unSpun_ Finding Facts in a World of Disinformation - Brooks Jackson [13]
WARNING SIGN: A Story That’s “Too Good”
WE SHOULD APPROACH CLAIMS CAUTIOUSLY WHEN THEY ARE TOO dramatic, especially when we want them to be true. Consider a case involving what we might call “destruction of mass weapons.”
The book Arming America: The Origins of a National Gun Culture was greeted with celebration by advocates of gun control. The author, Michael A. Bellesiles, a professor of history at Atlanta’s Emory University, claimed that household gun ownership had been rare in colonial and pre–Civil War America. Michael Barnes, the then president of Handgun Control, Inc. (now known as the Brady Campaign to Prevent Gun Violence), lauded Bellesiles’s “discovery” and proclaimed: “By exposing the truth about gun ownership in early America, Michael Bellesiles has removed one more weapon in the gun lobby’s arsenal of fallacies against common-sense gun laws.”
If, indeed, weapons ownership wasn’t widespread in colonial America, then the picture of a nation of “Minutemen” with muskets over the fireplace was false. This also meant the National Rifle Association would have less room to argue that the Second Amendment was written to guarantee the right of individuals to own guns privately, and not just to bear arms as members of a regulated militia. Bellesiles offered as proof what he described as a painstaking, ten-year study of 11,000 probate records showing what people owned when they died. Columbia University seemingly endorsed the finding, awarding Bellesiles’s book the coveted Bancroft Prize in 2001.
For those favoring gun control, Bellesiles told a story that was way too good to be true—literally. “The data fit together almost too neatly,” noted Professor James Lindgren of the Northwestern University School of Law. After checking a portion of the same records on which Bellesiles said he had based his conclusions, Lindgren and his colleagues concluded that he had “repeatedly counted women as men, counted guns in about a hundred wills that never existed, and claimed that the inventories evaluated more than half of the guns as old or broken when fewer than 10% were so listed.” In eight different sets of probate records, guns appeared in 50 percent to 73 percent of estates left by men, figures several times higher than Bellesiles had claimed. Lindgren found guns were twice as common as Bibles in estates from 1774.
Other discrepancies were noted; for example, Bellesiles claimed to have examined probate records from San Francisco, but all such records had been destroyed in the 1906 earthquake and fire. Emory University placed Bellesiles on paid leave and asked a panel of outside historians to investigate. The investigators found “evidence of falsification” regarding the “vital” table summarizing Bellesiles’s probate data and said “his scholarly integrity is seriously in question.” He resigned from Emory, still protesting that he was guilty of nothing worse than innocent mistakes. Nevertheless, in December 2002 Columbia University withdrew the Bancroft Prize, saying that Bellesiles had “violated basic norms of acceptable scholarly conduct.”
There was plenty of reason to question Bellesiles’s data. Given how starkly it contradicted what had been accepted for two centuries, it amounted to an extraordinary claim demanding extraordinary proof. Gun-control advocates were too quick to swallow it because it seemed to help their cause, and they paid the price in embarrassment when the facts of Bellesiles’s deception were uncovered. When a claim seems “too good,” it should be a warning to withhold judgment until we get a close look at the evidence.
Data in the Service of Ideology
Extravagant claims are just too easy to accept when they match biases. In 1991, we were told something shocking, which seemed to confirm the view that women are victims of a sexist society. In her book The Beauty Myth, Naomi Wolf claimed that 150,000 women die annually from anorexia nervosa. This was a preposterously high number, more than five times the number of Americans who died of AIDS that year, for example. But it