So reports a press release posted at ScienceDaily.com, which also says,
The study estimated that people with a gun were 4.5 times more likely to be shot in an assault than those not possessing a gun.
“This study helps resolve the long-standing debate about whether guns are protective or perilous,” notes study author Charles C. Branas, PhD, Associate Professor of Epidemiology. “Will possessing a firearm always safeguard against harm or will it promote a false sense of security?” ...
Penn researchers investigated the link between being shot in an assault and a person’s possession of a gun at the time of the shooting. As identified by police and medical examiners, they randomly selected 677 cases of Philadelphia residents who were shot in an assault from 2003 to 2006. Six percent of these cases were in possession of a gun (such as in a holster, pocket, waistband, or vehicle) when they were shot.
These shooting cases were matched to Philadelphia residents who acted as the study’s controls. To identify the controls, trained phone canvassers called random Philadelphians soon after a reported shooting and asked about their possession of a gun at the time of the shooting. These random Philadelphians had not been shot and had nothing to do with the shooting. This is the same approach that epidemiologists have historically used to establish links between such things as smoking and lung cancer or drinking and car crashes.
This was promptly echoed in the Philadelphia Daily News.
Conspicuously missing from the press release and the news story were two critical limitations that were admitted in the original study. These qualifiers mean that the press release headline, as well as all the other statements and implications of causation, were quite mistaken. Perhaps defensive possession and carrying of guns helps protect the possessor and carrier, and perhaps it doesn’t. But the study sheds virtually no light on the subject.
1. To begin with, there’s the obvious causation/correlation problem. Maybe, as the authors speculate, carrying a gun increases your chances of being shot with a gun (as suggested by the framing of the issue as “whether guns are protective or perilous”), or at least fails to decrease them (“guns did not protect”). Or maybe a third source — perhaps some people’s being the targets of death threats, or being in a dangerous legal line of work, or being gang members or drug dealers — causes both higher gun carrying among those people and higher risk of being shot.
By way of analogy, we don’t suggest that pacemakers cause heart attacks, or don’t protect against heart attacks, just because we find a correlation between the presence of pacemaker and the incidence of heart attacks. Obviously, people might get pacemakers precisely because they’re at risk of heart attacks. Well, people might get guns precisely because they’re at risk of attack. (Stewart Baker makes a similar point.)
One can try to control for this in some measure — but while the study controls for some relevant attributes (race, sex, age, neighborhood, having a “high-risk occupation,” and having at least one arrest on one’s record), it leaves a vast range of factors uncontrolled. You’d think that gang members are more likely than others to carry guns and to get shot, even controlling for the presence of an arrest record. (Lots of law-abiding people carry guns, but I expect that more gang members do.) But the study doesn’t control for that, or for many other things.
Let me illustrate this with a deliberately oversimplified model. Let’s begin by assuming a total population of 100,000, that’s divided into two groups, a 10% high-risk group and a 90% low-risk group. Let’s say that the high-risk group has a 60% risk of being attacked, and as a result 40% of its members carry guns. And let’s say that the low-risk group has a 5% risk of being attacked, and as a result 3% of its members carry guns. Let’s also imagine a total population of 100,000 (just to make the numbers easier), and let’s assume that possessing a gun has a modest protective effect for both groups — it reduces the risk of being injured when attacked from 75% to 60%.
Here’s what this turns out yielding, with “A” meaning “armed subgroup” and “U” meaning the unarmed subgroup.
| Group | Number of people in group | Probability of being attacked | Armed subgroup fraction | Armed subgroup number | Armed subgroup injury risk | Armed subgroup number injured | Unarmed subgroup number | Unarmed subgroup injury risk | Unarmed subgroup number injured |
| High-risk | 10000 | 0.6 | 0.4 | 4000 | 0.36 | 1440 | 6000 | 0.45 | 2700 |
| Low-risk | 90000 | 0.05 | 0.03 | 2700 | 0.03 | 81 | 87300 | 0.0375 | 3273.75 |
| Total | 100000 | 0.067 | 6700 | 0.227015 | 1521 | 93300 | 0.064027 | 5973.75 | |
| Odds | 0.293686 | 0.068407 |
The result: The armed subgroup has 3.5 the risk of injury compared to the unarmed subgroup, and the relative odds ratio between them is 4.29. And this is so even though in the model gun possession decreases the injury risk for both the high– and the low-risk group.
Naturally, this is just a model; the real numbers are likely very different from the ones I give here, and in fact no-one knows what the real numbers are. (The model also doesn’t precisely fit the numbers in the study, though I’m pretty sure one can make a similar model that would fit them more closely.) My point is that one just can’t infer from an odds ratio of over 4 to the judgment that “guns did not protect those who possessed them,” much less that they were actually “perilous” to the possessors. The high odds ratio is just as consistent with the model I describe as with a model where gun possession increases the risk of injury.
2. But wait, there’s more. The research model works only to the extent that you actually know who possesses guns and who doesn’t. Both the cases (people who were shot) and the controls (people who were called on the phone) might want to conceal their gun possession. The cases might have thrown away their guns before the police arrive (sometimes easy, sometimes hard or dangerous). The controls might have lied to the stranger who calls them to ask them, “Where were you at 10:30 pm two nights ago?,” and “Were you possessing a gun at the time?” (always easy and safe).
And both the cases and the controls might have plenty of reasons to lie. They might have been possessing guns in public without a concealed-carry license. They might have been felons who didn’t have the right to possess a gun even at home. (People with arrest records made up 53% of all cases and 37% of all controls; the study doesn’t tell us how many had felony conviction records, but I suspect that quite a few of those with arrest records did.) Or they might not be sure what the questioner is getting at. And that’s true even if the questioner claims that he’s just an academic researcher.
