Voters fill out their ballot on January 23, 2024 in Loudon, New Hampshire. With Florida Governor Ron DeSantis withdrawing from the race two days early, the Republican presidential candidates, former President Donald Trump and former UN Ambassador Nikki Haley, are competing against each other in what is a first in the country.
Tasos Katopodis | Getty Images
2014 was the first year Lonna Atkeson remembers receiving hate mail.
Atkeson, a political scientist who studies election polls and public opinion, has been conducting voter polls since 2004. She is currently a professor at Florida State University and has written several books.
But ten years into her election work, the angry messages started pouring in, she said.
“I got letters from people saying, ‘You’re part of the problem. You’re not part of the solution. I’m not going to answer your surveys anymore. You’re an evil academic trying to brainwash our children.'” recalled Atkeson spoke in an interview with CNBC.
For Atkeson, those notes marked a shift: A more polarized electorate began to lose faith in institutions like polling places, and voters may no longer be as willing to talk to her.
At the same time, technology was advancing and landlines or mail were no longer foolproof ways to contact survey respondents.
“People weren’t answering their phones,” Rachael Cobb, a professor of political science at the University of Suffolk, told CNBC. “Even in the last decade, you could try 20 callers to get the one you need. Now it’s double that: 40 callers to get what you need. So each poll takes longer and is more expensive.”
Polarization and technology are among the obstacles that pollsters say complicate the task of conducting accurate voter surveys.
As a result, poll workers have made some big mistakes over the past several election cycles.
“If you look at some of the big misses, I mean they’re pretty big,” Atkeson said.
Investigate blind spots
One of the big mistakes that has scarred the voting industry is the 2016 presidential elections, when several headlines spread the news claiming that Democratic candidate Hillary Clinton’s chances of winning against Republican candidate Donald Trump were around 90%.
An industry-wide post-mortem identified several key causes of the 2016 election flop.
Certain factors were beyond pollsters’ control.
For example, according to the American Association for Public Opinion Research, some voters decided whose name to write on their ballot at the last minute, making accountability difficult.
And some voters were shy about their support for Trump because of his controversial rhetoric during the 2016 campaign. As a result, they didn’t always admit their voting intentions to pollsters.
But other factors were direct results of methodological supervision.
“People weren’t taking representation in education into account,” said Matin Mirramezani, chief operating officer at Generation Lab, a polling organization that specifically targets young voters. “Education is a lesson learned from 2016.”
White, non-college-educated voters, who made up much of Trump’s base, were undercounted in the 2016 polls, in part because people with higher education are “significantly more likely” to respond to surveys than those with lower training, according to AAPOR.
Despite identifying these problems, polls during the 2020 election produced the highest margins of error in four decades, once again underestimating Trump’s support, AAPOR found.
And the 2022 midterm elections never saw a “red wave” of voters, which the media was convinced would overwhelmingly put Republicans back under control of Congress. Democrats retained their majority in the Senate and ceded the House by a narrow margin.
Course correction 2024
Heading into the 2024 rematch between Trump and President Joe Biden, pollsters are trying different strategies to avoid a repeat of history and accurately capture the elusive Trump vote.
First, pollsters have adjusted their approach to “weighting,” a method that assigns each respondent a multiplier to change how much their answer affects the overall outcome of the survey.
Pollsters have always used weighting to create survey samples that accurately reflect the electorate in terms of gender, age, race or income. But after 2016, they pay special attention to weight education.
Atkeson suggested that pollsters go beyond the 2024 education weighting and consider variables such as how someone voted in 2020, or even whether they rent or own a home, or whether they are a blood donor.
“You just start tagging everything you can,” Atkeson said. “Anything that can tell us, ‘What does the population really look like?'”
In addition to the weighting, pollsters are paying more attention to survey respondents they previously ignored.
“Some people will start a poll, they’ll tell you who they’re going to vote for and then they say, ‘I’m done. I don’t want to talk to you anymore. Goodbye,'” said Don Levy, director of Siena College Research Institute, which helps conduct polls for the New York Times, told CNBC. “We did not count those people in 2020 and 2022.”
But this time, Levy says they’re counting the drop-offs.
They found that if they had included those impatient respondents in 2020 and 2022, their poll numbers would have moved “about a quarter point toward Trump,” Levy said, eliminating about 40% of their errors.
Levy added that SCRI is also taking an extra step to target Trump voters by modeling their sample with a higher survey quota for people considered “high probability Trump voters in rural areas.”
“If you think of them as M&Ms, let’s say the Trump M&M ballot is red,” Levy said. “We still have a few extra red M&Ms in the jar.”