Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
9 The Challenge of Urban Governance The preceding analysis of the demographic, social, and economic aspects of the urban transition has emphasized three important features. First, we have demon- strated the inevitable tendency for the global population to become urbanized. Second, we have illustrated the growth of a system of urban places at both an international and national level dominated by very large urban regions that are responsible for generating a significant proportion of national wealth. And third, we have argued that these large urban regions are functional because they have been integrated through improvements in the transactional environment of the flows of people, commodities, capital, and information. One of the least well- understood elements on this list is information. The information requirements for the management of large cities are impor- tant conditioning factors in the cities' institutional elaboration. With the spread of democratization and decentralization now under way almost everywhere, lo- cal governments are increasingly being required to operate with the speed and efficiency of private business while facing ever more complex political and regu- latory issues. Local governments must digest an immense amount of information to perform their duties in a fair and efficient manner. Yet data are severely lack- ing in many cities throughout the developing world. The typical 10-year interval between censuses presents a problem for policy analysis and planning for large cities, where the metropolitan population may easily grow by 1 or 2 million in- habitants in 5 years. Moreover, demographic data at the subnational level typically are not processed or made available for several years after a national census has been conducted. Governments could also benefit greatly from data on population size, growth, and composition; differential fertility and mortality rates; and socioeconomic char- acteristics for numerous small-area units, which could then be aggregated to an- alyze trends within various parts of a metropolitan region. Employment data are particularly important here, since population growth has outstripped job growth 355
356 CITIES TRANSFORMED in urban places in most developing countries. These data are also lacking or at least not readily available to planners and policy makers. Planners often have only rudimentary knowledge of the numbers and characteristics of recent migrants in most large developing-country cities. Moreover, cities' own projections of their future population growth often have been widely off the mark. Planning is further hampered by limited information about local land markets. Most large cities lack sufficient, accurate, and current data on patterns of land con- version and infrastructure deployment. Frequently, urban maps are 20 to 30 years old and lack any description of entire sections of cities, particularly burgeoning periurban areas. In most large cities in developing areas, each municipal agency or department typically maintains its own database, often using differing standards and rarely sharing data. Computerization of such data is still relatively uncommon. Many municipal agencies continue to rely on paper files and maps, which are often stored in formats and at scales so diverse that they cannot be compared or col- lated and are not easily updated (Bernhardsen, 1999~. There are almost no exam- ples of integrated databases for the constituent parts of large metropolitan regions. This is hardly surprising given the large number of political entities constituting most metropolitan regions, with their differences in age, socioeconomic needs, and financial and managerial capacity. Typically, there are reasonable data for the central city, with data (often noncomparable) for the outlying municipalities being of varying quality. This chapter is concerned with the political and institutional implications of the urban transition. For some commentators (Ohmae, 1990; Scott, 1998), the urbanization process is a result of powerful global forces that are, at the same time, leading to a reduction in the political power of the nation state. Whether or to what extent this may be the case, there appears to be ample evidence that the urban transition of the twenty-first century will involve a significant renegotiation of political relationships between national and urban governments. Just as the present system of governance grew out of the replacement of city states by national states (Mumford, 1961), new systems of governance may very well emerge from the current urban revolution. Underlying these political questions is one fundamental fact. The way in which present patterns of urban growth are being framed territorially differs from that of earlier periods of urbanization. Previously, government was organized on the basis of rural-to-urban responsibilities that were defined both spatially and structurally. Government was organized vertically so that tasks were divided among the national, provincial, and county levels, as well as highly local units of urban administration such as cities and towns. Each of these levels of gov- ernment was defined in terms of discrete spatial areas, although there were often overlaps in administrative responsibilities. Today there is general agreement that the rapid urbanization of developing countries should involve some reorganization and even reconceptualization of
THE CHALLENGE OF URBAN GOVERNANCE 357 systems of governance. For example, the increasingly close interaction between central cities and their rural hinterlands (made possible by improvements in trans- portation and information transmission, among other things) suggests that rural and urban jurisdictions need not always be administered separately. And the fact that very large urban regions (including their hinterlands) are in many countries the source of both major economic initiatives and advances in national productiv- ity suggests that these city-regions should not remain under the restrictive control of other levels of government if they are to reach their full potential for devel- opment (see Scott, 2001~. Finally, the emerging importance of cities and city- regions in the world economy underlies the generalized shift from power relations based on pyramidal and hierarchical structures to those grounded increasingly in networks and horizontal relationships (Castells, 1997~. This chapter begins with a brief discussion of the concept of urban gover- nance and the issues involved, which are illustrated by the case of the Bangkok Metropolitan Region. We then undertake a detailed review of the major challenges of urban governance in developing countries. Next we examine the question of whether there is, in fact, a single "best" model of urban governance. The final section of the chapter presents conclusions and recommendations. THE CONCEPT OF URBAN GOVERNANCE While the concept of governance is widely used today, its common usage in the social sciences is a product of the last decade and a half. In a lengthy discussion of governance as applied to urban examples throughout the developing world, McCarney, Halfani, and Rodriquez (1995) find that an important element in the development process, explicitly lacking in many official and agency-based def- initions, is the connection of government, and particularly local government, to emerging structures of civil society. Accordingly, they define governance as "the relationship between civil society and the state, between rulers and ruled, the gov- ernment and the governed" (McCarney, Halfani, and Rodriquez, 1995: 95~. Im- portant elements of this definition were adopted by other researchers writing about comparative local governmentin developing countries (Wilson and Cramer, 1996) and were eventually incorporated into the United Nations Development Program (1997a: 2-3) current definition: Governance can be seen as the exercise of economic, political and administrative authority to manage a country's affairs at all levels. It comprises the mechanisms, processes and institutions through which citizens and groups articulate their interests, exercise their legal rights, meet their obligations and mediate their differences [emphasis added]. One reason for the emergence of the concept of "governance" or "urban gov- ernance" is that the context within which local government operates has become
358 CITIES TRANSFORMED much broader and more complex. In the United States, researchers dealing with metropolitan problems increasingly use the term "metropolitan governance" rather than "metropolitan government" because of the more inclusive connotations of the former (Stephens and Wikstrom, 2000: 47~. Similar ideas have begun to take root in Europe. In an important article on France (where a Law of Decentralization was first passed in 1981), Le Gales (1995) argues for a shift in nomenclature from "the government of cities to urban governance." While the term "local government" is associated with a formal description of powers and responsibilities of urban au- thorities, local politics and the way in which French cities are administered are changing rapidly. According to Le Gales (1995: 60), "the term 'governance' sug- gests . . . functions and actions of government, but without the idea of uniformity, rationality, or standardization. The term 'urban governance' implies a greater di- versity in the organization of services, a greater flexibility, a variety of actors, even a transformation of the forms that local democracy might assume, and taking into account citizens and consumers, and the complexity of new forms of citizenship." The emergence of a discussion of "governance" in a continental European context is of more than purely scholarly interest, since for some time the concept was considered an Anglo-Saxon term, and as such inappropriate outside English- speaking countries. Some Latin American researchers believed the more apt terms to be "governabilidad" and "governabilidade" in Spanish and Portuguese, respec- tively, which essentially mean "governability." But as Coelho and Diniz (1997) argue, the problem in the case of Brazil after formulation of the 1988 Constitution was not decision-making incapacity (or "ingovernability"), but rather the inability of leadership to achieve sufficient support and legitimacy to implement a whole host of technical measures. As a result, the authors propose retaining the concept of governability (governabilidad), but at the same time accepting a new concept, governance, to denote the state's command and steering capacity, its capacity to coordinate among politics and interests, and its capacity to implement measures from the level of the center to the local area. The concept of local governance, they add, brings to bear the political dimension "and places the interdependence of state and civil society at the center of the debate" (Coelho and Diniz, 1997: 113~. The intersection of state and civil society, in this view, takes place at all levels of government from the most local to the national. Discussing Africa, the French planning writer Jaglin (1998) first notes how the concept of governance permits the incorporation of a wide variety of actors and groups in both the formal and informal sectors, as well as local, national, and international groups and agencies. The situation in South African cities in the mid-199Os, however, was extreme. In the transition from an apartheid system in which the majority population was totally excluded from power to a (potentially) fully democratic local government system, an elaborate process of dialogue and negotiating forums ensured a relatively smooth movement from exclusion to in- clusion. Jaglin argues that the complexity and multisectoral nature of this process can best be described under the rubric of governance. "In imposing a form of
THE CHALLENGE OF URBAN GOVERNANCE 359 representation of the political class, the private sector and local residents (through the 'civics' based in black townships) at the center of the required forums, the law made governance official, since it obliged the different actors to negotiate before local policies could be considered" (Jaglin, 1998: 31~. Beyond the pure concept of governance, which to us denotes the neutral rela- tionship between government and the governed (mediated for some through cul- ture and political practices), many writers and policy makers now speak of "good governance." While one must be careful not to reproduce in such a value-laden concept all the sociocultural and institutional prerequisites of government in de- veloped countries, the notion of good governance carries with it a premise of institutional design that is at once open and accountable to civil society in gen- eral, and effective in terms of financial management and policy implementation. Good governance involves an effective balance between the raising of revenue and the proper expenditure of this revenue on services and investments that are based on accountable decisions. This model, in turn, implies that many levels of government and many local stakeholders and social groups will be involved. The task of analyzing these changes in the patterns of governance in develop- ing countries is complicated by the diverse paths and scenarios of urbanization and socioeconomic development that exist in the developing world. Yet the expansion of the urban built environment is occurring everywhere, and thus the development of systems of urban governance that can cope with urban expansion is a major priority. The issues associated with urban governance are illustrated by the example of the Bangkok Metropolitan Region (BMR), which consists of the Bangkok Metropolitan Administration (BMA), with a population of 8 million, and the five adjacent provinces. In 2000, the BMR was estimated to contain 11.5 million peo- ple. Adding the adjacent industrial heartland of Thailand creates an extended metropolitan region (EMR) of 17.5 million people, making up 28 percent of the total population of Thailand in 2000 (Webster, 2000b: 8~.i This region has been the focus of much of Thailand's economic development, involving a major change from an agricultural rice-exporting economy to an economy open to foreign direct investment (a high proportion from Japan) in the 1980s and a very rapid increase in export-oriented manufacturing. The result was a period of hypergrowth that made the Bangkok region one of the fastest-growing urban economies in the world, with an annualized growth rate of 17.5 percent in the period 1990 to 1996. With gov- ernment encouragement, this economic expansion also led to the rapid outward sprawl of the region from the city (Webster, 2000b: 9), resulting in increased traf- fic congestion; environmental pollution; and serious infrastructure problems, in- cluding inadequate provision of water and sewerage and a substantial proportion of the population in the periphery living in substandard housing. iMuch of the discussion in this section is summarized from Webster (2000b).
360 CITIES TRANSFORMED As the region grew in size and territory, no adequate system of governance emerged that could solve these problems in a coordinated manner. The city core, the BMA, is under the leadership of an elected governor; 60 elected councilors (one per 100,000 people), making up the Bangkok Metropolitan Council; and a large number of elected councilors in 50 local districts who advise district directors appointed by the governor. While it has a huge staff complement (num- bering 82,950, including teachers, in 2001) (Bangkok Metropolitan Administra- tion,2001), the Metropolitan Administration has limited powers although much greater than is the case for other local authorities in Thailand. Moreover, many of its key functions, such as water and electricity service, are under the control of national agencies. Outside the BMA, in the BMR, matters are even more com- plicated as provinces control most of the planning, and there is no overall plan for the EMR. Webster points out there are more than 2,000 local government authorities (many of them serving small villages) in this outer region of urban activity. The period of hypergrowth in Thailand was abruptly halted by the economic crisis that began in July 1997, which was focused on Bangkok. Between 1997 and 1998, the city product fell by one-half. By 1999 the official unemploy- ment rate had increased from 1.4 percent in 1997 to 5.1 percent, reflecting a decrease in construction employment. But the major effect was a decline in in- come among the urban informal sector, such as taxi drivers. Inflation caused in- creases in food and gas prices that also adversely affected the poor. Estimates suggest the slum population of the BMA grew from 1.2 million in 1996 to 1.5 million in 1998, and there appear to have been deteriorating social conditions, including increasing crime, drug use, and suicides (Webster, 2000b). In con- trast with the Indonesian case, however, there was no massive social and polit- ical disruption, for several reasons. First, households proved very adaptable- adopting strategies of outmigration, sharing of income, and increased participa- tion by women in informal-sector activities. Second, many foreign companies, particularly in the manufacturing sector, did not lay off workers but stopped giving bonuses; this was the case, for example, with a number of Japanese companies. As the baht devalued, these firms were able to take advantage of this strategy to export more cheaply. Finally, the government remained stable during this period. In fact, some writers have argued that the crisis benefited Bangkok, first because the city has been able to rebound more quickly than could either Vietnam or Indonesia, its chief competitors for export-oriented manufacturing in Southeast Asia; and second because the cost of living (including rent) has been substantially reduced, making Bangkok more competitive as a site for foreign investment. In addition to these factors, public works programs designed to provide employment have led to a decrease in pollution and traffic congestion. And the national govern- ment has been forced to accelerate its program of administrative decentralization to the local level despite very weak institutional capacity at this level. In this case,
THE CHALLENGE OF URBAN GOVERNANCE 361 important attempts to develop new systems of urban governance are being made during a particularly volatile phase of contemporary globalism. The urban form of the larger EMR of Bangkok is thus assuming many similar- ities to other large mega-urban regions in the developing world (see McGee, 1991, and McGee and Robinson, 1995, for a description of mega-urban regions in the Asian region.) Basically, the EMR is divided into three zones the core, suburbs, and exurbia on the basis of typical characteristics of each zone. While these three zones function as an economic region, there are sharp differences among them (see Table 9-1~. Is there a preferred governance model that best fits such a mega-urban region? There are essentially four categories of mega-urban governance. These categories reflect attempts in other parts of the world to manage large metropolitan regions, a subject to which we return in the penultimate section of the chapter. The first cat- egory, which can be called the fragmented model, is characterized by a myriad of autonomous local government units, each with jurisdiction over a particular func- tion and/or territory. There is sporadic and poor coordination among the various units. This model is the most typical of the American approach to metropoli- tan governance, though some examples exist in developing countries as well. A second category, which can be termed the mixed model, encompasses regional governance approaches in which both central and local government play a role in the administration of a region. This approach is typical of most mega-urban regions in developing countries, including the case of Bangkok discussed above. A third form of administration is the centralized model. This model, still found in transitional societies such as Vietnam, is dominated by a central government. Finally, we should mention the comprehensive metropolitan governance model, although no pure examples of this approach currently exist in developing coun- tries. In this model there is either a single coordinating governance unit for the whole mega-urban region or a two-tier system in which local governments (or mu- nicipalities) perform a number of local functions, but cede to a higher metropoli- tan (subnational) authority the performance of region-wide functions. A version of this model has been operating in Abidjan, Cote d'Ivoire, since 1980, and a "unicity" model has been emerging in South Africa since the local elections of November 2000. Four major Chinese cities that are governed as provinces also fall into this category. While there is no consensus on the type of mega-urban governance that may emerge in this new era of urbanization, there is general agreement that as the wealth of these regions grows, some new form of governance will be required. As mega-urban regions grow in wealth, however, they will need to develop sys- tems of taxation and governance that operate at both a regional and a highly local level, a process that will involve both decentralization and privatization. The next section reviews some of the major issues involved in both designing and implementing governance structures at the metropolitan level in developing countries.
