‘Shy Trump Voters’ Re-Emerge as Explanation for Pollsters’ Miss
(Bloomberg) -- Donald Trump’s unexpectedly strong election showing against Joe Biden is reviving the notion of the so-called shy Trump voter among pollsters, who had consistently yet inaccurately predicted a far weaker performance by the president before the Nov. 3 contest.
The outcome of the race marked the second time in four years that pollsters understated Trump’s support in the months leading up to the election. Now, polling firms are faced with another possible culprit: a legion of voters who either refused to say they were voting for Trump or declined to participate in polls at all.
In the aftermath of Trump’s surprise 2016 victory over Hillary Clinton, pollsters dismissed the idea that any voters had hidden their support, instead concluding the surveys had underestimated White voters without a college degree.
While pollsters say they fixed that underestimation and ultimately picked the right winner, Biden’s margin of victory was half what had been predicted, and Trump’s 73 million-vote strength was surprising to many who had closely followed the surveys.
Biden won the election with 79.4 million votes to Trump’s 73.5 million and with 290 Electoral College votes so far to Trump’s 232, according to unofficial numbers from the Associated Press. Vote counting is still under way in several states.
“There were a number of polls this year that on paper looked fabulous. They did everything right — but they were still among the polls that had problems. And so that does suggest that we have more work to do,” said Courtney Kennedy, the director of survey research at Pew Research Center. “I was really hoping we solved it with the 2016 autopsy,” a report from the American Association for Public Opinion Research that sought to examine what went wrong and that Kennedy co-authored.
Pollsters say they are now questioning whether the problem goes deeper than voters just being wary of admitting they back Trump, especially after his four years in office. Some Trump supporters, they worry, may be so distrustful of polls that they’ve become a blind spot in surveys.
“Shy Trump voters are only part of the equation. The other part is poll deniers,” said Neil Newhouse, a Republican pollster. “Trump spent the last four years beating the crap out of polls, telling people they were fake, and a big proportion of his supporters just said, ‘I’m not participating.’”
In a survey conducted after Nov. 3, Newhouse found that 19% of people who voted for Trump had kept their support secret from most of their friends. And it’s not that they were on the fence: They gave Trump a 100% approval rating and most said they made up their minds before Labor Day.
Suburbanites, moderates and college-educated voters — especially women — were more likely to report that they had been ostracized or blocked on social media for their support of Trump. Some, Newhouse said, even confessed to lying to pollsters in the past — although the numbers weren’t large enough to draw conclusions.
The shy Trump effect is a new iteration of a phenomenon known as social desirability bias that pollsters have observed for decades. In the parliamentary elections in the U.K. in 1992, conservatives did better than the polls suggested — leading to the discovery of a “shy Tory” effect. And in U.S. elections where Black candidates have run in mostly White districts, polls have overestimated opposition because more voters would claim to be undecided.
University of Arkansas economist Andy Brownback conducted experiments in 2016 that allowed respondents to hide their support for Trump in a list of statements that could be statistically reconstructed. He found people who lived in counties that voted for Clinton were less likely to explicitly state they agreed with Trump.
“I get a little frustrated with the dismissiveness of social desirability bias among pollsters,” said “I just don’t see a reason you could say this is a total non-issue, especially when one candidate has proven so difficult to poll.”
Indeed, some of Trump’s most hardcore supporters aren’t shy at all.
“They’ll show up at his rallies. They just won’t talk to us,” said Monmouth University pollster Patrick Murray.
That has him worried that public skepticism of polls could become a self-fulfilling prophecy.
“This could all be purely a Trump factor and this could all disappear on its own once Trump’s not on the ballot,” Murray said. “But I am less concerned about whether we’re going to get the next execution right or not than whether Donald Trump has systematically undermined our ability to have a serious conversation about politics.”
Historically there’s been no consistent partisan bias in political polling, and as recently as 2012, polls overestimated Republican support. But high-profile misses in two election cycles in a row could undermine public confidence in polls, particularly among conservatives.
Those are doubts that Trump himself has been eager to stoke. He tweeted last week about “fake pollsters” at ABC News and the Washington Post after a poll two weeks before the election showed Biden leading Wisconsin by 17 points. The current margin is 0.7 points and the Trump campaign on Wednesday paid $3 million for a partial recount.
The Washington Post said most of its polls fall within 2.8 points of the final result, earning it an A+ rating from ABC-owned poll aggregator FiveThirtyEight. Spokeswoman Molly Gannon Conway said the newspaper is reviewing all its polling.
A full accounting of the performance of 2020 polls is only possible once final results are in. Already, late-counted votes have made the polls appear more accurate than they did on election night.
Surveys of Florida, Iowa, Ohio and Wisconsin voters significantly underestimated Trump’s support in 2020. But polls in other states — often by the same polling outfits — were spot on. The polling averages in Arizona, Pennsylvania and Nevada appear to have predicted the final result to within a fraction of a percentage point.
AAPOR said it’s waiting for final results in December to figure out how far off the pre-election polls were and what caused any errors. “It is premature to make sweeping judgments on the polls’ overall performance before all the ballots are counted. Patience is necessary,” the group said in a statement.
©2020 Bloomberg L.P.