Un-retraction in the John Lott affair: along comes Mary

If you’ve been following the John Lott fracas, most of what follows will be old hat. The rest of you, gather round and listen closely. It’s a complicated story, and it’s also a character test. If you find it sad, you have a fine tragic sense and a good heart. If you find it funny, you should be ashamed of yourself, and I want you on my side in a political knife-fight.

Background: Lott, who made the famous (and vigorously disputed) finding that states that liberalize the rules for issuing permits to carry concealed weapons experience reductions in crime as a result, has been accused of fabricating a survey about “defensive gun uses.” Lott claimed to have found that 98% of incidents where people used a gun to defend themselves involved brandishing only, rather than firing or actually wounding. Other surveys found much higher rates of firing, and the sample size Lott asserted was much too small to support the precision implied by the 98% claim. Lott also claims to have lost all his data, kept no financial records, and forgotten the names of the students who allegedly helped with the purported survey.

Based on what I had read, I opined that some time ago that (1) Lott was probably making it up and (2) if so, that ought to count as a very serious — career-ending — instance of scientific misconduct. [Previous post and links here. Tim Lambert keeps a running box score.]

Imagine my dismay, then, when someone came forward who recalled having been one of the interview subjects for the survey in question. Lott’s original accuser, James Lindgren (the same scholar who delighted the gun-rights folks by demonstrating the fraudulence of much of Michael Bellesiles’s Arming America), interviewed the person and found him convincing. Lott’s defenders were relieved, and announced that the controversy was over.

What a revoltin’ development! If the survey had actually taken place, I had done Lott a serious injury by asserting that the balance of probabilities was against it and drawing personal conclusions about him (albeit tentatively) on that basis. It still seemed to me, as it does to others, that the 98% claim was unjustified and that Lott must surely have known that, but publishing a dubious interpretation is a long way from making up data.

It was put to me in an email that I owed Lott a retraction, and I procrastinated, trying to decide whether to eat my crow with salt & pepper only or try to dress it up with a sauce bordelaise about confidence intervals and Lott’s curiously variable account of where the 98% number came from in the first place.

Then Atrios (oh, come on, you didn’t really think you were going to get through a story like this without encountering Atrios, did you?) dropped the bombshell. Not only was the supposed interviewee a gun-rights activist (which Lindgren knew when finding him credible) and former NRA board member, he was the practitioner of an especially sleazy trick by which a gun-rights group appropriated the name of a gun-control group that had neglected file its annual report on time, thereby (unbeknownst to its organizers) losing its charter and leaving its name free for anyone to grab.

That fact, combined with the fact that no other survey respondent, and none of the workers, had come forward, suggested to me that the survey question was open again. Someone willing to cheat — and it seems to me that the maneuver described, which everyone seems to agree actually happened that way, amounts to cheating, despite its legality — might well be willing to lie (which, after all, is also perfectly legal, as long as you’re not under oath and not selling anything).

Then came the part that’s either very sad or hilarious, depending on your enlightenment level. Julian Sanchez, who like many others had more or less concluded that the survey was real, noticed a coincidence of IP addresses between Lott himself and a woman named Mary Rosh. Rosh had vigorously defended Lott, and sharply attacked his critics, in a variety of Internet forums (Lott having told various people that he personally never engaged in on-line controversy of that sort), had offered warm praise of his teaching performance, and had written an extremely enthusiastic review of Lott’s book on Amazon.com. (“SAVE YOUR LIFE,READ THIS BOOK–GREAT BUY!”)

After a little back-and-forth between the pro-Lott and anti-Lott sides about what could and couldn’t be inferred from IP addresses, Lott acknowledged that “Mary Rosh” was his pseudonym, and that he had been making the posts himself. [Tim Lambert offers a selection of “Mary Rosh” comments, with links.]

