Trump’s most loyal media ally promised a pro-Trump poll. It didn’t deliver — and then pulled its story.


June 11, 2020 at 6:42 p.m. EDT

CNN’s release of a poll this week showing President Trump trailing former vice president Joe Biden by 14 points nationally clearly rattled the president and his reelection campaign. In short order, Trump tweeted out a memo making various allegations about how and why CNN conducted the poll, each assertion ludicrous and easily debunked. On Wednesday, the campaign escalated its efforts to portray CNN’s poll as unfair, demanding that CNN retract the poll and issue an apology.

CNN’s attorneys, with complete and understandable justification, declined to do so.

The network is one of Trump’s most frequent targets for criticism. He has repeatedly bashed CNN’s reporting and on-air talent, disparaging the network as hopelessly biased against him. Trump’s taste in television coverage runs more along the lines of Fox News’s Sean Hannity, a fervent supporter of the president, and, in recent months, One America News.

The Fix’s Eugene Scott breaks down how presumptive Democratic presidential nominee Joe Biden has reacted to the death of George Floyd. (JM Rieger, Blair Guild/The Washington Post)

OAN is a small network, clearly seeking to make its name in part by promising endlessly obsequious coverage of the president. The network has often explicitly asked Trump to watch it instead of Fox, as chief executive Robert Herring did Wednesday.

Trump does watch. It was an OAN report about the protester in Buffalo who was injured by police that prompted Trump to speculate wildly that the elderly man who suffered a head injury had faked his fall in service to a murky network of left-wing anarchists.

Herring defended that report Thursday, writing on Twitter that his network had “put out our facts” on the incident, while the mainstream media hadn’t done so. There is no evidence for the claims made in the report, though OAN reporter Kristian Rouz did certainly offer up some allegations.

That tweet was followed up with another attempt to goad the mainstream media.

Robert Herring@RobHerring

Here’s a message for the mainstream media: We have put out our facts on the Buffalo protestor that you claim are wrong. Where is your evidence? @OANN

We did see CNN’s @donlemon interview a lady that knew him. Do you want to stand behind her?

Robert Herring@RobHerring

Just to add some fuel to the fire, did you see our poll on Florida? @OANN

200 people are talking about this

Herring had on Wednesday pledged an upcoming poll, as the Trump campaign’s feud with CNN over its national poll was heating up. He promised that OAN would be “releasing a poll concerning the 2020 presidential race” that “looks as though it will be in favor of” the president.

Early Thursday afternoon, the poll came out. Conducted by Gravis Marketing, a pollster that earns a C in FiveThirtyEight’s ranking of pollsters, it was focused solely on Florida.

It had Trump and Biden tied in the must-win state for Trump, a state Trump won narrowly four years ago.

OAN produced a video segment on the poll, again featuring Rouz, which appeared on its website. Soon after it was published, though, the report was pulled. A tweet from Herring that apparently announced the results was also deleted.

Image without a caption

Both the page and the video report were captured, however. The image above comes from Google’s cache; the report can be seen on YouTube thanks to writer Arieh Kovler.

The poll itself and Rouz’s report on the poll immediately raise red flags in a variety of ways.

For example, it shows a 50-50 tie between Biden and Trump. Polls don’t normally result in perfect 50-50 ties, unless respondents are forced to choose between the two candidates. That appears to be what happened in this case, as Kovler notes; at another point in Rouz’s report, he reveals that Trump leads Biden 53 percent to 47 percent among undecided voters.

What does that mean? It means that Biden necessarily leads among voters who have made up their minds. Imagine that half of voters say they’re undecided. To get a 50-50 result overall, Biden would need to lead Trump 53 to 47 among decided voters if Trump led by the same margin among undecideds.

Rouz didn’t mention that detail.

Lots of bad news for Trump is glossed over in the same way. Trump is presented as having an edge over Biden on handling the economy. But this point, one central to Trump’s reelection bid, downplays that Trump is under 50 percent among respondents and leads Biden by only four points, basically a tie.

Many of the other data points that Rouz highlights are based on leading questions (“Is activating the national guard an effective way to prevent further rioting?”) or are presented by Rouz in ludicrously loaded language.

At one point, he shows the results of a question about where blame lies for violence that emerged following protests throughout the country.

“The OAN/Gravis poll reveals 43 percent of Floridians blame far-left protesters for the latest violence and looting,” he said, “while only 36 percent blame the police.”

First, Rouz’s presentation of the question doesn’t match what’s shown on the screen, nor did the response options. (The question asked only about “protesters,” not “far-left protesters.”) Second, and more important, a seven-point difference is hardly anything definitive, given that the difference is probably not statistically significant, depending on the margin of error of the poll.

Or it’s like when Rouz claimed that “at least 50 percent of Floridians would vote to reelect their president if the election were held this week” — an impressive use of “at least” and an equally impressive effort to ignore that the same thing could be said about Biden. And Rouz’s framing works only if you assume those undecided voters actually vote for Trump at the margin they indicated they would.

The OAN report tries to suggest that this result is somehow a repudiation of CNN’s poll. It isn’t, for a variety of reasons. The most obvious, of course, is that the CNN poll was national and the OAN poll conducted only in Florida. Florida polling from other outlets shows Biden with a narrow lead in the state of about three points on average.

Meaning that if, say, a third of respondents in OAN’s poll said they were undecided, Biden could be leading Trump 52 to 48 among voters who have made up their minds. That puts the OAN/Gravis poll very much in line with other polling in the state.

That Rouz is so sloppy and presents the findings so dishonestly should not be a surprise. He has another report that was published by OAN on Thursday. It is a buffet of allegations that would resonate with Trump: The “deep state” is working with Democrats to produce polls making Republican voters demoralized. It’s lifting up one of the dumber points of contention Trump’s team raised in its criticism of the CNN poll but somehow manages to do so in an even more ridiculous way.

This entire polling effort by OAN is a remarkable, if unintentional, window into how the network works. Its chief executive trumpets a poll that he promises is likely to show Trump doing well. His on-air reporter gins up a thoroughly misleading presentation of results that are far from great for the president. For some reason — but probably exactly that reason — the story and the report get deep-sixed.

After all, Trump’s not going to want to watch that. We don’t want him switching over to Fox, now, do we?

Headshot of Philip Bump

Philip Bump is a correspondent for The Washington Post based in New York. Before joining The Post in 2014, he led politics coverage for the Atlantic Wire.Follow

osdlfm(-1,”,’BisbK_Q3kXtPOEIzVogaBqrTAAQDpte67-gIAABABOAHIAQngAgDgBAGgBiGIBwHSCAQIABAC’,”,4176054298,true,’la\x3d0\x26xdi\x3d0\x26′,3,’CAMSeQClSFh3xx0nyeUcV2rXLKcO7mCRrwb_q8JTqr_cqY-nfBLCluGtWEfzQNkZn0CO43sq35sJec-XuroFxVBsDuFR_tm6EZttNLzccm2ahx5aBOMgm1oGvDpAr4Wzs5C3Ga2PHcOCi3XtZz9aKcJQWmvtNpD_MlAxuG4′,’https://pagead2.googlesyndication.com/pcs/activeview?xai\x3dAKAOjst6Hq8n2rAk7QzeV08FyFY7kWs2hbA-T4FaoRTgBadFphN5dBut3ikpZwYqEZ-Oc1QXVdUmbNKUym3vX6sm6krDPW3q1NvWhA\x26sig\x3dCg0ArKJSzDNS_fvEkLykEAE’,”);googqscp.init([[[[null,500,99,2,9,null,null,null,1]]]]);""

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: