Dynamic Model of Housing Supply

Housing markets often exhibit a high degree of volatility in prices and quantities, with significant economic consequences for both homeowners and the construction sector. While aggregate housing market patterns have been well-documented, our understanding of the micro foundations of housing markets, particularly the timing of housing supply responses to demand shocks, is limited. In order to shed light on the primitives underlying these patterns, I develop and estimate the first dynamic microeconometric model of housing supply. In the model, parcel owners choose the optimal time and nature of construction, taking into account their expectations about future prices and costs. While retaining computational tractability, the model includes a large state space, allows profits to vary at a very fine level of geography, and incorporates a much broader definition of costs than previous work. Estimates of the model, using a rich data set on individual land parcel owners in the San Francisco Bay Area, indicate that changes in the value of the right-to-build are the primary cause of house price appreciation and that the demographic characteristics of existing residents are determinants of the cost environment. I find that fluctuations in broadly defined cost variables, which include regulation costs, are key to understanding when and where building occurs. A counterfactual simulation suggests that without pro-cyclical costs and forwardlooking behavior, construction volatility would be substantially greater.

For more information, contact Petra Todd.