Journal Issue

Funding the Metropolitan Areas of South Africa

International Monetary Fund. External Relations Dept.
Published Date:
January 1995
  • ShareShare
Show Summary Details

SOUTH AFRICA’S major cities, long divided along racial lines, are being restructured as the country moves toward democracy. How and where will the boundaries of the relevant municipalities be drawn, and how will their governments be financed? This article proposes a two-tier metropolitan structure that would, the author suggests, maximize both the autonomy and the fiscal integrity of municipal governments.

The economic destinies of South Africa and its metropolitan areas are intertwined. Overall, the urban sector accounts for about 65 percent of the nation’s population and more than 80 percent of its GDP. Nowhere is the concentration of people and economic activity more visible than in its major metropolitan areas—the Johannesburg and Pretoria complex, Cape Town, Durban, and Port Elizabeth—which represent almost 45 percent of the nation’s population and an even higher share of its GDP. Designing a comprehensive urban strategy is, therefore, an important national priority for South Africa. Inevitably, the framework for financing the metropolitan areas is central to such a strategy. The challenge is to design an urban fiscal strategy which will underwrite the political unification of the country’s racially divided cities; contribute to the national goals of poverty alleviation and redistribution; and, finally, sustain the economic bases of the cities.

Apartheid and local government

Apartheid reserved 87 percent of the country’s land for the whites, who currently constitute only 13 percent of South Africa’s population. The black community was relegated to ten artificial jurisdictions called “homelands.” Initially, influx control laws regulated the migration of blacks from the homelands into the white urban areas; such migration was conditional upon employment in the formal sector. Complementing the influx regulations were strict residential laws restricting blacks to “dormitory townships” at the urban fringe—often one to two hours away from the centers of employment.

By the mid-1980s, the metropolitan areas were divided, along racial lines, into black and white local authorities (BLAs and WLAs). (Local jurisdictions were also created for the Indian and colored (mixed-race) population. These will be amalgamated into the metropolitan structure described later on in this article.)

The BLAs and the WLAs differ sharply from each other in terms of their fiscal autonomy, provision of basic urban services, and governance.

Financing of BLAs. The BLAs are dependent on fiscal transfers, primarily from the central government, for maintaining the operations of the townships. These transfers account for more than 80 percent of the revenues of the BLAs and, since the early 1980s, have taken on the character of deficit-financing grants.

Several factors have resulted in this state of affairs. In general, racially based zoning policies designed to regulate the location of households and firms in the urban areas have forced the concentration of lower-income households, primarily in the townships, while preventing the growth of formal manufacturing and commercial employment in the BLAs. In addition, restrictions on property ownership have left the BLAs without a property tax system. In the long run, this has also prevented blacks from using land and housing as a form of collateral to generate a growing economic and taxable income base. Consequently, user charges and rents from government-owned housing were expected to be the main sources of revenue for black local authorities. Rent and service-charge boycotts aimed at dismantling apartheid, however, crippled the ability of BLAs to access even this source of limited revenue.

Financing of WLAs. The fiscal story of WLAs is in sharp contrast to that of BLAs. For example, in the largest metropolitan area in South Africa, Central Witwatersrand, grants from other tiers of governments account for approximately 3 percent of the total income of WLAs, compared with 95 percent for BLAs in the same area. Property taxes; “hidden” surcharges which implicitly tax the consumption of electricity, water, and other services; and interest income provide the main sources of income for WLAs.

To ensure their access to revenues from trading accounts, local authorities were given “monopoly” rights to distribute services such as water and electricity and impose implicit surcharges on their consumption. For example, in 1990 and 1992, the surcharge on water and electricity in the WLAs in Central Witwatersrand has averaged 11 percent and 20 percent, respectively. These revenues have generally been used to cross-subsidize other services and reduce residential property taxes in WLAs through a system of rebates. By law, the rebates cannot be applied to taxes on businesses or commercial property in WLAs or any taxes in BLAs. In addition, given that most formal sector jobs and retail stores were located in WLAs, the potential for exporting part of this tax burden from WLAs to BLAs was high. Ultimately, the system provided adequate revenues for the WLAs and was well adapted to the needs of municipalities in which equity played no role, and expenditure responsibilities were limited to municipal services such as water and electricity.

Regional Service Councils. In response to increasing fiscal transfers from the center, and to rectify the inequities of the local government tax system, the central government created, in the early 1980s, Regional Service Councils (RSCs) with jurisdiction over groups of BLAs and WLAs. Among other functions, the RSCs were responsible for funding capital investments in the BLAs, using turnover and payroll taxes levied at very low rates. In several regions, even these resources came to be used to finance the operating expenditures of BLAs. In effect, then, central and RSC transfers were the fiscal underpinnings of South Africa’s racially based local government system.

