Changing Patterns of Residential Segregation in a Prismatic Metropolis: A Lacunarity-Based Study in Houston,1980–2000

The role of race versus class in shaping residential segregation patterns has been a contentious issue in segregation studies for decades. Despite the voluminous, interdisciplinary literature, scholars have reported conflicting evidence on the role of race versus class in residential segregations. We attribute the current inconclusive literature partially to the failure to consider scale explicitly in residential segregation measures, and partially to the growing complexity of a multiethnic melting pot in most cities in the United States. Inspired by new metrics employed by landscape ecologists to measure landscape heterogeneity, residential segregation is reconceived as a scale-dependent social phenomenon in this paper. We also present an alternative to existing structural or spatial segregation measures, considered as less efficient because most of the existing indices measure only a few dimensions of segregation at a single scale. We have developed a multiscale, lacunarity-based segregation measure, and have used it to examine the role of race versus class in residential segregation patterns in Houston, Texas. Using census-tract-level data from 1980 to 2000, we found that race is still the most important factor in explaining residential segregation despite the overall decline of segregation by both income and race. It was also found that the changing segregation patterns over time are contingent upon the scale as well as the race or income group considered.