Okay, folks. Donald Trump is not Hitler. Donald Trump is not Stalin. Donald Trump is not Mussolini. Donald Trump is not Berlusconi. Donald Trump is not Putin (though he’s weirdly similar to Zhirinovsky) and its way past time to dump all those stupid analogies.
The other day I was assured that Trump can’t win the presidency because women and people of color will rise up against him if he wins the GOP nomination. I don’t know if Trump can win a general election, but I’m pretty sure that prediction adds nothing to the debate either way. You would think that after a year in which everything people said about how social systems work has proven inutile that we’d see a bit more humility but we don’t. So I’m going to say something and put it on record: to date, Donald Trump is sui generis and we’re about to learn a shit ton about ourselves.
My Love Affair and Partial Breakup with FiveThirtyEight
Four years ago I was infatuated with Nate Silver. No, not in person, but as a writer. For those who missed the story then, or who absorbed one of the weird retellings circulating on reddit (can’t find the link) Nate Silver took a bank of computers to the polls and came up with very good predictions in a series of elections leading up to the defeat of Mitt Romney in 2012. More importantly, he published (most of) his methods.
Silver’s approach was simple: he looked at the polls that had come out in the past, labeled components of pre-election polls (how they were conducted, who they covered, who developed the questions etc) and then conducted an extensive analysis using their accuracy at predicting an actual vote as the criterion. From this not only was he able to assign “accuracy scores” to the polls themselves, he was able to disassemble and reassemble parts of the polling to interpolate results where no poll had been conducted. This required a lot of mathematical heavy lifting and Silver has always been a bit coy with his models, but the results, especially in 2012, speak for themselves.
To some extent Silver had it easy. Polls are considered self-report predictions, which are the fatal flaw in every goddamn “stages of change” study I read (you know the ones: on a scale of one to five, how likely are you to continue eating three cups of oat bran per day?) but the act of voting is much like the act of responding to a poll, plus it requires no more effort or social capital to vote X vs Y when the voting is secret.
Much has been made of how Silver “beat the pundits” but this is a bit of a misunderstanding. He wasn’t studying politics at all- he was studying the relationship between how people respond to polls, and how they vote. That was it. The “pundits” he was outguessing were trying to predict voting behavior from news stories, demographic trends, old wives’ tales, and other random bits of wisdom and weren’t playing the same game at all.
Then, sometime between 2012 and 2015, Silver lost the plot and decided to be a pundit himself. I won’t join the dogpile on his (persistent) Trump skepticism except to say that, had he looked at the polls blinded, the Trump phenomenon was there all along. What stood between the numbers and the prediction was the conviction, broadly held by Silver and a whole lot of other people who should know better, that a Trump victory in a party that didn’t want him was impossible.
So Who Was Hitler Anyway?
I actually think this is a history problem. I, and probably both of my readers, have grown up in an era in which the behavior of large groups of people is a scientific field of study, and we rarely reflect on how anomalous that is. Charles MacKay wrote The Madness of Crowds in the 1840s, and both Durkheim and Weber preceded the second world war, but for the most part the study of sociology, especially systematic sociology, has been a post-war phenomenon. That means it has also been a post-nazi phenomenon, and this has had an effect.
What do I mean? In 1945, Germany was in ruins, the world had entered the atomic age and the cold war, Americans were starting to realize exactly how many civilians had been exterminated in “labor” camps, and yet no consensus narrative had emerged how such an unthinkable sequence of events could have happened. (For a modern analogy, ask yourself how you would explain to an eight year old the causes of WWI- if you have a good and honest answer I’m curious to hear it.) Within a few years, though, Adorno had published The Authoritarian Personality and a decade after that we had Eichmann in Jerusalem and the Milgram obedience experiments and the first rumblings of an antipositivist philosophy trying to act on the lessons of fascism in a meaningful way.
