A while back, I posted about Michel Podhorzer’s election analysis showing that Harris’ loss was driven by stay-at-home anti-MAGA voters who sat on their couches rather than voting. Now another analyst, Eric Levitz at Vox, disputes three of Podhorzer’s conclusions:
1) Voters who backed Biden in 2020 — and then stayed home in 2024 — are not necessarily resolutely anti-Trump.
Podhorzer’s argument assumes that Biden voters who stayed home in 2024 could not have done so out of sympathy for any of Trump’s messages. But there’s little basis for that assumption. Low-propensity voters are less ideological than reliable ones, and voters often choose to sit out elections because they are conflicted, agreeing with some of what each party has to say. There’s reason to think that this dynamic drove part of the Democrats’ turnout problem in 2024: Both polling and geographical voting patterns indicate that low-propensity voters became more Republican-leaning during the Biden era.
2) Young, first-time voters turned against the Democratic Party.
The electorate’s youngest voters appear to have been far more right-wing in 2024 than in 2020. This is not a problem that can be attributed to mobilization. Republicans seem to have simply had greater success in appealing to first-time voters last year than they have for a long time.
3) In the Biden era, American voters did become more conservative in some of their values and priorities.
Contrary to Podhorzer’s suggestion, there is considerable evidence that voters grew more right-wing in their attitudes toward immigration and criminal justice and more likely to prioritize those issues. Meanwhile, the electorate also grew more confident in the GOP’s economic judgement.
Since I posted about the Podhorzer piece, I thought I’d also post a piece that pushes back a bit on his conclusions. That said, Podhorzer’s piece is more numbers-driven (relying more on real turnout data), and Levitz’ piece is more polls driven (relying on both pre-election surveys and exit polls). The best case he makes is for (2) since it also relies on registration data (specifically in North Carolina). He also thinks that Democrats don’t necessarily need to moderate their positions on crime and immigration (since they already have) and that they should avoid over-reaction (such as what’s happening with the Laken Riley act.)
Finally, I think the main conclusion that I’ve come to in reading the serious post-election pieces (by which I mean those that use real numbers to back their arguments) is that a tight election like this one is open to a lot of different interpretations that seem plausible depending on which numbers you pick.