362 · - ~ so i Q o o ~ .~ no so ~ ~= m v ~ O O ~ to .~ .e O ID pi ._ ~ O E¢-4 ~ ' ~ 3 ~ ~ E ~ so ~ I ~ ~ c c , E cow ~ ~ ~ E o ~ ~ ~ E ° ~ E ~= ~ E v ~ o ~ ~ ~ ~ ~ '> E 2 D 2 E ~3 j E ~ ' ~ E .. no o Cal o L~ ¢ . . V V:
THE CHALLENGE OF URBAN GOVERNANCE MAJOR CHALLENGES OF URBAN GOVERNANCE IN DEVELOPING COUNTRIES 363 As the example of Bangkok illustrates, the governance challenges of large metro- politan areas in developing countries are both diverse and complex. It is arguably the case that this diversity and complexity are increasing with population growth and globalization, and that as this process unfolds, new approaches to governance and the management of cities will emerge. To impose some order on a very large subject, we examine these major urban challenges along five major dimensions: capacity (with a focus on urban services and service delivery);financial resources (with emphasis on generation of local revenues); diversity (in particular, issues of inequality and fragmentation, often leading to violence and a failure to regulate social conflicts); security (involving crime and violence, and approaches to the preservation of public order and the alleviation of violence); and authority (with a focus on decentralization and distribution of powers, local jurisdictional configu- rations, and political participation). The literature on these interrelated topics is voluminous, but a certain disci- plinary specialization tends to attach itself to each of these dimensions: geogra- phers (with some economists) are more likely to focus on urban services, public finance economists on the financial dimension, sociologists and criminologists on the diversity and security dimensions, and political scientists and public adminis- tration specialists on the authority dimension. Each of these dimensions can be related to demographic dynamics. The sectoral policies adopted to address the challenges associated with each have implications for such demographic variables as urban migration, the differential treatment of gender and age groups in the population, and the quality of life of urban residents. The Capacity Dimension During the 1980s, rapid urban growth throughout the developing world began to seriously outstrip the capacity of most cities to provide adequate services for their citizens. Beginning in the 1960s, this incapacity was made visible through the increasing number and extent of slum and squatter settlements in the cities of de- veloping countries. The 1960s and 1970s were a period of extensive contention between municipal (and national) authorities and low-income urban residents. While the authorities sought to limit the use of urban land to the purposes for which it was usually zoned (i.e., high-income residential, commercial, and indus- trial uses), low-income populations attempted to build individual shelters on some of this land or to organize "invasions" for more concerted attempts to convert the land to their use. Clearly, the supply of cheap, serviced urban land fell far behind the demand for land on the part of a rapidly growing low-income population a
364 CITIES TRANSFORMED population that, in addition, had to be close to centrally located sources of income and employment to subsist economically. In many countries, government response to this demand originally took the form of setting up centralized housing banks and construction agencies. For ex- ample, the National Housing Bank of Brazil (established in 1964 and closed in 1986) produced around 4 million units; two major agencies in the Ivory Coast produced close to 40,000 units during the 1960s and 1970s; in Egypt, public hous- ing agencies built 456,000 units between 1960 and 1986; and in Singapore, some 460,000 units were built from 1969 to 1985 (UNCHS, 1996: 219~. While this housing made up a substantial proportion of low-income housing in a number of countries by the 1980s, there were major problems: maintenance was poor, public subsidies were high, it was difficult to avoid corrupt practices entirely, and the pace of construction was in any case inadequate to respond to the level of inmigration (Cohen, 1974; Mayo and Gross, 1989; Perlman, 1976; Stren, 1978~. Notwithstanding exceptions, such as Hong Kong or Singapore, in which high-quality housing was successfully developed and maintained (Yeung, 1998a: Chapter 4), international agencies turned to more collaborative approaches. Two policy responses stand out. The first is the sites and services projects of the 1970s and 1980s (Cohen, 1983), in which minimally serviced plots in large subdivisions were allocated to low-income applicants, who were expected (with some assistance in the form of training and loans of materials) to build their own homes. The second is squatter upgrading projects, which regularized land tenure and improved services and infrastructure in "slum" areas to encourage the orderly improvement of neighborhoods without displacing existing residents. Variants of these approaches are still operative today, but both involve complex, costly planning and administrative organization, as well as the design of an incentive system to encourage investment by the poor and discourage "leakage" to higher- income groups. From the building of housing units, public policy approaches in developing countries have shifted to "enabling" strategies that encourage land and infrastructure development, as well as support for medium-sized and small- scale enterprise (UNCHS, 1996~. Prominent among these strategies are reforms in the governance of urban services from arrangements based entirely within the public sector, to arrangements based on partnerships with private and nongovern- mental groups (Fiszbein and Lowden, 1999; Freire and Stren, 2001), to various kinds of arrangements with private service providers (Batley,1996~. Overall, these enabling and partnership strategies are consistent with what some observers see as a shift toward "neoliberal" economic policies, as applied at the local level. While the typical package of neoliberal policies (such as fiscal dis- cipline, reduction of public expenditures, deregulation, open exchange rates, and trade liberalization) was generally applied at the national level when first intro- duced, new approaches to reform include social safety nets and decentralization. In the case of the former, many countries began in the late 1980s to develop redis- tributive programs and agencies targeted at the very poor; these programs were
THE CHALLENGE OF URBAN GOVERNANCE 365 inevitably operated (at least partially) through institutional mechanisms set up locally though not, in most cases, within existing local government institutions. As for decentralization, while most studies of national reform policies in the 1980s and 1990s relegate it to a minor role, some authors claim it is an essential element of neoliberalism. Vilas (1996), for example, sees the elaboration of what he calls "neoliberal social policy" in the 1990s as having three basic characteristics: pri- vatization, targeting of the poor, and decentralization. Lowi (2000), with special reference to the United States but by implication including developing countries as well, argues that decentralization (which he qualifies as Revolution and links to privatization and deregulation) is a fundamental part of strategies to "address the spillover effects of extreme inequalities" caused by neoliberalism. Behind this assertion is the argument that local elites and political institutions in the United States are better able to manage the "fallouts" from continuing and increasing inequalities than are national institutions. One example of a large city confronting its service and infrastructure chal- lenges in an energetic and innovative fashion is Shanghai. With a population in its metropolitan area of more than 13 million (not including the transitory, unreg- istered population, which may account for a further 3.3 million) and a total land area of 6,340 square kilometers, Shanghai is the largest city in China and one of the largest in the world. The metropolitan area consists of 14 urban districts in the city proper (2,057 square kilometers) and six suburban counties, all within a single area with the status of a province within the Chinese system of government (Wu, 1999a: 207~. While Shanghai was a major industrial and commercial center from the mid- 1850s until the takeover of the Chinese Communist regime, it experienced neglect and disinvestment after 1949. Overall, from 1949 through 1983, some 87 per- cent of Shanghai's revenue was remitted to Beijing, leaving 13 percent for local allocation. By contrast, Beijing and Tianjin averaged 30 percent during the same period. As a result, Shanghai was known as the "golden milk cow" of the planned economy of China (Yeung, 1996: 9~. Limited investment in infrastructure and housing meant that urban amenities suffered, so that "by the 1970s and 1980s, the central city's infrastructure was near collapse. For instance, in the former French Concession, nearly 700,000 dwellings were without flushing toilets. Compared to the national urban average and Beijing (a city of comparable size), Shanghai lagged in several important indices of urban infrastructure, including per capita living space and per capita paved roads" (Wu, 1999a: 208-9~. The central government's neglect of Shanghai began to reverse itself in the early to mid-1980s, culminating in a 1988 agreement between the two to give the city more autonomy in revenue collection and expenditure. In the same year, the city established a foundation to mobilize funds for urban construction; in 1992, this became the Shanghai Urban Construction Investment and Development Company (Wu, l999b: 2277~. This state-owned company allocates funds in many different areas of infrastructure and services and to many different local agencies.
366 CITIES TRANSFORMED According to Wu (1999b: 2278), the company "has displayed an impressive record of achievement in infrastructure financing since its creation." Wu goes on to say that the company ... has employed a wide range of financing mechanisms, particu- larly through such non-state channels as international capital, bank loans and credits, construction bonds, stock markets and service con- cessions. It has entered into concessions with profit-making enter- prises to operate the three bridges and a tunnel across the Huangpu River. It also has established a number of subordinate entities, mainly in charge of water supply, which are listed on the Shanghai Stock Market. Available official information shows that in 1995 and 1996, funds mobilized by the company accounted for about 76 per cent and 90 per cent, respectively, of Shanghai's total urban maintenance and construction revenue. These systematic investments had begun to show results by the mid-199Os in terms of road construction, park expansion, and wastewater treatment. Some massive infrastructure projects were completed, including the beginning of a new subway line, a tunnel, and three large bridges across the Huangpu River. The most spectacular outcome of the new emphasis on modern infrastructure in Shanghai has been the development of the Pudong New Area, a virtually new district across the river from the old commercial center to the east of the city. Pudong has been designated the country's financial center, "the Wall Street of China" (Saywell, 2000: 58~. Major financial institutions, such as the Shanghai Security Exchange, as well as many large international corporations (including Philips, General Motors, Eriksson, and Xerox) have already located their offices in the Lujiazui area of Pudong (Wu, 1999a: 214-15~. By the end of 2000, some 70 foreign and joint venture companies had established themselves in Pudong's industrial parks, and some 80 foreign and domestic financial institutions were em- ploying an estimated 150,000 white-collar staff in 20 large office towers (Saywell, 2000: 56). As impressive as they are, these infrastructure achievements have done lit- tle to improve the living conditions of the large majority of Shanghai's working population, whose per capita living space is no more than 8 square meters on av- erage, with 10 percent of the population living in a space of less than 4 square meters (Wu, 1999a: 208~. One of the main problems with housing construction is the cost of urban land. And yet in the future, as land prices increase with ma- jor infrastructure projects and international investments, the cost of housing will rise even higher. The 1996 decision of the Shanghai municipal government to grant residency permits to those from outside the city who purchase an apartment in Pudong worth at least 500,000 renminbi (US$60,200) can only reinforce the polarization of the housing market.
THE CHALLENGE OF URBAN GOVERNANCE 367 In addition to the rising cost of land, a major factor behind the lag in housing has been a large increase in the city's population, caused partly by migration of rural peasants to urban areas. Most of these migrants (often referred to as the "floating population") do not have official urban status and live in substandard housing. Many live in so-called "villages" named for the provinces from which the migrants originally came. As documented by Solinger (1999), recent rural migrants in major Chinese cities survive under very depressed conditions with respect to access to such major city services as water, electricity, transport, and even cheap food. As one Shanghai citizen complained, "The urban environment is ruined. Wherever there's a large concentration of mobile population, shacks are erected at will, cooking is done outdoors, the streets are used as public toilets, structures are soiled, and public facilities are destroyed" (Solinger, 1999: 119~. Box 9.1 describes the extremely marginal conditions under which some migrants subsist in Shanghai. The living conditions of rural-to-urban migrants in some of China's largest cities will be a major public policy issue in the years to come. Yet in some respects, the Shanghai case is unique and very promising at least from the point of view of urban services and infrastructure improvement. But on the housing side, with severe backlogs and pressure for improved accommodation for lower-income groups, it is more typical of the situation common in many cities of the developing world. By the end of the 1990s, a huge gap still existed across the developing world between the effective supply of adequately serviced land and minimal standard housing on the one hand and the demand for accommodation as expressed through what poor families are able to pay on the other. Because land and minimal standard housing are, in general, not supplied quickly enough to keep pace with demand on the part of low-income groups, the backlog is expressed in dramatic fashion through the continuing existence of slums, squatter areas, and other forms of informal housing or impermanent structures. These areas, often totally bereft of services and infrastructure (at least in the early stages of their development), are growing in many cities and declining in few. Estimating the growth in the proportion of the urban population living in "self- help" housing in four major cities of Latin America, Gilbert (1998) shows that the proportion in Mexico City increased from 14 to 60 percent between 1952 and 1990; that in Lima, Peru, increased from 8 to 38 percent from 1956 to 1989; and that in Caracas, Venezuela, increased from 21 to 42 percent from 1961 to 1991. Only Bogota, Colombia, was able to decrease the proportion of the population living in such settlements in this case, from 40 to 26 percent from 1955 to 1991 (Gilbert, 1998: 82~. The Global Urban Indicators Database of the United Nations Centre for Hu- man Settlements (UNCHS), based on a major survey of 237 cities around the world (some 176 of which were listed as being in "developing" countries) reflects the same trends in housing quality. According to 1993 figures, among cities in Africa, on average only 49 percent of dwellings could be considered in compli- ance with local building regulations; for the Asia Pacific region (which in the
368 CITIES TRANSFORMED BOX 9.1 Living on the Edge in Shanghai A stone's throw from the glistening skyscrapers that tower over the eastern Chinese city of Shanghai, three-year-old Wang Kun lives with his parents in a hovel surrounded by festering garbage. A 15-strong tribe of rubbish collectors live in three rows of jerry-built shacks at the side of a rubbish dump in Shanghai's southeastern Nanhui one of scores of makeshift villages built on the outskirts of dumps around China's gleaming economic capital. Wang Kun and his father Wang Jun are from neighboring Jiangsu province while many of their fellow scavengers come from central China's impoverished Sichuan province. Their numbers are growing.... "Rubbish collectors do play an important role in cities and authorities have an ambiva- lent attitude towards them," said Sophia Woodman, research director of Hong Kong-based Human Rights in China. Crackdowns on garbage collectors can be triggered for arbitrary reasons ranging from National Day celebrations to a senior leader driving past a cluster of migrant dwellings and "thinking we don't want this in our gleaming modern city," she added. For 26-year-old Liu Tian, his wife and four-year-old son, who live on a rubbish tip in Sanlitang village in Shanghai's southern suburbs, the insecurity is bearable. The 800 yuan (97 dollars) Liu can rake in every month operating a smoke-belching machine that turns old plastic bags into rubber is four times more than he would earn in his Jiangsu village. "It's not a bad life. I earn much more than I would at home and when my son is old enough I want to send him back to school in Jiangsu," Liu said.... The Sanlitang rubbish dump is run by an entrepreneur who hails from Suxian, the same Jiangsu village as Liu Tian and his co-workers. Liu said his boss is doing well for himself but the authorities have asked him to relocate. "We have to move on in a few weeks, some of the other workers have already moved but we don't know where we're going yet. We have to wait for the boss to tell us," he explained.... However, for Wang Jun. the smell of the refuse and filth as well as the threat of being driven out by police is no deterrent. "Although work here is dirty and hard, I have more freedom and human dignity than in a factory job," he said. SOURCE: Agence France-Press (2001) survey does not include Korea, Singapore, or Japan), the figure is 59 percent; and for Latin America, the comparable figure is 74 percent. These figures show that access to adequate housing is still a distant goal for large numbers of people living in these three regions; they also show all too clearly that there is a strong rela- tionship between the per capita income of countries (and cities) and the level and quality of housing and urban services supplied to their people (UNCHS, 1999~. In general, writes Angel (2000: 320), . . . while the formal housing sector [has] produced the bulk of hous- ing for high-income groups, and while public housing [has been] of
