The tragic shootings at Virginia Tech have already led some to ask whether more gun control — or more private gun carrying, including at universities — would help avoid such crimes in the future. They have also led some (for instance, Eric Muller (IsThatLegal?)) to fault those who are publicly discussing such policy responses so soon after the deaths.

It seems to me clear that such discussions are generally sound, even worthy. Using the attention created by a tragedy to try to prevent similar tragedies strikes me as in principle an eminently proper response, a way to allow at least some good to come from the evil. Preventing the tragedy from leading to unsound reactions likewise strikes me as an eminently proper response. (Complaints that legislative proposals triggered by the tragedies "politicize" the tragedies thus strike me as misguided, though of course complaints that particular proposals are practically or morally unsound may be eminently sensible.) But the question is whether we should pause before engaging in such discussions; in Eric Muller's words, "Let's wait at least a day before trying to score political points, shall we?"

I'm not sure what the answer is, but I thought I'd pose the question here (hoping that at least there's nothing wrong with using the tragedy as an occasion for asking this meta-question). I don't think the answer is clearly "yes, wait," the way it is as to critical obituaries of writers whose work one dislikes; responding to death using unpersonalized policy discussion is different from responding to death using personalized criticism of the dead person. On the other hand, I don't think the answer is clearly "no, go ahead," at least as a matter of first principles; perhaps we ought to have a social ritual of grief and condolences first, policy analysis (even of the most cerebral sort) later, and perhaps the very immediacy of the tragedy may lead to unsound first thoughts about the policy questions.

One extra piece of the puzzle: Even if we think that in the abstract the right approach would be to wait a day, should our analysis change because others will surely start talking about legislative responses right away?

The Brady Campaign, for instance, responded quickly (at the latest by 3:30 pm Eastern time the day of the murders) with condolences coupled with a call for more gun control. [UPDATE: So has the Violence Policy Center, as of 7 pm Eastern or earlier.] I first learned of the incident when I got a call from a French news agency that wanted to ask me all about American gun controls, and how the tragedy would likely affect the political debate; I expect that American news outlets will likewise discuss this in the coming hours.

Should this reality, coupled with the plausible expectation that there will be many pro-gun-control sentiments expressed even today, lead pro-gun-rights forces to speak up at the same time as the pro-gun-control forces are? Or would that just be practically counterproductive, as well as in bad taste (assuming that one thinks as a matter of first principles that talking about legislative responses right now is indeed in bad taste)? Two wrongs don't always make a right, but sometimes the right answer for one side is indeed altered by what the other side is doing. (That's why, for instance, advocates of campaign finance reform might both (1) prefer that all candidates fund their campaigns only using small donations, but (2) when their adversaries are getting big donations, conclude that it becomes proper for pro-reform candidates to seek out such big donations, too, at least until bilateral disarmament is achieved.)

In any event, I thought I'd pose this question, and see what our readers thought.

I'd heard some suggest that mass shootings are more common now (in the post-Columbine era) in the U.S. than before; does anyone have some data as to whether this is so?

I have one seemingly reliable piece of data handy — the list in Gary Kleck's Targeting Guns, p. 144, reports on all the mass shootings Kleck knew of from 1984-1993, with mass shootings "defined here, somewhat arbitrarily, as an incident in which six or more victims were shot dead with a gun, or twelve or more total were wounded" (pp. 124-25). That list reports 15 mass killings, roughly evenly distributed from 1984 to 1993. (For those who want to check for completeness, the murderers are Ferguson, Ferri, Hennard, Doody & Garcia, Abeyta, Pugh, Wesbecker, Purdy, Farley, Simmons, Schnick, Cruse, Sherrill, Huberty, and Thomas.) My sense is that the frequency has not gone up materially since then, though I should note that this is just based on my likely quite faulty memory.

On the other hand, only one of those shootings (Purdy, in Stockton) was at a school, and it did not involve a student, unlike the Columbine murders and some of those that followed. My sense is that schoolyard shootings are indeed up since Columbine, but again I don't have handy data about how much. I'd also love to hear about data from before 1984; of course, Charles Whitman's murders in 1966, were at a university, but I do not know of any pattern of school or university mass shootings after that. (I would bracket the 1970 Kent State shootings, simply because they seem so radically different in motivation from the other killings that it's hard to see what sound policy analysis one could engage in that would group these shootings together with the other shootings I mention.)