Just as wars provoke developments in trauma medicine, cataclysmic social crises provoke developments in sociology. The third reich was a very good reason to go out and learn more about how humans behave in groups.
By the time I was old enough to give a damn, “we,” meaning the western intellectual tradition, thought that we had a thorough understanding of how the nazis came to be, how they had commanded such adoration and power during their brief reign, how they had compelled such horrific acts from the German people, etc. We had a system, custom-built to explain the nazis, that explained the nazis. A side effect is that now, every large-scale bad social movement looks a bit like nazis, but that’s overfitting for you.
We Don’t Know What Will Happen With Trump
So why did WWI? That frustrating sense of not really getting it is what you should be thinking about the future career of Donald Trump. Everything we think we know about how “women and people of color will rise up” is based on the same kind of modeling that assured Nate Silver that the party would decide, or that Bush’s endorsement hat trick would prove critical, or that Rubio’s early lead over Clinton in head-to-head matchups would guarantee his eventual nomination. These are narrative rules based on prior observation, but we don’t actually know the moving parts of how individuals behave, or even how large groups behave.
One weird analogy I’m seeing compares Trump to The Mule, from the Foundation books. If they were written today, Foundation would be the embodiment of Hubris and the Mule would be the unprovable truth that demonstrates the incompletness of the psychohistory model. Asimov wrote a long time ago, though, so the Foundation survives the anomalous Mule. There’s a whole history of positivism and anti-positivism wrapped up in these two interpretations that’s not worth going into here, but what makes Trump The Mule so compelling is how he’s anomalous. He isn’t anomalous in how he behaves, he’s anomalous in that he’s invisible.
Lets get back to Nate Silver. Since I first thought to look at the numbers, Trump has led in the polls in virtually every state, consistently and by large margins. A mechanistic understanding of history would see this as an indication that people were going to vote for Trump and Trump was going to win. However, we’re all too smart for that- we want a narrative understanding in which the forces and powers and trends that we believe in can derail mechanisms, rather than vice versa. We want to believe that parties decide, that there is a political force called “women and people of color,” that there exists a trait called “presidential” that somehow matters here. This is supposition on top of humbug (or, consistent precedent, but that can be the same thing) but it blinds us to the mechanisms of what we’re seeing. Frontrunners tend to win- Trump is behaving like a frontrunner, and his supporters are acting like supporters of the frontrunner, and somehow when people try to write up the election in terms of social systems and organizational theory, all this becomes invisible.
And don’t scoff at the general yet either. If turnout to primaries and caucuses is any guide, the next president will be a republican. If head-to-head matchups in polling are a better guide, Clinton wins against Trump and loses to anyone else (Sanders beats Trump, but its hard to find a poll that shows him winning the nomination). I’m not sure what I think will happen here.
If a Trump presidency is as catastrophic as some pundits are predicting- and remember, they don’t know- then one thing is for certain: we’re about to learn a whole lot more about how humans behave in groups.
By the way…
Apparently my new project is called “clumsy writings about why history doesn’t work the way you think it does.” I tried this earlier, trying to explain how whatever is happening that feels like a “collapse” its outlines, structures, mechanisms and causes will only be apparent in retrospect, and we should stop looking for analogies in the past. Actual historians don’t tend to think history repeats itself, or if they do, they find celebrated yet incomplete examples that don’t assume the world began a century ago and only one bad thing every happened in it. I welcome challenges to this but please don’t limit your scope to Western Europe in the Twentieth Century, okay?
Oh, and whenever I mention the Nazis I have to mention two things: Godwin’s Law and my enduring admiration for the German people’s willingness to accept, over several generations, collective responsibility for the fact that the most intellectually and scientifically advanced country on earth went crazy, declared war on the planet, exterminated twelve to thirteen million people for being “undesirables,” and left their entire continent a smoking ruin. Genocide is not uncommon in human history, a willingness to confront responsibility for genocide is, sadly, quite rare.