Some of Lott’s defenders have tried to make the issue here Lott’s use of an internet pseudonym. By itself, that is certainly inoffensive, though a really scrupulous honesty might insist on the use of an obvious pseudonym or anonym: Publius, Atrios, Concerned Citizen, or simply “John Smith,” with the quotation marks. But Mary Rosh went way beyond pseudonymity: “she” made statements that her non-existence except in Lott’s imagination made contrary to fact, such as that she had taken Lott’s course and found him fair-minded. I find reading “her” stuff, knowing that she is just Lott’s persona, rather creepy. This isn’t academic dishonesty, since it doesn’t involve data or results, but it’s plain old ordinary dishonesty of a fairly garish hue.

And the Amazon.com review in particular is very near the line that separates ordinary lying from legally actionable, and perhaps even criminal, fraud. The claim that the book might “save someone’s life” is obviously silly, since it’s about the statistical effects of legislation, not the mechanics of self-defense. But by posting it in a false name, Lott conveyed to readers (it’s the lead review on Amazon.com, by the way) the impression that some real person had read his book and found it highly valuable, when that was not the case. That the review heading ends “GREAT BUY” makes the commercial context clear. Of course, no prosecutor would be bothered with such a petty fraud, but it’s really infra dig. for someone who’s supposed to be in the truth-telling business.

Moreover, “Mary Rosh” makes it easier to believe that Lott would not only invent a non-existent survey, but also coach a witness to confirm its existence. And even if, contrary to what now seems most likely, the survey turns out to be a reality and not a fantasy, we’re still left with Lott as the inventor of “Mary Rosh” and the publisher — not once, but many times — of that unjustifiable 98% claim.

As I’ve noted before, you don’t have to believe the extreme Lott claim that liberal concealed-carry laws actually reduce crime to support such laws. The strongest case against such laws is that they will tend to increase crime in a noticeable way, and that doesn’t seem to be true, either. If liberalizing concealed-carry laws has any effect on crime, either way, that effect is too small to be obvious statistically. [See John Donohue’s chapter in the forthcoming Brookings gun-policy book.] And after several years of several hundred thousand Texans being allowed to carry concealed, the Violence Policy Center, trying its hardest, can’t come up with much in the way of actual gun violence committed by that group of people. [Note to VPC: you don’t need a gun, or a license to carry one concealed, to drive drunk, so counting drunk-driving incidents by concealed-carry-permit holders doesn’t really tell us anything interesting about the laws in question.]

So here’s my plea to my gun-rights friends in academia. If you admit that John Lott lacks the basic honesty required for citizenship in the Republic of Science, that doesn’t mean you have to give up your guns and join the Brady Campaign. You’re perfectly free to believe in an individual-rights interpretation of the Second Amendment and oppose gun registration. So you ought to be willing to call this one on its merits.

The inclination to circle the wagons against a co-partisan under attack is a strong one. But Lott is now beyond reasonable defense. Those who were most vigorous in pursuing Bellesiles, and most contemptuous of those who kept defending Bellesiles as the evidence of his duplicity mounted, ought to be the first to say “Enough is enough, already.”

Cut your losses, fellas.

UPDATE: The story keeps getting weirder, and sadder; now Lott is blaming the book review on his thirteen-year-old son.

Author: Mark Kleiman

Professor of Public Policy at the NYU Marron Institute for Urban Management and editor of the Journal of Drug Policy Analysis. Teaches about the methods of policy analysis about drug abuse control and crime control policy, working out the implications of two principles: that swift and certain sanctions don't have to be severe to be effective, and that well-designed threats usually don't have to be carried out. Books: Drugs and Drug Policy: What Everyone Needs to Know (with Jonathan Caulkins and Angela Hawken) When Brute Force Fails: How to Have Less Crime and Less Punishment (Princeton, 2009; named one of the "books of the year" by The Economist Against Excess: Drug Policy for Results (Basic, 1993) Marijuana: Costs of Abuse, Costs of Control (Greenwood, 1989) UCLA Homepage Curriculum Vitae Contact: Markarkleiman-at-gmail.com