A bit later, he retweeted Texas Attorney General Ken Paxton, who hyped the same numbers with an all-caps intro: “VOTER FRAUD ALERT.”
Paxton’s presentation of the argument was at least nuanced in a way that Trump’s wasn’t. He pointed out that the 95,000 noncitizens had been identified as such by the Department of Public Safety. In fact, as the world quickly learned, it was even less firm than that.
The name matches were weak (as the notice to counties indicated in an all-caps warning of its own), and in short order the state and individual counties started clearing names from the list as people’s statuses were confirmed. As our fact-checkers noted, it’s also more than possible that people on the list obtained citizenship since the time they first presented documentation to the state about their status. In 2016, more than 110,000 people in Texas were granted citizenship. Over the decade from 2007 to 2016, nearly a million people became citizens in the state.
This wasn’t a mystery at the time of Trump’s tweet. The Texas Tribune had already written an entire thread on Twitter urging caution after the state’s initial announcement.
“You might be seeing headlines or tweets tonight that claim Texas says 58,000 non-citizens have voted in Texas,” the paper wrote on Jan. 25. “That is not true. That is not what the state has said.” Two days later, watching “Fox & Friends,” that’s what the president tweeted anyway.
Trump and the Tribune are not in the same business. The latter is interested in sharing accurate information about what’s happening in Texas; the former is interested in sharing information, regardless of provenance, that advances his political goals. Often, during his time in politics, that has meant hyping misunderstood, misrepresented or completely untrue claims about voter fraud. Most infamously, Trump championed an apparently entirely fabricated claim of millions of fraudulent votes cast in the 2016 election with zero evidence — his goal being to cast doubt on the popular vote margin that year, in which he placed second. (That, too, is why he isolated California as being another example of alleged fraud: Hillary Clinton’s wide margin of victory there alone gave her more total votes than Trump.)
By now, Americans should be conditioned to applying a wait-and-see approach to claims of rampant in-person voter fraud. Beyond the fact that such claims have been repeatedly investigated without turning up examples of significant fraud, there have been repeated announcements from authority figures (generally agencies led by Republicans) about thousands of questionable votes or voters — assertions that, in short order, fritter away into dust.
There was a similar push in Florida a few years ago, which gained newfound attention with the state’s contested Senate and gubernatorial races last year. Initial news reports suggested that perhaps 200,000 voters in the state might not be citizens; the actual number was 207, one one-thousandth of that number. There, too, an initial list was quickly whittled down to 1 percent of its initial size, and further investigation determined that only a tenth of that group were ineligible.
A lawsuit filed in Pennsylvania last year claimed that perhaps as many as 100,000 voters in that state were noncitizens ineligible to vote. The state determined that there was actually a pool of only about 11,200 voters about whom there might be questions, a group then narrowed further to about 8,700 before being sent to counties for evaluation. Results of those evaluations do not appear to have been finalized.
Trump’s narrow loss in New Hampshire in 2016 prompted him to claim that voter fraud had blocked his victory, a claim seemingly bolstered when the state discovered more than 6,000 voters who’d registered with out-of-state driver’s licenses. An investigation found that only 66 of those individuals (many of whom were believed to be college students) didn’t have their identities verified -- and only four people were found to have voted illegally, mostly out of confusion about where to vote.
Last year, Georgia’s then-secretary of state, Brian Kemp, flagged 53,000 voter registrations that failed the state’s strict “exact match” verification system. Kemp at the time was facing off against Democrat Stacey Abrams for governor. That system was thrown out by a judge shortly before the election. Four years earlier, Kemp had flagged for additional scrutiny 85,000 applications submitted by a voter-registration organization founded by Abrams. Of that total, about 50 were identified as fraudulent.
The most recognizable advocate for cracking down on voter registrations was Kansas’s former secretary of state, Kris Kobach. At one point, he claimed that changes to registration laws that he had championed might have prevented as many as 18,000 people from voting illegally — though after a fervent effort on his part to uproot illegal voting during his tenure, his office could document only 127 ineligible individuals who voted or tried to vote. By the end of 2017, he had obtained nine convictions for voting illegally. One of those convicted was a noncitizen.
Kobach also ran for governor last year but, unlike Kemp, was unsuccessful. In the wake of his departure from the capitol, the state appears to be winding down his more aggressive anti-fraud efforts.
It’s the specter of rampant voter fraud that is important from a political perspective. Much of the national political debate is centered on unseen and uncountable threats: fraudulent voters, people being smuggled across the border with Mexico for human trafficking, terrorists swarming into the United States. It seems, at first glance, possible that thousands of people might have voted illegally in Texas, especially for those inclined to take Trump’s warnings about fraud at face value.
Time and again, we have seen such claims collapse or erode to almost nothing. But the political utility of those initial big numbers is simply too hard for some to resist.