How Well Does the American Community Survey Count Naturalized Citizens?

BACKGROUND Citizenship status among the foreign born is a crucial indicator of social and political incorporation, yet there are good reasons to suspect that citizenship status is inaccurately reported on U.S. surveys. OBJECTIVE This paper updates research carried out in the mid-1990s by Passel and Clark (1997) on the extent to which foreign-born non-citizen respondents in U.S. government-sponsored surveys misreport as naturalized citizens. METHODS We compare demographic estimates of the resident naturalized foreign-born population in 2010, based on administrative data, to estimates from the 2010 American Community Survey (ACS). RESULTS Similar to previous research, we find that misreporting in the ACS is especially high among immigrants from all countries/regions who report fewer than five years in the U.S. We also find that among longer-term foreign-born residents, misreporting is concentrated only among those originating in Mexico, especially men, a finding that diverges from Passel and Clark in that we find no evidence of over-reporting among immigrants from Central America and the Caribbean. Finally, the estimated magnitude of misreporting, especially among longer-term Mexican-born men, is sensitive to assumptions about the rate of emigration in our administrative-based demographic estimates, and assumptions about coverage error in the ACS, though altering these assumptions does not change the conclusions drawn from the general patterns of the results. CONCLUSIONS For applications that use citizenship as an indicator of legal status, we recommend that self-reported data on citizenship be accepted at face value for all groups except those with less than five years of U.S. residence and Mexican men.