This tears it for me, I think. I’m not paying attention to any polls from anyone or anywhere, ever.
I don’t know why anyone ever would, except, well, you know. Most people are stupid and gullible with flat learning curves, and our political parties and the news media depend on it.
An alleged polling groups called Median Strategies released a poll before the Wisconsin gubernatorial race that showed Democratic Socialist whack-job Francesca Hong with a lead of more than 20 percentage points. Hong lost: It was close, but metaphorical miles away from winning by 20 points.
But wait! There’s more!
Then Median Strategies issued a poll this month showing Los Angeles Mayor Karen Bass leading City Councilwoman Nithya Raman by nearly 12 percentage points ahead of the November mayoral runoff. Bass was apparently ecstatic, declaring on social media, “Doing the work, showing up, and gaining momentum. Let’s do this, LA!”
On Monday, the company acknowledged on its website that it had produced the fake polls as a”short-term social experiment examining how purported polling information could enter and spread through the political information ecosystem without independent verification.”
Oh! That’s all right then! No, it’s not, and I don’t care what the alleged justification for releasing fake polls was. The incident proves one thing, and one thing only: we can’t trust polling organizations, we can’t trust polls, we can’t trust the news stories that cite polls (Trump’s support among Hispanics drops 50!” was a headline this week…I bet they polled illegal immigrants), and we can’t trust politicians who use them either.
We should have already known this after the miserable performance of polling during the last few election cycles, but the Median Strategies “experiment” should put the question to bed permanently. All polling is a tool for manipulating elections, discouraging some voters and energizing others. Ignore them all. If a poll is correct, it’s luck.
We owe Median Strategies a debt of gratitude for clearing this question up for all time. But I think their real experiment was designed to find out how many people will still take polls seriously even after this farce. My guess: lots. Because…you know.
[Update: A 21-year-old recent college graduate named Rahil Prakash revealed that he was the one behind the fake polls and “Median Strategies.” In a report published this week he admitted that Median Strategies was his invention, using AI. “I wanted to see if fake polls could really penetrate the ecosystem that easily,” Prakash said. “And as it turned out, it could.”
_______________
Ethics Alarms 
Well, the polls have been consistently and significantly wrong for at least the past 10 years. We were told on several occasions that it was shy Trump voters or some such, but obviously that doesn’t hold water.
The Michigan Senate race was supposed to be a landslide but the DSA candidate just barely pulled it out. And then the Wisconsin race, which again was supposed to be a landslide for the DSA candidate. I’ve also read speculation about why Hong conceded so quickly on what was a razor thin margin.
But they won’t suffer any real consequences — they never have all these other times.
I like Real Clear Politics as a website, but their model will break down if many of the polls are essentially rigged or biased.
As an Iowan, I don’t have to look much further than the Anne Seltzer that was published here the weekend before the last Presidential election. It “revealed” that VP Harris was leading in Iowa by three points over President Trump. It was picked up and trumpeted by every single outlet from ABC to Brian Tyler Cohen. Seltzer was “the Gold Standard” of polling, we were told repeatedly. The Gold Standard. The. Gold. Standard.
About one hundred hours later, the voting was done. President Trump was re-elected and he won in Iowa by fifteen points, the widest spread of victory in any Iowa vote since…Nixon in ’72 maybe?
Was Seltzer putting out a fake poll just to see who would grab the narrative? I highly doubt it, so it’s a slightly different scenario. But it again highlights how pollsters’ biases, along with the nebulous way questions are often asked and who actually gets polled, can skew results away from the truth.
Interestingly, we are in the midst of a gubernatorial election, as a highly-popular Republican governor who won by fourteen points four years ago is retiring. And now we are being told repeatedly by pollsters that the race between the two candidates is neck-and-neck.
Is it? Maybe.
Do I believe the pollsters’ conclusions? Absolutely not!!
I’m annoyed with myself. The Ann Seltzer episode was supposed to be part of the post, and I left it out. Thanks for remedying my negligence.
I was once polled and asked which poll was the most important to me. My answer was the one that comes out a the end of election day.
When I Initially read the article about Median Strategies’ fake polls a couple of days ago, it strengthened my resolve to never put any stock in any political poll regardless of the source.
It’s likely that I was unknowingly biased by a Marquette poll leading up to the Democratic Party primary for Governor. The Marquette Poll has been one of the better political polls for years, not so much anymore. I genuinely thought Francesca Hong was going to have a landslide victory in the Democratic Party primary for Governor, everything seemed to be leading to that outcome; however, Hong lost with 39.33% of the total vote which was 3,796 votes (.5%) less than the winner of the primary. I choose not to vote in partisan primaries, even if they are open primaries like in Wisconsin, because I’m not affiliated with any political party.
Political polls planned by political activists with a goal in mind are used as a motivation tool instead of just putting together a reasonably accurate partisan free poll and simply sharing the findings.
The baldfaced liars at Median Strategies claiming (after the fact) that the fake polls they released were some kind of “short-term social experiment examining how purported polling information could enter and spread through the political information ecosystem without independent verification” is another baldfaced lie. You don’t perform a secret social experiment on an entire population in a single state during an election cycle without knowing full well that what the fake polls are literally doing is intentional manipulating the public perceptions of political candidates. These liars at Median Strategies lied to everyone and now they claim that they’re not lying with their social experiment narrative? I don’t believe a word of it.
I think this one could be an interesting addition to the Rationalization list, “it was only a social experiment”, kinda like “it was only a joke”.
“I think this one could be an interesting addition to the Rationalization list, ‘it was only a social experiment‘, kinda like ‘it was only a joke.”
More like it was only Schrödinger’s Douchebag
PWS
Jack wrote, “Hong lost, and it wasn’t all that close either.”
I don’t know about that; I’d consider .5% difference between the top two candidates to be a pretty darn close race, heck it’s even close enough for a state funded recount in Wisconsin.
Yeah, that was stupid. I’ll fix it.
INTERESTING OBSERVATION
It was reported that “Median Strategies released a poll on the Wisconsin gubernatorial race that showed Democratic socialist Francesca Hong with a lead of more than 20 percentage points”. Median Strategies admitted to everyone that they released fake polls.
The Marquette Poll (renowned for having the most accurate political polls in the USA) released the following poll on July 29th showing that Hong had a 31 point lead…
Logically speaking; if Median Strategies admitted that the polls they released, that were far from reality, were fake and Marquette presented a poll that was even farther removed from reality than the fake Median Strategies polls, inquiring minds want to know what that information logically says about the Marquette Polls?
CORRECTION
That “Hong had a 31 point lead…” should read “Hong had a 31 point lead over Crowley…”.