Fortunately, the study helpfully tells us what would happen if there’s concealment of gun ownership by some fraction of cases and controls — though of course the press release and the newspaper article are silent about this. If only 1% of controls and cases who are reported not to have had guns are randomly recoded to having guns, two of the three results (“all gun assaults,” “gun assaults where victim had at least some chance to resist,” but not “fatal gun assaults”) end up yielding statistically insignificant results. If 3% are so recoded, all three results lose statistical significance.
If we assume that 1% of controls were concealing their gun possession and 0% of cases were concealing it — not implausible, since it’s easier for a control to conceal gun possession than for a case to do so (since the cases may be too injured to get rid of the gun, may need a gun for continued self-defense, and in any event are the subjects of a police investigation in which the police might learn the truth) — all three results lose statistical significance. The numerical value of the non-significant odds ratios falls as well; if 5% of controls conceal their gun possession buy 0% of cases do, the odds ratio falls to 2, which of course reflects a considerably lower relative risk. And these are pretty low percentages of false reporters, given the incentives that many people might have to hide their gun possession.
And all this is in addition to the possible confounding factors discussed in item 1 above. If there were no such confounders, then perhaps even a low odds ratio might be telling, or perhaps even a statistically insignificant odds ratio above 1 might in some measure undermine the “guns as protective” theory. But these two problems put together — the possibility that the result stems from the existence of a high-risk group whose members are especially likely both to carry guns and be the targets of attack, and the possibility of even slight misreporting dramatically affecting the results — make the study highly uninformative.
So it’s possible that gun possession was “perilous,” in the sense of increasing the risk of the possessor’s being injured. It’s possible that it “did not protect those who possessed guns,” in the sense that it didn’t reduce the risk of the possessor’s being injured. But it’s also possible that it was “protective,” in that it reduced the risk of the possessor’s being injured, but this result is swamped by the other phenomena I point to. The study doesn’t give us much extra information about which theory is correct. And yet it is publicized, and it’s reported, as if it did robustly show the causal relationship.
David Schwartz says:
This is a horrible abuse of statistics, utterly inexcusable.
Quote
October 5, 2009, 4:22 pmPeteP says:
Liars, damned liars, and statistics.
These idiots take 40 examples of ‘people who were shot while carrying a gun’, and think they can draw some kind of conclusion from it that applies to a population of 300,000,000 + , with 100,000,000 + gun owners ????
I wonder how much taxpayer money went into this little piece of garbage ?
Quote
October 5, 2009, 4:25 pmDaily Pundit » More From the Sorts of Scientists Who Bring You All Those Global Warming Studies says:
[...] The Volokh Conspiracy » Blog Archive » “Guns Did Not Protect Those Who Possessed Them from Being... [...]
Rob M says:
It also does not seem to take into account the numerous cases of people defending themselves with a firearm but no shots were fired (e.g. homeowner with a gun stops a burglar who has a knife).
Quote
October 5, 2009, 4:32 pmGordo says:
1. Based upon the second factor, that people might lie about whether they had a gun or not, NO ONE would ever get a valid statistical sample regarding this question.
Which means, not only that no one could ever prove that gun carry is more dangerous, no one could ever prove the opposite, that gun carry is safer.
2. It appears that the researchers did indeed control for a lot of relevant factors, which Professor Volokh briefly acknowledges before trumpeting the lack of control for “gang members.” I suspect that even if the researchers had controlled for many, many more factors than they did, someone would pop up with an uncontrolled factor and claim the whole study was therefore invalid.
Also, if you think people are going to lie about whether they have a gun, do think they also might lie about whether they were gang members? So no study will ever be conclusive.
Quote
October 5, 2009, 4:37 pmVeracitor says:
There is a much worse flaw in that study. It does not tell us anything about people who did not get shot. Suppose having a gun reduces the chance that you will be shot by an assailant from 50% to 10% (numbers hypothetical). For example, assailants might commonly back off when their victims pull out guns for defensive use. The “control group” in that study is not composed of assault victims, but rather of random citizens. Since the researchers did not study victims who escaped being shot,* they cannot say whether possessing a gun helps reduce the chance of being shot!
The editors promoting that crap should be deeply ashamed of themselves.
*Which would’ve been difficult, since successful self-defenders might not have reported assaults to police.
Quote
October 5, 2009, 4:37 pmJack Burton says:
Gordo sez: Based upon the second factor, that people might lie about whether they had a gun or not, NO ONE would ever get a valid statistical sample regarding this question.
Jack replies: Yes, a great many people DO feel a compelling need to lie when asked questions about guns, gun possession, ownership, and use of them.
“Guns? What guns?” is often a common expression if there’s even the slightest whiff of the legitimacy of the questioner.
Quote
October 5, 2009, 4:54 pmIke says:
Gordo is quite right in his assertion that no study will ever be conclusive; at least, no study which relies upon self-reported information. Self-reporting is one of the fundamental flaws in most such studies these days; see, e.g., self-reporting in studies of childhood food allergies most often peanut butter.
The difficulties of proving either guns are useless or useful is one of the reasons why it is not legally required that either proposition be proven. The Second Amendment to the federal constitution says that we each, as individuals, have the right — note, not a privilege granted by some politician, but a right — to possess and use firearms. So, studies seeking to prove or disprove the utility or dis-utility of firearms possession are ... well, superfluous, comes to mind. It isn’t important and provides neither a justification for “allowing” gun ownership and possession nor a justification for “banning” gun ownership and possession. As we are human, we have certain inalienable rights and gun ownership and possession is one of them which is recognized and listed in the Bill of Rights. Most states’ constitutions have a similar bill of rights, in one form or another. So much for studies of issues which are irrelevant.
