President Donald Trump wrongly claims that “58,000 non-citizens voted in Texas, with 95,000 non-citizens registered to vote.” This latest voter fraud misinformation from the president is based on the state’s efforts to match driver’s license and state ID card applications from noncitizens to voter registration rolls.
However, the Texas director of elections warns the lists provided to county elections officials represent “WEAK matches” (the capitalization and emphasis is the state’s) and does not account for the possibility that many people on the list may have simply become naturalized citizens after having obtained a driver’s license.
History tells us to be wary of these numbers. In Florida, when officials similarly matched driver’s license and voter registration data in 2012 in an attempt to remove noncitizens from voter rolls, an initial list of more than 180,000 names was ultimately whittled down to less than 100.
Nevertheless, Trump, who has long warned — without evidence — about widespread voter fraud, claimed via Twitter that Texas had found evidence that tens of thousands of noncitizens have voted illegally in U.S. elections. Trump claimed Texas’ discovery was “just the tip of the iceberg” and that “voter fraud is rampant” around the country. It “Must be stopped,” he wrote, with “Strong voter ID!”
The president’s tweet came after Keith Ingram, the state’s director of elections, released an advisory on Jan. 25 explaining that since March 2018 his office has been compiling data from the state’s Department of Public Safety on applications for driver’s licenses or Personal Identification Cards in which the applicants supplied a green card or work visa indicating they were noncitizens. The elections office, which is in the secretary of state’s office, then matched that list against the voter registration rolls.
A press release issued by Texas Secretary of State David Whitley announced that about 95,000 people identified by DPS records as noncitizens matched voter registration records in Texas, and that approximately 58,000 of them have voted in one or more elections between 1996 and 2018. The secretary recommended counties send notices to those on the list seeking proof of citizenship in order for the individuals to keep their voter registrations. The press release notes that voting in an election when one is knowingly ineligible — and noncitizens are ineligible to vote in Texas and federal elections — is a second-degree felony in Texas.
U.S. & World
The day's top national and international news.
Texas Attorney General Ken Paxton shared those findings in a tweet (retweeted by Trump) beginning “VOTER FRAUD ALERT.”
But none of those on the lists have been confirmed as noncitizen voters, and experts told us it is likely that further investigation will reveal that the vast majority are not.
Chris Davis, the head of the Texas Association of Elections Administrators, told us it is “certainly possible” that there may be matching errors in the state-generated lists or that many or even most of those on the lists could be people who were noncitizens in the country legally when they applied for driver’s licenses and became naturalized citizens between the time they got the licenses and when they registered to vote. Davis said the vetting of the lists has “only just begun” and that he hopes it ultimately “will shed light on how many voters registered to vote legally.”
Zenén Jaimes Pérez, advocacy and communications director for the Texas Civil Rights Project, said matching those lists is not evidence that these people were not eligible to vote — as the president claimed — and that many likely became citizens after getting a driver’s license.
“The full picture has not been presented,” Pérez told us in a phone interview. “It’s pretty dangerous to claim all of those people registered to vote illegally.”
According to the Department of Homeland Security, nearly 180,000 people became naturalized citizens in Texas in the three years from 2015 to 2017.
The League of Women Voters boasted in 2016 that it had attended hundreds of naturalization ceremonies nationwide and had helped more than 50 percent of these new citizens to register to vote. In Houston alone, the League of Women Voters said it had attended about 11 naturalization ceremonies a year and that the League had helped more that 90,000 new Harris County citizens complete their registration forms.
It’s not a stretch, then, to think that many of the matches identified by the secretary of state’s office were legal immigrants who obtained driver’s licenses or state IDs — which is allowed — and then sometime later became naturalized citizens. At that point, these new citizens would be eligible to vote, and the Department of Public Safety data would not necessary reflect that.
“They are using old, outdated information to make really stark assumptions,” Loyola Law School Professor Justin Levitt, an expert in election fraud, told us.
In addition to failing to account for people gaining citizenship after obtaining a driver’s license, the data are likely to contain matching errors, Levitt said. In his advisory, the state elections director acknowledges, “All records submitted through this process will need to be treated as WEAK matches” and counties should further investigate as they see fit.
“That means they aren’t sure that people matched from one list to the other one are the same people,” Levitt said.
When similar matching exercises have been done in other states, he said, numerous matching errors were uncovered.
At best, the information provided by the state is a list of potential noncitizen voters. But given the uncertainty of the data, Levitt said, “It is wildly irresponsible to announce the potential scope at the outset when it is almost certainly inflated.”
Inevitably, Levitt said, once lists like the one created by Texas are scrutinized, the amount of voter fraud discovered amounts to a small fraction of the original list. That’s what happened in Florida in 2012.
Stop Us If You’ve Heard This Before
Much like the recent headlines in Texas, an Associated Press story from Florida in May 2012 ran under the headline “2012 Election: Nearly 200,000 Florida Voters May Not Be Citizens.” But that alarming figure turned out to be wildly inflated.