THE CHALLENGE OF URBAN GOVERNANCE TABLE 9-2 Percentage of Urban Households Connected to Utility Services, by Region Region Water Sewerage Electricity Telephone Africa 37.6 12.7 42.4 11.6 Asia (Pacific) 63.2 38.5 86.1 26.0 Latin America and the 76.8 62.5 91.6 41.2 Caribbean Industrialized 99.4 97.8 99.4 89.1 SOURCE: United Nations Centre for Human Settlements (1999~. limited scope everywhere except in Hong Kong and Singapore- most low-income housing in the developing market economies was and is produced by the informal housing sector . . . occupying land il- legally (squatting) or built in land subdivisions that do not conform to zoning ordinances and planning regulations . . . without adequate infrastructure, without any state subsidies, and without any formal housing finance. 369 This picture of housing supply lagging badly behind demand runs parallel to a variety of other statistical measures of urban service delivery. Based on the same UNCHS database of 237 selected cities, Table 9-2 reveals a major differ- ence between developing and industrialized regions. In terms of water, sewerage, electricity, and telephone connections, the proportion of urban households receiv- ing urban services is much lower in developing regions. By region, however, there is a clear gradation for each of these services from a low in Africa, through Asia and Latin America, to a high in the industrialized world. The poorest countries (in this case those located in Africa) claim the lowest level of urban services, re- gardless of which service is being measured. A similar gradation of urban service levels from a low in the poorest cities of Africa to a high in the industrialized world obtains for virtually all other major urban services not included in this table, including waste disposal, expenditures on roads per person, education, and health services (UNCHS, 1999~. In the specific case of health statistics, the UNCHS database shows that African cities lag considerably behind cities in all other regions. Thus in the 87 cities representing Africa in the database, there is an average of 954 persons per hos- pital bed, compared with 566 for Asia Pacific (which includes China), 288 for Latin America and the Caribbean, and 132 for the industrialized world. Child mortality (under age 5) is also high in African cities, at a rate of more than twice that in the Asia Pacific region (UNCHS, 1999: socioeconomic develop- ment tables), and 30 times greater than that reported for the industrialized world. The low quality of health services in African cities is closely related to poor sanitation. The introduction to a major recent comparative study of waste
370 CITIES TRANSFORMED management in African cities (Onibokun, 1999: 2-3) makes the following ob- servation: The rapid rate of uncontrolled and unplanned urbanization in the de- veloping nations of Africa has brought environmental degradation. Indeed, one of the most pressing concerns of urbanization in the de- veloping world, especially in Africa, has been the problem of solid-, liquid-, and toxic-waste management. Recent events in major urban centres in Africa have shown that the problem of waste management has become a monster that has aborted most efforts made by city au- thorities, state and federal governments, and professionals alike. A visit to any African city today will reveal ... heaps of uncontrolled garbage, roadsides littered with refuse, streams blocked with junk, disposal sites constituting a health hazard to residential areas, and in- appropriately disposed toxic wastes. As many commentators have observed, the poorest cities in the world (which are also among the most rapidly growing) have acute capacity problems when it comes to servicing their populations. Not only is there generally a severe short- age (verging on absence) of trained, professional staff to deal with complex urban management problems, but the overall resource levels of poor municipalities are extremely low. Since water and sewer reticulation systems, to say nothing of elec- trical and telephone networks, cost no less in developing than in industrialized countries, substantial financial resources must be found to finance major infra- structural investments. But with per capita municipal revenue figures at such lev- els as US$13.20 in Nairobi, $2.60 in Lagos, $17.10 in Delhi, $27.70 in Dhaka, and $31 in Abidjan, only the most elementary municipal activities and local services can be supported. Latin American revenue figures are somewhat higher again, depending on the wealth of the country in question but in rapidly urbanizing countries such as Bolivia (where the revenue per capita in La Paz is US$108) or Guatemala (with Guatemala City having a per capita revenue of $26) (UNCHS, 1999), services and infrastructure cannot even come close to keeping pace with population growth. When services (such as water or electricity) are either par- tially or fully privatized, the new owners have difficulty raising rates to finance new infrastructural investment. Cities in the industrialized world (such as Toronto, with municipal revenue per capita of US$2,087; New York, with $5,829; or Amsterdam, with $4,559) have a much larger pool of local resources from which to finance needed infrastructure. While the individual returns may be somewhat unreliable, the UNCHS survey indicates that in 1993, the average per capita revenue received by municipal gov- ernments in Africa was US$15.20, in Asia (Pacific) $248.60, in Latin America and the Caribbean $252.20, and in the industrialized world $2,763.30. The ratio between the lowest and highest regions is on the order of 1: 182. This ratio is much higher than that between per capita income in sub-Saharan Africa and that in the
THE CHALLENGE OF URBAN GOVERNANCE 371 "high-income" countries (1:51) listed in the World Development Report (World Bank, 2000c: 275~. As discussed in the following section, one important means of dealing with local service capacity problems is to devolve more power to local authorities. While there have been problems in many countries with undertaking Revolution before the necessary financial resources are available, local governments are of- ten in closer touch than more central authorities with community and other civil society groups that can support service initiatives. In the Philippines, for example, the Local Government Code of 1991 trans- ferred responsibility for overseeing the implementation of primary health care to local government units. While the Ministry of Health apparently resisted devo- lution at first, the decentralization of many health services in fact took place on January 1, 1993, at which time 595 hospitals, 12,859 rural and urban health units and health centers, and most health programs were (at least formally) transferred from the national to the local level. Significant financial challenges resulted at the local level since revenues were allocated by the central government on the basis of demographic and poverty measures, not the existing distribution of health facilities and programs (Stover, 1999~. Here, civil society helped buffer the shock. Since the central Ministry of Health had expended considerable effort on developing relations with nongovernmental organizations (NGOs), volunteer health workers, and other community organizations at the local level before the advent of man- dated decentralization, local governments were able to find partners to carry out their primary health mandate, and according to some, the transition in many areas took place relatively effectively (Bautista,1998~. The Financial Resources Dimension While it is impossible to separate the capacity (or incapacity) of cities to deal with the service needs of their population from the financial resources they have at their disposal to do so, it is also the case that the level of resources available is not immutable. While wealthy countries can direct much higher levels of taxes and fees for service to their municipalities and/or private service providers than can poorer countries, the elasticity of supply of municipal revenue in relation to per capita income is clearly greater than zero. One of the reasons for this is that cities and their surrounding regions are often the most dynamic economic units in the country; the more effectively municipalities can build services and infra- structure to facilitate productive economic activity, the more both the country and individual urban citizens will benefit. The logic behind this relationship was at least one of the factors underlying the major decentralization reforms of the 1980s and 1990s throughout the developing world. There is some discussion among researchers as to the fundamental factors behind decentralization in different set- tings and different countries, as we have seen above, but the balance of opin- ion on these initiatives appears to be positive. In any case, as a result of these
372 CITIES TRANSFORMED decentralization reforms, local governments (and cities in particular) have been given substantially more power by central governments in many countries around the world. As Manor (1999: 1) points out: Decentralization has quietly become a fashion of our time. It is be- ing considered or attempted in an astonishing diversity of developing and transitional countries ... by solvent and insolvent regimes, by democracies (both mature and emergent) and autocracies, by regimes making the transition to democracy and by others seeking to avoid that transition, by regimes with various colonial inheritances and by those with none. It is being attempted where civil society is strong, and where it is weak. It appeals to people of the left, the center and the right, and to groups which disagree with each other on a number of other issues. The nature of- these decentralization policies varies tremendously from in- cremental changes in protocols for intergovernmental relations on the one hand to major constitutional amendments or even new constitutional dispensations on the other. Three large countries gave new constitutional powers to municipalities during the period. In Brazil, a new constitution in 1988 considerably increased the power of mu- nicipalities in relation to the states, assigning to them control of intracity transport, preschool and elementary education, land use, preventive health care, and histor- ical and cultural preservation. On the participatory side, municipalities in Brazil were given the right to establish councils of stakeholders (termed in English "mu- nicipal boards" or "community councils"~. These bodies, established in most of the largest cities in the country, include unelected representatives of community groups and deal with such important matters as urban development, education, the environment, health, and sanitation. In India, an important constitutional amendment in 1992 provided an illustrative list of functions that were henceforth to be considered appropriate for municipal government; among these functions were planning for economic and social development, alleviation of urban poverty, and even urban forestry. The amendment also limited the degree to which state governments are able to suspend democratic local government (a practice that, until then, had frozen democratic lo- cal governments in nearly half of the largest cities in the country), provided for a revision of state-local fiscal relations, and required that no less than 33 percent of all elected local councilors be women. The new South African constitution of 1996 devotes a whole chapter (con- taining 14 separate articles) to local government. Among other things, this chap- ter (Section 152) states that the objectives of local government (including mu- nicipal government) are "(a) to provide democratic and accountable government for local communities; (b) to ensure the provision of services to communities in a sustainable manner; (c) to promote social and economic development; (d) to
THE CHALLENGE OF URBAN GOVERNANCE 373 promote a safe and healthy environment; and (e) to encourage the involvement of communities and community organizations in the matters of local government." There are two clear messages in these reforms. The first is that municipalities (and other local governments) are now expected to undertake and to finance a much broader range of services and other economic and social activities. The second is that community and important local stakeholder groups must be engaged in the local governance process. In spite of this general assignment of greater powers and responsibilities to municipalities in many countries, revenue has not kept pace with expenditure re- quirements. Not only are most local authorities dependent for up to one-third of their revenue on other levels of government, but their own resources are inade- quate. One important reason for this inadequacy of own-source revenue is that it is usually based on fees for service and property taxes, rather than on more lu- crative and collectible taxes (such as gasoline and income taxes) (UNCHS, 1996: 178-81~. In many countries, even mandated central government transfers are not reliable. In poor countries, collecting fees and property taxes is notoriously diffi- cult. Most municipalities in developing countries cannot borrow, and they cannot run a deficit. In addition, local revenue is highly dependent on both macroeco- nomic factors (for example, whether national economies are prospering or in dis- tress) and on the level at which municipalities are permitted to borrow through the market. Under recent conditions of crisis in such countries as Argentina, Brazil, and Indonesia, local authorities have been negatively affected even though they have been able to borrow and their potential sources of local revenue have been augmented by improving the tax base. Comparative data on local government revenues and expenditures are partic- ularly difficult to obtain. Most countries do not supply consistent and reliable revenue and expenditure data. Even the Government Finance Statistics Yearbook published by the International Monetary Fund, the primary and most authoritative source of international data on government finance, gives information on local government finance for only 52 of its 114 listed countries in its 1998 edition, with time-series, detailed data being provided for only 12 developing countries (International Monetary Fund, 1998~. Revenue and expenditure figures reported in the UNCHS database, referred to above, appear questionable in the case of some developing-country cities. Havana, Cuba, for example, reports revenues of US$1,418 per capita and capital expenditures per capita of $1,318, while Toronto reports, respectively, $2,087 and $253, and New York reports $5,829 and $582 (UNCHS, 1999~. For the most part, as argued above, the level of revenues and expenditures for all cities in poor countries is very low. Given the increase in Revolution of powers and responsibilities that has char- acterized the last decade, figures on municipal expenditures are slowly rising in comparison with national government expenditures, but the levels are still low. Figures published in the 1996 Global Report on Human Settlements covering a sample of 18 wealthy countries; 4 transition countries; and 16 countries in Asia,
374 CITIES TRANSFORMED Africa, and Latin America show that on average, local government expenditures are 22 percent of total government expenditures in the first group, 20 percent in the second group, and only 9 percent in the third group (UNCHS, 1996: 174~. Unfortunately, many of these figures are over a decade old. Under Bolivia's Ley de Participacio'n Popular (the implementation of which began in 1994), municipal governments automatically receive 20 percent of all central government revenues through fiscal transfer. And starting in the early l990s, Colombia began to trans- fer an increasing fixed proportion of national recurrent income to municipalities- from 14 percent in 1993 to 22 percent in 2002 (Hoskin, 1998: 105~. Looking at municipal spending as a proportion of total public-sector spending between the 1970s and the early l990s for Colombia, Argentina, Chile, and Peru, one finds in- creases of, respectively, 10.5 to 15.7 percept, 5.4 to 8.6 percept, 4.7 to 12.7 percent, and 2.2 to 9.2 percent (UNCHS, 1996: 167~. There is good reason to believe as the Colombian example of fiscal transfer already indicates that these proportions will rise further in the early years of the twenty-first century. As a result of changes in intergovernmental relations, the financial health of municipalities is slowly improving at least in some countries. A compilation of data from case studies carried out in the mid-199Os for Brazil, Bolivia, Colombia, Chile, and Peru shows that transfers from other levels of government to the local level are now relatively high, representing 62.7 percent of total local revenue in Brazil, 54.7 percent in Bolivia, 47.8 percent in Colombia, 42.7 percent in Chile, and 58.1 percent in Peru (Aghon and Casas, 1999: 77~. At the same time, property tax reforms in some countries have resulted in a much higher level of recovery. In La Paz, for example, after the Ley de Participacio'n Popular (1994) devolved property taxes from the central government to municipal jurisdictions, property tax recovery increased considerably from US$4.6 million in 1993 to $9.3 mil- lion in 1996. In Bolivia, the new law also created "vigilance committees" at the central municipal level, composed of local citizens who were to monitor the use of resources by elected mayors and councils, could propose investment projects, and in some cases could censure and dismiss the mayor (Grindle, 2000: 95-7~. The creation of these groups contributed to a heightened sense of local "ownership" in municipal government, and probably to the higher rate of tax recovery as well. And in Bogota, Colombia, recovery rose from approximately 9 million pesos in 1993 to 30 million pesos in 1996 once a system of self-evaluation of property tax had been instituted (Aghon and Casas, 1999: 78-9~. That local tax collection and revenue generation may be as much a governance as a technical issue (requiring, for example, changes in tax codes, property attri- butions, and accounting systems) is illustrated by examples in China and Brazil. In the case of China, market-oriented reforms since the late 1970s have resulted in the Revolution of important decision-making powers from the center to localities (Wang, 1994), as well as in changes in the fiscal relations between the two levels of government. While there have been many twists and turns in the redefinition of the fiscal relationship between levels of government, a decentralized market
THE CHALLENGE OF URBAN GOVERNANCE 375 environment has clearly emerged. Local governments now enjoy considerable "off-budget" revenues, generated from such sources as donations by individuals and enterprises to specific public projects, profits from township-owned enter- prises, fees for services and fines, and revenue from the leasing of public land to enterprises and developers (Gang, 1999~. Indeed, the sale of land use rights has constituted the largest source of income for many cities in the coastal region since the late 1980s (Yeung,1998b: 177-78,271-74~. In 1992, the last year in which off-budget revenues were officially counted as fiscal revenue, such revenues had reached 93 percent of total formal budgetary rev- enues at the local level, and they are presumably still rising, at least in the aggre- gate. Although observers remark on the difficulty of estimating the contribution of off-budget revenues to large government structures (such as city government), that contribution is undoubtedly considerable. By 1996, such revenues accounted for close to 40 percent of all funding for urban infrastructure (including local pub- lic utility and construction charges, transfers from the central government, and regular municipal budgetary allocations) (Wu, l999b: 2271-72~. While there is controversy over the legitimacy of the way these revenues are allocated (see be- low), local governments use informal revenue generation systems with the in- volvement of their citizens to generate necessary resources for the provision of public goods. Gang (1999: 234-5) remarks on the close connection between local participation and effective mobilization of local tax resources: A real challenge ... is to develop a new type of local governance. Case studies in various regions show that in most places where off- budget revenue makes up a large share of local public revenues, some kind of democratic mechanism has emerged spontaneously. For ex- ample, people have started to use the local People's Assembly as the mechanism for decision making and monitoring of local public fi- nance. Qingde town of Zhejiang province established three years ago an Assembly subcommittee, a "Financial Committee," as the de- cision making body for local public finance. Almost 30 percent of the Assembly delegates joined the committee, whose main task is to discuss and decide whether public projects should be initiated, how they should be financed, how the resources should be mobi- lized and how the revenues (mainly off-budget revenues) should be distributed and used.... In Shahe town of Shantou city in Guang- dong province, the local people set up three years ago a special local "Board of Directors for Public Projects." It functions as the decision making (legislative) and administrative body for local public projects, mainly physical projects for infrastructure, including roads, schools and hospitals. The board consists of most of the "elite" of the town, elected by the local people, but includes no current local government officials. It not only makes decisions on projects and fund-raising but
