Patterns and Correlates of Racial Residential Segregation: A Spatially Refined Approach

Barrett A. Lee, Pennsylvania State University
Glenn Firebaugh, Pennsylvania State University
Chad R. Farrell, University of Alaska at Anchorage
Sean F. Reardon, Stanford University
David O'Sullivan, University of Auckland

Despite its presumably spatial emphasis, the literature on racial residential segregation tends to be aspatial in nature. We respond to this limitation by pursuing a new approach that incorporates GIS-based proximity calculations to estimate the extent of segregation across a range of scales (i.e., for local environments of varying radii). The measure featured in our approach is a spatially informed version of the information theory (H) index. Using 2000 census data for the 100 largest U.S. metropolitan areas, we: (1) report means and variances in black-white, Hispanic-white, Asian-white, and black-white-Hispanic-Asian segregation at different spatial scales; (2) supplement metropolitan structural characteristics drawn from Farley and Frey (1994) with an indicator of school district fragmentation to determine what predicts the scale of segregation; and (3) identify the metropolitan characteristics associated with changes in segregation as one moves from smaller to larger local environments.