Both Donald Trump and Tammy Baldwin translated their slender statewide victories into hypothetical 1-seat majorities in the Wisconsin State Assembly, each winning a majority of the vote in 50 of the chamber’s 99 districts.

Under the map used in the 2022 election, I estimate that Trump would have won a majority in 64 districts and Baldwin in just 36.[i] The table below shows the outcome in various maps proposed during the 2022 and 2024 redistricting disputes.

Although drawn by Gov. Tony Evers, the current maps are slightly less favorable to Assembly Democrats than the plans submitted to the Wisconsin Supreme Court by the progressive firm Law Forward and the Wright Petitioners.

However, the current map is far more favorable to Democrats than the map drawn by Evers’ own People’s Map Commission. That group, which used a nonpartisan criteria, proposed a map under which Trump would’ve won 55 seats.

Click here to receive John’s blog posts in your email inbox.

Of course, actual Democratic candidates only won 45 Assembly seats, falling 4 seats short of the total won by Kamala Harris and 5 short of Baldwin. This is because Republican Assembly candidates were (with exceptions) more popular than Donald Trump.

The graph below shows how each Assembly race compares with the outcome of the presidential race in that district.

Five GOP Assembly candidates won districts that Trump lost. They are as follows:

• 88^th (Green Bay area): Benjamin Franklin won 50.3% of the two party vote, compared to Trump’s 49.8%
• 61^st (Southwestern Milwaukee suburbs): Bob Donovan won 51.6% to Trump’s 48.9%.
• 53^rd (Neenah/Menasha): Dean Kaufert won 50.6% to Trump’s 47.8%.
• 51^st (Dodgeville area): Todd Novak won 51.7% to Trump’s 48.2%.
• 21^st (Southern Milwaukee suburbs): Jessie Rodriguez won 51.4% to Trump’s 47.9%.

One Democratic candidate carried a seat that Trump won. Longtime incumbent Steve Doyle carried the 94^th district (north of La Crosse) with 50.3% of the two-party vote, compared with 51.1% for Trump.

There were no split results in the State Senate. The closest race came in the 8^th district, covering Milwaukee’s northern suburbs. Here, Democrat Jodi Habush Sinykin won 50.8% of the two-party vote, compared with Harris’ 51.2%.

This graphic also shows the 17 odd-numbered districts where elections will be held in 2026. They are visualized as empty circles, plotted at the zero point on the y-axis.

Of the likely competitive seats in 2026, Harris and Baldwin both won all of them.

• The 31^st (Eau Claire) is currently held by a Democratic incumbent. Harris won it by 2.2 points, Baldwin by 4.8.
• The 17^th (Southwestern Wisconsin) is held by a Republican. Harris won it by 1.1 points, Baldwin by 4.7.
• The 21^st (southern Milwaukee metro) is held by a Republican. Harris won it by 1.3 points, Baldwin by 2.3.
• The 5^th (western Milwaukee metro) is held by a Republican. This is the rare district where Harris outperformed Baldwin. Harris won it by 6 points and Baldwin by 5.

Democrats will need to win 3 of the 4 districts listed above to win a majority of the State Senate in 2026.

As the careful reader may have noticed, the current State Senate map actually has a structural advantage for Democrats. Despite losing the state by about a point, Kamala Harris won a majority of the vote in 18/33 Senate districts.

This table shows the number of senate districts won by Trump and Baldwin under selected alternative map proposals. Again, the difference is striking.

Under the map used in 2022, Trump’s narrow victory in 2024 was enough to win a supermajority of senate districts. Under the new maps, he wins fewer than half.

Under the nonpartisan People’s Map Commission, he would’ve won 21 of 33 seats.

Two common stories about the 2024 election are (1) that the election reflects a global anti-incumbent wave, and (2) that it was a reflection more of enthusiasm for Donald Trump than Republicans generally. Regarding the latter point, Wisconsin GOP strategist Bill McCoshen argued, “This was a Trump wave, not a red wave. And Republicans should be very careful how they interpret the results of this.”

There is good evidence in support of both those perspectives. It’s true that incumbent parties—left, right, and center—have lost elections across both hemispheres. Donald Trump has enjoyed electoral success by turning out infrequent voters who show up less reliably during elections when Trump isn’t on the ballot.

Still, the data from Wisconsin’s state legislative elections show the limits of both those narratives. Of the 82 contested state assembly races, Donald Trump was more popular than the GOP assembly candidate in just 12. While incumbent parties and executives may be unpopular, some degree of incumbency advantage still clearly endures.

Data Note

This post uses unofficial 2024 election results. The results of state legislative races are taken from the Associated Press. I collected presidential and US Senate results from each voting tabulation district in the state (available here). To calculate the results of these races within each state legislative district, I intersected the centers of voting tabulation district polygons with the polygons defining state legislative districts.

Access source data for this post from this GitHub Repository.

