It goes without saying that Democrats broadly exceeded pre-election expectations last Tuesday. Kentucky Governor Andy Beshear won reelection comfortably, improving relative to his 2019 performance across the state. In Mississippi, challenger Brandon Presley came closer to winning the gubernatorial race than any Democrat since 1999.
Their success wasn’t limited to gubernatorial races. In Virginia, Democrats held their Senate majority and flipped the House of Delegates — narrowly denying Republican Governor Glenn Youngkin a trifecta. In New Jersey, Democrats held both chambers and reversed their 2021 losses in the Assembly with a five-, possibly a six-seat gain. Pennsylvania Democrats held a Supreme Court seat and made further gains down ballot.
In all, we called just three legislative districts incorrectly in New Jersey and Virginia combined, and got all of our statewide predictions correct, making 2023 our most accurate election forecasting cycle to-date.
Perhaps the biggest surprise of the night came in Mississippi. Incumbent Republican governor Tate Reeves put up the worst performance of any Republican gubernatorial candidate in a generation, squeaking out a shockingly-close four-point win over Democratic challenger Brandon Presley, who nearly pulled off an unprecedented win in the Deep South through an exceptionally strong campaign buoyed by excellent Black support and turnout.
The aftermath of the gubernatorial election in neighboring Louisiana led to many Democrats sounding the alarm over catastrophically bad turnout and a sharp dip in Democratic support observed among Black voters in the state. Coupled with presidential polls finding Biden’s standing with nonwhite voters at an all-time low, this led to fears of a sharp and sudden exodus of Black voters from the party (a hypothetical phenomenon nicknamed “Blexit”). Presley’s performance, however, rebukes the apparent inevitability of realignment this theory promotes.
Why did the two races turn out so differently? To begin with, Democrats actually invested in Mississippi, where Presley nearly matched Reeves in spending, as compared to Louisiana, where Democrats spent just $1M in response to $26M of GOP spending. In other words, Democrats put up a campaign in Mississippi, whereas they effectively ceded the Louisiana governorship from the beginning.
Investment drives competition, and competition drives turnout. In the South, this can be observed through turnout and Democratic support shares among Black voters, and so it should come as no surprise that Democrats saw extremely robust numbers with them in Mississippi. In fact, Presley actually slightly improved on Jim Hood’s 2019 margins in majority-Black counties while simultaneously holding his own with whites (a feat in itself, considering that Hood only lost the Trump +17 state by 5).
In contrast, Democrats lost support among every group of voters possible in Louisiana, thanks to their invisible campaigns. Even majority-Black counties, which gave John Bel Edwards a combined 86% of the vote in 2019, saw a massive drop in Democratic support.
The difference in margins is only one component of the story, of course. Usually, crashing Democratic margins with Black voters are accompanied by a corresponding decline in turnout, and comparing 2019 to 2023 turnout levels for both states paints a very clear picture. Namely, while turnout was down across the board, the turnout crash was much worse in Louisiana, and it was most acute with Black voters — a county analysis shows that the more Black the county was, the greater its turnout decline from 2019 tended to be. Mississippi, meanwhile, showed no such correlation, as Black turnout was remarkably robust, albeit lower than 2019, across the state.
Looking at the data, we feel comfortable saying that the differences observed between the Mississippi and Louisiana gubernatorial races can be attributed to the differing levels of investment, outreach, and competition in the two elections. This probably doesn’t come as a surprise. After all, voter support is earned, not granted — there is nothing mandating that Black voters support the Democratic Party, and if the party does not make an effort, then they can expect declines across the board with this demographic. But if Democrats do put in the work, then there is no reason that they cannot maintain their remarkably high levels of Black support.
In Kentucky, incumbent Democratic governor Andy Beshear outran all of his fellow statewide Democrats by over 20% en route to a five point win over challenger Daniel Cameron, the attorney general. While nationwide partisan polarization has only increased since 2019, Beshear mostly managed to meet, and in some cases exceed, his previous margins in Kentucky’s rural counties.
