State and Local Government Utility Maximization According to Garp

Abstract

Many recent state and local expenditure determinant studies have modeled public choice as the result of the constrained maximization of a utility function. This model has been tested using particular utility functions, but rejection of constrained maximization under such a test may imply only the inconsistency of the data with the chosen functional form. A nonparametric test, based on the generalized axiom of revealed preference (GARP), exists that shows whether a price-quantity data set may be rationalized by the constrained maximization of any utility function. This test is applied to various state and local government cross-section and time series data sets. It appears that the constrained maximization model can explain the bulk of public price-quantity observations in each data set tested. The results tentatively suggest that time series data are more likely to be consistent with utility maximization than are cross-section data.