376 CITIES TRANSFORMED also takes responsibility for project implementation and fund expen- ditures. In some sense, the board functions like a "parallel govern- ment" in charge of special fields of local public affairs. Meanwhile, the local administration plays the role of an outside monitor: it still carries out other government duties and at the same time monitors the work of the board and reinforces the formal government policies which need to be taken into consideration. In both cases described ... we can see the development of demo- cratic mechanisms and a new type of local governance. Such new mechanisms minimise the problems that have arisen with off-budget revenues. As it develops spontaneously, based on solid economic motivations, this local democracy holds promise, with an obviously far-reaching impact on local governance. While these new participatory mechanisms undoubtedly open some local eco- nomic systems to participation from the community, there is another side to this process. On the one hand, as researchers have pointed out, local governments and officials in China have been developing large collective enterprises, whose income (described above as "off-budget") can be used in a constructive fashion for administration, infrastructure improvement, and the establishment of new en- terprises that can expand the revenue base of local governments (Oi, 1992,1995; Walder, 1994~. On the other hand, as Sargeson and Zhang (1999) point out in a case study of Xihu, a suburban district of Hangzhou, the provincial capital of Zhe- jiang, some local officials have taken advantage of these new enterprises to enrich themselves. They have done so by translating their political influence into stock shares, preventing a more balanced distribution of enterprise benefits throughout the local area. The costs and benefits of this decentralized fiscal system in China clearly warrant further evaluation. Another and much better-known case is the development of "participatory budgeting" in Brazil. While versions of this system have been operating through- out Brazil (see, for example, Singer, 1996), the most well-known example of a city practicing the participatory budget system is Porto Alegre, a city of about 1.3 million in the south of the country. The system which essentially involves ordi- nary citizens in planning for the yearly capital budget is based on the work of 16 forums based on local regions of the city. In addition, there are 5 thematic forums (created in 1994), addressing education, health and social services, transportation, city organization, and economic development. There is also a municipal budget council with representatives from the regional and thematic forums. The system was originated in 1989 by the Union of Neighborhood Associa- tions. By 1995, some 7,000 people were participating in the regional assemblies and 14,000 more in further meetings to negotiate compromises among regions' conflicting demands. The system is complex and continues virtually throughout the year. The regional forums even micromanage the actual implementation of
THE CHALLENGE OF URBAN GOVERNANCE 377 capital projects (Abers,2001~. According to the municipality, more than 70 cities elsewhere in Brazil and throughout the world (including Buenos Aires, Barcelona, and Saint Denis) have adapted this system to their own needs (Porto Alegre, 1998: 10~. A study of the city management process in Porto Alegre (Pozzobon, 1998) reveals that even from 1992 to 1995, the city increased its total tax receipts by 34 percent. The former mayor of the city claims the popularity of the budget- ing system contributed to a tripling of the city's tax revenues during the period 1989-1999 (Pont, 2001~. Abers (2001: 140), examining the participatory budget process, emphasizes the widespread democratic learning that was entailed: Along the way, [citizen] participants have developed a series of demo- cratic skills. The most elementary are the basic habits of collective decisionmaking holding coherent meetings, allowing all to speak, and learning how to debate and vote on complex issues where choices are multiple. Participants have also gained critical skills in negotiat- ing with the administration. They pressure agencies to produce infor- mation about government actions and to demystify technical rules. They often successfully force administration officials to talk in or- dinary people's terms and, in doing so, unmask attempts to veil in technical complexity the real reasons for rejecting or changing the demands prioritized. Abers also argues that participatory budgeting would not have been possible with- out the active, time-consuming, and persistent efforts of local government officials to work with neighborhood groups. Pont (2001), the mayor of the city during the period 1996-2000, implies that the participatory system in Porto Alegre was so popular that it resulted in the mayor's party (the PT or "Worker's Party") winning the state governorship in 1996: In the last election [19961 we really guaranteed Olivio Dutra's [the governor's] victory with a huge landslide in Porto Alegre. Rio Grande do Sul has a population of almost 10 million; there are about 7 mil- lion voters, and the victory was by 80,000 votes. It was a very tight contest overall, but in the capital we won with a very large margin. So we can say that it was the voters in the capital [Porto Alegre] who guaranteed the victory (Pont, 2001: 146~. While Porto Alegre is the best known of the Brazilian cities practicing par- ticipatory budgeting, it is not the only or even the first city to do so. In Piraci- caba, a municipality in the state of Sao Paulo, a Citizens Budgetary Committee- in which popular organizations had seats and votes was established as early as 1980. The initiative died when a new mayor was elected, but other participatory
378 CITIES TRANSFORMED initiatives were undertaken in such cities as Lages, Fortaleza, and Recife in the 1980s (Souza, 2001~. At the time, these cities were not administered by the PT, but by other leftist parties or coalitions. Sao Paulo had a less-than-successful ex- perience with participatory budgeting under mayor Luisa Erundina during 1989- 1992 (Singer, 1996~. Since the election in 2000 of another PT administration under Mayor Marta Suplicy, the participatory budgeting process has been revived. On the more positive side, Belo Horizonte, a relatively affluent city in Minas Gerais State, has organized "priorities caravans" consisting of budget delegates making local bus visits to check directly on the problems identified at subregional meetings. Overall, Belo Horizonte has indicated on its Web site that some 200,000 people have already taken part in participatory budgeting (Souza, 2001: 15~. In Recife, a much poorer city in the northeast of Brazil, participatory budgeting be- gan with the election of Mayor Jarbas Vasconcelos in 1985 and has continued to the present, including a recent period when the city was administered under a coalition government. Nevertheless, in Recife as in other cities, a relatively small proportion of the total budget is under discussion in the participatory bud- geting process (Melo, Rezende, and Lubambo, 2000~. The needs of the poor are a long way from being solved by this process. As a knowledgeable Brazilian writer claims with reference to the process across the country, "although some of the claims and results [of participatory budgeting! deserve more research, the experience does allow low-income segments of neglected areas ... to decide on investment priorities in their communities. Furthermore, it also argues that en- couraging participation in highly unequal societies like Brazil should be valued more for heightening citizenship rather than for the material gains it may bring to some areas of the cities" (Souza, 2001: 1~. The Diversity Dimension As they grow, many cities become increasingly diverse with respect to both the cultural, ethnic, and even religious characteristics of their populations and the na- ture of their economies. The extraordinary internal diversity of cities has always been a source of stress and conflict, but also of creativity and originality. Hall (1998: 7) suggests that, contrary to gloomy predictions of decline, many of the largest cities of the Western world have also served as platforms for the highest levels of innovation. While "no one kind of city, nor any one size of city, has a monopoly on creativity or the good life . . . the biggest and most cosmopolitan cities, for all their evident disadvantages and obvious problems, have throughout history been the places that ignited the sacred flame of the human intelligence and the human imagination." It can be argued that one of the most important measures of a successful city in both the developed and developing worlds is the ability to deal effectively with social diversity and to create structures of governance that integrate diverse communities and economic interests into a functioning system (Polese and Stren, 2000~.
THE CHALLENGE OF URBAN GOVERNANCE 379 Studies of the morphology of cities in developing countries almost always comment on the fragmentation of the population, from both a social and infra- structural point of view. In his classic article on urban planning in developing countries, Balbo (1993: 24-5) points out that, although Western industrial cities are characterized by a certain coherence that is amenable to master planning, almost all cities in developing countries are more complex, both spatially and socially: the city of the Third World is a city of fragments, where ur- banisation takes place in leaps and bounds, creating a continuously discontinuous pattern. In the fragmented city, physical environment, services, income, cultural values and institutional systems can vary markedly from neighbourhood to neighbourhood, often from street to street. An aerial view of the city shows a spatial structure made up of many different pieces drawn together in a rather accidental way. There are more of some kinds than others. Those in the periphery are incomplete and more "fragile," while older areas are well established with clearly defined boundaries. Even in Latin America, where most cities date from the sixteenth century, a process of "tribalisation" seems to be under way: the city is splitting into different separated parts, with the apparent forma- tion of many "microstates." Wealthy neighbourhoods provided with all kinds of services, such as exclusive schools, golf courses, tennis courts and private police patrolling the area around the clock inter- twine with illegal settlements where water is available only at pub- lic fountains, no sanitation system exists, electricity is pirated by a privileged few, the roads become mud streams whenever it rains, and where house-sharing is the norm. Each fragment appears to live and function autonomously, sticking firmly to what it has been able to grab in the daily fight for survival. This depiction of the situation in the early l990s has been reinforced by ev- idence of the further fragmentation of urban and interurban networks in both developed and developing cities. These networks involve such essential urban services as transport, water and sanitation, electrical supply, communications, and even security. As Graham and Marvin (2001:383) argue, the urban world of the last decade has seen a "splintering urbanism" in which "standardized public or private infrastructure monopolies ... laid out to offer broadly similar services at relatively equal user charges over cities and regions, are receding as hege- monic forms of infrastructure management." While a major factor behind these changes in developing-country cities is the inability of service providers to keep pace with demographic growth, more general factors behind these changes include "the widespread retreat of the idea that networked services are 'public' services that should be available to all at standard tariffs" (Graham and Marvin, 2001: 96)
380 CITIES TRANSFORMED and the pervasiveness of the idea that all forms of ownership (public and private) should compete for the supply of local services. As a result of many of these factors, knowledgeable commentators were still observing social segregation in Latin American cities toward the end of the l990s. Thus, Gilbert (1996: 91-3) observes that "Latin American cities remain highly segregated.... In Lima, everyone knows that San Isidro and Miraflores are rich while Comas and Villa E1 Salvador are poor; in Santiago, the extremes are found in affluent Providencia and Vitacure in the north-east and poor La Pintana and La Granja in the south. In Rio de Janeiro, the rich live in Leblon and Ipanema and the poor in the Baixada Fluminense...." While residential segregation is becoming more complex, he argues, there is no sign that it is declining; "indeed, in some respects there is greater polarization." The complex nature of this pattern can be seen in a number of cities, as described below. Rio de.laneiro Rio de Janeiro represents a story of "emerging dualization in a historically unequal city" (Ribeiro and Telles, 2000~. The Rio metropolitan area, with a population of some 9.7 million spread over 17 municipalities, grew at a rate of only 1.1 percent from 1980 to 1991. As many of the city's population migrated to medium-sized cities in the State of Rio de Janeiro, the outer parts of the city region grew much more quickly than the central portions. The authors' observations, based on cen- sus data, show that the poor, the less educated, and nonwhites are concentrated in the periphery and the favelas (spread throughout the metropolitan area), while the white, educated population is concentrated in the city center. During the period 1980-1991, income concentration increased the poor be- came more concentrated in the periphery and the indigent poor in the central parts of the city, while the rich increased their share of total income. At the same time, as salaried employment for the poor was becoming more scarce, and many at the lower end of the income scale were moving to the city center to take advantage of opportunities to work in the personal services sector, large real estate companies "intensified central city polarization, transforming the South Zone into an exclu- sive citadel, fortified by Rio's steep hills and an ominous public and private police presence. The termination of the housing finance system in the 1970s, the reduced income of workers, and upper middle class desires to live in exclusive zones have made elite housing by far the most viable sector of the real estate market" (Ribeiro and Telles, 2000: 86-7~. As these tendencies toward a duality among the population according to socio- economic status and location evolved, a parallel division emerged in the associat- ions maintained by the two groups. Thus, those who have little education and are poor are much more likely to become involved in neighborhood community and religious associations than in unions, professional associations, and political par- ties. Those with high income and education, by contrast, join unions, professional
THE CHALLENGE OF URBAN GOVERNANCE 381 associations, and cultural and sports associations and play a more active role in formal political parties. The periphery and the favelas are also characterized by what the authors call "the expansion of criminal and perverse forms of sociability" (Ribeiro and Telles, 2000: 93~. These social patterns make coherent governance much more difficult to achieve. Sao Paulo Sao Paulo is another well-documented example of social fragmentation. The mu- nicipality of Sao Paulo, with some 10 million people, is one of the world's largest cities. Greater Sao Paulo, which includes an additional 38 municipalities in a continuous urban area of 1,500 square kilometers, contains 17.8 million people (United Nations, 2001: 84) and is the second-largest metropolitan area in Latin America (after Mexico City). As was the case in Rio de Janeiro, the growth rate in the City of Sao Paulo and the greater Sao Paulo region slowed during the 1980s, to the benefit of smaller municipalities elsewhere in the State of Sao Paulo as man- ufacturing firms relocated from the center to the periphery. Total employment grew more rapidly outside the metropolitan region as well, with the exception of the financial services and communications sectors, which grew more rapidly in the City of Sao Paulo (Santos, 1996: 229~. During the 1980s and l990s, pock- ets of poverty and rundown housing developments bedeviled the generally well- equipped central areas of the city; at the same time, more and more of the housing in the peripheral areas of the city was classified as "precarious," with low-quality materials, often on marginal land, with little or no infrastructure, and with few social services. Sposati (1996), a professor of social work and currently an elected municipal councillor, developed a series of maps of social inclusion/exclusion in the mid-199Os. In the introduction to her collection of maps (which depict the presence or absence of various facilities and services throughout the city), she makes the following observation (Sposati, 1996: 5~: If the city of Sao Paulo had eyes and could see itself in a large mirror, it would see the broken and shocking view of its inequalities. How- ever, its two eyes would not be similar either. As in the popular song, maybe while one of them was staring, the other one would be floating. Because everything here seems unequal. The largest city in Latin America, with circa 10 million inhabitants, is famous for the pres- ence of the most advanced forms of technological development and the turnover of finance capital. It houses gardened neighborhoods, lavish mansions, dauntless buildings, excellent teaching centers and first class hospitals. A numerous fleet of luxury and imported cars cir- culates through its avenues or stops in traffic jams. Nevertheless, Sao Paulo also lives with the most severe forms of deprivation and human suffering. A deprived and unemployed population seeks refuge in slums and shack houses, or is abandoned on the streets. They are
