House price convergence: Evidence from US state and metropolitan area panels

Abstract

We study house price convergence in panels of US states and metropolitan areas. Our analysis is centered on three issues. First, we test whether the US house prices are converging over time using log t convergence test by Phillips and Sul (2007). The test results suggest that there is little evidence of overall convergence. The second issue is to investigate the possibility of a convergence club where the cross-sectional dispersion of house prices of the club members decreases over time. We utilize a clustering algorithm and the results support that there is strong evidence of multiple convergence clubs. Finally, we examine the general characteristics of the various convergence and divergence subgroups as well as some important driving forces of convergence clubs. We find that housing supply regulation together with climate are important determinants of convergence club membership.