The effect of property taxes on vacation home growth rates: Evidence from Michigan ¬リニ

a r t i c l e i n f o The Tiebout model assumes that individuals sort to the jurisdiction which best matches their fiscal preferences. However, there is a paucity of reliable estimates for the impact of tax changes on household mobility. We utilize a state mandated school finance reform and temporal differences in vacation home densities to provide a unique test of this fundamental Tiebout assumption. The results show that changes in property taxes explain a significant amount of the variation in vacation home growth; a 3–4 mil decrease in property tax rates is associated with an increase of approximately one vacation home per square kilometer. The ability of individuals to sort across locations and choose their preferred tax–benefit combination is an essential assumption of Tiebout (1956)'s work. Indeed, sorting has been a fundamental assumption in modern public finance 1 and leads to a simple empirical hypothesis: holding everything else constant, a decrease in property taxes will cause in-migration to a jurisdiction and a corresponding increase in the quantity of housing. In spite of this clear prediction, there have been few reliable empirical estimates of the effect of property tax changes on household mobility. This primarily is driven by empirical challenges that include the potentially endogenous relationship between population and tax levels, attenuation errors due to spatial and population aggregation, and difficulties in controlling for changes in public good levels associated with changes in tax levels. The dual relationship between population and taxes is a classic 'chicken-or-egg' problem: Do tax changes attract in-migrants to a jurisdiction or do new in-migrants bring with them preferences for changes in tax levels? 2 Thus, to correctly identify the effect of taxes on household mobility it is important that the determinant of the tax change under consideration is not affected by changing demographics. A second issue common in the literature concerns the impact of aggregating data to the state or county level. Aggregation introduces attenuation issues associated with taking demographic and tax averages across diverse populations and locations. 3 Further, high degrees of aggregation may mask the presence of an inelastic housing supply. In this case, shifts in housing demand due to property tax changes may not be solely observed through changes in housing quantities , but may also be capitalized into housing values. 4 A third issue is that tax levels are typically associated with spending local public services. In …