During a campaign rally Republican presidential candidate Donald Trump reads a statement made by Michelle Fields, on March 29, 2016 in Janesville, Wis.
Photo by Scott Olson/Getty

Poll denialism makes a comeback as impeachment, 2020 elections loom

Fox News released a national poll last week that was rather brutal for the White House. The data from Donald Trump’s favorite network showed, among other things, that 51% of Americans support the president’s impeachment and removal from office. The same results pointed to large segments of the population criticizing Trump’s “dealings with the Ukrainian president.”

Not surprisingly, a variety of news outlets, including the New York Times, took note of the Fox poll. According to Trump, that’s a mistake.

President Trump on Monday blasted The New York Times for using in one of its stories last week’s Fox News poll that said a majority of respondents wanted his impeachment.

“The Fox Impeachment poll has turned out to be incorrect,” he tweeted Monday. “This was announced on Friday. Despite this, the Corrupt New York Times used this poll in one of its stories, no mention of the correction which they knew about full well!”

In reality, there was no correction to “mention.” The president appeared to reference this New York Post piece, which argued that the sample in the Fox News poll included too many Democrats.

“Princeton, New Jersey, pollster Braun Research, which conducted the survey, noted 48% of its respondents were Democrats,” the New York Post’s analysis read. “But the actual breakdown of party affiliation is 31% Democrat, 29% Republican and 38% independent, according to Gallup.”

There are two important angles to this. The first is the flaw in the New York Post’s analysis, which may make Trump feel better, but which is nevertheless mistaken. The second is the re-emergence of poll denialism.

Let’s tackle these one at a time. At first blush, the effort to discredit the results from the Fox News poll may seem compelling: if Gallup says 31% of the electorate is Democratic, while Fox’s poll says it’s 48%, then many might look at the results as weighted and misleading. The trouble is, the New York Post may not have looked closely enough at the Gallup data its analysis cited.

Gallup asked self-identified independents, “As of today, do you lean more to the Democratic Party or the Republican Party?” With these “leaners” added to the mix, the Democratic electorate goes from 31% to 49%.

All of which tells us the sample in the Fox News poll wasn’t especially slanted after all. When Trump asserted yesterday that that poll “turned out to be incorrect,” this was “announced,” and there was a “correction,” he was completely wrong on each point.

But it’s the bigger picture that’s worth keeping in mind in the coming months, because the president is likely to keep trying to discredit survey results that fail to satisfy him.

In April 2018, Trump called into a radio talk show and argued that when it comes to understanding his public support, the responsible thing to do is take his approval rating and then “add another 7 or 8 points to it.” The president said this accounts for those people who like him, but are too embarrassed to admit to pollsters.

Soon after, at a rally in North Dakota, Trump pointed to a non-existent group of people to bolster his practice: “They said, some great people, they said, ‘Any time Trump gets a poll, add 12 to it.’”

More recently, Trump argued that people can add 15 percentage points to independent data on his public support, simply because he says so.

If this sounds vaguely familiar, it’s because poll denialism was a staple of Republican thought seven years ago. The Atlantic had this report in November 2012:

Mitt Romney and Paul Ryan both believed the public polls were wrong, and that they’d win on Election Day. Their wives did, too. “I don’t think there was one person who saw this coming,” a senior adviser told CBS News’ Jan Crawford. An advisor said of Romney, “He was shellshocked.” When Romney claimed on Election Day that he hadn’t written a concession speech, it sounded like trash talk. Apparently it wasn’t. How could they not have seen it coming? […]

Conservatives began claiming the polls were wrong, that they vastly overestimated what turnout levels would be among blacks, Latinos, and young people. UnSkewedPolls.com changed the number of Democrats and Republicans in polls to show Romney leading everywhere. You’d expect Romney’s campaign to play this up publicly to maintain supporters’ enthusiasm – like when political director Rich Beeson said the Sunday before the election that Romney would win more than 300 electoral college votes. But you don’t expect them to actually believe it. But Romney, his wife, Ryan, and his wife apparently did.

In the months leading up to Election Day 2012, many Republicans managed to convince themselves that independent pollsters simply didn’t know what they were doing. A site called UnSkewedPolls.com became quite popular in GOP circles because it took real results and shifted the data in a Republican-friendly direction, based entirely on the idea that all professional pollsters were relying on an inaccurate turnout model.

Those beliefs were ridiculous; Barack Obama cruised to a relatively easy re-election; GOP officials were stunned, and UnSkewedPolls.com became a punchline and a cautionary tale about the dangers of poll denialism.

Seven years later, some Republicans – including the one in the Oval Office – apparently missed the point of that cautionary tale.