 2/ This assumes a minimum of l0U square meters per Lnew resident, for all purposes. - 65 - The case for progressive disengagement is as easy to justity economically as it may be difficult to implement, politically. Misguided and unaffordable efforts at equating equity with subsidies, rather than with access to services, have failed across Africa. In country after country, central governments are facing up to the hemorrhaging of public funds directed at subsidized urban shelter and public services for the few. At a minimum, and often as a result of Bank-funded demonstration projects, countries like Senegal and the Ivory Coast are negotiating performance contracts which specify better performance accompanied by sharply declining infusions of central government funds. In both countries, transport and housing parastatals have been disciplined and, in some cases, scheduled for divestiture. From the building codes in Nairobi to the sites and services developments in Bamako and in Bauchi state (Nigeria), Bank projects have managed, in little more than a decade, to make significantly lower and more appropriate standards acceptable across much of urban Africa. This 'has a significant impact not only on the quality of resiaential developments but also on the efficiency of the productive sectors, especially informal activities, that function in their midst. The small-scale businesses that are so prominent in these local economies require careful tending; for the informal entrepreneur unaffordable infrastructure standards are as damaging as ao services at all. For the end result is that most neighborhoods are neglected because cost recovery strategies are made impossibly burdensome. Thle stage is now set to introduce, on a wide scale, the other components naeded to ensure the success of a strategy of rapid serviced land development. The first, engaging the energies of the private sector, is slowly falling in place. The very first Bank urban project, for sites and - 66 - services work in Senegal, was meant to leave much of the investment up to the individual households, and these households responded. For every US dollar invested by the Bank as part of the project, more than eight US dollars were invested by households benefitting from the project. 1/ What's more, this "multiplier" effect encouraged the growth of sectors with relatively high domestic value added, and diverted funds that would otherwise have been spent on imported goods. This has obvious macroeconomic implications and supports the structural adjustment process. In Zimbabwe, the governmen-t has engaged the energies and financial resources of private savings and loans building societies to help finance serviced land development. Where once municipalities were mortgage holders as well as developers of serviced plots and housing solutions, now the building societies hold the mortgages, financing up to two-thirds of recoverable development costs. Because local governments can now recover most investment costs attributable to serviced land development immediately, the same resources can finance up to one-third more infrastructure investment per year than before. 2/ The same general principle is being introduced in the bank's first urban project in Malawi. Kenyan authorities have gone one step further. As part of the Bank's third urban project, a subdivision owned by a private developer in Eldoret is part of an experimental program. The developer is ceding riants of way and land for public facilities to the municipality, which, in turn, provides the requisite infrastructure. The developer then gets private 1/ C. Tager, "Sub-Saharan Urban Africa in Macroeconomic Perspective: West Africa...", op.cit. 2/ Staff Appraisal Report: Zimbabwe Urban Development Project, op.cit. - b7 - financing for plot development costs and repays the municipality at the time of sale. All these mechanisms provide interesting preparatory steps for the more ambitious effort to establish consolidated revolving infrastructure development funds. The fund agencies would toLlow many of the steps that the Bank itself requires, as part of project preparation. Both project proposals and local government financial and institutional capacity would be appraised when fund requests arrived from individual municipalities. Weaker municipali- ties would be provided with technical assistance in the initial phases of such an operation. Eventually municipalities would self-select themselves for participation in fund proceeds, by competing with one another in terms of both projects and municipal government performance. This concept, already being adopted in countries like Brazil and the Philippines, is now under consider- ation by many more. Nigeria, Zimbabwe and Kenya may provide the first test of this approach in Africa. There is one final item on the agenda: an improved urban management process. 1/ Administrative and institutional reform at the local level must parallel conventional infrastructure project efforts, focusing on self- contained hardware investments. As new local resources are mobilized, more attention must be paid to creating a local development strategy laying priorities for municipal development, the policies required to implement them, and the investments needed to assure their realization. This framework should inform single sub-sector project development so that they can mutually-support one another. 