— Former secretary of state Hillary Clinton, at the annual “Bloody Sunday” commemorative service, Selma, Ala., March 3, 2019
“Just think about it: Between 2012, the prior presidential election where we still had the Voting Rights Act, and 2016, when my name was on the ballot, there were fewer voters registered in Georgia than there had been those prior four years.”
Clinton claimed 40,000 to 80,000 Wisconsinites were turned away at the polls because of their skin color, age or “whatever excuse” in 2016. This matters because President Trump won Wisconsin by a narrow 22,748 votes. So this speech in Selma had an air of “what might have been.”
It was the first presidential election since the Supreme Court struck down a key part of the Voting Rights Act, Clinton noted. This “made a difference in Wisconsin,” she said. Total voter registration declined in Georgia once this law was gone, she added.
Some studies show that voter ID laws favored by Republicans have a deterrent effect, especially on minority voters, who tend to vote for Democrats. Wisconsin enacted a voter ID law before the 2016 election. A study from the University of Wisconsin found that the new requirements deterred thousands of eligible voters in two counties from casting ballots.
Nevertheless, Clinton made several factual errors, offered questionable claims about a couple of studies, and ended up giving a misleading assessment of her loss.
In 2013, the Supreme Court in a 5-to-4 decision struck down Section 5 of the Voting Rights Act. This section of the law was designed to safeguard minority voters’ rights in places where they historically had faced discrimination at the polls. The law required nine states and several counties to seek authorization from the federal government before making changes to their election procedures. Writing for the majority in Shelby County v. Holder, Chief Justice John G. Roberts Jr. said Section 5 was based on decades-old data that no longer applied.
Wisconsin was not one of the states covered by Section 5 when the court ruled in 2013, so, right off the bat, Clinton’s claim that this “made a difference in Wisconsin” is unfounded. Georgia was covered by Section 5, but Clinton’s claim that total voter registration declined in that state from 2012 to 2016 is false; it increased. (Update: We previously said the Supreme Court struck down Section 4 of the law; it was Section 5.)
Wisconsin did enact a voter ID law under former governor Scott Walker (R). It requires voters to show a U.S. passport or a photo ID issued by Wisconsin state agencies, the military, Veterans Affairs, a university or college, or a federally recognized Indian tribe. Voter fraud cases are extremely rare in the United States. Experts say these voter ID laws are often attempts to tamp down the Democratic vote.
It’s a case Clinton made often during the 2016 campaign (though we note she did not visit Wisconsin after winning the Democratic nomination). “If there’s one place where we were caught by surprise, it was Wisconsin,” she wrote in her post-mortem book, “What Happened.” “Polls showed us comfortably ahead, right up until the end.”
Did this Wisconsin law prevent 40,000 to 80,000 voters in the state from casting ballots in 2016? We asked Clinton spokesman Nick Merrill to weigh in, and he pointed to several studies.
For the low end (40,000), Clinton was relying on a 2017 study led by Kenneth R. Mayer, a political scientist at the University of Wisconsin at Madison.
Mayer and researcher Michael DeCrescenzo surveyed 293 registered voters who didn’t vote in Dane and Milwaukee counties. That covers “the two largest metro areas in the state (Milwaukee and Madison) and have the largest low-income and minority populations, which research has shown are most likely to be affected by voter ID requirements.”
Their study mentions voters who were deterred, meaning “they lack qualifying ID or mention ID as a reason for not voting. Voter ID could be a nominal reason or the primary reason for not voting.” Mayer and DeCrescenzo estimated that anywhere from 11,701 to 23,252 voters were “deterred.” In his responses to The Fact Checker, Mayer said “14,000 is the equivalent of the most likely number.”
The researchers also analyzed a subset of these voters who were prevented from voting. This is a “stricter definition” for those who “lack qualifying ID or list voter ID as their primary reason for not voting.”
So, the “deterred” pool includes those who mentioned the voter ID law as a reason for sitting out the election. But those in the “prevented” pool mentioned it as the main reason. Both pools include voters who said they didn’t have a qualifying ID under Wisconsin’s law.
Mayer and DeCrescenzo estimated anywhere from 5,250 to 14,101 “prevented” voters.
Clinton mentioned voters being “turned away,” but the study defines in different terms what it means to be deterred or prevented. And the study says “the estimates cannot be extrapolated to the state of Wisconsin as a whole.”
Here’s where things start to get hairy. Mother Jones interviewed Mayer about the study and paraphrased him making just such an extrapolation.
The study’s lead author, University of Wisconsin political scientist Kenneth Mayer, says between roughly 9,000 and 23,000 registered voters in the reliably Democratic counties were deterred from voting by the ID law. Extrapolating statewide, he says the data suggests as many as 45,000 voters sat out the election, though he cautioned that it was difficult to produce an estimate from just two counties.
“We have hard evidence there were tens of thousands of people who were unable to vote because of the voter ID law,” Mayer told me.
Merrill pointed to the key line in the article: "Extrapolating statewide, he says the data suggests as many as 45,000 voters sat out the election, though he cautioned that it was difficult to produce an estimate from just two counties.”