Just as Texas did recently, Florida officials, at the behest of then-Republican Gov. Rick Scott, culled a list of noncitizens from a state database of driver’s licenses and matched those names against the state’s voting rolls. That resulted in a list of about 180,000 Florida residents registered to vote who the state said might not be citizens.
But that number fell apart under scrutiny. As in Texas, the driver’s license database didn’t include when legal residents later became citizens. The list sparked a 2012 lawsuit. Two of the plaintiffs, Karla Arcia, a Nicaraguan immigrant, and Melande Antoine, who is a Haitian American, found themselves on the list even though both are naturalized citizens who were eligible to vote in the 2012 elections.
As the updated Associated Press story explains, the initial list of about 180,000 names was later reduced to about 2,600 (which turned out to include a World War II Army veteran). That list was later pared to 207 when checked against a federal database. County elections supervisors found errors in that list too, and when PolitiFact Florida in 2013 asked the division of elections for data, it provided a list of just 85 “noncitizens” ultimately removed from voter rolls. (It was unclear how many of those people may have actually voted illegally.)
“It was sloppy, it was slapdash and it was inaccurate,” said Polk County Supervisor of Elections Lori Edwards in 2013 of the original list of 180,000 names.
We recently wrote about all this when Donald Trump Jr. plucked the sensational AP headline from the history bin, and tweeted it with the comment, “Amazing, but not shocking at all anymore. Nearly 200,000 Florida Voters May Not Be Citizens.” But as we wrote, readers only needed to click on the link in Trump Jr.’s tweet to debunk this whopper. An editor’s note atop the story explains that the initial list of 180,000 names was ultimately whittled to 207, and that an “Aug. 1, 2012, state elections document showed only 85 noncitizens were ultimately removed from the rolls out of a total of about 12 million voters at that time.”
Levitt said he expects to see the same pattern in Texas.
“To be clear, there are probably a few people who were in fact noncitizens when they registered to vote,” Levitt said. “That wouldn’t surprise me at all.”
But he doubts many of those people actually voted. When a legal permanent resident applies to become a naturalized citizen, it is standard practice, Levitt said, to check the candidates against voter registration and voter lists. In other words, he said, you’re likely to get caught if you have voted illegally. And the consequences are severe. In addition to torpedoing the citizenship application, he said, such candidates would face a penalty of up to five years in jail and a $10,000 fine. That’s a lot of disincentive for a single vote.
Among those noncitizens who are on the voter rolls, he suspects most of them got on there by mistake; people who didn’t know that as noncitizens they shouldn’t be registered.
“But it’s hard to believe that number approaches anywhere near 95,000, or that the number who voted is anywhere close to 58,000,” Levitt said.
It may take months for counties to sort through the lists. According to the Texas Tribune, only one Texas county, Galveston, indicated that it would immediately begin sending letters requesting proof of citizenship to people on the list provided by the secretary of state’s office. Most of the large counties contacted by the Texas Tribune indicated “they were still parsing through thousands of records and deciding how best to verify the citizenship status of those flagged by the state.” Some expressed concerns about the quality of the state data.
In Harris County, officials said they would do their own research before sending out letters demanding proof of citizenship. For example, they will try to match the names against lists of recently naturalized residents in the county.
“We are going to proceed very carefully,” Douglas Ray, a special assistant county attorney in Harris who specializes in election issues, told the Houston Chronicle. “We’re going to make sure we don’t improperly disenfranchise anyone.”
Meanwhile, lawyers with 13 civil rights groups in the state have sent letters to county officials and the secretary of state warning that “any actions taken based on this list are likely to violate federal law.”
Not a First-Time Offense
This is, of course, not the first time Trump has made claims about widespread voter fraud. But as we have written on numerous occasions, he has failed to provide any proof. Several months into his presidency, Trump created the Presidential Advisory Commission on Election Integrity — which was to examine “fraudulent voting,” among other issues. The commission was dissolved just months after it was created, and there has been no evidence to date of widespread fraud.
In the past, the president has cited a disputed study about noncitizen voting, which we wrote about here. Numerous voting experts, however, say that voter fraud is exceedingly rare, especially in-person voter fraud.
That doesn’t mean there have been no cases of noncitizens voting illegally. Last year, a 39-year-old Texas woman was sentenced to eight years in prison for voting illegally as a legal permanent resident (but not a citizen). As we wrote previously, a 2015 report from the conservative Heritage Foundation documented fewer than a dozen individual cases of noncitizens convicted of registering or actually voting since 2000.
As for Trump’s call for “Strong Voter ID,” Texas already has a voter ID law that requires most to present a photo ID at the polls (there are exceptions). According to the Texas Tribune, Texas lawmakers are considering bringing up a proposal this legislative session to require some proof of citizenship in order to register to vote, though the U.S. Supreme Court has rejected similar efforts in the past.