1/ This section draws heavily on D. Williams, "T'he Role of the b3ank in UrDan Management," Wiashington, D.C., Water Supply and Urban Development Department, The World Bank, processed, October 1965. - 68 - In the past, governments and donors have emphasized the preparation of physical investment projects by agencies often only tenuously connected, if at all, to the local government whose presumed responsibility it is to coordinate local development and to maintain the resulting works. The result was a proliferation of weak or missitng links between conventional project investments and the rest of the area's infrastructure. Many of the problems identified earlier in the section on infrastructure deficiencies are symptoms of weak urban management capacity at the local level. It is also important to put into place a systematic strategy to manage miunicipal assets and services, including the development of measures to monitor, assess, and modify operations and maintenance performance. To date inadequate attention has been given by the Center and the donors alike to issues of operations and maintenance. Yet the conventional approach to projects development is often guilty of triggering a management crisis at the local level by failing to coordinate the creation of assets with the definition of responsibilities and revenue sources to preserve them thereafter. Over time these issues can only grow more important. Many cities in Sub-Saharan Africa can be expected to grow very large over the next decade or two. Capital investments will proliferate in number and grow in size as more cities begin to experience the problems of a Lagos or a Kinshasa, an Accra or a Dar es Salaam. The ability of local governments to select, coordinate, and preserve infrastructure service packages will be tested continuously and the results will help determine whettier efficient economies can develop in Atrica. - b9 - D. The Need to Improve the Operation of Factor Markets Many of the measures discussed so far require that attention be paid to factor markets, especially la.id. There is a serious need to develop land information systems that will not only make resource mobilization more effective, but will also allow entrepreneurs and households to freely trade land whose ownership is not in question., Land information systems can be devised which are progressiveLy upgradeable. At its simplest, the responsible officials can record only the information needed for property taxation. The name of the owner, owner's agent, or occupant can be assigned to each plot, along with the rough size class corresponding to the land area occupied. A size class and quality of construction category can also be assigned to any structure on the lot. This type of system has been used successfully in Burundi, as part of the Bank's first urban project. It will also be utilized in the Lagos Solid Waste and Stormdrainage Project. More elaborate systems can aiso be utilized, where feasible. A good example of the type of effort required can be found in the case of the Monrovia urbant development project. As part of that project, US $2 million have been allocated to help the government create a survey and ownership registry for the city. In Liberia, the state is the residual claimant to any land plots whose private ownership cannot be established. On an area-by-area basis, an adjudication unit of the Probate Court is racording tentative owners and staking out presumptive property borders. Title searches are being made, where possible, back to the original grant or sale from the state. tWnere disputes arise, hearings are lield and referees make final assignments. These are followed by official pcoperty surveys and the establishment of permanent ownership records. - 70 - More modest, but also interesting, is an effort underway in Bamako, under the Bank's second urban project, to survey a substantial port,4n of the city, and to provide tenure regularization, through the systematic regis- tration and compilation of all occupancy permits, which can be bought and sold on the market as elementary titles. Such efforts deserve more support from central government and donors alike. As with land market issues, those involving loans to smaller-scale enterprises can be tackled if accumulated experience is used as a guide. I/ Improvements in capital market access for smaller firms begin at the macroeco- nomic level. As uith most forms of subsidization in poorer countries, subsi- dization of interest rates on loans to small businesses constrains access to funds by making such loans unremunerative to the supplier. Interest on such loans will have to be higher than those for larger businesses if only because, during a "breaking in" period, the administrative costs and the risks involved w7ill be self-evidently higher. Lenders also face costly and time-consuming .