The Weekly Sift

The kinds of lessons Democrats are learning from the 2024 election may not matter any more.

---

A post-election article I found very challenging and important is “Republican Victory and the Ambience of Information” by Nathan Heller in The New Yorker.

Heller begins with an observation I keep banging my head against: All those accounts where Harris lost because she didn’t do something — almost invariably she did do the thing they’re claiming she didn’t do. She talked about kitchen-table issues, she had detailed policy proposals, she gave interviews, she reached out to every kind of voter, and so on. She raised money, she advertised, she had a great ground game. But for some reason the things she said and did didn’t register with some large chunk of the electorate.

This seems to me like the central problem for Democrats to wrestle with. Sure, work on the Party’s message, work on the outreach to Latino men, come up with more popular policies. But none of that is going to matter if your great message describing your great policies goes in one ear and out the other.

But why would it do that?

On the other hand, Trump seemed to do everything wrong. His campaign speeches were boring and largely unwatchable. He didn’t have a ground game to speak of. Any policy ideas — there weren’t many of them — were vague. (Does he want a 10% tariff or 20%? It seems like that should matter.)

Trump’s speeches at rallies, many people noticed, had a curious background-music quality: they went on forever, aimlessly, and people would come and go at will. The actual speeches didn’t seem to matter; they existed simply to set a vibe and keep certain broad suggestions (immigration big problem! Biden Administration so corrupt!) drifting into the ether. … Planting ideas this way isn’t argument, and it’s not emotional persuasion. It’s about seeding the ambience of information, throwing facts and fake facts alike into an environment of low attention, with the confidence that, like minnows released individually into a pond, they will eventually school and spawn. Notions must add up to a unified vision but also be able to travel on their own, because that’s how information moves in a viral age.

… The communications researcher Pablo Boczkowski has noted that people increasingly take in news by incidental encounter—they are “rubbed by the news”—rather than by seeking it out. Trump has maximized his influence over networks that people rub against, and has filled them with information that, true or not, seems all of a coherent piece. This is the opposite of micro-targeting. The goal is for voters to meet ideas coming and going so often that those notions seem like common sense.

So when Harris described policies (or even Biden administration accomplishments) that benefit the working class, it didn’t register, because people “know” (from having run into the notion over and over again) that Democrats are elitists who look down on the working class. They didn’t listen, because they “knew” that Harris wasn’t talking to people like them.

Conversely, when Trump said immigrants were “eating the dogs … eating the cats”, maybe people eventually heard that this story was false. (Or maybe they didn’t.) But the idea that immigrants are causing problems all over the country was seeded. When you heard it again, you’d heard it before.

That’s how you wind up with a result like this: Harris won handily among people who were paying attention, but got clobbered among voters who just “knew things” without checking them out.

Heller points out that if you’re trying to seed the world with ambient information, it helps to have your own dedicated media organizations like Fox News, Truth Social, and ultimately X/Twitter, where your factoids can be repeated endlessly without contradiction. Democrats have the so-called “liberal” media, but the message discipline just isn’t there. As often as not, “liberal” outlets like the New York Times and Washington Post would pass along Trump-oriented ambience: Biden’s too old, the economy feels bad, Harris’ campaign doesn’t have enough substance, and so on.

In the old days, campaigns imagined that even fairly uninformed voters had an issue checklist: abortion, inflation, immigration, climate change, education, and so on. Just before the election, they’d find out which candidate agreed with them on those issues, and then vote for that candidate.

For a large (and probably growing) chunk of the electorate, that’s not what happens any more. This is how you wind up with results like we saw in Missouri: The same electorate that voted for Trump 59%-40% also passed ballot measures protecting abortion rights and raising the minimum wage.

What does happen is much harder to get your hands around. But we need to figure it out.