UPDATE: A Better Where To Find has a long list, not claimed to be complete, of multiple-victim shootings, though with a somewhat different selection criterion than that given above, and limited to schools.

It appears that the Virginia Tech murderer was a resident alien. Did this mean someone broke the law by selling him the guns? (Naturally, the killer showed himself to be unconcerned about following the law, but the premise of some gun control laws is that they may deter generally law-abiding people from selling guns to those who might be dangerous.)

I don't think so. Federal law bars most nonresident aliens, including illegal aliens, from possessing a gun, but treats resident aliens the same as citizens. Virginia law likewise doesn't bar gun ownership by resident aliens, and even allows them to get concealed carry permits.

Some states or cities have banned gun ownership by aliens, or at least aliens who hadn't declared an intent to become citizens. (See here for my criticism of one such law, in Omaha.) Such laws may violate the federal Equal Protection Clause, by discriminating against noncitizens without adequate justification; but the one case I've seen on the subject, State v. Hernandez-Mercado, 124 Wash. 2d 368 (1994), rejected such a challenge.

The laws may also violate state constitutional provisions, if the provision isn't limited to citizens. The Washington Constitution's provision is limited to citizens; but the Michigan Constitution is not — it says "Every person has a right to keep and bear arms for the defense of himself and the state" — and the Michigan Supreme Court accordingly struck down a ban on gun possession by noncitizens in People v. Zerillo, 219 Mich. 635 (1922). (At the time, the Michigan provision differed slightly — it omitted "keep and" — but not materially.)

The Utah Supreme Court upheld a ban on gun possession by noncitizens in State v. Vlacil, 645 P.2d 677 (Utah 1982), despite a constitutional provision stating, "The people have the right to bear arms for their security and defense, but the legislature may regulate the exercise of this right by law." It is possible (though in my view mistaken) to read "the people," as opposed to Michigan's "every person" or Nebraska's "all persons," as referring to individuals — which Utah courts have done as to the right to bear arms — but only to individuals who are full members of the American polity. But the Utah Supreme Court didn't even do that; it simply cited to State v. Beorchia, 530 P.2d 813 (Utah 1974), which reasoned that "It is quite evident from [the constitutional text] that the Legislature had sufficient power to enact thestatute in question." In my view, it is not at all evident that the power to regulate the exercise of a right includes the power to deny the right to whatever group the legislature chooses.

If banning gun ownership by noncitizens is constitutionally permissible, is it a good idea? I don't think so. It seems to me that resident aliens, at least, and perhaps legal but nonresident aliens, are just as morally entitled to try to defend their lives against crime as citizens; their ability to do so is just as valuable to society as citizens' ability to do so; and they seem no more likely than citizens to use the guns to cause harm. Moreover, in a nation with over 200 million firearms in private hands, it seems highly unlikely that the rare noncitizen with mass murder on his mind will indeed be stymied by laws banning gun possession by noncitizens, and will be stopped from getting a gun on the black market. That unlikely possibility of social benefit is substantially outweighted by the cost of denying millions of law-abiding noncitizens of their ability to effectively defend themselves.

Please correct me if I've misunderstood the often quite complicated federal and state gun laws that govern resident aliens; but I'm pretty sure I read them correctly.

UPDATE: A correspondent informs me that "there is one difference between treatment of citizens and resident aliens. Resident aliens have to provide extra identification to prove 90 days' residence in the state. See 27 CFR sec. 478.124(c)(3)(ii)." I appreciate the extra information, though it does not much affect the big picture analysis. (One could argue against even this requirement, given that it creates a fairly long waiting period for buying a gun in certain situations; but this nonetheless seems a relatively small detail relative to the broader question of whether non-citizens should be able to own firearms.)

Professor Liviu Librescu:

David Bernstein mentioned him below, but I thought this was worth a picture, and a longer excerpt, this one from an AP story:

Relatives said Liviu Librescu, an internationally respected aeronautics engineer and a lecturer at the school for 20 years, saved the lives of several students by barricading his classroom door before he was gunned down in Monday's massacre ....

Librescu' students sent e-mails recounting the last moments of their teacher's life to his wife, Marlena, his son, Joe, told The Associated Press on Tuesday.

