Local Quantile House Price Indices

Locally weighted quantile regressions allow the coefficients of hedonic house price functions to vary over space. Using data on all house sales in Cook County, Illinois, for 20002011, I show how the full distribution of appreciation rates changed over time in small geographic areas. The estimates reveal significant spatial variation in appreciation rates both geographically and across the distribution of house prices. During the boom, house prices rose most rapidly among lower-priced homes, particularly on the South and West sides of Chicago. Prices then declined most rapidly afterward in these same areas. In contrast, high-priced homes in the Near North Side of the city and in the far North suburbs had only moderate declines in prices after 2006. The results clearly indicate that standard approaches to estimating house price indices oversimplify what is actually a rich set of spatial and temporal variation in appreciation rates.