The Effects of Property Taxes and Local Public Spending on Property Values: A Reply and Yet Further Results

Abstract

In addition to presenting some findings of his own, Henry Pollakowski contends that my earlier empirical study (1969) of local public budgets and property values suffers from misspecification on two different counts and from an improper use of two-state least-squares estimation. I want to consider each of these points in turn. First, Pollakowski is quite legitimately disturbed by the fact that the reduced-form equation I estimated contained a variable representing only one of the public services, namely primary and secondary education, that is provided by local governments in New Jersey (although it is the case that this accounts for roughly two-thirds of municipal spending among the communities in my sample).' The logic of a Tiebout-type model suggests that residents of a community should have some concern for the levels of other public services as well and that, as a result, these outputs too might influence local property values. In addition, levels of spending on other public functions surely do affect local tax rates so that we would expect differentials in such spending across communities to be correlated with local rates of property taxation. Therefore this omission will, as Pollakowski indicates, probably result in biased estimates of the coefficients of my earlier equation. To correct for this likely bias, it is necessary to reestirnate the equation incorporating variables representing levels of output of other public services. As those who have worked on these problems know, it is extremely difficult to develop reliable operational measures of levels of output of most public services. The only readily available measures typically relate to spending, although, as we are all aware, expenditures can be a very unsatisfactory proxy for the level of service actually provided.2 It may be, however, that across a group of residential municipalities in the same general section of a metropolitan area levels of expenditure may represent a reasonable measure of output.