Property Tax Distortions as an Explanation of the Flypaper Effect

Leonard Wade Locke


This study investigates how local government behaviour is altered when property taxes are permitted to distort both the median voter's housing consumption decision and the spatial location of firms. It is demonstrated that optimizing politicians, in the presence of either or both of these property tax distortions, will, contrary to the predictions of standard median voter models, respond asymmetrically to increases in local resources and increases in intergovernmental lump--sum aid. Moreover, the model predicts that property tax distortions will cause the local politicians to allocate more of an increase in aid to public sector expenditure than it would for an equivalent increase in income. That is, the theoretical model offers an explanation of the so-called "flypaper effect", which is contingent upon neither the coercive power of the bureaucracy nor the mistakes of the pivotal voter.

In addition to developing an optimizing model which generates the flypaper effect, the predictions of the model are tested by applying White's (1980) least-squares-covariance-matrix estimator to the per capita expenditure equation derived from the model. Instrumental variable estimation is also utilize to correct for simultaneous equation bias that might result from the property tax distortion variable. The bins is due to the fact that the ax distortion term is a function of the property tax rate which, in turn, is endogenous to the local government's optimization problem.

The key finding of the empirical test is that the property tax distortion variable, as predicted by the model, is both negative and significant. This result is particularly encouraging and provides support for the property tax explanation of the flypaper effect.