Estimating Defensive Gun Uses Reasonably

By now, we hope to have laid to the rest the absurd notion that there are 2.5 million defensive gun uses (DGUs) annually, or anything even close to it,  in the U.S.  But that still leaves unanswered just how many there really are. Is it possible to make a more reasonable approximation than those suggested by the various “studies” that indicate at least 65,000? Indeed it is; and it possibly could be done using real numbers – i.e. actual statistics – rather than survey results.

As we noted, there are at least 200 DGUs per year; we know this because there has been a rather consistent listing circulated during the past few years, of media accounts of such events. But we also noted that, which has reported between 200 and 300 per year for the past decade or so, has recently shifted gears. In recent weeks, the rate of occurrence has been around 1 or 2 per day. Assuming the trend continues and they’re all genuine and non-duplicated, 2 per day would add up to a whopping 730 annually. Better buy another Uzi, ma.

So does the higher incidence lately indicate  an increase (perhaps temporary) in frequency, or does it just mean that the team of volunteers at has become more diligent in its detective work? There’s reason to believe the latter.

There’s another website called Armed Citizen (not the same as the one affiliated with the NRA) that is currently on hiatus, but which compiled DGU stories from 2003 to 2010. This site was co-founded by Clayton Cramer, who’s connected with the prominent right-wing “think tank” The Cato Institute.  I laughed out loud when I saw the interactive map of DGUs he’s now compiling. Supposedly millions of the suckers every year, and he expects to indicate them all with dots on a map? They’d be layered so thick that even at maximum magnification they’d be spilling out of the computer screen onto the floor.

Actually, he’s a bit more modest than most gun propagandists, averring only “tens of thousands” per year. But even that is going to be hard to pile up on a map, especially since he’s depicting several years’ worth.

This Armed Citizen site does have a rather extensive archive of  “nearly 4000”. Which isn’t quite “tens of thousands”. And oh yes, it’s not just for one year – it’s for the entire period of a little over 6 years. Which comes out to between 600 and 700 per year. Since “tens of thousands” suggests a minimum of 20,000, then Cramer evidently believes that at least 96 percent of these incidents go unreported. The interesting thing is that his estimated total is far lower than the Kleck-Gertz total, yet he is suggesting that a much higher percentage of them slip under the media radar. Go figure. (Somebody certainly needs to.)

Still, it does appear that (assuming these incidents are all genuine non-duplicates) there are 600 or more DGUs reported in the media annually, and is just catching up.

Now according to the Kleck-Gertz figures, 36 percent – roughly a third – of such incidents go unreported. I don’t buy that, of course; one reason is that Kleck and Gertz are extremely loose in their standards for defining a DGU. A defensive gun use means that someone uses a gun to prevent a crime – or, okay, an animal attack. It does not mean a pissing contest that one person settles with a firearm. If a guy breaks into your house and you greet him with a shotgun, that definitely qualifies as a DGU. If you get into an altercation with a guy over a parking space and he becomes disproportionately aggressive to the point of threatening violence and you pull out a gun, chances are that qualifies as well. If you’re arguing with a guy, or even having a fistfight, with more or less equal ferocity and you whip out your Luger just to get the upper hand, that probably does not qualify.

But since we have no other number to use, let’s take it. If there are about 600 reported DGUs, and that’s about two-thirds of the total, that would give us 900 or so. Call it an even grand if you like.  We’re still short of 2.5 million, and we’re still outnumbered by gun crimes at least 400 to 1. But at least we’ve surpassed the frequency of lightning strikes.

We still should allow for the possibility, however, that there are other incidents out there that have been reported, but that these websites have overlooked. So let’s cross-check it.

FBI statistics show that for the five-year period ending in 2010, there was an average of 213 justifiable homicides per year by firearm. (A justifiable homicide is not necessarily a defensive use, but the majority apparently fit that description.) The Kleck-Gertz figures indicate that the defenders wound or kill their assailants only 8.3 percent of the time, but this is surely far too low – especially given that many alleged defensive gun uses involve nothing more than mentioning the existence of a gun! And the figures don’t specify how many are fatal. But if indeed there are 2.5 million DGUs per year, then the fatal shootings would account for only .0085 percent!  So let’s just skip Kleck altogether and stick with real numbers.

In a random selection of 200 DGU stories, I found that at least 69 resulted in the death of an alleged offender. (I say at least because I only read the headlines, which may or may not specify death – which often does not occur immediately in a fatal shooting.) If there are 213 such killings per year, and that represents about a third, then once again we are left with between 600 and 700 annual DGUs. In other words, it appears that there is no significant number being overlooked.

My exercise was not a controlled scientific study, just a suggestion or two about how one might be conducted. My figures may not be precise but they are, if not in the ballpark, at least on the right planet. If I had to wager whether the true DGU tally is closer to (a) 2.5 million, (b) 65,000 or (c) 500 to 1000, I’d bet the deed to the Ponderosa on the latter.

AFTERWORD: (2-22-12) In order to get an accurate estimate of DGUs, there are still three questions that need to be answered: (1) How often do the defenders actually report the incident to police? (2) How often do the police follow up on it? (3) How often do the media report the incident? I’m willing to bet that the answer to each question – for genuine DGUs – is “the great majority of the time”. But whether that’s true or not, researchers would be better served seeking answers to those questions and comparing those answers to the verifiable incidents, than in questioning gun owners about how often they’ve brandished their goodies. Even if the answer to each question is only 50 percent, that would mean no more than about 5000 annual DGUs.


  1. First off, for an interesting take on Clayton Cramer see:

    I’m not in any position to cogently criticize Mr. Cramer’s second amendment scholarship, and the general consensus seems to be that he knows his stuff on that score: I’m happy to assume that that’s correct. But based on Mr. Cramer’s creative approach to the truth back when I had the inclination to personally joust with him (just to save you a bit of time: the money quote from the above-linked exegesis is where he manages to dig up an article from from Redbook Magazine in the 70s that claims that homosexuals perpetuate themselves via assaultive pederasty; this is presented as ipso facto proof the subseqent scholarship on the subject is a coverup), I would suggest that if Clayton Cramer claimed that the sky is blue, you’d probably want to rent a spectrometer and double-check to be sure.

    Anyway, given that there would be some form of injury, if not death from actually using a firearm. Perhaps even a police report due to the fact that a firearm was involved since some innocent bystander might misunderstand what happened, you’d think there would be a lot more of these reported in officialdom, not just the press.

    As this lot like to say, they would prefer to be judged by 12 than carried by six.

    Unfortunately, the 12 jurors might disagree with their version of reality.

    • Blah blah blah. More endless reference to “studies” (polls), and more polarization (“a darling of our opponents”). I still prefer to stick with real numbers. As I’ve demonstrated, we can expect about 650 media reports of DGUs each year. That is only ONE PERCENT of even the SMALLEST total projected by the “studies”. Those huge numbers would require that a huge percentage of DGUs be suppressed by each of three filters standing between them and the media: failure of the defender to report to police, failure of the police to follow up, and failure of the media to follow up on the police follow-up. Kleck and Gertz claim that the first filter occurs about a third of the time -which as I’ve stated, is a tremendous exaggeration, since so many incidents their subjects called DGUs were really not. But even if that figure is accurate, and even if each of the other three filters also stops about a third, that would still give us only about 2000 DGUs. In order to obtain the fabled total of 2.5 million, we would need a scenario in which, for example, each filter suppresses about 93 percent!

      • Yes, because the media is SO good at reporting news that helps out conservative causes… I personally know 2 people last year that used their firearms defensively (Both by brandishing their weapons) to stop crimes. Guess what? Neither of them reported their actions. Why? Because it would have been nothing more than a hassle to deal with police paperwork. If you’re relying on the media to furnish your statistics, you’re nothing more than a fool and a shill. Nice try, bub, but you’re so poorly informed that you should shut down your blog, sell your Macbook, and go join a hippie commune. Statistical failure with a VERY poor baseline premise.

      • As a matter of fact, the media are VERY good at reporting (and concocting) news that helps “conservative” causes – while convincing the gullible that they have a “liberal” bias. But that’s really beside the point, since the issue here is not sociopolitical ideology, but gun defense. As I already stated, the Kleck-Gertz survey (which the gun culture loves to tout as the ultimate authority on the matter) projects that about two-thirds of the defenders DO report their cases to police. And if those incidents really are self-defense against crime, they’ll probably end up in the news. This means that if Kleck and Gertz are correct, about a third DON’T get reported, so your two buddies, even if they really were acting in self-defense, hardly contradict this proportion. Review my comments about “filters” while I retreat to my commune for a while. (Macbook??? Get thee behind me, Satan!) I might add that, whatever the actual figures, if there is indeed criminal intent then it would be highly irresponsible NOT to report it; and we know that gun owners are all very responsible individuals who always do the right thing, don’t we?

