Association for Public Policy Analysis & Management

Studies of residential segregation tend to focus on characteristics of households that shape neighborhood outcomes, rather than on housing options available in the neighborhoods where households seek to live. The spatial configuration of housing determines the housing opportunities in each neighborhood, the backdrop against which households’ resources, preferences, and constraints play out. In this study, I use Census and American Community Survey data to document trends in the segregation of housing units by type and by cost at multiple geographic scales, and I test whether housing segregation predicts residential segregation by income in the 100 largest metropolitan areas from 1990 to 2014. Housing segregation between single-family and multifamily units increased during this time, while segregation between higher- and lower-cost housing units declined. The majority of housing segregation occurs between neighborhoods, within places. Housing segregation varies among metropolitan areas, across geographic scales, and over time, and residential income segregation is markedly higher when and where housing segregation is greater. This study complements research that implicitly assumes but does not examine the spatial configuration of housing. As long as housing opportunities remain segregated, residential segregation will change little, with urgent implications for urban and housing policymakers.