Comments

THE UNITED STATES has an enormous number of local governments. Even when the focus is limited to municipalities, as it is in this paper, the number is impressive. There were 19,372 municipalities in the United States in 1997.¹ Most large metropolitan areas in the United States are divided into a hundred or more distinct municipalities. Even metropolitan areas of moderate size typically have fifty or more distinct municipalities.

For much of the twentieth century, there appears to have been a relatively broad consensus reaching across disciplines that this multiplicity of local governments was undesirable—a manifestation of disorganization and a prescription for inefficiency in the provision of public services.² This consensus spawned a metropolitan government movement that met with little success and much resistance. At about the same time the political movement for metropolitan government was beginning to run out of steam, the intellectual underpinnings also came under challenge––thanks largely to the work of Tiebout, and Ostrom, Tiebout, and Warren.³ The trade-offs between the benefits of decentralized choice and the potential gains from coordination have been the subject of debate ever since.

The lack of popular support for consolidation of metropolitan governments is mirrored in data affirming the permanence of municipalities. Municipal boundaries, once drawn, are highly resistant to change. Vigdor reports that in Massachusetts, no new cities or towns have been created since 1920 and none deleted since 1938.⁴ In Pennsylvania between 1980 and 1990, there were only eighteen annexations or mergers, involving less than one square mile and fewer than 500 persons.⁵ Using data from the decennial boundary and annexation surveys of the 1970 and 1980 Census, Epple and Romer find that 98 percent of boundary changes are those in which an existing municipality adds previously unincorporated land.⁶ Annexation of one municipality by another and mergers of municipalities have become increasingly rare. Nationwide there were an average of fewer than ten such consolidations per year in the decade of the 1970s.

Strong advocates of metropolitan consolidation remain, but they acknowledge that success is unlikely.⁷ While efforts to promote metropolitan consolidation have largely disappeared, the outcomes that emerge under the current system of highly decentralized local government garner widespread criticism. Broadly, there are two issues. One centers on the distributional consequences of the tendency toward income stratification observed in U.S. metropolitan areas. One consequence is unequal quality of public services across municipalities, with quality inversely related to income. The other concern is that independently chosen tax and expenditures lead to large distortions at the metropolitan level, when central city governments engage in taxation for both redistribution and provision of public services. These issues are well articulated by Mieszkowski and Mills. They note the practical difficulties of maintaining a separation between the allocatable and distributive activities of governments. They point to the inefficiencies associated with redistributive taxation in the central city, and they offer the judgment that “tax and government service level considerations inhibit central city redevelopment.”⁸

Recent years have witnessed growth of a body of research in political economy seeking to characterize equilibrium when both the populations of local jurisdictions and their policies are endogenous.⁹ In these models, policy outcomes within jurisdictions are determined by “democratic federalism,” to use the terminology coined by Inman and Rubinfeld; policies are chosen by simple majority rule.¹⁰ Because of the well-known difficulties with characterizing equilibrium with multidimensional policy choices, each of these papers focuses on a unidimensional policy space.¹¹ Thus these models are unable to capture the tensions between public service provision and redistribution, and the associated impacts on housing markets, household location, and welfare that are at the center of the debate about the consequences of metropolitan fiscal decentralization.

Our contribution is to analyze metropolitan equilibrium when localities have multiple tax instruments and can engage in both public service and redistributive expenditures. Moreover, when voting, households consider the impact that these public policy choices have on housing prices and the population of the jurisdiction in which they reside. We do this by considering...