The Justice Department says it requested the change because it needs more granular data on who is and isn’t a U.S. citizen so that it can better enforce protections for minorities in the Voting Rights Act.
Civil rights groups say the census already provides reliable data on citizenship in a separate survey. The administration’s plan might backfire, they add, if it causes immigrants to avoid the census for fear they might be targeted.
On both sides, many of these claims attempt to predict the future: This might happen, or this might stop happening. We won’t know the true impact of the citizenship question until the 2020 Census is in the rearview mirror.
But some of the claims flying around are clearly inaccurate or misleading now, or they lack key context. We wanted to flag these claims for readers and unpack some of the arguments behind them. As is our custom with quick roundups, no Pinocchios will be handed out today.
“Having these data at the census block level will permit more effective enforcement of the [Voting Rights] Act.”
As commerce secretary, Ross oversees the U.S. Census Bureau. ProPublica reported that Ross overruled career staff in the bureau when he approved the citizenship question on March 26.
Administration officials say the citizenship question will improve the quality of the data the Justice Department uses to monitor and enforce compliance with Section 2 of the Voting Rights Act. This section “prohibits voting practices or procedures that discriminate on the basis of race, color, or membership in one of the language minority groups” recognized by law.
When the Justice Department sues under Section 2, it often does so to block attempts at “vote dilution.” Let’s say more than 50 percent of eligible voters in a district are part of a minority group. If that district has rules or practices that prevent these minority voters from electing a representative of their choice, that’s vote dilution, and the Justice Department might file a lawsuit under Section 2.
The key to these cases is measuring minority voters as a share of the total voter population. Only citizens have the right to vote, so figuring out the total number of voting-age citizens is the same as figuring out total voter population.
The census itself has not included a citizenship question since 1950, so the Justice Department for years has taken citizenship data from the American Community Survey. This long-form questionnaire, sent out as a census supplement, includes a citizenship question. But the Census Bureau sends it out only to 3.5 million households a year, or one out of every 38.
The crux here, then, is whether the ACS produces adequate data for Justice Department lawyers trying to stamp out biased elections.
The Justice Department says no and that’s why the citizenship question is needed in the decennial census itself, not just the ACS.
“This data is critical to the Department’s enforcement of Section 2 of the Voting Rights Act and its important protections against racial discrimination in voting,” Arthur E. Gary, general counsel of the department’s Justice Management Division, wrote in a letter to the Census Bureau dated Dec. 12. “To fully enforce those requirements, the Department needs a reliable calculation of the citizen voting-age population in localities where voting rights violations are alleged or suspected.”
Civil rights groups and many election law experts disagree. They say the ACS gets the job done just fine.
“I am not aware of any cases where civil rights were not able to be enforced because there was some deficiency in the ACS data on citizen voting-age population,” said Wendy R. Weiser, director of the Democracy Program at the Brennan Center for Justice. The Justice Department does not need granular, block-by-block data on citizenship to estimate whether a minority group might be large enough to elect a representative of its choice, she said.
“There’s a broad consensus that this is going to get us wrong data, which is going to dramatically undercount minority communities, which is going to be horrible for civil rights enforcement,” Weiser said.
Joseph R. Fishkin, a law professor at the University of Texas, said the “ACS data has not interfered at all with the DOJ’s ability to enforce Section 2 of the VRA.”
“The reason is that citizenship data is only used in VRA enforcement in one way: as a piece of predicting whether minority voters are going to end up winning or losing in a given district,” Fishkin said.
All kinds of numbers and estimates go into predicting the outcome of an election, he added.
“In particular, the error bars on rates of voter registration, and turnout among those voters who are registered, along with effects of incumbency and other variables, are all much, much larger than any difference in precision between ACS and Census citizenship data,” Fishkin said. “In other words, compared to other variables that we are putting into the same equation, ACS citizenship data is more than sufficiently precise.” (He wrote more about this issue in a blog post.)
We asked the Justice Department whether the quality of the ACS data has ever prevented it from filing or winning Section 2 cases. We received no response.
It’s worth mentioning that the Justice Department has not filed any Section 2 enforcement cases since Trump took office, and filed only five such cases during Barack Obama’s administration, according to a DOJ list available online. Only two of those were vote-dilution cases, Weiser said.
We also reached out to several lawyers who have dealt with Section 2 cases.
“I have full confidence in ACS data,” said Nina Perales, vice president of litigation at the Mexican American Legal Defense and Educational Fund. Perales works with individual plaintiffs filing Section 2 lawsuits and has argued all the way to the Supreme Court.
ACS data are fully reliable at the state, local and precinct levels, she said. But every so often, lawyers have to zoom in to the block-by-block level to see whether vote dilution is happening.
“There is a widely agreed-upon methodology to estimate citizenship at the block level in the rare instances when you might need it,” Perales said. “DOJ uses that methodology, states use that methodology, and private plaintiffs such as those represented by MALDEF also use that methodology. So this has been a nonissue.”
Think of cities or school districts. In those cases, the ACS “has always been close enough” to estimate whether a minority group has enough voters to elect a representative of its choice, said Justin Levitt, a voting rights expert at Loyola Law School.
“There are some cases, few, where it’s right on the cusp, and more precise information would be useful — vanishingly few,” he said.
But Levitt, who worked on Section 2 cases at the Justice Department under Obama, said the citizenship question in the census would not salvage those cases. Whatever improvements you might see in the citizenship data would be defeated by the inaccuracies resulting from fewer immigrants responding to the census, he said.
