Ari Armstrong's Web Log (Main) | Archives | Terms of Use

Gun Policy after Uvalde

"Assault" guns, rifles, confiscation, background checks, suicide, liberty vs. security, and more.

Copyright © 2024 by Ari Armstrong
June 6, 2022; ported here on June 4, 2024

In my last article here, "A Swiss Cheese Approach to Curbing Gun Violence," I listed my main proposals for addressing gun-related violence. Here I continue the discussion about guns, violence, and gun policy.

What's an 'Assault' Gun?

Recently I Tweeted,

Perhaps journalists would be a bit more curious about terminology if conservatives promoted laws to ban "assault abortions." Perhaps it would occur to them to ask, and report, what that means. (I'm not saying that never happens, but it often does not.)

An article on proposals of various Boulder County cities, which discusses "banning the sale and possession of assault weapons," prompted my Tweet. I've seen many such articles that leave the matter vague.

In response, someone posted a 2018 CNBC article, "Definition of what's actually an 'assault weapon' is a highly contentious issue." This shows (as I allowed) that sometimes journalists do take the matter seriously. Yet the article proves my broader point, that "assault gun" (or "assault weapon") is a made-up category with arbitrary boundaries. There is no substantive difference between a semi-automatic "assault weapon" and any other semi-automatic gun of the same caliber that accepts magazines. (If we limited "assault gun" to full-autos, things would be different. But these days no one does that.)

Of course, the ambiguity is a feature, not a bug, for many gun-control advocates. The entire point is to start banning some semi-automatics with the intention of banning all semi-automatics down the road. Once some guns are banned, they'll declare, "It so happens that these other semi-automatic guns operate the same way as 'assault guns,' so we should ban these other guns, too."

Colorado Newsline reports that Congress is considering a variety of gun-control measures, but an "assault weapons ban not part of the package." One measure would restrict magazines to ten rounds (with a grandfather clause). Although I do not favor such magazine restrictions, at least there is an argument for them. There's some reason to think that restricting magazine capacity might slow down a mass-murderer a little. At least magazines are easily defined and easy to recognize. Unlike proposals to ban "assault guns," magazine restrictions do not rely on ambiguity and obviously arbitrary lines. Sure, the number of rounds permitted is arbitrary, but at least magazines are not treated differently depending on cosmetic features.

As I've mentioned, I'd be willing to consider certain restrictions on all semi-automatic guns—or, better, all semi-automatic guns larger than a .22 that accept external magazines. For example, I think age limits to buy such guns (with exceptions) probably make sense.

Right now, U.S. law severely restricts full-autos but basically treats all semi-autos the same as any other gun. That doesn't make sense. Unless you think government should not restrict full-autos either, there's no reason to think government should recognize only two categories of guns (full-auto and everything else). Obviously a semi-auto AR-15 with large detachable magazines is different from a single-shot hunting rifle or a six-shooter.

Is the AR-15 a 'Weapon of War'?

Paul Eaton claims, "For all intents and purposes, the AR-15 and rifles like it are weapons of war." What's his case? It's similar to other guns commonly used in the military. Unlike military-grade weapons, Eaton concedes, the "AR-15s can only shoot single shot" (one round per pull of the trigger). But people in the military often use their select-fire rifles to fire single shots. Hence, the AR-15 is a "weapon of war."

The problem with Eaton's case is that it applies to all larger-caliber semi-automatic rifles that accept detachable magazines, including many guns commonly used for hunting. (Some people do use the AR-15 for hunting, although usually not for large game.)

Civilian-owned AR-15s are not "weapons of war" simply because their owners do not use them and do not intend to use them for war. They intend to use them for target practice mostly, for emergency home defense, and for general readiness against large-scale threats. Sure, an AR-15 could be used as a "weapon of war," as could practically any other gun. But civilian-owned AR-15s are not used that way.

True, the AR-15 as typically configured fires a relatively "high velocity" round, because it uses a small-diameter .223 bullet packed with quite a lot of gunpowder. By comparison, the .22 bullet is nearly the same diameter but shorter and shot with a lot less gunpowder. Then the velocity is a matter of physics. If you pack the same amount of gun powder behind a larger bullet, the larger bullet will move slower.

This image from Wikipedia shows a .22 LR (long rifle) round next to a 5.56 Nato round, which is close to the same size as a .223 round. A typical large-game hunting round would be something more like the .308 (fourth from the right).

