Why Pollsters and Pundits Systematically Underestimate Extremism, and Overestimate Democracy
This is going to be a very simple, very irritating, and very smarmy essay, with a very simple, very stupid, and very annoying point. You, voting (and proving history wrong, which today are the very same thing).
Imagine that, in a troubled land, a house was burning down. A grand old house, which, having stood for a thousand years, weathering storm, flood, and famine, a little weathered, cracked here and there, still welcomed all, a place of shelter and succour, a haven for the weary, a refuge for the infirm. The last place, perhaps, in this troubled land, which was a symbol of better times. Now, not everyone wanted this house to stand tall — after all the land was troubled, and people, being selfish and blind things, wanted more for themselves, and less for anyone else. Hence, there were some who believed that in it’s place, should rise a gallows, or a concentration camp, or at the very least, a temple to ignorance and hate. “Only the strong should survive!”, cried such people.
So that night, in the crackling firelight, amidst the jumping embers, a bright mind had an idea — why, let us take a poll! So around they went. “Excuse me, good sir — do you want the old house to burn down, Y/N?” The results of the poll said, reassuringly, that there was nothing to worry about. Ahhh — what a relief! And so many of those who wanted to save that old burning house, instead of racing there with water and sand, stayed home, comforted, smiling, and watched television, or played with their children. But many of those who wanted the house to burn right down into ashes, raced there, with gasoline and matches, laughing and cheering. And so even though more people wanted the house to stand than to burn — the great old house which held the land together burned down, at last. And because it was the last such symbol of better times, it seemed to take something larger and truer with it.
Are you beginning to see my point? Let me put that another way.
Why is that polls and pundits never predict a fascist-authoritarian meltdown — and instead always tell you there can’t be one? Even when, well, you’re in the middle of one? Isn’t that funny, strange, and weird? That’s what they’ve done for the last several years — so much so that it’s become comical at this point. That’s kind of a fatal flaw with this way, model, set of ideas. What good is a doctor who tells you you’re perfectly healthy — at the exact moment you’re about to buy the farm? If you want to put that in more sterile terms, it goes like this: why do polls, pundits, and political scientists err so badly on the side of Panglossian “everything will be fine!” that they’ve demonstrably failed systematically when it comes to predicting, explaining, or understanding the death of democracy?
First there was Brexit — a catastrophic mess the British are still trying to clean up, and will be for the rest of their lives: the polls said, “forget it, don’t worry, it’ll never happen!”, and then the very next day, the nation woke up in stunned silence. Then there was Trumpism — a whole election year of laughing, mocking pundits — and pollsters who preened and dismissed the very possibility. Bang! An authoritarian meltdown.
The polls and pundits, my friends, when it comes to extremist movements, are wrong — because they cannot be right. I’ll come to that part, because I want you to really understand it. First, let’s discuss why they have been wrong time and time again — systematically, predictably erring on one side, but never the other, overestimating the health of democracy, underestimating the allure of extremism — with disastrous results.
What happens when you believe the pollster who says — “Relax! There’s nothing to worry about! The bad guys won’t win!!”? Three things, which are like a nuclear bomb going off in the heart of a democracy. First, you, the good and decent person, breathing a sigh of relief, don’t vote. Or at least you’re less likely to — after all, everything’s going to be ok. The statistics say so! Your incentives have been dulled, in other words. But the opposite is true for the extremists. They’re more likely to vote now, aren’t they? After all, they don’t want everything to be ok. So bad polls, based on bad science, which offer spurious conclusions, both overconfident and inaccurate, lead to disastrous results for democracy.
The result is that polls and pundits miss, over and over again, the secret hate vote — while vastly overestimating the amount of decency and goodness and humanity a society has to safeguard itself with. What do I mean by “secret hate vote”? Think about Brexit and Trumpism again — a large percentage of people voted just the opposite of the way they said they would. Hence, such upside down results, which are still “shocking” and surprising” to establishments. (Why would people do such a thing? Well, there are three reasons. One, they’re ashamed to admit their extremist leanings and tendencies to pollsters and focus groups. Two, rage overcomes them at the voting booth. Three, they’re deliberately trying to skew the results — and hack democracy, in just the ways above. Which is it? Probably some combination of all three.) But again, that is what extremism is — people acting in bad faith, to the mechanisms of democracy, whether institutionally, culturally, or, in this case, informationally. Yet pundits and pollsters are gullible enough to suppose that people will always happily tell them the truth they want to hear — and ask us to be that gullible, too. “Sure, I love democracy!”, said the Trumpist in 2016 — and then dropped a ballot like a bomb right into the heart of it.
What are we likely to do when we think we know what everyone else is likely to do? But can we ever really know that, accurately? Do you see the different in these questions, and “how are you likely to vote?”? The problem is that pollsters and pundits are neither intelligent nor sophisticated enough to understand the problem above — what you might call their second order effects. They believe that polls are a mirror which reflects democracy — purely, accurately, reliably — but the fact is that ever-present polls and punditry, at least in this age, are more like hammers which are shattering the window of democracy. What we believe to be “true” goes on to impact, shape, mold how we behave at crucial junctures — how we vote, whether we do, who we align with — whether or not it is “true” at all, just the result of a foolish, erroneous, inaccurate model. But this catastrophic feedback effect is largely, if not totally, ignored, in the way that the people we pay to understand democracy fail to understand democracy. Let me put that another way, so it’s clearer.