382 CITIES TRANSFORMED daily victims of violence and do not have access to their rights and justice. They hang on overcrowded buses and trains, and if they get ill, the health service is poor. Their children, when they manage to, attend deteriorated schools, and evasion takes place very soon.... To some writers, this pattern is systemic. Rolnick (1999), a senior planner, argues that the coincidence of precarious housing, low income, and low formal employment that characterizes many peripheral zones of the city actually consti- tutes a pattern of "territorial exclusion," whereby those groups consigned to live in these areas cannot fully realize the benefits of citizenship and economic growth that are available to people living in other areas of the city. She shows that there is a marked relationship between levels of violence and homicide in the city and levels of basic services available. Using 1991 census statistics showing housing conditions, location, infrastructure availability, and number of rooms per house- hold, she constructs a measure of territorial exclusion for 118 cities within the state, showing an inverse relationship between the ranking on territorial exclusion and homicide rates in both 1991 and 1994. "Territorial exclusion," she explains, "makes daily life insecure and risky. It blocks access to jobs and educational and cultural opportunities, which are concentrated in small and protected enclaves within cities. Since most residences in excluded areas are illegal and mixed use is generally forbidden by municipal land regulations, people are denied the possibil- ity of using assets, such as home ownership, to generate money and create jobs.... [In these excluded areas] living in a permanent condition of denial of basic hu- man environmental needs makes inhabitants feel as if their lives are worthless" (Rolnick, 1999: 17~. In particular, the peripheral areas of the city of Sao Paulo exhibit many of the preconditions for violence, in that resentment and anger build up when in- frastructure and services are not easily available. As Cardia (2000) describes the situation, violence in the metropolitan area is significantly greater in the periph- eral areas, where the population growth rate is the highest. Because of inadequate infrastructure, these populations are cut off from services and public authorities in a variety of ways (Cardia, 2000~. While residential segregation and unequal in- frastructural development do not in themselves produce violence and crime, these writers clearly believe that such conditions raise the likelihood that a culture of violence will be reinforced. Another approach to describing this pattern of increasing fragmentation of so- cial groups and neighborhoods is offered by Caldeira (1996: 63), who discusses the development of social and spatial segregation in Sao Paulo from the early twentieth century to the present. During this period, Sao Paulo went from being a "concentrated city" (from the early part of the century to the 1930s); to a city with "a rich centre and a poor periphery" (from the 1940s to the 1980s); to its present form in the l990s (and presumably later), which she describes as "proximity and high walls." Her study of a rapidly growing wealthy suburban neighborhood
THE CHALLENGE OF URBAN GOVERNANCE 383 (Morumbi) shows that fear of violence and crime is a pervasive subject in every- day conversation. However, talk of crime is common throughout all social groups in the city, as "people from all social classes fortify their homes, change their habits, and end up transforming the city and its public areas." The result is a new urban landscape made up essentially of "fortified fragments" from which the poor and marginalized are physically excluded (Caldeira, 1996: 64-5~: As the spaces for the rich face inwards, the outside space is left for those who cannot afford to go in. In fact, the public is treated as leftover both by the design of the enclaves and by the citizens who create the new private order. The modern public space of the streets is increasingly the area abandoned to the homeless and the street chil- dren . . . fragmentation enforces separation and expresses not simple differences but irreconcilable inequalities. Newspaper advertisements appearing between 1985 and 1995 illustrate this trend. These advertisements aimed at the rich both reflect and promote an image of security, isolation, racial and class homogeneity, and high levels of services and facilities (Caldeira, 1999: 120~. Manila A final example of the tendency toward social and institutional fragmentation is the case of Manila, the capital of the Philippines. Metro Manila (or the National Capital Region [NCRj), with a population in 2000 of 10.87 million (United Na- tions,2001: 89), is by far the largest city in the country, almost 10 times larger than the next-largest city, Cebu City. Metro Manila incorporates 17 municipalities (or local government units) occupying a territory of 636 square kilometers. Accord- ing to Laquian (2000: 3), "the urban field of Metro Manila rightfully includes 18 adj acent [local government units] that add another 4 million people resident within a territory of 1,681 square kilometers within the so-called Calabarzon region ... as well as the provinces of Bulacan, Zambales and Pampanga in Central Luzon." In this mega-urban region, the outlying towns have been growing more quickly than the national capital region since the 1960-1970 intercensal period. Some of these towns registered an annual growth rate of over 6 percent during the period 1980-1990 (Ocampo, 1995: 285~. In the process of growth, the net movement of migrants within the region has been outward rather than inward, creating major problems of sprawl and ribbon development. Sprawl has put an extreme strain on public utilities and services, extended travel distances for workers living out- side the NCR, and "resulted in social disparities and conflicts of various kinds" (Ocampo, 1995: 287~. One knowledgeable writer has described Metro Manila as a "city of villages" in which "autonomous local units ... have traditionally resisted efforts to cen- trally control their activities" (Laquian, 2000: 1~. Local resistance to higher-level
384 CITIES TRANSFORMED controls is a result of three elements: a relatively entrenched localism that de- fines neighborhoods as distinct units within the wider metropolitan area; strong local government units, reinforced by the 1991 Local Government Code; and a strong political culture of particularism that makes coalitions difficult to achieve (Laquian,2000~. Fragmentation at the local level starts with the division of the city into four major types of neighborhoods. First are the barrios of the urban poor, often on marginal lands and even huge garbage dumps. Second are the "villages" of the rich, in which the residents lead protected, isolated lives with a full range of public services available for their use. Third are the traditional neighborhoods that house the majority of the population, bound closely by the observance of cultural and re- ligious festivals, and participating in groups such as parent-teacher and neighbor- hood watch associations. Fourth are the traditional villages on the city's periphery, although these semirural areas are becoming pervaded by pollution and industrial enterprises, and their young are leaving for the city (Laquian, 2000: 1-2~. The fragmentation of neighborhood cultures is reinforced, one author argues, by the control retained by powerful local families in many Manila cities and mu- nicipalities. It is further exacerbated by the competitive behavior of certain local units (such as Makati, Quezon City, and Mandaluyong), which vie for large infras- tructure investments and buildings that can serve the global economy. One result of this differentiation is that there is a wide disparity in the revenue collection of Metro Manila local governments. In 1997, revenue per capita ranged from 486 pesos in Malabon to 7,656 pesos in Makati, the upscale commercial and financial center (Shatkin, 2000: 2368~. Coordination and planning at the metro level are weak. While there is an upper-tier authority with responsibility for the whole metropolitan area, its pow- ers and even its budget are limited, especially relative to the local government au- thorities, whose powers were increased under the 1987 Constitution and the 1991 Local Government Code. The power of the metropolitan authority was not en- hanced by the fact that Imelda Marcos had been appointed governor at the outset of the authority's life, presiding over the Metro Manila Commission until President Ferdinand Marcos was toppled in 1986. The agency, renamed the Metro Manila Authority in 1990, was assigned a rotating chairman, chosen every 6 months from among the 17 city and municipal mayors of the NCR. In 1995, the Metro Manila Authority became the Metro Manila Development Authority, although it has "re- mainedaweakgovernancestructure" (Laquian, 2000: 8~. While NGOs end civil society groups are highly active in Manila as well as throughout the Philippines, these groups have not been able to overcome the innate localism and particularism that obtains at the metropolitan level (Laquian, 2000: 23~. The Security Dimension Urban security became one of the dominant issues of the 1990s in many devel- oping countries. Since September 11, 2001, an international dimension has been
THE CHALLENGE OF URBAN GOVERNANCE 385 added to the domestic sources of insecurity generated by some of the urban so- cial conditions considered above. We have chosen to discuss this important sub- ject here since it is closely related to issues of social fragmentation and social policy, and because we wish to explore the governance implications of different approaches to dealing with the problem. There is no question from both macro- and micro-level studies that urban vio- lence has increased in a great many parts of the developing world. At the global level, crime and violence are generally much higher in Africa and Latin America than in Asia. A dataset (United Nations Interregional Crime and Justice Research Institute, 1998) covering 18 developing-country cities2 shows that Asia consis- tently ranks the lowest for all types of crime, while Africa and Latin America share the highest ranking for all types of crime. When the "contact crime" rate (i.e., incidents involving violence) is analyzed separately, Latin American cities outstrip African and Asian cities with regard to sexual assault, with a rate of 5 percent of all types of crime (compared with 2.4 percent in the African cities and 1.6 percent in the Asian cities). For assault with force, African cities show a rate of 3.1 percent, compared with 2.7 percent in Latin America and 0.8 percent in Asia. For robbery, cities in Latin America have a rate of 8.1 percent, compared with 4.2 percent in Africa and 1.4 percent in Asia. Aside from the human and sociological effects of crime and violence, there is a significant economic cost to the countries in which rates of crime and vio- lence are high. A groundbreaking study by two Colombians (Londono and Guer- rero, 2000) estimates the cost of violence against people and property in terms of loss of potential urban gross domestic product (GDP). The authors claim that E1 Salvador, Colombia, Venezuela, Brazil, Peru, and Mexico lose, respec- tively, 24.9, 24.7, 11.8, 10.5, 5.1, and 12.3 percent of their urban GDP annually as a result of violence (Londono and Guerrero, 2000: 27~. Likewise, a study by the Mexican Health Foundation (Lozano, Hijar, Zurita, Hernandez, Avila, Bravo, de Jesus Ramirez, Carrillo, Ayala, and Lopez, 2000), based on statistics for Mex- ico City collected in the mid-199Os (when the city's violence rates were increas- ing), suggests that the annual cost of violence in the city (taking both direct and indirect costs into consideration) is a staggering $1.897 billion. Data on homicide in cities are often cited as the most striking indication that urbanization in the developing world produces a decline in social cohesion and an increase in conflict and insecurity. Table 9-3 summarizes some data on homicide rates for different cities in Latin America, indicating major differences among countries, as well as rather important variations within countries. Rates in Brazil 2This dataset, collected through the International Crime Victimization Survey, constitutes the most detailed global dataset on urban violence available. Crime was measured according to five categories: vehicle-related crime, break-and-enter crime, victimization experienced by the respondent personally (including robbery, theft of personal property, assault/threat, and sexual incidents), consumer fraud, and bribery/corruption. Data were collected in the largest city in each of the selected countries. A1- though not representative of the urban population in each country, this dataset provides a starting point for regional comparisons of urban crime (bearing in mind such issues as underreporting).
386 CITIES TRANSFORMED TABLE 9-3 Homicide Rates per 100,000 Inhabitants in Selected Latin American Cities City Country Year Homicide Rate Medell~n Colombia 1995 248.0 Call Colombia 1995 112.0 Diadema Brazil 1997 146.1 Belford Roxo Brazil 1997 76.5 Sao Paulo Brazil 1998 55.8 Rio de Janeiro Brazil 1998 52.8 Lima Peru 1995 25.0 Ciudad de Mexico Mexico 1995 19.6 Santiago Chile 1995 8.0 Buenos Aires Argentina 1998 6.4 Caracas Venezuela 1995 76.0 Guatemala City Guatemala 1996 101.5 San Salvador E1 Salvador 1995 95.4 SOURCE: Carneiro (20004. and Colombia, for example, are very high by worldwide standards, while rates in cities such as Santiago and Buenos Aires are relatively low. Trend data from case studies show a general increase in levels of violence in cities of Latin America over the last two decades, although some cities have also experienced a decrease in recent years. Nevertheless, trends may not be consistent even within a country. In Brazil, for example, the critical period for increases in violence was the second half of the 1980s, coinciding with the reestablishment of democracy. More recently, however, Rio de Janeiro has seen a decline in the rate of homicides, while this has not been the case in Sao Paulo. This finding suggests that urban violence has both local and more systemic explanations. How to respond to the increasing violence is a major challenge at a number of levels national, local, and social. Given the broad contours of a model that regards violence as a result of the erosion of physical, human, natural, and so- cial capital, it is possible to identify policy options with the potential to restore these community assets. Of the four types of capital, the most amenable to pol- icy intervention at the local level is social capital. In Chapter 2, we defined social capital as "networks and local associations . . . that might support collective action, enforce norms, generate expectations or reciprocity, or foster feelings of mutual trust." What is the relation between social capital and urban governance? Among the reasons social capital is important is that both the size and density of social networks and institutions, as well as the nature of interpersonal inter- actions, significantly affect the efficiency and sustainability of development pro- cesses. Violence erodes social capital when it reduces the trust and cooperation within formal and informal social organizations and among their members that are critical for a society to function (Moser, 1998; Moser and Holland, 1997~.
THE CHALLENGE OF URBAN GOVERNANCE Social capital in formal institutions 387 In contexts characterized by human rights violations and high impunity rates, violence often erodes faith in the relevance and governability of many formal social institutions. When judicial, educational, health, media, and security insti- tutions are no longer able to function appropriately and transparently, democracy itself is challenged. In some contexts, violence-linked industries associated with drugs, diamonds, or other natural resources can erode the state by corrupting in- stitutions and dividing the population. For instance, drug traffickers' systematic threats and attacks against the communications media effectively suppress the ef- fective and peaceful participation of civil society institutions in political decisions at the community and national levels. Social capital in informal community-level institutions The capacity for community-level organizations to function depends on levels of cohesion and the ability to meet locally, which in turn hinge on personal safety issues. Sustained violence creates fear and reduces trust among neighbors and communities. Fear of crime is greater where rates of contact crimes are higher (United Nations Interregional Crime and Justice Research Institute, 1998~. The response, particularly for women, is frequently to avoid certain places after dark. Such restrictions break down community cohesion. As noted earlier, fear increases urban fragmentation, resulting in a new ur- ban landscape made up essentially of fortified areas from which the poor and marginalized are excluded. Rodriguez and Winchester (2001) discuss how feel- ings of insecurity in Santiago, Chile, have led to reduced interaction across so- cial class levels, resulting in neighborhoods more homogeneous in income level and reduced social mobility. Agbola (1997) describes a crime prevention strategy in Lagos, Nigeria, based on environmental design, whereby the physical envi- ronment is 'manipulated' to deter crime more effectively through the installment of fences, residential enclaves, and other segregational measures. As discussed earlier, fear of crime in Sao Paulo has led middle- and upper-income citizens to segregate themselves spatially from the rest of the city (Caldeira, 1996~. In some contexts, violence both contributes to and is the result of the creation of "perverse" social capital.3 A primary example of perverse social capital is gang involvement, whereby young people bereft of strong family and community support form mutually reinforcing groups. In many poor neighborhoods, gangs form the main nexus of socialization for children, who join as young as age 12 or 13. Often gangs may be at war with rival groups involved in robbery, theft, drug 3A useful distinction can be made between productive and perverse institutions. While productive institutions aim to provide benefits in order to improve the well-being of the community, perverse institutions benefit their members but are usually detrimental to the community or society at large (Rubio, 1997; Moser and McIlwaine, 2000a: 78).
388 CITIES TRANSFORMED distribution or consumption, and assaults. In some communities, gangs protect their neighbors, committing crimes elsewhere; in others they prey on their neigh- bors, creating a climate of fear (Moser and McIlwaine, 2000b; Rodgers, 1999~. In Nicaragua, for example, almost half of all crimes and delinquent acts are attributable to youth gangs. Rodgers (1999) highlights five key factors linked to youth gangs in Latin America: their context is generally that of urban poverty; their behavioral patterns and formation are highly localized; their relationship with the local community may be protective or violent; when involved in drug traffick- ing, they tend to be more violent; and postwar migration and deportation of illegal immigrants are affecting their formation. Social capital in household relations Violence erodes beneficial household relations when it reduces the households' capacity to function effectively as a unit. High levels of stress in conflict zones, for instance, where many men join illegal guerrilla or paramilitary groups, seriously disrupt family life. In poor urban communities, many women identify a direct linkage among male unemployment, alcohol abuse, and increased domestic vio- lence, which may result in increases in female-headed households. In Guayaquil and Budapest, for example, domestic violence was found to be the single most important cause of household restructuring (Moser, 1998; Moser and McIlwaine, 1999~. Family members both male and female are put in a vulnerable position when communities are displaced by violence. Women are often more vulnerable than men at the moment of eviction, when they are exposed to the risks of flight and separation from their homes. While men appear better equipped to cope at such times, the reverse is true when displaced households restructure their lives; then the impact is greater for men, who become unemployed and experience a loss of status as breadwinners and a shattering of their sense of masculine identity. Women appear better equipped to develop support networks so they can continue the routines of daily survival and find new ways of earning an income, creating social capital not with other women originating from the same area, but with those sharing the same history of displacement (Meertens and Segura-Escobar, 1996~. There are many initiatives under way to deal with urban violence. Currently in many cities, one of the biggest constraints is not the absence of interventions but the lack of a coordinated approach to reducing violence. The impact of im- pressive menus of initiatives implemented by government, private entities, and NGOs is often limited by fragmented approaches involving narrowly focused, independent programs. If informed policy decisions regarding the relative fiscal returns of different programs are to be made, one of the most important priori- ties is the development of a cohesive policy that integrates and combines different objectives and instruments for reducing violence, and encompasses monitoring of key indicators as well as rigorous evaluations. This is largely a governance issue.