If you want to limit magazine capacity, that at least is a coherent position. If you even wanted to limit the amount of gunpowder packed into a .223 round—still more than adequate for Ken Buck's raccoons!—I would oppose that, but that too would at least be a coherent policy. Saying you want to ban AR-15s because they are "weapons of war," unless you also want to ban larger-caliber semi-automatic hunting rifles, is not a coherent position.

Rifles Versus Handguns

Remember when the political focus was on handguns? The Brady Center used to be called Handgun Control, Inc. Now it seems like all anyone can talk about are AR-15 rifles.

Pew reports:

In 2020, handguns were involved in 59% of the 13,620 U.S. gun murders and non-negligent manslaughters for which data is available, according to the FBI. Rifles – the category that includes guns sometimes referred to as "assault weapons" – were involved in 3% of firearm murders. Shotguns were involved in 1%. The remainder of gun homicides and non-negligent manslaughters (36%) involved other kinds of firearms or those classified as "type not stated."

Robert Gebelhoff is upfront about the aims of sales bans on select semi-automatics:

Based on the evidence we have, banning these weapons probably won't do too much to curb overall gun deaths. We know this because in 1994, Congress passed legislation to outlaw the sale of certain types of semiautomatic guns and large-capacity magazines, and the effect was unimpressive. Gun homicide rates declined during the ban, but they also continued to fall after the ban expired in 2004. One federally funded study of the ban found that the effect on violence was insignificant, partly because it was full of loopholes.

But banning so-called assault weapons was never meant to reduce overall gun deaths. It was meant to make America's frustratingly common mass shootings less deadly — even if these horrific events represent a small portion of gun violence.

And, in fact, mass shooting casualties dipped during the ban, although a review of studies by the Rand Corporation found the role the ban played in the dip to be inconclusive. . . . [E]xperts say focusing on reducing large-capacity magazines might be more effective.

I think these considerations support some age restrictions on sales of semi-autos and of magazines.

'No One's Coming for Your Guns'

Gun owners are crazy—absolutely bonkers!—to think some people want to send out heavily armed federal agents to confiscate their guns. Noah Smith thinks gun owners worried about this are suffering from racist fantasies. Michael Shermer says it is a lie to say "the left is coming for your guns." Except that many people say that is precisely what they want to do.

Beto O'Rourke promised, "Hell yes we're going to take your AR-15." The New York Times looks to Australia's gun confiscation program as a model for the U.S. The Times also salivates over a Canadian proposal to confiscate more guns. New York Governor Kathy Hochul wants to confiscate various gun magazines. I've seen many variants on Twitter of the confiscation theme, including the remark, "We need to use state power to take peoples AR15s by force."

The goalposts have shifted. To me, it's a Pretty Big Deal if the federal government attempts to confiscate any guns that people currently legally possess. But some people now pretend that, so long as the federal government does not seek to confiscate all guns from all people, there's nothing to worry about.

Jay Andersen takes this attitude: "Gun advocates will continue to gripe that the 'liberals' want to take their guns away. No. No one is suggesting you can't own a hunting rifle, shotgun or pistol for sport or self-protection." But he wants to ban possession of AR-15s. And what about other semi-automatic rifles of equal or greater caliber with detachable magazines? He doesn't say, but he indicates those should be banned too. Where exactly he'd draw the line is unclear.

Worrying about people trying to confiscate your guns, when they tell you that's precisely what they hope to do, is hardly irrational.

Background Checks as Registration Schemes

As noted, I endorse a sort of background check system that does not involve the government collecting records on gun owners. What about the system we currently have?

When Colorado extended its background checks to gun shows years ago, I had two main complaints. First, the check system sometimes delays sales, and that can prevent a person being threatened from purchasing a gun for self-defense when needed. Second, the current system results in partial gun registration.

Sure, the federal government is not supposed to keep permanent records of all sales. Thankfully, federal agencies never illegally collect records.

Most gun records collected by gun stores eventually end up in federal hands. The ATF summarizes:

When an FFL discontinues business, the FFL must send their firearms transactions records to the National Tracing Center (NTC). The NTC receives an average of 1.2 million out-of-business records per month and is the only repository for these records within the United States.

Records can be mailed to the NTC or, alternatively, they may be delivered to your local ATF Office in order to comply with laws for surrendering records (which include all bound log books/acquisition & disposition books and computer printouts, ATF Form 4473's, Theft/Loss Reports, Multiple Sales Reports, and Brady forms).

The Giffords group explicitly wants to convert the current background check system into a system of universal gun registration. Then it is but a short step to calls for confiscation.

