Many Americans of all political persuasions long ago lost confidence in public opinion polling. The repeated errors and misses by a host of polling firms, almost always in favor of liberal or progressive politicians, led many Americans to conclude polling is often intended to influence public opinion (and elections), rather than report on the mood of the electorate. In the last few months, four separate polling firms were revealed to be illegitimate (or at least suspect), further hammering the polling industry’s credibility.
American Polling Already Lacked Credibility
Public opinion polling in the USA has long been a debacle. In fact, I dedicated three chapters of my book Fake News Exposed About Trump, Volume One to the shoddy polling of the 2016 and 2020 Presidential race and the biased media reporting promoting it. Perhaps, you remember this colossal miss from The New York Times in 2016 (the NYT projection was “based on the latest state and national polls”):
After the complete polling debacle in 2016, the American polling industry repeated the same mistakes in 2020, as this article in Fortune highlights quite well:
I also wrote about polling misinformation and associated media propaganda on this substack. See here:
Shortly after that article, I encouraged readers to remain critical of 2024 polling in another article here:
The pollsters favored by legacy media in this Country simply can’t get it right.
The Polls Almost Always Favor Democrats
The repeated polling errors alone are enough to cause voters to question the legitimacy of polling, but making matters worse is the fact the errors almost always favor Democrat politicians. These one-sided errors led many to conclude the errors are not errors at all, but intentional efforts to influence public opinion in favor of Democrat candidates for office. A few illustrations will inform readers.
Polling of Wisconsin in 2016 called for Hillary Clinton to win the state comfortably:
The result? Donald Trump won Wisconsin in 2016 by 1% or 27,257 votes.
In 2020, polling of the Presidential race in Florida had Joseph R. Biden winning the state handily:
The result? Donald Trump won Florida in 2020 by 4.3%.
I could go on and on listing one polling miss after another with virtually all of the errors favoring the Democrat candidate, but the point is made: the polling relied upon and promoted by legacy media almost always tilts the playing field in favor of the Democrats. (The two substack articles of mine linked to above cover a wide range of similar inaccurate polls for anyone wanting more examples).
What Happened Now?
This election cycle four separate polling firms were revealed to be illegitimate, some with completely fabricated data.
Median Strategies. The first domino to fall was “polling” firm Median Strategies in August 2026. Median Strategies outed itself, so to speak, in an unsigned statement claiming (in part):
Median Strategies was created as a short term social experiment to examine how purported polling information could enter and spread through the political information ecosystem without independent verification.
The “polls” from Median Strategies showed Los Angeles Mayor Karen Bass leading councilmember Nithya Raman by 12 points in the mayor’s race, the Democratic Socialist Francesca Hong leading David Crowley by more than 20 points in Wisconsin’s Democratic primary for governor, followed by a poll of the Nevada governor’s race showing Republican Governor Lombardo leading his Democrat challenger, NV AG Aaron Ford, 51-46%.
The Public Sentiment Institute. Around the same time, The Public Sentiment Institute or TPSI admitted, in a survey of the Florida governor’s race, it altered the results of its survey to realign respondents who indicated they favored Florida Lieutenant Governor Jay Collins to incorrectly reflect they favored Florida Republican gubernatorial candidate James Fischback. TPSI publicly disclosed the switch in a report issued on August 18, 2026. As a result of this misdeed, polling aggregator FiftyPlusOne announced it was “permanently discontinuing use” of TPSI surveys.
Wedgewood Polling. Wedgewood Polling apparently did not poll anything at all. Yet, its “fabricated numbers” from non-existent polls were included in prominent polling aggregators and databases, as Newsweek reported on October 21, 2026:
Wedgewood Polling issued a public statement outlining the reasons these fabricated polls were supposedly created and shared:
Wedgewood Polling was prompted to reveal itself by the polling aggregator FiftyPlusOne after an investigation into the pollster. FiftyPlusOne wrote about the investigation here. Interestingly, Wedgewood Polling issued its public statement on August 20, 2026 (and folded up its operation) literally minutes after responding to a request by FiftyPlusOne for polling data. FiftyPlusOne shared Wedgewood’s response:
One cannot help but wonder if Wedgewood’s motives were more nefarious than conducting an “experimental project” to demonstrate “vulnerabilities” in polling.
Poliagg.com compiled the Wedgewood Polls and put them up on its website. Take a moment to review the data and you will find Wedgewood routinely overstated support for Democrat candidates vs. Republicans and for Democratic Socialists over traditional Democrats in poll after poll. These skewed results call into question Wedgewood’s “experimental project” excuse, making the operation look more like pro-leftist propaganda instead.
SoCal Strategies. We once again have the guys and gals at FiftyPlusOne to thank for the investigation into this polling firm. They reached out to SoCal Strategies after receiving information from an anonymous source calling into question some of SoCal’s data. I cannot direct readers to anything directly from SoCal Strategies, because it scrubbed its online footprint from existence after FiftyPlusOne reached inquiring into its operations. SoCal Strategies literally shut down its website, newsletter (also here at substack) and deleted its social media accounts. That seems completely legitimate, eh?
FiftyPlusOne reported on SoCal Strategies and the investigation in an article posted on October 22, 2026:
FiftyPlusOne announced it will no longer collect surveys published by SoCal Strategies as a result of the investigation.
I was unable to locate a compilation of all of the polling data published by SoCal Strategies. Its data is spread across several years (2024, 2025 and 2026) and was reported by numerous media outlets and polling aggregators. Some of it is suspect, such as the poll showing the far left Abdul El-Sayed ahead of Haley Stevens in the Democrat primary for Michigan Senate by 15 points (56-39) in a survey conducted from July 28-30, 2026. A few days later, on August 4, 2026, Abdul El-Sayed defeated Stevens in the Democrat Senate primary by 1% (48.5% to 47.5%). It takes real effort to miss that big.
Conclusion
The American people don’t trust polls for good reason. The public opinion polls relied upon by legacy media are routinely wrong and almost always prop up Democrats. They are probably headed for another series of big misses in the mid term elections. We will find out soon enough. The difference this election cycle is now we have 4 separate polling firms either openly admitting their data was bogus or tacitly admitting as much, after much of the data was widely disseminated and/or incorporated into polling aggregates. These disclosures are devastating blows to the remaining credibility of the legacy media pollsters.
Keep in mind, there are some good guys conducting polls (though they are often ignored by legacy media). Check out Rich Baris and Big Data Polls, Trafalgar Group, and Atlas Intel. There are a few more, but these are a good start.
There will probably be more to report about inaccurate polling after the mid terms, so stay tuned!