Quote
October 5, 2009, 5:01 pmRob M says:
@Gordo
1) The questions and method of asking this study used are unlikely to ever give anything approaching statistical validity. However it is true that it is incredibly hard for any study pro or con to get good information and control for relevant factors. That does not mean that no study ever could by asking different questions or using different methods. Also even if no study ever gets 100% on the mark it doesn’t mean all studies are equivalently flawed.
2) Controlling for “a lot” of factors but missing one that could and probably does completely skew your outcome does make the study invalid. Its not the quantity of the factors controlled for but the quality. It would be like a doctor bandaging some scrapes and bruises but missing the fatal gunshot wound on a patient.
Just because these guys failed doesn’t mean its impossible.
Quote
October 5, 2009, 5:03 pmhappycynic says:
The majority of gun violence in this country is due to gang members shooting other gang members. This study fails to control for this factor. In other words, the study is almost certainly comparing a population comprised largely of violent felons to the general population. Such a failure is completely inexcusable given the subject matter and raises the question of whether this was a serious study to begin with, or simply an attempt to generate statistics favorable to the latest anti-gun fad floating around medical circles. I won’t even get into the absurdity of trying to use methodology used to measure disease on what is quite obviously a question of social science.
Quote
October 5, 2009, 5:13 pmKazinski says:
I think the problems you find in the study make it pretty much meaningless, except for one thing: it doesn’t fit the narrative.
Quote
October 5, 2009, 5:16 pmpmorem says:
There is no reason why I cannot own a firearm if I so choose.
That said, unless I have some compelling legal reason to do so or the person is already known to me (and trusted), I generally won’t honestly answer the question of whether or not I own one or more firearms.
I might decline to answer, or I might deliberately mislead the person asking, ie “I may or may not own a Remington Model 760 .30–06″.
I don’t remember where I picked that up from, but it wasn’t my original idea. I’ve heard others advocate the same thing, some of whom have disclosed lawful firearms ownership to me. I suspect it’s common among lawful firearms owners, of whom I may or may not be one.
In some regards, this is a deliberate effort at confounding questioners.
How does one control for that?
Quote
October 5, 2009, 5:28 pmpete says:
So if there is in fact a correlation, the most logical response is to ban police officers from carrying guns to reduce the chance that anyone will shoot them. Perhaps we should ban soldiers from carrying guns to protect them as well since I suspect that unarmed soldiers are less likely to be shot than armed soldiers.
Quote
October 5, 2009, 5:30 pmMikhail Koulikov says:
The Second Amendment to the federal constitution says that we each, as individuals, have the right – note, not a privilege granted by some politician, but a right – to possess and use firearms.
Um...the Second Amendment — like all the others — *is* a privilege granted by politicians. Laws and “rights” are not something that just exists in the atmosphere; they care created. And if *enough* politicians get together and are of the same mind, they can always take the privilege away. In case you forget, that’s how human society is always organized.
Quote
October 5, 2009, 5:31 pmKevin P. says:
This is another “study” funded by the Joyce Foundation which is a primary funder of gun control groups and has funded several studies in the public health field, almost all of which find guns to be bad for health.
Most criminologists who seriously study the issue find that widespread ownership of guns makes little difference to the crime rate or lowers it to a small extent.
Public health activists on the other hand, start with the premise that owning guns is immoral, conduct studies to “find” the effects of owning guns, and voila! find that guns are bad!
Read this interesting article by criminologists — and physicians — on public health studies of guns:
Guns and Public Health: Epidemic of Violence or Pandemic of Propaganda?
Interesting excerpts:
In 1979 the American public health community adopted the “objective to reduce the number of handguns in private ownership,” the initial target being a 25% reduction by the year 2000.[3] Based on studies, and propelled by leadership from the Centers for Disease Control and Prevention (CDC), the objective has broadened so that it now includes banning and confiscation of all handguns, restrictive licensing of owners of other firearms, and eventual elimination of firearms from American life, excepting (perhaps) only a small elite of extremely wealthy collectors, hunters, or target shooters.
... and ...
We believe we have documented an emotional anti-gun agenda in the treatment of firearms issues in the medical and public health literature. While the anti-gun editorials and articles discussed had the superficial form of academic discourse, the basic tenets of science and scholarship have too often been lacking. We call them “anti-gun health advocacy literature” because they are so biased and contain so many errors of fact, logic, and procedure that we can not regard them as having a legitimate claim to be treated as scholarly or scientific literature.
Quote
October 5, 2009, 5:44 pmXenocles says:
The study fails in the very first line, since guns don’t “protect” from anything. All they do is augment a person’s ability to make credible threats. If a person has no will to violence, going armed will probably be counterproductive since his weapon will be like a talisman whose power will fail at the critical moment.
Just as it takes a person to kill it also takes a person to protect. The mind is the actual weapon; objects like guns and knives are merely tools that extend the power of that weapon. How many of the people in this study got hurt because they thought all they had to do was buy a gun and they’d be safe? How many failed to train to use that gun? How many simply lacked that necessary will to take a life?
Quote
October 5, 2009, 5:49 pmStateOfNature says:
@Mikhail Koulikov
Why can’t a natural right be the antecedent of a legal right?
On a more snarky note,
Police power *is* a privilege granted by the people. Governments are not something that just exists in the atmosphere; they care [sic] created. And if *enough* people get together and are of the same mind, they can always take the privilege away.
Wait a sec–that sounds a lot like the rationale for the Second Amendment.
Quote
October 5, 2009, 5:53 pmrmd says:
I’m going to commit the cardinal sin of commenting on a topic without having read the source material but I hope you’ll forgive me.