      • “As a matter of fact, the media are VERY good at reporting (and concocting) news that helps “conservative” causes – while convincing the gullible that they have a “liberal” bias. ”
        The media have covered up issues like Fast and Furious and did so for a year and a half. Only one or two major media outlets let information out against Obama and his cronies. Hardly necessary to ‘convince’ anyone of any liberal bias when the media has gone out of its way to crucify conservatives and completely overlook liberal criminal behavior.

      • For an issue that was “covered up”, I certainly did hear a great deal about it. Not only that,, but I heard a great deal about the bogus right-wing claims that it was a major scandal of Watergate proportions.

      • “For an issue that was “covered up”, I certainly did hear a great deal about it. Not only that,, but I heard a great deal about the bogus right-wing claims that it was a major scandal of Watergate proportions.
        Did anyone die in Watergate?
        I guess human life isnt much in the eyes of your sort.

      • Ding, ding, ding! Straw man alert. People die all the time as a result of presidential decisions, but that doesn’t mean the president is directly to blame. The real question is whether President Obama is guilty of malicious, unethical or unlawful actions, as Nixon was. And the answer, contrary to the ceaseless spin of the “liberal” media, is no.

    • Yes, there are no doubt quite a number that don’t make it to the media, which is why I make allowance for them. But I would bet that THEIR numbers are greatly exaggerated as well (particularly when so many incidents that are reputed to be defensive gun uses turn out to be not so defensive on closer inspection). Many people would read the comments on a blog like this and see your claim of a DGU and think, “okay, if one or two people who comment on this blog have had DGUs themselves, then surely that’s only a tiny slice of the total”. Not necessarily so. It’s more likely that those who have had DGUs (or believe they have) are greatly overrepresented on the Internet because they are attuned to opportunities to get their story out there.

      • “quite a number” indeed. more like ‘most’ dont make mainstream media.
        I have to admit,however, that as concealed carry has become more accepted that the media does seem to be actually covering more and giving a fair presentation of facts than they used to.

      • I really doubt if “most” don’t. But even so, in order for the bogus DGU numbers to be accurate, the number of unreported incidents would have to be more than “most”; they would have to be a hugely overwhelming majority — at least 90 percent or so. And that’s just not realistic.

      • I’d say your conclusion about the rate of reporting at each step is unbelievably flawed. From the DGU’er’s point of view, depending on the jurisdiction, reporting a defensive use may, itself, be admission of a crime.

        As I have quite a few friends and family in law enforcement, I’m at least aware of what the police blotter looks like in my city and, as well, what it looks like in the city where my eldest son attends college (and became a victim of a gun crime that went completely unreported in the press). If the papers had to report everything in the police blotter, including every instance of a homeowner racking a shotgun and saying “Leave now, I have a gun! (and maybe some pot and paraphernalia that I’d rather not have to hide from the police)”, then we’d be out of ink.

        But, your much vaunted press spent years perpetuating the myth that domestic violence calls (and one presumes occurences) increase up to 40% on Superbowl Sunday. Now we know that the statistic was completely made up, but to the folks responsible for reporting the news, it was perfectly reasonable and their response was to raise their reporting on domestic violence ORDERS OF MAGNITUDE around Superbowl Sunday.

        That said, there’s certainly no reason why there would be a DOWNtick in the occurence of Domestic Violence, which is interesting in this context because it’s all too common and, yet, the assumption at least in the press is that it is vastly underreported, as are most violent sexual crimes. Now, assuming CDC numbers are correct, they estimate between 960,000 and 3 Million incidents of domestic violence per year in the US. Let’s make it an even million for discussion’s sake. These million incidents occur in a country with an estimated 250 million privately owned guns. Criminologists expect that offenders tend to be repeat offenders and that victims often tend to be repeat victims. So, let’s say 25% of these victims have been abused by the same abuser in the past. 250,000 incidents a year where the victim knows it’s coming and you posit that fewer than 600 have the sense to grab a gun and say “No.”

      • I don’t believe that those incidents in the news are the only ones that occur, nor even necessarily most of them. But there needs to be a base of confirmed incidents from which to make a calculation. (Police blotters, incidentally, might also be a good source to extrapolate from.) And the claims made by the surveys (“studies”) are far from confirmed. And one can’t make assumptions about the rate of defenses versus the rate of crimes.

  2. Your theory of only four defensive gun uses per state per year is simply unbelievable to me. When I lived in California I was beneficiary to two DGU’s in one year (not my gun, but I was benefited just the same by the guy who DID have it). Neither was reported for the very reasons Kleck suggests — he probably shouldn’t have had it; but then again, neither should have the poachers that were hunting us in the mountains. All’s well that end’s well and nobody got reported.

    • I always hear from the exceptions, but the same questions remain: What exactly made your incidents defensive gun use? How typical or atypical were they? What segment of the genuine unreported DGUs are represented by those who, like yourself, come forward and make it known online? I have yet to see any evidence that my rough estimates are very far off.

      • Okay, I guess your comment deserves a bit more detailed response.
        The figure of “four per state per year” is based on the absolute minimum that can be verified; and as I’ve indicated, that’s almost certainly too low. The more realistic estimates I’ve given would be in the range of 20 to 30 per state per year. But that’s an average, and it will vary greatly. California, being the most populous state, will surely have far more than Alaska or Montana. Even so, I’ve spent nearly 20 years in San Francisco and Los Angeles, and I don’t recall ever encountering anyone who’d experienced a DGU. If they occurred with anywhere near the frequency claimed, there’s no way I could have avoided hearing about NUMEROUS instances.
        My theory, if you want to call it that, hinges in part (though by no means entirely) on the very loose definition of “defensive” that is generally applied. It’s not defensive if you use a gun to settle a barroom brawl, and it’s not necessarily defensive if you use one to chase a trespasser off your property; yet such incidents are often included in the tally.
        The litmus test I’d propose is this question: In the absence of a gun (or other means of forcible defense) would an innocent person have incurred harm or loss of property? If the answer is yes, then it was genuinely defensive. If no, then probably not.
        When you examine it closely, you find that there are relatively few justifiable DGUs in which the assailant did not also have a gun. Now there are roughly 400,000 crimes successfully committed with a gun each year. Let’s say that there is an equal number of crimes using some other type of weapons which also pose a serious threat (quite unlikely). That would make 800,000 in which the use of a gun would have been justified. According to Kleck and Gertz, at least THREE TIMES that many were PREVENTED by a gun. This is just plain nuts, especially when you consider that fewer than half of the intended victims will even have access to a gun, much less have it on hand at the moment and use it successfully. I could believe, in other words, as many as 400,000 DGUs per year, though it would be quite a stretch. But 2.5 million is fantasyland.
        I cannot guarantee that my figures are in the ballpark, but the evidence is that they are much closer than most of the “studies”.

      • “Now there are roughly 400,000 crimes successfully committed with a gun each year. ”
        Which only proves one thing…that law abiding citizens are victims 400,000 times a year counting JUST gun crime alone…and they should absolutely have a tool to defend themselves from these scumbag felons.

      • Actually, it proves many things. And assuming your point is correct (and I won’t argue with it), that doesn’t necessarily mean that the “tool” needs to be yet another gun.

      • “Actually, it proves many things. And assuming your point is correct (and I won’t argue with it), that doesn’t necessarily mean that the “tool” needs to be yet another gun.”
        Actually it does.
        Law enforcement has figured that common sense conclusion out, yet Brady fans seem to think that armed felons are run off by harsh language and dirty looks for some reason.
        Ive tried all that…it doesnt work….seeing a gun aimed at their faces, however, is a very SURE way to give them the understanding that they should run.
        And if it doesnt actually shooting the dirtbags is an acceptable alternative to a deterrent.

      • Having a gun also creates many violent situations that otherwise wouldn’t arise. The real debate is whether these risks are worth the risks one runs of being a victim of crime. And I’m not at all convinced that it does. I am convinced, however, that those individuals who like to refer to perpetrators as “dirtbags” or similar labels are generally not people that I feel comfortable about being armed.

