Tammy Baldwin, a Democratic senator, managed to grasp victory on Tuesday’s re-election despite her association with Kamala Harris. This favourable outcome, peculiar to most of Wisconsin’s smaller counties, flourished extensively in areas where the commendable former President Donald Trump received an overwhelming support. These counties had demographics mostly delineated by lower income and education rates. The assertion of a high number of Baldwin-Trump voters in these areas would be an overstatement.
The distinction between the Senate and Presidential elections in Wisconsin was a mere whisper, less than 2 percentage points or 60,000 votes. The Republican’s triumphant stride in the Presidential race was by a narrow margin of 0.9%, while Democrats claimed the Senate with a comparable margin. Exit polls showed a minuscule crossover, with 4% of Trump voters opting for Baldwin, and a lower 3% of Harris’ supporters switching to Republican Senate nominee Eric Hovde.
It is interesting to note that Hovde, despite being part of the party that nominated Trump for President, secured about 54,000 less votes than Trump. On the other hand, Baldwin surpassed the votes for Harris albeit by a smaller gap. Indeed, merely 4,500 more votes statewide. A reflection of the impact of the small fraction of the Trump votes which Hovde failed to achieve.
There are three possible fates for these missing Trump votes. They may have gone to Baldwin due to party crossover. Alternatively, they might have been distributed to the Senate third-party candidates Phil Anderson and Thomas Leaguer, noting they together gathered about 71,000 votes or 2% of the total vote. The third possibility, yet equally plausible, is that some Trump voters elected to bypass the Senate race altogether, as evidenced by approximately 21,000 fewer votes in the Senate race.
Determining the exact distribution of these votes amongst these categories, merely through election results, is however a challenge. Despite this ambiguity, one can derive inferences about areas where Baldwin significantly surpassed Harris, and Hovde fell short of Trump. Post-election analysis by Marquette Law School fellow John Johnson, combining demographic data with ward-level election returns, provides insightful revelation about this disparity.
Mystery shrouded around, where Baldwin outperformed Harris? It was astonishing to find Baldwin stepping ahead of Harris by three points in election units with lower rates of education. While losing overall, Baldwin managed to curtail her margin of loss compared to Harris. Filtering for higher educational wards, Baldwin and Harris enjoyed a comfortable victory, with no noticeable divergence between their individual performances.
We noted similar patterns within income brackets as well. Lower income areas saw Baldwin take lead over Harris and, interestingly, no such gain was registered in high income areas, where both Democrats nevertheless performed decently. Baldwin also presented a slightly superior performance than Harris in both majority-white wards, home to a vast majority of Wisconsinites, and the majority-Black wards. However, this divergence climbed significantly in the majority Latino wards, which further tilted towards Trump in this election.
Albeit contributing a tiny share to the overall Wisconsin votes, the majority Latino wards can’t account for much of the difference between the two races. Similarly, wards with younger populace saw Baldwin pull ahead of Harris, spectator to the Democratic shift towards Trump, and accounted for a smaller divide between Trump and wards with older inhabitants.
Analyzing these patterns brings us to the conclusion that Baldwin relocated the gaps to areas where Trump demonstrated significant strength or massive inroads in this election. This association is even more evident across Wisconsin’s 72 counties. Small counties like Lafayette, Crawford, Buffalo, Ashland, Forrest, and Clark, witnessed an outperformance of Baldwin over Harris, simultaneously underperforming Hovde in comparison to Trump.
The Democratic margin in the Senate race in these counties outweighed their Presidential race margin by 5 to 8 points. In Lafayette, the Democrats loss difference was 20 points for the president and just 12 points for the Senate. Similar figures were noted in Crawford County as well, with a loss difference of 14 points for the President and only 8 points for the Senate. These counties marked some of Trump’s largest gains in 2024.
Lafayette witnessed the largest enhancement in Trump’s point margin from 2020 of any Wisconsin county, marking an increase from just under 14 to just over 20. Crawford closely followed with the third most significant increase, from an 8-point edge to a 14-point edge. Prior to 2016, both these counties leaned Democratic for president. Perhaps the alluring charisma of Trump drew new voters to the polls, or enticed some traditional Democrat voters who were not as committed to voting Republican down-ballot.
Contrasting examples include Republican Congressman Derrick Van Orden, who secured a tight re-election victory in western Wisconsin’s Third Congressional District but lagged noticeably behind Trump. However, it is worth noting that counties where neither Baldwin leapfrogged Harris nor Hovde lagged Trump reveal a starkly different scenario.
Hovde even had a better margin than Trump in few counties: Menominee Indian Reservation, suburban ‘WOW counties’ like Waukesha, Ozaukee and Washington surrounding Milwaukee. All three WOW counties supported Trump, but were among a handful of locations in Wisconsin where Trump faced a diminished performance compared to 2020.
These counties are home to many suburban communities, with a high count of college graduates, that have slowly distanced themselves from GOP in the era of Trump. It is thus logical to consider these territories as areas where a slightly larger share of Republican-leaning voters backed Hovde but refrained from supporting Trump.
Alternatively, Trump notably outperformed Hovde in the more rural north, west, and center of the state, witnessing the most substantial voter shifts in Wisconsin over the past decade, similar to Republican Ron Johnson in 2016. In conclusion, the outcome in Wisconsin was historically uncommon – different parties winning for the president and Senate on the same ballot, a spectacle unseen in the last 56 years.
Although, there was nothing unusual about the small size of the gap between these two races. The act of ticket-splitting did not re-emerge significantly in this election, but with elections as close as these, it takes minor ticket-splitting to cause a divided result. Interestingly, it wasn’t just Wisconsin experiencing this pattern, Michigan and perhaps Nevada and Arizona seem to follow suit as well, where the presidential race losers emerged victorious in the Senate race.