Mitt Romney and Barack Obama after presidential debate at the University of Denver. File/Associated Press

Rasmussen Reports used to be a fairly creditable and credible political polling organization, good enough to be included among the pollsters relied on by services such as FiveThirtyEight to give a broad-spectrum gauge of voter sentiment in the run-up to state and federal elections. It’s true that Rasmussen had a detectable pro-Republican “house effect,” in polling parlance — but one that was consistent enough to compensate for in published polling averages. But something has happened to Rasmussen in recent years. Not only have its results become more sharply partisan, favoring Republican and conservative politicians, but it also has increasingly promoted right-wing conspiracy theories on topics such as race relations, election results and — perhaps most troubling — COVID vaccines and COVID origins.

Earlier this month, Rasmussen tweeted the results of polls it conducted in June 2023 and last month, claiming to find that one in five Americans believe they know someone who died from a COVID vaccine. There are many reasons to disregard any such poll asking people what they think about a scientifically validated fact — in this case, that the record shows overwhelmingly that the COVID vaccines widely used in the US are safe and effective. But Rasmussen has doubled down on its findings. In a series of tweets on June 9, it declared, first: “If the numbers implied by our COVID polling are correct, the vaccines killed more people worldwide than Jews killed in the Holocaust.” Then it tweeted: “China lied. Fauci lied. People died.” And followed that with: “The government takeover of medicine was as deadly as always predicted.”

In other words, Rassmussen has morphed from a quantifier of public opinion into a participant in the spread of noxious propaganda. It still tries to validate its results by claiming that they’re “relevant, timely and accurate,” citing its “track record.” But that track record has been sprouting gray hairs. The most recent election polling cited by the web page documenting its track record is from 2010. More recently, 538, now owned by ABC News, dropped Rasmussen from its polling averages in March. ABC took that step after Rasmussen failed to respond suitably to a questionnaire 538 submitted asking Rasmussen to explicate its polling methodology. Rasmussen published ABC’s query on its website under the headline, “ABC News: ‘Answer Our Questions — Or Else!’”

I asked Rasmussen Reports by phone and email to comment on its tweet and its polling, but received no response. Rasmussen’s veer to the far right has been noticeable for several years. Founded in 2003 by pollster Scott Rasmussen, the firm’s forecasts received high marks for accuracy in the 2004 and 2008 presidential elections. But it fell short in 2012, predicting victories for Mitt Romney over Barack Obama in several states that Obama won. As my colleague James Rainey observed in the aftermath, the Rasmussen polls had been used by conservative media outlets “to prop up a narrative in the final days of the campaign that Romney had momentum and a good chance of winning the White House.”

In 2013, Scott Rasmussen left the firm due to unspecified business disagreements with its owner, the private equity firm Noson Lawen Partners.

In recent years, the firm has resembled a pollster-for-hire appealing to conservative organizations and authors. During the Trump administration, it became known for “a social media presence that embraced false claims that spread widely on the right,” Philip Bump of the Washington Post observed in March.

The firm’s treatment of the 2022 Arizona gubernatorial election, in which Democrat Katie Hobbs defeated Republican Kari Lake, is a good example. In March 2023, Rasmussen reported the results of a poll it had conducted four months after the election, purportedly finding (according to a headline on its website) that “most Arizona voters believe election ‘irregularities’ affected outcome.”

According to Rasmussen, 51% of Arizona voters chose Lake and only 43% voted for Hobbs. The poll placed the election turnout at 92%; actually it was 62.6%.

On Steve Bannon’s War Room podcast, Mark Mitchell, Rasmussen’s lead pollster, said its results showed that “people in Arizona, by and large, think that cheating happened.” That unsupported assertion, of course, is the core of the long, fruitless campaign to overturn the election by Lake — who gleefully cited the Rasmussen results.

Rasmussen polls on COVID vaccines and other such topics aren’t entirely worthless. They may not tell us anything useful about scientific research or electoral results, but they do offer a window into how propaganda and claptrap have penetrated deeply into our political discourse, at least within the right-wing fever swamp.

That brings us back to its polling on COVID and COVID vaccines. Rasmussen’s methodology seems to include wording its questions as if they are stating a fact, no matter how dubious. For its May 2024 poll of 1,250 American adults, for instance, it asked, “Do you know someone personally who died from side effects of the COVID-19 vaccine?” Rasmussen reported that 19% replied in the affirmative; the poll had a margin of error of 3%.

Such questions have obvious flaws. The most important is that most respondents have no way of knowing whether an acquaintance’s death was related to the vaccine; nor does Rasmussen, which conducts its polls with robot calls, have any way of authenticating the respondent’s answer.

Blaming the COVID vaccines for a tide of undocumented injuries and deaths is a popular theme in the anti-vaccine community.

For them, it has the virtue of being suggestive and unverifiable; with nearly 700 million doses of the Moderna and Pfizer vaccines having been administered in the U.S. alone, the law of large numbers implies that “by random chance alone ... there will be a large number of people who die within, say, 30 days of being vaccinated even if the vaccine has absolutely nothing to do with their deaths,” in the words of veteran pseudoscience debunker David Gorski.

It’s not unusual for the death or illness of a prominent entertainer or athlete to provoke swarms of anti-vaxxers to assert that the victim must have been recently vaccinated. Florida Surgeon General Joseph Ladapo, who I earlier identified as “the most dangerous quack in America” and a “card-carrying member of the anti-vaccine mafia,” misrepresented published research to claim that the COVID vaccine presented an elevated threat of cardiac problems for young men.

The research said no such thing; on the contrary, it said that the risk of cardiac death from the vaccines was statistically nonexistent and, indeed, lower than the risk of cardiac death resulting from catching COVID-19 itself.

Despite all that, conjectures by laypersons that the illness or death of acquaintances can be traced to the vaccines are legion. One promoter of the idea, economist Mark Skidmore of Michigan State University, even concluded from an anonymous database of 2,840 respondents compiled by a third-party survey firm that the number of respondents who said they knew someone who had died from the vaccine meant that the number of deaths from the vaccine in the US “may be as high as 278,000.”