unSpun_ Finding Facts in a World of Disinformation - Brooks Jackson [47]
But tragic as slow death from ALS certainly is, the story of the three celebrity football players is a bit like our tale of the great crow fallacy. An EPA epidemiologist who headed a team to study the matter, Patricia A. Murphy, said the ALS-Milorganite connection “has been blown up out of some groundskeeper’s imagination.” She found no evidence of an increase in ALS in the Milwaukee area or Wisconsin as a whole, and concluded that “the anecdotal stories linking ALS and Milorganite are purely that, i.e., anecdotal with no basis in scientific fact.” Dr. Henry Anderson, Wisconsin’s state environmental epidemiologist, agreed. He reported that scientific evidence against Milorganite was lacking and recommended that the state not give high priority to further study. The three 49ers might not even have been exposed to Milorganite. No records were ever found to support the groundskeeper’s recollection. Even if the three players had been exposed to Milorganite, they appear to constitute what statisticians call a random cluster, meaning an unusually large grouping that occurs purely by chance. Epidemiologists are trained to look at the whole picture without being misled by such coincidental clusters. Sooner or later, if you toss a coin enough times, it will come up heads ten times in a row. The odds of it coming up heads on any given toss, however, are always 50–50. And the odds that Milorganite had anything to do with ALS are somewhere between very low and zero.
Mitch Snyder’s “Meaningless” Numbers
Real evidence, unlike attention-grabbing anecdotes, is generated by systematic study. But not all studies are created equal, and some hardly deserve to be called studies at all. Consider a claim made in 1982 that more than 3 million Americans could be homeless. It was widely reported as an estimate that 3 million were homeless, and it was widely believed. Critics of the Reagan administration saw the figure as evidence that the Republican president’s policies were creating a social calamity on a scale not seen since the days of the Dust Bowl and the Great Depression. Had 3 million been an accurate number, it would have meant that one American in every seventy-seven was living on the street.
In fact, the figure was close to an outright fabrication. The source was Mitch Snyder, a former advertising man and ex-convict (he served a federal prison term for grand theft, auto) who had turned himself into an advocate for the homeless and an unrelenting detractor of Ronald Reagan. His methods were hardly scientific. In 1980, he and others at the Community for Creative Non-Violence had called up one hundred local clergy, city officials, and others involved in aiding the homeless in twenty-five cities, asking them for a quick estimate of the number of homeless persons in their locality. But an estimate is not a count, and some estimates are more reliable than others. For one thing, CCNV rejected estimates that Snyder deemed too low. On this highly subjective basis, the group concluded that one percent of the entire U.S. population, or about 2.2 million people, were homeless at the time of the survey. Then, after a recession hit the economy, CCNV said, “We are convinced the number of homeless people in the United States could reach 3 million or more in 1983.” But that was nothing more than their guess piled on top of a dubious “average” of their cherry-picked estimates.
While testifying before the House Subcommittee on Housing on May 24, 1984, Snyder was challenged to support his estimate; he all but admitted that he had pulled the number from thin air. He said: “These numbers are in fact meaningless. We have tried to satisfy your gnawing curiosity