Doesn’t it matter whether the guns were being carried legally or illegally? Without that distinction, even if the result is statistically valid, then it’s useless for formulating policy.
As an aside, if you’re going to use statistics to try to convince me to do or not do something, you’d better not let me find out that you buy lotto tickets.
Quote
October 5, 2009, 5:58 pmChrisTS says:
I am not a libertarian on gun ownership/carrying, but I find this study very strange. I cannot quite figure out how the researchers thought it would provide [valid] data about ...anything. Quite apart from the tiny sample size.
Quote
October 5, 2009, 6:07 pmchris says:
Not once do they mention how many of the people in the study were involved in illegal activity when they were “assaulted”
Quote
October 5, 2009, 6:29 pmpete says:
A more insteresting study might have checked for other forms of assault and attempted assault. What percentage of rape/stabbing/mugging etc. victims in the city were carrying guns while assaulted? How many reported attempted victims were carrying?
Quote
October 5, 2009, 6:49 pmDisintelligentsia says:
Why does this not remind me of the Lancet study that “found” that approximately 100,000 Iraqis died due the war and this finding was published just three days before the election in 2004? Here we have the SCOTUS taking up the second amendment’s application to the states and, viola! a new study pops up telling us how very silly those pro-gun rights people are and that guns are a societal evil that hurts even those who use them only for self-defense. Of course we know the SCOTUS is supposed to ignore such sociological data and rule on the law, right?
Quote
October 5, 2009, 7:07 pmDisintelligentsia says:
The ScienceDaily article said that the study was funded by the National Institutes of Health — where did you get that it was funded by the Joyce Foundation?
Quote
October 5, 2009, 7:16 pmGordon Langston says:
The CDC quit doingstories about the effectiveness of firearms regulation. I guess studies like this are the next line of offense.
Quote
October 5, 2009, 8:01 pmyankee says:
It’s not irrelevant: for one thing, many of us would find it useful in evaluating whether or not the Second Amendment is a good idea.
Quote
October 5, 2009, 8:04 pmtraveler496 says:
If forced to wildly speculate, I would wildly speculate that, on average, owning a gun does increase one’s likelihood of being shot; but I realize that I have no hard data to support this, either before or after exposure to this apparently execrable study.
Thanks for ripping it to shreds, Eugene. IMO the American Journal of Public Health has some ’splainin’ to do (hedging a bit only because I couldn’t access the .pdf).
Quote
October 5, 2009, 8:20 pmandrewpaterson says:
This reminds me of the old Redd Foxx routine on ‘pocketknives’ He mentions how this ‘academic fella’ claimed that his study proved that a disproportionate number of ‘blackfolk’ possessed ‘pocketknives’ compared to ‘whitefolk.’ he then went on and on about all the black friends of his he had known over the years and none of them owned a ‘pocketknife.’ He said he was indignent about this ‘academic fella’ and was thinking about giving him a piece of his mind. Why in all his years he — and he was a ‘blackfolk’- and in all his years he had never owned no ‘pocketknife.’ Yes he was gonna get ahold of that ‘academic fella’ and let him know he was wrong — none of his friends owned no ‘pocketkives’ and he didn’t own no ‘pocketkife.’ What kind of study could that ‘academic fella’ have done?
Then he concluded: ‘Now I ain’t got no pocketkife — that ‘academic fella’ is just a fool, yeah I ain’t got no pocketknife... but I gots’ this really nice switchblade I’ve had for 40 years, you should see it, its really fine”
Quote
October 5, 2009, 8:21 pmJeff Walden says:
Here’s a better link to the original article.
Quote
October 5, 2009, 8:38 pmRicardo says:
Social science literature (especially in labor economics) now has many good examples of studies where there is either pure randomization or “quasi-randomization” so that the control and treatment group are comparable. A study like this is almost never going to give a reliable result because those things they did not — and maybe cannot — control for are likely to be significant. That’s life.
An example of a good study would be this: convince a state to randomly deny concealed carry applications or at least defer them for a year or two. Then, in statistics jargon, you can use approval as an “instrument” for gun ownership. People who are approved for concealed carry are more likely to be carrying a gun than people who were denied. Then you have comparable treatment and control groups and can compare murder rates among these two groups if your sample size is large enough and if the crime rate is high enough in the state. You can then definitely say whether carrying a gun lowers your chance of being a crime victim or not.
Now a study like this is unlikely to ever be implemented in the U.S. But that is, in fact, how a properly designed study would look. These guys didn’t “help resolve” any debate.
Quote
October 5, 2009, 9:49 pmJ.S.Bridges says:
Ahhh...Mikhail — to put it as politely as possible*: Your understanding of the true nature of “rights” and the U.S. Constitution — and of their relationship to politics and politicians — is approximately 180-degrees bassackwards.
The Second Amendment — like the other nine of the first ten Amendments — states an inherent right of all citizens of the U.S. that is specifically retained by those citizens. This is not a “privilege granted by politicians” — it’s a more-detailed iteration of one of those “inalienable rights” the founders wrote about. Politics may “regulate” the free exercise of that right — within certain limits. Most of the current legal debate regarding possession and/or use of guns by individuals is over where those limits are, or should be. However, no politician or group of politicians may legally obviate — or “infringe” upon — arms possession/use, in the end. It is a right — inherently and lawfully — not a “privilege” granted by benevolent politics.
*A less–polite way of putting it is that you — along with anyone else who thinks the Second is a “privilege” — are simply full of s**t.
Quote
October 5, 2009, 11:54 pmMatthew Carberry says:
A “good idea”?
Apply some logic at least.
