Just before Wisconsin’s Democratic gubernatorial primary, polls painted a very different picture from the final result. One of the state’s most trusted pollsters had Francesca Hong ahead by a wide margin, but on election day David Crowley eked out a narrow victory. The miss has renewed questions about how pollsters model primary electorates.
The Marquette University Law School poll, widely regarded as historically accurate, conducted its final survey July 22-27 of that year. It showed Hong with 38% support, while Milwaukee County executive David Crowley trailed far behind at just 7%. By election day, Hong captured 39.3% of the vote, but Crowley surged to 39.8% to take the nomination.
According to PJ Media, the polling failure wasn’t about bias. Rather, it was a “massive methodological error.” The turnout models used by pollsters were off by a wide margin. The miss echoed the 2016 general election, when pollsters failed to catch the final wave of support for Donald Trump from voters who hadn’t been polled in the last 72 hours of the race.
The same issue appeared in Michigan’s Democratic Senate primary, where Abdul El-Sayed vastly underperformed his polling numbers, making the race far closer than surveys suggested.
Several dynamics complicated this year’s Wisconsin race. Crowley had dropped out in July, only to re-enter nine days later after Lt. Gov. Sara Rodriguez ended her campaign. That created what PJ Media calls an “X factor” that pollsters wrestled with until the end.
The Marquette poll was within the margin of error for Hong’s support, but it also showed an unusually high 34% of voters undecided. Older voters were the least supportive of Hong and the most undecided age group, according to the poll.
“It is well known that voters in Wisconsin tend to make up their minds, tend to start paying attention to primary elections pretty late in the game,” Anthony Chergosky, an associate professor of political science at the University of Wisconsin-La Crosse, told PJ Media.
The poll’s turnout model also assumed younger voters would turn out at much higher rates than they did. Instead, the electorate looked like a typical Wisconsin primary: older and more moderate. Those voters were more likely to remain undecided until close to the end, said the Marquette pollster, Charles Franklin.
Late Endorsement Proved Pivotal
Another factor was the very late endorsement of Crowley by Gov. Tony Evers. In the Marquette final poll, more than one-third (36%) of Democratic primary voters said that endorsement made them more likely to support Crowley. Evers carried a sky-high approval rating of 85% among Wisconsin Democrats.
Even as an internal poll showed Crowley trailing Hong last week, his campaign manager predicted the polls wouldn’t reflect the primary outcome. In an Aug. 5 memo to supporters obtained by The New York Times, campaign manager Garren Randolph argued that Wisconsin surveys overrepresented voters aged 18 to 45. In some cases, polls expected twice as many young voters as actually showed up in 2018 and 2022. Hong was strongest among young voters, according to the internal poll.
Pollsters also grapple with what is known as nonresponse bias, where certain voters are more or less likely to respond to a poll, The New York Times reported. A similar phenomenon likely affected Michigan’s race, said Adam Carlson, a Democratic pollster and founding partner of Zenith Research.
“El-Sayed’s supporters were more fired up and willing to take polls down the stretch, but more enthusiasm does not necessarily equal more votes,” Carlson wrote in a post-mortem analysis Monday.
The polls were spot on for Hong’s support, PJ Media notes. What they missed was the surge of undecided voters moving strongly to Crowley. As with Trump in 2016, the DSA-backed candidate may represent a new kind of political force that pollsters haven’t yet learned to measure.
Source: pjmedia.com — https://pjmedia.com/rick-moran/2026/08/13/why-were-the-polls-in-the-wisconsin-democratic-governors-race-so-wrong-n4956097
This article is for informational purposes only and does not constitute financial, investment, tax, or legal advice. Do your own research and consult a licensed professional before making financial decisions.