"My father blocked the doorway with his body and asked the students to flee," Joe Librescu said in a telephone interview from his home outside of Tel Aviv. "Students started opening windows and jumping out." ...

When Romania joined forces with Nazi Germany in World War II, [Librescu] was first interned in a labor camp in Transnistria and then deported along with his family and thousands of other Jews to a central ghetto in the city of Focsani, his son said. According to a report compiled by the Romanian government in 2004, between 280,000 and 380,000 Jews were killed by Romania's Nazi-allied regime during the war....

As a successful engineer under the postwar Communist government, Librescu found work at Romania's aerospace agency. But his career was stymied in the 1970s because he refused to swear allegiance to the regime, his son said, and he was later fired when he requested permission to move to Israel.

After years of government refusal, according to his son, Israeli Prime Minister Menachem Begin personally intervened to get the family an emigration permit....

In the late 1980s, Librescu moved from Israel to America.

Ross Douthat at Andrew Sullivan's blog writes (paragraph break added):

Eugene Volokh wonders how soon is too soon to start the inevitable post-Virginia Tech dialogue about gun control, and Joshua Claybourn chimes in. Obviously, this kind of meta-debate is somewhat academic, since nobody -- from the New York Times editorial page to Michelle Malkin -- seems interested in waiting even a day before trotting out their hobby-horses.

I'm extremely skeptical, though, that there's actually anything significant to learn about gun policy from yesterday's violence: Extreme, unpredictable events like this one seem like precisely the kind of thing that shouldn't dictate lawmaking decisions (though of course they inevitably do). If there's a case for gun control, it's in the daily run of shooting deaths that don't make the front page; if there's a case against gun control, it's in the daily run of crimes deterred by an armed citizenry (and in more abstract questions of personal liberty), not in the faint chance that a kid with a conceal-and-carry permit might have taken the Virginia killer down.

I would add "stopped" to "deterred"; and I would agree that, even if one thinks that either gun control or gun decontrol would have helped in this instance, we shouldn't make broad gun policy based on these highly unusual incidents -- which, tragic as they are, represent a tiny and extraordinarily unrepresentative fraction of all the homicide that's out there. (This of course is a separate question from whether it is improper or disrespectful to discuss policy questions immediately following a tragedy such as this one.)

That's basically what a foreign reporter asked me today, suggesting that the answer must be some new gun control proposal. After all, if someone murdered all these people with a gun, what is America going to do about guns? After past shootings, I got similar questions. Likewise whenever I do talks or debates about gun controls: OK, people say, you say these various gun controls don't work; so what do you propose to do instead, given that there's an undoubted problem out there to be solved?

Well, it turns out that yesterday, about 25 to 40 people were killed in alcohol-related homicides, not including those that died because of their own alcohol consumption. Each year, between alcohol-related drunk driving deaths and alcohol-involved murders, about 10-15,000 Americans (not including the responsible drunk drivers themselves) die. That translates into roughly 25 to 40 deaths per day (the range is wide because the source numbers are necessarily back of the envelope estimates), about the same number as the extra homicide deaths from yesterday's mass shootings. If you counted people whose alcohol consumption killed themselves, the total would likely be far more.

So what are we going to do about it? When are we going to ban alcohol? When are we going to institute more common-sense alcohol control measures?

Well, we tried, and the conventional wisdom is that the cure was worse than the disease -- which is why we went back to a system where alcohol is pretty freely available, despite the harm it causes (of which the deaths are only part). We now only prohibit alcohol abuse, generally allow alcohol purchase and possession, and regulate alcohol purchase and possession fairly lightly. Some of the regulation, such as bans on sales to minors, are quite likely wise (at least as applied to minors; I express no opinion on bans on sales to 18-to-21-year-olds), though imperfect. Others, such as bans on Sunday alcohol sales, are pretty clearly unwise. Others are closer calls, but on balance the answer to "what are we going to do about alcohol-related deaths?" is "not much, other than trying to catch and punish alcohol abuse."

Now the likely pathologies of gun prohibition -- or even of many regulations that fall short of prohibition -- would probably differ in some ways from the likely pathologies of alcohol prohibition. I've talked of some of those likely pathologies elsewhere, but this post is not about that. My point is simply that the right answer to "so what are we going to do about it?," even when the "it" is horrible, is sometimes "not much."

