The Increasing Income Elasticity of Housing Prices

Housing prices can be decomposed into two parts: the value of physical structure and the value of land upon which the structure stands. Extant empirical evidence indicates that real land values tend to be volatile and respond significantly to changes in housing demand. In contrast, real structure values generally are stable over the long run. These observations results in a prediction that, in a growing economy, the share of land value of total housing prices increases as time passes, because of which the income elasticity of housing prices trends upwards over time. We apply the long historical time series for housing prices and GDP provided by Knoll et al. (2017) and Jorda et al. macrohistory database to investigate the long-run trends in the income elasticity of housing prices at the country-level for a number of developed countries. We estimate time-variation in the elasticity with FMOLS using rolling estimation windows. The empirical results are in line with our a priori expectations: For all the 11 countries for which sufficiently long time series are available, the sensitivity of housing prices w.r.t. GDP has exhibited a significant growth trend over the long run. The findings suggest that housing prices become more and more volatile as the economy grows – unless the real economy becomes less cyclical.