Abstract

The aim of this article is to develop a methodology for a spatial or regional cost index of housing that considers spatial differentials across regions. Using micro data from the Chilean survey CASEN 2003, it is showed that a spatial or regional housing price index based on the weighted mean or the estimators of hedonic price equations might lead to biased results due to spatial heterogeneity. The potential bias is reduced by matching the houses in a region with a clone in the Metropolitan Region, according to own and neighbors' characteristics using propensity scores. As a result a very different pattern of spatial cost of housing arises. Finally, using a Fisher ideal price index, the paper proposes a spatial or regional cost index of housing that shows price differences for homogeneous houses across regions.