Heterogeneity within Communities: A Stochastic Model with Tenure Choice

Abstract

Standard explanations for the observed income heterogeneity within communities rely on differences of preferences across households and heterogeneity of the housing stock. We propose a dynamic stochastic model of location choice where households differ according to income only, and homes are identical within locations. Households choose whether to own or rent their home motivated by concerns over housing expenditure risk. The model highlights how differences in the timing of moves generate income heterogeneity across homeowners within neighborhoods, in particular in cities that experience strong positive demand shocks. US Census data provides evidence in favor of the income mixing mechanism we identify. In communities that have experienced strong price growth, the heterogeneity of homeowners' incomes is positively correlated with the heterogeneity of the times since they bought their homes. Homeowners who moved in more recently earn higher incomes than homeowners who bought earlier, more so in cities with strong housing price growth. These relationships do not hold for renters.

JEL Classification: D31, R12, R21