Keywords

JEL Classifications

In his famous paper, Charles Tiebout (1956) argued that there were realistic conditions under which local public goods would be provided efficiently; an efficient allocation would emerge as each household selected the community providing the public good levels most closely aligned to its preferences. This has come to be known as the ‘Tiebout hypothesis’ and the related Tiebout community-choice mechanism has been dubbed ‘voting with your feet’. This article focuses on research linking the Tiebout hypothesis to local political jurisdictions.

Oates (1969) gave Tiebout’s hypothesis empirical content. He reasoned that if households selected among communities in the way Tiebout conjectured, ‘capitalization’ should result. That is, ceteribus paribus, housing prices should be higher in communities with high levels of public good provision and lower in communities with high tax rates. Oates tested and found support for these predictions using data for municipalities in New Jersey. His paper led to an explosion of research on capitalization and launched research into a variety of related aspects of the Tiebout hypothesis.

Choice Within Jurisdictions

Tiebout was largely silent about how communities would settle on their levels of public good provision, though he emphasized parallels with market provision. Study of this market-based approach is largely the domain of the theory of clubs. Another approach, initiated by Barr and Davis (1966), focuses on collective choice within communities. In the terms of Albert Hirschman (1970), Tiebout emphasized ‘exit’ while Barr and Davis emphasized ‘voice’. Much research followed. Bergstrom and Goodman (1973) formalized estimation of demand for local public goods. Romer and Rosenthal (1979) investigated the role of agenda setters. Micro-level estimation of demand for local public goods was undertaken by Bergstrom et al. (1982).

Goldstein and Pauly (1981) observed that neglect of self-selection of households into communities would potentially bias estimates of demands for local public goods. Work followed linking intra-community choice and inter-community choice, beginning with Rubinfeld et al. (1987), and continuing with research on models of multi-community equilibrium.

Equilibrium Among Jurisdictions

Not surprisingly, households with higher incomes tend to prefer communities with high levels of local public goods. It is natural to ask whether income stratification can be sustained in equilibrium. Building on the work of Ellickson (1971), Westhoff (1977) proves existence of equilibrium in a model with sorting across communities and voting within communities. Westhoff’s model is extended to incorporate housing markets by Epple et al. (1984), who demonstrate that income-stratified equilibria can be sustained by differentials across communities in the price per unit housing of services. Fernandez and Rogerson (1998) add an important dynamic feature, with community education spending by each generation affecting incomes of the succeeding generation.

While households tend to sort by income across communities, there is much income variation within jurisdictions, even within small neighbourhoods (Hardman and Ioannides 2004). Several approaches seek to capture this intra-community heterogeneity as an outcome in multi-community equilibrium. One approach emphasizes heterogeneity in preferences as well as incomes. Structural estimation of multi-community equilibrium models embodying such heterogeneity is undertaken by Epple and Sieg (1999) and Epple et al. (2001). This framework is applied to study large-scale policy change by Sieg et al. (2004).

An alternative approach emphasizing heterogeneity and durability of housing is developed by Nechyba (1997). Nechyba has extended and applied this framework to study important policy issues, with particular emphasis on school choice and vouchers (Nechyba 2000). Structural estimation taking Nechyba’s model as a point of departure is undertaken by Ferreyra (2005) who extends the model to include heterogeneity in household tastes, including tastes for sectarian and non-sectarian schools.

Still another approach, by Bayer et al. (2004), permits detailed investigation of the way a household’s own demographic characteristics affect preferences with respect to the demographic composition of communities and the quality of local public goods. Bayer et al. (2005) apply this framework to estimate household preferences for community composition and to investigate the extent to which sorting within a metropolitan area is driven by preferences for education.

Residents of a jurisdiction may affect the public goods provided therein via ‘neighbourhood effects’ and ‘peer effects’. Peer effects are introduced into a multijurisdictional model by deBartolome (1990). While research on the Tiebout model emphasizes choice among municipalities in a metropolitan area, there is also population sorting within municipalities, especially central cities. Benabou (1996) and Durlauf (1996) study how peer effects influence such sorting, and the economic consequences of such sorting. Multi-community models increasingly emphasize peer effects (Nechyba 2000; Epple and Romano 2003; Bayer et al. 2004; Rothstein 2006; Sethi and Somanathan 2004; Ferreyra 2005).

Equity and Efficiency

Local governments impose many restrictions on land use (Fischel 1985). Hamilton (1975) emphasizes the potential efficiency-enhancing role of zoning. Other research investigates ‘fiscal zoning’, the allegation that jurisdictions use zoning to restrict entry by households who would contribute less in taxes than the cost of public services they would consume. Early contributions are in Mills and Oates (1975). Multi-community models with community residents choosing zoning by majority rule are developed by Fernandez and Rogerson (1997) and Calabrese et al. (2005). Computational results in the latter reveal that zoning can enhance efficiency, but also support critics who argue that fiscal zoning benefits wealthy households at the expense of poorer households. Henderson (1985) emphasizes the role of the private sector in community development. Henderson and Thisse (2001) focus on the role of developers in determining the character of the housing stock in communities. Glaeser and Gyourko (2002) conclude that land use restrictions play a major role in driving up housing prices in some areas of the United States, particularly California and some eastern cities.

The essence of the Tiebout hypothesis is that localities provide differing public good bundles to reflect variation within the population in tastes and incomes. Education is arguably the most important locally provided good, and efficiency arguments favouring decentralized provision in Tiebout equilibrium lie in uneasy juxtaposition with equity arguments favouring more centralized provision to increase equality of educational opportunity. US Courts have mandated intervention in many states to achieve greater equality of spending (Evans et al. 1998).There is growing emphasis in research (Duncombe and Yinger 1998) and the courts on policies designed to yield greater equality of educational outcomes. There is also increasing emphasis on incentive systems that might stimulate efficient provision of public education (Ladd 1996) and increasing recognition that interventions intended to achieve greater equity may affect both political support for public education and effectiveness of provision. Oates (2006, p. 42) puts the matter succinctly: ‘There seems to be an inevitable tension here.’ Policies promoting equity need to be designed to harness local incentives for effective provision while also recognizing their impact on political support for public provision.

See Also