One of the most damning bits of evidence in the Lott controversy had been that the 98% figure for defensive gun brandishings he said he found in his 1997 survey had been previously attributed, not just to “national polls” (consistent, if misleadingly so, with his intending the poll he’d conducted) but also to FSU gun scholar Gary Kleck. The idea was that perhaps Lott had invented the 1997 survey to cover for his mistake in reading Kleck to support the 98% brandishing figure (as some others had done). Except now it appears that maybe the only instances in which Lott apparently cited Kleck as the source for the number weren’t Lott’s doing after all.

The two references were in versions of the same article which were published by the Independence Institute and Heartland Institute, and the second was basically just copied from the first. And Independence’s David Kopel now says that he pasted in the mistaken reference to Kleck while editing the piece. Here’s what he said in an email Lott forwarded to Tim Lambert:

I’ve got no specific recollection of editing the piece, but the evidence seems to indicate that attributing the 98% figure to Kleck was an error by the
Independence Institute, rather than an error on the author’s part.

and then later he wrote to Eugene Volokh‘s mailing list:

Since the piece is several years old, I don’t have a precise recollection of the editing process, but I do dimly recall adding the attribution. The description of Kleck as a “Florida State University criminologist” is a common phrasing of mine.

Now, I was somewhat concerned that what this meant was something like: “Lott tells me he didn’t say that and I took his word for it, so it must’ve been my mistake,” an impression strengthened somewhat by an email in which Kopel wrote:

I always knew the editing error must have been mine, since I edited it. I supplied more details at Lambert’s request. Studying the particular
phrasing, before my second email, clarified my recall, because of the Kleck locution. [Emphasis mine]

Still, I searched both Google and Lexis-Nexis, and found that while Kopel often uses the “FSU criminologist” wording, Lott never has, as far as I can tell. Also, Kopel does now seem to genuinely recall making the addition, and I just don’t believe that he’d lie about it.

Mac Diva thinks I’m naive to put any stock in this, but I don’t know. Kopel has a scholarly reputation to protect, regardless of whatever personal or ideological affection he might have for Lott. And if the number didn’t come from Kleck, where did it come from? Either Lott just pulled it out of thin air, or there actually was a survey. Now, there’s no knockdown evidence one way or another, but this most recent development inclines me to think the latter is the more probable hypothesis. It’s still not entirely kosher to cite such an unusual result (and not mention all the much lower numbers) when you’ve lost the data, and certainly not when it was based on a uselessly small sample. But it’s not the same as fabricating a study that didn’t happen.

[UPDATE] Kopel adds:

Today I searched the web to figure out why I would have thought that the 98% came from Kleck. The answer: Kleck’s seminal 1988 article in Social Problems (which was a very major source of data cited by pro-rights folks like myself for many years thereafter), said that 98% of DGUs don’t involve wounding.
When I saw the 2 million DGU figure in Lott’s article, I added an attribution to Kleck, and I mistakenly thought that Lott was also citing Kleck’s 98% figure, and I re-wrote the sentence to attribute the figure to Kleck.
My error, or course, is that Kleck’s 98% is non-woundings/killings, whereas Lott’s entirely different 98% is non-shootings. Kleck’s data showed a very large fraction of DGUs to involve misses or warning shots.