Demographic Analysis (DA) is one of two methods the U.S. Census Bureau uses to estimate net coverage error in the census. The DA estimates are comprised of data on births, deaths, Medicare data, and international migration that provide independent estimates of the population. Differences between census counts and DA estimates have revealed differential coverage of the population in the census by race. A portion of the differences between DA and census may be driven by (1) race allocation in DA input data, (2) race classification in census data, or (3) changing attitudes on race identification in the census. This paper explores coverage error in the census by accounting for differences that stem from race allocation methods used in the 2010 DA estimates. We will use these findings to refine and improve 2020 DA estimates of coverage differentials by race in the 2020 Census.
Presented in Session 23. Racial/Ethnic Identity