President Trump and his administration are waving the white flag on the census citizenship question for the second time this month — this time apparently for good. Trump announced Thursday that he’ll push to get citizenship data by some other means, rather than attempting to avenge his administration’s Supreme Court loss or trying an executive order.

The news will hearten those who feared the census question would be used to help Republicans politically.

But what if it already has?

Opponents of the move have long argued that this whole thing is an attempt to dissuade undocumented immigrants from responding to the census. And if undocumented immigrants don’t respond, the argument goes, that would dilute the power of the largely Democratic areas in which they live and shift power to more Republican-friendly areas.

Some opponents, though, have suggested this might happen even if the question isn’t ultimately on the census. They said that could happen simply by virtue of the protracted debate we’ve had — which could render the undocumented immigrants suspicious of the process and the census.

Denise Hulett, a lawyer for the Mexican American Legal Defense and Educational Fund (MALDEF), broached the possibility with a federal judge last week, after President Trump tweeted that the census fight would continue.

“The president’s tweet has some of the same effects that the addition of the question would in the first place and some of the same effects on the 18-month battle that was just waged over the citizenship question,” Hulett said. “It leaves the immigrant communities to believe that the government is still after information that could endanger them.”

And there is reason to believe it could have just such an impact.

Response rates to federal government surveys have been dropping for decades, leading to significant fears about particular groups being undercounted. A survey by the U.S. Census Bureau ahead of the 2010 census found 85 percent of people said they would “probably” respond to the 2010 census, yet only about three-quarters did. In the same survey conducted ahead of the 2020 census, just 67 percent said they were at least “very” likely to respond. (The surveys used different adjectives and don’t provide a direct comparison.)

Among the groups least likely to participate? The ones that logically include undocumented immigrants. The pre-census survey did not ask people about their citizenship, but Hispanics, immigrants and non-English speakers were all less likely to fill out a census form.

To wit:

  • While 30 percent of all respondents were “extremely likely” to respond, just 23 percent of Hispanics were.
  • While 68 percent of people who were proficient in English were at least “very likely” to respond, just 55 percent of nonproficient people were. Just 13 percent of those without English proficiency were “extremely likely” to respond.
  • While 68 percent of nonimmigrants were at least “very likely” to respond, 63 percent of immigrants were. Here again, very few were “extremely” likely to respond, at 19.5 percent.

These groups were also more likely to be worried about census data being shared with other government agencies — possibly including Immigration and Customs Enforcement (ICE) — and about it being used against them.

While just 24 percent overall worried the information would be shared across agencies, 32 percent of Hispanics did, 32 percent of immigrants did, and 37 percent of non-English speakers did.

While just 22 percent overall thought their answers might “be used against you,” 32 percent of Hispanics did, 34 percent of immigrants did, and a whopping 39 percent of non-English speakers did.

Believe census info might be shared across agencies
Believe census info might be used against them
Non-English speakers

Even if some undocumented immigrants are already scared away, the new decision means the census won’t provide something else Republicans have coveted and could help them redistrict their way into more power: the ability to draw districts according to citizens, rather than total population. You need granular data to attempt that; it’s not clear it will be possible any other way.

There is an argument to be made that an information campaign could set the record straight and prevent a decline in survey response rates among undocumented immigrants. That’s because it turns out they believe in the importance of the census even more than groups that are more likely to respond. While 52 percent of American-born people think filling out the census could benefit their communities, between 62 and 64 percent of all three groups mentioned above think it could help their communities.

And at this point, relatively few of them think their responses could “harm” them personally — 17 percent of Hispanics, 16 percent of immigrants and 20 percent of non-English speakers. All are similar to the 19 percent overall who say this.

But this survey was conducted before the census citizenship debate came into focus, and the above data suggest there is considerably more suspicion about the process among groups that are likely to contain undocumented immigrants. The idea that the government might be trying to root out undocumented immigrants using the census is something they may already be suspicious of. In fact, 10 percent of people in the survey already did think the census was used for this purpose, while another 37 percent thought it might be.

Layer on top of that some nationwide ICE raids, and it’s not difficult to see why undocumented immigrants would steer clear of responding to a survey from the federal government — even if it doesn’t actually contain a question about their citizenship.