THE CHALLENGE OF URBAN GOVERNANCE 389 Ultimately, the debate on preventing urban crime in the developing world can- not be separated from broader issues of improving urban governance and manage- ment. This linkage implies closer integration of crime prevention and reduction imperatives within the context of more general city planning and management strategies. In some contexts, coordinated initiatives to respond to the increasing problem of violence have emerged at the city level. The role of local govern- ment in reducing violence and crime is increasingly recognized as a practical and effective solution (South Africa, 1998~: Local government, the level of government which is closest to the cit- izenry, is uniquely placed to actively participate in social crime pre- vention initiatives and to redirect the provision of services to facilitate crime prevention. Local municipal government tends to represent the lowest level at which plan- ning can take into account the needs of local communities and their particu- lar crime problems, and can therefore create effective linkages among locally elected officials, municipal departments, and the national police service (Straw, 1998~. Nevertheless, budget constraints and a lack of capacity often characterize developing-country municipal governments, and the consolidation of local gov- ernment structures presents an opportunity to integrate crime prevention into the line functions of municipal departments (Shaw,1998~. Along with government, civil society and local communities have an impor- tant role to play. Cohen and Swift (1993), for instance, advocate a public health approach that prioritizes coalition building and the development of networks of service providers. Thus, for example, one local health prevention program en- courages family violence services, alcohol abuse agencies, child abuse prevention agencies, rape crisis services, conflict resolution services, drug abuse agencies, and suicide intervention services to form a network of support to coordinate com- prehensive prevention services: Coalitions can accomplish a broad range of goals that reach far be- yond the capacity of any individual member organization. The work of an active coalition can range from information-sharing and co-ordination of services to advocacy for major environmental or regulatory changes. (Cohen and Swift, 1993: 62) Finally, the community can play an important role in reducing violence by creating its own authorities or groups to tackle the problem. Guha (1998) cites the example of the "peace committees" that are frequently formed in Indian commu- nities after riots. These committees often include members from both conflicting communities, considered to be "respectable citizens" who work to reduce the con- flict. The approach was found to be particularly effective in Calcutta for avoiding and containing many incidents that had the potential to escalate into major riots
390 CITIES TRANSFORMED (Guha,1998~. However, since the causes and manifestations of violence are con- text specific, the details of integrated intervention frameworks need to be tailored to the requirements of different situations. The Authority Dimension The basic parameters of urban governance in many developing countries have un- dergone a major transformation over the last decade and a half. Contributing to this transformation have been the relentless demographic growth of cities, with associated economic and geographic expansion of the urban base; the worldwide trend toward Revolution of both power and financial resources from the national to the local level in most developing and transitional countries (as discussed earlier); the spread of powerful local movements, such as social and environmental move- ments in Latin America, the "people's power" movement in the Philippines, the anti-apartheid movement in South Africa, and the workers' movement in Poland; and the powerful trend toward democratization and political pluralism, along with the emergence of robust civil society institutions (such as NGOs) at both the lo- cal and international levels (UNCHS, 1996: Chapter 5~. These driving forces have been reinforced by specific institutional reforms in particular countries, such as the constitutional reforms in Brazil, India, and South Africa discussed above; the Ley de Participacio'n Popular in Bolivia; the Local Government Code of 1991 in the Philippines; the extension of communal status to towns and cities, small towns, and all rural settlements in Cote d'Ivoire in 1980, 1985, and 1996, respectively (Crook and Manor, 1998: 141~; and a host of other reforms, well documented in the literature on Mexico, Colombia, Uganda, Indonesia, Thailand, and China. The particular governance structure currently in place in individual countries and cities is a result of the intersection of the decentralization process; the level of democratic reforms expressed through municipal restructuring; and local circum- stance, including leadership and the general involvement of civil society groups. Overall, urban governance appears most advanced in Latin America with regard to finding new modalities to manage and relate to growth, followed by Asia and then Africa (at least outside a few selected leaders in urban governance reform, such as South Africa and Cote d'Ivoire). While the focus of this chapter is on large metropolitan regions, the governance structure in these regions is invariably influenced by the overall approach to decentralization and local government in the country as a whole. Some examples follow. Decentralization in the sense of genuine devolution of power and financial resources (or the ability to raise finances) from the national to the local level has had a rather limited trajectory in Africa. One of the reasons for this is certainly the legacy of the centralized, statist bureaucratic regimes of colonial powers that did not cede independence until the 1960s and even later (whereas formal decol- onization took place 20 years earlier in South Asia and Indonesia, and a century earlier in Latin America). Another reason is that most postindependence regimes
THE CHALLENGE OF URBAN GOVERNANCE 391 in Africa were ruled initially by powerful leaders and single parties who relin- quished little ground to autonomous social formations or localities. By the 1980s, this picture had begun to change across the continent, and by 1991, only 8 of 54 countries for which reliable data were available could be con- sidered clearly "authoritarian," with 3 more being governed by a system that could best be described as "directed democracy." In addition, 8 countries were "demo- cratic," while 35 were said to be in various stages of "transition" to more demo- cratic forms of government (Africa Demos, l991~. Through the next decade, the full democratization of South Africa in 1994 and the return to democracy of Nigeria in 1998 were the major events, even though a number of transitional countries became politically unstable. At the end of the decade, the respected NGO Freedom House reported that of 53 countries evalu- ated,8 were "free" (i.e., full electoral democracies), 21 were "partly free," and 23 were "not free" (i.e., authoritarian systems) (Karatnycky, 1999~. Overall, there was little change over the decade. Abidjan A former French colony, Cote d'Ivoire has always had one of the highest levels of urban development and most elaborate urban policy systems of any francophone country (Cohen, 1974~. As of 2000, its largest city, Abidjan, had a population estimated at 3.3 million (United Nations, 2001: 86) over a land area of 627 square kilometers (Attahi, 2000: 10~. This coastal metropolis, often referred to as "the pearl of the lagoons," represents some 40 percent of the country's total urban population and about 75 percent of its formal employment (Attahi, 2000: 10~. Like other major cities in the West African francophone region, and in spite of its relative affluence in comparison with other cities of the region, Abidjan suffers from insufficient housing and infrastructure (20 percent of its population lives in irregular or "spontaneous" housing in unserviced neighborhoods), strug- gles to remove ever-increasing amounts of household and industrial refuse, and cannot build enough roads to keep up with the increasing number of motor vehi- cles. Even though importations of cars and trucks were restricted in 1998 to vehi- cles no more than 7 years old, the number of these vehicles has been growing at a rate of 9 percent per year (Attahi, 2000: 16) several times higher than the rate of population growth, estimated at 3.16 percent per annum for the period 2000-2005 (United Nations, 2001: 97~. Abidjan's fabled high-quality infrastructure and large formal employment base have been deteriorating since the late 1980s, as popula- tion growth has fallen from a high in the 11 percent per annum range up to the late 1970s to a figure closer to 3 percent today, a level lower than that of middle-sized towns in the region (Dubresson, 1997: 266~. Like other ax-colonial states in Africa, Cote d'Ivoire emerged after indepen- dence (in 1960) as a relatively centralized country. Then, beginning in 1978, a de- centralization process directed by the President saw, in succession, the restoration
392 CITIES TRANSFORMED of "commune" (or municipal) status to the major cities in the country, including 10 communes in Abidj an; the amalgamation of the 10 Abidjan communes into a second-tier government known as the City of Abidjan; and the extension of com- munal status to 98 smaller towns (in 1985~. Finally, 3 years after the death of President Houphouet in 1993, his successor extended municipal status to the re- maining rural areas of the country, increasing the total number of local government units from 136 to 196 (Crook and Manor, 1998: 141~. In terms of the fastidious details of its planning, as well as the follow-up of support to local councils, the Ivoirian decentralization exercise stands as one of the most thoroughgoing and successful in Africa. Two aspects of this exercise are worthy of mention here. First, beginning in 1990, the Ivoirian government loosened control over the political system, permit- ting parties other than the government party to contest both local and national elections. As a result, independents and opposition parties (based on a list sys- tem) won 9 of the 135 communes in the 1990 communal elections; in the 1996 elections, the proportion grew to 27 of 196 communes (Crook and Manor, 1998: 149~. The real possibility of opposition parties winning local elections and ad- ministering local councils clearly enhanced the legitimacy of local government in many parts of the country. Second, while it decentralized, the government also strengthened its advisory and central administrative systems, so that local councils could receive admin- istrative and technical support as needed. The Department of Local Government (within the Ministry of the Interior) grew impressively, its professional staff reach- ing 108 in 1995 (Attahi, 1996: 122~. The governance of the Abidjan metropolitan area consists formally of a two- tier structure: at the lower tier are the 10 communes, of differing size and wealth, each having elected councillors and an elected mayor; at the second tier is the City of Abidjan, consisting of the mayors of the 10 constituent communes, plus four more councillors from each commune. The mayor of the City of Abidjan is elected by the mayors of the 10 communes at the first meeting of the collective legislative body, called the "grand council." After his election, he resigns from his communal position, handing the position over to his "assistant mayors" at the local level. The major functions of the upper-tier government are waste disposal and man- agement; public lighting; sanitation; traffic regulation; maintenance of roads, parks, and cemeteries; and town planning. The communes administer markets, allocate plots for public purposes, deal with the maintenance of primary schools and clinics (but not school or health policy, let alone the supervision and payment of professionals, which are national responsibilities), operate social centers, and share functions with other levels of government with respect to pollution and hy- giene. Major services, such as waste removal, electricity, and water, are in the hands of private companies, albeit under some level of surveillance from either the local or national government.
THE CHALLENGE OF URBAN GOVERNANCE 393 Problems with revenue exist at all levels, even in relatively affluent Abidjan, where the differences among communes are sharp. In the early 1990s, for exam- ple, the 3 wealthiest of the 10 communes of the City of Abidjan spent, per resident, an average of 49 times the amount spent by the 3 poorest communes on recurrent expenditures and 6 times the amount for capital expenditures (Dubresson, 1997: 285~. Although the metropolitan system has been functioning for 20 years, observers identify a number of problems. For example, the national government has been less than fully cooperative in permitting the City of Abidjan to exercise some of the functions for which it has a legal mandate. These include the inspection of construction sites, the issuing of drivers' licenses, and the control of firefighting and rescue operations. Since the main source of the City's revenues, the property tax, is collected by the national government and then remitted to the communes (which then pay a fixed proportion to the City), the City has little direct influence over its finances. Major service operations are controlled by private corporations that cannot easily be controlled by the City. Furthermore, since the mayor of the City is not elected at large but chosen by the mayors of the constituent com- munes, he does not have an independent political base from which to promote metropolitan-wide policies. Already the mayor of the wealthiest commune has resigned from the grand council on the grounds that the obligatory funds (40 per- cent of revenues) paid to the City could be used more effectively locally. On a more positive note, a number of communes have set up neighborhood man- agement committees and endowed them with an administrative budget. These initiatives have taken place without higher-level support. In the 1996 local elections, 1 of the 10 municipalities in the Abidjan area Adjame returned a mayor, Marcel Amondji, who was a member of one of the opposition parties. This mayor has since set up 19 neighborhood management councils to involve citizens in local affairs. Each has an annual budget equivalent to US$4,500 to cover the costs of the local office. According to a New York Times observer (French, 1998: 4), Adj ame has been all but transformed. Once trash-filled streets are now kept clean by broom-wielding city workers [a program called "extreme hygiene". A multistory African-style market is rising to replace a warren of cluttered and dangerous side streets that served as the neighborhood's informal bazaar. And white-smocked inspectors reg- ularly make the rounds of the community's innumerable cafes ensur- ing that food served to the working class population is not only cheap, but sanitary. The mayor attributes these changes, at least in part, to the competitive electoral system. "What is being done here," he says, "represents a night and day change from the past, when elections were formalities and the office of mayor was largely an honorific title. We have managed to triple our budget by raising local taxes,
394 CITIES TRANSFORMED and nobody has complained because they see that local government is giving them valuable services for the first time." (Amondji quoted in French, 1998.) In Abid- jan, as throughout the country, decentralization has been a relative success, but local resources are still very limited, and the level of activity of civil society is barely perceptible. Mexico City Our second example is one of the largest metropolitan areas in the world Mexico City. For many years, Mexico City's massive, expansive growth was regarded as virtually inexorable. In 1991, for example, the population of the agglomeration was predicted to reach 25-27 million by 2000 (Rowland and Gordon, 1996: 173), although that prediction was scaled down considerably when the 1990 census fig- ures were released. The notion of an urban "leviathan" somehow out of control was reinforced by the city's fabled scarcity of water, its high levels of air pollution and other toxic chemicals, its extreme traffic congestion, and the struggles of its poor citizens to obtain decent and affordable housing and accessible infrastructure. While the city's service and infrastructure problems remain severe, many well- documented studies cited in earlier chapters show a decrease in its population growth during the 1980s and 1990s, a dispersion of its population from the cen- ter to the peripheral areas of the metropolitan region and beyond, and a gradual reduction in the pattern of socioeconomic differentiation among its major areas. Paradoxically, although there appears to have been a decline in social disparities in the city if one takes delegaciones as the unit of analysis (Rubalcava and Schtein- gart,1987,2000), at the very local level (Ward, 1999: 30), the segmentation, separation, and dividing line (barriers) between rich and working class neighborhoods is probably increasing. Mex- ico City, like many Latin American cities, has seen a dramatic rise in violence levels in the 1990s. Private security firms are increasingly hired to secure the perimeters of upper and middle income residential neighborhoods, making them effective no-entry zones for working class and outsider populations. Moreover as a growing number of urban services are contracted out to private operators, this serves to segment still further the transactional separation of rich and poor. Currently, the population of Mexico City which includes both the Federal District (with 16 delegaciones) and 41 municipalities to the north, east, and west in the states of Mexico and Hildago is estimated at 18.13 million (United Na- tions, 2001: 88), with an annual growth rate that has been shrinking for many years. Rates of both annual natural increase and inmigration have been declining since the middle of the past century. From a high of 6.7 percent in the 1940s (Ward, 1998: 48), the overall population growth rate for the metropolitan area has
THE CHALLENGE OF URBAN GOVERNANCE 395 declined to an estimated 1.81 percent, and it is projected to fall to as low as 0.35 percent over the next 5 years (United Nations, 2001: 99~. The reasons for this reversal in trends are discussed elsewhere in this vol- ume, but Mexico City remains a relatively poor city with serious problems in both production and distribution of public goods and services. Its current highly lim- ited metropolitan governance system is a product of both the country's history of decentralization and democratization and the unwieldy nature of a jurisdictional division between the Federal Capital district on the one hand and the adjacent municipalities of the State of Mexico on the other. For decades, governance of the Mexico City metropolitan area was a relatively low-key affair, involving a presidentially appointed regente of the Federal District and the mayors and administrations of the adjacent (but much less developed) towns and settlements of the periphery. Because of the importance of the capi- tal city, the regente position was a cabinet-level post in the federal government. Given the highly centralized and even personal nature of the Mexican political system until recently, ultimate control of the capital city was in effect vested in the president of the country. In practice, what planning and coordination existed between the two jurisdic- tions occurred often through the involvement of federal government ministries in both the State of Mexico and the Federal Republic, reinforced politically by the relationship between the governor of the state and the regente, the latter being, until the l990s, from the same governing party (although the regente was a much more senior position in terms of protocol). During the 1970s and 1980s, plan- ning initiatives and institutions tended to operate either at the national level (for the whole country), or in one or the other of the Federal District or the State of Mexico. Metropolitan institutions, such as a Conurbation Commission, a secre- tariat for planning and budgeting, and a Council for the Metropolitan Area, came and went according to the political inclinations of different regentes. But with little or no public participation in planning and a multiplicity of competing and overlapping jurisdictions, the economic and social forces inherent in the growth of the city took their own paths, with little guidance from planners (Ward, 1998: Chapter 5~. Aside from the appointed regente and his staff, the city was further subdivided into 16 delegaciones (municipalities before 1928), whose administrative directors (delegados) were in turn appointed by the regente. During the 1970s and 1980s, various local and neighborhood councils were set up to give residents a say in local matters, but these groups had at best a consultative role relative to the senior appointed administrators (Ward, 1998: 118-9~. In 1994, an effort was made by the dominant party, the Partido Revolucionario Institucional (PRI), to introduce "citizen advisers" as elected, paid officers at the level of the delegacion, mandated to assist with local community service requests. Without explaining its action, however, the government discontinued this reform 3 years after its initiation (Eck- stein, 2000: 192~. The arbitrary nature of local government in the capital city,
