Donald Trump is citing unsubstantiated urban myths and a contested academic study to paint a false narrative about rampant voter fraud in the U.S. and the likelihood of a "rigged" election.
For weeks, Trump has been warning about rigged elections. He urged his supporters in Ambridge, Pennsylvania, on Oct. 10 to monitor polls and "watch other communities, because we don't want this election stolen from us."
In a speech in Wisconsin on Oct. 17, Trump provided some detail and purported evidence to back up his claims about the prevalence of voter fraud, particularly by noncitizens and people casting ballots on behalf of deceased voters. But we found that his evidence is lacking.
One of Trump's principle claims of voter fraud is that "dead people" are voting in large numbers.
"People that have died 10 years ago are still voting," Trump said in his Wisconsin speech.
Later, Trump cited a Pew Charitable Trust report as evidence of "dead people" voting in large numbers. But that's not what the report says.
"The following information comes straight from Pew Research, quote, 'Approximately 24 million people — one out of every eight — voter registrations in the United States are no longer valid or significantly inaccurate.' One in eight," Trump said. "More than 1.8 million deceased individuals, right now, are listed as voters.' Oh, that's wonderful."
"Well, if they're gonna vote for me, we'll think about it, right?" Trump joked. "But I have a feeling they're not gonna vote for me. Of the 1.8 million, 1.8 million is voting for someone else."
Trump accurately quoted from the report, "Inaccurate, Costly and Inefficient: Evidence That America's Voter Registration System Needs and Upgrade." But the report did not allege the 1.8 million deceased people actually voted. Rather, Pew said that it is evidence of the need to upgrade voter registration systems.
In fact, researchers say voter fraud involving ballots cast on behalf of deceased voters is rare.
The Help America Vote Act of 2002 encouraged states to improve the accuracy of their registration lists and to audit their election results. As a result, Minnite told us in a phone interview, a number of states compared their voter lists to the Social Security Death Index, and in some cases they turned up hundreds or even thousands of apparent instances of "dead people" voting.
But with a bit of digging, almost all of those turned out to be due to clerical errors or as a result of people who legally voted via absentee ballots or the early voting process but later died before Election Day, Minnite said.
For example, in 2012 South Carolina's attorney general notified the U.S. Department of Justice of potential voter fraud after finding 953 ballots cast in the 2010 election by voters listed as deceased, in some cases as long as six years. The finding ran in the Augusta Chronicle at the time in an Associated Press story under a headline, "South Carolina attorney general informs Justice Department of voter fraud."
But a subsequent review by the State Election Commission found no evidence of fraud and that mostly the cases were clerical errors.
In a letter to the attorney general, the executive director of the State Election Commission wrote that it only had the resources to investigate 207 cases from the most recent 2010 election. Of those cases, it found 106 cases were the result of clerical errors by poll managers; 56 cases were the result of bad data matching, meaning that the person in question was not actually dead; 32 cases were "voter participation errors," including stray marks on lists erroneously indicting they had voted; three cases were absentee ballots issued to registered voters who cast ballots and later died before Election Day; and 10 cases contained "insufficient information in the record to make a determination."
Cases of people actually voting fraudulently on behalf of deceased people are rare — though isolated examples have occurred, Minnite said.
"There are a handful of known cases in which documentation shows that votes have been cast in the names of voters who have died before the vote was submitted," wrote Justin Levitt in a 2007 report, "The Truth About Voter Fraud," for the Brennan Center for Justice. "It is far more common, however, to see unfounded allegations of epidemic voting from beyond the grave."
Much of the misinformation about "dead people voting" is due to "flawed matches from one place (death records) to another (voter rolls)," Levitt found. Levitt explored five reports of widespread fraud regarding "dead voters" and found all of them were unfounded or greatly exaggerated.
The second leg of Trump's allegation of widespread voter fraud rests on "illegal immigrants voting," another practice that experts told us is fairly rare.
In Wisconsin, Trump cited a 2014 Washington Post article titled "Could non-citizens decide the November election?" It was a piece penned by Old Dominion University professors Jesse Richman and David Earnest about research the two later published in the journal Electoral Studies. It turns out to be a disputed and very controversial study.
Trump accurately quotes from the blog post. But the authors' results are contested by a number of academics, including those who administer and manage the data on which it is based.
The study relied on data from the Cooperative Congressional Election Study, which is administered by YouGov/Polimetrix and managed by Harvard University and the Massachusetts Institute of Technology. Richman and Earnest estimated the number of noncitizens who voted nationwide based on those in the survey who self-identified as noncitizens who voted.
"Our best guess, based upon extrapolations from the portion of the sample with a verified vote, is that 6.4 percent of non-citizens voted in 2008 and 2.2 percent of non-citizens voted in 2010," Richman and Earnest wrote in the Post.
In a blistering rebuke of that study, the managers of the database on which the article by Richman and Earnest was based wrote in Electoral Studies that "measurement errors" in the survey led to a "biased estimate of the rate at which non-citizens voted in recent elections. The results, we show, are completely accounted for by very low frequency measurement error; further, the likely percent of non-citizen voters in recent US elections is 0."
