The polls are probably an accurate picture of what’s happening right now—Trump is getting trounced—but the election is more than two months away, and anything can happen.
By Seth Masket
Donald Trump speaks in front of a capacity crowd at a rally for his campaign on April 10, 2016, in Rochester, New York. (Photo: Brett Carlsen/Getty Images)
Republican leaders are in a unpleasant situation right now. The numbers look bad for them for this fall’s elections, but they still want their supporters to feel engaged and encouraged. Otherwise, they’ll stay home in November, and predictions of doom will become a self-fulfilling prophecy. How do you keep supporters motivated when the math looks this bad?
It seems clear that quite a few conservative elites and journalists have approached this in entirely the wrong way. That way is to simply say that polls are wrong and other evidence, such as turnout at rallies, is a better predictor of election results.
Eric Bolling at Fox News’ The Five show is one of the most notable recent perpetrators of this myth, explicitly stating that polls are unreliable and the 10,000 or so people who reliably show up at Donald Trump’s rallies are a strong indicator that he’s going to win. Radio host Bill Mitchell added to this approach by tweeting one of the most profoundly weird critiques of polling I’ve ever seen. Honestly, if there’s a book of campaign poetry published this year, I hope this is featured prominently:
A poll is a hypothesis. Rallies and social media presence are empirical evidence. A hypothesis unsupported by empirical evidence is wrong.
— Bill Mitchell (@mitchellvii) August 16, 2016
This argument is, of course, complete crap, and is a reliable go-to for losing campaigns that can’t admit they’re losing. Peggy Noonan famously suggested that rally size indicated Mitt Romney was about to defeat President Barack Obama in 2012. Despite hundreds of polls showing a modest but persistent advantage for Obama, “Nobody knows anything. Everyone’s guessing,” claimed Noonan. All the “vibrations” suggested a Romney win.
Such willful self-delusion is not solely the province of Republicans, of course. Democrats tried to convince themselves that Walter Mondale’s huge crowds meant he was heading for a victory in 1984. (He got trounced, as the polls predicted.) And let’s not forget the inspirational 80,000-strong crowds gathered to hear John Kerry (OK, and Bruce Springsteen) a few days before Kerry’s 2004 loss to President George W. Bush.
It should be obvious, but crowd size is a very poor predictor of vote share. Hans Noel explained this very nicely in one tweet:
https://twitter.com/ProfHansNoel/status/766270063875850240Now, there’s a good chance many of the people making silly arguments dismissing polls and talking up crowd size actually know what they’re saying is silly. They’re just doing it because it’s boring and depressing (and bad for ratings) to say over and over again that your side is losing, and because they want to give their supporters hope. That’s not nothing. If Republican voters become convinced that their side is doomed this year, many might not even bother to vote, making the situation even worse for them.
But there are different ways to give people hope. Bolling et al.’s approach is to simply lie. This, as his colleague Dana Perino forcefully argued, is self defeating and irresponsible. It fills those voters with false expectations and makes them even more cynical, or even suspicious, once their side loses.
The other way is to note that, yeah, the polls are probably an accurate picture of what’s happening right now, but the election is more than two months away. Quite a few things can still happen. Hillary Clinton still has high unfavorables, third-party candidates are having their best year in decades, Trump was consistently underestimated in the primaries, and the debates are still to come, and no one knows quite what those will look like. Democrats certainly have reason to be confident, but they’re not complacent, or shouldn’t be, because they know November is still far away and it’s been a really, really weird year.
Republicans are facing difficult math. But they don’t have to treat their voters or their viewers like marks.
||
