Well, as my old journalism profs used to say, this one appears to have acquired “legs.” Arthur Silber blogs an interview with Lott on the Larry Elder show, though by that account it seems like the main issue raised there was the use of the pseudonym, which on its own isn’t that big a deal, even as the various uses to which that pseudonym was put appear increasingly sketchy. Slate’s Chatterbox has a broader report that (rightly) identifies the pseudonym business as a “bizarre side issue.” It also reminds me of two things I’d thought of a little while ago and then buried deep in my subconscious because I told myself I was done with this story. Here they are.

(1) David Gross says he got a tape of this lecture in January of 1999 at the Minneapolis Athletic Club. He says that during this talk, Lott referenced the 1997 survey, and it was then that Gross knew he’d been one of the respondents. Well, that’s perfect. Because the first reference to the survey we’d previously been aware of was in May of 1999, in response to a query Otis Duncan had sent back in January. (As you’d expect, the background info is all on Lambert’s page.) Now, if Lott was telling audiences, unsolicited, about the 1997 survey well before that — and maybe even before Duncan first started asking questions — we’d have very strong reason to believe that the survey couldn’t have been fabricated as a response to those questions in an attempt to save face. Moreover, we’d have evidence, not in the form of an uncertain recollection, but in the form of an audiotape! So we now have a pretty straightforward means of testing our hypothesis, assuming DG (or, for that matter, anyone else who attended) kept the tape. If I’m misreading, and it wasn’t a tape of that lecture he’s talking about (though I don’t know how else the tape would help to pinpoint the date), it still should at least be possible to find others who attended the same talk. As far as I’m concerned, that would be all the evidence that we’d really need that the survey took place.

(2) Lott apparently told Tim Noah, as he previously had the Washington Times, that he “couldn’t reconstruct the survey.” But he also told Jim Lindgren that he had tally sheets from the survey, and that these were lost when he moved to another school. Well, if you’ve got the tally sheets, my admittedly limited understanding of how these things work would be that one could just get an undergraduate to spend a week punching that data back into a computer, just as it had originally been entered. Since Lott was at Chicago until 1999, when he moved to Yale, why exactly couldn’t he reconstruct the survey at some point in that two year period? I don’t mean to ask this with a “j’accuse!” tone, but it is the kind of thing one would like clarified. Possibly the answer is just that he moved his office or home within Chicago soon after the survey.

And broken record time, because this part seems to keep getting missed: a precise percentage claim on the basis of a really tiny sample size is still misleading. And a repeat of that result with another equally small sample wouldn’t count as “confirmation.” The best case scenario here is that Lott really did a survey, not that it was legitimate to make the 98% claim on the basis of that survey.

Update: I’m wrong about the tape, as David Gross makes clear in an email.

At the time of his talk in January 1999, I was unaware that the survey reference was to HIS (John Lott’s) survey, as he didn’t say it was his in the talk; just
that I had participated in the survey he had referenced. At the time of that communication on the morning of the 19th, and even later during the conversation with James Lindgren, where I recalled having purchased the tape, I couldn’t quite recall exactly what it was that prompted me to make the comment to John Lott after the talk. But I DID know that I had commented to Lott that I had been a respondent in the survey he referenced. […]
He does not, anywhere that I could find in the tape, say that he was the one who had developed the figure and its nuances. But it was clear to me, then, whoever did conduct the survey, that I had been a respondent in that survey; and I mentioned it to John in passing.

So scratch that route to confirmation. If anything, it now seems weirder that Lott didn’t mention that he’d conducted the survey even after Gross identified himself afterwards as one of the respondents. I’ll add that Gross sounds quite lucid and credible, and I feel a little rude asking for some independent confirmation in addition to his recollection of being a respondent, but given his stake in the question, it would be nice to have something more. Lindgren’s report, which should be ready soon, may have more developments.

Update 2:You know what? I was partially wrong about the IP address thing — apparently that one IP does cover a bunch of Comcast users in a certain service area. Given the other indicators, it still would’ve been a heck of a coincidence, and I still would’ve posted more or less the same thing, but the people who raised that objection were right about it. Irrelevant now, of course (especially since “Rosh” also posted to Usenet from an AEI IP address) but figured I should cop to the error.

Update 3: I cannot even begin to tell you how disturbing I find it that I’m now providing zingers for Paul Begala.