The Last Postmortem

Swear I'll stop after this

“Trump’s Rampage”, Video game from GOP Arcade at the 2016 Republican National Convention, via The Guardian.

A popular postmortem sentiment is the idea that what happened in the US in the November election is what has happened all over the place in developed democracies, in the wake of the inflation that accompanied the recovery from the Covid pandemic all over the world, when angered voters punished the ruling parties in a frustration with the way democracy is seen to be failing:

“There’s an overall sense of frustration with political elites, viewing them as out of touch, that cuts across ideological lines,” said Richard Wike, director of global attitudes research at the Pew Research Center.

He noted that a Pew poll of 24 countries found that the appeal of democracy itself was slipping as voters reported increasing economic distress and a sense that no political faction truly represents them.

and even though inflation in the US wasn't nearly as bad as in some of those other countries, our voters just did the same thing.

I wanted to check that hypothesis out for the flood of big elections between 2021 and 2024, and found evidence for a much more nuanced picture: a bad time for a few ruling parties, but not so great either for Trumpies and like-minded individuals on the nationalist, anti-immigrant, authoritarian side of the right (loosely characterized as "fascist" below, sue me if you don't like it):

• fascists threatened but incumbents left or right retained control in Austria, France (in a settlement that has just fallen apart), Netherlands, Spain

• incumbents left or right kept control where there were no conspicuous fascists in Mexico, Belgium, Finland, Japan (though the LDP lost its majority, following inflation that peaked in 2023 at 3.27%), South Korea (though they elected a rightwing president to a five-year term in 2022, the leftish Democratic Party increased its parliamentary majority in 2024, so the new president was never able to do anything, and in the end he was frustrated enough to attempt a hopeless coup), Taiwan

• incumbents right and fascists both lost in Germany (2021, to the awkward leftish coalition of Olof Scholz that has now, sadly, collapsed, and called a new general election for February), UK (to Sir Keir Starmer's barely left Labour), and Colombia

• fascists beat incumbents in Argentina (personality-cult austerian president), Italy (a weird coalition of fascists sometimes eager to be seen as "moderate"), and US

• fascists were incumbents, and thrown out, in Brazil and Poland

• fascists were incumbents, and won, in India and Israel

What happened to us last month is not something typical of what happened to the industrialized countries in the wake of Covid and inflation. It's something we share uniquely with the land of perpetual hyperinflation ("Don't cry for me") in the Southern Hemisphere. Everybody else seems to have coped with Covid and inflation (often higher than in US, as noted) and some of them with unpopular immigration (especially Italy, France, and Britain in this list), more maturely than we did (not too sure of this but I think Spain and Germany deserve actual praise on immigration, except for Germany's former communist East, which ironically remains the only part of the former Third Reich where fascism still has a serious pull).

Only in America did people end up voting for the triple threat of fascism, the lowlife gangster criminality of the New York real estate industry, and invincible stupidity. How did that happen?

Kamala Harris got close to seven million fewer votes than Biden did in 2020, and Trump got almost three million more than he did four years ago, but that's not exactly it, because in the battleground states, the only ones that count in our weird system, she actually did better than Biden in Wisconsin, Georgia, North Carolina, and Nevada. Trump's victories in the three states that mattered most, Michigan, Pennsylvania, and Wisconsin, were almost as narrow as in 2016; a total of around 230,000 votes. If Harris had won those three, and thus the presidency, she'd still most likely have lost the popular vote, and it's fun to imagine how the Republicans would be screaming to get rid of the Electoral College, and The New York Times darkly concerned about the new president's legitimacy, in exactly the way they weren't at all concerned when Republicans won that way in 2000 and 2016.

---

It was another turnout election, like all American elections, in which the big winner was the nonvoters, 93 million out of a total voter-eligible population of 245 million, compared to almost 75 million for Harris and just over 77 million for Trump; and the biggest story nationwide was that of those seven million voters the Democrats lost.

On the other hand, from the consultant-brain standpoint of the campaign strategists, focused on the seven states that mattered to the Electoral College outcome, they really didn't do so badly, just not quite well enough. The turnout was quite high in those states (it's possible that the low turnout for Harris in the other states was mostly because voters realized their votes really mattered even less than usual), and Trump just captured more of it than Harris did.

Newly emerging stories in The New York Times suggest that the Trump strategy was a lot more effective than I imagined it would be: a lot more like the 2016 campaign, in fact, if without whatever assistance they'd had then from Cambridge Analytica and/or the late Yevgeny Prigozhin's Internet Research Agency, but everything Chris LaCivita and Tony Fabrizio had learned from that experience: making their money go as far as possible by targeting their advertising at the narrowest possible set of voters, as identified by Fabrizio's polling.

