|by David Luberoff|
To do so, he uses Zillow data on housing costs in over 11,000 ZIP Codes going back to the mid-1990s. While these data don't cover the entire country, they do cover more than half of the nation's residents. Hipsman focuses most closely on house prices in the more than 500 ZIP Codes that cover areas on the borders between two states. Using the TAXSIM model developed at NBER, Hipsman considers, for each tax year, the total income tax bill (federal plus state) that the same household would face were they to live in another state. Analyzing relative changes in these tax bills over time allows him to estimate whether and how changes in state tax burdens affected home values.
At first glance, the data seem to indicate that notable changes in state tax rates (or differences in bordering states' tax rates) could have dramatic impacts on home values. One of Hipsman's models, for example, indicates that a one percent drop in taxes might cause as much as a five percent increase in home values. However, Hipsman cautions against making too much of that finding. "Ultimately," he writes, "the evidence is inconclusive; standard errors are large, and different specifications lead to different conclusions."
This inconclusiveness, he adds, shows that while border-pair studies, such as his analyses, can offer important insights about policies governing taxes and spending, the results of those findings should be carefully tested before they are used for policymaking. "Obtaining a good estimate" of how changes in taxes affect home values, he concludes, "is important of further study."