How One Survey Question Changes Who Answers: A Look at Citizenship Queries and Data Gaps

When a census asks about citizenship, some people stay quiet. This is a big deal because missing voices can twist the numbers. In one large study, researchers looked at how a citizenship question on a decennial census affected who replied.

The team used an experiment. Some households got the question, and some did not. They also had access to government records for each address. This let them peek at the makeup of each home without asking directly.

Homes with only U.S.-born White people had a high response rate. About 70 out of every 100 answered the survey when there was no citizenship question. But homes with someone who was likely undocumented? Only around 27 out of 100 replied. That is a 43 point gap. When the citizenship question was added, the gap grew another 2.4 points. The change was small, but real.

The story does not end there. People also left off household members on the form. In homes with a likely undocumented person, members were left out six times more often than in all-U.S.-born White homes. Add the citizenship question, and the rate jumps to over ten times higher.

Why does this matter? Official numbers often come from these surveys. If certain groups are missing, the data will not match what is really out there. Records from the government can show a different picture. They tend to count more non-citizens than surveys do. The gap in who answers helps explain that mismatch.