  3. I have been involved in 6 DGU’s, all of which were reported to the police. Only 2 of the incidents made the newspaper, and one just stated that the thieves assaulted the homeower, (me) and stole tires and parts off a jeep in the backyard, and that they were apprehended based on fingerprints left at the scene. I only found out about the fact that the person who attacked me had been in prison for killing someone else because I had a friend on the police force. The other one that made the news was the one time out of the six that I had to fire. That was reported as “…a passerby disarmed the gunman and held him for police.” No mention of me being armed.

    • Wow, sounds like you live an exciting life. What can I say but repeat what I’ve already said? Having 6 actual DGUs is HIGHLY atypical. I seem to hear from the exceptions rather than the rule, which is quite understandable. If these incidents all occurred at or near your home, I’d REALLY consider moving, now matter how inconvenient.

      • “It does indeed. But there are certain places where it is FAR more likely to happen.”
        Ive come to accept the reality, having been the victim a few times of crime, that when crime is happening to you personally you really dont care about the odds anymore. After youve been a victim you know that it “more likely” just doesnt matter anymore…you begin to understand the necessity to ALWAYS be safe and ALWAYS have a tool that might save your life….ALWAYS be prepared.

      • I think POP is being very realistic in his numbers and anyone can come here and say they had a DGU… Hell i live in Miami and leave my back door open and I live just across the highway from a “bad area”… I do have however two chows! 🙂 Many people here “pack heat” and there have been many instances of “road rage” where if people DIDNT have heat there would have been no shootings…so the easy access to the guns actually can INCREASE the number of fatalities in an area without much gun control.

      • Quite so. I’m thinking that I will do a follow-up to this post, since so many people have misconceptions about my argument.

  4. Ive had 4 times where I was legally allowed to use lethal force and had a gun on my person.
    Between 2 of your posters here thats a total of 10 defensive uses of firearms.
    I’d say your math is a bit off.

    • Possibly. Or else the posters’ concept of what constitutes a DGU is a bit off. Or else they are outliers rather than the norm. There are many possibilities.

      • Or…..there are just excuses being made by anti gunners who dont want to admit the usefulness of a gun that the cops have managed to figure out 😉

      • The usefulness of a gun is not the issue. They sure as hell are useful, no doubt about that. “Anti-gunners” are generally just concerned about when they should be used and by whom.

  5. “Actually, he’s a bit more modest than most gun propagandists, averring only “tens of thousands” per year. But even that is going to be hard to pile up on a map, especially since he’s depicting several years’ worth.”

    Actually, seeing that most people dont carry guns its simply a math issue. No gun, no defensive use of a gun. Easy enough to comprehend?

    As more and more law abiding citizens have started buying and carrying guns the number of defensive uses has risen.

    Heres what we know as absolute fact.
    Most of the USA now has concealed carry with more and more law abiding people carrying firearms…and yet FBI statistics show that violent crime has been steadily DECREASING….including yet again in 2011.

    While we cannot prove that carry laws have caused the decrease in crime we DO conclude two absolute conclusions from the data.
    1. Guns havent caused an increase in crime, and,
    2. Gun packers HAVE defended themselves quite a lot with their firearms….thus we make the easily concluded observation that hte law abiding citizen carrying a gun is an ASSET to the individual and hardly any detriment at all to society as a whole.

    In fact, guns can also be used to take down those intent on committing mass killings…

    • You are correct that violent crime has been decreasing. But that trend has been going on much longer than the recent gun hysteria. And there is no evidence that the number of DGUs has increased. I’ve already been over that.

  6. “Even if the answer to each question is only 50 percent, that would mean no more than about 5000 annual DGUs.”

    So what, exactly, is the point of your article….that those ‘only’ 5000 potential victims that possibly saved themselves from being raped and or murdered are expendable since you claim its such a low number?

    You must have had some agenda in writing this article…and it certainly doesnt come off as being pro gun.

    • My point is that the claims about the number of DGUs have been vastly overestimated. And my agenda is to demonstrate that. I thought that was rather obvious.

      • Im sorry but there has to be an underlying reason for even bothering with the topic. In and of itself it is rather pointless. Who cares how many defensive uses there have been when in the overall picture we see that law abiding citizens carrying guns hasnt been any issue at all, nor have crime rates risen because of our carrying guns.
        If ONE law abiding person saved themselves with a gun then that ONE is enough justification for us to carry guns.

        Here it comes…I hear the wheels turning….”but what about how many people are killed by criminals with guns every year compared to that one life that was saved?”

        Ah…so we DO have an underlying motivation beyond just proving lesser numbers in defensive uses of a gun in writing this article, dont we?
        Youre simply comparing illegal use to legal use and trying to convince someone, possibly yourself, that if more crimes are committed with guns than lives are saved then its a fair tradeoff to let those defensive users be defenseless victim in order to (supposedly) make the death tolls smaller.

        if only it would work out that way….
        It wouldnt.

      • To put it as simply as possible, the number of guns is proportional to the number of deaths and injuries caused by guns. I know, difficult concept, huh? But that’s one reason we need the perspective of more accurate numbers that relate the benefits of an armed populace. Another reason for questioning the bogus numbers is that this is a blog about propaganda, and the DGU numbers are propaganda. And it would be wise to question why the gun lobby and right-wing mouthpieces push these claims so hard.

  7. You are vastly wrong, a DGU is anything from displaying your weapon because ofa threat to actually shooting to defend, the vast majority of DGU’s don’t get reported because nothing really happened, firearm displayed, threat disappeared, most don’t want the hassle of reporting it to the police. How do I know this? Because I’ve been a police officer for 30 years and I am quite familiar with this topic..

    • The sticky word in the term “defensive gun use” is “defensive”. In many purported DGUs , the alleged defender was being, in fact, another offender. I don’t know what proportion of the total this represents, but I have read many accounts in which this was the case. In other incidents, though the gun operator was on the defensive, it was also clear that the use of a firearm was not necessary. So when you try to analyze how much of an impact guns might have on self-defense, you really need to consider only those cases in which they were both necessary and defensive. And that number is much smaller than the gun lobby would have you believe.

      • That’s the final blow to all this nonsense about large numbers of DGUs. Once we have a final figure, we have to reduce it by a certain percentage to allow for all the false ones. Considering human nature, with its penchant for self-justification and rationalization as well as the mentality of gun owners, I would say that percentage is rather high.

  8. Well, Mr. P.O.P., I don’t really know where to start. I guess I would give you a B for effort because I admire the use (or at least attempted use) of critical thinking skills – a talent that is not applied very frequently in our society. I especially like testing all of the various studies by trying to fit the numbers into the real world and attempting to make adjustments and assumptions to make sense of them. But I am afraid I will have to give you an F on application.

    Let’s start … well I guess with the media. Think for a few seconds about where most violent crime occurs … yeah, in the bigger cities. So while the news outlets in your typical 20k population county seat might be able to cover most of the crime reports that occur in the area at large – it is not even remotely possible for news organizations to cover all of the crime that occurs in cities (and surrounding areas) of larger cities. There are over 200 cities in the US with populations greater than 100k. You are deluding yourself if you think that we can get even a remotely accurate count of crime or DGU using the media as a source. They best they can do is provide anecdotal evidence.

    Next, let’s look at reporting to the local police. As you mentioned, Kleck and others claim that many DGU are unreported – and I think that is true. If you look at when these studies were performed – I think in the early 90’s, the relaxation of gun control was just beginning. Most states allowed ownership of firearms even if they restricted the carrying of arms in public. But laws for using a gun for defense of self and family were still pretty murky. In Texas, for example, until a few years ago – one section of code allowed for defense without retreating while another required retreat before deadly force could be used. Even displaying (i.e. not firing) a gun in the home to fend off a criminal meant putting yourself at the mercy of the prosecutor as to which section of the law they would choose to follow. And, while perhaps you might still win at the end of the day, most of us cannot afford the legal fight. So yeah – I think in the early 90’s if a homeowner scared away someone in the act of entering their home by cycling the pump action of their shotgun – it is entirely plausible that 1/3 of them would have left that detail out of their report.

    But, the problem with local police statistics is not just whether or not the victim reports the use of the gun – but if no shots were fired, how does that get counted? Is there a check-box on the police report labeled DGU without shots fired? I am sure the actions of the citizen get recorded with verbs and adjectives – but how does that reliably get counted in the DGU category? There are 15,000 state and local law enforcement agencies in the United States and so I am quite sure that there are about 15,000 different ways in which incidents are tallied and then reported. It’s much easier to count when shots are actually fired and easier still to count bodies. But it is much harder to define the threshold for a DGU. It’s a lot like sport statistics. We count the things that are easy to count – because they are easy to count. They do not always reflect the abilities of the individual player or team.

