Match quality in housing transactions. What can we learn from comparing buyers and sellers?

Abstract

Match quality, the part of housing value to the buyer which is unique for each buyer-house match, is important in several housing market matching models, but measuring it is difficult for an econ-ometrician.

I suggest that similarity between buyers and sellers (at the time they bought) may be used to measure match quality. Successive owners of houses should share characteristics if observable characteristics of a buyer are correlated with the buyer's preferences for housing. A buyer could expect to have a high match quality if similar to the seller. I use a simple matching model to show this mechanism. I test this prediction using unique data with information on buyers and sellers (at the time they bought), and show that their similarity can be used as a proxy for match quality. Buy-ers who resemble sellers are paying more, also when a large number of observable housing characteristics are controlled for.

Supplementary analyses strengthen my claim that the distance between seller and buyer can be used as a proxy for match quality. Matches with low distance lead to slightly reduced hazard rate of reselling the house, and an increased probability of having children, both of which would be expected in a high quality match.