Racial Inequality and Segregation Measures: Some Evidence from the 2000 Census

This paper examines the extent to which income differences across the racial groups can account for observed residential segregation. The authors adopt an approach to the decomposition of segregation measures that combines the method of indirect standardization with the idea that some degree of segregation is the outcome of purely random processes. The authors use the dissimilarity index as a measure of segregation and data on race and income from US metropolitan areas for 2000. The study finds that the role played by racial income inequality in accounting for segregation is modest but varies significantly across the cities. The role of racial income inequality as a determinant of segregation will increase over time. For those cities, where the share of black households is relatively small, it may be useful to adjust existing measures of dissimilarity to better capture the role of non-income factors in determining segregation. [CDE-DSE WP no.177]