Pollsters fear to start again in 2022

It’s not that pollsters haven’t tried to address the issues that plagued them during the recent election. From state-owned companies conducting investigations for the media and academic briefs to private campaign consultants, they’ve spent the past two years fine-tuning their methods to avoid a repeat in 2020.

But most of the changes they have made are minimal. Some pollsters are hoping that since Trump doesn’t show up midterm, problems with Republicans’ vote share underestimation will go away with him. But others worry that Trump’s continued dominance of the news cycle — from the FBI’s seizure of classified documents at Mar-a-Lago to litigation against his companies in New York — will make him the central political figure. before election day.

“There is no doubt that the polling errors in [20]16 and [20]20 worries the electoral profession, worries me as a pollster,” said Charles Franklin, poll director at Marquette Law School in Milwaukee and a longtime pollster in the battleground state of Wisconsin. « What’s troubling is how unique it is when Donald Trump is on the ballot, compared to mid-terms when he’s not on the ballot. »

After 2016, pollsters said the problem was that their samples included too few voters without a college degree. Polls were better for the 2018 midterm elections, though they were still too Democratic overall.

Then came 2020 – which was worse than 2016, and for which pollsters have yet to find a definitive explanation for exactly what happened. As a result, an easy fix proved elusive. But pollsters mostly agreed that, particularly in 2020, the polls missed a portion of Trump voters who declined to participate in the polls.

The current 2022 ballot is hugely pro-Democratic. FiveThirtyEight’s « lightweight » prediction model, which is based only on the latest polling data, indicates that Democrats have a 79% chance of retaining control of the Senate. This probability clashes with the expectations of both parties and most independent handicappers, who consider the battle for the chamber to be closer to a draw.

And the New York Times noted that some of the strongest numbers for Democrats are coming in the states that had the biggest failures in the polls in the last election.

Celinda Lake, a prominent Democratic pollster, told POLITICO that her firm, Lake Research Partners, is working hard to get the right balance of voters in its samples — but a certain segment of Trump voters are increasingly elusive, especially as the former president’s exploits have been in the headlines lately.

« It was less [of an issue] for a long time,” Lake said. « It seems to us that it’s becoming more and more of an issue recently, with the Mar-a-Lago thing, with its candidates winning a lot of these primaries, with the Jan. 6 committee. »

His firm is not alone in trying to alleviate these problems. Quinnipiac University in Connecticut, a longtime political pollster, released polls last week showing Sen. Raphael Warnock (D-Ga.) Leading GOP nominee Herschel Walker by 6 percentage points, despite other polls showing a tied race or even a narrow Walker advantage, and Democrats with big leads in the two big races to statewide in Connecticut.

Doug Schwartz, director of the Quinnipiac University poll, said his pollsters had changed the way they asked respondents about their voting choices, being careful to separate those who said they were undecided from those who refused to answer outright. the question.

« Ultimately, we hope to reduce that percentage of people who don’t give us an answer on the horse race, » Schwartz said, pointing out that while they accurately reflected Biden’s vote share in their pre- 2020 election polls, the large number of denials led them to underestimate Trump’s. Quinnipiac now had President Joe Biden leading Trump in two states Trump would carry, Florida and Ohio, on the eve of the election.

Another Northeast college pollster, Marist College, released two polls last week: one showing Warnock ahead of Walker by 5 points, and the other showing Ohio GOP Senate candidate JD Vance at the elbow to elbow with the Democratic representative. Tim Ryan.

Marist also had some setbacks in 2020, showing Biden leading by 6 points in North Carolina (which he lost) and 5 points in Pennsylvania (which he won by 1 point) in polls conducted for NBC News. . Since 2020, Marist’s Lee Miringoff said the school has diversified its sampling by reaching out to voters not only by phone, but also by text message and online interviews.

Miringoff told POLITICO he’s not as worried about the same nonresponse bias — the segment of Trump voters who won’t participate in polls, consistently skewing the results toward Democrats — showing up this year. That’s because, he said, Trump himself isn’t on the ballot and Democrats have largely wiped out the advantage of GOP enthusiasm this summer.

« I’m pretty comfortable that what might have been the case in previous elections may not be the case this time in terms of misfires, » Miringoff said.

In the last midterm elections, the most prolific pollster was Siena College in upstate New York, thanks to an ambitious approximately $2 million « live poll » project with the New York Times to survey dozens of congressional districts. In total, the Times and Siena conducted just under 100 polls that accurately depicted the momentum of the Democrats in their successful bid to overthrow the House majority.

Siena is doing polls in swing states this year with and without The Times, including two new polls last week in Wisconsin and Texas conducted for Spectrum News, the local news outlets of cable company Charter Communications.

Don Levy, director of the Siena College Research Institute, said he was « as careful as possible » to increase the share of Trump voters, both in their sampling (who is called upon to participate) and their weighting (making them count for more after the interviews have been carried out in order to correct their under-representation). It’s not enough, Levy said, to just call more Republicans because it’s a specific type of Republican they’re having trouble reaching.

“This is not a partisan non-response. This is a reinforced non-response from Trump supporters,” Levy said. “A small majority of them are self-identified Republicans, but a significant number of them are independents or self-identified Democrats. You can’t fix that by saying, « Let’s raise the Republicans. » It does not work.

Monmouth University in New Jersey is trying a different approach — eschewing horse racing polls for polls that measure each candidate’s level of support without pitting them against each other. Patrick Murray, the school’s pollster director, said his analysis of the 2020 results revealed no « silver bullet » to fix their poll, which also failed to predict the race of the Governor of New Jersey closer than expected last year.

Murray warned pollsters who haven’t given up on the horse race that the dynamics of this year’s midterms are different from the last election – and will likely be different from the next two years from now. « If the 2022 polls are good, » he said, « that doesn’t necessarily mean we fixed what went wrong in 2020. »

Franklin, the Wisconsin pollster, said he made « moderate or marginal adjustments » to the Marquette Law School survey methodology, including increasing the percentage of respondents contacted by cell phone. He’s also paying close attention to the response rate of his polls in Wisconsin counties that have been more favorable to Trump in the last election — but so far voters aren’t turning out in fewer numbers.

Another major failure in the polls « will continue to damage the reputation of the polls, » Franklin said. « I think it’s just obvious and undeniable. »

It can also be unavoidable. Partisan campaign pollsters from both parties have suggested that Trump voters are once again hard to capture as this election approaches.

« There’s a good chance that many public surveys overestimate Democratic strength, » said Glen Bolger, a Republican pollster with the firm Public Opinion Strategies.

Amanda Iovino, a Republican WPA Intelligence pollster who worked on the race for current Virginia Governor Glenn Youngkin last fall, added, « It’s even easier to get college-educated voters on the phone » than voters. who have not graduated from university.

Lake, the Democratic pollster, said she sees the steps her colleagues are implementing to get the right mix of voters. But she’s not sure they’ll be enough to avoid another failed 2020-style ballot.

« I’m convinced these are the right things, » Lake said. « I’m not sure they’re enough. »

But some Democrats not only dare to believe the polls — but hope the party could actually outperform in November, highlighting two special congressional election victories last month in Alaska and New York, where polls showed Republicans ahead of the polling day.

« You just saw the polls underestimating wins in Alaska and upstate New York, » the rep said. Sean Patrick Maloney (DN.Y.), chairman of the Democratic Congressional Campaign Committee, said in an interview during a POLITICO Pro Premium roundtable earlier this month. « So if anything, the polls may show conservative bias right now. »