Political polls; bah humbug!

Polls do not predict the future; they only record public opinion at a slice of time.

That truism can’t be repeated often enough. Sadly, it’s not. People still look to polls to deliver the low-down on which candidate will win.

That’s reinforced by the never-ending, over-reported, ceaseless and relentless candidate preference polls leading up to the Republican presidential primary. Check out realclearpolitics.com to see how many rain down on us. Unfortunately, the site reinforces the predictive idea by publishing a “scorecard” to compare how close they came to the actual vote. count.

I know a thing or two about polling because as a reporter at the Chicago Sun-Times I worked with Gallup to publish a poll about the Chicago mayoral race in 1983.

The Democratic primary was a heated three-way race by three candidates: incumbent mayor Jane Byrne (no relation); Richard M. Daley, son of the legendary Mayor Richard J. Daley, and Harold Washington, a long-time black local and congressional lawmaker.

Gallup used the latest, scientific methods to conduct a number of polls as the primary election neared. My job was to turn the social science gobbledygook into readable stories.

The polls consistently showed that Washington trailed the two white candidates—Byrne and Daley. Thanks to the polling, the “smart” betting was on those two; Washington was given a slim chance.

Washington won.

Byrne trailed narrowly, and Daley somewhat more in third.

Came the questions from all sides: Why were the polls so far off? What was wrong with the polls? Were black voters reluctant to tell a stranger on the phone that they favored the black candidate? Were black voters harder to find? Did the pollsters miss the 50,000 newly registered black voters? Was there a last-minute surge of black voters?

Or was nothing wrong with the polls? I don’t know. But a predictive tool, it was not.

So what do we make of the daily revelations of polls that show former President Donald Trump a runaway winner? Could they be so far “wrong” if Ron DeSantis or Niki Haley win the upcoming Iowa caucuses in an “upset?” Would it really be an upset if the polls were so flawed in accurately portraying the race?

I don’t know. But I’ve learned not to trust them, and certainly not use them as a basis for whom I would vote.

A footnote: The 1983 elections was the first time that the Sun-Times used a professional polling operation. Prior to that, the paper unleashed a flood of interviewers at grocery stores, train stations and elsewhere to gather the data. It lacked the methodological rigor about such things as the size and quality of the sample. And yet…they almost always more “accurately” measured the outcome.

Dennis Byrne is a retired Chicago journalist, author and writer. Email: dennis@dennisbyrne.net