Ignore everyone making political predictions

Pollsters and pundits do not deserve your trust or attention.

Ignore everyone making political predictions

Pollsters and pundits do not deserve your trust or attention.

When a story’s audience knows something that the characters do not, this situation is called “dramatic irony.” The classic example is Oedipus Rex — we know some meaningful information about the king’s parents that he does not find out until the end of the play. Sometimes, we are forced to view our own pasts through this lens. There are times we behaved in ways that would have made no sense at all if we knew then what we know now.

This is how I remember the events of November 8th, 2016. There was no doubt in my mind what would take place on that day, the day of the previous presidential election. I was not the only one. Pollster Frank Luntz tweeted, “Hillary Clinton will be the next President of the United States.” The most up-to-date predictions from the most reliable sources were in agreement. Clinton’s chances of winning were up to 99 percent, according to some analysts. That night, I happened to be following the New York Times Live Presidential Forecast, which showed the shifting probability of which candidate would win. I watched in disbelief as the chances of victory dropped for Clinton, from their 85 percent start, and grew for Trump. The final result, which still stands today, shows a 95 percent chance of Trump winning. Even then, approaching midnight, I thought to myself, five percent isn’t zero.

The experience has left me with a lack of faith in the prognostications of pundits and pollsters, an awakening unfortunately not shared by those pundits and pollsters. Indeed, the very people who massively fucked this up the last time around are desperate to convince us today that they know exactly what is going on. One of the most outspoken is New York magazine’s Jonathan Chait, a man who is so high on his own supply that in January 2017, he published a book called Audacity: How Barack Obama Defied His Critics and Created a Legacy That Will Prevail — three days before Trump’s inauguration.

You might say that a liberal who believes that Barack Obama’s legacy has prevailed is delusional. But that is the description Chait applies instead to his worst enemies, namely anyone to his left. “New Poll Shows Democratic Candidates Have Been Living in a Fantasy World,” says a recent piece of his. Indeed, Chait seems to consider leftists a greater threat than conservatives, sometimes devoting many more column inches to the dangers of campus activism and redistributive economic policies than the rising tide of the right.

Let’s look back at Chait’s record, choosing a random year. How about 2016. His greatest hits include, “Why Liberals Should Support a Trump Republican Nomination” (the reason being, “he would almost certainly lose”), “Donald Trump Is Not Going to Win Michigan” (he did), “Hillary Clinton Isn’t Very Popular, But She’s Winning Anyway” (she didn’t), and “Why Hillary Clinton is Probably Going to Win the 2016 Election” (all together now).

Chait’s latest refers to polling data from the New York Times, focusing on swing states. The results “ought to deliver a bracing shock to Democrats,” he says, given that Trump “trails Joe Biden there by the narrowest of margins, and leads Bernie Sanders and Elizabeth Warren.” (It’s worth noting that a recent national poll from the Washington Post shows all three leading Trump.)

It is indeed true that Trump is a more formidable foe than many, for example Jonathan Chait, realized the last time around. But his claim is not that Democrats need to fight him harder — it’s that they need to appear more like him. Characteristically, and in spite of evidence to the contrary, he gestures towards middle ground:

The “center,” of course, is a somewhat hazy concept, subject both to overinterpretation and misinterpretation. Capturing the center isn’t the only reason politicians win elections, and some policies that Washington elites consider “radical” are in fact popular. Nonetheless, it really is true that there are a bunch of persuadable voters who can be pushed away from a party based on their perception that it’s too radical.

According to Chait’s interpretation of the data, “Biden’s paper-thin lead over Trump in the swing states is largely attributable to the perception that he is more moderate than Elizabeth Warren or Bernie Sanders.” He suggests that “Biden’s name recognition and association with the popular Obama administration” have placed him ahead of other centrists, like Amy Klobuchar, but doesn’t factor it into his (declining) lead over Warren and Sanders.

It is ridiculous — hilarious, even — to suggest that voters examined the policies closely and concluded that Trump’s moderate, measured policies made him the best candidate.