"Their finding is entirely due to measurement error," one of the authors, Stephen Ansolabehere of Harvard and the principal investigator of CCES, wrote to us in an email. "Measurement errors happen. People accidentally check the wrong box in surveys. The rate of such errors in the CCES is very small, but such errors do happen. And when they do happen on a question such as citizenship, researchers can easily draw the wrong inference about voting behaviors. Richman and Earnest extrapolate from a handful of wrongfully classified cases (of non-citizens)."
How do the database managers know? Ansolabehere explained, "We asked people in successive years their citizenship. That minimizes the error. Upon doing so we find NO INSTANCES of voting among people stating consistently that they are non-citizens."
"The CCES conducts a panel (repeated interviews of people asking the same questions) and vote validation," Ansolabehere said. "We found that NONE of the 85 individuals in the 2010-2012 panel survey who indicated that they were non-citizens in 2010 and again in 2012 in fact voted.
"The Richman and Earnest study is an incorrect use of the survey that we manage, and a false claim of evidence of non-citizen voting. It's a dangerous, stray false-fact."
We reached out to Richman, who told us that he and Earnest "stand by our study, but we encourage people to read the critiques too."
Richman said Ansolabehere's figures were off, and that there was "one individual in 2012 who identified as a non-citizen in both 2010 and 2012 and also cast a validated vote in 2012." And, he said, "There were also ten individuals who twice stated that they were non-citizens and also stated that they voted in 2012. There were also in 2014 two out of about 26 individuals who three times confirmed that they were non-citizens who also said they voted."
Nonetheless, Richman credited Ansolabehere and his colleagues for their "valuable contributions" to the issue, which he said "highlighted the fact that response errors could have biased our estimates upwards." Richman said he remains "unconvinced" that Ansolabehere's evidence shows there was no voting by noncitizens, which he said "simply [does] not fit with the data."
"Concerning the broader fight Trump has raised, I think our results are getting taken out of context in important ways by people on the right who want to make an unsupported claim concerning massive vote fraud," Richman told us via email. "We found low levels of non-citizen participation in elections. These levels are sufficient to change the outcomes in extremely close elections. But one should keep in mind that such elections can be swayed by any number of factors that arguably bias election results toward, or against, particular parties and candidates. Put another way, our results suggest that almost all elections in the US are not determined by non-citizen participation, with occasional and very rare potential exceptions.
"Both sides of the debate on non-citizen voting have exaggerated our findings concerning non-citizen representation," Richman added. "There are many on the left side of that debate who have relentlessly sought to discredit our results and want to push the level of estimated participation to zero. On the right there has been a tendency to misread our results as proof of massive voter fraud, which we don't think they are."
Researchers who have studied the issue have found relatively few cases of noncitizens voting. That's not to say it never happens. It has. A 2015 report from the conservative Heritage Foundation documented less than a dozen individual cases of noncitizens convicted of registering or actually voting since 2000. And in Texas, a city councilwoman was sentenced in 2007 to five years in prison for registering noncitizens to vote.
But such cases are rare, experts told us. Sarah Pierce, an associate policy analyst with the Migration Policy Institute, said there's very little evidence of it, in part because the disincentives are enormous. It is an illegal offense for an unauthorized immigrant to vote — a deportable offense that makes a person permanently inadmissible for return to the U.S., she said.
Trump argued that as a result of dead people and noncitizens voting, "voter fraud is very, very common." It isn't, according to numerous researchers who have studied the issue — including the Old Dominion professor whose work is cited by Trump.
To be sure, voter fraud happens. After the 1982 election, in Chicago, 62 people, most of them precinct captains, were indicted by a grand jury for stuffing ballot boxes and buying votes, including a scheme in which they would identify registered voters not voting on Election Day and forge ballots in their name.
But in a speech in Manheim, Pennsylvania, on Oct. 1, Trump urged his supporters to "watch your polling booths because I hear too many stories about Pennsylvania, certain areas. … We can't lose an election because of, you know what I'm talking about." On Oct. 10 in Ambridge, Pennsylvania, he asked supporters to "watch other communities, because we don't want this election stolen from us." That suggests a problem with what voting experts call voter impersonation.
In an interview on CNN's "State of the Union" on Oct. 16, former New York Mayor Rudy Giuliani, a Trump surrogate, talked about busloads of people voting numerous times in some big cities. He also quipped that "dead people generally vote for Democrats, rather than Republicans."
"We have people who cheat in elections," Giuliani said, particularly in "inner cities" controlled largely by Democrats.
And on CNN on Oct. 17, Rep. Steve King, a Trump backer, insisted, "There is significant evidence out there that there is voter fraud."
But is there? Many election experts say the kind of voter fraud Trump is talking about — voter impersonation — is extremely rare, and not enough to tip even a close presidential election. And there is plenty of research to back that up.
"The best facts we can gather to assess the magnitude of the alleged problem of voter fraud show that, although millions of people cast ballots every year, almost no one knowingly and willfully casts an illegal vote in the United States today," Lorraine Minnite writes in her book, "The Myth of Voter Fraud."
