A New Approach to Estimating Hedonic Equilibrium Models for Metropolitan Housing Markets∗

We formulate and estimate an equilibrium model of metropolitan housing markets with housing differentiated by quality. Quality is a latent variable that captures all features of a dwelling and its environment. Two applications illustrate the value of our new technique. First, we develop a semi-parametric approach to estimate the model for Chicago and New York, obtaining housing prices for each quality level for each metropolitan area, stocks of each quality, and compensating variations required for a household of a given income in Chicago to be equally well off in New York. Second, we estimate rents and prices across the quality spectrum in Miami during the dramatic run-up in housing prices in the years preceding the financial crisis.