When Republican pollster Frank Luntz appeared on CBS News the day after the first GOP presidential primary debate, he declared that Donald Trump’s performance guaranteed the “destruction” of his campaign for the party’s nomination.
The basis for that conclusion was the opinion of a few dozen Ohio Republican voters in one of his fast-paced made-for-TV focus groups. While many participants had said they came in supporting Trump before the debate, only three or so remained supporters after it was over.
“Those focus groups are accurate,” Luntz said.
But the days that followed seemed to prove Luntz wrong. Trump, who is leading the Republican field, has seen his national poll numbers either go up or remain mostly unchanged since the debate.
On the day of the debate, an average of the national polls used to determine who would make it on stage showed Trump with 23 percent. This week, 12 days after the debate, a national CNN poll was released showing Trump at 24 percent.
Some say Luntz’s method of predicting the future has a clear flaw.
“The idea of taking a poll in a group of conservatives to get a representation [of the electorate] is completely bogus,” Dr. Robert Shapiro, a Columbia University political science professor, said in an interview Tuesday with the Washington Examiner media desk. “You don’t know whether [Luntz] has a random sample of conservatives or likely voters. These are no more than straw polls of a random group of people.”
Shapiro, who specializes in public opinion research, said Luntz’s work makes for compelling and entertaining television, but that focus groups in general are meant to be a step toward developing more reliable surveys.
“Don’t interpret a focus group as representative of what’s going on,” Shapiro said. “But take it as examples and illustrations of how some individuals are reacting. So, it’s a range of reactions but you don’t know how representative it is, what proportion of voters are thinking the same way.”
Luntz has made a name for himself within Republican political circles and the news media for his work on projects like the GOP’s 1994 Contract with America, which helped in the party’s takeover of both houses of Congress.
But his work has also stirred controversy. The National Council on Public Polls censured Luntz in 2000 “for allegedly mischaracterizing on MSNBC the results of focus groups he conducted during the Republican Convention,” according to the Washington Post.
Since the Fox debate two weeks ago, a feud has brewed between Trump and Luntz. Trump accused the pollster of deliberately trying to harm his campaign and claimed it was because Luntz had once sought business from the real estate tycoon, only to be turned away.
And this week, Politico reported that Luntz, ahead of the first debate, told a group of Republican donors in Southern California that Trump was “turning what we believe into a joke.”
Luntz has maintained that he merely offers a professional interpretation of his focus groups. He did not return a request for comment for this story.
Dr. Lonna Rae Atkeson, director of the Center for Voting, Elections and Democracy at the University of New Mexico, said Luntz’s conclusion of the first debate focus group is not something that should have been applied to the American voting electorate as a whole.
“We can only make it about those people in the focus group,” she said. “If we started out at 10 supporters and you go down to three, that’s all you can say. You can’t say that same decline would be seen in the broader electorate.”
“The unique thing about a focus group is they all sat there and watched,” Atkeson said. “Whereas 90 percent of the public didn’t watch the debate or think about it at all. That’s a huge difference.”
Atkeson stressed, however, that she doesn’t “think there’s a chance in hell” that Trump will be the Republican nominee.