There’s another school of thought. Christian Adams, who worked on Section 2 cases at the Justice Department under Obama and President George W. Bush, said the block-by-block citizenship data would have made a difference in some of the cases he encountered.
“It doesn’t make any difference if you’re doing a voting rights case for African Americans in, say, Cleveland, Ohio,” he said. (Blacks are a majority in Cleveland.)
“If you’re doing a case in South Florida like I did, and you don’t have CVAP [block-by-block data] you’re going to have a much more difficult time analyzing the case, and anybody telling you otherwise is not being straight,” Adams said.
“Latest absurd freak out is over #census2020 citizenship question. In every nation citizenship matters, so shouldn’t we know how many we have? And districts apportioned based on # of people not here legally dilutes the political representation of citizens & legal residents.”
The last part of Rubio’s tweet suggests the Trump administration is adding the citizenship question to exclude undocumented immigrants from the formula states use to apportion congressional and legislative districts.
Rubio also says including undocumented immigrants in apportionment counts “dilutes the political representation of citizens & legal residents.”
This is debatable. Let’s suppose the citizenship question ends up spooking immigrants, as many Democrats, election law experts and civil rights groups predict will happen. These immigrants might stonewall the census takers knocking on their doors or ignore the census form in the mail, leading to an undercount of both the total population and the immigrant population.
As a result, an immigrant-rich state such as California might shed one or more of its 53 congressional districts. All California residents would see their political power diluted in that case, since they would lose one or more votes in the House. The state would lose the same number of electoral college votes in presidential elections.
In short, Rubio says citizens and legal residents lose some political power within districts by sharing their representatives with undocumented immigrants. But these immigrants can’t vote, and some states overall might lose House seats if they were excluded from apportionment counts.
It’s worth reviewing what the Constitution has to say about all this. The 14th Amendment says, “Representatives shall be apportioned among the several states according to their respective numbers, counting the whole number of persons in each state, excluding Indians not taxed.”
Careful readers will note there’s nothing in the amendment about counting only citizens or legal residents. Instead, it’s “the whole number of persons in each state.” (The part about Indians is an old vestige.)
In deciding the 2016 Texas redistricting case Evenwel v. Abbott, the Supreme Court said apportioning districts based on total population “promotes equitable and effective representation.”
The court was not ruling on the scenario Rubio references: apportioning districts based on the number of eligible voters. However, the justices made several statements supporting the use of total population instead.
“As the framers of the Constitution and the Fourteenth Amendment comprehended, representatives serve all residents, not just those eligible or registered to vote,” Justice Ruth Bader Ginsburg wrote for the court. “Nonvoters have an important stake in many policy debates — children, their parents, even their grandparents, for example, have a stake in a strong public-education system — and in receiving constituent services, such as help navigating public-benefits bureaucracies. By ensuring that each representative is subject to requests and suggestions from the same number of constituents, total population apportionment promotes equitable and effective representation.”
Justice Samuel A. Alito Jr. wrote in a concurrence: “The decennial census required by the Constitution tallies total population. These statistics are more reliable and less subject to manipulation and dispute than statistics concerning eligible voters.”
A Rubio aide said the tweet was meant to rebut claims that adding the citizenship question to the census would lead to voter suppression.
“Voting is limited to citizens,” the aide said. “While citizenship is not the basis for apportionment, the real voter suppression is the fact that the more the 435 congressional districts are based on people who are not here legally, the less representation the people who are here legally will have in Congress as a proportion of our overall population.”
“This is a question that’s been included in every census since 1965, with the exception of 2010 when it was removed.”
Several wires got crossed in this quote.
First, there was no census in 1965. The closest ones were taken in 1960 and 1970.
Second, the census itself has not asked a citizenship question since 1950. Third, the American Community Survey included a citizenship question in 2010, although it was not in the census itself.
Let’s take a quick stroll through the history. In 1950, the decennial census form asked respondents to enter their birthplace and whether they were naturalized.
In 1960, the birthplace question appeared again — but not the naturalization question.
In 1970, 1980, 1990 and 2000, the census sent households a separate, long-form survey that included a citizenship question. This was in addition to the standard questionnaire, but only a fraction of U.S. households would receive it, about one out of every six.
In 1996, the census began to send out the ACS, which it continues to use today. This in-depth annual survey of 3.5 million households has contained a citizenship question since 2005. Contrary to what Sanders indicated, the citizenship question appeared in 2010.
It may not have been in the census itself, or in the long-form survey (which was sent out in 2000 but discontinued by 2010), but the citizenship question did appear in the ACS.
“Asking about immigration status is a transparent attempt to discourage immigrants from responding, and reduce representation in states with large immigrant populations. This violates the Constitution and the government’s responsibility to conduct an accurate census.”
It’s not clear from his tweet what exactly Schiff thinks would be unconstitutional: asking the citizenship question or attempting to discourage immigrants from responding to the census. His staff did not respond to our questions, so it will remain a mystery for now.
Let’s assume Schiff meant that asking the citizenship question would violate the Constitution.
It probably wouldn’t, since the census included a citizenship question until 1950, then once every decade from 1970 to 2000 in the long-form survey, then every year from 2005 onward in the ACS.
Send us facts to check by filling out this form
Keep tabs on Trump’s promises with our Trump Promise Tracker
Sign up for The Fact Checker weekly newsletter