Presidential Polls Are Useless. Here’s Why

In the months leading up to the most consequential presidential election in history, it’s easy to see why everyone’s jumpy. 

Every passing hour seems to bring a new poll showing former President Donald Trump ahead of incumbent Joe Biden, or Biden ahead of Trump. And with it, another round of BREAKING NEWS SCOOPING SCOOP alerts about the polls, plus a round of discussion on the news about whether those polls matter. 

For the audience, this constant barrage of non-news not only foments anxiety but disseminates misinformation. At a time when journalism claims to be locked in a vital fight against bad info from bad actors, the industry is reserving its greatest resources for a kind of story designed to  confuse us and tell us less, not more. 

Let’s take a recent CNN poll story, touting its own numbers that pit Trump and Biden against each other: 

Trump’s support in the poll among registered voters holds steady at 49% in a head-to-head matchup against Biden, the same as in CNN’s last national poll on the race in January, while Biden’s stands at 43%, not significantly different from January’s 45%.

Looking back, 55% of all Americans now say they see Trump’s presidency as a success, while 44% see it as a failure. In a January 2021 poll taken just before Trump left office and days after the January 6 attack on the US Capitol, 55% considered his time as president a failure.

Panic-inducing! Maddening! Terrifying! More than half of Americans want to return to office a belligerent bigot who encouraged an armed uprising of fascists attacking the U.S. Capitol to overturn a legal election. More than half of Americans think banning Muslims from immigrating and militarizing the southern border and letting another pandemic rage is a-OK. Half the country is so thoroughly marinated in right-wing propaganda that they’d let an incompetent criminal sex predator run the country and control their children’s futures. 

But it’s irrelevant. Despite its angina-inducing headlines and the reams of reactions and reflections it induced when it was first published, the poll doesn’t matter, in part due to the fact that it was a national head-to-head Biden vs. Trump poll that drew from “a random national sample of 1,212 adults drawn from a probability-based panel, including 967 registered voters.” 

Our presidential elections are determined by the electoral college, which means it doesn’t matter what people in general think. If it were, I’d be in line to visit the Al Gore Presidential Library right now. 

Recall that in 2000, then–Vice-President Gore won the popular contest by more than 500,000 votes across the nation. But the highly contested state of Florida—with its 25 electoral votes— was awarded to George W. Bush, giving him the victory. Hillary Clinton’s popular vote margin was even greater, more than 3 million votes, but Donald Trump’s securing of several Midwest states tipped the electoral college in his favor. 

So why bother polling, when a significant portion of those polled aren’t even registered to vote, much less likely to do so? National head-to-head polls are useless in this country for measuring anything except the general vibe. 

And the general vibe, despite what pundits may tell us, is not predictive of much if anything. In 2016, the media and the conventional wisdom told us that Trump was a gigantic fool who was going to lose horrifically enough to take the entire GOP down in flames with him. The general vibe during the 2022 midterms was that Democrats in races from Georgia to Oregon had no shot in hell. The general vibe, as understood by the politics sages at the corporate press, stands no chance against actual voters doing their actual voting. 

Which is the second reason so many of these polls are pointless. We’re six months away from the general election. Repeatedly, we’ve seen huge swings in presidential elections in the final months, from the movement away from Hillary Clinton to Trump in swing states like Wisconsin and Ohio, to the cratering of former president Jimmy Carter’s support in his contest against Ronald Reagan in 1980. 

Everyone reporting on politics in America knows these two things. So why do we see so many polls and poll stories? The answer is a combination of laziness and greed. 

They want something to post to feed the ever-hungry internet beast that is the 24-7 news cycle. They want something to drive clicks, drive Discourse, and drive the kinds of acrimonious online arguments that lead to trending topics that lead to panel discussions on TV that lead to promotions that lead to covering bigger and bigger stories. 

They want to hear “CNN poll” over and over in other news outlets, and they don’t really fear being held to account for misleading the American people because, as we’ve pointed out earlier in this space, there is no mechanism to hold them to account. A poll says one thing one day, and another thing another, and both stories are treated the same way. It’s no wonder people throw up their hands and decide to stop paying attention. 

Poll stories are easy to do. Lay out the numbers, find a poli-sci professor to comment on them, and you’re done. But when that’s the substance of a news outlet’s political coverage, who’s up and who’s down like it’s all a big wacky game, the impact of politicians’ decisions get reduced to their effect on the numbers. Candidates themselves become flattened in this kind of storytelling, reduced to two equal opposites, falsely equivalent.

Poll utility in general is dubious for the general public. Reporting that centers state-by-state polls, that prioritizes explanations of how polls were conducted, is mostly useful for conveying to campaigns where they should allocate resources: Oh no, Wisconsin’s looking shaky, better open up a few new offices and plan another event there! Looks like Iowa’s solid, no need to attend the umpteenth county fair and be photographed awkwardly eating a corn dog. 

Is that useful, though, for anyone who’s working a job and trying to get home on time to pick the kids up from daycare and find out even a little about what’s going on? Should as much airtime and resources be devoted to measuring the general, like, feeling in the room as the candidates’ actual promises and track records? While talking heads are freaking out on TV hourly about slight variations in numbers, are they actually teaching anyone anything? 

At best, polls would be showing who cared the most about what issues, and how political candidates at many levels were attempting to meet those needs. But they’re just making us all stressed out and miserable and feeding the panic that, frankly, doesn’t need another meal.