On Monday, the secretary of the United States Department of Commerce directed the Census Bureau to ask about the citizenship of respondents to the 2020 census. The secretary, Wilbur Ross, cited a request from the Justice Department for help in enforcing the Voting Rights Act in explaining his directive. “[H]aving these data at the census block level will permit more effective enforcement of the Act,” he wrote. On Thursday, the bureau submitted its final list of questions to be asked on the census and for the American Community Survey to Congress.
In that listing, the Census Bureau writes, “Knowing how many people reside in the community and how many of those people are citizens, in combination with other information, provides the statistical information that helps the government enforce Section 2 of the Voting Rights Act and its protections against discrimination in voting.”
Opponents of the move see a more political cast to the decision, fearing that such a question will scare off immigrants from responding to the census at all, thereby diluting population counts from states with large immigrant populations such as Democratic strongholds like California and New York (or even Republican ones like Texas). That Ross’ decision came at the bureaucratic 11th hour for amending the census, and which makes testing out the effects of adding the question a practical impossibility, adds to these suspicions.
While the current political climate may stoke such fears, asking about citizenship is hardly unknown. Past censuses, up until 1950, included a citizenship question for all respondents, and from then until the 2010 Census the ‘long-form’ questionnaire included such a question (although no long forms were distributed the 2010 Census), and the American Community Survey administered by the Census Bureau still does.
Despite that history, many academic groups that use U.S Census data for research fear the negative effects of including the question. “Adding a new citizenship question to the 2020 Census would destroy any chance for an accurate count, discard years of careful research, and increase costs significantly,” an umbrella group, The Leadership Conference, wrote to Ross in January.
In his letter, Ross dismissed such speculation:
[N]o one provided evidence that reinstating a citizenship question on the decennial census would materially decrease response rates among those who generally distrusted government and government information collection efforts, disliked the current administration, or feared law enforcement. Rather, stakeholders merely identified residents who made the decision not to participate regardless of whether the Census includes a citizenship question.
In the letter below, the Consortium of Social Science Associations, or COSSA (a Social Science Space partner), explains why it opposes the citizenship question. COSSA described itself as “a nonprofit advocacy organization working to promote sustainable federal funding for social and behavioral science research and federal policies that positively impact the conduct of research.”
The Consortium of Social Science Associations (COSSA) is disappointed in the decision by the Department of Commerce to add a question on respondents’ citizenship to the 2020 Decennial Census. This decision circumvents the Census Bureau’s routine research and testing processes to ensure potential questions do not affect the quality of responses and could compromise one of the most valuable data resources the government produces.
It is simply too late in the cycle to contemplate adding a question to the census. In the decade leading up to a decennial census, the Census Bureau conducts years of rigorous research and testing to ensure that even the smallest changes to design and wording will not impact the accuracy of the responses received. This research and testing phase culminates in the “dry run” of the census, the 2018 End-to-End Census Test, which is being conducted now—without a question on citizenship. Even Secretary of Commerce Wilbur Ross concluded that his department “is not able to determine definitively how the inclusion of a citizenship question on the decennial census will impact responsiveness.”
Census data is too important to risk an untested question that has not been fielded since 1950. The decennial census is an irreplaceable source of data for researchers across the social sciences who use it to generate valuable findings about the U.S. population that can be used to inform evidence-based policies. In addition, information from the decennial census undergirds numerous other surveys and data sets at the Census Bureau and beyond, so a problem at the source would have far-reaching implications across the statistical system. We have no way of knowing what future insights will be lost if this data is compromised. Obviously, the value of reliable Census data extends beyond its use for research. It is used to allocate more than $800 billion in taxpayer dollars to programs across the country, by businesses in the private sector to guide investment decisions, and by state and local governments to make better policies.
COSSA hopes that Congress will exercise its oversight authority to safeguard the integrity of Census data and enact legislation to prevent untested, unresearched questions from being added to the decennial census. We look forward to working with our partners in Congress and across the stakeholder community to ensure a fair and accurate 2020 Decennial Census.