A Republican canvasser visited Tiffany Blythe in Mooresville, North Carolina, last month. Photo / Travis Dove, The New York Times
Republicans insist that millions of Americans want to vote for Trump but won't admit it. Polling experts tell a different story.
It wasn't the most obvious spot for a flag that people usually buy to make a big statement. But there it was, peeking out from the inside wall of a garage, the white "Trump 2020" lettering just visible from the street in this suburban Charlotte neighbourhood.
From the front porch, Tiffany Blythe, a stay-at-home mom, said that she and many of the people she knows would be voting for President Donald Trump in November — but that many of them were nervous talking about it. And that hesitation is why Blythe doesn't trust the polls that are now forecasting losses this fall for Trump and other Republicans in North Carolina and beyond.
"I'm not buying it," Blythe said. "There are a lot of silent voters, and more will come out before the election. I think a lot of states are turning red from blue, but you don't hear about that in the media."
The belief that Americans aren't getting the real story about Trump's chances for reelection has taken hold among many of his supporters. For Trump loyalists, it is an appealing story and one with some validity: The news media, which largely failed to anticipate Trump's victory in 2016, are undercounting his voters, many of whom are even more reluctant today than they were four years ago to declare themselves in his camp.
Trump makes this argument often; on Saturday evening, he told reporters that "we have a silent majority the likes of which nobody has seen." One of his pollsters, John McLaughlin, has even put a name to this supposed flaw in the data, predicting that the "hidden Trump voter" will prove the news media wrong.
But the idea that there are substantial numbers of Trump voters who will emerge from hiding on Election Day, large enough to sway the outcome, is not supported by the latest public opinion research — or by a proper understanding of what happened in past elections where the voter surveys were off, said pollsters who work for Republican and Democratic candidates.
This does not mean that Joe Biden's lead, with the most recent national polls putting him ahead by as much as 10 points, won't tighten. And public opinion experts said there is growing evidence that Americans across the spectrum have become more skittish about sharing their political preferences outside a trusted group of like-minded people. But it would be a huge leap to conclude that the country's tense political dynamics are causing people to lie to pollsters in large enough numbers to explain Trump's poor standing.
"There are many people who are voting for Trump who are in environments where it's politically untenable to admit it because he's become so toxic," said Whit Ayres, a Republican pollster. "But I'm still not convinced that not telling your business associate or the people in your Rotary Club or the people in your country club is the same thing as not telling a pollster."
The possibility that Americans are hiding their true intentions from pollsters has provided an irresistible sense of intrigue to presidential elections before, even though there are few confirmed examples where it made a difference. Political experts compare such speculation to the quadrennial predictions of a brokered convention, which has not occurred since 1952.
In 2008, there were questions about whether support for Barack Obama was being inflated in the polls by people who didn't want to say they weren't supporting him. That did not happen.
Four years ago, some suggested there might be a similar phenomenon at work with Trump supporters who were too embarrassed to reveal themselves. And when Trump won by squeaking out victories in a few battleground states, his backers argued that shy voters were a reason the polls missed his strength in those places.
"The idea that people lie, it's an interesting theory, and it's not like it's completely off-the-wall," said David Winston, a pollster who works with congressional Republicans. "But it's obviously a very complicated thing to try to prove, because what do you do? Ask them, 'Are you lying?'"
Winston said that many proponents of the theory about hidden Trump voters rely on what is known as the Bradley effect, named after Tom Bradley, the former mayor of Los Angeles who lost the 1982 California governor's race despite polling consistently ahead of his white opponent. Among political scientists, the theory that emerged to explain the gap between the polls and the election results was that white voters were worried about appearing racist if they did not say they were supporting Bradley, who was Black.
But some have questioned the validity of the Bradley effect, including Blair Levin, one of Bradley's former advisers, who has argued that Bradley lost because of a complicated mix of factors, among them a robust Republican absentee voting campaign and an unpopular gun control initiative on the ballot, both of which turned out a surge of Republican voters.
If voters were indeed afraid of voicing their support for the president, Winston said, other numbers in the poll would reflect that, like seeing an uptick in the percentage of undecided voters rather than a rise in support for Biden.
"It would not be people saying they are voting for Biden," he said, "but that they're undecided."
