On Tuesday, Wisconsin voters will elect a Supreme Court justice in an election that could decide the ideological balance of the court until 2028. Although we will not be releasing a model for this race, the early vote data we have extensively analyzed and modeled suggests Democratic-aligned liberal judge Susan Crawford is a clear favorite to win this election.
While the GOP-aligned conservative Brad Schimel certainly could pull off an upset (and it would be foolish to rule him out this early), the early vote suggests that this would be a monumental feat for him, and he is staring down the barrel of an immense deficit that he is unlikely to overturn on election day. The early vote data shows that even relative to 2024, Democrats are significantly more likely to show up than Republicans are — and in a state that Donald Trump won by less than a percentage point in 2024, this makes the GOP’s job much harder.
Across the board, Democrats are simply showing up more. For instance, among voters who cast mail ballots in 2024, Democrats are more likely to have cast a mail ballot in 2025 than Republicans are.
It’s a similar story with in-person early voting, where among those who voted early in 2024, Democrats seem much more likely to vote early in 2025 than Republicans are.
As a result, our own internal modeling suggests that Schimel would likely need a massive amount of persuasion in order to flip this race, because he might well be facing an electorate that would have backed Kamala Harris by double digits. Based on the data available to us, a sizable Crawford victory is significantly more likely than a Schimel win is, and operatives we have spoken to on both sides concur with this assessment.
Why?
It’s simple: an election like this generally comes mostly down to turnout (and if Crawford wins, the vast majority of her gains over Kamala Harris will almost certainly be from that alone). And in the Trump era, Republicans have bled massively with higher-propensity white voters, replacing them with lower-propensity ones in the process. While this is fine for presidential elections, it has major consequences in every other election, when only the most engaged voters tend to show up. Increasingly, those voters are backing Democrats, thanks to the party’s firm edge with white, college-educated voters.
It is abundantly clear from even a cursory look at county-level data that education drives turnout — and this suddenly makes the Democratic turnout edge in Wisconsin extremely easy to understand.
This fits in with a broader trend, where areas that are high-income, high-education, and high-employment all tend to vote more, especially in off-cycle elections. While this would have spelled doom for Democrats just a decade ago, it has turned into a net benefit for them in the Trump era.
At a geographic level, the places that tend to show up the most in elections are areas like Dane County (Madison) and the Milwaukee suburbs (also known as the WOW counties — Waukesha, Ozaukee, and Washington). These regions are some of the most highly-educated and affluent ones in the entire state. They are also uniquely repelled by Trump and his brand, and the last eight years have seen these types of areas drift steadily towards Democrats, with Kamala Harris actually gaining on Joe Biden’s margins in the WOW counties.
This creates a math problem for Republicans. When GOP voters in the Northern Wisconsin and the Driftless fail to show up, their share of the electorate gets swallowed up by the Madison and Milwaukee suburbs —and the Madison suburbs, in particular, punch above their weight the most in off-year elections, yielding tens of thousands of “buffer” votes for Democrats. In 2024, Madison was just 16.5% of the electorate — but in 2023, it was 18.8%. That 2% gap, creating a boost in Democratic votes, is something Republicans can ill afford to see if they want to win in Wisconsin with their current coalition.
For the clearest case study of how Wisconsin Republicans have struggled to turn out voters in non-presidential years, just look at what happened in the Driftless area: a blue-collar, white region that has taken a hard turn to the right of late. Republican gains have been disproportionately concentrated among non-college, low-income whites, and while those voters turn the area red in presidential elections, they simply don’t show up in off-cycle years. This is why Trump’s 2024 gains were greatest in this area, whether compared to 2022 or 2023.
The Democrats in these areas are much more high-propensity than the Republicans are, and so the voters who do show up in non-presidential races, like ones held for the State Supreme Court, tend to be a lot bluer than the region as a whole. Add on the factor of downballot lag (these counties are ancestrally much more Democratic) and it explains why, between 2022 and 2023, the Democratic gains in the Driftless were some of the largest in the entire state.
The combination of poor turnout dynamics and renewed Democratic enthusiasm, fueled by anger over Elon Musk and Donald Trump’s recent actions, create a very difficult environment for Republicans. While we certainly do not rule out an upset, such an event would cut against every bit of available information we have. Virtually any Democratic candidate over the last decade would win with the electorate that is likely to materialize for this race.
For this reason, Susan Crawford is the favorite to win this election. Looking ahead to 2026, Republicans might find themselves in an even tougher spot for the gubernatorial race and the election for the third Congressional District, located in the Driftless Area (where Rebecca Cooke is likely to have a rematch against Derrick Van Orden). They will likely need to make substantial persuasion gains to offset the unfavorable turnout dynamics they face, because the Wisconsin GOP’s coalition doesn’t seem to show up in off-cycle elections.
I’m a computer scientist who has an interest in machine learning, politics, and electoral data. I’m a cofounder and partner at Split Ticket and make many kinds of election models. I graduated from UC Berkeley and work as a software & AI engineer. You can contact me at lakshya@splitticket.org
I develop election forecast models and share data-based political analysis. You can find my insights on X at @giaki1310 and contact me at giaki1310@gmail.com