The polling data in this case does show that about half of Biden supporters consider Warren too far to the left. The question is whether this would become a decisive factor in the choice of whom to vote for. Biden voters who find Warren too politically extreme, says the Times, “are relatively well educated and disproportionately reside in precincts that flipped from Mitt Romney in 2012 to Mrs. Clinton four years later.” These are, in fact, the very same voters that were prioritized in Hillary Clinton’s 2016 campaign, as Democratic Senator Chuck Schumer summarized in a now-notorious statement: “For every blue-collar Democrat we lose in western Pennsylvania, we will pick up two moderate Republicans in the suburbs in Philadelphia, and you can repeat that in Ohio and Illinois and Wisconsin.” They weren’t enough to win.

Chait and Schumer aren’t the only ones peddling this doctrine. At Vox, Matthew Yglesias has argued that Trump’s victory shouldn’t have surprised anyone. (2016-era Yglesias, however, believed other pundits underestimated Clinton’s odds and wrote an essay commemorating her victory in advance.) According to him, “it’s not true that Trump ran and won as an ideological extremist.” While he may have had “extremely offensive rhetoric,” his platform was “almost paint-by-numbers pragmatic moderation.” This leads him to a tautological conclusion: “candidates who take popular positions on the issues are more likely to win than candidates who take unpopular ones.”

It is ridiculous — hilarious, even — to suggest that voters examined the policies closely and concluded that Trump’s moderate, measured policies made him the best candidate. We don’t know each individual voter’s mind, but Trump’s stance, as a political outsider with an anti-establishment orientation, was a stark contrast to Hillary Clinton’s self-presentation as the most “qualified” possible candidate for the office. Indeed, Yglesias has made the same point himself:

Voters don’t like “the system,” and the baggage that comes with it. The best way, by far, to make Trump own his corruption is to pick someone from outside the swamp to run against him, rather than letting him continue to position himself as the scourge of the establishment.

The prevailing argument of centrist pundits is that regardless of the merits of progressive policies, Americans object to them, and would only support a candidate they saw as falling in the middle. But again, polling data only backs them up if selectively chosen. An Ipsos/Reuters poll that looked at policy preferences found that voters rated the leftmost candidate, Bernie Sanders, as their favored choice on some of the very issues centrists cite as divisive, including health care.

As helpful as it may be to consider the totality of statistical information, it’s also important to remember that probability is not reality. Even now, the New York Times probability meter leaves Trump with a 95 percent chance of being the president. The inherently hypothetical nature of probability is why it leads to counterintuitive conclusions for those of us who are not theoretical mathematicians, and always runs the risk of favoring outcomes that don’t happen.

What makes policies popular? Public opinion is not some kind of geological formation, solid and immobile.

Even simple predictions make countless assumptions about circumstances — any event is contingent upon infinite factors in its environment and process. But even more troublesome is the observer effect — the name given by science to the phenomenon when measuring a result affects the outcome itself. The results you get are affected by what question you ask, like if you have a preceding suspicion that Biden voters think Warren is too far left. And in the most uncomfortable reality for the political press, when those results are publicized, they have a ripple effect on public opinion.

The idea that moderate candidates do well simply because they represent the will of the electorate, rather than because they are endorsed, supported, and popularized by the party establishment and the media is self-evidently absurd. This is plainly observable in a variety of ways. The media adores Pete Buttigieg, who fulfills their wildest fantasies of a dignified, erudite statesman, but who appears to the rest of the world as a barely sentient robot. His showings in any and all polls fall far short of his rapturous reception by pundits, but his lead over competitors may be a result of his attempted coronation. Meanwhile, the political press continues to comically underreport poll data that shows favorable outcomes for Bernie Sanders, outright declining to mention him even when he comes in first or second.

What makes policies popular? Public opinion is not some kind of geological formation, solid and immobile. You might describe politics as the process of popularization, both of policies and principles. Consider the idea of socialism, which, since the last Sanders candidacy, has risen significantly in popularity among the general population — particularly young people. Probability is hard to measure and even harder to predict, but it’s not the best basis for political action. It’s not necessarily even the best way to play poker.

Reducing politics to electability turns it into little more than a card game, and the pundits playing the odds are just as superstitious as any compulsive gambler. But politics is not a game, it’s a fight. It’s only worth participating if you believe something.

At the end of the Oedipus Cycle, it finally dawns on our hero that he killed his dad and has been screwing his mom. In response, he gouges his own eyes out. You wouldn’t take Oedipus’s advice. After 2016, we shouldn’t trust the pundits either.

Shuja Haider is a writer-at-large at The Outline.