Share This
Artwork for Dillinger/Fay Article

Search Finance & Development

Advanced Search
About F&D


Back Issues

Write Us

Copyright Information

Use the free Adobe Acrobat Reader to view a pdf file of this article

Free Email Notification

Receive emails when we post new items of interest to you.

Subscribe or Modify your profile

Finance & Development
A quarterly magazine of the IMF
December 1999, Volume 36, Number 4

From Centralized to Decentralized Governance
William Dillinger and Marianne Fay

Decentralization can foster political stability and economic development—if transfers of resources and responsibilities are carefully coordinated and intergovernmental relationships are clearly defined.

As the twentieth century draws to a close, people around the world are clamoring for a greater voice in the way they are governed. Groups that have historically been denied power now demand it, and central governments are increasingly unable to resist their demands.

This trend has many causes. It has been attributed to the economic failure of the centralized, authoritarian state (with the consequent alienation of important support groups); the decline in the threat of war and external aggression in most of the world (with the consequent rejection of strong authoritarian government); and the emergence of educated, urban middle classes (with the consequent decline in traditional patron-client relationships between the government and the governed).

National governments have responded in several ways. Some have become more democratic. Twenty-five years ago, only one-third of the world's countries held competitive elections. Today, 60 percent do. Governments are also decentralizing—shifting responsibilities and resources to subnational units of government. Both measures provide a means of maintaining political stability and conceding political power within a formal, rule-bound decision-making system.

Although the overriding objective of decentralization may be political stability, the devolution of authority and certain functions to local governments also has implications for more traditional developmental goals. On the positive side, decentralization can improve the efficiency and responsiveness of the public sector by bringing decision making closer to citizens. On the negative side, decentralization raises the risk of macroeconomic instability. The issue is not whether governments should decentralize or not—this is dictated by political circumstances—but rather how to accommodate underlying political pressures so that the developmental potential of decentralization can be realized and the risks minimized.

Strategies for decentralization vary according to the circumstances of individual countries. Nevertheless, the experience of the past 15 years provides some lessons that are applicable everywhere. Perhaps the most important is that a system that is based on a coherent, explicit, and stable set of rules works better than one that is not. These rules need to encompass three aspects of the intergovernmental relationship: the division of national political power between national and subnational interests; the functions and resources assigned to subnational governments; and the electoral rules and other political institutions that bind local politicians to their constituents.

Balancing regional and national interests

The national government makes the rules under which subnational governments operate. The power of subnational interests in the national government therefore has a key bearing on how intergovernmental relationships are enforced and changed. Democracies have long struggled with the question of how to represent regional interests at the national level. On the one hand, giving too much power to regional interests makes it difficult to defend national interests against regional ones when the two are in conflict. On the other hand, too strong a central government will stifle the political openness that decentralization is intended to achieve.

At present, national governments tend to be more concerned with avoiding political instability than with encouraging greater openness. One way to address their concerns is to strengthen the executive branch in relation to the legislature. The powers granted to the executive branch, which include the power to rule by decree (to promulgate laws without legislative approval) and to dismiss the legislature and call for new elections, vary from country to country. Electoral rules also influence executive power. Systems that encourage the proliferation of political parties may weaken the power of the executive branch of government by forcing candidates to form coalitions to win office. Party rules also have an important impact on the balance between national and regional interests. Strong party discipline and the power to select candidates for the legislature can strengthen the hand of a national executive even in a system that is constitutionally highly decentralized.

Subnational functions and resources

The second major component of the intergovernmental relationship concerns the functions and resources assigned to subnational governments.

Functions. The traditional "fiscal federalist" approach to decentralization calls for a subnational government structure with several tiers, each delivering services that benefit those residing in the subnational government's jurisdiction. This model, while useful, has some limitations. As a practical matter, it is often difficult to define the scope of benefits of a specific service and to determine which specific jurisdiction will reap these benefits. Education, for example, can be considered to have extremely localized benefits, given its impact on the future incomes of pupils. But, as a vehicle for poverty alleviation and political acculturation, it also has national benefits. Administrative constraints are another limitation. Public services are subject to economies of scale, particularly in the use of specialized personnel. Local governments may therefore find it advantageous to contract with larger, higher-level authorities to provide such services. By the same token, a central government may find it cost-effective to contract with local governments to carry out field-office functions on its behalf.

A successful division of functions is therefore often characterized as a complicated set of principal-agent relationships, in which subnational governments act both as agents of higher levels of government and as principals (or, more precisely, as agents of their own constituents) in the delivery of local services. To accomplish this, individual services are often unbundled, with each level of government performing the role that best reflects its interest or comparative advantage. For example, because education is used as a tool of poverty alleviation, national governments often provide financial support to ensure a minimum level of education in all jurisdictions, while the day-to-day management responsibilities are assigned to local governments or school boards.

