Using 2020 Self-Response Rates to Measure Census Quality

A new working paper from the Georgetown Center on Poverty and Inequality found that “the 2020 Census likely will contain similar inaccuracies seen in past censuses.”

Authors Bill O’Hare and Jae June Lee analyzed “self-response rates as an early indicator,” albeit an imperfect one, “of differential census data quality (i.e. the gaps in census coverage between groups and geographic areas).” These kind of “census process indicators… can provide early evidence about the likely differential quality of the census.” The paper examined “whether historically undercounted groups have relatively low self-response rates to the 2020 Census” to try to “uncover early evidence about whether historical patterns of unequal coverage in the census were likely repeated in the 2020 Census.”