CNS Kept Pushing Pro-Trump Poll Findings Before Election

We documented how CNSNews.com was interested in reporting on pre-election polls only if they showed President Trump winning. It kept that pattern up all the way to Election Day.

Indeed, the day after we published our original report, an Oct. 19 article by managing editor Michael W. Chapman gushed over another poll by the very Trump-leaning Trafalgar Group having “place[d] Trump slightly ahead of Joe Biden in Michigan and Florida, and gives Biden a slight lead in Pennsylvania.” As before, Chapman added that Trafalgar was “the only group to foresee a Donald Trump win in Michigan in 2016, and one of the few groups to forecast a Trump win in Pennsylvania in 2016.” Two days later, Chapman pushed more Trafalgar polling that “currently places Trump ahead of Joe Biden in Arizona and in Florida.”

On Oct. 28, Chapman was Trafalgar’s PR person again: “The Trafalagar [sic] Group polling firm, the only pollster to accurately foresee a Donald Trump win in Michigan in 2016, and one of the few groups to forecast a Trump win in Pennsylvania in 2016, currently places Trump ahead of Joe Biden North Carolina, Ohio, and Pennsylvania.” And on Election Day, Chapman wrote as if he was drawing a check from Trafalgar:

The Trafalgar Group polling firm, which was the only group to accurately predict a Donald Trump win in Michigan in 2016, and one of only a few firms that predicted a Trump 2016 win in Pennsylvania, gives Trump a slight lead today — Election Day — over Democrat Joe Biden in Florida, North Carolina, Arizona, Michigan, Wisconsin, Ohio, Georgia, and Pennsylvania. 

Trafalgar’s chief pollster Robert Cahaly, who accurately predicted Donald Trump’s 2016 win and his electoral vote total (306), believes that Trump will win the 2020 race with an electoral total in the “high 270s.”

Of course, as it turned out, Trump appears to have lost at least four of those states, suggesting there was something wrong with their model, which weights its polling to account for the “social desiriability bias” of purportedly shy Trump voters. Chapman hasn’t written a thing about Trafalgar since the election.

For all of this love toward a shiny new polling firm, CNS didn’t completely forget its first love, the Trump-skewing Rasumussen. On Oct. 22, Craig Bannister touted how “Rasmussen’s daily presidential tracking poll released Thursday shows that the percentage of U.S. likely voters approving of President Donald Trump’s job performance has risen above fifty percent for the first day this month,” adding that “Black voters registered 37% approval.”

On Oct. 27, Bannister served up another piece of Rasmussen’s biased polling:

A majority of voters believe it’s likely that, while vice president, Joe Biden knew about and may have benefitted from his son Hunter’s controversial foreign business dealings, a new Rasmussen poll shows.

In a national survey of U.S. likely voters, conducted October 22 and 25, 2020, Rasmussen asked the following question:

“How likely is it that Joe Biden was consulted about and perhaps profited from his son Hunter’s overseas business deals including at least one involving a company in mainland China?”

More than half (54%) of voters polled say it’s at least “somewhat likely” Biden knew and may have profited, with 44% calling it “very likely.”

Chapman even tried to frame a finding from normally reputable pollster Gallup as somehow pro-Trump:

As Election Day approaches on Tuesday, Nov. 3, it is interesting to note that, according to Gallup, at this time in their first terms Americans’ job approval ratings of President Barack Obama and President Donald Trump were very close — Obama, 50% and Trump, 46%.

[…]

For Trump, his job approval rating on day 1,376 in office is 46%. (Poll conducted Oct. 16-27, 2020)

For comparison, Obama, at 1,377 days in office, had a job approval of 50%. (Poll conducted Oct. 22-28, 2012.) 

Gallup also reported that Obama’s first-term average job approval rating was 48%.  (Similar data for Trump have not been posted yet by Gallup.)

If job approval ratings are an indicator of election performance, it’s clear that Gallup was much more correct than Rasmussen. Chapman and Bannister won’t admit that, of course.

UPDATE: Chapman got in one more pre-election poll to fit his template, using a Nov. 2 article to tout how “a new Des Moines Register/Mediacom Iowa Poll shows that Trump has pulled into the lead with a strong 48% support compared to Biden’s 41% support.”