Trump, Brexit and the State of the Race
Before I left for vacation, I put out a piece on the “scary groupthink” of the punditry, as it relates to Donald Trump. The basic idea was this: Commentary on the 2016 election has broken down somewhat because both the online right and online left opposed the Trump candidacy. Because of this, we analysts find ourselves in something of an echo chamber, which makes us more susceptible to bad arguments, and more likely to overlook good ones that point in an intellectually uncomfortable direction.
Judging by my inbox and Twitter mentions, this was interpreted by many as an apologia for Trump, or a prediction that he would win. It was intended as neither, so it seems useful to offer a follow-up piece, to clarify the argument. Such a piece is especially timely, given that analysts across the pond found themselves in a similar position with respect to the Brexit, and botched the call for many of the same reasons that I worry about here.
To be clear, I definitely did not mean that conservatives were under some sort of obligation to write nice things about Trump. At the very least, Trump represents a deviation from the brand of conservatism that has dominated the GOP intelligentsia for the past couple of decades. One would not expect those who hold to that brand of conservatism to cheer its grip being weakened.
What I mean is simply that this state of affairs has consequences, regardless of whether it makes sense. In the early days of blogging, it was a common refrain on both the left and right that the internet was self-correcting. If someone on the right wrote something dumb, you could count on someone on the left to rip the argument to pieces. Likewise, conservatives were poised to jump on any untrue argument made by liberals. It was the marketplace of ideas at its best; bad ideas withered, while good ideas flourished, and for people who were eager to read both sides, we all moved closer to the “truth.” This is what we’re missing out on this cycle.
We can see how the breakdown of this sort can be consequential with the all-around failure of predictions regarding the Brexit vote on whether or not the United Kingdom should leave the European Union. There’s been a lot of ink spilled on this, so I won’t belabor the point too much. What I will note is this: Most analysts got the basic framework correct. It was commonplace to hear that the Brexit vote was “likely to be close,” was “too close to call,” or “could go either way.” This much was correct, given the very close polling.
Yet almost all analysts ultimately concluded that “Remain” would win. To my knowledge, betting markets never had “Leave” ahead before the vote (if they did, it was for a very short time), even when “Leave” was surging in the polls.
This is the intellectual equivalent of pundits declaring (correctly) that the first possession in the Super Bowl will be decided by coin toss, but then unanimously asserting that the coin will come up heads.
What happened was a massive outbreak of “unthinkability bias.” To the class of people who engage on Twitter, advise banks, or bet on outcomes, Brexit wasn’t just a bad idea. It was catastrophic. It operated as a rejection of an ideal that transnational elites hold dear, regardless of whether they are on the right or the left, one we might just call the idea of “Europe.” A rejection of this idea was not something upper-middle-class analysts could accept, absent absolutely compelling evidence that “Remain” was going to lose.
You can see how this cross-party groupthink played out in some of the analysis. Take this piece by the normally prescient Bagehot at The Economist. That something was off should have been obvious from the conclusion, where he relayed a conversation with a taxi driver; this mode of argument is usually reserved for Tom Friedman of the New York Times.
But consider the other arguments. First, he called a poll showing a significant lead for “Leave” an outlier; fair enough, but there were outliers on both sides. He then made a series of arguments that, in essence, were just a single argument: In referenda, undecided voters tend to break for the status quo.
This argument has the benefit of being true, and it was widely shared across the internet whenever “Remain” fell behind in polls. But because most people wanted it to be true, there wasn’t much critical thinking surrounding it. It was an empirical argument, it gave the desired result, and so that was the end of the story.
What was missing? Well, possibly something I tweeted about in the weeks leading up to the vote: it wasn’t clear what the status quo vote was here. The argument against “Remain” was that the EU was a radical institution that was rapidly transforming the U.K.; from this point of view, a vote to “Leave” was a vote to keep the U.K. as it was, while a vote to “Remain” was a vote for more change.
In a more balanced reporting environment, I think more people would have picked up on this, and perhaps our predictions would have been better. But the sort of folk who felt this way – who viewed the EU primarily as a vehicle for (unwanted) change – are grossly underrepresented among analysts. And so groupthink carried the day.
This is the root of my concern about the commentary about Trump. To be clear, I don’t think there’s a straight-faced argument that Trump is leading Clinton right now, and even when he was up in the national RCP Average, there were good arguments that his lead wouldn’t hold. I thought those were the better arguments, which is why I only gave him a 30 percent chance (or so) of winning in November.
Most people agreed with this ultimate assessment, but then made arguments of the type you would make when, deep down, you think (or hope) that the subject only has a two or three percent chance of winning. This just strikes me as off, and by losing the normal back and forth, we might collectively miss things pointing toward a Trump win.
Consider the polling that came out over the past weekend. My timeline was filled with tweets from both conservative and liberal commentators about the ABC/Washington Post poll showing Clinton up 12 points on Trump. This was obviously an important poll, and it deserved some attention. But there was comparatively little attention paid – at least by the most well-known writers – to the NBC/Wall Street Journal poll showing Clinton up five, and up just a point in a four-way race.
Along those lines, as we are usually reminded, the race is ultimately decided in the states, not by a national vote. To win the Electoral College, Trump simply needs to flip Florida, Ohio and Pennsylvania. The state polling shows Clinton’s leads in those states to be 3.4 points, 2.7 points, and 0.5 points, respectively. This isn’t consistent with a 12-point Clinton lead, or even the six-point lead that the national average shows.
Now there are reasons both of these sets of polls could be the truth: It could be that state polls lag, for example, or that Trump’s weak showing in some very red states (think recent polls from Arkansas, Utah and Kansas) lowers his national vote share while keeping him competitive in the Electoral College. I lean toward the former explanation (I’m sure there are other explanations), but we should also keep in mind that in years where there has been an outright discrepancy between state and national polling (such as 1996, 2000 and 2012), it has been the picture painted by state polling that has proved the more accurate.
Again, nothing here should be taken as a prediction that Trump is favored. But when I say Trump probably has a 20 percent chance of winning (down from the 30 percent chance I saw a month ago), I really mean that if you ran this election 100 times, Trump would win 20 of them.
That isn’t an “outside shot.” As my colleague David Byler noted over the weekend, if you toss a coin twice and get a head, then a tail, an outcome with just a 25 percent chance of happening just occurred. Or as Emory University political scientist Drew Linzer observed, the probability of Brexit occurring, according to betting markets, was as low as 10 percent, while the chances of the Cavs winning the NBA championship fell to five percent at one point.
While polling can give us a good take on where things stand today, I feel somewhat like a pilot flying without gauges when trying to figure out how things are likely to play out. Because just about all of the major analysts aren’t just against Trump – they loathe him – we’re basically situated like the British analysts were when looking at the Brexit election. Good arguments are quickly dismissed, while bad arguments slip through all too easily. This increases the chances that we will miss things we might have otherwise seen, and will latch on to things that we shouldn’t. It just adds an awful lot of uncertainty to this election. That should make us all nervous.