While the effects of a hidden Trump vote are certainly overstated by the president's allies, that does not mean that no evidence exists that polls are missing some of his voters. A small percentage of his support is probably being undercounted and has been in the past, public opinion experts said. And in states like North Carolina, where the margin of victory could be narrow, the undercount could make a difference between a poll being right or wrong.
"We assume the race will tighten, and as that happens, the size of the shy Trump vote could very easily come into play," said Neil Newhouse, a Republican who led Mitt Romney's polling in 2012.
In 2016, Newhouse said that Trump tended to score 2 or 3 points higher in phone surveys when respondents were asked to press a button to record their preferences rather than talk to a live person. In postelection polling, when he asked people if they had ever been unwilling to talk about their vote, 35 per cent of Trump voters said yes. And they tended to be women from Democratic-leaning counties.
Newhouse has picked up further evidence of such reluctance recently. In polls he conducted late last month in North Carolina and Iowa, he found that one-quarter to one-third of voters answered "yes" when asked if they knew someone who is voting for Trump but would not say so to anyone but their closest friends.
"This totally confirms the notion of 'shy Trump voters,'" Newhouse said.
But, he added, if polls are undercounting some Trump voters — a group that tends to be uniquely expressive and adamant about their support for the president — no one can say by how much.
And in any case, pollsters said they have corrected one of the biggest mistakes they made in 2016, when they failed to account for the high numbers of voters without college degrees who turned out, many of whom voted for Trump. And they are including a larger pool of possible voters in surveys — not just people who say they are likely to vote, as pollsters often do — because they anticipate historic turnout.
One variable that public opinion experts are still grappling with is how the polarised political climate is affecting the accuracy of their work. Recent research has shown that conservatives fear they are more likely than moderates and liberals to be targeted for being honest about their political beliefs, although self-censorship appears to be rising among most Americans.
In a survey conducted last month by the libertarian-leaning Cato Institute, 77% of conservatives said they felt unable to share their political opinions because others might find them offensive, an increase from 70 per cent who said so in 2017. Among liberals, 52 per cent reported those concerns, compared with 45% three years ago; among moderates it was 64 per cent versus 57 per cent previously.
Interviews with voters like Blythe suggest the unease over political speech has become highly politicised.
"I see Democrats not protecting our freedoms, and that includes freedom of speech," she said.
Still, Geoff Garin, a Democratic pollster, said what really explains Trump's low standing is not flawed methodology but rather the president's inability to expand his coalition.
"The problem for the Trump campaign is not the silent Trump voter but the disappearing Trump voter," Garin said. "And there are a lot more disappearing Trump voters than there are silent ones."
The flip side of that dynamic — people who did not vote for Trump in 2016 and who have now become supporters — is barely a factor.
"Among Clinton voters, it is less than 5 per cent," Garin said.
Henry Olsen, a senior fellow at the Ethics and Public Policy Center, has noted another warning sign in the data for anyone holding out hope that a hidden pocket of voters will save Trump: There aren't many people backing a third-party candidate right now who could potentially switch to the Republican ticket. Olsen said in an email that while some conservative voters often park themselves with the Libertarian candidate in the months before an election, many eventually come home to the Republicans.
But the Libertarian Party candidate, Jo Jorgensen, is registering minuscule support in the polls now compared with the high single-digits that Gary Johnson, the party's 2016 nominee, was seeing at this point four years ago.
Still, among Republicans in North Carolina, none of these facts seem to dent their confidence or convince them that they should worry about polls showing Trump losing to Biden — or others that have the state's Republican senator, Thom Tillis, behind his Democratic opponent, Cal Cunningham.
Tim Moore, the Republican speaker of the North Carolina House of Representatives, said, "Here's the thing I wonder: If you think about how divided we are as a country, how polarised the two sides are at the national level when it comes to the presidential race, how many people are truly undecided?"
Moore believes that voters who say they are undecided today will decisively break for Trump and help him win the state.
In Blythe's suburban Charlotte neighbourhood, many people who answered the door were willing to say they supported Trump for reelection. But when asked for their names, few of them agreed.
"I'd like to keep my friends," one woman said.
At another residence down the street, a man who opened the door barely enough for a reporter to hear him say he was voting Republican joked, "Don't burn down my house."
Written by: Jeremy W. Peters
Photographs by: Travis Dove and Anna Moneymaker
© 2020 THE NEW YORK TIMES
Six people have now died in a suspected mass methanol poisoning event.