Though the Commerce Department asserts its decision to include a question about citizenship status in the 2020 census will produce more accurate data and ensure fair representation for all U.S. residents, census experts and civil rights groups warn the impact could be exactly the opposite.
“We do not know the answers to any of those very important questions that would tell the Census Bureau whether or not this can be done well without pulling the rug out from under the entire census,” said Terri Ann Lowenthal, who served as staff director for the House subcommittee overseeing the census from 1987 to 1994 and covered census issues for President Barack Obama’s transition team in 2008.
In a memo late Monday, Commerce Secretary Wilbur Ross directed the Census Bureau to place the question last on the decennial census form that will be sent to every household in the country in 2020.
Though the Census Bureau submitted its planned questions for 2020 to Congress last March, the Department of Justice asked the Commerce Department in December to add citizenship to the survey. In a letter, DOJ General Counsel Arthur Gary claimed the answers would be critical to “enforcement of Section 2 of the Voting Rights Act and its important protections against racial discrimination in voting.”
Citizenship was included in most decennial census surveys for the whole population from 1820 to 1950. After that, it remained in a long-form survey that was sent to one in six people until 2000. The American Community Survey (ACS), which replaced the long-form questionnaire, still includes the question, but it is only sent to 2.6 percent of the population.
According to the DOJ, the ACS does not provide information in small enough units for redistricting purposes and census-level population and citizenship data is needed. However, some experts predict demanding citizenship status will lead immigrants, including many in the U.S. legally, to refuse to participate at all.
“It’s a false flag from our standpoint that they’re saying the greater good is to talk to and enumerate 100 percent because it’s not going to be 100 percent. Participation rates are going to drop off rapidly,” said Phil Sparks, co-director of the Census Project, a coalition of stakeholders that includes business, labor, government, civil rights, and child advocacy organizations.
Ross acknowledged that the Census Bureau itself was among those concerned that adding the question to the decennial census will result in a significantly lower response rate for non-citizens. Internal analyses indicated noncitizens and Hispanics were less likely to answer inquiries about citizenship, but none definitively proved a direct causal link.
According to the Ross memo, self-response rates for the ACS were 3.1 percent lower than for the 2010 decennial census, but the Census Bureau determined that could be attributed to the fact that the census survey is much shorter or that more outreach and follow-up is conducted for it since it is constitutionally mandated.
Nonresponses to the citizenship question on the ACS from 2013 to 2016 were around 6 percent for whites and about 12 percent for blacks and Hispanics. Those rates were comparable to nonresponse rates for several other questions, though.
The Census Bureau also reviewed self-response data from the 2000 short-form and long-form surveys from citizens and non-citizens. For the long-form questionnaire, which included citizenship, the decline in response rate for non-citizens was 3.3 percent greater than for citizens. Again, analysts were unable to determine how much of that was attributable specifically to the citizenship question.
“The citizenship data provided to DOJ will be more accurate with the question than without it, which is of greater importance than any adverse effect that may result from people violating their legal duty to respond,” Ross concluded.
Democrats have blasted the announcement as an attempt to slant the census in favor of Republicans, and civil rights groups and advocates for communities of color promptly sounded alarms about the damage it could do.
“This untimely, unnecessary, and untested citizenship question will disrupt planning at a critical point, undermine years of painstaking preparation, and increase costs significantly, putting a successful, accurate count at risk,” said Vanita Gupta, president and CEO of the Leadership Conference on Civil and Human Rights, in a statement.
Gupta, who was head of the DOJ civil rights division under President Obama, said the question is “unnecessarily intrusive,” will result in “deeply flawed” data, and will raise doubts about the confidentiality and use of census data.
Many saw the announcement as just the latest of many efforts by President Donald Trump’s administration to intimidate and disenfranchise immigrants.
“This is an attempt to racially rig the census and undercount communities of color, and goes against the fair representation our democracy relies on,” said Karen Hobert Flynn, president of Common Cause.
Census data has enormous implications for congressional representation, electoral votes, state legislatures, and federal funding. Data collected in 2020 will also form a baseline for other research and corporate business decisions over the following decade, so some fear undercounting of immigrant populations will cause long-term harm for many communities.
“It’s the cornerstone of our democracy in determining both representation in Congress, but, perhaps just as importantly, federal funding that goes out to states and communities,” said John Yang, president and executive director of Asian Americans Advancing Justice.
Critics of the Commerce Department decision, including organizations that fight for equal voting rights, are highly skeptical of the DOJ claim that existing ACS data is now insufficient for enforcement of the Voting Rights Act.
