Tagged: polls

Poll Dancing-My Next Move

In my semi-retirement I’m enjoying my part-time freelance gigs that keep my brain from turning to grits and thanks to this election cycle, I think I’ve decided on my next endeavor. I’m going to be a pollster.

What I’ve learned from watch actual, professional pollsters is it seems like you can make some decent money while never actually being accurate. As a journalist, this goes against all my ethics.  Then again, news organizations are among the biggest spenders on polls in order to manufacturer news stories that may or may not be true, but every time the poll is referenced in a story the name or names of the sponsoring news organizations are mentioned, providing some effective promotion.

We’ve seen from both the 2016 and the 2020 presidential election cycles that pollsters can swing and miss by a mile the eventual results. Guess all that victory party planning by Hillary Clinton’s campaign based on polling that she’d wipe Trump’s butt in the election was a big oops. Maybe they should have charged the pollsters the costs of streamers, confetti and caviar. 

They blew it again this year, prediction a big blue wave where the Democrats took back the Senate, widened their majority in the House and Joe Biden would sashay into 1600 Pennsylvania Avenue. The Dems won’t regain the majority in the Senate, their majority in the House narrowed and days after votes were cast, Joe Biden still can’t tell the post office to begin forwarding his mail as of Jan. 20, 2021, even though it seems inevitable. It wasn’t supposed to be this close…according to the polls. 

The irony is, despite their total whiff, pollsters will still make big bucks for what really amounts to an attempt at legal jury tampering. The supposition by political organizations that buy polls is if voters see their candidate as a winner in the pre-election polls, they’ll be likely to support him or her with real votes. Turns out voters may enjoy reading news stories about the latest polls but when they cast their ballots they think for themselves. 

If I ran a polling agency I’d be more honest about it. I’d run the poll and report the results with a margin for error of plus or minus 100 points. The client would get the numbers they paid for and if they turned out completely wrong I could always say, well…they were within the margin for error. 

I would give my new polling agency the appropriate title, “I’ve Got Your Numbers” or IGYN. Can’t wait to pick up the New York Times and read the lead, “In an IGYN-NY Times poll, 78% of those on the Acela Express Amtrak agreed that railroads take people places. 17% said they wandered on the train looking for packs of Saltines and the rest had no opinion and asked to return to their naps. ‘This poll is conclusive evidence people depend on Amtrak for something,’ said Amtrak spokeswoman Dee Rail.” 

See? I think this could work out. In fact I polled my family on the idea. 94% nodded their heads while muttering “yeah, sure,” 2% smirked and 4% asked me to bring them beers. None responded negatively. Margin for error, 100%. I’m goin’ with it. 

When 2+2=BS

statisticsOne of the best books I ever read was a slim little paperback thing published in 1954 titled “How to Lie With Statistics,” by Darrell Huff. It was required reading in my “Ethics in Journalism” course at the University of Arizona when I attended grad school there in 1978.

I bring this book to your attention because it should also be required reading for anyone who takes any stock in the myriad of public opinion polls tossed in our faces during this dreadful political season.

Huff warns us, “The secret language of statistics, so appealing in a fact-minded culture, is employed to sensationalize, inflate, confuse, and oversimplify,”

Indeed. If you don’t already know this, polls are not the same as elections. News organizations buy polls to give them something to report, regardless of their accuracy. Polls are also useful for earning publicity for the purchasing news organization because every time the poll is cited in another news organization’s story, the purchasing network, station or publication’s name is mentioned…like the CNN/Wall Street Journal Poll, or the Mad Magazine/Hustler Poll. I made that one up. Doesn’t matter if the polls reflect reality. They can always tout the “margin for error,” to explain away the fact the poll’s results could be full of crap.

Political candidates buy polls to convince voters they’re winning. Corporations purchase polls to prove the world can’t live without their products or services.

It’s all in the wording of the questions. Sure, there can be the simple choice of candidate listed. But then the questions become even more leading. Say, “If Donald Trump wasn’t a misogynistic, lying creep, how much more likely would you be to vote for him?” Or. “How much does the fact that Hillary Clinton may very well be indicted affect your decision whether or not to vote for her?”

A company touting, say, its new miracle product  might ask consumers identified as ex-felons, “Agree or disagree that your personal well-being would be enhanced with a product that could completely dissolve the serial number from a weapon used in a crime.”

Huff covers that possibility with the declaration “there is terror in numbers.”

You may recall the polls appeared to predict Mitt Romney unseating Barack Obama from the White House four years ago, only to be handily disproven when actual votes were counted. The polls showed that because those cited were “internal polls” taken for Romney, and paid for by Romney’s organization. Gotta keep the customer satisfied, until poor Mitt let his polls blind him into deciding not to write a concession speech “just in case.” Unfortunately for him, the real poll, known as the election, didn’t square with his self-serving survey and Mitt had to concede to the fact he was unprepared to cogently concede.

This is why I completely disregard any sort of poll plastered on the screen or on the page, no matter the subject. I learned long ago, courtesy Darrell Huff’s 144 pages of truth, the margin for error, is the poll itself.