Fighting the Last War on Trump and Polls

Michael Brendan Dougherty asks, “Is there a reluctance to acknowledge that Donald Trump is a serious underdog for reelection?” It depends whom you ask, of course: Most liberal or progressive Trump critics are pretty stridently certain by now that Joe Biden will win in November. But there is unquestionably, even in some liberal quarters, a lingering kind of post-traumatic reaction to the fact that Trump won in 2016 after most everyone expected him to lose. The good news is that we should all have learned some humility along the way about how precisely polls could predict the state-by-state outcomes. Not so regarding national polls: In the national popular vote, Hillary Clinton’s approximately two-point margin over Trump was in line with the polls. We should also have some humility about how certain we could be in declaring a likely winner when the race was close. In fact, even retrospective analysis of 2016 remains much-contested, given the likelihood that exit polls may still have understated the proportion of non-college-educated voters in the 2016 electorate. Moreover, the wild events of 2020 should caution us that it may yet be early to try to accurately model what the fall electorate will look like, and how it will react to these two candidates.

What concerns me, however, is when humility turns into surrender. It’s one thing to say “the polls got some things wrong in 2016,” and another to treat Trump as possessing some sort of special immunity from polls, such that no evidence could possibly convince people that he is losing. At some point, throwing up your hands and declaring all poll data useless smacks of denial.

Menu