24, 24 hours to go....

...I wanna be sedated

A few thoughts as we approach the fateful hour:

1)As MattYglesias wrote this morning, it’s been a rough couple of years for incumbent parties globally, almost all of whom have lost, often badly, in their countries’ most recent elections. This has included center-left parties and right-aligned parties, from Japan to Portugal and many places besides. The pandemic and associated shocks related to inflation and other disruptions have put voters in a bad mood and incumbents are paying the price.

From that perspective, the surprise isn’t that Trump could well win the election. It’s that Harris is as competitive as she is. One reason is that, by almost any macroeconomic measure, the United States has performed better economically these past three years than any of its wealthy peers. But the Democratic Party also forced out its deeply unpopular incumbent. His replacement has been able to run as a hybrid challenger.

It may well be true that another Republican, say Nikki Haley, would be solidly ahead, not in what pollsters tell us is a dead heat. In other words, Trump is paying for his outrageousness, one could argue. Whether he’s paying enough to cost him the election remains to be seen.

2)part of the point of making that observation is to demystify the appeal of Trump. His emergence on the national political stage, his presidency and its aftermath have been jarring and dismaying in so many ways. That he ever ascended the heights he has is, of course, a terrible indictment of so much about our political system and culture. But he’s never earned the support of a majority of Americans. He won 46% of the vote in 2016 against a *very* unpopular opponent. He won just under 47% of the vote in 2020. He’s likely to win about that proportion in 2024, which will include many people who don’t like him but are voting for him because they are Republicans and Trump is the party’s standard-bearer.

Indeed, were it not for the ridiculousness of the electoral college, Trump would never have been president in the first place. It’s useful to remember that not because we can do anything about it, but because the fact of it has fundamentally altered our understanding of Trump’s appeal. Since he won the presidency, many imputed to him a magical hold that he didn’t actually possess, at least not over a majority of voters. He still got to wield power and, of course, would be able to do so again if he wins another minority vote presidency. But it doesn’t follow that “most” Americans are in thrall to him. That’s very much not the case. David Schanzer has a really good post today about Trump’s aura of indomitability, its self-reinforcing nature and the importance of resisting it.

3)As we all know, most pollsters and modelers underestimated support for Trump in 2016 and 2020, leaving a lot of egg on their collective face. Leaving aside some bad faith actors in the field, most pollsters want to get it right and have taken measures to correct for the biases that led them to underestimate support for Trump the first two times he ran. But since this is all far from an exact science, they may well have ended up overshooting the mark in the other direction, as a result.

4)the possibility that pollsters have overcorrected for their previous misses brings us to the poll that the Des Moines Register published Saturday night, showing a three point Harris lead in Iowa. The significance of that result is due to the legendary status of the pollster, Ann Selzer. Selzer has a long, strong track record of accurately polling Iowa, both in its consequential presidential caucuses and in statewide races, including for the presidency.

Beyond the top line number, a couple of (unoriginal) observations. One is that Selzer has chosen not to use the kinds of weights other pollsters have deployed to ensure they’re better capturing Trump voters. Selzer may well feel she doesn’t need to because she has not missed those voters before. When, in 2020, polls were showing Biden heading for a big, perhaps even blowout victory, Selzer’s Iowa poll right before the election, which had Trump up by 7 in Iowa, was a sign (or so it’s now interpreted as such) that the polls were off. That’s because the interest in Selzer’s poll isn’t so much in what it tells us about Iowa, a now solidly red state. It’s in what Iowa tells us about other midwest states, like Wisconsin and Michigan, that share enough demographic attributes with Iowa for an Iowa poll to provide meaningful information about those battlegrounds.

And Selzer nailed Iowa in 2020 (Trump won it by 8). So, if Selzer thinks Iowa is actually in play for Harris, when others don’t, it *could* mean that she’s in far better shape in Michigan, Pennsylvania and Wisconsin, for instance, than the conventional wisdom suggests.

A second interesting point here is that Selzer is willing to publish what appears to be such an outlier result. Analysts have talked more than ever this election cycle about “herding,” the practice of pollsters to massage what look like outlier results for fear of being embarrassingly off the mark. Better to have results that look a lot like everyone else’s (primates are social animals, after all, including those who administer and analyze surveys for a living). The herding discussion derives in part from the extraordinary consistency of the polls this cycle, which have overwhelmingly pointed to nearly deadlocked races across the battleground states and nationally. Of course, we could end up with an excruciatingly close race. And Selzer’s poll is just one poll. But it has an almost unique capacity to cause close observers to question their assumptions.

5)Politics are dynamic. Just when we think we have a handle on which groups are coalescing in which parties around which issues and candidates, the ground shifts. After Obama won reelection in 2012, the Republican National Committee published an “autopsy” that lamented the fact that, as the party of white voters, Republicans were going to have a hard time winning future presidential elections. The report urged the GOP to soften its ethnocentrism to better appeal to non-white, especially Hispanic voters, whose share of the electorate continues to grow. Many liberals, myself included, believed that ‘demography is destiny,’ that the changing face of America spelled future doom for Republicans if they didn’t adapt to an increasingly diverse country. Four years later, Republicans responded by nominating Donald Trump, of all people. And he won the presidency, largely by attracting an enormous share of non college educated voters, including a substantial number who’d voted for Obama (yes, Obama-Trump voters are a real thing).

Trump’s racism and immigrant bashing, political observers assured us, would accelerate the racial polarization of the parties, as Democrats increasingly became the home of nonwhites and Republicans of white voters.

Lo and behold, reality proved a bit more complicated. It may be overstated, but there appears to be some shift of Black voters to the Republican Party (though they are still overwhelmingly Democratic) and a more consequential shift of Hispanic voters to the GOP (though still not a majority in most places). Meanwhile, college educated whites, who constitute a much bigger chunk of the electorate than they did two decades ago, continue to gravitate to the Democratic Party, especially women. So, in the era of Trump, the most overtly racially inflammatory major party presidential nominee we’ve had in modern times, our electorate is racially *depolarizing,* at least to some extent.

6)And what the race will come down to, which no poll or model can perfectly anticipate is who will show up to vote. As Trump has lost ground among better educated, higher propensity voters, he has increasingly sought out lower propensity, especially male voters, convinced that appealing to women and college educated voters would be an uphill battle. Apart from his own native misogyny, this shift may help explain the particularly egregious sexism of the Trump/Vance campaign in 2024. It’s a high risk strategy, in some ways, though one he may have no choice but to follow (at least given the instincts that he cannot shake).

7)I am writing this in part to preserve my sanity and in part, I hope, to provide some broad strokes context for how to think about an electorate that will include perhaps 160 million or more voters, when all the counting is done.

8)For those of you who don’t want to just sit around and wring your hands, there's still time to canvass, phone bank, etc.

9) If Harris does win, we’ll take up separately the attempts by Trump to overturn the election results, as he certainly will try to do (I don’t think he would succeed but, again, we’ll take that up separately).

Deep, cleansing breaths.