    I notice also, that you keep going back to the Kleck-Gertz figures. I guess they are an easy target because of perceived bias. Why do you not attempt to tear apart the 1994 US CDC study that found 498,000 DGU per year? In the comments you correctly list approx 400k crimes per year in the US in which a gun was used (actually, that is the total for violent crimes -assaults, robberies, sexual assaults, rapes, and murders – but we will give it to you). Then you attempt to estimate the number of such that are committed without the use of a gun and show that the Kleck-Gertz figures would mean that DGU’s prevent three times more violent crime than actually occurs. But you don’t have to guesstimate those numbers; the Department of Justice is kind enough to provide them for us. In 2008, there were over 5.3 millions violent crimes in the US (that’s just the violent ones – mind you). So that would put the Kleck-Gertz figures at stopping 1/2 as many violent crimes as are successfully committed. Or, if you add them you could say the citizens prevent about 1/3 of attempted violent crimes via DGU’s. Certainly plausible given that about 1/2 of homes in the US have a gun in them. (Actually, perhaps that means we are slacking off a little – we should be preventing 1/2 of the attempts … lol).

    Ok, ok – I know that is not entirely fair as I am sure there is some overlap. I am sure that some of the assaults that are counted in the DOJ numbers are probably counted in the DGU figures as well. In fact, without the DGU, some of them might have been rapes or murders instead of just assaults – but it would not change the overall 5.3 million totals – so I know that is not fair. But the point is that the Kleck-Gletz numbers are entirely plausible given the true scale of violent crime in the US – which your ‘math’ significantly underplayed.

    Finally, your self-selection of 200 DGU is so obviously flawed that I can’t believe you even attempted to pass it off. Of course the cross sample of DGU headline stories will have a MUCH higher percentage of shots fired and casualties. That’s what makes the news. ‘If it bleeds it leads’ – that’s the saying in the news business. It can’t possibly be a representative survey so any attempted extrapolation is fallacy before it even begins.
    Me personally, I think the 2.5 million estimate is too high also. Most studies I have seen seem to put it at between 500k to 1M per year. I think that is probably the correct range and certainly much closer to the actual ballpark than the badminton court you’re playing on.

    • It’s refreshing to see a gun advocate actually make some good points, but I believe I’ve already addressed and made allowance for them. I noted, for instance, that about 42 percent of Americans have some degree of access to a gun (not quite “half”, but close). But in order to assume that this means they prevent 42 percent of the crime, you’d have to assume, among other things, that these people are armed at all times, and can always use their guns effectively. Sorry, I can’t buy it. I also can’t buy your contention that crime is under-reported in metropolitan areas. (I’ve lived in San Francisco, Los Angeles, Memphis and Detroit, so I have some personal familiarity with this.) Sure, there’s more crime, but there’s also more media; and sensationalism sells. The reason I focus on Kleck-Gertz is that it has become the darling of the gun culture.

      • My reference that DGU’s should be preventing one-half of all attempted crimes was meant to be humorous. I was merely attempting to show that with 5.3 million violent crimes in the US every year (as opposed to the 800k you estimated) – the Kleck estimates of 2.5 million DGU’s become plausible.

        But, as I mentioned, I think true number of DGU’s is likely somewhere between 500k to 1M. That would put the number of violent crimes prevent vs occurred at between 9% – 19%. Now, that sounds just about right, doesn’t it?

        Or maybe that is even a little too low when you consider the 1982 survey of felons which found:
        • 34% had been “scared off, shot at, wounded, or captured by an armed victim”
        • 40% had decided not to commit a crime because they “knew or believed that the victim was carrying a gun”
        • 69% personally knew other criminals who had been “scared off, shot at, wounded, or captured by an armed victim”

        And as far as under-reporting by the media in large cities; lets take a closer look:

        You mentioned your personal familiarity with some cities – for example: Los Angeles. Ok, well the crime rate of Los Angeles is 113,549 violent and property crimes per year (according to

        That equates to an average of 311 incidents per day. When you lived there and watched the news, do remember them reporting over 300 crime incidents each night? When you received your weekly addition newspaper, did you find 2183 incidents listed in the crime blotter? Of course not – it is not physically possible for the media to report every incident in larger cities.

        And you are right that sensationalism sells! That is exactly why the media has not choice but to report the bloody horrors of gun-crime and ignore the ‘yawn …’ another homeowner scared away a criminal by racking the slide on their gun.

        So I am afraid I cannot agree that you have made the allowances you claim you have. Your math simply doesn’t add up. Between the choices you offered up: (a) 2.5 million, (b) 65,000 or (c) 500 to 1000 – I’m thinking you owe me one super-sized Ponderosa.

      • Los Angeles is an unusual case: It’s actually many cities spread out over a very wide area. So yes, it’s possible that all the crimes are reported in the media, though not necessarily the same media. But I doubt if literally all of them are. Many are just not newsworthy enough. Of the statistics for L.A. you cited, the great majority (more than 91,000 out of 113,000) are property crimes. That includes offenses like shoplifting or stealing a lawn ornament — not good candidates for media coverage or gun defense. Furthermore, even many of the more serious property crimes — e.g., auto theft — are committed when the owner is not even present. I extrapolated rough estimates for DGUs from media stories, because very broadly, crimes that are serious enough to warrant gun defense are also serious enough to warrant media attention — in a community of any size. I’m afraid I’ll have to hold onto the ranch until further notice.

      • That retort has some validity. It is true that many property crimes occur to parked vehicles are homes when no one is present. But it still doesn’t quite address the entire issue. Just the violent crime alone averages to 59 incidents per day. Again, it is simply not possible for all of those crimes to be reported via the media outlets. And even if it were, it would not be possible for a single human to be able to digest them enough to get glean an accurate picture of crime (or DGU’s) in the area.

        Again, the main error of you logic is in assuming that: “because very broadly, crimes that are serious enough to warrant gun defense are also serious enough to warrant media attention”. That is simply not not not true. You can say it over an over again, but you would have to suspend logic to do so.

        If you have 59 violent crimes committed daily (where people were hurt or severely threatened) and some unknown number of either property or violent crimes that were stopped non-violently via a non-firing DGU; you simply cannot publish them all. It is simply not possible. Which are you, as the news editor, going to pick?

        Your dodge that Los Angeles is made up of smaller communities and thus the news is spread among many outlets is irrelevant. First, it is not that unusual. Most large cities are that way; Houston, Chicago, Dallas. The crime figures we are using are ones that occur in Los Angeles proper (population of 3.7 million) and not the greater Los Angeles metro area (population 18 million). So the majority of them would fall under the major media outlets in LA. And actually, most media outlets in major cities cover more important (read violent) crimes that occur in the suburbs. So there is even more time pressure to focus only on the sensational.

        And even if it were the case; how would you, a single human, be able to glean data from the literally dozens of surrounding media organizations? Do you drive around and collect all of the community newspapers? Do you regularly go the web of each bedroom community and read through the stories (most of which would require you to purchase a digital subscription)?

        No my friend, I am afraid reality is not on your side of this particular argument.

        And then, you didn’t even address my 9% – 19% calculation. Does that not sound just about right to you? The difference is that your supposed “accommodations” estimated total crime in the US at 800k/yr. But given the number of violent crimes committed is actually 5.3M/yr (not even including property crime) – that makes 2.5M DGU’s plausible – but more importantly – it makes 500k to 1M DGU’s not only plausible but entirely likely.

        You keep saying that you have made accommodations (which I do applaud). But I have shown you how you used erroneous figures as the seed for you accommodations – thereby rendering your conclusions invalid.

        BTW – I have enjoyed this debate. It is nice to finally have a discussion with someone from your side that at least attempts to look at real numbers. We will see if you can stay intellectually honest enough for us to boil this down to the lowest points of contention. Thank you.

      • Least important things first. My point about Los Angeles is not that it is made up of smaller “communities”, but that it is made up of many de facto separate cities . It’s a distinction that may not be so evident to anyone who hasn’t lived there; but some of these cities (e.g. Hollywood, Venice) were once independently incorporated municipalities that were later annexed to L.A. – which may not be unique in this regard either, but what seems to make the City Of Angels unusual is the sheer number of such cities – about 25 by my count – and a correspondingly large number of media outlets, even for such a large population. Thus, it seemed feasible to me that most of the serious crimes could be noted in the media in some fashion. But it’s a minor point.

