The 2024 presidential cycle has been defined by two things. The first is its length; in previous cycles, the matchup only became clear in March or April, but this time around, the scene was arguably already set by the beginning of the year. That means that this has been the longest general election “season” in recent memory.
The second is its sheer stability — seemingly every national poll consistently shows a rightward swing of roughly six points, on average. And while Biden has lost less ground in the older, whiter Midwest than he has in the Sun Belt, both sets of states have seen rightward swings at the presidential level.
This has understandably led to a rather static manner of coverage; when virtually every single survey says the same thing, the tone is easily set, and much of the recent analysis has focused on why Biden has seemingly been unable to make any headway against a candidate he beat four years ago as a challenger.
I like to describe this as the “Nothing Ever Happens” worldview. Proponents of this lens tend to believe that everything is fixed, viewpoints are set in stone, and there is little variance in what happens, no matter what candidates or campaigns do. In their minds, voters do not respond to campaigns or advertisements, and filter their understanding of every single macro issue through a partisan filter before forming judgments.
That sentiment probably does describe a few things accurately, especially when it comes to partisans. But it misses one critical point: even though independents and soft partisans may be a shrinking group, they still exist. Large swaths of this voter group simply haven’t made up their minds yet. And they do respond to major news and campaign events, especially as the election nears and its salience grows.
This is why events like Donald Trump’s felony conviction can serve as important “inflection points”. For disaffected soft partisans, it reminds them of why they vote for the side they typically back, and for independents, it helps push them in a certain direction based on the tone and the nature of the event.
I’ll admit that I didn’t think Trump’s conviction would change much regarding the national picture. In some sense, I think I bought into the “Nothing Ever Happens” camp’s worldview more than I’d care to admit. But I was wrong: there has been a small, but clear shift towards Biden since the conviction. This has made the polling picture far closer than it was pre-conviction.
FiveThirtyEight’s averages do a good job of illustrating this. They’re quite sticky, meaning that they don’t bounce around nearly as much as other polling averages do. Additionally, they do adjust state averages for national movement, meaning that they can better control for a paucity of polling in states like Nevada.
The data suggests that the picture has shifted by approximately two points towards Biden since Trump’s conviction. That lines up with voter recontact surveys done by Echelon Insights and the New York Times/Siena College, where voters previously surveyed were asked about their current vote choice immediately post-conviction. Both surveys found a two point shift towards Biden — Echelon went from a tie (47-47) to a Biden +2 lead (49-47), while the Times went from a Trump +3 lead (48-45) to a Trump +1 margin (47-46).
This effect is small, but it is non-trivial and has now shown up across a few surveys and aggregates using differing methodologies. Given the tight margins that the election is likely to be decided by, it is worth taking seriously.
Theories for this effect abound. To me, the most convincing one is that Biden’s improvement is fueled by an improvement with independents and “soft partisans”; the Times survey found that roughly a quarter of the Biden 2020/Trump 2024 voters in previous surveys switched back to Biden following the conviction, while a fifth of the “double haters” (voters who disliked both candidates) moved away from Trump as well.
Critically, the more disengaged that voters were, the more pronounced the movement was; in the Times survey, 15% of Trump’s voters who said they pay attention to politics only “some of the time or never” switched to supporting Biden after the conviction. While it might be a bit excessive to expect this to last, it also suggests that these voters are gettable, and it shows the path forward for Biden and his campaign.
The conviction may currently place Trump’s re-election in much greater jeopardy than many casual poll enthusiasts may have guessed; given the margins in the rust belt, Biden is suddenly less than a point away from an electoral college victory, whereas this number was 2.5 points on the day of Trump’s conviction. Biden leads in Michigan and Wisconsin, and current polling suggest he is less than a point away in Pennsylvania as well.
There remains a lot of time for things to change in either direction, and seeing Trump regain the lead rebound wouldn’t be much of a surprise. But for the first time in a while, polling suggests the presidential election would probably be a pure coin-flip if it were held today.
That’s probably not much comfort for Joe Biden, especially given that he still faces critical persuasion and turnout issues with large swaths of his former voters. But it does suggest that they have responded favorably to him, perhaps for the first time in months. Whether that lasts could determine who the next president is.
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