We should certainly consider proposals that aim to ameliorate the problem, and weigh their costs and benefits. But we should not presume that there's somehow a moral imperative to Do Something. In fact, there's a moral imperative not to do something that's likely to make matters worse.

Consider two scenarios:

1. An armed madman comes to a place and starts shooting people. None of the people who's around is armed.

2. An armed madman comes to a place and starts shooting people. Several (say, five) people in the vicinity are armed.

Which madman is more likely to be stopped quicker — the one who outguns everyone else 1-0, or the one who is outgunned 5-1?

If this weren't a madman but Jack Bauer — or even an average highly trained soldier — the five may well be unable to stop the one. But otherwise, the odds would seem to be more against the madman in situation 2 rather than 1, no?

No-one can prove anything, of course. Maybe the five would be the first to be shot. Maybe they'd run away. Maybe they wouldn't be around. Maybe they'd shoot and miss. Still, if you had to bet, which would you bet would be the worse scenario for the madman, and the better one for his victims?

Now of course if arming the five people for the extremely rare situation when they'll need to stop a madman will end up causing more harm than good in the much more common situations when there's no madman around, that might be a bad tradeoff. That is the argument I've heard against letting students possess weapons on-campus: They're young, they drink a lot, they'll start shooting when they get into a hot argument in class or at a debate. I'm not sure that's right, but let's say it is.

What, though, is the argument against allowing professors and other university staff to possess weapons, if they choose? (Assume the professors lack criminal records, and assume they go through whatever testing and modest training is required to get a concealed carry permit, or perhaps even some extra training.) One argument is that it's just dangerous for law-abiding citizens to have weapons, because they'll start shooting over arguments or fender-benders. But that's precisely the argument that has been rejected by the 38 states that allow any law-abiding citizen to get a concealed carry license (or, in 2 of the 38 states, to carry without a license). What's more, as I understand it, people who get such licenses have in fact almost never committed unjustified homicide or attempted homicide (or even lesser crimes) using their guns. Whatever the pluses or minuses of shall-issue, the "licenseholders will start shootouts over petty slights" theory has not been borne out.

If Virginia and other states have found that it's safe to let law-abiding citizens carry guns on streets, in shopping districts, in parks, and the like, why wouldn't it be equally safe to let law-abiding professors and staff to carry guns in the university? What magic is there about a university that makes guns in law-abiding citizens' hands (again, let's even set aside college students, if we think they are unusually likely to behave foolishly) more dangerous at a university than elsewhere? I know there are some university professors who are, er, a bit odd. But wouldn't the average professor — or average university employee generally — who wants a concealed-carry license to carry on campus be at least as responsible as the average citizen who wants a concealed carry license to carry outside campus? Given that licensees don't start shootouts over fender-benders, and that gun store employees, police officers eating lunch, and other law-abiding people who are routinely armed don't start shootouts over arguments, why should we think that armed professors (to be precise, that small group of professors who chooses to get concealed carry licensees) would start shootouts at faculty meetings?

What then is the downside? One possibility is that if mad killers know that professors and staff may be armed but students won't be, the killers will shoot the professors first. It's hard to see why this would increase the total death toll, though, especially in cases such as this one. I doubt that the typical mad killer who's willing to shoot any university employees he comes across just in case they have guns would have spared them, and just shot a selected subset of unarmed students (and staff and faculty), if he assumed the employees were unarmed. And in any event, it seems to me that this modest risk is worth running, just as the risk that armed security guards would be shot first is worth running in order to provide the protection that armed security guards might offer.

Another possibility is that the mad killers would just start shooting lots of people very quickly, rather than at the relatively leisurely pace that we've seen in many mass killings. "I was going to walk around killing people over 15 minutes or half an hour," the killer might think (Kleck's Targeting Guns reports two mass killings that happened over 2-4 minutes, but seven that happened over 10 minutes or longer, including five that took 30 minutes or longer) — "but since there's now a substantial chance that I'll be stopped, I'll just shoot as many as I can as quickly as I can." But this strikes me as pretty unlikely; the pace of killings seems to be driven more by the killer's own mad desires rather than by a calculation such as this one.

