Fact check: Was there election fraud in Wisconsin in 2020?

The 2020 United States election produced a whole family of claims that the process failed. Claims like that deserve to be checked rather than shared, and the two loudest ones concern Wisconsin, where Joe Biden held the lead.

The two claims:

• How could turnout in parts of Wisconsin exceed 90%, even 100%?
• Why did Wisconsin’s turnout jump so drastically - by “up to five standard deviations” - compared with previous years?

Does turnout above 100% prove the election was stolen?

Mike Coudrey posted a tweet that read: “BREAKING: Wisconsin has more votes than people who are registered to vote. Total number of registered voters: 3,129,000. Total number of votes cast: 3,239,920. This is direct evidence of fraud.”

Coudrey claimed that Wisconsin recorded more votes than it had registered voters. He later deleted the tweet, but an archived copy is preserved here.

Two things about American elections matter here: most states require voters to register in advance, and Wisconsin also allows registration on election day itself.

The first number Coudrey relied on, 3,129,000 registered voters, is wrong. According to the Wisconsin Elections Commission, the state had 3,684,726 registered voters as of 1 November 2020. That is 555,726 more than Coudrey claimed, and more than enough to cover the gap he was pointing at.

And that registration figure is itself only a snapshot from 1 November. Because Wisconsin offers same-day voter registration, anyone who had not registered by then could still do so at the polling place on 3 November. The final registration total was therefore higher still than the commission’s 1 November snapshot.

So the claim of election fraud is simply false. Basic arithmetic and a basic understanding of how Wisconsin’s registration system works are enough to show it.

How could the town of Russell report 101% turnout, and Sheboygan County over 90%?

The same explanation applies. The pre-election registration counts date from 1 November; on election day itself, many more voters registered on the spot, which can push turnout measured against the old registration snapshot past 90%, or past 100%.

Why did Wisconsin’s turnout deviate so sharply from previous years?

The president’s son, Eric Trump, amplified a series of posts, among them one from Harlan Hill, an adviser to President Trump (the tweet).

If the link no longer resolves, here is a screenshot:

A jump to 89.25% turnout, when previous years hovered around 70%, does catch the eye.

But check the arithmetic before leaping to the conclusion that happens to support your side of the argument, left or right.

Turnout is calculated by dividing the number of people who actually voted by the number of people eligible to vote. The eligible population (broadly, everyone over 18) is always larger than the registered population. In Wisconsin, the eligible-voter population was 4,536,293 (election commission data, 2019). Of those, 3,684,726 had registered by 1 November, roughly 850,000 fewer than could have.

Turnout, then, is the votes cast (3,278,963 as counted on 5 November) divided by the eligible population (4,536,293):

3,278,963 / 4,536,293 = 0.7228. Wisconsin’s turnout was 72.28%, far lower than the 89.25% claimed by the president’s adviser.

And 72.28% is no meaningful deviation from earlier years. In fact, by Hill’s own figures, turnout in 2004 was higher: 73.24%.

Hill divided the wrong numbers. For previous years he used the eligible population as the denominator; for 2020 (deliberately or not) he used the registration snapshot instead, and the “anomaly” comes from that substitution.

This analysis was first published in Croatian on 7 November 2020, while the count was still under way. The Croatian original remains in the archive, preserved as written.