A national polling organization disclosed on Tuesday that most of its published data had been fabricated, describing the effort as an "experimental project to demonstrate vulnerabilities" in the political polling industry. Wedgewood Polling, whose surveys had been cited by several news outlets, issued the statement after the analysis site FiftyPlusOne questioned the legitimacy of its results.

The firm said on X that it sought to test "how easy it is for an entirely fabricated polling firm to gain traction, be taken seriously, and expose how little scrutiny exists in media and public commentary." The short answer, it added, was "far too easy."

Wedgewood’s data had appeared in aggregations used by FiftyPlusOne, The New York Times, MS NOW, Newsweek, Decision Desk HQ and Mediaite. Following the admission, those outlets removed the figures from their databases. FCC Commissioner Brendan Carr subsequently warned that broadcasters could breach regulatory obligations by airing "fake polls," though he did not cite specific examples.

The firm identified itself as a "fictional entity run by a small group of industry veterans" and noted that its X account opened in June 2026, while its website now returns an error. FiftyPlusOne and The New York Times said they had contacted Wedgewood for verification before publishing its data, receiving no response that would have prompted further scrutiny.

Independent news outlets had previously flagged Wedgewood’s practices. Straight Arrow News and Urban Milwaukee highlighted the lack of contact information on its site and questioned the credibility of a poll showing Wisconsin Democratic gubernatorial candidate Francesca Hong at 30%, a figure that contrasted with other surveys and preceded her primary loss. Bruce Murphy, editor of Urban Milwaukee, described the outfit as "shady" and noted that it had earlier admitted to fabricating results as part of a "short-term social experiment."

Democratic Michigan state senator Mallory McMorrow, who paused her U.S. Senate campaign in July, detailed how a Wedgewood poll showing her support in single digits allegedly caused a drop in donations and volunteer morale. She wrote on X that while the experiment proved a point about polling flaws, "tens of thousands of people poured their hearts, time, energy, hard-earned money into something they believed in… for a year-and-a-half, and it was treated like a game."

Analysts pointed to methodological issues in Michigan’s primary polling. Courtney Kennedy of Ipsos explained that many surveys used a text‑to‑web approach, which reaches younger, progressive voters efficiently but struggles to capture moderate respondents. Phone‑based polls performed better among centrists, she said, illustrating how data collection methods can skew results.

FCC Commissioner Anna Gomez pushed back against Carr’s warning, stating that broadcasters reporting publicly available polling data in good faith should not be penalized for a poll’s inaccuracy. She emphasized that FCC rules on news distortion require proof of deliberate misrepresentation, not isolated errors.

The incident has reignited discussion about best practices for evaluating polls. Experts recommend checking a pollster’s track record, transparency about methodology and the presence of a margin of error. Lee Miringoff of the Marist Institute of Public Opinion noted that probability‑based polling requires every participant to have a known chance of selection and that reputable surveys adjust samples through weighting to achieve representativeness.

Polls are designed as snapshots of public sentiment at a given moment, not forecasts of election outcomes. While they can offer a high‑level view of attitudes, they cannot reliably predict winners in tightly contested races, according to Kennedy.

A free press remains essential to democratic discourse, and readers are encouraged to support trustworthy journalism.