#196 Blogpost – Wednesday 25 October 2023
http://dennyhatch.blogspot.com/2023/10/196-political-polls.html
Posted by Denny Hatch
Direct Marketers vs. Political Pollsters:
Why We Are a Lot Savvier Than They Are.
Nate Silver, fivethirtyeight.com,
America’s Most Trusted Pollster.
Chart Right Published: 7:05 P.M. Election Eve, 7 November 20016.
Nate Silver made his bones as a statistician and baseball analyst who allegedly once made $400,000 over a three-year period playing online poker. Founder and editor-in-chief of the polling organization fivethirtyeight.com, Silver was named one of the world’s 100 Most Influential people by Time after he successfully predicted the outcomes in forty-nine of the fifty states in the 2008 election. Subsequently his forecasting system predicted the outcomes of the 2012 and 2020 presidential elections with a high degree of accuracy.
Nate Silver Blew It Big
Time in 2016 When He Predicted
Hillary Clinton would Cream Donald Trump 71% to 28%.
Being a political junkie for over 50 years, like the majority of Americans I was stunned everybody got it all so very wrong.
Same Thing 76 Years Ago in the 1948 Election.
—Photo: W. Eugene Smith. LIFE Magazine.
Quick Backstory of This Extraordinary Photo
— One of
The Most Famous Journalistic Gaffes in American History.
In the summer of the 1948 election President Harry Truman had a sad-sack 36% approval rating. Throughout the summer the Gallup and Roper polls declared Truman’s opponent, Thomas E. Dewey, was an absolute shoo-in guaranteed to win the election. In the last two months of the campaign Truman virtually disappeared from public view. He was on a whistle-stop trip on a private train where he could be seen and talk in person directly to the American people from the back of his train. Some real early mornings he'd be in his pajamas, robe and slippers and talk to a handful of voters. At the huge Iowa State Fair in September he roared over the public address system to mesmerize an audience of 72,000. In those final two months of the campaign Truman crisscrossed the country over 31,000 miles and made 352 speeches. All media coverage was local. The pollsters never caught the shifts in public opinion. Robert R. McCormick, owner of the Chicago Tribune insisted on being first to announce Truman’s defeat starting with the earliest edition of paper.
What triggered this blog post are some current projections about the 2024 elections coming up twelve months hence.
The Total Incompetence of College Student Pollsters
May Show Up Yet Again on the Same Error-Strewn Path.
This Marquette Law Students’ poll of “690 Likely Voters” is a teensy-weensy
microscopic sample out of a humongous universe of 168.42 million Registered American Voters.
Preposterous!
What’s more, The above illustration is a large slide with full color photographs, big headlines and a small muddy caption at bottom. It was shown for a few seconds full sized on network and local TV news programs along with a bunch of other pollsters’ projections.
These sample
numbers are disguised in the small line at bottom printed in blurry medium blue
mouse type printed over a light blue background. It is guaranteed to be
overlooked by virtually all TV viewers. I missed it clean and noticed it only
after I photographed the slide for this blog post. The numbers:
About Amorphous Political Polls vs. the
Hard Data of Direct Marketers’ Tests
Political pollsters are mining the mushy world of “Who would you vote for if the election were held today?” The answer is in the slippery world of emotions, of “if” and “how do you feel right now?”
When launching a new product, service or financial opportunity direct marketers rent lists of buyers or subscribers or members of analogous products and services. Or we advertise in media our potential buyers read and view.
Below is the order card for INTERNATIONAL LIVING, a publication that did not exist. After three failures and $70,000 in debt, young copywriter Bill Bonner sent out a “dry test” mailing with a powerful 8-page letter describing how you can retire overseas in luxury for a fraction of what it would cost in the U.S. The dry test mailing pulled 300% of breakeven and Bonner was in business. Today he is owner of The Agora, a billion-dollar publishing empire out of Baltimore, Maryland.
This offer was the real deal — NOT a weak and wifty, “if this newsletter existed do you think you might buy it?” This was offering a solid commitment from Bonner to the recipient urging a change in behavior that will be very beneficial. (And not change his mind tomorrow.)
• We start with a series of small tests. Then follow up with medium size confirming tests. We track results every step of the way.
An Example of Some Basic Direct Marketing Arithmetic.
I got into direct
mail marketing in the 1970s, the formula was primitive and simple. The minimum
test was 5,000 names. Way back then the average response to a direct mail
campaign was 2%. A two percent response from 5,000 names is 100 orders — the
absolute minimum cell size of paying customers just large enough to be statistically
dependable.
Direct marketing brings immediate results that you can take to the bank — literally. The success or failure is measured precisely by test results and money, promised or sent. If a test bombed, you sweetened the offer or tested different pricing until you got it right.
In 1984 Peggy and I started the monthly WHO’S MAILING WHAT! newsletter and archive service. It was a publication that existed in our heads only. We did a “dry test” offering this mythical publication. We rented 5,000 names from Pete Hoke’s Direct Marketing magazine and 5,000 from Adrian Courtenay’s DM NEWS for a total of 10,000 names. Our offer:
Take the first issue FREE and the next twelve issues for $99. Cancel any time and receive a full refund no questions asked. No risk. We got 452 responses and became publishers.
In Political Polling, No Interim Results.
Unlike direct marketers, political pollsters have to wait till the very end of the campaign when the polls close and the votes are all in and counted before they know for sure whether they are champs or chumps.
These are the Amateur
Collegiate Political Pollsters Who Are
Poaching In the Worlds of Fox News, CNN and Five
Thirtyeight
Emerson College, Boston, MA
Iowa State Civiq (News Service), Ames, IA
Marquette College Law School, Milwaukee, WI
Quinnipiac University, Hamden, Ct
Winthrop University, Rock Hill, SC
My personal opinion these minor colleges and universities are in the political polling business for publicity and public relations. Any coverage in state and national media (e.g., the Marquette Law School poll shown in this post) may well be noticed by wealthy alumni who take pride seeing their beloved alma mater in the forefront of the news. Maybe they'll increase their giving. Plus it gets minor colleges' names out to prospective students interested in political science.
Here's another cheapsy-weepsy political poll I photographed off the TV.
NOTE: “Among 868 Registered Voters.”
Takeaways to Consider
• Any time I see a national or state political poll, I now look for the size of the sample it was based on.
###
Word count: 1117
