More on Defensive Gun Use
It probably won’t surprise you to learn that the NRA and the gun lobby have profited immensely from the most recent school massacre du jour. It probably won’t surprise you to learn that they always profit immensely from mass shootings. And you’re probably familiar with the fact that part of their strategy involves whipping up hysteria in their constituents about how President Obama (or fill in the Democrat of your choice) wants to “take away your guns”. What you may not realize, however, is that during the past election season while they were publicly demonizing the president for his supposed hostility to supposed “Second Amendment rights”, they were privately rooting for his reelection so they could ratchet up the paranoia and the profits even more. Nice work if you can get it.
You’ve probably heard about Wayne LaPierre and others declaring that the solution to gun violence is… well, a hair of the dog, natch. (Actually, they’re likely to insist that “gun violence” is merely a semantic contrivance of the “gun control” advocates — that “there is no such thing as gun violence, only violence committed with a gun”. You just can’t make up this kind of stuff.) They have an ample supply of cutesy slogans to promote their panacea, but their primary mantra is “more guns, less crime” — which is even the title of a popular book by soundly discredited gun guru John Lott. This slogan expresses their belief (or at least the belief they want their clientele to maintain) that guns are used more often for defense than for crime.
Even if that were perfectly true, employing it as an excuse to put more guns into circulation is rather like saying we should start more house fires because more people will collect insurance than will die. But there’s no evidence their premise is correct; in fact, the evidence strongly suggests that the reverse is true. The best they can come up with is statistics that seem to show a correlation in certain localities between stricter gun laws and higher crime and/or looser gun laws and lower crime. (In reality, such correlations aren’t nearly as clear-cut as they assume; and in any case, correlation is not the same as causality. But that’s a topic for another day.)
Gunsters often tout “studies” (i.e., surveys) that supposedly prove that there are hundreds of thousands, or even a few million, defensive guns uses (DGUs) every year. But not even all of these surveys support the “more guns, less crime” motif. Two of the best known figures, both projected by the Dept. Of Justice, are 108,000 and 65,000 DGUs. Both of which fall far short of the documented 400,000-plus crimes committed with a gun annually. No wonder the gun culture has such a love affair with Dr. Kleck, who posits 2.5 million. He calculated this by interviewing just 222 individuals. Furthermore, as stated by the Virginia Center for Public Safety:
Kleck’s Interviewers do not appear to have questioned a random individual at a given telephone number, but rather asked to speak to the male head of the household. Males from the South and West were oversampled. The results imply that many hundreds of thousands of murders should have been occurring when a private gun was not available for protection. Yet guns are rarely carried, less than a third of adult Americans personally own guns, and only 27,000 homicides occurred in 1992.”
I’ve discussed some of the problems with these surveys in two previous posts, Make My Day: Mention Gun Defense Statistics and Estimating Defensive Gun Uses Reasonably. Since then, I’ve heard from a great many gunsters who really, desperately want my analysis to be wrong. Which is not surprising; if we burst the bubble of “more guns, less crime”, the very foundation of the gun culture starts collapsing. Some of them are obsessed with trying to establish that I’m a “liberal”, which one gathers would instantaneously vaporize all the inconvenient facts. But not one of them has been able to provide any evidence that my “theory” about the true number of DGUs (as if I actually had one) is mistaken. Except for saying something like “I’ve had 27 DGUs of my own, and none were reported by the media”. But such anecdotes, even if perfectly true, hardly contradict anything I’ve said. The gun culture insists upon the existence of a legendary beast of titanic proportions, but the only evidence is the most miniscule of footprints.
Potential vs. Probable
Mostly, these folks just argue in a circle: “The numbers produced by the surveys must be accurate because there are surveys that produce them”; or “I believe these figures are right because they seem reasonable to me.” They also tend to confuse potential with probability. Their reasoning is that if we know the number of gun owners in the U.S. and the number of annual crimes committed, then we can make an educated guess about the probable number of DGUs. Not so. There are many other factors that can play a role. Many of these factors we do not know; some, indeed we probably will never know. But two of the most obvious are: (1) The criminals usually have the element of surprise on their side, and (2) Few gun owners are armed at all times. A third factor that’s not so obvious, but which nonetheless has the potential for a significant impact is that humans have a documented tendency to compensate for an added safety measure (e.g., a gun) by indulging in riskier behavior, so the net level of danger remains the same. Which might help explain why some alleged DGUs, on closer inspection, turn out to be aggressive rather than defensive.
Mind you, these are just some of the major factors. And even a number of very minor factors can make an enormous difference in the final product. Which might help explain how Kleck’s 222 interviewees mutated into 2.5 million. It’s called the butterfly effect. Look it up. Or see the movie.
The Hawthorne-Rambo Effect?
The overriding fatal flaw of those DGU surveys is that, contrary to what they purport, they really don’t even attempt to determine how many DGUs really occur. Instead, they attempt to determine how often participants say they occur. And they obviously have difficulty succeeding even at that task, as evidenced by the extremely wide range of results. Gallup, for instance, conducted a poll in 1991 in which it concluded that the annual DGU count was just under 800,000. Two years later, the organization conducted another poll on the subject, and came up with a total more than twice that high! And we’re really expected to take such estimates seriously? If so, which one?