396 CITIES TRANSFORMED at least until nearly the end of the decade, led the author of a textbook on local government in Latin America to state that "few major cities in the world have less local democracy than Mexico City" (Nickson, 1995: 199~. Traditionally, cities had little status in the Mexican system of government and politics. Article 115 of the 1917 Mexican Constitution gave states power over local government but spoke of the municipio libre as the cornerstone of territo- rial administration. It was not until 1983 that a new phase in intergovernmental relations began with a number of amendments and additions to Article 115. In proposing these changes, the president, Miguel de la Madrid (1982-1988), sug- gested a major "sea change" was in the offing. As one commentator observed of de la Madrid's campaign for the presidency, the candidate made clear in public meetings throughout the country that (as cited in Rodriguez, 1997: 69) his purpose is to eliminate, or at least minimize . . . excessive central bureaucracy and population concentration, tardiness and inefficiency in some sectors of the federal government, dependency and financial weakness of the municipalities, and so forth. The time has come to "federalize" the "National Life." Each and every aspect of the state's activities which before were forced to centralize, are now obliged to decentralize. Although de la Madrid went on to establish a system that was increasingly de- centralized in administrative terms while remaining highly centralized politically, the changes to Article 115 were the beginning of an altered political environment for the municipalities. The new article approved by Congress in February 1983, to take effect on January 1, 1984, specified the public services for which munic- ipalities had prime responsibility and suggested that the municipalities should be governed according to organic laws passed at the state level. The municipalities were also guaranteed their own sources of revenue, in particular a property tax. For many years thereafter, municipalities still relied largely on transfers from the federal government for the bulk of their revenues, but the potential existed to ex- pand their own-revenue accounts. After the reform was passed, the president set up a National Center for Municipal Studies to serve as an information and research center for reform-related issues. The most recent constitutional reform relevant to Mexican cities (and to Mex- ico City) is a revision of Article 115 passed by the Mexican Congress in June 1999. This revision, a result of negotiations among the three major political parties, rec- ognizes local government as an essential level of Mexican government. It not only states that local governments (municipios) are to be governed by elected councils consisting of regidores and syndicos, but also specifies a list of functions and pow- ers that are under their exclusive jurisdiction. Such functions cannot henceforth be limited by the state governments. They include the provision of drinking wa- ter and all related services, public lighting, cleaning and waste disposal, markets, police, streets and gardens, planning and land regulation, and other services the
THE CHALLENGE OF URBAN GOVERNANCE 397 municipalities judge to be within their capacity to administer. The municipalities are further empowered to set tax rates and to collect the revenues necessary to carry out the above functions (Guerrero Amparan and Lopez, 2000~. While Mexicans have traditionally elected their mayors and councillors (ex- cept in the Federal District), the elections had little meaning for many years as long as the PRI controlled the selection of all candidates. But political pluralism (albeit constrained at first) began to erode this situation as, beginning in 1983, a significant number of municipal elections began to be won by opposition parties. While some of these victories were orchestrated by the PRI itself and some re- sulted from splits within the dominant party, others (such as those in Durango, Ciudad Juarez, and Chihuahua) reflected genuine support for the Partido Accion Nacional (PAN) (Rodriguez, 1997: 53~. The story of opposition politics at the subnational level in subsequent years is one of conflict between emerging opposition sentiment and (often) PRI tactics to discredit or undermine the local victories of opposition candidates (Rodriguez, 1998: 174-83~. By 1996 the PRI controlled 1,551 of the 2,412 municipalities in the country (or 64 percent); the PAN controlled 225 (or 9 percent); the Partido de la Revolucion Democratica (PRD) controlled 181 (or 8 percent); and "others" controlled the rest (19 percent), including Oaxaca state, where indigenous groups elected their civic leaders in 413 municipalities (Rodriguez, 1998: 177~. At this point, most major cities were under opposition control, with PAN mayors being significant in the more-developed north and the PRD more important in the area around Mexico City and in the south. Rodriguez (1998) argues that the advent of more competitive politics has brought more scrutiny of candidates by voters. This in turn means that more competent individuals are selected and that the quality of governance is enhanced by greater transparency (Rodriguez, 1998: 183~. On the other hand, many municipalities especially smaller, poorer towns and towns in the less industrial and developed areas of the country are not competitive. Did the new political pluralism make a difference in urban management? A comparison of three cities with opposition regimes in the 1990s Tijuana and Cordoba under the PAN and Nezahualcoyotl under the PRD shows that dur- ing the decade, the new regimes were far from fully successful in implementing their campaign promises of administrative rationalization, greater efficiency, and improved citizen participation. Setting aside the case of Nezahualcoyotl, whose municipal council had been elected only a short time before the research com- menced, the other two municipalities had been able to improve the delivery of public services to their citizens (in comparison with the previous PRI regimes), and efforts were at least under way to modernize the management of the munic- ipal enterprise. While they did not observe regression, the researchers are far from categorically enthusiastic about the changes that have ensued (Duhau and Schteingart, 1997~. And a historical analysis of the politics and policies of var- ious mayors of Mexico City since the 1950s (Davis, 2002) reveals that different mayors have exercised considerable autonomy in developing their own signature
398 CITIES TRANSFORMED policies in spite of the fact that they were not elected until recently, and that this autonomy has been based at least in part on the fact that Mexico City as a political constituency of the national political elite must be heard and responded to (Davis, 2002~. A major reform in Mexico City in particular was the decision by the Mexi- can government to permit elections in the Federal District, first at the level of the whole district and then at the level of the constituent delegaciones. In 1997 for the first time ever, the post of "head of government" of the Departmento del Distrito Federal was made an elective one, to be supplemented by elections throughout the city for positions in a new Representative Assembly. Of the 40 elective posi- tions in the assembly, 2 were won by the PAN and the other 38 by the PRD. The new head of government or mayor, who won by an overwhelming margin, was 63-year-old Cuauhtemoc Cardenas, himself a former candidate for president and the son of a revered former president. At the same time that the PRD won the local government elections, it also captured 28 of the 30 single-member Mexico City congressional district seats in the national elections. A careful interpretation of the vote shows that the main supporters of the PRD were working-class vot- ers who were strongly antagonistic toward the ruling party. Many switched their intended votes during the campaign to align with the party that best expressed their antigovernment sentiments. As in other elections, the Mexico City elec- tions showed that "the prototypical PRI voter was a housewife with limited ed- ucation; the classic PAN supporter an educated Catholic; and the paradigmatic PRD voter a politically engaged member of the working class" (Lawson, 1999: 151~. Among the watchwords of the new administration were the cleaning up of cor- ruption and crime, efficiency in service delivery, and popular participation. Be- cause the national government did not support, in terms of either resources or policies, the opposition PRD government in the Federal District, the mayor could achieve success in these particular objectives without calling on assistance from the center (Davis, 2002: 251~. Nevertheless, Mayor Cardenas left his position to run (unsuccessfully) for the presidency in 2000. The fact that the mayor had early on demonstrated his interest in national political issues (such as the situation in Chiapas) may have contributed to the widespread feeling among poorer people in Mexico City that, although electoral democratization was important, parties still operated as "clans" and did not represent them. "Residents had learned the hard way to be cynical about political change" (Eckstein, 2000: 194~. Change, however, continued. In 2000, not only did the Federal District mayor stand for general election (won again by the PRD candidate, garnering 1.7 million of a total of 4.3 million votes cast) (Gomez Tagle, 2000: 661), but so, too, did the councillors within individual delegaciones (Rodriguez Araujo, 2000: 657~. Despite the new, more democratic system in the Federal District, however, no new structures of metropolitan governance have been instituted, and the larger metropolitan region is still almost totally bereft of a framework for planning.
THE CHALLENGE OF URBAN GOVERNANCE Local participation in Chinese cities 399 Although Western-style pluralist electoral politics is absent in China, major changes in the governance of Chinese cities have been proceeding since the late 1980s. These changes have taken place at the local level of the urban community in Chinese cities and, according to one estimate, affect 200-300 million of its cit- izens (Choate, 1998: 6~. While these changes are complex, they can be discussed from two vantage points: the formal structure of urban government and the op- eration of the local residents' committees that have become so prominent in the larger cities since the mid-199Os. During the period of centralized planning and administration in postwar China (roughly 1949-1978), the nation's municipal administrative and political structure was characterized by the penetration of local communist party structures into all levels of urban administration. The secretary of the Communist Party's city com- mittee, representing the party, was the most powerful individual, notwithstanding the fact that he had no position in the city government. The mayor, as the head of the city administrative bureaucracy, was second in importance, but at the same time was usually the party committee's deputy secretary. Decisions were nor- mally initiated by the party and then carried out by the city government. The city government had control over land and construction, but many functions related to urban services (such as health care, housing, and primary education) were under the control of the work unit, or danwei, which could be a large government office or an industrial plant. Beneath the level of the city government, cities were divided into districts (of which the largest cities had up to 14), subdistricts (or street offices), and neigh- borhood (or residents') committees. In this system of tight vertical control, only individuals who were formally working for a government-approved danwei were permitted to live in the city. Rural-to-urban migration was tightly circumscribed, and only those with a resident's permit had access to housing, rationed commodi- ties (including staple grains and oils, meat and fish, cotton cloth, and most con- sumer durables) and social services. "The government managed not only urban development, but also urban residents' lives" (Zhang, 2001: 187~. With the introduction of market-oriented reforms beginning in 1978, local governments (as discussed above) were given more freedom to raise and distribute financial resources, responsibilities were transferred from the national to the lo- cal level, and the government stopped managing every detail of urban citizens' lives. The formal structure of municipal government at the city level (including the relationship between government and party leaders) did not change; around 1985-1986, however, the government began to place much more emphasis on the structure and functioning of submunicipal government, in particular urban resi- dents' committees (Choate, 1998: 10~. Behind this decision were a number of important factors: massive movement of rural migrants into the cities, a rapid in- crease in urban under- and unemployment caused by enterprise restructuring, and
400 CITIES TRANSFORMED a withdrawal of the social safety-net functions formerly provided by large urban work-based units. From 1989 to 1993, a number of laws reorganizing residents' committees were passed, and a 1992 joint policy paper and a 1994 major government cir- cular laid the groundwork for a more elaborate and stronger role for the local committees. By the late 1990s, these residents' committees (which catered to modal populations of some 2,000 each) were systematically involved in some or all of the following services: information and record keeping; public safety and security; mediation of local civil and family disputes; public health and family planning; environment and sanitation; legal education of the public; provision of "convenience" services, such as fast food restaurants, public transport, and pub- lic telephones; social welfare (especially for young children and the elderly); and employment placement (Choate, 1998: 16-25~. According to observers, the aver- age age of committee members is declining, and the committees are hiring more educated, experienced, and professional staff; paying more attention to the needs of the people through social surveys and feedback mechanisms; and connecting with newly formed associations, such as volunteer social service groups, propri- etor and land development associations, and cultural organizations (Choate, 1998; Read, 2000; Ying, 2000; Zhang,2001~. By the end of the 1990s, there were some 119,000 of these local committees throughout urban China. In Beijing and its suburbs in 1997, for example, there were 10 districts, l l 8 street offices, and 5,026 neighborhood committtees (Read, 2000: 807-8~. In many of their local initiatives, committees organize to collect fees and de- velop local projects. Ying (2000: 8-9) gives a good example of the new local management style in Shanghai, a style that often involves city residents actively reporting their complaints to the residents' committee office, which follows up with energetic and probably financially advantageous measures: The Quxi Road Market for Agricultural and Non-staple Products un- der the jurisdiction of the Wuliqiao Subdistrict [Street Office] in Shanghai [is a good] example. In 1997 the market was still dirty, chaotic and jammed with traffic. In addition, it was a market without effective management where some small private retailers ran ram- pant, gave short weight and beat up administrators. The local 2,000 households or more reported all this to the subdistrict office for ac- tion. Cadres of the office immediately held a meeting to discuss the matter and took measures. They cut the number of stands in the market from over 300 to 144. They also invested 260,000 yuan in building unified, standard permanent stands. They also organized an all-weather six-member sanitation team which worked on two shifts a day. The law enforcement team also entered the market to exercise supervision and strict management at regular intervals, thus reducing the number of problems. In 1998 the market was elected an advanced
THE CHALLENGE OF URBAN GOVERNANCE exchange at the city level, and acclaimed by local residents. Find- ing problems through letters and calls and helping the people allay their worries and tide over their difficulties have become a distinc- tive feature of the Wuliqiao Subdistrict Office in its effort to offer good community service. According to statistics, in 1998 the office processed 253 letters from the people and received over 4,000 calls. These changes were not present in or before the early 1990s. They are also unprecedented in the development history of Chinese cities. so they are of great historical and practical significance. 401 While not all local committees may be as effective as this particular exam- ple, Choate (1998: 28) reports that, based on his observations and interviews in 14 cities over a 3-year period, "it appears as though the work of residents com- mittees is reasonably well-regarded by the relevant populations themselves." Per- haps as a result, many cities are developing full-service community centers at the street office and district levels. They are assisted by citizen "boards" or "man- agement committees" consisting of representatives of major mass organizations and the party (Choate, 1998: 29-31~. While these groups are not elected, they constitute an interesting local solution to the challenge of representing a range of opinions and professional skills in an increasingly complex social and economic environment. IS THERE A "BEST" MODEL OF URBAN GOVERNANCE? In this chapter we have attempted a limited review of some of the more notable changes in urban governance in very large cities in the developing regions of Asia, Africa, and Latin America. This review, while far from comprehensive and sys- tematic (inasmuch as more extensive case studies and comparative data do not yet exist in the research literature), indicates that new institutional forms are emerg- ing in municipalities across the developing world. Common to most if not all of these new institutional approaches are three main elements: greater involvement of NGOs and community groups in local governance, often through a more plural and democratic electoral system; greater transparency and accountability in both the planning and implementation of local policy; and the devolution of more legal and constitutional responsibility for urban affairs from the state or national level to the local level. From the election of mayors and local councillors across Latin America, to the increasing pluralism of the political process in Africa, to the incor- poration of massive numbers of new actors in the Indian and Philippine municipal systems, to the involvement of nonstate actors in service and infrastructure pro- vision in China, a massive opening of political space is taking place at the urban level. Behind this enhancement of the municipal political role are multiple and com- plex structural changes: the emergence and more active participation of civil
402 CITIES TRANSFORMED society at the local level (often together with or as an offshoot of other social movements, such as those involving human rights, environmentalism, and indige- nous rights); the connection of municipalities and their activists with networks of other municipalities and elected officials; the decentralization of powers and func- tions from national to local government units; and the new power and influence that cities especially large cities are assuming in a competitive and globalizing world. Given the diversity of the new institutional arrangements coming into active use, as well as the fact that each country has its own historical and political cir- cumstances that vest local governance with a special logic and legitimacy, can common elements be discerned? This question is particularly pertinent since cities and their surrounding regions increasingly see themselves as being in competition with other cities and regions around the world (Scott, 2001~. At stake are poten- tially large investments on the part of multinational institutions (both private and public) that can make a major contribution to the employment prospects, and thus the economic welfare, of any city. Since the overall quality of governance is re- garded as contributing to a city's ability to manage its infrastructure and services and maintain a certain quality of life for its citizens, alternative models of good governance frequently figure in discussions of ways to improve the competitive position of cities. Reforms in local governance and the technical support needed by developing municipalities to put these reforms into practice effectively in their constrained economic circumstances have been the object of significant program assistance from international agencies. These agencies include the Urban Management Pro- gram of UN Habitat (formerly UNCHS); the World Bank; a number of inter- national municipal and local government associations; and various bilateral aid agencies, such as the U.S. Agency for International Development and Germany's Deutsche Gesellschaft fur Technische Zusammenarbeit (GTZ). In 1999, Dinesh Mehta, current director of the Urban Management Program, compiled three lists of "good governance" criteria to be incorporated in advice proffered by interna- tional agencies: (1) a Habitat II list (including accountability, transparency, par- ticipation, the rule of law, and predictability); (2) a United Nations Development Program (UNDP) list (including participation, the rule of law, transparency, re- sponsiveness, consensus orientation, equity, effectiveness and efficiency, account- ability, and strategic vision); and (3) a list compiled by Konrad Adenauer Stiftung for the "Better Cities Network of East and South-East Asian Cities" (including accountability, responsiveness, management innovation, public-private partner- ships, local government-citizen interaction, decentralized management, network- ing, and human resource management) (Mehta, 1999~. Of course, each program has developed explanatory text as to how these crite- ria might be applied, and there are many other lists. Virtually no research of which we are aware has been done on the relative effectiveness of different elements of these lists in either controlled or comparative situations. What is clear from these
