Do City Borders Constrain Ethnic Diversity?

Résumé 0

U.S. metropolitan areas, particularly in the industrial Midwest and Northeast, are well-known for high levels of racial segregation. This is especially true where core cities end and suburbs begin; often crossing the street can lead to physically similar, but much less ethnically diverse, suburban neighborhood. While these differences are often visually or "intuitively" apparent, this study seeks to quantify them using Geographic Information Systems and a variety of statistical methods. 2016 Census block group data are used to calculate an ethnic Herfindahl index for a set of two dozen large U.S. cities and their contiguous suburbs. Then, a mathematical method is developed to calculate a block-group-level "Border Disparity Index" (BDI), which is shown to vary by MSA and by specific suburbs. Its values can be compared across the sample to examine which cities are more likely to have borders that separate more-diverse block groups from less-diverse ones. The index can also be used to see which core cities are relatively more or less diverse than their suburbs, and which individual suburbs have the largest disparities vis-\`a-vis their core city. Atlanta and Detroit have particularly diverse suburbs, while Milwaukee's are not. Regression analysis shows that income differences and suburban shares of Black residents play significant roles in explaining variation across suburbs.