Eugene Volokh posts a four part chorus of support for John Lott from ex-colleagues, all attesting to the veracity of his claim about suffering a 1997 hard drive crash. That would be extraordinarily compelling evidence if it were true that, as one of the letter-writers says, the Washington Post, in their article, “unfairly casts doubt about whether John Lott suffered a hard disk crash on his computer.”
Except the article doesn’t question whether there was a crash; it questions whether Lott lost a survey in that crash. In fact, just about everyone, including Lott’s most dogged critic Tim Lambert, has acknowledged that the fact of the crash is beyond doubt at this point. I remain agnostic on whether the survey was done — though in cases like this, where evidence runs in conflicting directions, I’m inclined to give the benefit of the doubt — but asserting again that a crash happened is clearly beside the point. So why harp on it, especially when the crash clearly isn’t central even in the Post’s cursory account, unless it’s to give the false impression that confirmation on this point constitutes some kind of unambiguous vindication? Ditto comments about how the results have been “reproduced.” I guess that’s true, insofar as the first survey’s measure of “mere brandishing” rates was statistically meaningless, and so is the second’s, but I don’t know what we’re supposed to conclude from those signally unhelpful facts.