        My figure of 800,000 that you keep quoting was just a wild hypothesis, for illustration purposes, of the number of attempted crimes with a deadly weapon. It was not intended as an accurate estimate of violent crimes (which could include one person slapping another), which certainly would be higher – in fact, it’s nearly twice as high – the most recent figure I’ve seen is 1.3 million. That’s a far cry from your 5.3 million, which I assume is based on the common assertion that the vast majority of violent crimes go unreported.

        In any event, it seems, we’re straying too far afield here. The real issue is not how many crimes are committed or reported, but how many of them are prevented by a firearm. I don’t know the answer to that, and neither does anyone else – which is precisely why I have been questioning those who claim to.

        Your approach to solving the problem, it seems, is to say that if there are x crimes committed and y people who have guns, then some calculation involving these two numbers alone must produce the likely number of DGUs. You ask whether I don’t consider that logical; my response is yes – but no. There is admittedly a certain intuitive logic to it, but if there’s one thing I’ve learned many times (sometimes the hard way) it’s that the truth can be counter-intuitive. (See the “Monty Hall problem” for a classic illustration. Although I should think it would be apparent that if we all based our beliefs on our instincts, we’d still believe the world is flat.) There are surely many other factors involved than just a gun on one side and criminal intent on the other. What you are really estimating is the (more or less) absolute maximum, which is hardly the same as the probable or even the plausible. It’s especially crucial to question such intuitive assumptions when they run contrary to the evidence — or in this case, the lack thereof.

        Previously I discussed the incident last year in which a gunman strode into a Nevada restaurant, opening fire and killing and wounding several people. Let’s suppose there were 50 diners present (it was a peak time, so the actual total may have been higher.) Statistically, we could expect that 21 of them were gunsters. Yet not a single one stopped the carnage. This is just a slight hint of the hitches we run into when trying to apply your optimistic formula to the real world.

        Yes, I’ve done some calculations involving crime and gun ownership myself, but I was just trying to underscore and expound the real thrust of my argument, which is this: DGU totals must necessarily be speculative; and such speculations stand to be more realistic if they are projected from a foundation of confirmed fact than if projected from a foundation of unsubstantiated claims – as Kleck and Gertz have done. (Am I implying that their interviewees might have provided less than reliable responses? The short answer is, hell yeah.)

        What constitutes confirmed fact? Well, I’m willing to be unusually lenient here. It could include media accounts. It could include government statistics. Heck, I’ll even accept anecdotes supplied by gun devotees’ forums that attempt to track such things – even the NRA, though I don’t regard it as exactly the most forthright of organizations. But all these sources combined manage to come up with only about as many DGUs as the media alone: namely, a few hundred per year. Moreover, many of these stories are covered by media all over the country. Why would a newspaper in Indianapolis devote precious column space to such an incident in Omaha when it presumably has its own ample store of DGUs to draw on? It is factors such as these – including my own personal familiarity with how media operates – that have led me to the conclusion that defensive gun episodes are generally newsworthy. I’m not just making an intuitive guess.

        Yet you have chosen to adamantly take exception to that observation. Well okay. No harm will come to you because of it. But surely you will concur that those cases in which a gun was fired generally make it into the media. No, wait – I’ll more than meet you halfway. Surely you’ll concur that those instances in which the offender is killed or wounded will generally make it into the media. Well, those add up to roughly 300 per year. But let’s continue to be generous and call it 500.

        Now according to Kleck-Gertz, these comprise 8 percent of the total. As I’ve stated before, this figure is surely far too low, because Kleck and Gertz are extremely loose in their definition of a defensive gun use. But I don’t have a better figure to offer, so we’ll use it. If the generously high figure of 500 represents the generously low figure of 8 percent, then that makes the total around 6250. It’s simple grade school arithmetic. People who claim that my “math is off” are not only incorrect, they’re off track. It’s not exactly MY math that has become the basis of dogma.

        Bottom line: extraordinary claims require extraordinary proof; but the only “proof” ever offered for the extraordinarily high DGU counts is that some people say so. Sorry, but I find that quite inadequate. The 2.5 million claim is a legendary beast of titanic proportions, and if it existed, it would leave an appropriately large footprint. But the only footprint in evidence so far is miniscule. Show me the behemoth footprints and I’ll consider it more likely that the beast itself exists. Until then, I must remain skeptical.

      • Man, I think I am really starting to like you. I didn’t think that would be the case when I first replied. But I appreciate your reasoned and sans-emotional responses to what is an extremely complex issue by any measure.

        [Regarding LA – that is very interesting and a larger scale than I was aware. My primary experience is with Houston that has also annexed former townships – but also has a few that they failed to annex that are now islands surrounded by the sea of Houston. But at most I think the annexed and the surrounded probably number in the 10’s.]

        The 5.3 million figure is based upon a composite of the FBI Uniform Crime Report, which is raw data, and the DOJ National Crime Victimization Survey (NCVS) which is gathered from interviews with crime victims and thus does contain extrapolated data. ( So, after further review, there may be more room for questioning than I initially allotted –although I suspect the number is fairly accurate. But we don’t need it to proceed.

        However, you state that my approach is: “to say that if there are x crimes committed and y people who have guns, then some calculation involving these two numbers alone must produce the likely number of DGUs.” But that is not correct at all. That is not my approach. I was merely counting your assertion that, with 800k violent crimes, 2.5M DGU’s is not really plausible (and with those numbers, I would agree). I was showing that with 5.3M crimes – it is entirely plausible and that with 500k to 1M DGU’s it is not only plausible but actually very likely.

        In fact you still seem to be locked in to rebutting the Kleck figures which I have admitted may very well be inflated. Again, the CDC estimated about 500k per year – and they are a government agency with no reason to inflate the statistic in my general direction. And I know we are not allowed to use Wikipedia for grading purposes (lol), but this seems to be a pretty good summation of the state of DGU studies (from

        “The vast majority of defensive gun uses (DGUs) do not involve killing or even wounding an attacker, with government surveys showing 108,000 (NCVS) to 23 million (raw NSPOF) DGUs per year, with ten private national surveys showing 764,000 to 3.6 million DGU per year.”

        So you claim that we are devoid of any evidence as to the number of DGU’s. It is true that we don’t have any hard data – but that is because determining what qualifies as a DGU is extremely subjective. It is somewhat easy for police to count the number of times a citizen fires a gun and easier still to count the number of dead criminals. But it is very hard to determine (and then count) when someone was scared away by something (was it the gun or the light that got flipped on?). So we are not likely to ever get hard data. But we are not devoid of data at all – in fact, there is plenty to be had if you don’t discount it out of hand.

        The Nevada restaurant incident is not an appropriate example at all. As you know, the vast majority of DGU’s occur at home. The figures of gun-ownership *may* begin to factor into home based DGU’s. But the number of citizens who actually carry in public places is just a fraction of the total number of gun-owner’s. So it really has little relevance in looking at a public incident. Again, that is not my formula. As I said in a previous post “My reference that DGU’s should be preventing one-half of all attempted crimes was meant to be humorous.”

        Then you again come back to the media and your expectation that the limited number of events they cover paint a reasonably accurate picture of defensive gun use. Let me answer a couple of your observations/questions individually:

        1. “Why would a newspaper in Indianapolis devote precious column space to such an incident in Omaha when it presumably has its own ample store of DGUs to draw on?”
        As I have stated before, they do this precisely because on that particular day – the story coming out of Omaha is sensational. Perhaps someone pulled a gun in a public place and someone who had a gun defended them. And there may have been 2 minutes of security camera footage associated with it – footage always makes the cut! While on that same day, 11 homeowners in Indianapolis and the surrounding area reported using their gun defensively. 5 of them heard a commotion outside or people banging on a door (or window) and saw people through the peephole and yelled “I’ve got a gun and I’m calling the police”. 4 of them racked their shotgun and the noise scared them away and they didn’t call the police. 2 had someone actually break in the home but ran as soon as they saw the gun. Boring boring boring! If it bleeds it leads – that is the media tag line.

        2. “It is factors such as these – including my own personal familiarity with how media operates – that have led me to the conclusion that defensive gun episodes are generally newsworthy.”
        But that is just the thing. Most DGU’s are very very boring and are not, in fact news worthy.

        3. “But surely you will concur that those cases in which a gun was fired generally make it into the media. No, wait – I’ll more than meet you halfway. Surely you’ll concur that those instances in which the offender is killed or wounded will generally make it into the media.”
        I can concur that they generally make it into the media at some level. I am sure that nearly all of them will make the police blotter in the local newspaper. But that is not the media you are counting – right? Many won’t make the national news or get picked up in web news at all simply because something more sensational happened that day. Some do get picked up, but the facts are not immediately known so it gets reported as a general gun-death or gun-shooting instead of as a potential DGU. By the time more information is available, it is history, old news, as something else has come to suck up the media spotlight.