Social Contagion

Researcher Jillian Peterson says, "[M]ass shootings are socially contagious and when one really big one happens and gets a lot of media attention, we tend to see others follow."

Peterson's colleague James Densley adds, "Mass shooters study other mass shooters. They often find a way of relating to them, like, 'There are other people out there who feel like me.'"

Paradoxically, the media dynamics of these stories are driven partly by the ongoing background political debate over guns.

Dylan Matthews has more on copycats.

Stats on 'Children'

Can we acknowledge that a child dying in a mass-shooting at a school is a basically different phenomenon than a young adult gang member dying in gang-related violence? I have seen various write-ups that intentionally conflate these categories.

A CNN headline states, "More US school-age children die from guns than on-duty US police or global military fatalities." So . . . we're talking about children dying at school? No. Many of the victims are adults, many are no longer in school, and most do not die at school. The relevant statistic involves "38,942 fatalities among 5- to 18-year-olds from 1999 to 2017." 32% of these deaths are suicides, a much different problem than homicide. 83% of the deaths involve "children [and adults] between 15 and 18 years old." Some of those victims are criminals with extremely dangerous lifestyles. For reference, at one point the Department of Justice found the "average age of the arrested gang offender is seventeen or eighteen years," which means many were even younger.

Every one of these deaths is a horrible tragedy. But we won't prevent tragedies by ignoring the various underlying problems.

Incidentally, does anyone really think that gang members are not going to be able to get whatever guns they want, regardless of the law? It's reasonable to think that gun prohibitions will limit the sale of black-market guns about as well as drug prohibitions have limited the sale of black-market drugs.

Mental Illness

Jim Sciutto writes, "Fact is, doctors say mass shooting are not principally a mental health issue. Data shows only 5% of shooters had a diagnosable mental illness."

First observe the double-standard here. What fraction of gun owners commit violent crimes with their guns? I'm not sure offhand, but the figure has got to be much lower than 5%. And yet many people today have no problem declaring "it's the guns" or demonizing gun owners.

Anyway, I replied,

It's pretty reasonable to think that many cases of mental illness go undiagnosed. Also there's a relevant broader category of poor mental health that probably encompasses many or even all mass murderers. Anyway, to say x can be a problem is not to stigmatize all people with x. . . . What we need to do is look at contributing factors.

Suicide and Statistical Mirages

Here's what the subhead of a 2020 Stanford Medicine article by Beth Duff-Brown states:

Men who own handguns are eight times more likely to die of gun suicides than men who don't own handguns, and women who own handguns are 35 times more likely than women who don't.

What many people take this to mean, but what it does not mean, is that if a normal/typical person buys a gun, the risks of suicide for that person go up by the stated amounts. If a non-suicidal person buys a gun, the risks of that person committing suicide with the gun (or by any other means) remains the same, zero. (Of course, it's possible for a non-suicidal person to become suicidal later.)

Much of what the stated statistics are capturing are two different phenomena. 1) Sometimes suicidal people purchase a gun with which to commit suicide. 2) Sometimes suicidal people who possess a gun use the gun to commit suicide.

Duff-Brown quotes Matthew Miller, an author of the study under review:

New handgun buyers had extremely high risks of dying by firearm suicide immediately after the purchase. However, more than half of all firearm suicides in this group occurred a year or more later.

Here are the big raw numbers: "A total of 676,425 cohort members acquired one or more handguns, and 1,457,981 died; 17,894 died by suicide, of which 6691 were suicides by firearm." In other words, just under one percent of the total group committed suicide with a gun.

Just the fact that someone commits suicide with a long-possessed gun does not imply the person would not have committed suicide but for the gun. The person might otherwise have committed suicide by some other means. True, people are more "successful" at committing suicide with a gun relative to other means. Also true, people more-serious about killing themselves are more likely to choose effective means.

Gary Kleck thinks people can very easily substitute methods of suicide. I agree its possible for people to substitute other means of committing suicide, but, as a practical matter, I do think the availability of a gun can make a suicide more likely in many cases.

Take a look at the study that Duff-Brown reviews for its nuances and qualifications. (I plan to discuss various details of this study in a future post.)

The broader point is that it is wrong to violate the rights of non-suicidal people in order to try to bring down the suicide rate. Instead we should try to get suicidal people the help they need and try to discourage gun possession specifically by suicidal people.

Are Armed Guards at Schools Counterproductive?

Here is a surprising finding:

[A]rmed guards [at schools] were not associated with significant reduction in rates of injuries; in fact, controlling for the aforementioned factors of location and school characteristics [see paper], the rate of deaths was 2.83 times greater in schools with an armed guard present.

