Record turnouts after voter ID law passed? Not really

Citing concerns over voter fraud, Wisconsin Republicans implemented a voter ID law in 2011 requiring photo identification before casting a ballot.

Opponents blasted this as disenfranchisement, saying it disproportionately affected certain groups of voters and imposed unconstitutional barriers to voting. That prompted a yearslong legal battle that concluded only when the U.S. Supreme Court rejected a challenge to the law in 2015.

With Republicans gearing up to push an array of new election law changes, state Sen. Duey Stroebel, R-Cedarburg, raised the voter ID topic in a recent TV interview.

"We’ve had more voter turnout since voter ID was passed than ever before," Stroebel, who is leading the effort to introduce the new limitations, said in a March 7, 2021, appearance on WISN-TV’s UpFront program.

In other words, Stroebel is asserting the photo ID requirements that opponents on the left condemned as voter suppression in fact were followed by an increase in voting.

Let’s see if he has a point.

The first question is how we define turnout — a more complicated question than it might seem.

It’s easy to bungle the data. For instance, while trying to make the case for widespread fraud in 2020, Donald Trump Jr. said voter turnout in Wisconsin had jumped from 67% in 2016 to 89%. But he calculated the two figures differently, and we rated the claim Pants on Fire.

In defending Stroebel’s claim, his policy director, Brian Sikma, pointed to raw vote totals. But that’s a poor measure since the number of potential voters changes dramatically over time, due to population growth and other factors. Stroebel’s claim was that recent turnout is the highest "ever," but Wisconsin has twice as many potential voters as it did in the 1960s.

So we need a rate statistic — some reflection of the percentage of eligible people who voted, which allows us to compare different eras.

Many states use registered voters for this calculation, but that’s a tough measure to use in Wisconsin since the state allows same-day voter registration. That creates a significant lag between knowing how many votes were cast and knowing how many people were registered, which clerks report well after the election. 

But there are other problems as well.

"Some have argued that a strict voter ID law could actually discourage people from registering if those individuals believe that they lack proper ID to vote. If this sort or behavior occurs, then a law that reduces voter registration could misleadingly appear to produce higher turnout because fewer people are registered," said Barry Burden, director of the Elections Research Center at the University of Wisconsin-Madison. "In a similar fashion, the actions taken by the (Wisconsin Elections Commission) over the past few years to remove outdated or incorrect registration records from the statewide database could lead to misleading conclusions about turnout rates."

Because of such concerns, the elections commission calculates turnout as the number of voters divided by the number of voting-age adults. We’ll lean on that for our analysis, along with a similar calculation that divides by the number of eligible voters (which removes non-citizens, felons and other ineligible voters from the total).

Stroebel’s assertion is noteworthy since it is at odds with traditional wisdom on the subject — the expectation that adding requirements to voting will reduce the number of voters.

Some form of voter ID is now required in 36 states, according to the National Conference of State Legislatures.

A 2014 report from the nonpartisan U.S. Government Accountability Office found the changes had a mixed impact on turnout. They examined 10 studies that analyzed turnout before and after voter ID was implemented in various states: five studies found no connection between the two, four found decreases in turnout and one found an increase in turnout.

In a Wisconsin-specific study in 2017, Kenneth Mayer, a University of Wisconsin political science professor, estimated between 11,701 and 23,252 people did not vote in the 2016 presidential election because of confusion over voter ID requirements or lack of proper identification.

Featured Fact-check

Threads posts

stated on November 24, 2024 in a Threads post

Donald Trump lost the popular vote by 2%.

Conservatives disputed that finding, however. Will Flanders, research director at the Wisconsin Institute for Law and Liberty, told the Wisconsin Center for Investigative Journalism in 2018 the study "pushed a narrative" of voter suppression but did not actually prove it.

So did Wisconsin really post its highest-ever voter turnout in the years since voter ID took effect?

Burden said presidential elections are the best way to gauge the law’s effect, since other elections have a lower turnout and involve voters who "tend to be older, wealthier, more educated, and whiter, all factors that correlate with the likelihood of knowing about the law and possessing IDs."

We’ll start there, but will review all elections since Stroebel’s claim wasn’t specific to presidential contests.

In 2012 — the last presidential election before voter ID — 70.1% of the voting-age population turned out. That dropped to 67.3% in 2016 under voter ID, then rose to 72.7% in 2020, according to the elections commission.

The 2020 turnout figure is the third-highest since 1948, topped only by turnout around 73% in 1960 and 2004, according to data assembled by the elections commission. Turnout figures from the U.S. Elections Project — which calculates turnout based on eligible voters — showed 2020’s turnout was the highest since at least 1980, slightly edging 2004.

But let’s look closer at 2016.

Turnout rose overall in the U.S. that year to 60.1%, up from 58.6% in 2012, according to the Elections Project. So Wisconsin was bucking the national trend by declining in the first year of voter ID. Wisconsin had the second-largest decline in turnout in 2016 among the 50 states from 2012 to 2016.

That’s a big mark against the narrative Stroebel is laying out.

Taking a broader view, here’s how the other elections since 2016 stacked up, according to elections commission data:

• The gubernatorial elections, i.e. mid-term elections, show results in line with Stroebel’s claim. The 2018 turnout of 59.43% was the highest in the state dating back to 1950. That was part of a national surge, however. Nationwide turnout in that election was the highest overall since 1914 amid the polarizing tenure of President Donald Trump, according to the elections project.

• Turnout for the recent April presidential primaries has not been the highest ever. The 2016 turnout was fourth-highest since 1948, and 2020 was 10th-highest. 

• The August primary turnout was even lower compared to historical norms. Among the 2016, 2018 and 2020 elections, only the 2018 race was in the Top 10 since 1948. 

• The 2018 and 2019 spring elections, which were topped by Wisconsin Supreme Court contests, ranked second and third for turnout among the 10 contested races since 2010, but they were far short of the 2011 turnout in the race won by Justice David Prosser. 

If we narrow our focus to just the few elections before and after voter ID (as in the chart above), we see turnout rose since 2016 in the gubernatorial elections, presidential primaries and partisan primaries, while it stayed largely stable in presidential elections and fell short of the recent high in the state Supreme Court race.

But Stroebel didn’t just say turnout rose, he said it was the highest ever.

And on that, he is off the mark.

Stroebel said, "We’ve had more voter turnout since voter ID was passed than ever before."

The implication baked into this claim is that the law must not have hampered voting if turnout was still historically high. But the data doesn’t make that case, especially since turnout rose in the U.S. overall in recent years.

Turnout in the 2018 gubernatorial election was the highest ever — as part of a nationwide surge — but the other elections held since voter ID passed don’t show such a historic jump. And many weren’t close. In the presidential elections, turnout rose in 2020, but it dropped in 2016 even as voter turnout rose nationally.

We define Mostly False as "a statement that contains an element of truth but ignores critical facts that would give a different impression." That fits here.