Note: A previous version of this post incorrectly described the 88th district and showed the D minus R margin of the vote including third parties. Those errors have been corrected.

[i] I calculated this by first allocating unofficial 2024 reporting unit results into census blocks. I disaggregated reporting units to blocks using allocation weights derived from registered voter addresses in the L2 Voter File. Then, I used the block assignment files for each plan to aggregate votes.

Voter turnout in Wisconsin this year was very high—among the highest ever in presidential elections in Wisconsin. Unofficial returns show 3,422,802 votes cast for president, up from 3,298,041 in 2020.[i]

About 125,000 more people voted in 2024 than 2020. Meanwhile, the state’s over-18 population grew by an estimated 56,000 between 2020 and 2023.[ii]

We won’t know the 2024 population estimate until sometime next month. If the over-18 population increase is fewer than 69,000, that will indicate that the share of adults voting in 2024 exceeded 2020, which was already an exceptional year.

The Wisconsin Elections Commission calculates that 72.9% of adults voted in 2020, which is the third highest rate since their data begins in 1948. In 1966, they estimate 72.0% voted and in 2004, 73.2%.[iii] Depending on the 2024 adult population estimate, the 2024 election will likely join this group of extremely high turnout elections where just shy of 3-in-4 adults participated.

Keep in mind that turnout as a share of the voting eligible population is even higher. Here are some estimates from the University of Florida Election Lab.

In 2024, the Election Lab estimates that 76.4% of eligible voters in Wisconsin cast a ballot, compared to 63.5% nationally. In 2020 it was probably 75.0% in Wisconsin, compared to 66.4% nationally. Turnout in Wisconsin is usually higher than US average, but this year the gap grew especially large. In 2020, turnout was 8.6 points higher in Wisconsin, growing to 12.9 points higher in 2024.

The true voter turnout rate can only be estimated because the denominator (actually eligible citizens) is unknown. At least in a month or two, we’ll have more recent data for the state adult population, from which we can further refine these estimates of the eligible share.

Estimates for the adult population of counties and municipalities will become available later in 2025. The kind of granular data needed to estimate ward-level adult population turnout rates won’t be available until the 2030 census.

The absence of high-quality data will not prevent the creation of turnout statistics, nor should they, necessarily. We cannot wait for the 2030 census to begin figuring out which voters were inspired by the last campaign and which weren’t. My point in elaborating these difficulties is to warn you that the first turnout statistics you see will not be the final word on the matter.

If you would like to receive John’s Faculty Blog posts in your email inbox, subscribe here.

Here is what the limited data we do have can tell us about voter turnout around Wisconsin.

• The number of votes cast for president increased in 71 of 72 counties. Only tiny Menominee was the exception (18 fewer votes cast).
• The number of votes cast grew by more than the 2020-2023 increase in the adult population in 65 of 72 counties. The 7 counties where vote change lagged adult population growth are Bayfield, Vilas, Forest, Ozaukee, Menominee, Ashland, and Burnett.
• In Milwaukee County, the number of votes cast grew by 4,346, despite the adult population falling by about 11,900 between 2020 and 2023.
• In Dane County, the number of votes cast grew by 21,153, compared with an adult population increase of about 15,600 between 2020 and 2023.
• In Waukesha County, the number of votes cast grew by 7,756, while the adult population grew by an estimated 6,900 between 2020 and 2023.

The Census Bureau’s Population Estimate Program doesn’t provide adult population estimates for individual municipalities, so we’re stuck with total population estimates.

• In the City of Milwaukee, the number of votes cast grew by 560 votes, while the total population fell by about 15,800 between 2020 and 2023.
• In the City of Madison, the number of votes cast grew by 11,740, while the total population grew by an estimated 5,100 between 2020 and 2023.

It is not the case that turnout fell in Democratic-leaning communities and increased in Republican ones. There is no correlation between a county’s change in turnout and the share of the vote received by Donald Trump. There may be a slight correlation between the increase in turnout and the increase in Trump’s share of the vote relative to 2020.

High turnout helped Trump because a group of infrequent voters showed up and disproportionately supported him. These “low propensity” voters were spread evenly across the state, contributing to Wisconsin’s small but remarkably uniform 1.5-point swing toward Trump.

Note: This post was updated to include a new analysis of the correlation between turnout and partisan vote share.

[i] The number of votes cast for president is slightly lower than the total number of ballots cast, which I have not collected. In 2020, 10,764 voters turned in a ballot but declined to vote in the presidential contest.

[ii] These numbers are from the Census Bureau’s Population Estimates Program and are estimated for July 1^st of each respective year. The estimates for July 1, 2024 will be released in December.

[iii] The WEC numbers use total ballots cast as their denominator (not presidential votes), and they use a slightly different estimate of the over-18 population. Their estimate comes from the Wisconsin Department of Administration which uses a somewhat different methodology and calculates their annual estimates for a different date than the federal Population Estimates Program.