        So regarding the media: I am sorry, but I think this is one we will have to agree to split on. And this is key because it is not really your math that I find lacking. It is that you are starting with the assumption that you can use the media as the seed for your math – and that is the primary flaw in your calculations. And as we are probably not going to agree on this one point, we may have to call this the primary point of contention unless you can find another method to support your thesis. IMO, using the media as your finger into the currents of either gun crime or defensive gun uses is equivalent to making a decision to purchase your next vehicle based solely upon car commercials.

        You conclude by stating “extraordinary claims require extraordinary proof; but the only “proof” ever offered for the extraordinaily high DGU counts is that some people say so. Sorry, but I find that quite inadequate.” Again, because it is subjective, we will never be able to get hard facts. But it is not just a few people who say so. 2.5 million might well be on the high side, but virtually everyone who has attempted this type of study, from the government to private firms, says that they are orders of magnitude higher than your estimates.

        Why do you think the Brady Campaign or VPC don’t publish their own studies for this particular issue? Do you think they haven’t looked at it? Of course they have – but they don’t like the numbers. So they find studies that ask different questions that support their thesis – ones that count overall gun deaths and the number of people killed by concealed permit holders. I am not faulting them for this – the NRA and other pro-gun groups do the same thing. So in my mind, the lack of contradicting data for this particular facet is proof that the numbers of DGU’s are higher than these groups care to admit.

        Let’s do another thought exercise. I can’t use LA for this because they do have significant restrictions on gun ownership and usage. I wouldn’t know how to account for that. So let’s take Houston. Now, this is not scientific – it is just intended to see if the numbers are plausible:

        The population of Houston proper is 2.1M. The population of the US is 311M. So Houston accounts for about .67% of the population of the United States. Now let’s split my estimates for DGU’s down the middle and call it 750k per year. Assuming even distribution, 5025 of those DGU’s would occur in Houston which would be about 13 per day. Now, the property crime rate in Houston is 117K per year which is 320 per day. Doesn’t it sound plausible, or dare I say even extremely likely, that at least 13 property crimes were prevented in Houston if 320 occur in any given day? And I would argue these estimates are extremely favorable to your side as I would think Texas has higher gun ownership rates than most areas of the country as well as the red-neck willingness to use or threaten to use them! Plus, they only include the property crimes and not the 61 violent crimes that occur daily.

        I would love to hear your reasons why the above thought experiment is invalid in your estimation.

      • Gee thanks, I appreciate being liked, and it happens far too rarely in this endeavor – especially from people who are supporting gun-ness and gun-ocity. It appears, however, that you have become so enamored of me and my musings that you want to hear me repeat myself ad infinitum. You keep parading the same concepts in slightly different garb.

        Of course the Carson City incident was different. They’re all different. That’s the point – each encounter provides a unique set of circumstances that might make it difficult to defend oneself. I don’t know where you come up with the conclusion that “the vast majority of DGU’s occur at home”. According to our old friend Dr. Kleck, the figure is 37 percent; and I’d say that’s probably rather high. (I also wrote a post about how home invasion is, like defensive gun use, vastly overestimated.) But even if your statement is correct, it doesn’t necessarily carry a great deal of weight. Sure, it seems logical that people would be in a better position to defend themselves on their own turf; if you keep a pistol under your pillow, you’re more ready to fire it than any of those people in the IHOP apparently were. But does that necessarily translate into a higher number of DGUs – especially given the relative rarity of home invasion? Maybe, maybe not. We still don’t have hard numbers, and we know that the truth can be counter-intuitive.

        You have (inadvertently, perhaps) hit a bullseye by noting that many Texans have a “red-neck willingness to use or threaten to use” their guns. One problem with guns is that owning them intoxicates many people, who then begin (consciously or not) to seek opportunities to use them. In other words many “defensive” gun uses are not defensive at all, because the use of a gun could have been avoided readily. (Bear in mind that when the “studies” interview a gun owner about an incident, they’re only getting one side of the story.) If a DGU is “boring”, then chances are it fits this category.

        There is nothing wrong with your logic or your math regarding Houston. But what kind of concession are you after here? You want me to acknowledge that the city has the potential for the numbers you estimate? Very well: it absolutely does. And we all have the potential to become millionaires every time we visit Las Vegas. But how often does it happen? There’s a difference between potential and probability.

        Approach it from another direction. Out of (to be generous) 500 media accounts of gun defense per year, we statistically would expect Houston to publish about 2 or 3. But you’re suggesting that this represents only (at the very most) one per 1700 or so – and what you regard as a more realistic estimate is more like one in 3500 to 5000. C’mon, really?

        I’m not suggesting that media coverage is the ultimate verification of reality. But for this situation, it happens to be the only verification we’ve got. Consequently, I don’t see any way to project an estimate of DGUs with any accuracy without using media as the foundation. This is true whether the media accounts represent 90 percent, half, one tenth or one five-thousandth of the total. The thing is, nobody knows for certain which of those is closer to reality. Therefore, people who repeat as fact claims of millions (or hundreds of thousands) of DGUs annually are just parroting gun lobby propaganda. I don’t know how I can state it any more clearly than this.

      • Again, POP, I thank you for taking the time to reply.

        This is an interesting comment as it certainly applies both ways: “You keep parading the same concepts in slightly different garb.”

        For example, you keep referencing Dr. Kleck despite the fact that the issue is settled for me. I think his 5.3M number is likely inflated so therefore I am not likely to take 37% of DGU’s occurring at home at face value either. (I had not heard that figure before; I will take you at face value that it was part of his study.) I believe that the vast majority of DGU’s occur at home simply because that makes the most sense to me. 47% of Americans have a gun in their home while the best guess I can get for Americans over the age of 21 with a CWP or CHL is no more than 5%. It seems reasonable to me to assume that most DGU’s would occur in or at someone’s home. That is also where someone is much less likely to have the opportunity to flee from a crime in progress. But you are right, we don’t have hard numbers.

        “But what kind of concession are you after here? You want me to acknowledge that the city has the potential for the numbers you estimate? Very well: it absolutely does. And we all have the potential to become millionaires every time we visit Las Vegas. But how often does it happen? There’s a difference between potential and probability.”

        Well, I guess that is another area where we differ. When I consider a city with a population of 2.1M and legal access to firearms that has 320 property and 150 violent crimes each day – I would be hard pressed to believe that there are not at least 13 legitimate DGU’s in that same period of time. The numbers are simply staggering – over 450 crimes committed each day – how can 13 of them not be resisted by force utilizing a firearm? So where you see potential, I see probability. This isn’t some table game with the odds stacked in favor of the house (nice try though). This is an effort to use critical thinking to see, not just if something is possible, but if it is likely. With those numbers, I think it is. But again, we can’t actually know for sure because DGU’s are extremely hard to actually count. So you are right again here.

        But then you go back to trying to use media accounts again. As I said before, if you can’t find another justification for you figures then this will be our point of diversion. Using media reports simply cannot work. You say that 500 annual nationwide media reports of DGU’s would be generous. I suspect there are many times that actually reported in print and web media publications (yet still a fraction of the total number). But there is simply no way to accurately scour them.

        You then make this claim: “I’m not suggesting that media coverage is the ultimate verification of reality. But for this situation, it happens to be the only verification we’ve got.” But that is exactly the thing – your guesstimate at the media reports you have managed to expose yourself too is not verification at all – not even close. It is at least as much of a phantom number as what I have put together and that you invalidate as not being a hard number. You wouldn’t build a house using lumber that might be rated at “half, one tenth, on five-thousandth” of the expected load of the house because it would probably fall. Your model is built upon a subjective personal sense of an unknowable quantity of nationwide media reports that capture an unknowable percentage of events with an unknowable degree of “first report” accuracy. It simply cannot stand.

        Then finally you state: Therefore, people who repeat as fact claims of millions (or hundreds of thousands) of DGUs annually are just parroting gun lobby propaganda”. But even with this you fail to acknowledge the fact that the 500,000 DGU figure is not from an NRA puppet – but from the US Centers for Disease Control. I can’t see any motivation for them to exaggerate the numbers in general direction of the gun-lobby. In fact, this may be closest we can ever come to a neutral source for estimating DGU’s and so, perhaps should be the starting point for common ground discussions. You haven’t even tried to explain those numbers away.