The Second concerns, at root, the natural right to general self-defense using any or all means available, not a utilitarian proposition about the “worth” of arms for such self-defense. Therefore studies based on utilitarian arguments are utterly irrelevent to the issue.
Further, even if the second did have a utilitarian base, this article doesn’t show an utter lack of positive societal or individual effect to gun ownership or carriage (indeed, no study yet has), it merely claims a proportional disadvantage. “Lack of a statistically significant positive” effect is, at worst, a possible negative on an individual basis, not a dismissive negative in all cases.
Therefore, though I hate the phrase and puerile mindset that conceived it, if (lawful) possession and carry have any positive effect, “saves even just one life” (insert eye roll), this study fails to help determine policy using even that asinine metric.
Quote
October 6, 2009, 12:08 amMatthew Carberry says:
Whoops.
They do claim a negative effect, don’t they. Mea Culpa.
However, “more likely” still doesn’t mean “will in every case”.
The “just one life” puerility defense still stands, since the negative effect is to the individual who chooses to carry, not a bystander or other uninvolved innocent.
At worst, under a utilitarian system absent evidence of harm to others, I should be free to risk the “4.5% more likely” harm to myself without government butting in.
Quote
October 6, 2009, 12:19 amLior says:
I assume everyone is familiar with this comic ?
Quote
October 6, 2009, 1:08 amKharn says:
The study definitely needs to be reexamined to compare licensed vs unlicensed carrying by both the shooter and the injured party. Also, were police shootings excluded?
I bet you will find a very high rate of unlicensed vs unlicensed, a very low rate of unlicensed vs licensed (either direction, obviously not counting police action), and a neglible rate of licensed vs licensed.
Quote
October 6, 2009, 4:34 amCarl from Chicago says:
The unfortunate thing is that “news” headlines and articles like this will then be trotted out as fact, and will become the newest canard.
Sound bytes move issues these days.
Watch for the “people with a gun were 4.5 times more likely to be shot” to crop up among the various gun control proponents. Watch also for it to crop up among our various elected officials.
Quote
October 6, 2009, 4:59 amCarl from Chicago says:
PeteP said:
I don’t know, but I understand that some was. You should know that the University of Pennsylvania School of Medicine’s work on “gun violence” is funded in part by Chicago’s Joyce Foundation (perhaps the single largest contributor of gun control funding).
Quote
October 6, 2009, 5:01 amSayUncle » More on that supposed study says:
[...] Eugene Volokh notes what we knew all along about the supposed study that was funded by an anti-gun group. [...]
Kevin P. says:
I misspoke when I said that the study itself was funded by the Joyce Foundation.
The entity and persons who performed the study are funded by the Joyce Foundation.
See this grant:
Three of the five authors of this paper (Charles C. Branas, Therese S. Richmond and Douglas J. Wiebe) are members of the Joyce-funded FICAP team.
Quote
October 6, 2009, 6:35 amSeaDrive says:
This study was clearly designed to generate some impressive sounding numbers that can be trumpeted out of context. I do wonder, however, about the effect of carrying a firearm on assumption of risk. For example, it’s possible that any particular person is more likely to go into a 7–11, or even just walk the street, in a bad neighborhood when he is armed than when he is not armed.
Quote
October 6, 2009, 6:52 amJohn Fast says:
I tried to post the following comment on Philly.com:
This study is flawed; by the same logic I can claim that having a pacemaker causes heart attacks, because people with pacemakers are more likely to have heart attacks than people who don’t.
I don’t blame the reporters, because I assume they’re not familiar with basic probability and statistics, but Professor Branas should be embarrassed, or more likely ashamed; an epidemiologist should know statistical analysis, and he probably does, so this is more likely from bias than incompetence. (I do economics and political sciente so I can crunch basic numbers — but the guy who spotted the logical flaw is a law professor, Eugene Volokh, on Volokh.com )
Quote
October 6, 2009, 6:53 amMark in Texas says:
In addition to the flaws that Professor Volokh pointed out, a number of people have also identified the problem of asking people if they posess guns, especially if it is not legal for them to do so.
If only there was a way to eliminate that issue of selection bias. If only somebody would compare homicide rates in counties where states allowed concealed hand gun carry to counties where states prohibited legal carrying of handguns. That study could also analyze the change within counties when states changed their laws on concealed carry.
Oh, that’s right. Somebody did.
Quote
October 6, 2009, 6:56 amTurk Turon says:
As criminologist Gary Kleck once remarked on “case-control” studies that imply causation: it’s like doing a study to determine how diabetics differ from non-diabetics, discovering that they are more likely than non-diabetics to possess insulin, and concluding that insulin possession is a risk factor for diabetes.
Quote
October 6, 2009, 7:55 amPintler says:
FWIW, for the limited set of people I know well enough to hazard a guess about, being armed makes them more likely to go dangerous places like having a smoke alarm makes them more likely to play with matches :-).
On the original post: Kudos to EV for the analysis — assuming slightly different data and seeing how it affects your results is done far too rarely. Doing so — creating what you think is a plausible range of data, and analyzing it before conducting a study — would prevent a lot effort being wasted on poorly designed studies.
For a ‘wisdom of crowds’ approach to the question, IMHE off duty police routinely carry, presumably because they think it protects them and their families. I wonder if the study authors, or Kellerman, think the police are making a mistake in doing so, and would in fact be safer leaving their gun at the station at the end of the shift.
Quote
October 6, 2009, 7:55 amCJColucci says:
It takes a fair amount of training and practice to use a handgun with enough proficiency to do oneself more good than harm in an actual emergency. Has any study controlled for competence?
