Why are electoral horse-race polls still a big story? Ted Diadiun

CLEVELAND - If we have learned nothing else about American politics over the last 10 or so years, it is this: The election-polling industry is broken.

At one point in our history, polls told the story of every election – before the election even happened. Pollsters delved into every American instinct; examined every American fear; plumbed every American desire. And they took all the drama out of election night, because we already knew what was going to happen. They were almost always correct.

Not anymore.

In the 2016 presidential election, when pollsters practically coronated Hillary Clinton months before Nov. 8, all came crashing down when the actual votes were counted, and to the bewilderment of pollsters everywhere and stunned broadcasters on election night, Donald Trump became our new president.

The polling companies arguably had an even worse showing in 2020, even though they successfully forecast the winner. Polls showed Joe Biden comfortably ahead, but then came Election Day – and only about 44,000 votes in Georgia, Arizona and Wisconsin out of the 154.6 million cast nationwide separated Biden and Trump from an Electoral College tie.

The pollsters got the popular vote largely correct, as Biden won by a 7 million vote margin – but whiffed badly on the state-by-state margin of his eventual victory.

Polls had shown Biden leading by 10 points in Wisconsin. He won by less than a point. Polls had him winning by eight in Michigan. He won by three. The pollsters had him with a five-point edge in Pennsylvania that shrank to about one on election night.

This year, as I write this, the polls show Kamala Harris with a slight edge in all three of those states. What could go wrong? Lots, as it turns out – which is why it pays to meet these reports with a bit of self-restraint.

Last month in The Atlantic magazine, senior editor Gilad Edelman offered a revealing deep dive into the past missteps noted above, and what polling experts think went wrong.

The short answer? They don't really know.

Oh, they've got plenty of theories. People are using cellphones instead of landline phones and thus are more difficult to find. Polling companies didn't make sure their research included enough white people without college degrees. Republicans – and especially people who intend to vote for Trump – are reluctant to participate in polls or to reveal their choice. (One polling executive quoted by Edelman calls that "anti-establishment response bias – the more someone distrusts mainstream institutions, including the media and pollsters, the more likely they are to vote for Trump.")

Then there was this delicious little theory from the story: "Some experts believe that Democratic voters were more likely to answer surveys in 2020 because they were more likely than Republicans to be at home with little else to do." (Hey, don't throw stones at me, my liberal friends ... I'm just the messenger here!)

Either way, it's well worth reading – it explains the challenges pollsters face without getting too deep in the weeds of poll-speak and takes you through the experts' guesses on what went wrong in the past and how that might be corrected.

"They are cautiously optimistic that they've learned from their mistakes," writes Edelman. "Of course, they thought that last time, too."

Given all of that, you would think that news executives and campaign operatives would approach poll results with caution, and a healthy dose of skepticism. But you would be wrong.

Why would they? It works. It must, because day after day reporters crank out stories about the latest poll, candidates and readers alike gobble them up, comment on them, and look forward eagerly to the results of the next flawed poll.

Poll stories have become clickbait.

It's a rare day that I don't see at least a few e-mails with breathless links to the latest poll report.

Here are a few examples that hit my inbox in the last couple of days:

"Are we at peak Kamala Harris?" asked NBC News in a Wednesday story, citing several positive poll results for the vice president and observing in the text underneath, "Since their only debate, there hasn't been a poll released — even polls with Trump ahead — showing anything other than improvement for Harris and problems for Trump."

"In particular," the story went on to say, "Harris has closed the gap on Trump's key issue advantages on the economy and the border. While Trump is still seen as stronger on those issues, his leads over Harris are lower than they were when voters were comparing him and President Joe Biden."

Maybe she has, and maybe she hasn't, but what Democrat wouldn't want to click on that and read happy news about his or her candidate?

Yet, just the day before, I opened my laptop to this: "Sun Belt Poll: Trump's policies helped, while Harris's would hurt," a Washington Examiner headline proclaimed over a story about last week's New York Times/Siena College poll that showed Trump ahead in the battleground states of Arizona, Georgia and North Carolina.

How can we be sure? But what Republican wouldn't be cheered by that "news," and want to devour every word?

Of course, even in these reports, there is something for everyone: The spin on the supposedly objective stories about the previous poll differed according to the political leaning of the news source, with the liberal New York Times carefully couching the results as "setting the stage for an extraordinarily competitive contest in three key states," concentrating on the narrow two-point Trump advantage in North Carolina – while the conservative Examiner put its interpretation in its headline, and presented it in the story as "Good news for Trump."

Closer to home, Newsweek threw a monkey wrench into the Ohio race for U.S. Senate, which has long been assumed to be well in hand for Democrat Sherrod Brown over Republican Bernie Moreno, with a story showing that two different polls had Moreno enjoying a two-point lead.

That's "news" sure to induce consternation in the Democrats' camp and renewed vigor on the Republican side, but what has changed, really? Only a report on a couple of polling agencies whose data might or might not be accurate – an answer we won't know for sure until Nov. 6.

"Like Olympic athletes, political pollsters spend four years fine-tuning their craft," wrote Edelman in The Atlantic, "but don't find out whether their preparations were adequate until it's too late to do anything differently."

We – both the purveyors and consumers of election news – can do something differently right now, however:

End our obsession with horse-race polls.

If you want to embrace polls that make your candidate look good, fine. Just keep it in perspective, and keep your expectations in check until after Election Day.

To reach Ted Diadiun:

Have something to say about this topic?