An application of a test for allocative efficiency in the local public sector

Abstract

This paper extends a test developed by Brueckner (1982) for Pareto-efficiency in the allocation of local public goods. The test is based on the result that property values are maximized when public goods are provided efficiently. By using cross-section data, the test is designed to locate local government expenditures on a total property value hypersurface and hence deduced if local public goods are provided in a Pareto-efficient manner. This paper attemps to generalize the test by using local spending data aggregated up to the country level. A country-level test is widely applicable and it accounts for public good spillover effects between jurisdictions. The empirical results suggest that local officials may not account for the presence of spillovers, leading to underprovision of certain public goods. The empirical results also suggest that smaller, more numerous local governments positively influence property values, other things equal.