Some pollsters are still grappling with the same problems that plagued battleground state surveys four years ago.
As President Donald Trump tumbles in the polls and falls further behind Joe Biden, his campaign has returned to a familiar refrain: The polls are underestimating Trump’s appeal again.
They have a point.
Pollsters aren’t deliberately skewing their surveys against the president and his party, as Trump’s orbit alleges. The national polls showing Trump trailing Biden by an increasing margin aren’t “phony” or rigged.
But some pollsters, especially the relatively few who conduct surveys in battleground states, are still grappling with the same problems that plagued those polls four years ago. In fact, most pollsters believe that, on balance, state polls are overstating the scale of Biden’s advantage.
That was precisely the problem in 2016: The national polls were largely accurate, to within the margin of error. But there were too few state polls, and many of those that were conducted failed to collect accurate data, especially from white voters without college degrees in key swing states.
And those issues haven’t been fixed.
“I would say that most, if not all, of the concerns that we expressed still hold — some to a lesser degree,” said Courtney Kennedy, the director of research at the Pew Research Center and lead author of the polling industry’s post-2016 autopsy. “But I think some of the fundamental, structural challenges that came to a head in 2016 are still in place in 2020.”
Polling errors are not uncommon in presidential elections. But pollsters see a real risk this year that the same 2016 mistakes will be repeated. Their colleagues are still not accounting for the fact that voters with greater educational attainment are more likely to complete surveys — and more likely to vote for Democratic candidates.
“There’s still a number of state polls, in particular, that are not fixing this issue,” said Kennedy.
Biden’s current lead over Trump is so large — over 8 points in the national RealClearPolitics polling average, and an average advantage of 3 points or greater in Arizona, Florida, Michigan, Pennsylvania and Wisconsin — that a 2016-level polling error wouldn’t matter. A lead that large would probably guarantee Trump would be denied a second term, and even a polling miss on par with 2016 wouldn’t be enough to overcome it.
But that doesn’t mean the president’s standing is quite as dire as it looks on paper — the same problem that pollsters identified in 2016 remains. Not enough surveys are being conducted in the battleground states, and those that exist are failing to account for a key political dynamic of modern politics, especially in the Trump era: the rapid movement of lower-income white voters to Republicans and upscale whites to Democrats.
Pollsters are looking for answers. One of the major takeaways of the American Association for Public Opinion Research’s post-2016 autopsy was that state polls that didn’t weight, or adjust, their samples to include more white voters who hadn’t graduated college missed a key element of Trump’s coalition. In previous elections, the differences in white voters’ preferences along educational lines were smaller, but began to grow last decade and accelerated with Trump on the ballot in 2016.
“Before 2014, it wasn’t that big of a deal because the reality is non-college white voters and college-educated white voters — the distinction between the two wasn’t as dramatic,” said Democratic pollster Jefrey Pollock. “But starting with 2014, that began to cleave a lot and is now obviously humongous.”
GOP pollster Glen Bolger said he believes a combination of pollsters’ inability to get the right educational mix and to convince potential Trump voters to respond and answer truthfully to phone polls is pointing their surveys in a slight Democratic direction.
“I don’t know how big the effect is. I also don’t know what the ratio is between it being ‘shy Trump’ voters and interviewing too many college graduates and not enough non-college grads,” Bolger said. “But I do think those are factors in some of the polls that show a particularly wide lead for Biden at this point in time. And I do think that things will be closer in the states that the polls indicate right now.”
The question of polling accuracy came to head last week, after CNN published a national survey showing Biden with a 14-point lead over Trump among registered voters, 55 percent to 41 percent. The Trump campaign immediately went on the attack against its frequent foe: First, Trump himself tweeted out an internal campaign memorandum — prepared by John McLaughlin, one of his campaign’s pollsters — calling the CNN survey “skewed” against the president.
Then, Trump’s campaign sent CNN a letter, demanding that the cable network retract the survey results and apologize. CNN stood by their poll, which generally follows best practices and does weight by education. The denouement of the episode was a contentious interview of Trump campaign attorney Jenna Ellis on a CNN “Reliable Sources” that descended into shouting. (The Trump campaign also demanded an apology from CNN over host Brian Stelter’s conduct in the interview.)
While the CNN poll shows a larger lead for Biden than others, they all show the presumptive Democratic nominee with a significant advantage. Trump hasn’t led a national poll in the RCP database since February, and a Washington Post analysis found that pollsters that have conducted multiple surveys this year have consistently showed Biden gaining and Trump falling in their most recent polls.
The bigger problem appears to be in state polls, as evidenced by CNN’s own polling standards. The network’s polling standards state it won’t report on election surveys that “do not ensure that respondents of all education levels are adequately reflected.”
Pollock, the Democratic pollster and president of Global Strategy Group, described making “a lot of investments” in updating his own firm’s education models. “And that in itself is like a seismic change. Because whenever you have one variable like that that is so critical, if you’re getting it wrong, then the whole poll is wrong,” he said.
At last week’s annual AAPOR convention — held online because of the coronavirus pandemic — Nate Cohn, the New York Times data journalist who has worked with Siena College on their multi-million dollar polling partnership, observed that the state polls leaned way too far toward Democrats in 2014 and 2016. In 2018, he said, the polls were more accurate but still showed a Democratic slant, especially “in a number of white, working-class states,” like Indiana and Ohio.
And, Cohn noted in his presentation, it might be happening again this year.
“So far in 2020, it sure seems like Joe Biden is faring particularly well in the states where the polls were most biased toward Hillary Clinton four years ago,” Cohn told the virtual attendees.
As if on cue, a new poll was released in Michigan on Tuesday: it showed Biden ahead by a whopping 16 points.
Source: Politics, Policy, Political News Top Stories https://www.politico.com/news/2020/06/17/trump-polls-biden-324210