Residential Property Tax Capitalization: Consistent Estimates Using Micro-Level Data

Abstract

Previous research on residential property tax capitalization has been dominated by models estimated with aggregate data. The property tax burden measure used in many of these aggregate studies is the effective tax rate, which accounts for differences in assessment ratios and nominal tax rates among taxing jurisdictions (Oates 1969; McMillan and Carlson 1977; Pollakowski 1973; Hyman and Pasour 1973). Other studies have touched on the advantages of individual property or micro-level data and have estimated residential property tax capitalization models based on such data (King 1973, p. 88; Edelstein 1974). These studies have not generally used the effective tax rate as the tax burden measure.1 A micro-level study that did use the effective tax rate measure of tax burden reported biased estimates of population parameters (Wales and Wiens 1974). In this article, we present a linear capitalization model for micro-level residential property data using the effective tax rate as the measure of tax burden, but estimated using an approach that yields estimates with desirable sample properties. A general specification of the model is: