The CDC Conspiracy
Should the media cover anti-vaccination claims? Conventional reporting would probably do more harm than good if it did so by just reporting both sides and creating the impression of equivalency between weight of evidence and false claims. But what if the latest claim is a complex statistical analysis—and a senior CDC vaccine researcher appears to back it up?
This piece was going to start with a simple pie chart, one showing the number of news stories that explained the statistical issues in a controversial study claiming a link between increased rates of autism in African American boys and time of vaccination and those that didn’t. The object was to illustrate whether the evidence underlying a shocking claim was explained in such a way as to be comprehensible and comprehensive enough for a reader or viewer to make an informed decision as to its validity.
But after a trawl through the media coverage of Brian Hooker’s paper, it appears that not one news organization made an effort to explain—as Rebecca Goldin does here on STATS—what, precisely, Hooker did to come up with this claim and what, precisely, he did wrong.
When you find nothing, change the search parameters. Were there stories with some statistical explanation to contrast against those that just asserted some statistical error? Mostly not— as in, there’s barely a pie to slice up. Hooker had evaded the diagnostic attention of The New York Times, The Washington Post, The Los Angeles Times, and the explainer journalism vanguard at 538 and Vox—even though his paper pointed to the deliberate omission of critical data and a cover up at the Centers for Disease Control (CDC).
More shocking still, shortly after Hooker’s reanalysis of the CDC data was published a veteran vaccine researcher at the agency who had worked on the original 2004 study claimed that the omission was intentional and wrong. “It’s the lowest point in my career that I went along with that paper,” William Thompson was recorded telling Hooker. “We didn’t report significant findings.”
The anti-vaccine movement had a smoking grail, a holy gun— and the surreptitious recording of Thompson was released in a YouTube video claiming that this “decade long cover up” was “arguably far worse” than Tuskegee, the notorious 40-year study of syphilis in poor African American men in which public health officials and, later, the CDC, not only failed to reveal the disease to those who had it, but failed to treat them.
Even after Translational Neurodegeneration, the journal that published Hooker’s paper, took it down in late August, pending an investigation, and then formally retracted it on October 4th, citing concerns about the validity of the methodology and statistical methods, the controversy failed to stimulate much journalistic attention.
A piece by CNN reported both sides of the controversy but provided little guidance to readers as to who was right on specifics; a few days earlier, its user-generated news feed, iReport had featured a piece headlined in flamethrower: “CDC Autism Whistleblower Admits Vaccine Study Fraud.” The iReport received 50,000 Facebook shares; the regular CNN news story, which ping-ponged between he said-she saids in an attempt to give each side its due, 23,000. Time expended much of its meager post on the retraction explaining what had been retracted rather than why, while an earlier piece on William Thompson’s claim provided more detail as to why it might be false, but not evidence as to why it was false.
There were two notable exceptions, the first by Russell Saunders (the pseudonym for a pediatrician) in The Daily Beast, which speculated on the errors in Hooker’s paper, and ventured that the whole thing was a “tempest in a teapot.” The second was a more substantial reported piece by health reporter Sydney Lupkin at ABC News. Even though Lupkin just summarized the statistical problems in Hooker, the headline suggested argued that against treating Hooker’s study as if it was just more anti-vaccine nonsense to be ignored: “How a Now-Retracted Autism Study Went Viral—Again.”
It wasn’t just the anti-vaccine movement’s homespun media that turned Hooker and Thompson into a viral phenomenon: the story also attracted the attention of conservative media outlets precisely because it had a news hook that freed the reporters from validating Hooker’s claims:
“It is highly unusual, if not unprecedented, for a sitting CDC senior scientist to blow the whistle on alleged scientific misconduct involving a study article that he co-authored,” wrote former CBS investigative reporter Sharyl Attkisson on her website. “In this instance, the impact of the charge is magnified by more than a decade of allegations from autism advocates who say the federal government and pharmaceutical interests have worked to downplay or hide associations between vaccines and autism.”
An article on Breitbart put it this way: “ a respected CDC researcher has alleged that the agency intentionally withheld important data from a critical study and then cited that study in testimony to Congress.” It ends by telling readers that there is reason to doubt the CDC assertion that vaccines are safe.
The Examiner ran several whistleblowing-focused articles, and a piece on Lew Rockwell.com criticized the media silence as evidence of complicity in “possibly the biggest government whistleblower revelation since disclosure of the Tuskegee Experiment in 1972…”
A post by the indefatigable surgeon blogger David Gorski, one of a several that covered the Hooker-Thompson story in great detail, argued that, actually, it was fortunate the news media largely ignored the story, and that, perhaps, journalists had learned that they “shouldn’t give undue credence to cranks.” One could agree if that coverage would have turned out to be more “on the one hand/on the other” reporting, shorn of any meaningful, which is to say statistical, discussion of the evidence.
But the one-two punch of a peer reviewed study and a credible whistleblower had just given the anti-vaccination movement an enormous boost. Thompson’s statements seemed to confirm that government was manipulating scientific data, a robust meme on the right thanks to climate change; they seemed to confirm longstanding claims of a government-CDC-pharma conspiracy on the left (promulgated by Robert F. Kennedy Jr. and resuscitated in his new book which was published in August); and, perhaps worst of all, they came after a summer of screw ups at the CDC involving anthrax contamination, flu viruses and, notably, the agency’s stumbling response to Ebola. As Joe Nocera noted in the New York Times, these mistakes threatened to undermine trust in the CDC and turn it into “just another federal agency that can’t get it right.”
In this light, does it seem wise that, with the partial exception of ABC News, no news source systematically analyzed the statistics in Hooker’s paper, even though doing so would have provided an objective way to decide whether the claims about an increased risk for African American boys and a CDC cover up were valid?
The entire controversy over vaccination, and its consequences—diminishing vaccination rates—has been produced by people refusing to accept “trust me, I’m a scientist and here’s the consensus” statements, people who by all epidemiological accounts are well educated, people who genuinely worry about their kids; and people who are trying, on some level, to be scientific. The Hooker-Thompson conspiracy can only buttress their fears by framing a complex argument about statistical manipulation with a well-worn and powerful journalistic narrative of government whistleblowing and missing data.
Who knows how the idea that the CDC conspired to suppress data will play out in the viral marketplace of half-baked, half-analyzed claims, but it will gain ground if people believe the agency is compromised by either politics or ineptitude. Who knows, too, what William Thompson was thinking when he made his statements? But who cares? We have statistics; we just need to look at them.
In commenting, please note that many readers are not professional statisticians and scientists. We’d very much appreciate you taking the time to unpack, as best as possible, any technical arguments.