November 5, 2020 1:08:23 pm
Garden signs supporting President Donald Trump and Senator Susan Collins (R-Maine) in Levant, Maine, on Wednesday. (Hilary Swift / The New York Times)
Written by Giovanni Russonello
As the results progressed on Tuesday night, so did a strong sense of déjà vu. The pre-election polls, it seemed, had once again been misleading.
While the nation awaits the final results of Pennsylvania, Arizona and other key states, it is already clear, no matter who ends up winning, that the industry did not fully take into account the missteps that led it to underestimate the support of Donald Trump. four years ago. And it raises the question of whether the polling industry, which has become a national fixation in an era of data journalism and statistical forecasting, can survive another crisis of confidence.
“I want to see all the results, I want to see where these deviations from the pre-election polls and the final margins are,” Christopher Borick, polling director at Muhlenberg College in Pennsylvania, said in an interview. But there is ample evidence that there were major problems again. We’ll see how deep they are. “
In some states where polls had projected Trump to lose by a narrow margin, such as Ohio, Iowa and Florida, he had already been declared the winner by a comfortable margin Tuesday night. And in states that seemed more than likely to go for Joe Biden, like Michigan and Nevada, the result was too close to call last night.
To some extent, it was clear that this process would be difficult to manage. With a large number of ballots being mailed and the first votes in person still unaccounted for, early results in most states gave an exaggerated sense of Trump’s strength, as voters in Republican areas outnumbered the Election Day, and those ballots were often the first to be tabulated.
It’s also possible, said Patrick Murray, polling director at Monmouth University, that Republicans’ efforts to prevent certain populations from voting easily had a considerable impact, a factor that pollsters knew would be immeasurable in their polls.
“We need to know how many votes were rejected,” he said. “I won’t know, until someone gives me some information, what happened. And we may never know. “
He added: “We will never know how many ballots were not delivered by the post office.”
But what is now clear based on the ballots that have been counted (and in almost every state, the majority have been) is that there was an overestimation of Biden’s support across the board, particularly with white voters and with men, indicate preliminary exit polls. .
While polls had heralded a departure from Trump among white voters 65 and older, that never fully took shape.
Partly as a result, Biden underperformed not just in polyglot states like Florida, but also in large white suburban areas like Macomb County, Michigan, where he was expected to do well.
Borick pointed out that while statewide polls had largely failed in 2016, they remained stable in the 2018 midterm elections. This led him to conclude that people’s opinions of Trump can be particularly difficult to measure.
“In the end, like so many things related to Trump, there can be different rules when voting an election with him on the ballot,” Borick said. “I am a quantifiable type of human being; I want to see evidence. And I only have two elections with Donald Trump in them, but both seem to behave in a way that others do not. “
Analyzing pre-election polls alongside exit polls is like comparing apples to apples: If one batch is rotten, the other probably is too. But exit polls can still provide some clues as to what the pre-election polls might have missed.
At the top of that list is Trump’s strength among college-educated white voters, particularly men. According to exit polls, the candidates divide white college graduates evenly, after an election season in which nearly every major nation and state polls on the battlefield had shown Biden ahead. with white graduates.
And if there is a tendency for polls to not sufficiently represent Trump’s support, this is not just affecting college-educated voters, as “shy Trump” theorists have often suggested. Some studies had posited that highly educated Trump supporters might be more likely to say they prefer their opponent because of social pressure. In many high-quality telephone polls before the election, Trump’s support ranged in the 50s to 50s among untitled white voters. But exit poll results put their support for this group firmly in the mid-1960s, almost on par with their totals in 2016.
There is also no certainty about what percentage of the electorate comprised these voters. Pollsters were perplexed by this question in the wake of 2016 and reached various conclusions; This year’s results are likely to reignite that discussion.
On the subject of the coronavirus pandemic, it is also notable that compared to most pre-election polls, exit polls showed a lower proportion of respondents favoring caution over a quick reopening. As of Wednesday afternoon, with final adjustments still anticipated to the data, there was only a 9 percentage point split between voters who said it was more important to contain the virus and those who said they cared more about rushing to rebuild the economy, according to the polls. In pre-election polls, the split had typically been in the double digits, with a sizable majority of voters across the country saying they preferred caution and restraint.
It appears that the virus was also a minor motivator for voters than many polls seemed to convey. This year, exit polls, conducted as usual by Edison Research on behalf of a consortium of news organizations, had direct competition from a new probability-based voter poll: VoteCast, collected through an online panel. put together for The Associated Press by NORC a University of Chicago research group. Looking at the divergence between the exit poll numbers and the responses to the VoteCast poll, we can see that there were far more voters who viewed coronavirus as a major problem in their lives than people who said it was the problem that They were. voting.
The VoteCast poll found that more than 4 in 10 voters said the pandemic was the number one problem facing the country when presented with a list of nine options. But in exit polls, when asked which topic had the greatest impact on their vote decision, respondents were less than half as likely to indicate it was the pandemic. Much more likely was the economy; behind that was the issue of racial inequality.
Not all pollsters did badly. Ann Selzer, long considered one of the nation’s top pollsters, published a poll with The Des Moines Register days before the election that showed Trump had a 7-point lead in Iowa; which seems to be in line with the actual result so far.
In an interview, Selzer said that this election season he had stuck to his usual process, which involves avoiding assumptions that a year’s electorate will resemble previous years. “Our method is designed so that our data reveals to us what is happening with the electorate,” he said. “There are some who will weigh their data taking into account many things: votes from past elections, what was the turnout, things from the past to project into the future. I call that survey backwards and I don’t. “
Inevitably, Robert Cahaly and his mysterious Trafalgar Group, which screened a series of close races on the battlefields, will also receive another look from curious commentators wondering why their polls have been so accurate, both in 2016 and this year.
The firm was one of the only pollsters to show Trump’s strength in the Midwest and Pennsylvania four years ago, and while its polls this fall may end up a bit rosy, they appear to have gotten closer to the final horse. Race results in states like Michigan, Wisconsin and Nevada than other pollsters, by not downplaying Trump’s strengths.
.