What Ward Data Shows About Shifting Support by Education, Race, Age, Income, and More

Donald Trump won Wisconsin’s 2024 presidential election by just shy of 30,000 votes, after losing the state by 21,000 in 2000 and winning by 23,000 in 2016. Meanwhile, Tammy Baldwin won reelection by a similarly slim 29,000 votes. This, after Ron Johnson won his 2022 reelection campaign by 27,000 votes.

While other parts of the country (e.g., New York, New Jersey) saw big swings to the right, Wisconsin shifted more modestly. Trump eked out a win, but not by enough to pull Eric Hovde along with him. Republicans won most of the competitive seats in the State Assembly, likely keeping a comfortable 9-seat advantage. But Democrats swept all 4 State Senate targets, making them marginal favorites to win the chamber in 2026.

What the 2024 Republican swing lacked in magnitude it made up in its breadth across Wisconsin. Communities of all kinds moved toward the Republicans. Trump improved over 2020 in all but 4 of the state’s counties.

Trump increased his vote share since 2020 in 70% of the state’s municipalities and 62% of its wards. Here are 5 graphs showing how Wisconsin wards have trended since 2016.

Click here for an interactive map of all Wisconsin wards.

Education

As everyone knows by now, our electorate is increasingly polarized by education. The least college-educated places vote more for Trump and the most educated places have shifted toward the Democrats. That was the trend between 2016 and 2020 anyway.

A big part of why Trump won Wisconsin in 2024 is that he continued to make gains in the wards with the lowest levels of college education, while Harris only matched Biden’s performance in the most educated wards.

Still, the net gap has expanded from 22 points separating how the most and least-educated wards voted in 2016 to 33 points in 2020 and 36 points in 2024.

Age

Young people are more liberal and old people more conservative, but in recent years these differences haven’t grown much—at least at the ward level. The oldest wards moved toward Trump by about 1 point in 2024, similar to the state average. Notably, the youngest set of wards, while still very Democratic, shifted 3 points toward Trump.

Population density

You can guess a lot about a ward’s political lean just by knowing how densely populated it is. But the gap between sparsely and densely populated places didn’t grow in 2024. Trump’s performance improved by a point or two in both the least and most dense places.

Race/ethnicity

Wisconsin doesn’t collect race in its voter registration data, so, to make this graph, I aggregated 2020 census block data into ward boundaries. By my count, there are 153 majority Black wards, 53 majority Hispanic/Latino wards, and 3,063 majority (non-Hispanic) white wards.

Taken as a whole, the majority white wards lean slightly Republican. Trump’s 2024 vote share fell halfway between his 2016 and 2020 performance. Baldwin’s vote share fell 9 points from 2018, but improved by 2 points over Mandela Barnes’ 2022 senate run.

In the majority Black wards, the Democratic vote share remains very high but has slipped by 1-3 points in each of the past elections.

The big change is in the relatively small set of majority Hispanic wards. The Democratic margin of victory fell from 61 points in 2016 to 52 in 2020 and 42 in 2024.

This sort of correlation runs the risk of the ecological fallacy, but the patterns here are consistent with trends in other cities and survey data.

Income

My final graph is possibly the strangest out of this set. It shows partisan support by per capita income, with the poorest wards on the left and the wealthiest on the right.

It is exactly these two kinds of places that are the base of the Democratic party. Democrats are strongest in the poorest fifth of wards, followed by the wealthiest fifth. Lately, Republicans have won everything in between.

Setting aside their baseline level of support, this graph also shows which places are growing more or less enthusiastic about the parties. Donald Trump and Hilary Clinton tied in the wealthiest fifth of wards in 2016. Biden and Harris each won them by 6 points.

Meanwhile, the Democratic margin of victory in the poorest fifth of wards slid from 19 points in 2016 to 16 in 2020 and 12 in 2024.

About this data

Throughout this article I use “wards” to refer to what are technically “reporting units.” In Wisconsin, municipal clerks in small towns are allowed, under certain circumstances, to combine multiple wards into larger reporting units. This reduces administrative load and helps protect voter privacy.

Ward data is published on election night by most county clerks, but it is not collected and standardized by the Wisconsin Election Commission for several weeks following each general election. Lacking data on the state’s 3,500-odd wards, most election analysis relies on the totals from Wisconsin’s 72 counties—a limitation which makes analysis like the above impossible.

To fill this gap, I have collected and standardized ward-level returns for the Presidential and US Senate race in (currently) 69/72 counties, covering 98% of voters. You can download a GEOJSON file with these ward boundaries, unofficial 2024 election results, and disaggregated past election results from 2012-2022.

Special thanks to Anna Balliekova, Charles Ruleis, Mitchell Henke, Charles Franklin, Marybeth McGinnis, Ben Welden, Kevin Goldfarb, Alex Ballwanz, Ben Iberle, and Dwight Maynor for their help tracking down ward files on election night.