She and other experts point to such situations as evidence that forecasts cannot accurately project specific voting outcomes, no matter how much the public would like them to, nor can they fully account for the impact of variables like the spread of misinformation on social media, voter intimidation efforts, or massive mail-in balloting during a pandemic.
It’s not even clear, in many cases, that pollsters are reaching the right mix of people. The sheer number of polls conducted during every campaign cycle — focused on everything from presidential preference to county-level ballot measures — leaves some voters, especially in swing states, feeling overwhelmed by incessant inquiries via phone, text message, and email. Experts say there is some evidence to suggest that the same, relatively small group of people responds to polls over and over, giving their opinions disproportionate prominence, while other cohorts are underrepresented. Pollsters try to adjust for those shortcomings with various methodologies, but it’s impossible to verify their accuracy.
“A big worry in political polling is that there are a group of people — distrustful of the media, academia, science, etc. — who do not want to engage with pollsters and survey research,” said Neil Malhotra, a professor of political economy at Stanford.
Rick Ector, a firearms instructor from Detroit who voted for Trump, told BuzzFeed News that he understands why people like him blow off inquiries from pollsters. “If we get a call from some agency asking us who we are voting for and we don’t know this person or where they are calling from and what they will do with that information, there’s high risks and stakes involved,” he said. “People put themselves at risk for getting targeted.”
Jason Chatwell, a Trump voter in Kouts, Indiana, said he goes even further. “A lot of times I will say the opposite of what I’m really going to do if I get polled,” he said. “Mainly because I don’t trust anybody.”
Where polls are both more accurate and useful, said Sunshine Hillygus, a professor of political science and public policy at Duke University, is at measuring public sentiment, across a broad swath of the American public, on issues such as universal healthcare or the Green New Deal.
According to Atkeson, polling data focused on voting outcomes tends to push the public conversation away from policy and may discourage candidates from even expressing their positions on important issues. With Biden leading Trump by high single-digit and low double-digit margins in nearly every national poll throughout the late summer and early fall, Atkeson said, there was little incentive for him to take risks by talking about his own plans. Instead, he mostly stuck to criticizing the incumbent for his handling of the coronavirus.
“I know Biden has policies out there somewhere, but as far as I can tell all he talked about was COVID, COVID, COVID,” said Atkeson. “You can see why they chose that strategy.”
Candidates also have become adept at using polling data for other ends, such as fundraising. Although Mark Kelly surveyed extremely strongly overall against incumbent Martha McSally in the Arizona Senate race, his campaign pursued an underdog strategy, sending supporters cherry-picked data from less favorable polls in order to generate donations.
“Three polls show Mark Kelly statistically tied with Martha McSally in Arizona,” read the subject line of one such email sent Oct. 14. In fact, polls taken at that time showed him up by as much as 11 points. It seemed to work: By Election Day, Kelly had raised $33 million more than his opponent. As of Thursday morning, Kelly led by a comfortable margin of 4.2 percentage points, although the race has not yet been called.
Over the past dozen years, polling has blossomed from a service providing data to campaigns and news outlets to a media industry in its own right. Forecasters strive to be accurate not only for accuracy’s sake, but also because of the financial incentives driving their own businesses, which extend far beyond political races. The irony is that this kind of data-driven analysis, which Silver has said he developed as an alternative to the subjective, narrative cast of so much horse race political coverage, has instead breathed new life into horse race coverage.
Some have suggested that the most recent polling debacle could spell the end of polling as we know it. But there were similar calls after the 2016 election, too. Most observers doubt that kind of change is coming anytime soon.
“I don’t see the pollsters go away,” said Atkeson. “It’s a great story and one that keeps giving. Every day you get to run new models. The polling itself becomes part of the election narrative.” ●
Friends, this isn’t the time to be complacent. If you are ready to fight for the soul of this nation, you can start by donating to elect Joe Biden and Kamala Harris by clicking the button below.
Thank you so much for supporting Joe Biden’s Presidential campaign.