A fake poll fooled the LA mayor’s campaign. It’s a 'wake-up call' for other politicians

Mark Baldassare has seen his fair share of eyebrow-raising political polls in his decades-long career as a public opinion researcher.

Never, though, has he encountered a completely fake survey with made-up findings.

Until this week.

“It’s disturbing,” said Baldassare, survey director at the Public Policy Institute of California, a nonpartisan, nonprofit think tank.

“It’s a wake-up call for all of us.”

The poll in question, published last week by a group called Median Strategies, purported to show Los Angeles Mayor Karen Bass leading City Councilmember Nithya Raman by double digits — a surprising result given Bass’s low approval rating and her administration’s recent setbacks on homelessness. It also claimed that Bass had won support from a vast majority of voters who supported insurgent Republican Spencer Pratt in the primary election.

But on Monday, Median Strategies retracted its poll and replaced its website with a disclaimer that the survey results were fabricated as part of a “short-term social experiment” and “should not be cited or treated as genuine polling data.” The group’s website also featured “polls” from the Wisconsin and Nevada gubernatorial races.

The fake results, first reported by the Los Angeles Times, duped the California Post, several local television and radio stations and even the mayor’s own campaign, which drew attention to the survey in a now-deleted X post.

The incident further underscores the importance of accuracy and precision in polling and presents yet another challenge for a public already struggling to separate fact from fiction and genuine content from AI-generated slop.

Bass’s spokesperson, Alex Stack, said campaign leadership had reviewed the polling memo that accompanied the results and decided it was legitimate given that “the poll was reported on by multiple news outlets.”

“Any bad faith attempts to influence elections should be investigated and prosecuted to the fullest extent of the law,” Stack said in a statement.

A scheme to rig betting markets?

It’s unclear who’s behind Median Strategies, what their true intentions were with the bogus data and whether they would’ve disclosed the deception to the public if journalists hadn’t intervened. The group did not respond to a list of emailed questions from CalMatters.

“This definitely isn’t a research experiment, right? It sounds like an excuse for very bad behavior,” said Mindy Romero, director of the nonpartisan, nonprofit Center for Inclusive Democracy.

“If we take them at their word, that this is some sort of social experiment, that’s not what reputable researchers engage in,” Romero added. “We always have to think about the harm that we’re doing.”

One alternate theory is that the false survey results were meant to manipulate prediction markets, such as Kalshi and Polymarket, in the same way that securities fraudsters “pump and dump” penny stocks. They buy low, release disingenuous information to drive the price up, then sell high and reap the windfall.

Median’s fake poll spurred high volumes of trades shortly after it was posted, though the price movements were small. The Associated Press tracked thousands of trades from roughly 20 different accounts on Polymarket within six minutes of Bass’s campaign posting about the poll on X. On Kalshi, the price of placing a “yes” bet on Bass winning the mayoral race rose roughly two cents, from 63 to 65, within 15 minutes of the post, the AP reported.

“If nothing else, Median Strategies proved the concept that a pump-and-dump scheme is easily transferrable to election prediction markets,” wrote David Dayen for The American Prospect, a left-leaning policy publication.

Could bogus polls become more common?

For Jonathan Underland, a political strategist who works with Democrat Xavier Becerra’s campaign for governor, the Median Strategies debacle reinforces how important it is to vet a poll’s reputability before sharing it on official platforms.

“I get the excitement of a poll that shows you in the lead,” Underland said. “Everything is an opportunity for your opponent to hit you. The last thing I want to do is promote a poll that’s not real.”

Baldassare detests that public opinion polls drive financial decisions, from betting markets to more traditional political decisions such as which candidates from powerful donors and political groups pour money into. He acknowledged that copycats and wannabes could follow Median Strategies’ example and pump more noise and disinformation into the political ecosystem.

But he noted a silver lining: This is happening in August, when only a fraction of people are tuned into politics, rather than October when voters have their ballots in hand.

He hopes the Median Strategies incident reminds journalists and campaign operatives to recommit to verifying the source of the information before sharing it.

When in doubt, “Call the pollster,” Baldassare said. “Make sure I’m legitimate. You know — a real person.”