Now, as Trump widens his lead over his GOP primary rivals, pollsters worry about the crowding of voters, making it nearly impossible to accurately measure his support.
Democratic pollster Andrew Bauman said next November “it looks like Trump will be in the polls.” “So it’s back with a vengeance.”
Not that Trump is some magical force. The problems are practical. In 2020, he singled out significant numbers of people who rarely — if ever — voted and who were either not included in polls or refused to participate in them. Trump trashed the polls that consistently trailed Biden. This created a backlash that made his supporters less likely to respond, and made the polls more inaccurate.
Bauman was among the attendees and presenters at this week’s American Association for Public Opinion Research’s annual conference, an annual gathering of pollsters from the academic, media and campaign worlds.
The organization has been grappling with the future of political voting for decades. A look at the polls from the last two federal election cycles will give you whiplash. By most measures, last year’s midterm elections represented a peak in election turnout. Five thirty eightAccording to post-election estimates, polls were more accurate than in any year since 1998.
But two years after the previous presidential election, national polls are further apart than they have been in 40 years, and state polls are the worst in recorded history.
The threat of another ballot defeat hangs over the industry’s continued efforts to reform its methods.
Among public pollsters, CNN made some dramatic changes to its methodology. Abandoning the long-standing process of random-digit phone sampling, the network and its polling vendor SSRS randomly selected street addresses for its national surveys and sent requests to complete the poll online or by dialing a number.
For most of their state polls — which require a quick turn-around to vote in statewide races like governor or Senate — some voters are surveyed from a file of registered voters and contacted by email or phone, depending on the best contact information. Others were recruited from the current pool of SSRS respondents who said they were registered to vote.
The results are significant. CNN’s polls correctly identified the winner in eight of the nine major statewide races they surveyed — missing only the Nevada Senate race — and half of the candidates’ vote shares were accurate to within one percentage point.
“We were within the margin of error on each one [poll] We did,” said Jennifer Azista, director of polling and election analysis at CNN. “So I feel good about how things have turned out. I would say that gives me some hope between now and 2024.
But Aziza said it’s too early to tell whether the same problems that plagued pollsters will resurface in 2020.
“I don’t think so [Trump’s] The views on the polls and the way he presented his views on the polls to his supporters helped in terms of response rates in 2020,” added Ajeesta, who began a one-year term as head of the organization of pollsters at this week’s conference. “But I don’t know if this will happen in future elections as well.”
Democratic polling firm Global Strategy Group is attempting a significant methodological shift on the back end. According to research presented at the AAPOR conference, their 2022 polls were made more accurate by using voters’ self-reported 2020 general election presidential votes as a variable — a practice many have followed, though still far from universal.
That, in addition to other adjustments that try to include less politically active voters, is an important finding because Global Strategy Group — like pollsters in the public and private campaign worlds — significantly underestimated Republicans in its 2020 poll.
“Between accounting for past voting history in the 2020 election and looking at how important politics is to a person’s identity — the things we’ve done to fix that — we think we can capture that and fix those biases that are going to bite us in 2020,” said Will, a partner at Global Strategy Group. Bauman said.
The phenomenon, which led a segment of Trump voters to boycott the polls, is similar to other trends. New poll this week from YouGovWith the exception of the conservative media, Republicans trust almost all media outlets less than Democrats.
“If anybody wants to be really honest, it’s going to be a huge challenge if it’s really going to be Trump. [against] Biden in ’24,” said Dan Levy, director of the Siena College Research Institute, which conducted the poll for the New York Times and local cable news outlet chain Spectrum News in 2022. “Because we know that voter doesn’t want to talk to us.”
But that’s only part of the problems pollsters have identified since 2020. Not only is it harder to reach voters closely aligned with Trump — all types of less engaged voters are less likely to participate.
And those who are likely to vote in high-turnout presidential elections like 2020, but less likely to participate in midterm elections — the best way to reach those voters remains the traditional and expensive form of telephone surveys. Some surveyors have changed or avoided the method altogether to save money.
Not all pollsters see a sharp dividing line between 2022 and 2024. Democratic pollster Anna Greenberg has 2022 clients including Sen. They include Mark Kelly (D-Ariz.), who, while Trump was not a candidate for office last year, is still a major issue in the midterm campaign.
“Trump was on the ballot,” Greenberg said. “Between the Mar-a-Lago documents and the Jan. 6 commission, and then of course at the state level, the so-called MAGA candidates: the plague masters, [Mehmet] Oz, [Doug] Mastriano … there was a lot of coverage that these were his candidates.
Overall, the mood at this week’s convention was largely positive, up from 2022 — despite the chances of another Trump-sparked miss in next year’s presidential election.
“I’m still worried” about 2024, Bauman said. “I don’t think any pollster should be overly confident that we’ve got everything fixed, because we believed we did after 2016. We haven’t.”