LOTT UPDATE: So is the Lott affair over, or not? A few days ago, it looked as if it had been laid to rest, but people are still talking. I’ve been waiting for Jim Lindgren to do a wrapup on this, which he was going to do last week but which he’s now promising for sometime next week. Lott critic-in-chief Tim Lambert summarizes the whole affair here, and while his dislike of, and distrust of, Lott is quite evident, he’s got all the links pro and con. There is now substantial evidence that Lott did in fact conduct a survey in 1997 — some factual corroboration and one person who says he was surveyed, which seems to be enough to satisfy most people, if not all. And certainly there’s no evidence presented, beyond inference and questions, that Lott didn’t conduct the survey. If that changes, I’ll certainly let you know.

But for now, the surest sign that this issue is largely settled is that Lambert is now arguing a different question — not whether the survey was conducted, but whether the 98% figure is accurate. I have no opinion on this at all. I’m not qualified to judge the statistical merit of this stuff, and at any rate, the question originally presented wasn’t whether Lott is a poor scholar, but whether he’s an honest one.

I’ve been uncomfortable blogging on this subject because I’ve been involved in trying to get to the bottom of it, though chiefly in the form of repeated and forceful admonitions to Lott to make as much information public as possible, as soon as possible. Lott has released his income tax data and all of the information on his forthcoming survey replicating the 1997 study to scholars for examination. They seem satisfied. (Interestingly, Lott, like Lambert, seems to think that the real question is whether his figure is ultimately right. They’re both wrong, in my opinion). Unlike the Bellesiles affair, where I was on the outside wondering why it was getting no attention, here I was (somewhat) on the inside — enough, at any rate, that blow-by-blow blogging felt wrong, especially on side issues like email pseudonyms.

I’ve been a bit annoyed by the efforts on behalf of many to make this into a Bellesiles-payback case. First, the Bellesiles case was in the email-list and scholarly inquiry phase for over two years. When I first blogged it, on October 3, 2001, the Bellesiles case had already been the subject of extensive detective work by Clayton Cramer, investigative reports in the Boston Globe and National Review Online, and over a year of back-and-forth on the same email list where people have been discussing Lott. My first post was, in fact, occasioned by Emory University’s demand to Bellesiles that he explain himself. I hadn’t blogged the issue earlier because it seemed premature; people were still looking into the matter.

As I said in my first post on this subject, even Lambert stated up front that this question didn’t call Lott’s main argument into question. Bellesiles was accused of, and eventually shown to have engaged in, outright fabrication of major data crucial to the essential argument of a major published work. Lott was accused, and not shown to have engaged in, false reports of conducting a study that was never published anywhere. The Bellesiles process went on for two years. The Lott process, by that standard, has taken place in an eyeblink. It’s also notable that Tim Lambert wasn’t ignored or dismissed in the way that Clayton Cramer was for years, and that some of those (including me) who have leaned hardest on Lott to explain himself are those who generally favor the results that his work shows. That’s in notable contrast to the Bellesiles case, too.

Greg Beato, who sometimes takes it upon himself to lecture me on fairness and decorum, has demonstrated his commitment to fairness and decorum by photoshopping Lott in drag and conflating Bellesiles’ false claims that a critic had forged emails attributed to him, with Lott’s use of a pseudonym in chat groups, two rather different things, on the dubious basis that both were “Internet-related.”

Lott has not covered himself with glory in this matter, and the pseudonymous-posting thing is kind of weird (though, um, certain bloggers are not in a position to criticize pseudonymous argument too much, and raising it after the main claim seemed to have been laid to rest seemed a bit cheesy to me). And I think that Lott’s reputation will suffer from all of this, and it probably should. But the desire of many people to have a Bellesiles-payback-on-the-cheap has done them no credit either. Accusing an academic of fraud is a serious matter, best done by those who — like Clayton Cramer, or Jim Lindgren — have done actual work, and have actual evidence relevant to the matter at hand. That’s one reason why I’ve waited on Lindgren, since everyone seems to agree that he’s honest, and he can hardly be accused of wanting a Bellesiles payback. And when Lindgren posts his findings, I’ll report those, of course.

UPDATE: David Levy, an economist at GMU, emails:

I’ve known John Lott for a long time and he’s been really good about data sharing. I require students to replicate published worked in my econometrics classes and one of them had the guts to get data from John. “Guts” only

because the data set is huge. It came on a zip disk (if I remember correctly) and probably in Stata format. No one at George Mason was using Stata then so it was a mild pain to get it converted.

People reviewing their own work and neglecting to sign their name has a long, wonderful tradition. One of the best reviews of Walt Whitman’s Leaves of Grass is by … yup … WW himself.

Levy also edits the excellent Library of Economics website.