A New Approach for Identifying the Parameters of a Tax Capitalization Model

Abstract

This study improves upon past estimates of the tax capitalization rate through two empirical innovations. First, the under-identification problem inherent in all capitalization studies is resolved by obtaining rental-data based estimates of the values of housing services instead of financial-data based estimates of the net user cost. The resulting estimate of the rate of tax capitalization is likely to be more reliable than those obtained in previous studies. Second, the methodology and the unique data set utilized mitigate biases resulting from unobservable publicly provided goods and services or other amenities whose levels are correlated with property tax rates. The extent of tax capitalization estimated in the current study is greater than estimates reported in most recent studies, and is statistically indistinguishable from full capitalization. The study also provides insights on the link between variations in the net user cost of housing and in several property characteristics.

JEL Classification: H71, R51