What is important is that these relationships—while complex—be clearly defined and stable. Otherwise, the division of functions may become the object of intergovernmental gamesmanship. In South Africa, for example, the central government and the provinces have joint responsibilities for health and education, but the exact responsibilities of each are not defined. The result is that provinces receive transfers to fund health and education but use them for other purposes, knowing full well that the central government will intervene to provide the needed services.

Resources. The guiding principle of revenue assignment is straightforward: finance should follow function. This is because subnational governments need resources commensurate with their responsibilities. In recent years, many African countries facing fiscal crises have devolved a broad range of government services to subnational governments without providing the necessary revenues. Not surprisingly, the quality of the decentralized services has declined sharply. The opposite occurred in much of Latin America: governments decentralized revenues without offloading corresponding responsibilities. In Colombia, central transfers to municipalities increased by 60 percent without a matching increase in responsibilities.

The "finance follows function" principle also applies to the choice of specific revenue instruments assigned to subnational governments: the mix of user charges, taxes, and transfers assigned to each level. Each revenue instrument has a different impact on consumer behavior and a different pattern of incidence. User charges, such as bus fares or charges for metered water, can be used to ration efficiently the consumption of goods whose benefits are largely private. Local "benefit taxes" can perform a similar role in pricing services whose benefits—while not private—are largely confined to local taxpayers. Intergovernmental transfers are needed in the variety of circumstances in which subnational governments operate—explicitly or implicitly—as agents of the national government. Because social safety nets are often considered a national responsibility, for example, intergovernmental transfers for education, health, and income support are common in both developed and developing countries.

The objectives of a given transfer and the administrative constraints under which it operates should determine how it is designed. These vary. What does not vary is the need to make transfers simple, transparent, and predictable, to help eliminate the uncertainty and bargaining that often plague intergovernmental fiscal relations.

Borrowing is an increasingly controversial part of the intergovernmental relationship. In countries with weak domestic capital markets, central governments have been the primary source of local credit, often providing it through specialized lending institutions. The financial record of these institutions is mixed. They have been successful in Europe, less so in developing countries. In countries with more developed domestic capital markets or better access to international markets, subnational governments are increasingly turning to the private sector for credit. In principle, such loans should be straightforward transactions between subnational governments and private lenders. But because subnational debt is often implicitly guaranteed by central governments, banks may be willing to lend to uncreditworthy jurisdictions. National governments have addressed this problem by attempting to regulate the level of local borrowing, but recent experience suggests this is not sufficient. What is critical is that central governments establish a reputation for maintaining a hard line against subnational debt relief. By alerting potential lenders that aid will not be forthcoming, governments can ensure that subnational borrowing is confined to levels that subnational governments themselves are able and willing to repay.


The third part of the intergovernmental relationship concerns the rules governing relations between local officials and their constituents. If decentralization fails to increase local influence over the public sector, the principal benefits of decentralization will be lost. One of the key determinants of local accountability is the system for electing local officials—governors, mayors, and members of subnational legislatures. Elections by district (rather than at large) can increase local influence by reducing the costs of running for office. Local control (as opposed to national party control) over the choice of candidates for local office may also increase accountability. But elections, in and of themselves, are not sufficient. The responsiveness of local officials to their constituents is also influenced by civic traditions and the strength and activism of community organizations.

Managing decentralization

Decentralization is a work in progress. A number of countries are experimenting with different approaches, and evidence on outcomes is still limited. Nonetheless, some lessons have emerged. The most compelling is the need to synchronize the elements of reform. Revenues must be decentralized in conjunction with the responsibilities they are meant to finance. Arrangements must be made for the transfer of central government staff and assets. The political impetus behind decentralization prompts central governments to take dramatic steps first—granting local elections and increasing revenue sharing. But what is critical—and takes longer to accomplish—is the working through of the new functional relationships between different levels of government. Administrative relationships between different units of the central government must be converted to arm's-length relationships between different levels of government.

Recent experience has also shown the need for central governments to demonstrate commitment to the new rules of intergovernmental relationships early in the game. Precedents matter. They affect expectations. Imposing hard budget constraints on subnational governments is especially important: central governments must stand firm against ad hoc transfers or debt relief so as to force subnational governments to live within their resources.

Strategies aimed at stopping decentralization are unlikely to succeed. The pressures to decentralize are beyond government control. In recent years, they have led to a broader distribution of political power in Latin America, Eastern Europe, and parts of Africa and East Asia. In the early twenty-first century, the same pressures are likely to emerge in the rest of Africa and East Asia as well as in South Asia. Rather than resisting these pressures, countries in these regions should learn from countries that have gone before.

This article is based on Chapter 5 of the World Bank's World Development Report: 1999/2000: Entering the 21st Century (New York: Oxford University Press for the World Bank).

William Dillinger was a member of the team that produced the World Bank's World Development Report 1999/2000.

Marianne Fay was a member of the team that produced the World Bank's World Development Report 1999/2000.