Fake Polls and Election Integrity
Jokes (and incidents) abound featuring dead voters voting, but what about dead pollsters polling? Last Tuesday, election analysis site FiftyPlusOne interviewed an affiliate of SoCal Strategies, a polling outfit that had gained recognition from major news organizations like The New York Times. By Thursday, SoCal Strategies had deactivated its website and social media accounts, and FiftyPlusOne never received the follow-up information for which it had asked. When its polls attracted scrutiny, SoCal Strategies simply went dark.
But the story grows even stranger. On May 22, 2026, SoCal Strategies filed Articles of Dissolution as a limited liability company registered in Florida. The given reason was exceedingly vague. “Its [sic] time for this to end and for us to move on. I just want to put this all behind.” SoCal Strategies published three polls after its dissolution.
FiftyPlusOne concluded that at least the two most recent SoCal polls it counted were fabricated, and it removed them from its database. The dissolution in May suggests that SoCal may once have been a legitimate polling outfit, but it is no longer.
SoCal Strategies appears to be connected to Wedgewood Polls, the second fake pollster outed this cycle. The same “anonymous source” tipped off FiftyPlusOne about the dubious nature of both outfits, prompting the election analysis group to launch a “parallel investigation” into both. One person identified with Wedgewood named Jack Sims (if that is a real name) said he used to work at SoCal Strategies, and indeed Jack Sims was the name listed on SoCal’s dissolution paperwork.
The incident that raised suspicions about Wedgewood was a poll allegedly showing Senator Cindy Hyde-Smith (R) up by only four points in Mississippi. The Washington Post pointed out this poll came “just hours after a prominent election handicapper changed its rating of the race from ‘Safe Republican’ to ‘Likely Republican.’”
When FiftyPlusOne obtained the raw polling data for the Mississippi poll, they noticed it was “surprisingly clean,” meaning it had too few skipped answers, and it too-closely mirrored the state demographics. FiftyPlusOne paid $900 to commission a poll from the same sample provider, Pollfish, with a questionnaire identical to that of Wedgewood. Their poll featured a far less representative sample of voters.
More importantly, they found major discrepancies in two columns containing survey metadata: in a column labeled “trans id,” the new poll gave respondents a number strongly correlated with the time they completed the poll, whereas in Wedgewood’s poll it had no correlation. Non-professionals may have neglected such a minor point, but to statisticians, this constituted fairly compelling evidence that Wedgewood’s poll data was fake.
It’s worth considering how someone could generate fake polling data. Most polls aim for at least 1,000 respondents, who will answer at least a couple dozen questions (counting demographic questions). To fabricate a facsimile, someone would have to enter data into tens of thousands of individual cells, including cells that contain unique strings of nine- and 10-digit numbers. Obviously, no one would attempt anything like this by hand. But an AI program might spit one out relatively quickly. An AI program might also generate data that was “surprisingly clean” based on the parameters given to it. In other words, even if this was not Wedgewood’s strategy, the age of AI makes the possibility of fake poll data much more plausible.
Several months ago, a polling firm called Median Strategies closed up shop, admitted it was fake, and declared that it had been a “social experiment” to test whether the media would believe its fake polls. But that amateurish operation had only a minor impact, compared to Wedgewood.
Wedgewood not only published fake polls, but it also generated fake polling data and took additional steps to deceive polling aggregators such as The New York Times and DecisionDeskHQ. According to the Times, Wedgewood “answered a series of questions regarding the identities of the pollsters; the methods they were using to draw their sample, contact respondents and weight their polls; and the source of their funding.”
In the midst of FiftyPlusOne’s investigation, Wedgewood also admitted its polls were fake and closed up shop. Just like Median Strategies, it claimed to be a social experiment. Days later, SoCal Strategies simply vanished with no explanation at all.
The good news here is that fake pollsters are getting caught. There are enough honest analysts to track down the legitimacy of unexpected data results and hound out at least the most obvious fake forecasters.
The bad news here is that fake pollsters exist. And, given the accelerating competence of AI, this is likely only the beginning of the Fake Polls Era.
To understand the damage this era can cause, let’s consider why political opinion polls exist at all. The primary purpose is to inform political campaigns. By taking a snapshot of the electorate, campaigns can strategize about which messages to emphasize, where to focus the candidate’s time, and even whether it is still worth staying in the race.
Because polls play such an informative role, fake polls can shape a campaign. For instance, in the Democratic primary for U.S. Senate in Michigan, establishment candidate Mallory McMorrow dropped out of the race against Abdul El-Sayed after being “hounded” by calls to do so, because Wedgewood’s polls showed her trailing.
Wedgewood’s confession had claimed that they “never intended to sway an election,” but the fact is that they did. “Sure, a point was proven,” McMorrow mourned. “But 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.”
A secondary purpose of polling is to inform the public of the state of the race. Rational voters like to know that their vote accomplishes at least some sort of purpose. If their favorite candidate is trailing in a distant fourth place, in single digits, they might make the rational decision to decide between the top-performing candidates instead. Some voters might even vote for the leading candidate because they like to back a winner.
The same logic stands for campaign donors.
For these reasons, political polls are not just conducted by and for the campaigns, but they are also disseminated widely through political media.
For the same reasons, fake polls could conceivably influence the behavior not only of campaigns but also of voters. Thus, fake polls present a novel threat to election security. They do not involve extra ballots, suppressed ballots, fake voters, or any of the more traditional forms of voter fraud, which have laws against them. Instead, fake polls exert their illegitimate influence on elections subtly, by nudging undecided voters towards or away from certain candidates. And, with so many elections decided by close margins, a nudge may be all it takes to flip a critical seat.
Political campaigns have been engaging in “dirty tricks” for as long as there have been elections. Fake polls just provide a sophisticated new façade to the sour old playbook.
AUTHOR
Joshua Arnold
Joshua Arnold is a senior writer at The Washington Stand.
RELATED ARTICLES:
Don’t Trust the Polls—Make Your Poll Count
EXCLUSIVE: ICE Drops New Batch Of Illegal Immigrants Charged With Voter Fraud Ahead Of Midterms
Breaking: Evidence Indicates Democrats Intentionally ‘Stacked’ the Ballot in Key Senate Race
RELATED VIDEO: Amir Tsarfati: Post 2020 U.S Election Special Update
EDITORS NOTE: This Washington Stand column is republished with permission. All rights reserved. ©2026 Family Research Council.
The Washington Stand is Family Research Council’s outlet for news and commentary from a biblical worldview. The Washington Stand is based in Washington, D.C. and is published by FRC, whose mission is to advance faith, family, and freedom in public policy and the culture from a biblical worldview. We invite you to stand with us by partnering with FRC.
Leave a Reply
Want to join the discussion?Feel free to contribute!