For four months, Wedgewood Polls was a ghost. No phone number on its Google Sites page. No employee names. No address. A website registered on June 26 and an X account created the same month. And yet its numbers were good enough to land on the polling pages of the New York Times, Nate Silver's operation, VoteHub, FiftyPlusOne, and Decision Desk HQ — and to be cited by Newsweek, the Hill, Breitbart, the Daily Beast and the Washington Examiner.
On Tuesday, the ghost confessed. In a statement posted to social media after the election-analysis site FiftyPlusOne contacted the firm for comment on its investigation, Wedgewood Polls admitted that it had fabricated its numbers. The firm described the project as a "social experiment" and a "wake-up call," insisting it was meant "to demonstrate the vulnerabilities, lack of oversight, and overall low barrier to entry in the modern political polling industry."
The cleanup started immediately. Decision Desk HQ said it had pulled 17 Wedgewood polls from its race-level polling lists and averages. The New York Times published an editor's note stating that all Wedgewood polls had been removed from its poll tracking project.
What Was Actually Published
The fabricated surveys were not confined to obscure websites. They were midterm horse-race numbers — the kind of data that drives fundraising emails, cable chyrons and campaign strategy memos in the final weeks before an election.
A June Instagram post tied to the operation described a Wedgewood survey of 600 likely Iowa voters. That post, whatever its reach, was built on data the firm now says it made up.
The warning signs were there for anyone looking. On July 10, a Washington Monthly columnist flagged that Wedgewood's first poll — of a Louisiana Republican primary runoff — had been "wildly wrong." In August, Straight Arrow News dug into the firm's background and found a website with no names, no phone number and no location. On August 16, an Urban Milwaukee journalist, after a Wedgewood poll showed Francesca Hong at 30% in the Wisconsin Democratic primary for governor — far above where legitimate surveys had her — called the outfit "a shady outfit." Adam Carlson, a Democratic pollster at a different firm, told Straight Arrow News he had communicated with someone connected to Wedgewood.
None of that stopped the aggregators from ingesting the numbers.
The Question Nobody Has Answered
The Washington Examiner reports that who exactly was running Wedgewood — and why — remains one of the biggest mysteries in Washington weeks before the midterms. Reports describe the operation being run by career pollsters claiming to be from both sides of the political aisle, while another polling aggregator, drawn into publishing the skewed results, said it was told the firm was founded by college students with scant professional experience as a grassroots project.
Both stories cannot be true. Neither has been confirmed. No named individual has been publicly tied to the operation as its operator, and no one has explained how a firm with no verifiable footprint cleared the vetting of multiple national aggregators.
The scale of the failure is the story. The New York Times took the polling. Nate Silver's site took the polling. Decision Desk HQ took the polling and had to strip 17 surveys once the confession landed. FiftyPlusOne's investigation forced the admission. The gatekeepers were not the ones who caught it — a columnist, a local journalist and a small news outlet were.
What It Changes
For campaigns, it means any strategy built on Wedgewood's midterm numbers in those 17 races was built on fiction. For the aggregators, it means a public accounting of what vetting actually looks like before a firm's data gets averaged into a national narrative. For voters, it means the polls they were shown this fall — and the coverage built on them — deserve more scrutiny than the people paid to protect that data gave it.
The "social experiment" defense does not answer the question the Examiner and others are still asking: who signed off on the experiment, and whether the point was to expose the industry's flaws or to exploit them. Until names are attached, the mystery stays open.
Our Take
Here is what the mainstream press will not say plainly: a fabricated polling operation walked straight through the front door of the most trusted names in election data, and the people who run those operations did not notice for four months. The New York Times, Nate Silver, Decision Desk HQ — all of them laundered made-up numbers into the national conversation and only pulled them after a confession.
This is the same institutional class that lectures Americans about "misinformation," demands censorship powers over social media, and tells voters which sources are legitimate and which are not. It turns out the watchdog has no fence. A Google Sites page and a two-month-old X account were enough.
If polling can be fabricated this easily, then so can the narrative the polls are used to sell. Every voter who was told a race was over, or a candidate was surging, based on Wedgewood data was sold a bill of goods. Demand the names. Demand the vetting standards. And remember, in the final weeks of this midterm season, who built the machine that let the fakes in — and who got caught.