Access to urban services. The expenditure on municipal services in WLAs not only exceeds the national average, it surpasses even the standard of delivery in countries with a higher GDP than South Africa. (See chart.) While equalization of standards between BLAs and WLAs may be difficult to achieve, even the simple upgrading of the municipal services in black areas will impose a substantial fiscal shock on the revenue base of local authorities. For example, in the townships of the four major metropolitan areas, the backlog in capital investment on municipal infrastructure—water and electricity distribution, roads, drainage, sanitation services, and solid-waste removal—is between 4 and 8 percent of one year’s national GDP, depending on the standard of services to be delivered. The costs of carrying out other local functions are not included in these estimates.

Selected major metropolitan areas: per capita public expenditures on residential infrastructure, 19901

Source: S. Mayo, Housing Indicators.

1 Residential infrastructure includes roads, drainage, water and electricity distribution, sanitation, and garbage collection.

Governance. Accountability of white local governments to their constituencies is ensured through both the electoral and local tax systems. Residents of WLAs have full political representation, and the cost of delivering services is partially passed onto the beneficiaries through increases in property taxes and user charges. In the BLAs, the local councils, often perceived by blacks to be part of the apartheid system, were frequently accountable to no one. In addition, the easy availability of central fiscal transfers has contributed further to the lack of local accountability.

The reform of the local authority system will, therefore, need to achieve several objectives. First, there is the political imperative of abolishing the racial basis of local governments by amalgamating the BLAs and WLAs. Second, the black community’s access to basic urban services needs to be increased substantially and immediately. Finally, an accountable local government system will need to be established.

Designing a fiscal strategy

Several important factors—the current fiscal system used by WLAs and BLAs, the location of urban households and businesses, and the national policy on fiscal decentralization—will influence the design of an urban fiscal strategy. Their implications are several and often contradictory.

Current system of BLAs and WLAs. The fiscal position of WLAs suggests a potential for horizontal financing of poorer communities within each urban area. With access to capital markets, and if investments were phased over a five-to-eight-year period, WLAs would be in a position to fund the capital costs of upgrading municipal infrastructure in existing townships to a significantly higher standard. Such a policy would require a pooling of fiscal resources by WLAs within each metropolitan area and relatively small increases in property taxes and existing surcharges and/or reductions in expenditures currently devoted to residential infrastructure in the white cities. The administrative and technical strengths of the WLAs also suggest that even if the backlog of projects in the townships were not funded at the local level, the delivery of the services could certainly be handled locally.

In the short-to-medium run, it will be difficult for the townships to rely on the traditional sources of local revenues—property taxes and user charges. Instead, other fiscal instruments—such as the turnover taxes of the RSCs or the value-added tax (VAT) being levied by the central government—might be needed to generate revenues from the black community. The broad base of the VAT and the inefficiencies of the turnover tax suggest that the VAT would be the better instrument to use. Its implementation would, however, make local authorities more dependent on upper-tier governments.

In addition, unless the efficiency (and accountability) of the delivery system in the townships is increased, existing operations and maintenance costs may still have to be covered through central transfers. The potential for horizontal financing of capital investments, therefore, does not automatically obviate the need for these transfers. Indeed, a simultaneous reduction of central support and implementation of a public investment program for municipal services may be difficult to sustain using the joint BLA/WLA local revenue base.

Finally, the culture of nonpayment that has emerged under apartheid suggests that only local governments that are viewed as legitimate and are able to deliver services rapidly can hope to implement user charges and other beneficiary taxes. Establishing a representative local government system is one of the more critical elements in creating a sustainable fiscal strategy.

Locational dynamics. With the pressures of urbanization and the formal removal of apartheid laws, the allocation of population and business activities in urban areas has been in a state of flux. In recent years there has been a rapid decentralization of employment, from the core white cities into surrounding white suburbs. In addition, in some of the urban areas, notably Central Witwatersrand, there has been an outflow of white residents into the surrounding white municipalities and an inflow of black residents from the adjacent townships and homelands. Nevertheless, a substantial portion of the black community is expected to remain in the townships in the foreseeable future while apartheid’s legacy of better facilities in WLAs continues to “lock in” industries and businesses, primarily in the white municipalities.

In this context, urban fiscal strategy should not further distort the economic choice of households and firms on where to locate in the urban sector. On the one hand, a central tax is not likely to create incentives for relocating within urban boundaries. Region-specific turnover taxes or local property taxes may, on the other hand, result in locational decisions being made based on fiscal, rather than economic, criteria. In addition, the mobility of factors of production at the local level places a strong constraint on the ability of cities to implement redistributional policies. Not only can households and businesses move outside a city’s boundaries but, if the tax burdens are sufficiently different, the movement may be from one city to another. Yet, given that the formal sector continues to persist predominantly in white areas, any fiscal strategy will have to devise a process for transferring fiscal resources back to the black communities.