Only two counties in this group flipped, voting Beshear-Cameron: Hancock and Carter. Even in Elliot, traditionally a Democratic bastion, Beshear won by 7 points despite a significant rightward swing. Beshear broadly improved his performance relative to 2019 in Kentucky’s southeastern coalfields, the western coalfields surrounding Henderson, and the traditionally Democratic Pennyrile.
Most interestingly, the value of leadership and incumbency in times of crisis was on full display in this election. Beshear defied realignment and made some of his strongest gains in the southeastern coalfield counties that were most heavily impacted by devastating flooding in 2022, including Breathitt, Perry, Owsley, and Lee.
Beshear also made major urban and suburban gains compared to 2019, with Jefferson (Louisville) and Fayette (Lexington) counties getting 5 and 11 points bluer, respectively. Oldham County, which contains several well-to-do suburbs of Louisville, got 4 points bluer — voting for Cameron by just under a point. And in Campbell and Kenton counties, both part of the greater Cincinnati metropolitan area, Beshear also made marked improvements.
Virginia and New Jersey
In 2021, Governor Glenn Youngkin pulled off a victory in Democratic-leaning Virginia, pulling Republicans over the line in races for lieutenant governor and attorney general. Youngkin’s victory also helped the GOP flip the House of Delegates, netting 7 seats to reach a 52-48 majority. Last night, Democrats won back the lower chamber.
Using the SHAVE method of combining the results from each House of Delegates seat and substituting in projected results for uncontested districts gives an approximate generic ballot of D +1.4%. That puts this election about four points to the left of the environment that brought Governor Glenn Youngkin in, where Republicans would have won the generic ballot by 2.7%.
In terms of seats flipped, Democrats gained 5 open seats compared to the Republicans’ 3. One incumbent, Republican delegate Karen Greenhalgh, lost reelection in HD-97. The GOP also had notable victories in HD-57, where Susanna Gibson was defeated, and HD-82, where incumbent Kim Taylor utilized incumbency to win reelection in a Democratic seat.
Democrats also held their Senate majority, though Republicans made a net gain of one seat. Long-time GOP incumbent Siobhan Dunnavant lost reelection comfortably in SD-16, which had a partisan lean too Democratic to overcome.
Democrats Russet Perry and Danica Roem picked up competitive open seats (SD-30 & 31) in NOVA. In SD-24, Republicans narrowly defeated sitting Democrat Monty Mason and in SD-27, delegate Tara Durant won an open seat, with independent Monica Gary pulling nearly 5% of the vote.
In New Jersey, George Norcross and his South Jersey Democrats had a triumphant return to center stage, contrary to our expectations. Ex-assemblyman John Burzichelli comfortably beat Senator Ed Durr in LD-03, avenging Steve Sweeney’s upset loss two years ago. The victory cancelled out Republicans’ automatic flip in LD-12.
In the Assembly, Democrats have gained five (and possibly six) seats. Most impressively, Democrat Avi Schnall won a seat in LD-30, a Trump +34 seat, thanks to bloc voting from Lakewood’s Orthodox community.
In Ohio, two Democratic-backed initiatives passed: the first enshrined abortion rights in the state constitution and the second legalized recreational marijuana eight years after a similar initiative failed by a wide margin. Issue 1 passed by 13.24%, while Issue 2 passed by a slightly-higher 13.65%, with a much more politically heterogeneous coalition than the former.
According to an NBC exit poll, Democrats were overrepresented in this electorate, which recalled voting for Biden by 2 percentage points. Both the crossover vote seen in Republican-leaning areas and the composition of the electorate suggest that abortion is a winning issue for Democrats, who can turn out base voters and persuade Republican and independent voters to give their candidates a second look.
In Pennsylvania, Democrat Daniel McCaffery comfortably held the seat of the late Justice Max Baer for his party, defeating Republican Carolyn Carluccio by seven points. His victory cements a 5–2 Democratic majority on the state’s Court. Democrats also won races for the Commonwealth and Superior courts, continuing their post-Dobbs winning streak in the state. It bears noting that Republicans have now comfortably lost each of the four major contested statewide elections since the Dobbs decision dropped — the closest they got was their four point loss in the Commonwealth court race.