Allegations of "busloads" of people going from polling place to polling place — such as Giuliani described — is a common urban myth, Minnite said. She has heard tales of busloads of college students coming into New Hampshire to vote, and about busloads of Mexicans from Oklahoma voting in Kansas. And in every case — including Giuliani's, she said — there is no evidence for them.
"These sort of fictions about busloads of people, you hear about it a lot on the right," Minnite said. "It is just very unlikely. Think about how and why it would happen. It makes no sense."
You'd have to know the person you were impersonating hadn't voted yet, and that the person at the poll doesn't know that person, she said. And in a busload of people, you'd have to count on every one of them keeping quiet.
"And for what? What is the benefit of it?" Minnite said. There is very little payoff with the potential for a felony conviction. And in the case of immigrants in the country illegally, the risk of permanent deportation."
A December 2006 report by the U.S. Election Assistance Commission interviewed more than two dozen researchers and experts on voter fraud and intimidation, including Minnite. That report concluded that "impersonation of voters is probably the least frequent type of fraud because it is the most likely type of fraud to be discovered, there are stiff penalties associated with this type of fraud, and it is an inefficient method of influencing an election."
In an Aug. 16, 2014, article for the Washington Post, Loyola Law School professor Justin Levitt, currently on leave to work with the Department of Justice overseeing voting, wrote that he has been tracking allegations of voter fraud for years, including any "credible allegation that someone may have pretended to be someone else at the polls, in any way that an ID law could fix."
"So far," he wrote, "I've found about 31 different incidents (some of which involve multiple ballots) since 2000, anywhere in the country. … To put this in perspective, the 31 incidents below come in the context of general, primary, special, and municipal elections from 2000 through 2014. In general and primary elections alone, more than 1 billion ballots were cast in that period."
In 2012, a team of students led by the Walter Cronkite School of Journalism and Mass Communication at Arizona State University analyzed 2,068 alleged election-fraud cases since 2000 and concluded that "while fraud has occurred, the rate is infinitesimal, and in-person voter impersonation on Election Day, which prompted 37 state legislatures to enact or consider tough voter ID laws, is virtually non-existent."
In October 2002, the Justice Department in the George W. Bush administration introduced the Ballot Access and Voting Integrity Initiative, which was charged, in part, with targeting voter fraud. But as Washington Post columnist Harold Meyerson pointed out in 2007, the efforts between October 2002 and September 2005 resulted in "just 38 cases [being] brought nationally, and of those, 14 ended in dismissals or acquittals, 11 in guilty pleas, and 13 in convictions."
Wrote Meyerson: "Though a Justice Department manual on election crime states that these cases 'may present an easier means of obtaining convictions than do other forms of public corruption,' federal attorneys have failed to rack up those convictions, for the simple reason that incidents of fraud have been few and far between."
We reached out to several other voting fraud experts for input on the claims made by Trump and his surrogates about voter fraud.
"These allegations are false. Fraud almost never takes place through in-person voting (and certainly not enough to swing an election)," Heather Gerken, a professor of law at Yale Law School and an election law expert, told us via email.
Gerken, who was a senior legal adviser to the Obama campaigns in 2008 and 2012, added, "There have been astonishingly few examples, and with good reason. It is far, far, far easier to steal an election by bribing an election official or mail-in/absentee voting. Unsurprisingly, all the evidence of serious fraud concerns the latter sort."
And Nathaniel Persily, a professor of law at Stanford University who worked on a bipartisan presidential commission formed after the 2012 election to address the issue of long lines at polling places, told us, "There is very little voter fraud. There is virtually no voter impersonation fraud. Where there is fraud it is usually done at the wholesale rather than retail level — usually with absentee or mail ballots."
Richard L. Hasen, a professor of law and political science at the University of California, Irvine School of Law, and author of "The Voting Wars," referred us to a piece he published on the subject on Slate on Oct. 17, which ran under the headline "Voter Fraud: It would be literally insane to try to steal an election in the way Donald Trump is alleging."
"The truth is, though, that not only does zero evidence exist that this sort of fraud has taken place on any regular basis, but multiple voting simply cannot happen in any practical sense on a scale to influence a presidential election," Hasen wrote. "To vote five, 10, or 15 times one would have to either register five, 10, or 15 times in different jurisdictions or with false names or go five, 10, or 15 times to polling places claiming to be someone else whose name is on the voter rolls, in the hopes that this person has not already voted and you would not get caught. And to do this on a scale for a presidential election, in a place such as Pennsylvania with millions of voters, you would need to pay tens of thousands of people, all without any way of verifying how they voted."
Hasen writes that in researching his 2012 book, "The Voting Wars," he "could not find a single instance anywhere in the U.S. from the 1980s onward where massive impersonation fraud was used to try to steal an election."
Minnite said Trump is simply "tapping into popular ideas about elections being rigged and corruption." They may seem believable, she said, but upon closer inspection, most of the claims of widespread voter fraud fall apart. Mostly, she said, "it's fallacy."
Count Trump's campaign manager, Kellyanne Conway, among those unconvinced about widespread voter fraud. Conway said that while, in the past, there have been reports "here and there" about people voting "a couple different times in places," she does not believe widespread voter fraud will mar the upcoming election.