Rasmussen Defends Design Amid Allegations of Pro-Trump Bias
Fran Coombs says widely followed pollster proved more accurate than most others during 2016 election, remains ahead of pack
President Donald Trump’s approval rating is at 51 percent — significantly higher than other polls indicate — according to the latest Rasmussen Tracking Poll, and the chief executive’s critics are not happy about it.
The crowd at “Morning Joe” on MSNBC ridiculed the poll Monday as an outlier or worse. Trump’s average in the RealClearPolitics job approval roundup is 42.4 percent. Trump scores no higher than 44 percent anywhere else.
FiveThirtyEight, which analyzed the historical accuracy of dozens of polls prior to the 2016 election, gave Rasmussen a C-plus.
But Fran Coombs, managing editor of the poll, said on “The Laura Ingraham Show” that the criticism his company receives is unwarranted.
“It’s nonsense,” he said. “All these pollsters that these guys like to rely on, as your listeners well know, had disastrous results in 2016. And to the best of our knowledge, I don’t think they’ve really corrected anything at all. So as far as I know, they’re probably still having disastrous results.”
[lz_table title=”Rasmussen Outlier” source=”RealClearPolitics”]Rasmussen Reports consistently records higher approval for the president than other polls
Coombs said he would put the Rasmussen poll up against any other. It was significantly more accurate than most polling companies during the last presidential election.
Rasmussen’s methodology is superior, Coombs explained, for three reasons. First, he said, the company screens out registered voters who are not likely to vote. Many pollsters this far before an election use registered voters even though including unlikely voters skews the results.
Second, Coombs said, Rasmussen is a daily tracking poll. It uses a 1,500-person sample, continually dropping out 500 people and adding 500 new people.
For that reason, Coombs argued, tracking polls were better able to pick up opinion swings and more accurate during the 2016 election than polls that simply took a snapshot in time.
And finally, Coombs said, Rasmussen uses “robo-calling,” an automated questionnaire that lets respondents avoid talking to a real person.
“We honestly believe, particularly in days like these where everybody is so partisan, that people are more likely to tell the truth to an anonymous machine, if you will,” he said.
It is not just television pundits who mock Rasmussen’s divergent results. Coombs recalled that Lanny Davis, a longtime friend of Bill and Hillary Clinton, sent him a nasty email three days before the 2016 election asking if Rasmussen was going to apologize for getting the race so wrong and skewing the RealClearPolitics average.
“We have not heard a word from Lanny ever since Trump was elected and we were proven right,” he said.
Coombs said Rasmussen conducts polls more or less the same way it did when Barack Obama was president, except for adding more cellphone users.
“In many instances, we had Obama, his favorables, higher than some of the other pollsters did,” he said. “And, of course, they didn’t complain about that.”