Abstract
There is a large degree of variation in the number of general-purpose local governments and school districts, per square mile and per person, among United States' metropolitan areas. Using data from 1982, this paper provides an empirical test of whether economic factors (especially variation in demand for local government services) partly account for these differences. The empirical findings show that after controlling for political, historic, and institutional factors, variations in the characteristics that affect demand for local government services do influence the number of local governments. This result is consistent with the hypothesis first put forth by Tiebout.