Pollster Behind Karen Bass’ 12-Point Lead Admits Poll Was Fake, Shuts Down Operations
“Median Strategies has concluded its polling project and will not publish additional polls. All previously published polling releases have been withdrawn and should not be cited or treated as genuine polling data.”
Among the biggest takeaways from the 2026 primary cycle has been the stunning inaccuracies of polling in many key races. One of the most astounding misses came in Wisconsin’s Democratic gubernatorial primary, where polls showed democratic socialist Assembly member Francesca Hong with a commanding 20-point lead over Milwaukee County Executive David Crowley. Crowley ultimately defied those predictions and narrowly won the race.
Polling in Michigan’s Democratic Senate primary was similarly off the mark. Surveys showed far-left candidate Abdul El-Sayed leading his opponent, Rep. Haley Stevens (D-MI), by double digits. Although El-Sayed ultimately prevailed, his expected 20-point advantage evaporated by Election Day, and he won the race by just 1 point.
Last week, polling group Median Strategies released a survey showing Los Angeles’ incumbent Mayor Karen Bass (D) leading her challenger, socialist city councilmember Nithya Raman, by nearly 12 points. Bass touted the poll on X with the caption: “Doing the work, showing up, and gaining momentum. Let’s do this, LA!”
Holy…
Karen Bass celebrated this poll showing her up by almost 12% in the LA Mayoral race. The poll received a bunch of attention.
Now the company behind the poll has admitted it was made up. pic.twitter.com/fon6xwRfQ0
— AG (@AGHamilton29) August 17, 2026
On Monday, however, the Los Angeles Times reported that the poll was fake. After numerous inquiries from a Times reporter, Median Strategies itself admitted the poll was bogus. Median called the survey a “short term social experiment to see how polling information could be distributed without verification.”
Afterward, Median published a statement announcing it was “discontinuing” operations and would be withdrawing all previously published polls.
The statement read:
Median Strategies has concluded its polling project and will not publish additional polls.
All previously published polling releases have been withdrawn and should not be cited or treated as genuine polling data.
Median Strategies was created as a short-term social experiment examining how purported polling information could enter and spread through the political information ecosystem without independent verification.
So this is pretty wild: A polling firm that came out of nowhere to show Francesca Hong with a 23-point lead over David Crowley in the Wisconsin primary has shut down and said its whole operation was bogus pic.twitter.com/QTmlXkvGPC
— Brett LoGiurato (@BrettLoGiurato) August 18, 2026
According to the Times:
The fake poll by Median Strategies showed Bass leading Raman by nearly 12 percentage points ahead of the Nov. 3 mayoral runoff, and said the results were based on a survey of 560 Los Angeles voters between July 30 and Aug. 5. It said it had a 4.1% margin of error.
The phony poll also claimed that 16% of its respondents said they had voted for Pratt, and that of those voters, about 85% were planning to vote for Bass in the general, while about 7% would support Raman.
“Faced with a choice between the incumbent and a Democratic Socialists of America-aligned challenger, most Pratt voters appear to be treating Bass as the more familiar, more moderate option rather than sitting out the race,” Median wrote in its analysis.
The Times reported that Median only “began posting on social media on Aug. 9 and had a total of about 20 followers.”
According to the Times, in its first post, Median had the audacity to write:
Polling has a credibility problem. Ours is the fix: no partisan clients, no campaign or party money, numbers that go wherever the data takes them even when it’s inconvenient. Trustworthy polling, rebuilt.
Ahead of last week’s Democratic gubernatorial primary in Wisconsin, Median released a poll showing Hong leading Crowley by more than 20 points (referenced above).
Following the primary, Median acknowledged on social media, “We got Wisconsin wrong- and by quite a margin too.”
I’ll say!
As astonishing as the Median Strategies story may be, it raises questions that extend well beyond one dishonest polling operation. How many other pollsters — even well-established ones — might be manipulating or even fabricating their results to achieve a desired outcome?
The consequences extend far beyond a few embarrassing Election Night surprises. Polls are increasingly driving the prediction markets, which in turn reinforce perceptions about which candidates are surging and which races are already decided.
Ultimately, how much does that feedback loop influence voters themselves — convincing some that their preferred candidate cannot win, dampening enthusiasm, and perhaps even persuading them that there is little point in bothering to cast a ballot at all?
Behind every poll are people and organizations with their own interests, biases, and agendas. Polls are intended to provide a snapshot of a race at a particular moment in time. But when they are used as a psyop — a tool to influence voters’ decisions and shape election outcomes, as they increasingly appear to be — they cease to measure public opinion and instead become another means of manipulating it.
At a minimum, the spectacular misses we’ve seen this primary season should remind us that polls deserve far more skepticism than they often receive. At best, they offer a snapshot in time; at worst, they can paint a profoundly misleading picture. Polls should never be mistaken for political reality.
Donations tax deductible
to the full extent allowed by law.
Comments
Isn’t fraudulent polling and voting a Socialist staple?
Fake but true?
You wrote, “Ultimately, how much does that feedback loop influence voters themselves — convincing some that their preferred candidate cannot win, dampening enthusiasm, and perhaps even persuading them that there is little point in bothering to cast a ballot at all?”
At what point does this cross over into legally (criminally) actionable activity? If it can be shown that fraudulent pollsters like this are consistently assisting one side or another, doesn’t that constitute a conspiracy to commit election interference?
Naw. Actual prosecution is reserved tor conservatives who text, “Vote for Hillary on WEDNESDAY!”