Dr. Kleck and others poo-poo the notion that interviewees in these studies might have been, um, less than truthful. While I won’t go so far as to say that they outright lied — well, at least not all of them — it’s clear that their responses are grossly distorted. And we don’t have to cast about a great deal for a scientific explanation. It might be a variation of the experimenter effect, epistemic feedback, the subject-expectancy effect and/or the Hawthorne effect — all of which involve the researcher somehow influencing the subject’s response. In addition, this particular topic of inquiry invites the subject to regale the researchers with tales of his derring-do, which openly invites embellishment, whether deliberate or inadvertent. (See Prof. Hemenway’s commentary for more factors of distortion.)
In preparing to write the previous posts, I scoured several gun-friendly websites that attempt to do their own tracking of defensive gun use, inviting their followers who’ve experienced one to submit their accounts. Among others, I combed through more than 50 years of records compiled by the NRA. But none of these sources ever racked up a tally of more than a few hundred per year. Why the gargantuan discrepancy between these totals and those supplied by the scientific surveys? The likely explanation lies in the difference in how the data are collected. In the surveys, researchers contact individuals directly and put them on the spot to deliver narratives of their heroism. The tracking websites maintained by gun communities, however, put out a general call; and perhaps only those individuals who’ve had a genuine experience are likely to respond. I’d also submit that perhaps these individuals feel bound by some unwritten code of honor to be truthful and accurate when dealing with an organization they belong to and/or respect. Moreover, they might fear, and perhaps rightly so, that their narratives will be subject to corroboration.
Virtually all of these latter incidents also were reported in the media. This, of course, does not prove that all DGUs that occur appear in the media, nor even that most of them do. But it does make you wonder why, if these (as often alleged) are only the tip of the iceberg, can’t millions of gun owners cough up more personal accounts to put on these websites and keep the public better informed about them. In any event, whether or not most DGUs really are covered by the media, most — are at least a very large percentage — are newsworthy; i.e., they would be reported in the news given the right circumstances. The less newsworthy such an event is, the less likely it is to be a bona fide defensive gun use. The media love this kind of story, and they’ll seldom just ignore it if it comes to their attention. And a great many genuine DGUs are difficult to keep secret even if one wanted to. Bear in mind that it’s irresponsible of the defender not to report the encounter to police, given that there’s a criminal running around who might attempt the same crime on someone else with tragic results. Of course, there is sometimes good reason why the alleged defender might keep the episode under his hat: he might, for example, be in possession of a firearm in violation of the law. But isn’t that all the more reason to question the defensive nature of the “defense”?
It’s particularly difficult to hush up those incidents in which a gun is fired; and doubly, triply, quadruply so for those instances in which the assailant is killed or wounded. For one thing, concealment itself would be a serious criminal offense. With the latter group, you can be virtually assured that such a sensationalist episode will make the news.
And here is where those surveys actually might be of some value. According to Kleck, 8 percent of the defenders wound or kill their assailants. This figure is certainly too low, given Kleck’s extremely loose standards for what constitutes a DGU. But even so, 8 percent of 2.5 million would mean it happens about 200,000 times per year. And yet only about one in 400 of these is reported in the news? Seriously? The National Crime Victimization Survey says it happens 3 percent of the time out of 108,000 DGUs per year. That’s 3240 in which the offender is wounded if not killed. And yet fewer than one in six is deemed newsworthy? Get real. One might argue that the media would have neither the capacity nor the interest to cover 200,000 such incidents per year. Perhaps not. But they would definitely have both the capacity and the interest to report 3240.
The most comprehensive listing of DGUs I’ve found is at keepandbeararms.com. Looking at the 75 most recent incidents listed (which covers a period of about 2 months), I see that, as best I can determine, there were 22 offenders killed, 38 merely wounded, and 29 were neither — some incidents involved more than one assailant. (And by the way, these incidents include “Man pulls gun on rowdy, line-cutting Black Friday shopper” and “Woman pulls gun on man who exposed himself at lake”. Very defensive, no? Note also that at least one of the stories is listed twice, and at least one actually details a case of unarmed self-defense — the defender had guns on hand but chose not to use them. How many other “DGUs” could be prevented if more people used their heads instead of their trigger fingers?) If these proportions are typical — and further research would be needed to declare that they are — then we could conclude that attackers wounded in a DGU die more than a third of the time. (In fact, this is quite consistent with a larger sample I examined in a previous post.) So if the NCVS is correct, then at least one percent of DGUs should involve the death of the suspect — which would amount to more than 1000 per year. But FBI statistics confirm that this only happens no more than about 300 times. Therefore, it seems that NCVS estimates are inflated by a ratio of more than 3 to 1, and thus the actual sum, by its own standards, should be about 25,000 to 30,000 DGUs per year. But bear in mind that this includes an unknown number of false positives. The Kleck survey would yield nearly 75,000 deaths, which would suggest it’s inflated by more than 250 to 1, which works out to fewer than 10,000 annual DGUs.
In fine, there is substantial proof of only a few hundred DGUs per year, and even inferential evidence of no more than a few thousand. The surveys that profess to demonstrate far in excess of that are, to say the least, highly unreliable. Crimes committed with a gun almost certainly outnumber gun defenses, probably by a large factor. And it’s staggeringly naive at best to propose that the remedy for gun violence (or whatever euphemistic circumlocution you choose to apply) is even more guns.
(NOTE: This post was revised on 1/7/13 to correct a couple of minor errors and insert a new paragraph of further explanation.)