Regional Housing Market Characteristics As Predictors of Diverse Areas: Addressing Segregation in America

Background: Although recent trends point to modest decreases in segregation across the United States, the magnitude of these changes vary across metropolitan areas.  Research suggests that characteristics of housing markets (e.g. market strength, types of available housing, and market fragmentation) play a role in addressing segregation.  This study examines how housing markets and income disparity shape levels of racial diversity across 319 metropolitan statistical areas.

Method:  Data for this study are drawn from three data sources: 1) the United States Census, the National Center for Education, and data from the Diversity and Disparities Projects from Brown University’s Longitudinal Tract Data Base.  Neighborhood diversity is measured using a dissimilarity index that is based on the percent of white residents that would need to move from one census tract to another to have evenly distributed white and African American populations throughout a particular region.  Housing market strength is measured as the percent of a region’s housing units that are occupied.  Market fragmentation is based on housing density by school district.  Type of housing available is based on the percentage of multifamily units present in a census tract, and income disparity reflects African American median household income as a percentage of white median household income.  The housing market characteristics are tested as a predictor of racial integration using ordinary least squares regression.

Results:  The average score on the mean dissimilarity index is 53.3 (s.d. = 10.4) across the 319 metropolitan statistical areas.  Meaning that on average, 53.3% of white residents would need to move to have an equal distribution of White and African American residents in each census tract in a region.  The average percent occupied housing units sits at 87.9% (s.d. = 5.6), and the average housing units per school district as percent of total housing units in a region is 16.8% (s.d. = 22.3).  Income disparity across race is quite large, with African Americans earning an average of 63.5% (s.d. = 1.7) of the median household incomes of white individuals.

Findings from the regression suggest that less market fragmentation by school district (β = -0.40, s.e. = 0.02), and less income disparity across race (β = -0.192, s.e. = 0.03) are associated with more diversity (p < 0.05).  Interestingly, racial segregation is more pronounced in areas that have a mix of housing types (β = -0.080, s.e. = 0.04).

Conclusions:  Two key lessons are that regional income inequality between African American and White residents and fragmentation within a region’s housing markets due to school district boundaries are barriers to creating diverse neighborhoods. Regional or state efforts to decrease income disparity,  reduce inequities in school district funding/quality, and market fragmentation based on school districts could lead to the development of racially diverse neighborhoods.  A third lesson from this research is that sub-regional and neighborhood level housing market characteristics may have a stronger effect on diversity than regional level housing market strength.  Identifying regional and sub-regional factors associated with neighborhood diversity can guide policy efforts aimed at addressing segregation in the United States.