August 7, 2018

Comments Submitted to U.S. Census Bureau on Citizenship Question

On March 26, 2018, Secretary of Commerce Wilbur Ross ordered that the 2020 Decennial Census include a question about the citizenship of all U.S. households for the first time since 1950. Contrary to the Secretary’s rationale, asking a citizenship questions on the census is not critical to enforcing the Voting Rights Act and enforcement of the right to vote has never depended on the collection of citizenship data via the census. Thus, the Secretary did not provide adequate reasoning or justification for the decision as is required by the Administrative Procedure Act.

The Niskanen Center believes that the collection of information is necessary for the proper functioning of the Department of Commerce and the U.S. Census Bureau. However, because the addition of a citizenship question is unrelated to the function of the agency and the information itself does not have any practical utility for the agency, the question should be removed.

An accurate count of every living person in the United States is required to adhere to our constitutional requirements. The purpose of the decennial census is to enumerate the total U.S. population; enumeration includes both citizens and noncitizens. Citizenship is immaterial to the goal of enumerating the population. Adding the citizenship question, therefore, bears no reasonable relationship to the goal of the taking the census — and may in fact harm efforts — and runs contrary to the Constitution. Furthermore, there is no practical utility in re-administering the exact proposed citizenship question already used in the American Community Survey, supplemented in precisely the same information and in the same manner the Department of Commerce proposes to use census, in which populations are difficult to quantify. This is because the population targeted by the citizenship question—the undocumented population—is less likely to respond to the more formal and intimidating inquiry of census takers than the ACS survey, which is targeted at fewer individuals.

Full comment available here.