Trump sees what he wants to see — and has people willing to provide it

Senate Majority Leader Mitch McConnell (R-Ky.) “should support [Trump] and join our objection on Jan 6th!” she crowed. “Look at how much President Trump’s endorsement tweet helped him!”

The accompanying photo showed a graph presenting a rough timeline of polling over the year in McConnell’s reelection bid. A poll in late May from RMG Research had McConnell tied with his Democratic challenger — and then Trump tweeted his support for McConnell on June 19. The next poll on the chart, from July 12, shows McConnell up by four points. By July 16, he’s up 22 points. A Trumpian miracle!

Except that it’s both cherry-picked and misleading. Trump’s tweet a month earlier somehow got McConnell to climb 18 points in the polls over a four-day period in mid-July? That’s the claim? And if we’re just using any old poll, why not include a Civiqs survey from June 15 that shows McConnell already up by 20 points?

Because it’s inconvenient and doesn’t tell the story that Trump wanted to tell. The story indicated below isn’t quite as compelling.

Nonetheless, Trump shared the graph more broadly a few hours later.

Now, an addition: An automated call Trump apparently made on McConnell’s behalf in late October that, the graph suggests, pushed McConnell up right at the end of the contest.

“Sadly, Mitch forgot” Trump’s help, the image declares. “He was the first one off the ship!”

McConnell didn’t “forget” that Trump help. McConnell, who has been in politics for a while, understands that he won without Trump’s help. He’s not some political novice who finds vast conspiracy theories about child trafficking credible; he’s a guy who was happy to have Trump help him raise money in June but who would have won either way.

That Trump would present these numbers as though they actually make the case being suggested is a bit amazing. He should know that this is a garbage argument, and perhaps he does. But this effort to reshape numbers to meet his needs not constrained to bad-mouthing a senator who has the gall to acknowledge the reality that Trump lost his reelection bid.

Earlier this month, Trump traveled to Georgia to campaign on behalf of two Republicans facing a runoff election early next month. He addressed his election loss during his comments.

“We thought that if we could get 68 million [votes], 67 million, that would be the end. All of our great, brilliant, geniuses said, uh, you’d win if you get 67 or 68, it’s over. We got 74 million plus, and they’re trying to convince us that we lost,” he said. “We didn’t lose. They found a lot of ballots, let’s be nice about it, they got rid of some, too. The 74 — let me tell you, the 74 could’ve been even higher.”

“As the great pollster John McLaughlin, who’s really a, a great pollster, one of the most highly respected,” Trump continued, “he said there’s no way this could’ve happened other than the obvious cheating or a rigged election. There’s no way it could’ve happened.”

This particular idea has come up before, that because Trump received so many votes, somehow the election must have been stolen. But it’s not complicated: Americans had very strong opinions on Trump, both for and (more commonly) against. They were motivated to vote more in 2020 than in 2016 or 2012, and turned out to do it.

Trump presents this argument as coming from McLaughlin, a pollster with whom he has been working since the 2016 campaign, when he was initially tasked with polling voters in New York. (Trump promised victory in his home state four years ago, as he did this year. He ended up losing it by a wide margin, as he did this year.) Two years earlier, McLaughlin came under fire for a poll for Republican House Majority Leader Eric I. Cantor’s reelection bid which was off by more than 30 points. In response, The Washington Post asked pollsters in Washington how wrong a pollster would have to be to have to find a new line of work. More inaccurate than that, apparently, though it likely helps to find a client who is more interested in what the numbers say than how accurate they are. A client like Trump.

Underestimation of Democratic turnout was one of the rationales McLaughlin offered in an interview with the National Journal for getting Cantor’s race so wrong. If you think turnout will be low and model your polling on that, but turnout is actually high, you’re going to be wrong. This is, in fact, probably one of the reasons that other polls underestimated Trump’s support in November: They assumed fewer Trump voters would come out, perhaps because those voters were less likely to respond to polls.

The difference between 2014 and 2020 is that McLaughlin acknowledged that the problem wasn’t that Cantor was cheated out of his win. It was that his estimates of what would happen were off the mark. This year, as Trump presents it, the numbers were wrong because something untoward must have happened.

Trump’s presidency has been defined by this sort of cherry-picking. It’s not new that Trump elevates sketchy, positive things to offset accurate, negative ones. The difference now is that reporting suggests that Trump is increasingly living in a surreal world where his delusions of the election being stolen are stoked and reinforced. Lots of people in Trump’s orbit have a reason to claim that he didn’t lose fair and square, including McLaughlin. But giving in to that impulse is increasingly reinforcing Trump’s distorted view of what happened. Storm clouds are forming in reality.

This is a moment when it’s important for people of good faith to be honest with the president and his supporters, no matter how hard it may be to provide that honest view. Instead, we get ridiculous graphs and false assertions and sketchy lawyers and hyperpartisan allies and a riotous, roaring base eating it all up.

This election year isn’t 2014 in one very important context. Then, the question was how inaccurate you could be and still have a career in Republican politics. Now, the question is how accurate you can be and still be able to appeal to Trump.