Making sense of the 2024 election
When nothing seems to make sense, social science can still help
All signs point to a close election in November – though a lot could change. And it seems wild that a president this unpopular could win reelection – or that a former president, twice impeached and convicted of a round of felonies – with more cases pending – could be a realistic contender.
The surreal nature of the election and polls so far has led some commentators to label it the “vibes” election, with data providing little guide to what to expect going forward. It’s true that there’s not a lot to help us discern who will win in what data tells us will be a close election. But it’s also true that the opposite of vibes is not data. The opposite of vibes is theory. And there are some ideas and frameworks that can help us break down what’s happening as the election unfolds.
For a few months, I’ve been saying that one of the main questions for this election is whether the Trump campaign will be successful at framing it retrospectively, as a referendum on the incumbent. The alternative for the Biden campaign is to try to frame things prospectively, emphasizing what a second Trump administration will do in the future and why this doesn’t reflect mainstream public preferences. Most political science models assume a simple retrospective calculus, in which the electorate rewards incumbents for a strong (usually economic) performance and punishes them for a poor one. But there’s evidence that the political environment – and campaigns themselves – can shape what considerations drive the election.
The other idea I’ve been playing around with, as we’ve contended with questions about how Trump’s felony convictions in New York will affect the polls – or not – is that Trump’s support levels are fairly baked in, while Biden’s are variable. There are a couple of data points to support this. There’s public approval data itself, though both presidents have been pretty consistent (consistently unpopular) and the main difference is that Biden enjoyed much more of a public opinion “honeymoon” than Trump did. On a panel a few days ago, Michael Tesler pointed out that Trump’s support among Republicans has remained strong while Biden’s has fallen. Trump’s character issues – his problems with women, accusations of dishonesty, statements rejecting democratic norms – these are all well-known. Attitudes about Biden appear to be much more closely linked to events and perceptions of his performance in office.
So the next question to ask is, how do these two questions intersect? If perceptions of Trump are largely baked in, why would it matter whether the Biden campaign can frame things prospectively? What if this second hypothesis is wrong, and attitudes about Trump are more variable? I consider a couple of possibilities below.
Prospective voting when attitudes about Trump are set
This is really a scenario in which a surgical approach to issues is required. Surveys show that the electorate trusts Trump/Republicans more on many issues. But that’s not the same as agreeing with issue positions. The challenge for the Biden campaign is to remind people – including wavering members of their own coalition – that, once in office, Trump will do things that they don’t agree with. This includes the abortion issue, where Democrats have experienced an unexpected political windfall. But it also includes tax cuts for the wealthy, and his own set of unpopular border policies. If it’s true that attitudes about Trump himself are fairly set, then the best approach is to get specific on issues.
Maybe I’m wrong and attitudes about Trump can change
One piece of data in favor of this view is that Trump’s support among independents declined after the May 25 guilty verdict in the New York cases. Maybe not everyone already has a firm opinion about the forty-fifth president, as wild as that may seem.
In 2016, Hillary Clinton made a lot of arguments about Trump’s character and fitness for office. And while this was enough to win the popular vote, it wasn’t enough to persuade crucial voters in the states that determined the Electoral College winner, as Vox’s Eric Levitz recently pointed out. But there’s another way to think about Trump’s personal characteristics as a candidate. There’s social science evidence that voters do care about presidential candidates as “strong leaders,” and Trump has often done well on this measure. But there may be room to frame the various indictments as evidence not of poor character (likely baked into existing evaluations) but instead evidence of a weak leader, unable to persuade others and thus forced to resort to illegal means to achieve his ends. Or a leader easily influenced by bad actors, or simply weak in the face of a mighty legal system.
There are undoubtedly lots of other ways to think about how these two ideas overlap, including how Biden might come out looking better or worse depending on different campaign approaches. But these are just a few that highlight a couple of key lessons. First, there are systematic ways of looking at the 2024 election, even if there’s no magic method to reveal who will win in November. There are ways to think in a structured way about how the public perceives candidates, issues, and priorities. More importantly, thinking about the possibilities for different campaign tactics reminds us that the election result isn’t foreordained, just waiting for us to find out what it is. The decisions made between now and November – and the belief that these decisions can make a difference – are what will determine the outcome.