Abstract
Two techniques, computation of segregation indices within income categories, and indirect standardizations based on income and two measures of housing cost, are used to assess the extent to which residential segregation in the St. Louis metropolitan area in 2000 may be attributed to socioeconomic differences between whites and African Americans. Both techniques indicate that the overwhelming majority of segregation is attributable to race, not class. In all income categories, segregation of blacks from whites is as great or greater than in the overall population. Indirect standardizations show that housing cost measures using estimated value of the house explain away somewhat more segregation than those based on monthly owner cost. However, the former type of measure may itself reflect consequences of racial discrimination in housing and lending. Despite a modest decline in the proportion of segregation attributable to race, the large majority of segregation remains tied to race, not class.