Needless to say, this is an exceptionally worrying trend for the Pennsylvania GOP in arguably the most pivotal state in the nation. They’ll hope they can do better when Trump is on the ballot next year, but the losing track record his ideological allies have in the state means that we are somewhat skeptical of that improvement occurring easily.
Observers of this election may be somewhat puzzled by the divergence between the results, which show fairly strong Democratic numbers, and the current presidential polling, which is nothing short of apocalyptic for President Joe Biden, even against former President Donald Trump.
The theory that we have seen most frequently tossed around is the idea that the Democratic overperformances can be attributed mostly to turnout. As the Democratic Party trades working-class voters for college-educated ones, they also increase the effective voting propensity of their coalition. The “turnout” theory posits that the party tends to therefore perform best in low-turnout affairs like midterms, specials, and off-year gubernatorial and state legislative elections, as Republicans have become comparatively less likely to turn out.
We don’t necessarily subscribe to this idea; in fact, we recently wrote a piece on how the special election results probably do give us some directional insight on the environment. In either case, polling is infamously noisy from a year out, and while surveys can inform us about the current mood of the electorate, they don’t necessarily do a good job of predicting things one year from now. Campaigns exist for a reason, and they tend to define candidates on key issues that differentiate the parties and coalesce their base in response.
Many voters are currently dissatisfied with their current options — of this, there is little doubt. But young and nonwhite voters still are some of the most Democratic-leaning groups in the country, and in virtually every single contested election of late, they have backed Democratic candidates by overwhelming margins. To us, this suggests that as the presidential race truly takes shape, voters will come to view it as a choice between the two parties. It also helps explain why so many post-Dobbs Republican campaigns centered around abortion or the 2020 election tend to underperform — Democratic messaging on these two topics is especially potent and defines GOP candidates very unfavorably, while framing the election as a clear choice in a manner that pulls swing and Democratic base voters alike.
If the 2024 election is a referendum on President Biden, he would likely lose; his approvals are nothing short of catastrophic, and it is unclear that they will get better anytime soon. But if it is defined as a choice between him and his opponent, then we would lean towards saying that he is favored, and the less stark the choice is, the more we expect Republicans to overperform. Recent polling by the New York Times and Siena College certainly contributes evidence to this sentiment, as Nikki Haley and a “generic Republican” both blow Biden out of the water in hypothetical surveys in ways that Trump simply does not.
But given Donald Trump’s lead in GOP primary polling, he currently seems the most likely nominee, and the legal albatrosses surrounding him could sink his general election candidacy. Even in the aforementioned Times/Siena survey, Biden surges to leads in key swing states should Trump be convicted.
Since 2020, whenever Trump’s preferred candidates have been on the ballot, they have paid massive and historically anomalous electoral penalties. So in order to believe that Biden is at a massive disadvantage for 2024 against Trump, we would have to believe that the archetype and founder of the MAGA movement would uniquely overperform in a post-Dobbs and post-January 6th world, despite the former president having run into both even more negative favorability ratings and a looming set of legal repercussions since the 2020 election. And while this is extremely possible, it is something we currently think to be unlikely, especially given the specter of his criminal trials.
I’m a software engineer and a computer scientist (UC Berkeley class of 2019 BA, class of 2020 MS) who has an interest in machine learning, politics, and electoral data. I’m a partner at Split Ticket, handle our Senate races, and make many kinds of electoral models.
I make election maps! If you’re reading a Split Ticket article, then odds are you’ve seen one of them. I’m an engineering student at UCLA and electoral politics are a great way for me to exercise creativity away from schoolwork. I also run and love the outdoors!
You can contact me @politicsmaps on Twitter.
My name is Harrison Lavelle and I am a political analyst studying political science and international studies at the College of New Jersey. As a co-founder and partner at Split Ticket, I coordinate our House coverage. I write about a variety of electoral topics and produce political maps. Besides elections, my hobbies include music, history, language, aviation, and fitness.
Contact me at @HWLavelleMaps or email@example.com