Are Fake Polls Good for Democracy? Not By Themselves

Opinion from the Left

For the second time this election cycle, a self-described “polling organization” has admitted to publishing completely fabricated polls that were prominently reshared by major media organizations.

As reported by Newsweek (Center bias), a website called Wedgewood Polls appeared in June and began releasing “polls” of 2026 midterm races. Outlets across the political spectrum republished Wedgewood’s fake polls, including Newsweek itself, The Hill (Center), Breitbart (Right), The Daily Beast (Left), and the Washington Examiner (Lean Right). Prominent polling aggregators including the New York Times (Lean Left), Silver Bulletin, and FiftyPlusOne included Wedgewood in their polling average calculations for the 2026 midterm elections.

After the news broke Wednesday, former Michigan Senate primary candidate Mallory McMorrow, a Democrat, claimed a fake Wedgewood poll that showed her trailing her opponents greatly diminished confidence in her campaign and contributed to her decision to withdraw.

The revelation harkened back to another fake polling “experiment” uncovered in August. A firm called Median Strategies released a “poll” showing Karen Bass with a wide lead over Nithya Raman in the Los Angeles Democratic mayoral primary.

At first blush, this would seem to be a straightforward story about traditional media behaving recklessly with unverified sources, and many commentators seem to be treating it as such. But if you look closer, the outlets that republished these polls arguably don’t deserve the heat they’re getting. They did their due diligence, but Wedgewood went to great lengths to conceal its deception.

Shortly after Wedgewood posted a since-deleted statement revealing the nature of their project, FiftyPlusOne said it had alerted Wedgewood to an investigation into some of their fishy-smelling numbers. FiftyPlusOne rebutted Wedgewood’s claim in the statement that they hadn’t been vetted appropriately by media outlets and polling aggregators; rather, FiftyPlusOne had contacted them for information about their operation and to confirm their methodology statement, and Wedgewood replied with false information.

Per Newsweek, other polling aggregators had reached out to Wedgewood by phone and received similar assurances, while several journalists were immediately suspicious of Wedgewood’s results and cautioned the public against putting much stock in them. The New York Times marked them as low-quality in their aggregator, meaning they were weighted less than established, reputable pollsters in the calculated polling average. Wedgewood even set up a crowdfunding campaign for a poll that they did not intend to actually conduct.

If you genuinely wanted to test the media’s credulity about polling, and the media immediately started scrutinizing you — would you keep the farce going as long as you could? Would you lie on multiple phone calls? If within a few months of existence, a polling aggregator discovered your deception of their own volition, would you release a statement not giving them any credit whatsoever? If after all that, you still declare that the media has failed your supposed good-faith test, how could they have possibly passed it?

As soon as the media started calling their bluff, Wedgewood should have ended the experiment and commended those who were appropriately suspicious. What they actually did looks more like an effort to discredit media organizations no matter how they responded.

Sure, more rigorous vetting could have exposed the fraud earlier; but what amount of vetting is worth the time for a media business? We have laws on the books preventing fraud like this, without which it would be difficult to do any sort of business at all.

That said, it’s truly a tough call how much blame to place on the media. Some outlets in this story come across as more credulous than they should have been; and we should absolutely hold the media to a high standard of trustworthiness.

However, it’s a bit frustrating to have this conversation knowing that digital platforms like X have annexed the lion’s share of Americans’ attention from traditional media but are held to no such standards — and if they were, would fall comically short. Just look at the recent viral claim that the Democratic candidate for Senate in Texas, James Talarico, had been hospitalized in a mental institution in Ohio, which despite being viewed millions of times on X had zero basis in reality.

Which brings me to my broader point.

Exposing weaknesses in editorial media institutions is a necessary part of rebuilding a healthier information ecosystem — sunlight is the best disinfectant, as they say — but it’s not sufficient. We need to also address the structural defect in today’s media ecosystem that prevents trustworthy institutions from growing.