Fiscal decentralization. Political change in South Africa may also lead to the creation of a more decentralized governmental system. Currently, South Africa is one of the most centralized countries in the world—80 percent of total government revenues accrue to the central government, which is also responsible for 70 percent of all government expenditures.

“The challenge is to design an urban fiscal strategy which will underwrite the political unification of the country’s racially divided cities; contribute to the national goals of poverty alleviation and redistribution; and, finally, sustain the economic bases of the cities.”

The current debate on decentralization has been focused on the new provinces. How the new intergovernmental system will affect the finances of the metropolitan areas has not been adequately assessed. The existence of this vacuum is surprising, especially since the design of the intergovernmental system may have several implications for the cities of South Africa. First, decisions made at the national level may change the very definition of local goods. For example, different aspects of health and education may be devolved to cities at precisely the time the process of amalgamation and the financing of the backlog of projects in urban services are about to begin. Second, the new provinces may acquire the right to regulate the level of fiscal transfers to the cities within their jurisdictions. Third, the choice of fiscal instruments available to the provinces may, in turn, affect the choice available to cities. All of these uncertainties suggest that there is some risk of a mismatch between available revenues and expenditure responsibilities at the local level. Any future fiscal strategy in South Africa should be designed to avoid such an outcome.

The discussion so far suggests several, often contradictory, implications for the design of an urban fiscal strategy. The nature of the problem—broadening access to local services—and the fiscal and administrative strength of the WLAs seem to imply that the local fiscal base should be used to fund the backlog. Using local financing would also contribute to greater local autonomy and accountability. In addition, given the backlogs in health and education, which are central government responsibilities, the availability of central transfers will be restricted. Nonetheless, limited access to revenues from the low-income households, the potential for local taxes to further distort the spatial structure of the cities, and the inevitable need to reduce poverty and achieve redistributional objectives suggest a strong role for upper-tier funding. Any successful strategy to fund the cities of South Africa must therefore strike a balance between these two approaches. Ultimately, such a balance will be determined by the way the boundaries of the local governments are drawn.

Reforming urban governments

In reforming local governments, policymakers will have to decide on new municipal boundaries and the organization of local authorities. In the context of South Africa’s large cities, a two-tier metropolitan system seems best suited to achieving the goals of the amalgamation process.

Redrawing boundaries. Two options concerning the structure and boundaries of local government are being considered. One is to merge WLAs and a few adjacent townships—a “twinning” program of sorts, with geographic proximity being the main determinant of these mergers. The second is to create some form of a metropolitan system with pooling of resources across several municipalities and townships. Ideally, economic factors such as labor market integration and efficient service boundaries would be the primary determinants of a metropolitan boundary.

The economic and fiscal implications of the two options are quite different. The drawing of narrow municipal boundaries, as proposed in the twinning option, would not provide a sustainable fiscal unit for funding the backlog of projects needed to facilitate delivery of urban services. The burden sharing among different WLAs would be highly unequal, and such policies might further distort the spatial structure of the cities as household and businesses sought to escape the incidence of higher taxes. In addition, narrow municipal boundaries would fail both to internalize the spillovers of costs and benefits that are common in an urban setting and to eliminate the exporting of tax burdens between WLAs and BLAs. Of course, central and regional fiscal revenues could be used to compensate for the unequal burden sharing or the spillovers if the twinning approach were adopted. Such policies would, however, perpetuate the dependence of local authorities on upper-tier governments and might result in reduced autonomy, transparency, and efficiency of lower-tier governments. They might also lead to the inefficient use of fiscal instruments to fund local services. For these reasons, this article assumes that a pooling of finances across a metropolitan boundary would best support the process of amalgamation of the cities and townships.

Delivery and finance. A metropolitan boundary, however, does not necessarily imply the need for a consolidated metropolitan government. Indeed, for South Africa, a two-tier metropolitan system—a metropolitan authority with municipalities under its jurisdiction—may be a preferable option. The metropolitan administration could play a fiscal role—along the lines of the St. Paul-Minneapolis model in the United States—by focusing exclusively on funding a minimum level of services across the municipalities. The municipalities, in turn, would be responsible for delivering urban services. This arrangement has the advantage of rapidly mobilizing the existing administrative capacity of the WLAs without creating a new tier of bureaucracy to deliver services.