When we reached out to Mayer, he said his study should not be interpreted like that. (“That’s not a claim that I made,” he said of the line in the Mother Jones article.)
“We were careful not to extrapolate statewide, as we do not have information about what the effects were outside of the regions we sampled,” Mayer wrote in an email. “Our goal was to understand the effects more precisely, among populations most likely to be affected. We also were careful, in that someone who did not vote because of the ID law might not have voted even if the law were not in effect.”
The Mother Jones article may have given Clinton a mistaken impression, but we always warn politicians: You are responsible for the claims you make, so double check the facts. We also told Merrill what Mayer said and didn’t get a response.
At the high end of the scale, the UW-Madison study estimates that 23,252 voters were “deterred” by the voter ID requirement. That’s just a hair above Trump’s 22,748 margin in the state. Mayer and DeCrescenzo did not ask survey respondents whom they would have voted for because their research was funded by the office of the Dane County clerk. In any case, Clinton said 40,000, not 23,000.
Where does Clinton get the 80,000 figure for the high end of her estimate?
Merrill cited a May 2017 memo by Priorities USA, a Democratic super PAC.
“While states with no change to voter identification laws witnessed an average increased turnout of +1.3% from 2012 to 2016, Wisconsin’s turnout (where voter ID laws changed to strict) dropped by -3.3%,” the memo says. “If turnout had instead increased by the national no-change average, we estimate that over 200,000 more voters would have voted in Wisconsin in 2016. For context, Clinton lost to Trump in Wisconsin by only 20,000 votes.”
But, as we previously noted in this Three Pinocchio fact-check, correlation isn’t always causation, and this study does not offer evidence of causation.
Rick Hasen, an elections expert at the University of California at Irvine, noted that black voters in Milwaukee, which experienced a dramatic decline in turnout in 2016, were not motivated by Clinton the same way they were for President Barack Obama in 2008 or 2012. Moreover, there was a general decline in the black vote in 2016, compared with 2008 or 2012, both in states that had voter ID laws and states that did not.
All that aside, how did 200,000 become 80,000?
“The 40,000 number comes from the more conservative estimate,” Merrill said. “The 80,000 number is also a conservative estimate well within the range of the two studies.”
Hasen was not buying Clinton’s claim. He told The Fact Checker:
"The claim is ludicrous because Wisconsin was not a state covered by the part of the Voting Rights Act, which the Supreme Court killed in the Shelby County case. So even if it had been in effect it wouldn’t have stopped implementation of Wisconsin voter ID, as it had stopped Texas’s voter ID.
"The numbers of people who were turned away from the polls is likely much lower than Clinton says. The most reputable study, which I think still overstates things, is the one by Ken Mayer at U. Wisconsin. But there is reason to be very cautious about trying to attribute the presence or absence of voter ID to turnout. A recent NBER study found no effect whatsoever.
“And it really is besides the point. The question is why Wisconsin is putting into effect a law which prevents no appreciable amount of fraud and puts stumbling blocks in front of voters.”
The National Bureau of Economic Research (NBER) study concluded that voter ID laws have had “mostly null” effects and that “the fears that strict ID requirements would disenfranchise disadvantaged populations have not materialized.” But the study also challenged the Supreme Court’s reasoning for upholding one such voter ID law in 2008. “We find no significant impact on fraud or public confidence in election integrity,” the authors wrote. “This result weakens the case for adopting such laws in the first place.”
This study comes with a caveat: “Because states adopted strict ID laws only 2 to 12 years ago, our results should be interpreted with caution: we find negative participation effects neither in the first election after the adoption of the laws nor in following ones, but cannot rule out that such effects will arise in the future.”
Other studies, aside from Mayer and DeCrescenzo’s, indicate that voter ID laws do have an effect on turnout. A 2014 study by the Government Accountability Office analyzed Kansas and Tennessee and concluded that their voter ID laws decreased turnout for the 2012 elections.
“As you well know, it’s going to be some time before we know exactly how detrimental these laws are to a free democracy, but suffice to say that she was being conservative in her estimates, and it still indicates a real threat to our basic rights as citizens,” Merrill said. “And I’d note that that was the theme of the Selma Jubilee this year. All the speakers addressed these issues.”
The Pinocchio Test
There’s an important debate to be had over voter ID laws and their effect on turnout, considering how rare voter fraud cases are in the United States and the risk of disenfranchisement. We’re looking at something different here. Clinton made a series of specific claims that were way off-base.
The Supreme Court’s ruling in 2013 had no bearing on Wisconsin. The University of Wisconsin study she relied on for her 40,000 estimate says its findings from two counties should not be extrapolated to form statewide conclusions. Her spokesman did not cite any study for the 80,000 estimate. Voter registration in Georgia did not decline from 2012 to 2016.
Wrong on multiple levels, seriously misleading and worth a cumulative Four Pinocchios.
(About our rating scale)
Send us facts to check by filling out this form
Sign up for The Fact Checker weekly newsletter
The Fact Checker is a verified signatory to the International Fact-Checking Network code of principles