Opinion: Trump-Clinton polls tell a story, but not the one that counts

Republican presidential candidate Donald Trump speaks during a campaign rally at Million Air Orlando in Sanford, Fla., on Tuesday.
(Joe Raedle / Getty Images)

To read some coverage of the presidential race, the election is already in the late Chick Hearn’s refrigerator, with the lights out, the butter getting hard and the Jell-O jiggling.

I’m as attuned to the polling trends as the next political junkie. But having watched how the contest between Donald Trump and Hillary Clinton has gyrated since July, and knowing how fragile some of the support for each candidate is, I’m wide open to the possibility of a game-changing late October or even an early November surprise.

Full disclosure — The Times’ editorial board has been unstinting in its criticism of Trump, and I completely agree. He’s not fit to be president.


But whether he wins isn’t my call. It’s yours (assuming you’re a voter). And Trump is right when he argues that media reports about Clinton’s lead in the polls could discourage some of his supporters from casting ballots.

“Horse race” coverage of politics is unavoidable. Poll results are news, even if they should be taken with a large grain of salt.

Here’s an example from Politico, in a piece by Nolan McCaskill about a Trump rally in Florida on Tuesday (with emphasis added by me):

“Trump accused the media of suppressing the vote by reporting that he is losing, although he is currently on track for a landslide loss given his standing in most national and battleground state polls. ...

“Even so, he vowed to win Florida, Iowa, Ohio, Pennsylvania and North Carolina. But while he leads in Iowa and narrowly edges out Clinton in Ohio, he trails by an average of 3.1 percentage points in Florida, 5.7 points in Pennsylvania and 2 points in North Carolina.”

I see what you’re doing there, Nolan! How many of those results were within the polls’ margins of error? By the way, Bloomberg just released a poll in Florida showing Trump with a 2 percentage point lead. Is that a sign of momentum shifting in response to Trump’s heavy campaigning in the state, or just an aberration?

Here’s the conundrum: “Horse race” coverage of politics is unavoidable. Poll results are news, even if they should be taken with a large grain of salt. Besides, Trump repeatedly cited his strength in the polls during the GOP primaries, and again in the general election campaign — until the momentum seemed to shift, at which point, he started disparaging them.

As The Times’ own Daybreak poll with USC Dornsife has shown, a poll’s result depends in part on its designer’s assumptions about who will turn out to vote on Nov. 8. That’s why the most popular prediction sites mash the various polls together into an aggregate picture. And while that aggregate shows Clinton holding a solid lead, RealClearPolitics’ chart of average poll results over time looks more like a sine wave than a straight trend line.


Like many (if not all) of you, I wish this campaign had ended weeks ago. But it hasn’t. To quote another philosopher from the world of sports, it’s never over until it’s over. The Times’ columnist Jonah Goldberg is rooting for a tie in the electoral college that throws the presidential election to the House in December, where conservative independent Evan McMullin conceivably could prevail. That’s about as long a long shot as they come. But at the very least, the winner won’t be decided before Nov. 8, no matter what you might read or hear otherwise.

Twitter: @jcahealey