| Tiebout (1956) proposed that households sort themselves between communities based on their search for the desired level of the public services. This thesis studies different implications of the Tiebout hypothesis and the effect of the Tiebout sorting on the provision of public services by local governments.;There are three chapters in this dissertation, each chapter is relatively self-contained and can be read independently. The first chapter investigates the relationship between intra-metropolitan public service-bundle differentiation and the heterogeneity of a metropolitan population in observed household characteristics. In the spirit of the Tiebout hypothesis, local communities should supply more differentiated service bundles in metropolitan areas where population exhibits greater heterogeneity with respect to the demand for local public goods. The findings indicate that municipalities indeed supply more diverse set of fiscal bundles in metropolitan areas with more heterogeneous population.;The second chapter investigates the relationship between within-community heterogeneity in observed household characteristics and the number of local governments in a metropolitan area. The contribution is to note that there are factors, such as geographic variations in housing prices, that cause households to stratify geographically and are unrelated to local public service levels. This stratification, referred to as "statistical sorting", implies that communities should again be more homogeneous if they contain a smaller proportion of the metropolitan population. Separate controls for the number of governments and the number of governments per capita distinguish between Tiebout and statistical sorting. They demonstrate that both types occur.;The third chapter estimates household demands for local public services. It derives the demand equation for local public expenditures, recognizing that community socio-economic characteristics can affect the cost of service provision, and estimates it using the 1990 Public Use Microdata Sample. Instrumental variables correct for the possible endogeneity of socio-economic characteristics, as implied by the Tiebout hypothesis. It demonstrates that socio-economic variables affect costs, and therefore household demands for local public goods. |
