Estimating the Effect of Asking About Citizenship on the U.S. Census Results from a Randomized Controlled Trial

Abstract

The 2020 U.S. Census will, for the first time since 1950, ask about residents’ citizenship status. The effect of doing so on census completion across different racial/ethnic groups is, however, unknown. We introduce the notion of contextual sensitivity to explain how seemingly innocuous questions can become costly to answer in certain political environments. Using this concept and a large survey experiment (n = 9,035 respondents), designed to mirror the appearance and substance of the 2020 Census, we find that asking about citizenship status significantly increases the percent of questions skipped, with particularly strong effects among Hispanics, and makes respondents less likely to report having members of their household who are of Hispanic ethnicity. When extrapolated to the general population, our results imply that asking about citizenship will reduce the number of Hispanics reported in the 2010 Census by approximately 6.07 million, or around 12.03 percent of the 2010 Hispanic population.