“The notion that it is necessary for Voting Rights Act enforcement is laughable, given that the census has not included a citizenship question since the enactment of the VRA in 1965,” said Dale Ho, director of the ACLU Voting Rights Project, warning that undercounting will result in less representation and federal resources for communities of color.
“Contrary to the Administration’s stated rationale, asking the citizenship question on the census is not critical to enforcing the Voting Rights Act,” said former Attorney General Eric Holder, now chair of the National Democratic Redistricting Committee. “As attorney general, I did not—nor did my predecessors—request the addition of a citizenship question to the decennial census to enforce the VRA. We did not need to: Data derived from the existing census process was perfectly adequate for any voting litigation that arose.”
While Ross maintained that he has been shown no conclusive proof that including the question will result in undercounting, his critics counter that he has offered no proof that it will not.
“I think that is one of the most egregious statements in the secretary’s decision memo, to suggest that stakeholders ought to have that evidence,” Lowenthal said.
She also observed that the Census Bureau estimates 28 to 34 percent of citizenship self-responses from noncitizens on the ACS are inaccurate.
“How therefore can he say that collecting the data through the same question on the census would be more accurate?” she asked.
Yang called Ross’ logic “self-contradictory” since the secretary prioritizes complete and accurate data while also admitting uncertainty about this question’s impact.
“The bottom line, everyone agrees that we want accurate data,” he said. “If we’re imposing an untested question, it calls into question the accuracy of that data.”
The state of California has already filed a legal challenge over the question and more litigation is likely in the weeks ahead. New York Attorney General Eric Schneiderman said in a statement Tuesday he will lead a multi-state lawsuit against the Trump administration “to protect New Yorkers.”
Advocates of stricter immigration controls dismiss claims that the question will cause underrepresentation, and they argue more accurate information on citizenship is important for the government to have.
“It has a huge impact on all sorts of things and it is very legitimate for us to try to understand how immigration is affecting the country,” said Ira Mehlman, media director for the Federation for American Immigration Reform. “Judging from the brouhaha on behalf of the mass immigration advocacy movement, it appears they have something to hide here because they don’t want this info even collected.”
FAIR has long called for congressional and state legislative districts to be determined based on legal population instead of total population.
“The idea that we are going to take representation away from citizens in one state and give it to illegal aliens in another state is something that I think most Americans would be scratching their heads about,” Mehlman said.
Research released Monday by the Center for Immigration Studies argued that nonresponses to the ACS—which includes the citizenship question but is only sent to 3.5 million households—have been rising for years and Trump’s presidency has not yet resulted in a spike in refusals to answer.
“This is contrived by advocacy groups trying to keep information that might affect the policies they want in place, keeping it from the American people,” Mehlman said, stressing that census responses are anonymous.
President Trump and his allies have done little to calm fears about the motivation behind the decision.
The President wants the 2020 United States Census to ask people whether or not they are citizens. In another era, this would be COMMON SENSE... but 19 attorneys general said they will fight the President if he dares to ask people if they are citizens,” Trump’s reelection campaign declared in an email to supporters last week.
“I suppose everything becomes political one way or another, but this is inherently important information that allows us to make good decisions on all sorts of things,” Mehlman said.
That argument is premised on the question resulting in better data, and some experts are not convinced it will improve what was already going to be an arduous and expensive task.
“This census is going to be more costly and less fair,” Sparks said.
The 2020 census will be the first to allow online responses and bureau analysts were already concerned about rising distrust of government and information security fears. Self-responses are expected to be significantly lower than in 2000, meaning more census-takers going door-to-door to follow up in a volatile political environment.
“It’s likely that’s going to be like pulling teeth,” he said.
Funding shortages forced the Census Bureau to curtail its large-scale field test of new counting techniques. Originally planned to cover 700,000 households in rural, suburban, and urban communities, the test is now being conducted with only 275,000 households in Rhode Island. Due to the late decision, the citizenship query will likely not be tested at all.
“That might have answered the question, but the cow is already out of the barn,” Sparks said.
In his memo, Ross cited nine other democracies that ask citizenship questions in their census surveys, a point Mehlman argued suggests this is legitimate information to collect and it will not spark panic. Lowenthal called that an “apples to oranges” comparison since some of those countries use different procedures, and she said it is incumbent on Ross to prove that the question will not have adverse effects before imposing it.
“It’s like opening day of a musical on Broadway and you add a new number that wasn’t ever practiced in the dress rehearsal,” she said.