Quote
October 6, 2009, 8:06 amJack Burton says:
Unfortuantely CJ’s statement of fact doesn’t fit actual real life data... Here are thousands of ordinary citizens who have done exactly what he claims is unlikely...
http://claytoncramer.com/gundefenseblog/blogger.html
This is a classic case of projection from CJ.
Quote
October 6, 2009, 8:48 amFloridan says:
“The majority of gun violence in this country is due to gang members shooting other gang members.”
The majority? Is this true?
Without searching out statistics, I would have thought domestic violence would lead to more gun violence. Seems like everyday I read about someone shooting his wife (and sometimes his kids) and then himself. Perhaps it’s just a South Florida phenomenon.
Quote
October 6, 2009, 8:50 amMark Buehner says:
It takes no training at all to brandish a weapon, and that is the most prevalent way a handgun is used in self-defense by far. That’s another problem with this study, it excludes those who were saved from assault in the first place.
Quote
October 6, 2009, 8:58 amPubliusFL says:
Perhaps you see such things in the news more precisely because it’s more unusual and attracts more attention. Domestic violence accounts for only about 10% of homicides.
Quote
October 6, 2009, 9:14 amDennis N says:
I suspect from personal experience, that the majority of firearms incidents are never reported and do not involve a discharge.
I was personally involved in two encounters in Washington State in the 1970s. One time my wife and I were accosted by two thuggy looking young men near Pike Place Market in Seattle. They very quickly discovered a need to be elsewhere, when I reached into my back pocket. There was nothing in particular to report, and dealing with cops is generally dangerous, so I just walked away.
The second involved some old teenaged girls who had been assaulted and kicked out of the car in front of my house. I heard the screaming, grabbed a shotgun, and investigated. Next thing I knew, there were two hysterical girls hiding behind me and a car disappearing in a cloud of dust.
The cops’ question about the shotgun was, “Where is the shotgun now?”
“It’s in the closet. It’s clear (That’s ‘armed professional’ for emptied of ammunition and checked.).”
“Oh. OK. Is there any more coffee?”
I eventually had some five cops, two EMTs and twenty six volunteer fireman making inroads on that coffee, but that’s another story.
While it is gospel among shooters that you do not draw a weapon unless you intend to shoot it, in fact it is seldom necessary. And thankfully so. But that makes for some pretty squidgy statistics in an area notorious for squidgy statistics.
Quote
October 6, 2009, 9:27 amA. Cooper says:
Can we take a moment to distinguish between the press release (likely written by someone in a press office who may have met the researchers a few times and probably didn’t read the study) and the news articles (certainly written by someone who didn’t read the study) and the study itself?
Science journalism has a tendency to be *really really* bad, worse perhaps than the reporting about judicial opinions. Don’t take a press-office writer’s or a journalist’s conclusions for being the conclusions of the authors of the study.
Quote
October 6, 2009, 9:45 amPintler says:
From the Bureau of Justice Statistics stats:
Male offender/Male victim 65.3%
Male offender/Female victim 22.7%
Female offender/Male victim 9.6%
Female offender/Female victim 2.4%
Unless there is a lot of same-sex domestic violence, DV related murders aren’t a majority. Maybe a third if all the M/F and F/M are DV related.
(I was quite surprised to see that F/M is almost half of M/F)
Quote
October 6, 2009, 9:45 amUpend, Coming says:
Boy, that statistics problem looks very similar to the example you used in your Academic Legal Writing book. That was about violence against gun owners too, right?
Quote
October 6, 2009, 9:46 amSeaDrive says:
I know what you’re saying, but leaving guns aside, people who feel protected by their safety gear do things they would not do otherwise.
Quote
October 6, 2009, 9:56 amNickM says:
It’s not just that it’s unusual. Crimes involving kids get extra publicity — and if the child is female, blonde, and generally considered cute, even more.
Nick
Quote
October 6, 2009, 10:05 amDennis N says:
rmd says:
It certainly effects the statistics. I have had the occasion to carry illegally a time or two. This was based on the “better judged by twelve than carried by six” principle, working late, and parking in a dark secluded area in a tough town. Anyhow, unless I was involved in an incident that involved someone assuming the horizontal position, there is no way I would report it. Fortunately, I never had anything to report.
Ricardo says:
You can also compare the rate of firearms crimes between states with and without CCW. The results of several studies I have seen, have been inconclusive, generally leaning toward a small decrease in violent crime. From this, I conclude that personal carry of weapons is not the problem that anti gun people claim it would be.
It is also instructive to note the reasons why CCW licenses are revoked. Usually, they are for procedural violations, things like carrying in inappropriate places, and not for discharging weapons under inappropriate circumstances. Again, it indicates that, at the worst, CCW does not cause a problem.
Pintler says:
As a gun toter, and someone who hangs around and talks with other gun toters, I think it is safe to generalize that the last thing we want to be involved in is a shooting. Shooters are almost phobic about it. Shooting someone does lots of bad things to you and your family, like remorse, being arrested, sued, etc. It is the absolute last resort.
CJColucci says:
More training is always better than less. But considering the absolutely abysmal level of police firearms training, and the relative low butcher’s bill therefrom, lack of training doesn’t seem to be the major problem. The average gunfight occurs inside twenty feet, and the hit rate from both cops and citizens is pretty low. Adrenaline appears to be the major confounding factor, much more so than formal range time. Few organizations actually train for that. The only ones I can think of, are the military and maybe some well endowed SWAT teams.
Someone did a study that showed that you were more likely to be the victim of an unjustifiable shooting by a cop than by a common citizen. This excludes being a victim of a criminal act, but covers things like being shot by accident, through mistaken identity, by a missed round or a shoot-through. I don’t think this is a condemnation of cops; their job includes going out and looking for trouble. But it does indicate that Joe Sixpack is nowhere near the menace he is sometimes painted to be by enemies of the gun rights. He really does not want that piece to go boom.