      • Little new territory here. Except to note that the estimate provided by the USCDC was, like all the others, based on a survey — i.e., it was projected from a relatively small sample of colorful anecdotes supplied by firearms enthusiasts, under prompting and prodding from researchers who really should know better. Half a million cases per year of intruders scared away with a gun, and virtually none can be verified by media accounts or police reports? Seriously?

      • “Little new territory here.” Correct, you still think your media estimates are accurate (total number of reports) and that they reflect reality (cover most DGU’s). I still think that model is ridiculously absurd.

        “the estimate provided by the USCDC was, like all the others, based on a survey” That is 100% true. But you are going to get a much more accurate picture doing it this way than try to build off of media reports. Why do you think no serious researcher attempts to use media reports as a basis? Because that is not part of any scientific method – you cannot control for all of the variables.

        Again, I thank you for your time and for a respectful discourse. Unless you can provide support for your hypothesis that doesn’t involve magical theories about numbers of media reports, I don’t think we can proceed any further.

      • Agreed. If you think I have a hypothesis of my own, or that I’m touting “magical” or any other kind of theories, then you’re missing the point. My purpose in writing this post was not to provide answers, but to challenge the answers provided by others, and suggest better approaches. Again, media reports are not the ultimate answer — they’re just the only answer at present. But if you can bring me any other evidence, I’d sure as hell love to see it. Not surveys masquerading as studies, not Monty Hall assumptions about probability derived from crime and gun ownership stats, but real evidence of some kind. That’s all I ask for, and despite all my research, I’ve never found it.

      • I understand and I appreciate the fact that you are trying – that’s more than many endeavor when they adopt a stance on an issue such as this one. But I am not the one missing the point.

        I understand you are trying to challenge answers provided by others and trying to suggest ‘better’ approaches. What I have shown (quite clearly, IMO) is that the model you have chosen as your ‘better approach’ is orders of magnitude less reliable than the models you are criticizing with that approach.

        That’s it – that’s the bottom line. Each of your replies to me end with essentially the critique that if it were true we would see a much higher reflection in the media. I have already shown why that is a fantasy. You haven’t been able to provide any other evidence as to why – so we can’t take it any further.

      • Media OR SOMEWHERE. But even if we have only the media, I am amazed that you would consider it less reliable than asking gun enthusiasts to regale us with tales of their derring-do.

      • “I am amazed that you would consider it less reliable than asking gun enthusiasts to regale us with tales of their derring-do.”

        And I am amazed you could possible think otherwise. I understand your concerns about using surveys and possible inflation of the given accounts – but it is certainly not as you describe above. Perhaps you meant that tongue-in-cheek – but conducting surveys is considered a scientifically viable model for discerning this type of information. The cross-section of those polled spans well beyond just ‘gun enthusiasts’.

        But again, you do have a leg to stand on to raise concerns as whether the models are designed and weighted well enough to limit that type of bias. However you are comparing a model/method with some potential bias to one where it is physically impossible to even gather the necessary information. A dozen people working non-stop for weeks to comb through all of the media outlets could not possible give you the data you seem to think you have at hand.

        It is ludacris on it’s face which is why I am shocked that someone of your obvious intelligence is unable to see past it.

  9. “One problem with guns is that owning them intoxicates many people, who then begin (consciously or not) to seek opportunities to use them.”

    (Citation needed)

    Particularly as this seems to be a minor lynchpin of your supposition that DGU reportings are inflated.

    • Actually, I’d call it rather major. And while I’m not aware of any study that estimates just what percentage of gun encounters fall into this category, I’ve already mentioned cases of alleged DGUs that do, and others that appear to.

      • So you’re broadly speculating? Ok just making sure. Try not to state your biased opinion as fact if you want anything but a heavily polarized audience. There are idiots in every group of people, but the idea that guns make you violent is most prevalent among the uninitiated.

        Going off topic: Guns stop being a magical talisman of violence and evil with familiarity. Much like anything else one might fear out of ignorance. Drugs, alcohol, certain races even all wind up demonized in the same way by people who only know about them through the media and urban legends. What most blows my mind is people who are pro drug legalization but anti-gun. If you can recognize that drugs are not inherently evil and are a matter of personal responsibility as to the potential consequences if their use, failing to see the same in any other inanimate object that can be used to harm or enjoy an afternoon with friends is willful ignorance.

      • By golly, you have a point. I really shouldn’t let my personal observations creep in, not even in my replies to other people’s comments, even though they are based on extensive exposure to gun addicts, for a period probably longer than you’ve been alive. I should ignore all the movie case covers depicting those sleek, shiny scepters of power (as mentioned in my post “Of Guns and Glamor”). I should ignore all the people who are obsessed by gun shows and target practice and who have made comments about how they fantasize about using their gleaming ejaculators to take down a commie or a Nigger or anyone else who might look at them the wrong way. I should ignore the dozens of magazines on the bookstore shelf that feature these nifty little pieces of machinery the way Playboy features centerfolds. I should ignore the people who kill each other in disputes over parking spaces, apartment rent, or the dark meat on the Thanksgiving turkey. After all, if they hadn’t had guns, they probably would have killed each other with drinking straws. I should ignore how many accounts I have read of alleged DGUs in which the “defender” was at least as much an aggressor as the “attacker”. I should ignore the gun owners who THEMSELVES acknowledge that they have an addiction to those compact and efficient little implements of Promethean prowess, which they often regard as “better than sex”. And shame on me for comparing them to narcotics. One of them was designed specifically to kill. I’ll try to keep a tighter trigger lock on my willful ignorance in the future.

    • The problem Lars is that a lot of gun proponents think that the gun-control proponents dont have any experience, knowledge or ownership of firearms…that is a logical fallacy…lots of people who own firearms dont believe in unfettered gun and ammo access. I am one. I believe that for Home protection there is a need…but I believe that the people who carry firearms just for their own “protection” wherever they go are the problematic ones…many people here in Miami carry them for thier “enjoyment” (protection)and as a result we have many road rage incidents and deaths.

  10. I began reading your article with interest because I would really like to know the truth, but when you descended into insults directed, I presume, at those you perceive as disagreeing with you, you lost me. If facts exist, they should be presented for discriminating readers to decide for themselves.

  11. Fantastic breakdown of the Kleck studies, thank you for that. I kept hearing the 2.5m number from gun advocates and appreciate your reasoning on the falsehood. I particularly enjoyed your long and civil correspondance with Frail Liberty. That exchange alone makes me hold a glimmer of hope for the future of debate on the internet.

  12. How is 8.3% “surely far too low?” Care to back this up, or are we to reject others’ presumption and accept yours as fact? What sociological or scientific premise do you rely upon? Sorry, but your reasoning is clearly opinion driven and thus no more reliable than the Kleck study.

    • I say it’s too low because, as I’ve pointed out many times before, Kleck is much too loose in defining what constitutes a DGU. As nearly as I can deduce, some 25 percent of his cases involved merely MENTIONING a gun. This is supposed to be a defensive gun use? How about just telepathically warning the bad guys that you’re armed to the teeth? If the total number of DGUs is lower than his estimate, then those verifiable cases in which an assailant was shot will constitute a larger percentage. It’s not opinion, it’s math.

  13. Professor, I think Frail Liberty has shown the importance of applying a standard social science method – the survey – to the DGU issue. I await a thorough debunking of the 1994 CDC study. I don’t consider the CDC sacrosanct, so I wouldn’t be surprised if it could be proven that the study was so poorly designed that it is worthless. But it must be worthless for your case to stand – even a 90% margin of error demolishes your DGU estimates. When you assert that the study had a participant selection bias towards gun nuts, it seems you could be right or possibly you haven’t seriously read the study. This piece of evidence remains on the table, and as of right now, it stands as far more credible than your anonymous blog. Categorically rejecting the survey method is an unscientific attitude that can’t be taken seriously, and it will cost you in the future by restricting you from invoking survey evidence or polling data that could be useful to you. On the other hand, I wouldn’t be surprised if you do refute the CDC study, as the survey method and social science are well known to lack the general level of rigor found in hard science. It would seem more reasonable to call on government and academia to settle this question once and for all through future research rather than to commit to a position that DGUs are a rare event if the available evidence is as poor as you claim.

    If you will indulge me, I would like to shift the discussion in a direction that I consider more constructive if you want to confront right wing propaganda. I will make a rational case for gun proliferation that shows just how seductive guns can be to a society that values freedom. In this argument, it turns out to be rational to exaggerate probability estimates in the absence of “hard numbers” – the precautionary principle is invoked.