Another concern might be that the universities would be held liable for their employees' misuse of guns. But, first, I take it universities already have liability insurance policies for possible misuse of weapons by university police officers. The employees who get concealed-carry permits (likely a small fraction of all employees, given that in shall-issue states the general pattern is that only a small fraction of all citizens get licenses to carry) could be added to such policies.

Given the pattern of safe use of guns by shall-issue licensees, I suspect insurance companies won't demand vastly higher premiums from universities for this. And the university could require some extra screening, testing, and training just to make sure that employees who get permits are as reliable as possible. Recall that many armed security guards are trained and screened only modestly, and they are likely to use their weapons more often (since they tend to guard places during high-crime times of day, and are more likely to be called over to the scene of developing crimes, which other armed university employees would rarely have to do). And if one thinks my analysis above is correct, and carrying by employees is pretty safe, but the liability system is unlikely to reflect this sound analysis, then some statutory liability limit — perhaps simply providing that professors and staff are armed on their own behalf, and the university thus wouldn't be liable for their actions — may be sensible.

I've also heard some arguments that suggest universities are different because they are places for reasoning, not violence: They should be gun-free zones (except of course for university police officers and security guards, who for some reason don't count) because that's needed to create the proper climate of peaceful inquiry. But the sad fact is that you can't make a university into a gun-free zone. Mad killers can bring guns, and use them, regardless of what policies you announce. The question is whether they will be able to use them against a disarmed population, or against a partly armed population. Allowing people the tools to defend themselves against the mad killers does not, it seems to me, worsen the climate.

[UPDATE: Finally, to give an even more essentialist version of the argument that universities are somehow unsuitable places for professors to be armed, let me quote a commenter: "There are a million reasons, but one will suffice for now. Schools and guns do not mix. Period. The more guns we inject into a school environment the worse it is for all involved...." It's hard to figure out how to deal with the argument that "Schools [including universities] and guns do not mix. Period." The commenter also mentions that professors "aren't trained for it, and I doubt they would be very good at it," but 38 states have taken the view that all law-abiding adults should be free to get a license to carry concealed weapons, and haven't had serious problems from the concealed-carry holders; the question is why those problems would suddenly appear at universities.]

Again, I should stress that this is not supposed to be a panacea, some guarantee that mad university killers will be immediately stopped. And I should say, as I've noted before, that there may be little point in planning for fortunately very rare events such as this one, which account for a tiny fraction of all homicides in the country. (The yearly average is that mass killings account for less than 0.1% of all U.S. homicides.)

But if we are trying to think what could have decreased the carnage, I've come around to the tentative view that allowing at least university employees to be armed is the likeliest solution. When one person — a not very well-trained person — comes into a place occupied by thousands of people, and kills over thirty, there's a simple explanation for why he could get away with it: Though he was outnumbered, they were outgunned.

If all university professors (and other employees) could be armed, and therefore some would be armed (again, I'm not saying that everyone will be armed, much less will be required to be armed, but only that some will exercise their right to get a concealed carry license allowing carry on campus), he would be both outnumbered and outgunned. Why wouldn't there be a very good chance that they would therefore be able to stop him earlier than he would have stopped otherwise?

Bloomberg News reports:

Illinois Representative Rahm Emanuel, chairman of the House Democratic Caucus, said the top priority of his party's lawmakers is hiring more police to fight crime, not tougher gun control.

Emanuel said the House "might" or "might not" re-enact an assault-weapons ban that expired in 2004. That legislation, which limited the capacity of handgun magazines, would have reduced the amount of ammunition used in a shooting rampage that killed 32 people this week at Virginia Tech University.

Setting aside the various other questions raised by assault-weapons bans, how can a news service say with a straight face that legislation limiting the capacity of handgun magazines would have reduced the amount of ammunition used by the murderer?

Recall that semiautomatic handguns are reloaded by popping out a magazine, popping in a new magazine, and chambering another round. If the shooter has preloaded several magazines — which the Virginia Tech murdered had — the process can take a second or two, even with no special training.

Banning 15-round magazines (which the Virginia Tech killer apparently had) and limiting magazines to 10 rounds — as per the expired assault weapons ban — or even to 6 rounds, would thus simply require the shooter to reload a little more often. This may limit a shooter who is in the middle of a firefight when one is shooting very quickly (6 to 10 rounds in a few seconds), but not a shooter engaged in a mass shooting such as this one, which took place over many minutes. It doesn't seem even very plausible that a smaller magazine size would have led to fewer shots being fired. It is certainly wrong to say that it would have reduced the number of shots (even if one recognizes that "would have" represents very high probability rather than just certainty).

