Andreas Benedictow, Erling Røed Larsen and Dag Martin Sundelius have written a new paper on “The Housing Phillips Curve”.
Using a high-resolution housing transaction data set, we examine the relationship between price changes and the unsold rate in the Norwegian housing market within a Phillips curve framework. Compared to the classical labor market Phillips curve the housing Phillips curve displays a reversed causality, from price changes to the unsold rate. We propose a consistent model based on microfoundations in which households decide on the sequence of buying and selling, i.e. whether to hold two or zero houses in the transition period. Data on individual sellers and buyers support the hypothesis that the propensity to hold two houses is pro-cyclical.