The Uptick in Income Segregation: Real Trend or Random Sampling Variance

Abstract

Recent studies have reported a reversal of an earlier trend in income segregation in metropolitan regions, from a decline in the 1990s to an increase in the 2000-2010 decade. This finding reinforces concerns about the growing overall income inequality in the U.S. since the 1970s. Yet the evidence may be systematically biased to show an upward trend because the effective sample for the American Community Survey (ACS) is much smaller than it was for Census 2000, to which it is being compared. There is a possibility that the apparent changes in disparities across census tracts result partly from a higher level of sampling variation and bias due to the smaller sample. This study uses 100% microdata from the 1940 census to simulate the impact of different sampling rates on estimates of several measures of segregation and to propose and test the effectiveness of approaches to correcting the bias. It then applies those approaches to publicly available data for 2000 and 2007-2011. The reduction in sample sizes associated with the ACS results in exaggeration of evidence for increasing income segregation for all measures tested here, especially for subgroups (African Americans are studied here as an example). The methods of correction applied here will yield more conclusive and unbiased results when applied to the original sample data that is held internally by the Census Bureau.