This is a classic policy analysis mistake, but one that I've found particularly common in gun control debates: assuming that when one enacts a law, that will change the subjects' behavior in the way the law contemplates, but with no compensating substitution effects. Sure, if by reducing magazine size, we get someone to load 4 10-round magazines rather than 4 15-round magazines, he'll have fewer rounds he could readily shoot.

But why on earth would we think that this is how people will react? Why wouldn't they just load 6 10-round magazines instead of 4 15-round magazines? (Another classic policy analysis mistake is simply not knowing the technical details of the items that one is discussing; my sense is that many people, likely including many reporters, just don't know how quickly one can switch magazines on a semiautomatic, or don't even know precisely what a semiautomatic is.)

I should say that banning semiautomatics altogether, and requiring handgun users to rely on revolvers, might theoretically have more of an effect; reloading a revolver does take somewhat more time. It's not vastly more, and if one has a backup gun handy, one won't even be particularly vulnerable while reloading the revolver; and there are other problems with the proposal, including the political problem that the ban would affect weapons that are owned by tens of millions of gun owners.

But at least there'd be something potentially plausible to talk about there. There is, on the other hand, no credible defense for the claim that "[the] assault-weapons ban that expired in 2004 ..., which limited the capacity of handgun magazines, would have reduced the amount of ammunition used in a shooting rampage that killed 32 people this week at Virginia Tech University."

UPDATE: In the original post, I described the process of replacing a magazine as removing the empty one, loading a full one, and then possibly chambering the round, unless one reloads while there is still a round ready to shoot. On reflection, I realize that one would almost always wait until all the rounds have been used before putting in the new magazine, so I changed the post to say that replacing the magazine requires removing, loading, and chambering. The bottom line is unaffected; reloading can still take a second or two, without any fancy training.

From an opinion piece in The Economist, Apr. 21, 2007:

[The Virginia Tech killer] had two guns: a Glock 9mm and a Walther P22. Both are semi-automatic: they fire bullets as quickly as you can keep pulling the trigger.

Actually, the other dominant form of handgun -- a double-action revolver -- also fires bullets as quickly as you can keep pulling the trigger. The rate of fire from a revolver is, I'm told, slightly less than from a semiautomatic (I take it because in a revolver the trigger pull needs to do more than in a semiautomatic), but only slightly. One can certainly fire a revolver at least once a second with no extra training; it's not a good idea, since one generally won't be accurate with a revolver when firing rapidly, but one generally won't be accurate with a semiautomatic when firing rapidly, either.

Certainly for someone who is shooting at unarmed targets, and thus doesn't have to shoot several times a second -- and apparently the Virginia Tech killer was shooting at a relatively leisurely pace -- the difference between revolver rate of fire and semi-automatic rate of fire is negligible. (The difference in time to reload might be more significant in some situations, though again not in this one.)

Why does this matter? One common argument made by some gun control proponents, expressly or implicitly, is that they're just proposing modest restrictions on just a few guns. After all, it's politically easier to ban something that fewer people own than something that more people own. We're not trying to ban all guns, just so-called "assault weapons." We're not trying to ban all guns, just semiautomatics. We're not trying to ban all guns, just large-capacity magazines. And in the process of making such proposals, they have to explain why this particular kind of gun or magazine is especially deadly.

The trouble is that "assault weapons" aren't really materially deadlier than unbanned non-assault-weapons. Semiautomatic handguns aren't really materially faster-firing than revolvers. Bans on over-10-round magazines will almost never limit criminals, especially the sort of mass killers whom the gun control advocates are discussing. The proposals will do virtually nothing to reduce crime; while I agree that they're not nearly as burdensome to law-abiding citizens as total gun bans would be, they also aren't burdensome to criminals. These modest proposals will fail. And what will gun control advocates propose then?

(More aggressive bans, such as total handgun bans or total gun bans, might actually have more of an effect, both for good and for ill. I think on balance the ill effects will exceed the good ones, but that's a separate matter; at least there's something more than pure symbolism or misunderstanding behind them.)