THE CHALLENGE OF URBAN GOVERNANCE 403 lists is that an idealized liberal-democratic model of urban governance (in which well-informed citizens protected by legal institutions make periodic democratic choices of teams of politicians and professional bureaucrats to manage and de- velop their local space) is the suggested institutional format. While improving governance is important in smaller and intermediate-sized cities, the analysis of governance models most commonly involves the largest cities, which have the most at stake in the new economic dispensation. By common agreement, achieving good governance in the largest metropolitan regions is more difficult than in smaller cities. Populations are larger and more diverse than in smaller cities, the level of resources necessary to institute and maintain services and infrastructure is much higher on an aggregate basis, civil society groups are more organized, and the range of functions for which the city is responsible makes the organization and logistics of urban management much more complex. As an element of his widely discussed argument that much of the world is moving from order toward anarchy, Kaplan (2001: 55) singles out urbanization: The 21st century is going to be the first century in world history when more than half of humanity will live in cities. Even sub-Saharan Africa is almost 50 percent urban. Urban societies are much more challenging to govern than rural societies. In rural societies people can grow their own food, so they are less susceptible to price increases for basic commodities. Rural societies don't require the complex in- frastructure of sewage, potable water, electricity, and other things that urban societies have. Urbanization widens the scope of error for lead- ers in the developing world while simultaneously narrowing the scope for success. It is harder to satisfy an urban population than a rural population, especially when that population is growing in such leaps and bounds that governing institutions simply cannot keep pace. As we have seen, very large cities do indeed present complex problems of urban management, but there is little or no evidence that they are sliding into an- archy. What we can conclude, which is also consistent (but not coterminous) with Kaplan's argument, is that the largest metropolitan areas will be more likely to house a highly diverse range of active civil society groups. Along with this or- ganizational diversity, large cities will harbor important movements of political opposition to the party or regime in power at the center. This tendency is most likely a function of both social and economic diversity and the pervasive pres- ence of global ties among the population. Thus in Kenya, the municipal elections of 1992 and 1997 reflected a largely antiregime sentiment, and both elections in Nairobi (the largest city) produced a majority of councillors representing oppo- sition parties. In Mexico City, as we have seen, the left-wing PDR carried the municipal election in 2000. In Brazil, the most recent municipal elections, in Oc- tober 2000, produced left-wing majorities (as against the center-left position of
404 CITIES TRANSFORMED the ruling party) in the three largest cities Sao Paulo, Belo Horizonte, and Rio de Janeiro. Political opposition may be one reason central governments are notori- ously reluctant to agree to municipal reform packages involving greater autonomy for their largest cities. One major issue arising in the design of governance systems for large metro- politan areas is that of coordination. Aside from the question of scale and size, the physical expansion of large cities in almost all cases extends the population and the rural (or ruraVurban) hinterland over which the city holds sway to include a large number of separate and even independent political jurisdictions. Coor- dinating these jurisdictions is both a political and an organizational challenge, particularly given the tendency of the effectively urbanized area to spread over an ever-larger territory. In one of the classic works on this subject, Jones (1942) argues that growing cities in the United States (including central cities and their suburbs) were increasingly being fragmented into a multiplicity of inadequately coordinated units. "This fragmentation ensured competitive behavior between local governments, an inability to solve regional problems, an uneven distribu- tion of tax resources, a lack of citizen control of local government, and the un- equal distribution of services, especially pertaining to mass transit, sewerage and garbage, water supply, public health, law enforcement, and firefighting" (Stephens and Wikstrom, 2000: 40~. To solve these problems, Jones argues, new arrange- ments to ensure metropolitan political integration should be undertaken. Since the time of Jones's study, a considerable literature has addressed this problem in both Europe and North America. Yet as metropolitan areas grow in both size and number around the world, we are no closer to finding suitable com- prehensive solutions. In an overview of four major metropolitan areas in Canada and five in the United States in 1995-1996, Rothblatt (1999) shows that the av- erage metropolitan area in Canada contains 40 municipalities, while that in the United States contains 157 municipalities and counties. This difference in North American cities reflects the American tendency to localized, individualistic so- lutions as compared with greater acceptance in Canada of a coordinating role for higher levels of government (Goldberg and Mercer,1986~. Nevertheless, there are exceptions to the trend in both countries, and there is today a disturbing tendency toward greater social polarization and fragmentation in both American and Cana- dian cities. This trend appears on the surface to be positively related to political fragmentation (which permits the separation of groups by income and race), but the relationship cannot be clearly demonstrated. (For the American literature, see National Research Council, 1999: Chapter 3.) As large cities in developing countries spread outward, the jurisdictional mo- saic becomes larger and more complex. As we have seen in the case of Mexico City, the capital region now includes some 41 municipalities in two states, a Fed- eral District, and 16 subunits within that district that are equivalent to municipal- ities. The metropolitan area of Sao Paulo consists of 39 separate municipalities, while Greater Santiago is made up of 34 separate communes. Rio de Janeiro
THE CHALLENGE OF URBAN GOVERNANCE 405 has 17 separate municipalities and Buenos Aires 20 local government units. In Africa, Abidjan consists of 10 communes and a second-tier government, but one must also include a number of privatized service agencies and central government ministries that administer important functions in the city region. In South Africa, there is currently a trend toward consolidation, but in the period immediately after democratization (in the early to mid-199Os), Cape Town was made up of 39 local government units and Durban as many as 69. The pattern extends to Asia as well, where, as we have seen, Bangkok consists of a Bangkok Metropolitan Admin- istration (divided into 50 districts) plus 5 adjacent provinces (which themselves include as many as 2,000 small local governments), while the greater Manila area includes 17 municipalities and up to 18 additional local government units in the surrounding mega-urban region. The governance of large (metropolitan) urban areas is currently a central issue in a number of Latin American countries as they try to reconcile a resurgent local democratic culture with the need to manage complex local functions efficiently so as to safeguard the economic benefits accruing to their city-regions. There are two parallel arguments in the literature on metropolitan governance a literature that, until recently, was based almost entirely on European and North American evi- dence. One argument addresses the question of whether there should be a unitary organization of government dealing with the entirety of the built-up metropolitan area. Assuming the answer to this question is positive, a second line of argu- ment has to do with whether there should be a single-tier or two-tier metropolitan government (Sharpe, 1995~. In principle, there are three possible organizational models arising from these options: (1) a loosely organized collaborative system involving relatively autonomous local governments that cover the metropolitan area; (2) a relatively unitary form of government incorporating the whole of the built-up area; and (3) a two-tier system of government covering the metropolitan area, with lower-level municipalities undertaking certain defined local functions and a higher-level council or metropolitan government dealing with functions of a regional nature. The first of these models parallels the fragmented model dis- cussed earlier in this chapter, while the second and third are variants of what we have called the comprehensive model. While the two-tier model was considered the most desirable system in prin- ciple for many years, it has been coming under attack for practical, political, and theoretical reasons (Sharpe, 1995: Chapter 2~. Among the practical problems are the difficulty involved in keeping up with the de facto extension of the boundaries of the metropolitan area, the problem of reaching agreement on exactly where the regional government begins and local municipal functions stop, and the distribu- tion of revenue both horizontally and vertically in large areas that are becoming fragmented and differentiated. There are, in addition, two major political prob- lems, one local and one intergovernmental. At the local, metropolitan level, those attempting to coordinate area-wide structures, whatever their democratic creden- tials (which are, in any case, usually limited) have great difficulty capturing the
406 CITIES TRANSFORMED loyalties and attachments of their citizens. Tasks are too technical, and local- ized political attachments to one's own commune or municipality tend to prevail. Area-wide sentiment, which the metropolitan structure could tap as a political resource, is in short supply; rather, tensions and conflicts among different local structures and communities within the overall metropolitan region appear increas- ingly common. At the intergovernmental level, there is almost always tension between a large and powerful local (metropolitan) government and a higher level of government such as a state, province, or central government that must de- centralize functions and allocate revenue to another unit with which it may very well be in political competition. Some would argue that the fragmented model is theoretically superior to the comprehensive model. An argument can be made, based on public-choice princi- ples, that a number of small local jurisdictions is superior to a single overarching government. Since a variety of small local government units can offer different baskets of services and taxes, the whole local area can operate as a quasi-market, supporting greater efficiency through a kind of competition among jurisdictions based on the choices of citizens about where to live (Tiebout, 1956; Ostrom, Tiebout, and Warren, 1961~.4 It appears clear that, whatever the challenges to governance in large urban re- gions in the developing world, no single or even dominant model of metropolitan governance is likely to emerge in the foreseeable future. This has not happened in Europe and North America (Lefevre, 1998) and is unlikely to happen in de- veloping countries, which are even more differentiated in history, geography, and culture. More effective structures for coordination need to be developed, but each solution will have to respond to a myriad of complex political, financial, and tech- nical problems. What can be said is that more participatory local solutions will be attempted for the myriad of challenges entailed in assessing and collecting taxes; improving transparency and justice in the allocation of capital funding; and involving com- munities in such local services as health care, primary education, and even the construction of basic infrastructure. As these cities grow (albeit more slowly than smaller, intermediate-sized cities), their professional staff will become more pro- ficient and their elected officials more experienced, and the new powers and re- sponsibilities devolved to them will be consolidated. One can only hope that all cities will see the value of democracy, the rule of law, and honest and effi- cient local administration, as well as the importance of supporting strong civil society organizations. However they are eventually constructed, cities must draw on the legitimacy of an emerging localism and commitment to democracy, and they must find nontechnical means to engage the imagination and energies of civil society. 4Virtually all these arguments came into play in Toronto when, in 1997, the provincial government of Ontario removed the well-known and much-admired upper-tier government and its six second-tier governments and replaced them with a single municipal government covering one-half of the greater Toronto area.
THE CHALLENGE OF URBAN GOVERNANCE CONCLUSIONS AND RECOMMENDATIONS Conclusions 407 The emergence of large urban regions, spread over many jurisdictions, presents acute problems of governance. Partly as a response, important institutional re- forms have taken place over the last decade. One way of characterizing these trends is to see them as a movement from "local government" to "local gover- nance." The term "urban governance" implies a greater diversity in the organization of services, a greater variety of actors and stakeholders, and a greater flexibility in the relationship between municipalities and their citizens. These trends in institutional reform can be illustrated by the example of the Bangkok Metropolitan Region. From this analysis the question may be asked: Is there a preferred governance model that best fits such a mega-urban region? There are four types of mega-urban governance in actual use around the world: the frag- mented model, the mixed model, the centralized model, and the comprehensive governance model. The first and last are the most widespread. The comprehen- sive model exists to some degree in Cote d'Ivoire, is emerging in South Africa with the "unicity," and is in effect practiced (albeit without local structures of democratic political representation) in four of the largest Chinese cities. In dis- cussing some of these variations on an overall theme of mega-urban governance, we looked in this chapter at some of the key challenges faced by all large urban areas: the capacity dimension (involving services), the financial dimension (with a focus on the generation of local revenues), the diversity dimension (where is- sues of fragmentation and inequality are central), the security dimension, and the authority dimension (looking at the distribution and allocation of power). As a result of decentralization reforms in many countries, local governments have been given more functions, as well as after an initial time lag more power to raise revenues. The range of these reforms is truly vast. Unfortunately, com- parative statistics are very incomplete, but what figures we do have show that municipal expenditure levels are slowly rising as a proportion of national govern- ment expenditure, although the level is still low. Local tax collection and revenue generation may be as much a governance is- sue as a technical finance issue, if we take into account some examples from China and Brazil. In the case of China, reforms since the late 1970s have decentralized decision making from the center to localities and have allowed local governments to enjoy the benefits of substantial "off-budget" revenues. With citizens' involve- ment, these revenues are a major source of infrastructural investment. In Brazil, many municipalities have adopted various versions of a "participatory budgeting" system, whereby cities are divided into regions or districts, and citizens in those areas participate in a process of determining the shape of the investment budget for each financial year. Security is an increasingly important dimension of urban governance. During the 1990s and beyond, urban security has become one of the dominant issues. As
408 CITIES TRANSFORMED for policy responses to violence, the emphasis in a myriad of programs throughout the world focuses on strengthening local social capital. Violence erodes social capital when it reduces trust and cooperation within formal and informal social organizations; conversely, when social capital is strong, development processes are supported. Since the late 1980s, major decentralization initiatives that have strengthened municipal governance have taken place in a large number of countries. Exam- ples such as Abidjan in Cote d'Ivoire, Mexico City, and some of the larger coastal cities in China illustrate a wide variety of institutional reforms that have been tak- ing place at the local level around the developing world. Common to most of these reforms are greater involvement of NGOs and community groups, greater openness and accountability, and the disposition of more legal and financial re- sponsibility for urban affairs at the local rather than at the state or national level. The issue of coordination of multiple jurisdictions is central to the governance challenge in large city-regions. The literature on metropolitan governance fo- cuses on the fragmented situation typically found in almost all North American metropolitan areas. But developing countries exhibit high levels of jurisdictional incoherence as well. Here there are two competing arguments: one suggesting that a wide variety of jurisdictions and local agencies can be beneficial to citi- zens since it permits local choice in services and living environments; and an- other promoting more coherent metropolitan-wide administrative and political solutions. These correspond, respectively, to the "fragmented" model and the "comprehensive" model of governance. As yet, no single model holds sway around the world, but discussions and analyses of competing visions are taking place almost everywhere. Recommendations The field of comparative urban governance is relatively new. But to improve our ability to explain and analyze urban governance reforms and policy options, we recommend more comparative studies and better interdisciplinary tools. This will involve, at a minimum: · Paying more attention comparatively to the local dynamics of policy reform, both among cities within individual countries and across national bound- aries. As urban populations increase to staggering dimensions in some cities, and as decentralization devolves substantial powers to municipal gov- ernments in the areas of health care, education, social assistance, and urban development, we need to incorporate information on and analysis of local governance into our development and policy discussions in a much more central fashion. Municipal development is real development for a large part of the population and must be treated as such in the research literature. Just as we have good cross-national information on GDP, demographic change,
THE CHALLENGE OF URBAN GOVERNANCE 409 and even income inequality, we need to disaggregate as much of this infor- mation as possible to the urban level for comparative purposes. · Incorporating local governance into our models and analyses of urban change. Until now, research on urbanization has focused primarily on the disciplines (such as sociology and geography) that deal with social and physical change in the urban environment. This was understandable as long as local governments had few powers and little ability to effect changes in the urban situation. But issues of metropolitan reform are on the table in many countries and need to be addressed using analytical approaches that link politics, administrative reform, and the other social sciences in the same . . . alscusslon.