Quote
October 6, 2009, 10:08 amRobtA says:
All I know is that my first teen girfriend (age 14, 1964) has detested me since 1965, and now she’s a well-known feminist professor with a gun.
Quote
October 6, 2009, 10:33 amJim Nielsen says:
Reductio Ad Absurdum. Nice.
Quote
October 6, 2009, 10:40 amJack Burton says:
Let’s hear from a true expert in the field... Massad Ayoob is one of the leading authorities about firearm usage, and has testified in front of hundreds of juries. He has written numerous books that are considered “must reads” by anyone who owns a gun.
People unfamiliar with the practice fear that “the trigger will pull the finger,” and armed citizens will go looking for a chance to exercise their deadly power. This, too, is a largely unfounded belief..
The collective experience of ordinary, law-abiding people who carry guns is that they don’t feel a sudden urge to go into Central Park at three o’clock in the morning and troll for muggers. They learn that being armed, they are held to what the law calls “a higher standard of care” and are expected to avoid situations like traffic arguments that could escalate and, with a deadly weapon present, turn into killing situations.
Quote
October 6, 2009, 11:01 ampete says:
What percentage of homicide/shooting victims are felons? From what I have read before in books like Homicide it was almost a majority of homicide victims (with felon on felon homicides being pretty common) and much higher than the percentage of felons in the general population. My quick search of DOJ statistics on homicide victims does not show any breakdown by felon status, although it does by race and gender, which both have a strong correlation to felon status. That is probably partly the reason why men tend to be more likely to be shot/killed then women.
Quote
October 6, 2009, 11:06 amOrder of the Coif says:
Not only did they NOT screen for unlawful possession, they didn’t screen for drugs/alcohol in blood at time of homicide/assault, nor for gang membership. Those are the three variables that crop up most often in violent crimes (just ask the Criminologists who report =/- 80% of incidents share these 3 characteristics). The MD’s always cherry pick their data base in order to secure the desired result.
In fact, this seems to be just another version of the original Kellerman study (43 times) with all the same flaws. I wonder if these “scientists” will share their dataset. Kellerman wouldn’t for 20 years.
Quote
October 6, 2009, 11:25 amJB says:
Will cops read this study and stop carrying firearms and in patrol vehicle to cut their potential to ‘get shot’ by 450% (4.5x)? A study of black Philly felons doesn’t really apply to us, and if an exchange, the chance of course goes up-crack a few eggs to make an omelet-where practice and training come into play.
Quote
October 6, 2009, 11:27 amSeaDrive says:
See Dennis N, above.
Perhaps if the tough town was NYC, Dennis would have stayed home.
Quote
October 6, 2009, 11:31 amlucklucky says:
And the assaults that did not happen because someone had a gun.?
Quote
October 6, 2009, 11:33 amGaius Obvious says:
This is similar to the statistic that establishing a new hospital in a neighborhood greatly increases the chances of death for the neighbors as the number of deaths recorded in the neighborhood skyrockets after a hospital is built there. Likewise the percentage of people who die while on the way to the hospital is far, far greater than the percentage of people who die on the way to a McDonald’s so we must conclude that if you are in danger of immediate death (heart attack, not breathing, stroke) since you have a greater statistical chance of dying while in transit to the hospital and a lesser statistical chance of dying while in transit to a McDonald’s, to increase your survival rate one should choose going to a McDonald’s instead.
Quote
October 6, 2009, 11:37 amDennis N says:
SeaDrive says:
No, I would have gone to work anyway, but without a weapon, would have been at greater risk. But with the weapon, I would not have gone out trolling for trouble.
Quote
October 6, 2009, 11:42 amPubliusFL says:
Not specifically about prior felonies, but here’s an article about murders for the first half of 2009 in Milwaukee:
“About 75% of the victims and about 86% of the suspects had prior arrests, according to the report, compiled by the Milwaukee Homicide Review Commission.
“Victims killed in the first six months of 2009 had an average of 12.4 prior arrests, the report shows. Suspects in those homicides had an average of 12.8 prior arrests.”
Quote
October 6, 2009, 11:42 amDennis N says:
These statistical comparisons remind me of the incident of the guy arrested trying to carry a bomb aboard the airliner.
TSA: “Why did you want to blow up the airliner?”
Bomber: “I didn’t want to blow up the airliner?”
“Then why were you carrying a bomb?”
“It was for safety.”
“Safety?”
“Do you know the probability against being on an airliner with a bomb aboard? Well, think of the probability against being on an airliner with two bombs.”
Quote
October 6, 2009, 11:47 ampete says:
I have seen similar stories before such as this one:
But it has always been for particular cities, not for the nation as a whole and it looks like DOJ may not collect that information.
Quote
October 6, 2009, 12:21 pmpete says:
I had a sociology professor who pointed out to us while explaining correlation/causation that there was a strong correlation between the number of fire engines that arrive at a fire and the total amount of damage the fire does, with more fire engines meaning more fire damage.
Quote
October 6, 2009, 12:35 pmTim McDonald says:
Here is a question. Of those people who were doing the assaulting, how many were injured by those who were armed and how many by those who were unarmed?
Another question? How many of those who were armed were not assaulted because they were armed?
This study seems to show a certain bias in the questions asked and not asked.
Quote
October 6, 2009, 1:01 pmtraveler496 says:
Having now skimmed the article (thanks for the better URL Jeff Walden) I’ll dial back my earlier “execrable” a bit.
