How Effective are Policy Interventions in a Spatially-Embedded International Real Estate Market?

Abstract

We introduce the role of ‘space’ in analyzing the effect of macroeconomic policy interventions on cross-country housing price movements. We build an empirically testable analytical model and test our theoretical predictions for a panel of European countries over the period 1985–2015. Our aim is to demonstrate that while macroeconomic policy exerts a significant impact on international housing markets, the magnitudes of such impacts may be overestimated in the absence of spatial frictions. To test our hypotheses, we employ a spatial dynamic panel method and quantify intra- and inter-country differences of the effects of macroeconomic policy interventions on spatially interdependent housing markets. Endogeneity issues arise in our estimation, which we ameliorate by employing the spatial Durbin model for panel data. Following this approach, we include spatial, temporal and spatio-temporal lags for the identification purpose. We show that a spatially-embedded model produces relatively smaller and correct signs for macroeconomic variables in contrast to the traditional non-spatial model. It is concluded that empirical estimates from the traditional model are consistently over-estimated. These have significant policy implications for the exact role of macroeconomic interventions in explaining housing price movements. A battery of robustness tests and evaluations of predictive performance confirm our results.