Can we really trust the 2024 presidential polls? Experts say, ‘Yes, but …’

DES MOINES, Iowa — After months of candidates cherry-picking them, bragging about them, and whining that they’re biased, the polls of the 2024 race are about to get their first reality check with Monday’s Iowa caucus.

Skepticism about political surveys are as much a part of the modern presidential campaign as visits to the Iowa State Fair and the Red Arrow Diner in Manchester, NH.

But both the pollsters themselves and experts in the field are confident that surveys are generally on point — if one knows how to analyze them correctly.

“Polls get pretty negative attention from most of the candidates. So when you put a poll out and it has a candidate at 4 or 5%, the first thing that they do is dismiss the credibility of the poll,” Spencer Kimball, a professor and executive director of Emerson College Polling, told The Post.

“The general industry takes a lot of hits every time they put out a poll because people who don’t like those results will find inconsistencies within the data to try to dismiss what the data is suggesting.”

Understanding the margin of error

The key to understanding any poll, the experts say, is understanding the margin of error.

“When we get down to that level of specificity, polling just isn’t the kind of tool that has that kind of laser ability,” Monmouth University Polling Institute director Patrick Murray said.

“It is subject to a certain amount of error, because it is a poll that’s trying to predict a population that doesn’t exist yet,” he added. “You have to accept these things with a grain of salt.”

Public perception and media reporting of polls often don’t put enough emphasis on the margin of error, which becomes more important as races across the country tightened, Murray argued.

4

“The 1-point and 3-point polls should be reported exactly the same way: It’s a close race and it could go either way,” he lamented. “But that’s not what happens.”

Samara Klar, a professor at the University of Arizona’s School of Government and Public Policy who studies political behavior, warned against gleaning too much from recent polls of the Nov. 5 general election.

“We’re seeing a lot of polls where really neither candidate is outside the margin of error. And I don’t know if we can really assume anything about who’s going to win a general right now,” she said.

“I think the polling can help us understand who the nominees will be.”

‘Snapshots in time’

While polls can give a window into the mood of everyday Americans, they are not crystal balls, according to Klar.

“What polling does really, really well is tell us what people think now. What polling is not necessarily designed to do is to predict what people are going to do in the future,” she explained.

A significant complication of election polling is the fact that some voters tend to change their minds late in any campaign.

“If we could predict whether polls are going to be reliable — we could predict the actual outcome,” Murray said. “The point of polls right now is to tell you where things stand at the moment.”

4

“Things change. There’s a lot of money being spent in these elections. And generally, people are not paying attention to candidates — a presidential [race] being more of an exception, but especially in down-ballot races,” Kimball added.

“As people start tuning in, and hearing things, we noticed that opinions do change and these snapshots in time that might be accurate at one moment — they don’t sustain themselves.”

A volatile 2024

This election cycle in particular is looking very volatile given the confluence of potential politically combustible factors — such as former President Donald Trump’s pending criminal cases.

“There’s never been as much interest in polling, and it’s never been harder at the same time,” said Don Levy, director of the Siena College Research Institute.

“Trump’s working aggressively to try to not go on trial in March. If the trial does take place and you’re getting a steady dose of transcripts from that trial, commentary on that trial — that could very well influence voters.

“And really, you’re coming down to influencing a small group of voters.”

4

Oustanding questions about Trump’s legal situation, a tumultuous geopolitical environment, and more could entice voters to change their calculus.

Still, if 2024 does wind up becoming a 2020 rematch of Trump vs. President Biden, voters may already be very dug in, according to Kimball.

“Looking at this from a 30,000-foot view, this should be a very volatile election. There’s a lot of things happening. There’s global wars going on, there’s issues here at home,” Kimball said.

“On the flip side, if it does end up being Biden and Trump, you’ve got two known candidates. There’s not much more for the public to know,” he went on. “We normally don’t see that type of rematch.”

But what about 2016?

Perhaps the greatest knock on the polling industry came during the 2016 election when a slew of polls seemingly missed Trump’s upset win over Hillary Clinton.

“We look at what we’ve done in the past. We know we’ve had missteps for a variety of reasons in 2016 and 2020,” said Murray, who rejected the notion there was a “hidden” vote for the 45th president eight years ago.

“It really came down to two issues — that the samples underrepresented people without a college degree, and there were a significant number of likely Hillary Clinton voters who stayed home,” he said.

4

To Levy, there was “likely an insufficient number of polls in some of the Rust Belt states” ahead of the 2016 presidential election and some polls didn’t sufficiently “take educational attainment into consideration.”

He does believe that in 2020, “there was a systemic reluctance on the part of avid Trump supporters to participate in polling” and that pollsters have generally recalibrated to account for that.

“Polling as a field got better,” he said.

“Did people overestimate Republicans in ’22? I don’t think we did,” he added of the midterm election, in which Democrats performed better than expected. “Perhaps some people were so very concerned about missing the Trump voter … that perhaps there was a tendency to weight in that direction.”

Even if pollsters do gauge this election cycle correctly, they’re not holding their breath for acclamation.

“I can give you countless cases of people calling me up to yell and scream at me before an election,” Levy said, “and suddenly when we get it right — as we do more often than not — after the election, somehow the phone calls don’t come saying, ‘Hey, Don, you got it right.'”