The paper does acknowledge the study’s limitations (wrt confounding variables, reverse causality, and misreporting); and IMO the study is not entirely useless (at least, it got me curious about the extent to which the sizeable odds ratios might be due to something real rather than to these limitations).
The problem remains that the paper’s primary conclusion (“On average, guns did not protect those who possessed them from being shot in an assault”) is simply unfounded, in that it totally ignores(!) these acknowledged limitations.
I therefore can’t give the paper better than a “mostly execrable” (while EV gets the apparently typical “exemplary” for both substantive analysis and diplomacy).
Quote
October 6, 2009, 3:03 pmOrder of the Coif says:
No, it does not. Virtually all self-defense shootings (as distinguished from murder or mayhem) take place at arms length, involve less than 3 shots, and are over in less than ten seconds. The only real physical skill is in getting the gun into your hand and pointed at the threat. The necessary mental element is the decision that “I will survive” (often made years before the incident). As usual, the brain is your weapon; the gun is merely a tool.
This is why competent self-defense instructors emphasize pocket holsters (so you can have the gun “in hand” before the threat grows imminent) or practicing your draw enough to make it instinctive and fast (always carry in the same location and work on smooth technique; speed come naturally with the practice).
Quote
October 6, 2009, 5:55 pmBob Kohn says:
This kind of work is common in the American Journal of Public Health. The editors are more concerned with how visibility in mainstream media like Science Daily will drive policy and futher public health (as they see it, of course) than in the scientific merit of the studies they publish. They make all epidemiology look bad.
Quote
October 6, 2009, 6:01 pmStill more birdshot « Blunt Object says:
[...] “Guns did not protect those who possessed them from being shot in an assault” (The Volokh Conspiracy) [...]
Debunked « Stuff From Hsoi says:
[...] | Tags: Guns, Politics | It wasn’t one of the sources I originally contacted, but still, debunking is debunking (and still a great and credible source for [...]
CJColucci says:
CJColucci: It takes a fair amount of training and practice to use a handgun with enough proficiency to do oneself more good than harm in an actual emergency. Has any study controlled for competence?
No, it does not. Virtually all self-defense shootings (as distinguished from murder or mayhem) take place at arms length, involve less than 3 shots, and are over in less than ten seconds. The only real physical skill is in getting the gun into your hand and pointed at the threat. The necessary mental element is the decision that “I will survive” (often made years before the incident). As usual, the brain is your weapon; the gun is merely a tool.
This is why competent self-defense instructors emphasize pocket holsters (so you can have the gun “in hand” before the threat grows imminent) or practicing your draw enough to make it instinctive and fast (always carry in the same location and work on smooth technique; speed come naturally with the practice).
Actually, that sorts of sounds like my point.
Quote
October 7, 2009, 11:22 amroad2serfdom says:
CJC:
Your point: A gun is more likely to do harm than good in a self defense situation when the user has little training.
His point: Even with little training, a gun is more likely to do good than harm in situtations. Of course, training can increase the odds for good even further.
Not one is saying training is bad, or that it has no marginal benefit. The fact that training increases the probility of good does not mean that without training P(bad)>P(good). Your claim has no evidence to back it. Given the fact that in the vast majority of self defense cases a gun is not even fired, just shown to the attacker who runs away, it would be really hard to imagine the statistics would support your claim.
Quote
October 7, 2009, 1:05 pmOtter says:
The facts of history back up the data from this study. During WW2, the jews were unarmed and very few were shot. But they all were gased. Your choice, shot or gased. I’ll take my chances by being armed. I may be shot, but I’ll take many of them with me.
Quote
October 15, 2009, 12:34 pmRichard says:
In 1982, I lived near University of Chicago and worked there. I was a jogger and a “night person”. I often jogged along the lakefront; hardly anyone bothered me because, to do so, they would have had to run as fast as I could run. All I had to do was speed up and they lost interest. I also carried a gun illegally. One night in December, 1982, I jogged up on five thugs as they dragged a screaming woman into the wooded area across from the Museum of Science and Industry, where the U-boat is now. I pulled the gun and confronted one, the lookout. I fired once but did not hit him. The blast and flash scared all of them off. I stuck my gun back in the holster and jogged home, very quickly. They thought I was chasing them and ran south as fast as they could. The incident was never reported, as most similar ones are not. No one was hurt and no one stuck around for the cops; thus, no statistics. If someone had called me for a poll, I would have assumed it might be the cops trying for an easy confession. Of course, had I not had the gun I would have been killed or seriously injured. Many U of C students and faculty might be cowards, but I am not. Indeed, I met many who were quite aggressively loud and proud of their own cowardice.
Quote
October 18, 2009, 9:38 amGaviota says:
Issac Asimov was a man of many talents. Famous principally for his prodigious body of science fiction literature, he began his career in chemistry, in which field he held a PhD. Back in those days, in order to work at the PhD level, a student literally had to become a scientist and produce work which added to the body of knowlege in that field. Asimov believed in his core that the ethical standards of a true scientist hold him to the pursuit of good, honest, and open science. He reserved his most venomous hatred for those scientists, particularly chemists, who perverted their science for unethical gain.
The five people who authored this study, led by Charles C. Branas, PhD, Associate Professor of Epidemiology at Penn, have done exactly what Issac Asimov hated most. They took money from the Joyce Foundation, a notoriously generous funder of anti-gun studies, and they sold their skills, talents, education, training, experience, and morals to the Joyce Foundation to produce an anti-gun study, and in the process, they deliberately perverted their scientific work for monetary gain.
In a just society they would be properly recognized and rewarded for their work. After all, people who sell themselves for money are not called scientists.
Gaviota
Quote
October 20, 2009, 7:13 pm