What Kind of Media Ecosystem Do We Live In?

In the 20th century, the media ecosystem was monocultural. A handful of newspapers and TV stations set the parameters of discourse across the country. Their reign was long and peaceful for a few reasons.

For one, they played their cards right. For a long spell after World War II, American news consumers were broadly content with the status quo, so alternative institutions struggled to find purchase. Then, when public faith in government diminished — starting with McCarthyism, accelerating through the Vietnam War, and peaking with Watergate — major publications did a decent job of questioning official narratives and prying the truth out of authorities, making populist critiques seem trite.

Most importantly, it used to take a lot of resources to scale raw information into major news stories: printing presses, on-the-ground reporters, industrial quantities of ink and paper, lights, cameras, broadcast licenses, etc. These were required ingredients for a certain “professional” aesthetic that was hard to fake. So it was a pretty reasonable expectation that any professional-looking news story you saw had passed multiple credibility checks on its way to your lap or living room.

But a monocultural ecosystem is vulnerable to external shocks. When digital communication entered the picture, all of a sudden the same professional polish that was once hard to fake became quite easy to copy, and distribution costs fell to near-zero. Spinning up a well-designed website required only a well-trained eye and a laptop. And now with AI coding, you don’t even need the well-trained eye anymore.

For a while, nothing much seemed to be changing, as digital media occupied a bottom-feeding position while the media titans continued to get all the sunlight. But it’s now clear that digital media has choked off a vital supply of nutrients to the old institutions. Raw information is now packaged into a narrative and weaponized for all manner of devious purposes before the New York Times can blink. A general news outlet founded today has minimal opportunities to break a story before half the internet is already fighting over it. Such institutions can no longer grow to maturity, and the existing ones cling to life in their old age.

So while we would all like media outlets to vet information for falsehoods with extreme care, that would amount to a fool’s strategy in the business today. They’re in a never-ending race to be the first to take advantage of new information; when they come across a juicy story that seems 95% legit, the temptation is very strong to just run with it. It’s not good that outlets do this, but in a sense it’s understandable.

We’re now entering an Ayn Rand-style situation in which the most trusted voices are individuals, cultivating personal brands and parasocial relationships with relatively small but devoted audiences. A lot of people seem to think this is a fantastic development with no downsides, especially people who are good at cultivating personal brands and parasocial relationships.

I beg to differ. This is a good way to make a buck, but it’s no way to build an institution, in the true sense of the word. Institutions are where humans do the difficult but essential work of cooperation. Institutions only function under norms of honesty, accountability, and loyalty to a mission. Cults of personality are one-to-many relationships; institutions are many-to-many.

Unfortunately, it seems the biggest multi-human entities that can emerge from the current ecosystem are siloed camps organized around various combinations of ideology and personality. There is little communication between these camps other than people dunking on other camps for status within their camp. One day, your camp gets dunked on so hard, its incoherences exposed so irrefutably, that it’s not fun anymore. You slink off into the wilderness to find another camp, eventually convincing yourself you always thought that camp was cringe anyway.

A trustless society is perhaps slightly better than an Orwellian society, but it's far worse than a healthy liberal society. The feudal camps I describe are clearly not sufficiently aligned with truth-seeking to be an adequate replacement for.

We can’t put the genie back in the bottle; the task before us is not to reinvent the media titans that ruled the 20th century. It is to engineer a new information ecosystem in which credible institutions emerge. What constitutes a hard-to-fake signal in the AI era? When publishing breaking news, what’s the right balance of speed and verification? These are some of the questions we should be asking ourselves. It doesn’t help to simply call out every mistake made by traditional media like a broken record player.

Evan Wagner is a Product Manager at AllSides. He has a Lean Left bias.

Henry A. Brechter, Editor-in-chief (Center bias) and Julie Mastrine, Director of Communications and Bias Ratings (Lean Right).