I upvoted you, because it’s a good reminder of the double-standard conservatives often face legally. But I do want to comment that trying to deceive voters about the correct date on which to vote is at the very least unethical and inappropriate. And yes, voters should definitely be better informed than to fall for such a stunt. Both things can be true at the same time.
It’s a freeking joke fer crying out loud! I heard I it first in the 60’s. Anyone falling for it obviously should not be voting.
This one is old enough to vote.
As far as I’m concerned, if you’re too stupid to know how to get a voter ID, but they let you vote anyway, you’re fair game for pranking.
The politicians are free to prank people like this 24/7, why aren’t I?
It wasn’t that old chestnut, it was a new variant that would actually seem plausible to low-information voters. Same idea, but dressed up more plausibly in context.
He was convicted, not of “election interference”, which isn’t a crime, but specifically of Conspiracy against rights: “If two or more persons conspire to injure, oppress, threaten, or intimidate any person […] in the free exercise or enjoyment of any right or privilege secured to him by the Constitution or laws of the United States, or because of his having so exercised the same.”
The theory was not that his advertisements interfered with the election, but specifically that they deliberately prevented people from voting, which is a legal right they have. Trickery was deemed to be included in “injure, oppress, threaten, or intimidate”.
His conviction was overturned because the prosecution had not proved that he had any conspirators. The Act specifically requires “two or more persons”, so a single person acting alone doesn’t violate it.
None of this has any connection to the current topic, because fake polling does not interfere with the act of voting. It may convince low-information people not to vote, but it would still be those people’s decision not to vote, whereas in Mackey’s case his victims (if any) wanted to vote and were prevented because he convinced them that they had already done so.
In principle what he did was no different than setting up an entire fake polling place, printing fake ballots, handing them out, and collecting them, thus preventing anyone who falls for it from casting an actual vote at an actual polling place. Or “harvesting” people’s votes on a promise of mailing them for the voters, and instead destroying them. That would certainly be a crime, and if done by two or more people in collusion would violate 18 USC 241.
There’s no such crime.
All polls are lies.
So communists and their allies lie about everything all the time.
Make two lists of things about Karen Bass:
1. A list of things that are true
2. A list of things that are false.
You will find that all the things that are true are negative whereas all the things that are false are positive.
Conclusion: There is not one positive thing that is true about Karen Bass
Weird. “… a ‘short term social experiment to see how polling information could be distributed without verification’ smells a lot like misfeasance, unreported in-kind campaign contribution, conspiracy to commit election fraud, and influence peddling.
None of those are actual crimes.
Particularly it could not be an “unreported in-kind campaign contribution”, because nothing was actually given to any actual campaign. By definition a campaign contribution must be (a) a contribution; something of value must be given. and (b) it must be given to a campaign.
Doing something on your own accord that helps a candidate, without consulting that candidate’s campaign (they are two separate entities) is not a contribution to that campaign, and is thus unregulated and doesn’t need to be reported to anyone.
As for the other charges, “misfeasance” requires that you owe a duty of care to some person or entity and fail in it. No one has a duty to the general public to give them truthful information.
There is no such crime as “conspiracy to commit election fraud”, nor would publishing a fake poll be “election fraud”.
Nor is there any such crime as “influence peddling”, but if there were such a crime it would require that you have influence, and that you peddle it; how does publishing a fake poll fulfill either requirement? What influence, and to whom is it being peddled?
Should I conduct a “short term social experiment” to blow up DSA headquarters and see if I get caught?
Blowing up a building is a crime. Publishing a fake poll is not.
If this is not illegal I smell a money making scheme selling desirable polls to politicians. Since politicians always tell voters what they think they want to hear it is only fair I sell them a poll that I think they would want to see.
You can do that all you like, so long as you don’t commit fraud against your customers by making them think you’re a legit polling organization and that they will be able to convince the news industry to publish the poll you’re selling. Without that it’s worth nothing to them, so by convincing them of this you’d be defrauding them of the money they pay you. Note that the voters would not be the victims of this fraud; you would have got nothing of value from them, which is a fundamental element of fraud.
If you did set up such a business the news industry would quickly learn about it and stop publishing your polls, thus making them worthless and triggering the fraud charge I described above.
If one ever wondered about the veracity of polls, here’s your answer.
“Median Strategies was created as a short-term social experiment examining how purported polling information could enter and spread through the political information ecosystem without independent verification.”
We were just kidding! It was a TEST, and you PASSED!
No, they’re saying it was a test and the news industry FAILED. They wanted to see whether the news industry would do some basic fact checking before publishing their results, and were surprised to find that they didn’t.
Lies, damned lies, and Polls.
Forget it, Jake. It’s Chinatown.
Funny thing that, as so far reported, Median Strategies in Cali, WI and MI limited bogus poll publication to only 3 ‘Rat Party Donk primary races and no others. Can’t say what that indicates, but them’s the facts. [???]
This is clear-cut election interference. Pretty sure that’s a felony.
Did these guys “conduct polls” in. The first dem primary? I’d insist—at minimum—on a recite with all three candidates.
No, there’s no such crime as “election interference”, let alone a felony.
No, the whole point of their experiment is that they didn’t conduct any polls, they just made them up and published them, and saw whether the news industry would gullibly report them, or do due diligence and ignore them. The industry failed.
A “recite”? What’s that? Like a concert?!
I am SHOCKED .. shocked I tell you
that anything associated with Dem politics is based upon lies or anything underhanded ,..
/sarc