El: Theory of a Vibes Election (Patreon)
Content
I have, over the past week or so, found myself in the curious position of continuing to be quite sanguine about the Presidential election even as large swaths of news headlines freak the fuck out. And since I can make a horse race politics post on Patreon without unduly mucking up my “above this shit” brand, I figured I’d offer an account of why I, at least, am still highly confident of a Harris victory.
Let’s start with polls. We’re now sixteen years out from the great FiveThirtyEight statistical modeling revolution of 2008, and all well-trained to refresh it/The Silver Bulletin/whatever other forecasting site you prefer daily and sweat the intensely close margin.
Stop that. Not even for your mental health. Just because it’s completely worthless.
I’m not anti-statistical modeling in the least. But even the best designed statistical model—and I’m hard pressed not to think that both FiveThirtyEight and Nate Silver have been prone to confusing making more complicated models with making better ones—is going to have the problem of garbage in/garbage out. And every poll that’s being fed into those statistical models is already a statistical model.
Eight years ago, the New York Times ran a bit where they gave their polling data to four different pollsters and asked them to interpret it. The result was a five point spread, from Trump +1 to Clinton +4. Off, again, the exact same set of 867 poll responses. That’s because a huge amount of modern polling, as response rates plummet, is about taking the responses you have and reweighing them against demographics so as to remove response bias. Basically, you say “OK, I expect the voting pool to be 67% white, so I’m going to take the breakdown of white voters and treat it as though it were 67% of the answers.” The famous big polling error of 2016 amounted to a failure of pollster weighing—the consensus is that they failed to notice they were undersampling non-college educated white voters.
But these days the modeling is increasingly complex and, increasingly, just weighing the population on partisan lean or near enough to it. (A common—and controversial—approach is to weigh based on how people tell you they voted in 2020.) You can see the problem here: at that point you’re all but deciding the poll’s outcome before you get your first response. So what we have with the statistical models is, effectively, a bunch of statistical models being fed into a statistical model. And that’s before you get into the fact that there are multiple pollsters who are flagrantly weighing their polls to get specific results so as to create propaganda pieces or, in some cases, to shift the statistical models.
And in all of these cases, nobody wants to look wrong. There’s a long documented problem of “herding”—a tendency for polls to cluster around the same basic results. I would put it to you that the models are also engaging in this, with a strong tendency towards declaring the election as close to a 50/50 election as possible so that you can declare you were right no matter the outcome.
So fuck it. I’m not saying completely ignore polls. That’d be silly. But the stats geeks don’t know more than anyone else this time; they just have emotional support numbers. Or, this year, emotional antagonism numbers so as to drive engagement.
No, no. The way to understand this election is just to vibe with it. Now, this has some obvious risks, confirmation bias chief among them. Loads of us spent 2016 thinking that the vibes were “ugh, not the fascist” when they were really “ugh, not the coronation of the woman who’s been widely hated for a quarter century now.” It’s easy to do this badly. What’s helpful, I think, is to frame the stakes clearly in your mind. Regardless of how this election turns out the planet is going to cook, the Palestinian genocide is going to continue, and capitalism will rumble along with its usual inevitability and apocalypticness. All that’s really up for a vote is whether I have to flee the country. Now, I’m still quite invested in this—I really don’t want to abandon my cats and most of my personal possessions to become a digital nomad kicking around a series of foreign AirBNBs. But it’s still helpful to be nice and fatalist about it, just so you don’t fall into the “of course Trump’s going to win, everything is awful” trap. It’s ok! Everything is going to be awful either way!
With that in mind, we can consider the situation on the ground without risk of getting overly enthusiastic.
Let’s start with the election that wasn’t, and the increasing sense of doom that was settling around the Biden/Trump election. The key factor in that election was simple: both candidates were absolutely despised. When he dropped out, Biden’s favorability was eighteen points underwater; Trump’s was twelve. Beyond that six point gap—which was significant—much of Trump’s advantage was simply that he was marginally more of the change candidate.
Then, on July 21st, that went up in spectacular smoke. Suddenly we had an election with an actual change candidate. Sure, the sitting Vice President isn’t an outright outsider, but Harris had a relatively low profile over the last four years. And running for President allowed her to reintroduce herself as a substantively new character, as evidenced by the fourteen point improvement in her favorability since Biden dropped out, to where she’s effectively level. Trump’s still at -9.
In addition to the massive favorability gap there’s the fundraising gap. Harris is sitting on a billion dollars; Trump is having to outsource his get out the vote effort to Elon Musk, who’s running it about as well as he does Twitter. Yes, third party PAC money evens out that gap, but there’s structural advantages to the money being held by the actual campaign.
More to the point, though, donations to the actual campaign speak to enthusiasm. Harris’s voters are opening their wallets. That tells you something about how invested they are, just like the failure of Trump’s voters to do the same tells you something about them. Because Trump has a huge structural problem here: he’s running for the third election straight. And the fact of the matter is, enthusiasm wanes in that timeframe. Even his—just look at the ratings history of The Apprentice. He’s offering a stale act for an increasingly small pool of dead-enders, and even they’re leaving his rallies early.
Ah yes. The rallies. That’s the other big flag. Trump’s just… cooked. He’s standing around swaying to music and ranting about Arnold Palmer’s penis. This just plain doesn’t look good. And not in the “grab ‘em by the pussy” way that things didn’t look good for him in 2016. He had scandals then, yes, but they weren’t scandals that mattered to his core voters, and, more to the point, so did Clinton, even if they were imposed on her by Comey. Trump was still out there delivering credible debate performances and enthusiastic rallies. He was projecting strength. And he’s just not this time.
Whereas Harris hasn’t had a single bad story that wasn’t either right wing fever swamp bullshit or Democratic operatives handling their anxiety disorders by confusing journalists with therapists, both of which are things that always happen and so count for rather less when they do. She’s projecting strength, making power moves like giving a Fox News interview or holding an event in Texas. Really, ask yourself, what’s Harris’s biggest problem or weakness here? Conventional wisdom pundits handwringing about lack of policy proposals or some sort of mythic “proper” sit-down interview she should allegedly be conducting? The slight Republican bias of the electoral college that we’ve known about for eight years now? McDonalds trutherism? Whatever it is, it’s not earth-shattering.
So let’s recap. One campaign leads in the polls, leads even more in approval ratings, has more money, has more voter enthusiasm, and is projecting strength. The other lags all quantitative measurements, has no turnout operation to speak for, and faces the significant logistical problem that its candidate keeps sundowning. Who do you think is going to win, given that?
Now, I could be wrong. It happens. I thought The Legend of Ruby Sunday/Empire of Death would be good. I might find myself in two weeks grimly looking up Canadian AirBnbs and figuring out how I’m going to handle my prescriptions. Nothing is guaranteed in the anthropocene.
But this ain’t fuckin 50/50.