-,cedures in dealing with many small loan applications, when compared with an established, larger firm. Per unit of funds lent, more applicants must be reviewed; more approvals must be aranted; more legal titles to collateral must be confirmed; and once approved, loans must be more carefully super-Vised, and higher-than-average expected arrears dealt with. Small loan borrowers also face serious problems. The formalities involved in loan applications are formidable. Countless'visits must be made to the banks, by the ownier-manager- 1/ Among useful review, see J. Levitsky, "Review of World bank Lenaiuig to Small Enterprises," Washington, Industry Department, The World Bank, processed, September 1985; and Zl. Harper, SmaLl b3usiness in the Tnird World: Guidelines for Practical Assistance, New York, Johin Wiley and Sons, 1984; and R. Blayney and M. Otero, "Small and Micro-Enterprises: Contributions to Development and Future Directions for ALD's Support, Washington, USAID, processed, November 1985. - 71 - supervisor, thus disrupting production. And appropriate collateral is often wanting. Once the interest issue is resolved, however, there are many, low costs ways of dealIng with the problems cited. Referrals and formalities can be filtered through local extension services, 1/ community banK managers, and trade groups or other nongovernmental organizations. Collateral bottlenecks can be alleviated by the development of guarantee or insurance fuinds co cover such loans. Financial institutions can also use self-selected mutual security group loans, where funds are guaranteed against default by several potential borrowers. Other features of the loans can be adapted to encourage a good credit record. The initial loans can be made both small and short-term, and directed at working capital needs; larger, longer-term loans can follow, based, in part, on payback performance. P'roblems with arrears can be dealt with by the use of substantial penalties and aggressive follow-up, leading to a restructuring, rescheduling or foreclosure of loans. 2/ The record of such carefully-crafted systems is generally very good, with high rates of return to income and with substantial employment effects. 3/ 1/ Extension services are needed, in any case, to provide advice on improved deployment of existing funds, efforts to separate business and personal resources, etc. Such services can be manned by carerully, but brietly, trained secondary school graduates, who are regularly and closely supervised, snd have access to specialist advice. 2/ Given the selective nature of this paper, it is impossible to discuss all the impediments to the growth and development of the small scale sector. Among other issues, two more can be mentioned: legal safeguards and regulatory impediments. Informal firms have no access to the courts to enforce contracts, to sue for damages, or to limit personal liability. These businesses have no access to sucn mechanisms as insurance. Finally, they do not participate in rulemaking involving zoning and health regulations. It is not unusual, as a result, for businesses to be formally banned from residential properties or be subject to removal during "resettlement" operations. 3/ Blayney and Otero, op.cit. - 72 - E. The Urban Sector and Public Expenditure Reviews Because of the strategic importance of urban areas in the narional economies of the region, it is important that the sector be considered in major analyses of investments such as those performed by the Bank in its public expenditure reviews. Unlike most examples cited in this report, those concerning these reviews are negative ones, illustrating the distance Detween perceptions of urban as a "social" sector and reality. 1/ The Bank reports vividly mirror the serious boundary proolem which obstructs the identification of intersectoral linkages both in the economy and in the investment programming. Furthermore these exercises appear to place on a featureless plain, so that geographic linkages are missed. Thus, for example, there is no effort to coordinate urban investments by area with the expected geographic focus of rapid agricultural growth. This occurs even where agricultural parastatals are being dismantled, as is the case in north- east Senegal's irrigated rice zones. In some cases, for example, Uauritauia, the bulk of agricultural investments are being directed at one set of regions (Gargol and Guidimaka), while new urban infrastructure is largely assiglned to other areas. 2/ Similarly, discussions concerning the agricultural 1/ This discussion is based on a cross-section of public expenditure reviews in Africa. Among the reports consulted are: Senegal: Public Investment Program 1985/86 - 1986/87, Review by The World Bank Staff, Washington, Western Africa, Country Programs Department II, The World Bank, draft report, November 1985; Ghana: Priorities for Public Expenditures 19bb-68, Washington, Western Africa Country Programs I, The Worid Bank, Report No. 5824-GH, August 13, 1985; Madagascar Review of the PUDliC Investment Program, Country Programs Department, Washington, Eastern Africa Region, The World Bank, November 15, 1984; Niigeria: Public Expenditure keview (1986-1990), Washington, Western Africa Rhegion, The World b3ank, Report No. 5920-UNI, October 29, 1985. 2/ "Republique Islamique de Mauritania: Development Urban: Elemnents Pour Une Strategie Sectorielle," Tjashington, West Africa Projects, urban Division, The World Bank, processed, elay 6, 1985. - 73 - development projects supported by the Bank in Nigeria make no reference to the urban centers in their midst which have a marketing and service role to play. The same applies to discussions involving Ghana's Bank-assisted Upper Volta Region Agricultural Development Project. Both in the countries themselves and within the donior agencies more effort is needed to identify which cities and towns have key developmental potential in support of agriculture; which policies and institutions neip or hinder the efficient symbiosis between town and country; and what bottlenecks exist that prevent the more efficient functioning of these towns in support of rural areas, especially with regard to infrastructure and services. Very recently, there have been some initial attempts to focus attention on these intersectoral issues. Studies have been undertaken by US AID and the European Economic Community to explore the rural-urban linkages. 