January 11, 2003
From the desk of Jane Galt:Bellesiles Again?
It looks like John Lott, the author of
More Guns, Less Crime, may also have fabricated data.
It's not quite on the scale of the Bellesiles fabrication: it looks like he made a stupid remark about having done a survey in order to deflect criticism of his using a Gary Kleck number indicating that 98% of gun uses do not involve a shooting; Kleck's number included shots that missed, or were warning shots, but Lott and others said it applied to merely "brandishing" the weapon. The distinction is a fine one, and doesn't overturn Lott's major or minor theses (unlike Bellesiles fabrications). Nonetheless, his story sounds fishy to me: he did a survey without funding, kept no records, and can't remember the name of a single student surveyor? If he can't substantiate, he should be slammed for it, and right quick. Fabricating data is fabricating data, even if the data only replicates data that's known to be good.
Of course, the left will jump on it and say it's just as bad as Bellesiles. I've no doubt that many will have some interesting reason why it's actually worse. That's silly. Bellesiles made up data and egregious mis-cites were his entire book, and the data were the central reason he got the various awards and prestige; Lott's assertion is a stronger version of another study that was itself pretty strong, and the distinction is fairly trivial. It's also a relatively minor contribution to his overall thesis. Nonetheless, scholars cannot tolerate other scholars who make up data. If it's true, embarassing as it is, the gun-rights folks can make this even more embarassing for the gun-control folks -- by showing that they, unlike the gun control people, do not succor and protect people who make up data to support their case. Just as every time the Republicans punish leaders who are guilty of wrongdoing, it hurts the Clinton Democrats on the stump, gun-rights people can wring victory out of defeat by refusing to countenance unethical activity in their ranks.
Posted by Jane Galt at January 11, 2003 03:05 PM | TrackBack | Technorati inbound linksI have no idea about this evidence fabrication thing, but after reading this a while back, I'm not sure why anyone takes his arguments seriously.
Posted by: Jason McCullough on January 11, 2003 06:17 PMPresumably part of the suspicion comes from both Kleck and Lott coming up with the same 98% figure. However, I would guess that defensive brandishing of guns far, far outnumbers defensive shootings. So, Kleck's figure for mere brandishing may well be in the 90%+ range. This makes it plausible that some other survey might have arrived at 98% for brandishing without shooting.
Posted by: David on January 11, 2003 06:27 PMIf you pull a gun and the bandit goes away, that is effective use of a gun. I suspect that the figure would be in the high 90's.
Posted by: Gene 6-Pack on January 11, 2003 06:58 PMLott has a new book coming out in which he lists another study which finds roughly the same thing.
If the second study proves out, then, as a practical matter, we can presume the first was legitimate.
Or he was an awfully lucky guesser.
Jason:
(and others, who don't see Lott's response: http://www.daviddfriedman.com/Lott_v_Teret/Response_to_Lambert.htm
My problem is that he's got even more of an agenda than Lott ("I've spent quite a bit of time on talk.politics.guns arguing with pro-gunners") and set out specifically to disprove Lott - and goes about it presuming fraud.
"Lott's diagrams do not show crime rates at all, but rather plot two quadratic curves that he fitted to the data. It is no surprise that there is a peak when the laws were passed--this is one of the few places where it is possible for the fitted curves to peak. "
I also would note that were he correct about the "increases" in crime after carry laws passed, I'd expect to see that all over every evening news, and front page in the papers. (Or else the Vast Right Wing Conspiracy is meeting again).
So yes, Jason, I still take Lott seriously.
I only wish we could get the Hollywood idiots subjected to this kind of scrutiny for their dumb comments. (wait, that would take more time than we've got)..
Addison
Posted by: Addison on January 12, 2003 09:42 AMIf it's the same Tim Lambert I (and David Friedman, Hal Varian, and Stan Liebowitz) debated on usenet's sci.econ, late last year, over Liebowitz's book, "Re-Thinking the Network Economy", then I'd suggest anyone relying on him, think again.
Look for the threads titled, "Liebowitz Unchained" and "In His Own Words" (Nov-Dec, 2002).
Posted by: Patrick R. Sullivan on January 12, 2003 03:07 PMDavid--According to Lambert, Kleck's 98% includes all missed shots and warning shots as well as brandishing.
So unless missed shots and warning shots are extremely rare, Lott's 98% for brandishing is higher than Kleck's data would suggest.
Richard--If it's plausible that Lott fabricated a survey once, it's plausible that he'll do it again. So, if Lott was in charge of adding up the numbers in the new survey, it doesn't prove anything one way or the other.
Posted by: Matt Weiner on January 12, 2003 09:08 PM> So unless missed shots and warning shots are extremely rare,
Why would they necessarily be common?
US police claim that the typical cop fires his/her gun outside the range once in a career. I suspect that they pull said gun far more times, and they rest a hand on a visible gun even more often.
None of the with-gun/not-police self-defense incidents that I know of involve shots fired and that seems to be the common case for other folks that I've talked to. Scientific? No, but it's enough to make me ask for empirical data when someone insists that it is unrepresentative.
Posted by: Andy Freeman on January 12, 2003 09:50 PMSince the burden of proof has been dropped on me, here's C.D. Tavares:
"A) Kleck says in 'Social Problems:'
'there were about 8,700-16,600 non-fatal, legally permissible woundings of criminals by gun armed civilians' annually, and 'the rest of the one million estimated defensive gun uses, over 98% involved neither killings nor woundings but rather warning shots
fired or guns pointed or referred to.'
"So I certainly screwed up by not including the warning shots fired, plus the times the victim simply missed. From Kleck's latest survey as reported here, the warning shots and misses are very significant (~14%)."
There's more at the link.
And Lambert:
Kleck's 98% "brandish, warning shot or miss" estimate was based on an indirect (and very rough) estimate of the number of defensive woundings. Subsequent research on defensive gun use allowed more direct estimates, and estimates of how often defenders fired their gun.
[ugly-looking table removed to below]
While there is some variation, the lowest number is 21%, and anyone who claims that only 2% of defenders fire their weapons is strongly contradisted [sic] by these surveys.
[ugly looking table]
Survey Percent firing Source
Kleck 24 Kleck 1995
NSPOF 27 Duncan 2000
NCVS 1987-1990 28 Duncan 2000
NCVS 1987-1992 38 Rand 1994
NCVS 1992-2001 21 NCVS online analysis system
Field 34 Kleck 1995
Cambridge Reports 67 Kleck 1995
DMIa 40 Kleck 1995
Ohio 40 Kleck 1995
The link to Tavares comes from Lambert.
(The link on "Lambert" seems to be FU not quite BAR; Google "Lambert Duncan Kleck Lott" and hit I'm Feeling Lucky.. At the moment the site is down, anyway. I copied the above from the Google cache of the page.)
Posted by: Matt Weiner on January 12, 2003 11:13 PMI looked at Lambert's page and, if what he says is true, it does look pretty fishy. If Lott 'fesses up and admits the original survey never happened, he'll at least have more integrity than Bellesiles, who has never admitted anything, despite reams of evidence against him.
Posted by: Ross on January 13, 2003 06:39 AMComments are Closed.