Matthew Rodier/Sipa USA via AP Images
Rampaging white mobs and paramilitary Proud Boys abolish any nuance, waving their Confederate, American, and Trump flags.
Just as the intelligence community may have missed the size, organization, and determination of Trump’s supporters to keep Donald Trump as president, the polls in 2016 and 2020 missed Trump’s ability to bring his base of white working class, evangelical, and rural voters into the electorate, to save white people from a changing America. Trump called Mexican immigrants “murders and rapists,” sent signals to the KKK, and instituted the Muslim ban. His political mission was defined by “good people on both sides,” closing the government to fund the border wall, and telling the Proud Boys to “stand back and stand by.”
Trump won the Republican nomination in 2016 by winning over the Tea Party, evangelical, and pro-life blocs, each of them determined to save America from Barack Obama, the first Black president. His enflaming racial resentment gave Trump an unassailable base in his party. But what the general elections reveal is that 40 percent of all Americans is fully part of an anti-establishment, God-first, racially resentful, anti-democratic bloc, who live in a right-wing media cocoon and adore Donald Trump. This bloc of white rural, evangelical, and working-class male voters rushed to the polls in both 2016 and 2020. And critically, they are three of every five Republicans.
More from Stanley B. Greenberg
Chanting “Make America Great Again,” “U-S-A.! U-S-A!” and “Stop the Steal,” Trump’s violent vanguard assaulted the Capitol to stop the counting of Electoral College votes. Rampaging white mobs and paramilitary Proud Boys abolish any nuance, waving their Confederate, American, and Trump flags. They aren’t sure anybody but Donald Trump will be as uncompromising, racist, and anti-establishment as they desire. And they don’t think anyone else on their side can win.
The proportion of whites without a four-year college degree eligible to vote fell a point from 45 to 44 percent since 2016, and the proportion of voters older than millennials fell 8 points from 71 to 63 percent, according to post-election analysis by the Voter Participation Project and Center for Voter Information. Only Trump’s ferocious campaign, which raised white working-class turnout by 7 points, was able to neutralize the demographic trends, discredit the polls, and make the election much closer than anyone expected.
I know America can fix the polls because I had to change my own to better accommodate anti-establishment and nationalist voters in Britain and Israel.
With that in mind, I am skeptical when I read a Washington Post headline blaring, “After the storming of the Capitol, Trump grows even less popular” or when poll after poll says that Trump left office with his lowest job approval of his presidency. Really, are Trump supporters completing polls these days?
Progressives, Democrats, and “Never Trumpers” will not govern successfully unless they see and understand that working people of all colors are frustrated with stagnant incomes, health care costs, growing inequality, and a politics rigged for the rich. But they also won’t be able to govern successfully unless there’s a strategy to marginalize this bloc of white anti-establishment, God-first, nationalist, racist, and anti-democratic voters determined to save the nation from its demographic destiny.
That starts with fixing the polls so you can have believable intelligence.
I know America can fix the polls because I had to change my own to better accommodate anti-establishment and nationalist voters in Britain and Israel.
Tony Blair and the Labour Party retained me, after the failure of the election-eve polling in 1992 was described as “the most spectacular in the history of British election surveys.” The polls missed the Conservative victory, according to post-election analysis, because of late deciders, not getting the right proportions of older and working-class voters, and above all “shy Tories”—Conservatives who refused to disclose their vote, or refuse to be interviewed at all.
I made changes to get the right proportion of older voters and self-identified Conservatives, but most important was incorporating the role of newspapers and tabloids. So in all my polls, 41 percent read Tory papers, particularly The Sun. It was wholly owned by Rupert Murdoch and it set a populist, anti-elite, anti-academia, anti-Europe, and anti-immigrant tone that working-class voters heard.
In 2016, The Sun led the charge for Brexit. On Election Day, its full tabloid-page headline was “INDEPENDENCE DAY,” with the teaser, “you can free UK from the clutches of the EU today.” The polls were not very far off, but they failed to pick the winner in a country highly polarized around “leave” and “remain.” And who won is all that matters. That is why my polls in Britain today have 52 percent “leave” voters, or nobody, including myself, will believe they are representative of the British public.
I became pollster for Ehud Barak and the Israeli Labor Party after the 1996 election, when Israelis “went to sleep with Peres and woke up with “Bibi”—Prime Minister Benjamin Netanyahu. The newspaper polls and exits missed the Netanyahu’s victory, but also underestimated the vote for Shas, the ultra-orthodox party led by a charismatic leader, who was critical of the Israeli Ashkenazi establishment and secular left.