    Just for the sake of argument, I accept that the lack of a total gun ban poses a serious net cost to society by elevating gun crime levels as much as 100% above the levels that would exist under a total gun ban, and I can accept that the murder rate is thus elevated by 200%. I wish to make a case for gun proliferation that concedes that gun proliferation incurs a large net cost to society in terms of the balance of its crime increasing and crime decreasing effects. My purpose is to direct attention to the constitutionalist argument that gun proliferation is good because it provides a net benefit to society in the form of insurance against unacceptable government behavior. The constitutionalists claim that this net benefit is so large that it exceeds all costs to society that can be associated with gun proliferation, including the possibility of a nearly third world crime rate. To focus in on the central question I’m getting at, I will assume that the original intent of the 2nd amendment is to prevent the restoration of Monarchy or the establishment of some other undesirable alternative form of government by allowing the masses legal access to military rifles like the M1 Garand that have at most an 8 round magazine. I will also concede a ban on private militias on the grounds that they cannot be “well ordered.” Other sets of assumptions could be explored, but I hope the point of these assumptions will be obvious.

    Constitutionalists make many alarmist arguments for the 2nd amendment, but I will present what I consider a modest constitutionalist argument:

    1) All forms of government bear risks of entering into a transition phase in which the form of government, no longer being viable, begins a catastrophic change to a different form of government. Virtually every instance of large scale government we know of supports this generalization. Arguably, the American nation has gone through several such transitions, so any reasonable probability assignment for the US federal government entering a transition phase within 500 years must be nonzero.

    2) A strong empirical case can be made that these transition phases frequently take a course, whether by revolution, unrest, civil war, or some totalitarian coup d’etat, that causes far more destruction in one year than 100 years worth of gun crime.

    3) Once an undesirable form of government is entrenched, the psychic damage done to the population is, according to some ethical systems, infinitely bad. Totalitarian governments are capable of auto-genocide. It is axiomatic to the ethical system of a constitutionalist that totalitarian government is infinitely bad or at least worse than an honorable death, so there is no point in trying to argue against this common faith. The US federal government already bears responsibility for a long history of human rights abuses that critics of American propaganda are aware of.

    4) The tradition of Anglo-Saxon self-governance is all about dampening the costs to society incurred over long historical cycles by the inevitable transition phases between forms of government. Wisdom requires us to endure a baseline level of pain caused by gun proliferation in order to manage the long-term risks that government poses to society.

    5) Rifle ownership remains a credible deterrent to extreme government dysfunction. In the event of the collapse of the rule of law and any semblance of federal legitimacy, spontaneous defections by army units led by junior officers, working in conjunction with elements of local law enforcement, would coordinate popular resistance to any overreach attempted by particular cliques of politicians or military officers. The “rifles can’t beat atom bombs” argument overlooks the military reality that ubiquitous semi-automatic rifles with 8 round magazines are good enough for the modern battlefield, if not ideal. Looking beyond the worst case scenarios to the everyday effects of the second amendment, the possibility of armed civilian unrest puts constant pressure on government to satisfy the demands of the governed.

    6) The second amendment is justified as good risk management. Over long historical cycles, the social costs of government dysfunction could be infinite depending on one’s ethical outlook. Thus, even the smallest nonzero probability of the form of American government changing to a kind that is infinitely bad justifies the second amendment, overwhelming the social cost of gun crime. If ubiquitous military rifles are treated as social insurance, we must admit that the American nation may be the exceptional case of a society that will never face catastrophic conditions where popular resistance is desirable. It must be admitted that the second amendment could be a big mistake – it could be like buying an insurance policy that never pays out. Constitutionalists speak from maturity and wisdom to argue for managing black swan risks rather than projecting current prosperity infinitely into the future, as liberals do.

    7) Every cost-benefit analysis of the short-term or medium-term effects of gun proliferation is irrelevant to the gun question because the long-run cost-benefit analysis is overwhelmingly likely to confer a net benefit to society, provided we do not think like economists and discount the utility of future generations. It is pointless to obsess about the damage done to society by guns because (1-6) demonstrate the overriding value of the second amendment. Policies that mitigate the damage done to society by guns without unduly restricting certain essential kinds of guns are the prudent way forward.

    8) It must be conceded that prudent gun control measures exist. These measures reduce crime without impairing the potential of the masses to bear arms at crucial moments in history. However, incremental increases in gun control aiming to achieve an eventual total gun ban are unacceptable.

    However delightful the rhetoric of labeling long-term social planning as pessimism, tin foil hatish, conspiratorial, lunatic, alarmist, hyperopic, or by some other pejorative term may be, this argument is rigorous enough to demand the attention of those with intellectual integrity. This argument is the core of the ideology of Americanism as it is traditionally understood, so it must be refuted or suppressed to defeat gun rights advocates. More immodest variants of this argument easily appear as incomprehensible or as symptoms of some bizarre social pathology to those who aren’t acquainted with the underlying political philosophy of the American right. Most liberals are unable to comprehend the modest argument stated here due to their faith in the federal government.

    I think the main hole in this argument is the decline in the quality of the American people. In a few years, I don’t think it will be possible for the American people to resist some kind of totalitarian or globalist takeover. So why not disarm the population and enjoy the promised benefits while we can? Also, this reasoning leads to the conclusion that, if the second amendment is as powerful as the constitutionalists believe, it must eventually fail to prevent the “infinitely bad form of government.” Constitutionalists are guilty of pessimism at times; they would be better off if they weren’t trapped by a conservative ideology that prevents them from seeing the total range of policies that can protect society from extremely bad government outcomes.

    I don’t really have a dog in this fight, as I consider American society so dysfunctional that a gun ban would ultimately backfire, failing to deliver the promised benefits. On the other hand, I find the naivete of liberals who try to battle constitutionalist rhetoric equally laughable. Financial collapse, epidemics, decadence and moral decline, nuclear war, natural resource exhaustion, fascism – none of this could ever happen in America, right? This is just stuff that happens on the history channel. Most of what alarmists like Alex Jones preach is false, so we can safely ignore the minority that is obvious truth. Let’s just worry about the culture wars. Let’s just fiddle while Rome burns. Let’s just focus inordinate brainpower on parsing the details of narrow debates within false dichotomies. Guns and immigration are the causes du jour, so let’s pretend the rest of social reality doesn’t exist.

    • You’ve served up a great deal to digest here, and perhaps comment upon in the future. Let me just address the first point here, since it’s the one relevant to this post: I have not faulted the methodology in the DGU surveys. I have just pointed out that they have not met the burden of proof by merely asking people whether they’ve experienced these events, and there is no evidence to substantiate the projected numbers. I wouldn’t go so far as to say that the surveys are biased toward “gun nuts”, but they do roll out the red carpet for gross distortion.

  14. I’d just like to thank the article’s author and several of his posters for giving was was perhaps one of the most intelligent and riveting discussions on a such an emotionally-charged issue that I’ve ever seen.

    P.O.P., Frail Liberty, and Jonathon (although others not mentioned have also added their own excellence to this debate) are to be commended for recognizing the need to substantiate opinions with fact, and doing so in a respectful manner and with regard to critical thinking over dogma.

    Although I do not agree with all points, you each have certainly given me an appreciated for the need to validate a viewpoint beyond personal preferences and false statistics, and I’m certainly going to try to carry that methodology over to my own discussions with others on gun-based debates. This was… refreshing.

    Thank you all.

    ~ Drew

  15. correction: “…an appreciation for the need…” I didn’t find the misspelling in time. My apologies.

    ~ Drew

  16. Has anyone pointed out yet that the number of DGUs might be significantly higher because it is probably impossible for one guy or even one organization to find and record every report of a DGU that does make the media?

    • I did discuss the fact that probably not all DGUs are reported. But just how much “significantly higher” is another matter.

  17. Well said and researched. the FACT is that these DGU self-reporting “surveys” are VERY unreliable, as are all other self-reporting surveys where the subjects are essentially in favor of a positive outcome. It’s laughable what the NRA claims re: DGUs.

    What we DO know as a FACT is that gun ownership increases one’s risk of death by gun.

    America is FOURTH in gun deaths, worldwide – after Thailand, Colombia and Nigeria. So, with regard to deaths by gun, we are a third world nation. Sad.

  18. […] The third type of anecdote is exemplified by accounts of “illegal immigrants” committing atrocious crimes or gun owners thwarting bad guys. While these episodes are (sometimes) true as reported, the amount of attention they are awarded is often far out of proportion to their significance. The intended (and often achieved) result is to give the impression that these things occur far more often than they really do. […]

Leave a Reply to Phil Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s