Last week brought a whirlwind of news about the 2020 Census. An omnibus spending was passed, providing much needed funding for the Census Bureau. And days later, it was announced that the census would include a question about citizenship, potentially discouraging undocumented non-citizens from responding. Here’s everything you need to know, in a handy Q&A format.
What’s the new citizenship question?
According to their latest announcement, the census will be asking respondents whether they are a citizen of the United States. Here’s how the question will look:
Why does the Census say they’re adding this question?
According to a by Commerce Secretary Wilbur Ross (the Census Bureau is part of the Department of Commerce), the Justice Department requested this data to help determine violations of the Voting Rights Act and help enforce the law.
Does the Census normally add questions based on requests from other departments?
Generally speaking, requests from other executive agencies are the primary mechanism for how questions are added to Census surveys.
Why are some people concerned about this question being added?
Because of the perceived anti-immigrant sentiment coming from the presidential administration, people are concerned that both documented and undocumented immigrants will be reluctant to answer this question, and potentially not respond to the census at all.
Because the census counts all residents of the United States, and not just citizens or legal residents, this could lead to an undercount. An undercount would mean misrepresentation in Congress, since districts are apportioned according to census results. It would also mean federal funds not being apportioned according to population, leaving undercounted areas short on resources. Furthermore, it could simply result in bad data, hurting anyone who uses decennial census data for research or decision-making.
Who stands to lose if this question suppresses responses?
Aside from researches and policymakers who rely on complete data, logic would dictate that this would affect any area with a population of undocumented immigrants (or even wary documented immigrants). States with large immigrant populations, such as California, New York, and Texas could lose seats in Congress. This would be even more of an issue for drawing congressional districts within states. In general, these areas tend to vote for Democratic candidates.
Could these responses be used to locate and deport undocumented immigrants?
By law, individual responses cannot be made public or shared outside the Census Bureau. According to the Census:
It is against the law for any Census Bureau employee to disclose or publish any census or survey information that identifies an individual or business. This is true even for inter-agency communication: the FBI and other government entities do not have the legal right to access this information. In fact, when these protections have been challenged, Title 13’s confidentiality guarantee has been upheld.
Of course, that’s not useful if respondents don’t believe it.
But that said, it’s certainly possible that government agencies could use the results to send enforcement resources to areas with high rates of non-citizens.
What happens if respondents either don’t respond, or leave the question blank?
Legally, respondents are required to respond, and answer each question. That said, no one has been prosecuted for not responding to a Census survey since 1970.
If a respondent doesn’t respond, the Census will have to send a human to the household to try to get a response. This could increase the overall cost of conducting the census. It’s unclear if the Census would send an in-person follow-up for a single unanswered question.
More problematically, if people decide not to complete the census form at all, the population will be undercounted. An area won’t receive its appropriate level of congressional representation and local services will be inadequately funded. This could result in overcrowded schools and hospitals, inadequate roads, under-resourced law enforcement, etc.
Has this question been asked in the past?
Citizenship was last asked to all census respondents in 1950, which asked where each person was born, and “If foreign born – Is he naturalized?” (Note the antiquated use of the male pronoun.)
From 1970 to 2000, the decennial census came in two formats, the short form and the long form. Five out of six households received the short form, which asked only basic questions, like age, race, and relationship status. The short form did not ask about citizenship. One sixth of households received the long form, which asked about citizenship, along with a slate of other more detailed questions.
Starting in 2005, the decennial census long form was replaced by the American Community Survey (ACS), an ongoing survey sent to a sample of households across the country. The ACS contains the citizenship question. The 2010 decennial census contained only the short form, and did not include the any reference to citizenship.
Does this affect the ACS?
Not directly. The household sample used by the ACS is independent of the decennial census. However, it’s often convenient to compare ACS data to census data, and a census undercount would make this more difficult. Of course, the ACS already asks the citizenship question, so one could argue that this brings them closer together.
Are congressional districts based on the total population, or voter-eligible population?
According to a recent Supreme Court case, Evenwel v. Abbot, states may draw districts according to total population. It is not clear whether this requires them to. (FiveThirtyEight had a good piece on this recently.)
Is this decision final? Is anybody fighting it?
The Attorney General of California announced that he’ll be suing the administration, and at least eleven other states have announced their intention to join. At least one Democratic Congressional representative announced plans to introduce legislation to block the question, although given how the question is predicted to help Republicans, it’s hard to see it passing.
It’s worth pointing out that other current administration proposals have been rejected by the courts, such as rollbacks to EPA regulation and DACA. The reasoning behind some of these decisions has been that the administration acted too quickly without following the necessary procedure. Given that a panel of experts assembled by the Census Bureau came to the conclusion that the citizenship question would be ill-advised, it seems this argument might come up again.
Is this the only new question being added?
No, there are a few others. For those who are married or in relationships, the census will ask whether these relationships are opposite-sex or same-sex. This will be the first census conducted since same-sex marriage was legalized in all states. Previous censuses asked about unmarried partners, paradoxically excluding same-sex married couples in states where marriage was legal.
Additionally, the census will ask respondents whose races are White or Black for their ethnic ancestry. This question, which has been on the ACS, is a fill-in-the-blank response, with suggested examples of “German, Irish, English, Italian Lebanese, Egyptian, etc.” for White respondents, and “African American, Jamaican, Haitian, Nigerian, Ethiopian, Somali, etc.” for Black respondents. Although this data is very interesting, we at PolicyMap have opted not to include it in our offerings, for reasons you can read about here.
Any other news last week?
Yes, very good news. After months of being underfunded, the Census Bureau received much needed funding as part of the omnibus federal budget, passed by Congress. This should put to ease concerns that the Census won’t have the necessary funding for the 2020 count.
Where should I go for more updates?
Mainstream news outlets have been covering the citizenship question well, so if there are any major updates, you’ll likely hear about them at places like the New York Times or National Public Radio. On Twitter, I’ve been following The Census Project (@censusproject), All Things Census (@allthingscensus) from the Pew Research Center, and NPR reporter Hansi Lo Wang from NPR (@hansilowang), whose been covering all of these topics extensively.
Of course, you can follow us at @policymap, where we’ll be posting and retweeting updates from wherever we see them.
What’s with the inconsistent capitalization of “census”?
When referring to the government agency, it’s the Census. When referring to the actual survey, it’s a census.
What if I have a question not asked here?