1/ Within the Bank, a joint mission of agriculture and urban sector specialists has identified a regional development project for Kivu province, in Zaire. The proposed interventions in each sector complement one another and include the rehabilitation of market areas and feeder roads, as well as technical and financial assistance to private cooperative savings and loans associations in the area. Much more reimains to be done. The same conclusion can be reached about linkages between urban areas and industrial, service, and artisan activities in general. with few exceptions, these ties are ignored in the preparation of macroeconomic adjustment packages, 2/ even when major urban bottlenecKs to eff4cient 1/ Examples are cited throughout the body of this paper. 2/ A singular exception to this sectoral myopia is the explicit acknowledgement of links between rehabilitationi of Accra's roads and macroeconomic objectives, found in the cited Ghana report. - 74 - operation of these subsectors are explicitly identified. 1/ In addition, the heavy volume of investments in interregional transportation facilities appears to be taking place in ways that are delinked from any urban development strategy. F. Next Steps When the Bank's urban program began in 1972, a deliberate decision was made to follow an experimental approach, one of "learning by doing." This effort embraced the experience from mistakes as well as successes, This strategy was deliberately intended to be experimental. Project staff would explore different technical solutions with borrower agencies.... This research and development stage would be difficult... and would require a substantial investment in staff time.... The urban... program would be closely monitored, and early... tresultsj would be evaluated to determine their applicability to similar situations in other countries. 2/ Today much is known about preferred technical solutions to particuLar ,'ity infrastructure and public service problems. There is a firm basis in experience to deal with complex institutionial and resource mobilization issues. At the same time, new issues have arisein that were not anticipated in the 1970's. The most critical new sectoral problem in Africa is how to integrate an urban perspective into the process of structural adjustment and the renewal of growth. Though both the BanK and the national oureaucracies are afflicted by sectoral myopia, the time has come to once again be daring and experiment with cross-sector approaches. 1/ See, for example, Nigeria: Public Investment aeview, op.cit. 2/ M. Cohen, LTarning by Doing: World Bank Lending for Urban Development, 1972-1982, Washington, The World bank, 1983. - 75 - The temptation might exist to apply mechanistic formulas such as those detailed above under "paths to avoid." While a spatial perspective is needed, the explicitly spatial instruments that were advocated in the past have proved bankrupt. Growth poles, migration controls, infant region promotion, and optimal city size controls -- all these have been tried and found wanting. Now what is needed is to test out impact analysis tools which seek to answer questicls about how different urban initiatives might help to achieve objectives of rural developmenit or transport planning; inuustriai development or export promotion; resource mobilization or rehabilitationi of infrastructure. By incorporating a broader perspective into the process, planning for rehabilitation and development in Africa would be more realistic and more rewarding. It would no longer be possible for "gatekeepers" to assign to the urban sector a percent of Bank or country investment funds based on past shares or on the relative priority granted narrow sectoral objectives treated in isolation. Instead, investment priorities should be determined first by the contribution each proposal, regardless of sector of origin, makes toward attaining critical development objectives. In the future the urban sector should be allowed to compete for funids on the basis of its demonstrable impact on the success of investmen-t programns in other, apparently unrelated sectors. A start has been made, in Accra's rehabilitation project and in Zaire's Kivu province regional development project, for example. elMuch more must be explored. The urban sector in Sub-Saharan Africa cannot be neglected either by governments or donors. Its level of efficiency affects the operations of most other conventionally defined sectors. To set urban issues aside as an unnecessary complication" is to deny the importance of intersectoral and - 76 - geographic linkages. In addition such inattention threatens the otherwise strategic contribution that the better maanagement of cities and towns can make to a host of structural adjustment objectives. Africa cannot recover uniess the cities and towns are made partners in the process. Africa will not develop unless the economic role of cities is acknowledged.