In its fiscal role, the metropolitan authority could have access to three sources of revenue: (1) turnover and payroll taxes that are currently in the hands of RSCs; (2) a portion of the property taxes and surcharges collected by the municipalities; and (3) central transfers currently designated for the BLAs. In the absence of user charges and property taxes, the turnover and payroll taxes would provide a “second-best” instrument for generating revenues from the townships. The sharing of residential property taxes would constitute the redistributive aspect of this fiscal arrangement; access to nonresidential property taxes and surcharges would compensate for the lock-in effect that past spatial policies have had on the location of industries. Finally, maintaining the central transfers would assist the municipalities in their process of restructuring and strengthen the fiscal role of the metropolitan area. The municipalities, in turn, would retain a portion of the property taxes and surcharges on services with which they could respond to the preferences of their communities.

During the implementation of such a fiscal arrangement, policymakers will have to determine the relative weights to be assigned to each source of revenue. A greater use of turnover, payroll, and property taxes will push the system toward local financing, while reliance on central transfers will foster further dependence on upper-tier governments. A similar choice will have to be made with regard to the metropolitan area and the municipalities. The sharing arrangement set up between the two tiers will determine the extent to which the metropolitan tier will be able to undertake redistribution and the degree of independence municipalities will retain to meet the demands of their constituencies.

Over a longer horizon, new sources of revenues will be required for the metropolitan system. Turnover taxes are highly inefficient, and property taxes are not a buoyant source of revenue. In addition, central transfers have the potential to reduce the autonomy and accountability of lower-tier governments. An alternative revenue source is the tax on personal income levied as a surcharge on the national income tax. This would constitute a buoyant source of revenue, concentrated in the urban areas, and could be administered nationally as a surcharge system. Metropolitan areas could also have access to a variety of automotive and excise taxes. The municipalities would continue to receive a portion of the property tax.

Capital market considerations. While the need to implement redistribution at the local level and minimize locational distortions have been important reasons for advocating a metropolitan structure, capital market concerns may also influence the structure of local government. Financial institutions in South Africa have indicated their willingness to finance service providers, independent of local governments. It is perceived that service providers that are part of local authorities run the risk of having their revenues siphoned off to meet the general obligations of municipalities, leaving the former with insufficient revenues to meet their own debt obligations. To attract finance from the capital markets, metropolitan authorities might have to use legal measures to isolate the revenues base of certain services such as electricity or water. A preferable solution might be to privatize those services for which direct user charges could be implemented. Distributional concerns could then be addressed by requiring cross-subsidies within services or—a preferable alternative—by guaranteeing fiscal transfers from the metropolitan budgets, thus separating distributional from operating concerns. Fiscal revenues from these services could be preserved by converting the hidden surcharges into explicit excise taxes.

The implications of privatizing municipal services go beyond the gaining of access to domestic capital markets. Privatized services also have the advantage of being able to attract domestic and international equity investments. Compared with debt financing, equity investments minimize the impact on cash flows and make services more affordable, thus requiring government to supply less in transfers to meet distributional objectives. In addition, through privatization, local governments would, in fact, eliminate some of the more capital-intensive municipal services from their debt portfolios. This might alleviate the fears of central authorities about the moral-hazard implications of local government borrowing.

The alternative option generally advocated is to restrict local borrowing to central public financial intermediaries (PFIs), on the assumption that centralizing borrowing powers would eliminate the potential for the central government to inherit the debt burdens of local governments. This would not only remove any possibility of financial markets exerting discipline on local politicians but also raise the specter of political allocation of credit and off-budget financing through the PFIs.

Ultimately, a strict control of local government borrowing could undermine the efficiency of local authorities and the fiscal autonomy gained through a metropolitan structure. Reducing the need for issuing debt by privatizing some municipal services may therefore alleviate central government concerns and provide additional autonomy to local authorities. Fostering a bond insurance market would further strengthen the case for establishing a more active municipal bond market and facilitating fiscal decentralization.


The design of a fiscal strategy for the metropolitan areas of South Africa will be influenced by the political boundaries of future municipalities. Amalgamation based on the drawing of narrow boundaries across adjacent WLAs and BLAs would require upper-tier governments to provide fiscal transfers as the main source of revenue for the new municipalities. A metropolitan boundary, however, would enable local authorities to depend primarily on local resources and to exercise greater autonomy. In advocating a metropolitan boundary, this article proposes a two-tier metropolitan structure. Under this arrangement, the metropolitan authority would not inherit any expenditure responsibility but would focus solely on assisting municipalities to fund the delivery of services. Concerns about their access to capital markets may also influence the organization of local governments. For South Africa, privatized delivery systems for municipal services that are regulated by metropolitan governments are most likely to effectively address the efficiency and equity concerns of policymakers. Ultimately, the reorganization of urban governments in South Africa will determine whether the unfortunate legacies of apartheid can be overcome. At stake is the success of the final step in the country’s transition to democracy—the holding of local government elections later this year.

Other Resources Citing This Publication