The Cross Section of Housing Markets

National Bureau Of Economic Research Inc, Cambridge MA

Investigators

Abstract

Abstract Proposal Title: The Cross Section of Housing Markets (Proposal No: 1156320) Principal Investigator: Piazzesi, Monika The projects described in this proposal focus on two key properties of housing: houses are (i) indivisible and (ii) heterogeneous by quality, even within geographical units. The proposal has both an empirical and a theoretical component. The empirical work exploits several large property-level data sets to document new stylized facts on the cross section of housing markets. We then proceed to build quantitative models that speak to these facts. The first project uses an assignment model to understand the cross section of capital gains during the 2000-2005 housing boom. The second studies the search process with heterogeneous houses. During the recent house price boom, there were large differences in capital gains across houses, even within the same metro area. In particular, cheap homes appreciated more than expensive homes. The first project studies several potential explanations for this pattern: cheap credit, shifts in the quality distribution of homes, or shifts in the distribution of mover characteristics (such as their income). The goal is to build a quantitative model to assess the role of these differences in the environment. The existing literature has investigated the recent housing boom with models in which housing is viewed as perfectly divisible capital. These models determine a single per-unit price of housing from optimality conditions of all households and thus do not have interesting implications for the cross section of house prices. Instead, this project studies models in which indivisible houses differ by quality. The price of a particular house is then determined by the optimality conditions of the particular subgroup of households who consider buying a house in that quality range. To the extent that a change in the economy (e.g., cheaper credit) affects this subgroup of households more (e.g., poor households depend more on credit), the house prices in their considered quality range will be more responsive (compared to existing models.) The second project studies models with heterogeneous houses and search frictions. In the housing market, sellers post prices without commitment, while buyers scan available houses. The search process is costly and makes houses illiquid. A key question for any quantitative implementation of such a model is ?what market segments do buyers search over?? This project makes progress on this question with databases on price histories from real estate listings (including time on the market as well as price reductions) and search queries on a popular real estate website. The preliminary data investigation indicates that buyers in the cheap segment of the housing market use narrower search criteria, explaining the longer times on the market and higher illiquidity for houses in this segment. The results of the proposed research will be informative about the quantitative role of frictions in housing markets. The recent house price declines have highlighted the importance of understanding the functioning of these markets, which makes these frictions worth studying for academics, policy makers, and the public more broadly.