These were (according to The Times's Shane Goldmacher) the lowest-propensity voters, the most genuinely undecided because the least informed about the campaign, mostly young and I think largely male and somewhat though not extremely low-income, reachable through ads on streaming services like Max, Tubi, and Roku, which provide advertisers with a lot more information than broadcast or cable TV do, right down to the specific individual, where the Harris campaign was stuck with the usual broad geographical targeting:

“In the seven states, we were talking to 6.3 million people — they were talking to 44.7 million,” explained David Lee, a top pollster for the super PAC, Make America Great Again Inc. “There’s roughly 38 million people that they’re hitting who’ve already made up their mind. So I don’t care how much more money you have than us to spend, you’re wasting 85 percent of your money.”

I thought this was a bad population to work on when it was originally reported, because they were people obviously not interested in voting, but the math of the actual outcome is inescapable: to get the 230,000 or so votes they needed on an investment in 6.3 million people, they had to have a conversion rate of just under 4%, and that's not so unachievable.

And then they didn't have so little money to work with as they expected, thanks largely to Elon Musk's gigantic infusion in the last months of probably upwards of $250 million, reported by Theodore Schleifer and Maggie Haberman. So while I was laughing at Musk and Charlie Kirk for their hilariously inept efforts at traditional get-out-the-vote activities and Trump for spending campaign money on his lawyers, Musk wasn't actually managing much of anything, beyond displaying his Musky sense of humor (for instance, he secretly funded the RBG PAC, named for the late justice Ruth Bader Ginsburg, pushing the message to women that Trump wasn't really as opposed to abortion as his work to overturn Roe v. Wade would suggest, to the anger of Ginsburg's family): rather, he was quietly bankrolling the approach favored by the professionals, at what you might call "levels we've never seen before!"

It's not really right to claim that one factor explains the whole election—there are just too many factors—but there is some evidence that this was an important one. One is the fact from the exit polls, which has been bothering me for a month, that Trump won the relatively low-income voters, with incomes between $30K and $100K, unlike in 2016 or 2020, when he'd been the favored candidate of the wealthy ($100K and up) and just even with the middle income ($50K to $100K), while the under-$50K had been solidly Democratic. Obviously there was a serious amount of falloff in Trump support among the wealthy, maybe mostly the same women frightened and offended by Trump who made the difference in 2022, but it also makes sense to suppose that some important number of the new voters were the ones targeted by the GOP strategy (like Joe Rogan listeners, disproportionately male, under 34 and unmarried, gamers, and middle income).

And many of them were new voters: 8% of the total electorate were voting for the first time, and Trump won them overwhelmingly, 55% to 44%—while I was welcoming the Taylor Swift registrations, nobody was noticing these other people. Then there's the phenomenon of the "bullet ballots"—people who voted only in the presidential race, ignoring all the downballot races from surrogate judges up to senators—which was briefly the subject of a full-blown conspiracy theory of voting machine manipulation:

in Arizona, Trump’s percentage of bullet ballots totaled 7.2%. In Nevada, 5.5%. In comparison, bullet ballots for Trump in Oregon, Utah and Idaho—the three states which border Arizona and Nevada, with equally fervent Trump voters—count for less than 0.05% in each state.

The same pattern continues across the other swing states, with an astonishing 11% of votes for Trump in North Carolina being bullet ballots....Critically, only 400,000 votes would be need to be added in strategic precincts in swing states in order to secure Trump’s victory. In each of these swing states, too, the number of votes for Trump takes the count just over the margin which necessitates a legal hand recount.

The leading proponent of the conspiracy theory, Stephen Spoonamore, has since dropped it, but the fact remains interesting: who are these people?

Exit polls obviously won't pick them up, but it's not difficult to speculate. They're very much nonpartisan, not concerned with partisan issues, extremely unlikely voters, attracted maybe by the Trump personality cult, maybe by Trump as smash-everything disruptor, maybe by Trump as a joke on the annoying world of serious people. Exactly the kind of people the campaign was targeting for the streaming-services ads, in exactly the precincts where they'd make the difference.

A conspiracy, in other words, but probably not terribly illegal (though there were certainly very questionable things like Musk's million-dollar fake lottery, and things that ought to be questioned but rarely are, like the question of collaboration between campaigns and super PACs, which would be really over the line if I've got this right). Kind of disgusting, too, in the contempt it shows for democracy—I'd love to see a politics that engages unlikely voters with a sense that they have a civic role to play, to make their voices heard in the councils of the mighty, but this picking up on a tiny subgroup who really don't care, as cheaply as you can with the minimum it takes to win the undemocratic Electoral College, is the opposite of that.