The Current Population Survey Voting and Registration Supplement Overstates Minority Turnout

The Current Population Survey (CPS) is a key source of information on who votes. Turnout estimates derived from the CPS are often cited in academic research on participation, widely used in the calibration of surveys, and central to ongoing legal and policy debates over the protection of voting rights in the United States. We compare CPS estimates to official voter turnout records from 2008–18 and document consistent, significant discrepancies that call into question the reliability of CPS turnout statistics. Specifically, the CPS overestimates black and Hispanic turnout relative to non-Hispanic whites, whether relying on turnout rates as a share of eligible citizens or the racial/ethnic composition of the voting population. Sampling error and commonly used adjustments to CPS estimates do not account for or correct this bias, and thus academics and policy makers should use discretion when judging recent shifts in voter turnout with survey data.