How did pollsters get the 2020 race so wrong – again – or did they?
Nov 10, 2020, 5:26 PM
SALT LAKE CITY — Until clerks count all the votes from the 2020 presidential race, it’s very difficult to judge how far off the mark the polls will end up. But ahead of the election, Democrats looked likely to take control of the White House and Senate with ease, according to polls.
As the votes rolled in, it became clear that the 2020 polls had underestimated Republican support in several states.
The polls this election cycle predicted a Democratic sweep with former Vice President Joe Biden leading by an average of 7.4 percentage points ahead of Mr. Trump in national polls, according to NBC News’ polling average.
President Trump on polling
On Nov. 5, President Trump claimed pollsters knew the election would be close but intentionally skewed the results.
“The pollsters got it knowingly wrong. We had polls that were so ridiculous, and everybody knew it at the time. There was no blue wave that they predicted … that was false, that was done for suppression reasons — instead, there was a big red wave.” he said.
TRUMP ON GOP WINS IN CONGRESS: “There was no blue wave that they predicted … that was false, that was done for suppression reasons — instead there was a big red wave.” pic.twitter.com/ClY1NQ2l1p
— Breaking911 (@Breaking911) November 6, 2020
The Pew Research Center says the most common question about surveys is: “Can they be trusted?” Perhaps, maybe more to the point: “Which polls can I trust?”
The answer to the first question is — you have to be realistic about what polls can and can’t do, say the survey methodologists and social science researchers.
A 2018 review of polling accuracy found that “relying on voter-intention polls from more than 200 elections in 32 countries over a period of more than 70 years, there is no evidence that poll errors have increased over time and the performance of polls in very recent elections is no exception.”
2016 was a swing and a miss, too
Pollsters faced a public reckoning after failing to predict then-candidate Trump’s win over Hillary Clinton in the 2016 presidential race. They underestimated support for Mr. Trump in the some Midwest states, such as Michigan, Pennsylvania and Wisconsin, which the candidate flipped from blue to red that year.
In 2016, state pollsters also didn’t adjust for the level of education of survey respondents and were left with a biased sample. College graduates broke for Clinton while high school grads went for Trump. But pollsters still disagree on how important this facet is in accounting for the error.
The average of the final national polls suggested that Clinton would win the overall popular vote by 3 percentage points; she ultimately won by 2 points, according to Pew Research Center.
Americans refusing to participate in polling has also hurt accurate predictions. In 1997, 36% of citizens routinely agreed to participate in pollsters’ surveys, but by 2019 the response rate had plunged to 6 percent because lots of people don’t answer phone calls from unknown numbers, a Pew Research Center study found.
2020 is a different beast for polls
The record turnout this year marked the highest rate among eligible voters since 1900 at nearly 67%, according to the U.S. Elections Project. Perhaps the polling error lies there.
“In 2020, rather than having that error [of 2016] focused in just a handful of states, this year really it looks like it was quite widespread. I can’t find any state where the polling error was in Trump’s favor,” Charles Franklin, who directs the Marquette Law School Poll, told CNBC.
“2020 is not the same kind of polling surprise as it was in 2016,” said Joseph Campbell, a professor of communication at American University and author of “Lost in a Gallup: Polling Failure in U.S. Presidential Elections.”
Independent pollster Richard Baris’ final poll of the crucial battleground state of Florida gave Mr. Trump a 2-point lead — 1 percentage point shy of the president’s 3.3-point victory in the state. In contrast, the New York Times-Siena poll predicted a 3-point win for Biden in the Sunshine State. Quinnipiac estimated a 6-point Biden edge.
“I think they bully each other. They herd. That’s when you start to mirror other pollsters because you’re afraid that Nate Silver [at FiveThirtyEight] or CNN is going to call you an outlier,” he told the New York Post.
“This industry is dominated by left-wingers. And a big, big problem is they’re trying to profile the voting behavior of people they don’t understand and may even despise,” Baris told The Post.