The Pollster Who Thinks Trump Will Win

WALL STREET JOURNAL – OPINION:

Joe Biden leads Donald Trump by an average of 7 or 8 points in national surveys, more narrowly in battleground states. Everybody remembers the shock of 2016, but can the polls be wrong again?

Ask the question in a different way: Are poll respondents telling the truth? Robert Cahaly, head of the Trafalgar Group, thinks a lot of people aren’t. Trafalgar polls accurately foresaw the outcome in 2016, calling Florida, Pennsylvania and Michigan for Mr. Trump. In 2020 the Atlanta-based consulting firm has generally shown Mr. Trump to be in a stronger position than the conventional wisdom would suggest.

Atlanta

Joe Biden leads Donald Trump by an average of 7 or 8 points in national surveys, more narrowly in battleground states. Everybody remembers the shock of 2016, but can the polls be wrong again?

Ask the question in a different way: Are poll respondents telling the truth? Robert Cahaly, head of the Trafalgar Group, thinks a lot of people aren’t. Trafalgar polls accurately foresaw the outcome in 2016, calling Florida, Pennsylvania and Michigan for Mr. Trump. In 2020 the Atlanta-based consulting firm has generally shown Mr. Trump to be in a stronger position than the conventional wisdom would suggest.

In an interview over a catfish supper at the OK Cafe diner, Mr. Cahaly won’t reveal much about his methods, but he says his polls mitigate what social scientists call “social desirability bias.” The mainstream media and other authority figures have openly and aggressively contended that Mr. Trump is a white supremacist, a would-be dictator, a cretinous buffoon and an inveterate liar. In such an environment, poll respondents who sympathize with the president, or who believe his administration has on balance done more good than harm, may be forgiven for not saying so to a stranger over the phone.

Do people lie to pollsters? “Yes,” Mr. Cahaly says, “but they’re not necessarily doing anything wrong. If a grandmother says, ‘This is my grandson, isn’t he a handsome boy?’ and you can see he’s anything but handsome—he’s sickly and weird-looking—you don’t say, ‘No, he’s sickly and weird-looking.’ You just say, ‘He sure is.’ ”

Social desirability bias is more pronounced among some demographics, Mr. Cahaly thinks, and he claims only that Trafalgar polls minimize it. “You can’t get rid of it,” he admits. To oversimplify his approach: If a poll respondent tells you he’s voting for Candidate A but that same person answers every other question in a way that suggests he’s voting for Candidate B, the pollster may wish to account for that oddity in the overall tally. And in a year when Candidate A is said by the cultural elite across the globe to be the Source of All Bad Things, the need to reckon with disingenuous answers is perhaps more important than usual.

“Take any retirement community in America,” he says. “Poll how many people watch ‘The Jerry Springer Show,’ and you’re gonna get one number. Contact the cable company and find out how many people at the same retirement community actually watch ‘The Jerry Springer Show,’ and you’re gonna get another number.”

He also thinks most pollsters rely too heavily on cold phone calls. “A lot of people aren’t gonna tell a stranger on the phone who they’re going to vote for, especially if they’re afraid that information might wind up on a website or a Facebook page for everybody to see.” These days many screen their calls and don’t even pick up those from unknown numbers.

Traditional polls fail in other ways, too, in Mr. Cahaly’s view. Some ask far too many questions. “Who’s going to answer 45 questions on a Tuesday night?” he asks. “People who know a lot about politics and like to express their opinions, that’s who. Most people aren’t like that.”

Read more at The Wall Street Journal

Buy on Amazon!