We were polling after the signing of the Oslo Peace Accords, when the parties of the right were competing, with more and more venom, over which leader was most distrustful of Palestinians and Israeli Arabs.
So in my first campaign polls in 1999, I made sure that 50.5 percent of the respondents had voted for Netanyahu. Going forward, I made more and more changes that favored the right, including locking in the proportion that were religiously Mesorati and Sephardi. I also locked in 45 percent identifying with the “right.”
In 2015, I conducted focus groups for Isaac Herzog’s Zionist Union Party and flew to Israel on Election Day, when public and campaign polls before the final weekend showed him ahead of Netanyahu’s Likud by about five seats. Exit polls then showed the lead parties with equal number of seats, and in the end, we lost by five.
That election had clear lessons for me: Don’t underestimate the ability of a charismatic leader like Netanyahu with an electoral base worried about becoming a national minority. In Israel, polls cannot be conducted on the final weekend. So after we stopped polling that year, Netanyahu declared there will be no Palestinian state and warned that buses were bringing Arabs to vote to support “the left.” He then passed the Nation State Law, researcher Dahlia Scheindlin writes, that made Israel “the nation-state of the Jewish people exclusively.”
Fixing the polls in Israel and Britain meant accommodating voters that were almost always working class, intensely nationalistic, hostile to meddling global institutions, and skeptical of secularism; they worried about Arabs or immigrants, and were distrustful of the establishment, elites, judiciary and mainstream media. They operated within their own media bubble, and their voting choices were shaped by a spiritual leader or media group. They worried about the victory of the secular left and their elites, and they embraced populist leaders who attacked democratic institutions and used racial fears to mobilize and consolidate their vote at the campaign close.
In America, the polls were wrong in 2016 and 2020 because we just did not have enough appreciation of the potential size of this anti-establishment, God-first, racist, and nationalist bloc, as well as Trump’s ability to enflame and grow their vote.
I wrote in The Atlantic in July 2020, “Believe the Polls This Time,” because national and state polls were interviewing a much bigger number of white non-college voters. Yet that did not save the polls from having almost the same scale of error they did in 2016.
Trump created the potential for polling error by building up massive support with the white working class in 2016, and by pushing away college-educated voters every day in office. Nate Cohn oversaw The New York Times’ polling in 2020 and he concluded that Trump voters hid from the pollsters, while Biden’s more educated voters trusted them and took their calls, particularly during the lockdown.
We are living in a time where the Republican Party is still battling with growing desperation to repudiate Barack Obama.
Trump held rallies during the campaign that sounded more like a “race war” than a conventional campaign, as I wrote in the Prospect. This strategy brought in millions of new voters and dramatically increased the turnout of rural and white working-class voters. They produced almost all of this election’s polling error.
The rest was produced by Trump’s back-to-back last weekend rallies, which were reminiscent of Netanyahu’s in that they shifted late deciders and the undecided.
In the end, Trump’s ferocious 2016 and 2020 campaigns won him 68 percent support with working-class men, as well as 71 percent with white rural voters and 84 percent with white Evangelicals. Together, these groups comprise 40 percent of battleground voters.
So, this is the lesson from Israel and Britain: Don’t believe any of new polls unless four in ten respondents are white working-class male, Evangelical and rural anti-establishment voters, and unless at least two-thirds of the white working-class men and white rural voters chose Trump in 2020.
You can see the power of that change when I simply raise the proportion of white voters without a four-year college degree in my survey conducted one month before the election from 43 percent to 48 percent, the correct number according to the exit polls. My survey would have shown Biden with only a 3-point lead in closely contested battleground states—very close to his actual margin on Election Day.
That “fixes” the polls, but you can’t join the battle to save America without a lot of imagination. If the proportion of white college and non-college voters matched the 2018 midterms, last November would have been a “blue wave.” But that will only push Trump or a Trump-like leader forward to run again a populist, anti-elite, nationalist, and racist campaign to defeat the liberal, secular left. The odds will perhaps be longer without Trump, but we can no longer underestimate their imagination.
We are living in a time where the Republican Party is still battling with growing desperation to repudiate Barack Obama and defeat his new American majority of Blacks, Latinos, unmarried women, and millennials. That has created a period of intense polarization, where small differences in voter turnout, like in Georgia, is the main determinant of each party’s fate.
Hopefully going forward, the Capitol police and political leaders will have the intelligence to win this battle for America.
