Abstract
Based on the theory of bureaucracy, in contrast to the median voter model, expenditure functions are estimated for various municipal services provided by some SMSA cities. Various econometric measures and diagnostics are applied to the sample data to test for the incidence of multicollinearity, heteroscedasticity, and influential observations. The test results indicate the absence of any econometric problems and the empirical results can be taken as statistically meaningful and to a large extent precise. Over the range of SMSA cities included in the estimated model, the presented econometric evidence points out that in these communities the municipal services are private in nature and there exist no economies of scale in the provision of these services. Although these services reveal 'no pure publicness' or 'quasi publicness' they are provided by local governments through a collective choice mechanism, perhaps due to the inherent problems associated with their provision, such as the cities' lack of knowledge of individuals' preferences and tastes and the anticipated potential free rider problem.