When a presidential campaign is losing, it’s difficult for aides to answers questions about the polls. Usually, they try to either point to the calendar (“There’s plenty of time for the polls to change, and we’re optimistic about breaking through”) or they’ll downplay the importance of polling itself (“The only poll that matters is the one on Election Day”).
But occasionally we’ll see a more problematic posture: a candidate’s supporters will argue that the polls are somehow skewed.
Donald Trump’s campaign manager, Kellyanne Conway, who’s scheduled to be on “The Rachel Maddow Show” this evening, sat down with the U.K.’s Channel 4, which asked about the Republican candidate’s current deficit. Conway pointed to “hidden” Trump backers who’ve been excluded from surveys.
The dozens of recent polls that show Hillary Clinton ahead of Trump both nationally and in battleground states are, according to Conway, “cherry-picked polling numbers that are put out there by media outlets that are also bent on his destruction.”“He performs consistently better in online polling where a human being is not talking to another human being about what he or she may do in the election,” she told Channel 4, in comments first flagged by MSNBC. “It’s because it’s become socially desirable, if you’re a college educated person in the United States of America, to say that you’re against Donald Trump.”“The hidden Trump vote in this country is a very significant proposition,” she added.
Asked if Conway, a longtime GOP pollster, has been able to quantify this, she said she has, but wasn’t prepared to discuss it publicly. “It’s a project we’re doing internally,” Conway said. “I call it the ‘undercover Trump voter,’ but it’s real.”
And while anything’s possible, it’s best to be skeptical about this.
We talked a bit last year about the so-called “Bradley Effect,” in which the public lies to pollsters about their true political preferences because they’re embarrassed to support unpopular candidates (or occasionally oppose popular candidates for embarrassing reasons). Some social scientists believe the phenomenon is real; others don’t. Your mileage may vary.
What matters in 2016, however, is that there’s no statistical evidence to bolster Conway’s assertions: “At the Huffington Post, David Rothschild outlined the data. In the primary, yes, online polls showed Trump with more support than live-caller ones. But that support was actually too high, Rothschild points out. And in the general election, the same split doesn’t exist.”
In fact, it’s worth emphasizing that the polling in the presidential race thus far has generally been quite good. Sure, pollsters famously got Michigan’s Democratic primary wrong, but this failure stands out precisely because it was so unusual: the rest of the results from the other major contests were largely correct.
And for Trump supporters, that spells trouble. To be sure, the Trump campaign isn’t the first to suggest the polls are overlooking its silent army of backers, but no modern presidential campaign has clung to this line and gone on to win.