Isn’t the feeling of now eerily like…the last time? Don’t worry — there’s no way the bad guys will win! Wait — but what if, just the same way, 5% of people are secret hate voters? And what if another 5% of people stay home, believing there’s no need to worry? Bang! Deja vu — all over again. The bad guys win — precisely because the good guys believed they couldn’t lose. Will such a thing happen all over again? Let’s think about it some more.
Why do pundits err so repeatedly, predictably, so badly, that such a systematic error — only on one side, never the other, underestimating extremism, overestimating democracy — has come to define an age? The last two years of American history go like this, after all. “LOL — he’ll never get elected!!” (he did). “LOL — he’s not going to, you know, put people in camps! Grow up.” (he did.) “LOL — he’s not going to issue white supremacist propaganda from the pulpit! Bro!! Grow up!!” (Reader, he did.) Do you see what I mean by a “systematic error”?
Pundits and pollsters don’t predict meltdowns and collapses because they can’t. (And because they can’t predict them, they can’t explain or understand them, hence they mostly pretend such things are not real at all, but I digress.) The reason for the systematic error in polling and punditry — underestimating the scale and intensity of fascism and authoritarianism in society — is very simple. Polls are based on extrapolating the statistics of the normal — inferring, essentially, that tomorrow will be much like yesterday, from a kind of smooth, predictable bell curve of social attitudes. But the arrival of extremism is like my metaphorical house burning down — not a normal event at all, but an abnormal one, like a plague, a fire, a flood, a meteor.
People’s attitudes yesterday, to, say, houses in general, or fires in houses, tell us very little about the house of democracy burning down. So when extremism arrives, what good is assuming statistical normality? Precisely the opposite is true: it is at just this juncture we must use yesterday’s old distributions and models with great caution and care — not assuming that the normal will prevail, but that it just might shatter. In other words, the tails of the distribution might be much higher, larger, fatter, than we believe them to be — and the middle, much narrow, lower, and slimmer. That is what extremism means, after all — that some social property has suddenly gone more extreme. Hence, pollsters are wrong time and time again when it comes to extremism — and pundits are really just salesmen whose job it is to sell you all this bad science masquerading as insight. But what happens if you believe it? Bang! Democracy melts down.
That is why, to answer the question, when the pollsters and pundits have been wrong, their error has been systematic — not random. In other words, it’s not that sometimes they’re wrong about this side, and sometimes, that side — but that they’re wrong in one direction: they haven’t underestimated the degree of liberalism, socialism, decency, humanity, and so on — they’ve underestimated, and badly, the amount of fascist-authoritarian extremism, viciousness, spite, and rage. In academic terms, we’d call that clear evidence of a failed model or theory, and in pragmatic terms, we’d call that a doctor who’s cure is worse than the disease.
That bias for the normal, in turn, is why polls and pundits never predict a fascist-authoritarian meltdown — they can’t. Instead, they err systematically, badly on the side of Pollyannaism, so much so that the result is systematic failures when it comes to predicting, explaining, or understanding the slow death of democracy, because they assume normality will prevail. It’s hardwired into their assumptions. What else can they do? Hence, tomorrow will be like yesterday — maybe just a little bit different. But not too much. After all, if your worldview is based on bell curves which never change shape — only people moving from one spot to another inside them — what really changes? It’s the pseudoscience of normality, in statistical terms. They simply cannot see the world any other way than as a largely unchangeable thing with the same shape — that is the self-imposed limit of a mindset which can only extrapolate quantitatively, and one-dimensionally, from recent history. But that is a lot, and dangerously, like, believing what you want to believe.
That is why polls and pundits are dangerously unreliable predictors of fascism and authoritarianism — so much so they’ve helped the bad guys take over much of the world, by now. They’ve underestimated extremism, and overestimated democracy — but never the other way around — over and over again — because they do not understand the world they are helping to break apart with that very lack of understanding is not a normal place, governed by smooth bell curves, which never really change. They think this layer of things — instant, endless “numbers” about “votes”, punditry about those numbers — are gleaming mirrors of democracy, which reflect it back to us. But in fact they are more like hammers, smashing through the window of democracy — because democracy is really a place from which we crowd and view the same green fields we hope to sow and harvest.
So. Will it be deja vu all over again? “Everything will be OK! Grow up, dummy!” Reader — it hasn’t been so far, has it? So the everything being fine part, my friend, is up to you. Don’t believe the polls. Ignore the pundits. They don’t account for exactly that. You, believing them. What happens next — how you behave, think, act, after you do. The best thing you can do, in my estimation, is pretend that they simply don’t exist. Just take a long, hard look at reality. Does the house of democracy appear to be burning down? Go forth and extinguish the flames. That orange glow, let me assure you, isn’t the dawn — it’s the embers of decency, goodness, truth, and wisdom, fading to black. And if you get this wrong, that’s all you’ll have left for warmth, light, and heat in the dark age that follows.