I'm worried that the study does not adequately account for the fact that schools at greater risk are more likely to hire armed guards. So maybe armed guards still make schools relatively safer, even if those schools remain more dangerous than other schools.

But, for sake of argument, let's assume that the paper is picking up on a real causal factor, and that armed guards really do make schools more dangerous. Why might that be? The authors write, "Prior research [see citation in paper] suggests that many school shooters are actively suicidal, intending to die in the act, so an armed officer may be an incentive rather than a deterrent."

Offhand I'm skeptical of this explanation. Maybe some school shooters are actively suicidal, but they also seek the global infamy that comes with murdering a lot of school children. If they just wanted to commit suicide, they could do that at home. And shooters are virtually guaranteed "suicide by cop" if they wait for the police to show up (which, as we've seen, could take quite a long time) and then shoot at police or refuse to drop their gun. So I doubt that any school shooter is drawn to a school specifically because they think a security guard might shoot them.

Still, this is an interesting paper, and perhaps future research can shed more light on the topic.

We Trade Safety for Liberty All the Time

Here is a partial list of the many ways that we routinely, and rightly, trade short-term safety for liberty. Of course, in the long term, maintaining our liberty does enrich our safety in important ways. Still, if we look at obvious and immediate policy implications, it is obvious that we gladly trade safety for liberty all the time.

If you want to claim that there are trade-offs between liberty and security, fine. If you want to pretend that liberty is irrelevant to the gun discussion but relevant to all those other discussions, you're not being serious.


The NRA is a total disaster, having torn itself apart with internal corruption and sold its soul to Trump. Zach Despart summarizes the views of many NRA convention attendees:

[T]hey attributed this attack and others to a broader breakdown in society wrought by the removal of God from public schools. . . . They described feeling ostracized for their beliefs, and not just those on guns. For their refusal to get the COVID-19 vaccine. For their objections to gay people serving as teachers. For their belief in disciplining children through spanking.

To clarify: Other, more-secular societies are much safer. If you refused to get a Covid vaccine, unless for some medically recognized reason, you are, objectively, an idiot. There's nothing wrong with gay people teaching kids. Spanking is wrong and bad.

Still, we shouldn't paint all NRA members with the same brush. Dave Kopel sticks up for the NRA.

The Uvalde Police

According to one officer on the scene of the Texas shooting, some officers entered the school to retrieve their own kids but did not engage the shooter. Although facts remain sketchy, signs so far point to a horribly botched police response.

Here is what CNN reports:

The 18-year-old gunman was not confronted by a school resource officer outside the building, officials now say, and he apparently entered through an unlocked door. Police officers who arrived and confronted the shooter called for more resources, an official said. After about an hour, tactical teams forced entry and fatally shot the suspect, officials said as questions about the timeline continue to surface.

The mother who rushed into the school after being briefly cuffed by police is not happy about the police response.

Police have not always done a great job in other active-shooter emergencies. After Columbine there's no excuse for that.

Quick Takes

Swearer: Amy Swearer offers a general critique of gun control proposals. She points to the practical problems with trying to put in place a stepped approach to implementing the age of adulthood.

Shermer: Michael Shermer has re-released his case for gun control. Unfortunately, he continues to rely on misleading studies that undercount defensive gun uses and that confuse causation regarding guns in the home by not adequately accounting for underlying personal characteristics. E.g., a gangster who buys a gun is at an elevated risk of being involved in a homicide; a normal person is not. Shermer also discounts the importance of an armed population as a bulwark against invasion and domestic tyranny. These are complex matters beyond my scope here.

Schools: Kirk Cameron and C. Bradley Thompson are among those blaming schools themselves for school violence. I think there are some causal forces along these lines at work in some of these cases but that we also have to look at all the mass shootings that do not involve schools.

Trump States: Ian Silverii points out that Trump states have higher rates of gun deaths (on average) than Biden states. But this doesn't mean much without the suicide breakouts and a look at crime in cities versus rural areas.

Porn: This is stupid: "Ohio Republican Senate nominee J. D. Vance has called for a total ban on porn—he suggests porn is responsible for mass shootings, along with abortion." This yet again illustrates that many of today's Republicans care nothing about liberty or the Bill of Rights.

Dildos: Yes, Texas restricts the possession of dildos. Yes, that's stupid. Yes, many Republicans are a bunch of hypocrites.

Security: Alex Tabarrok warns against turning schools into prison-like facilities. Matt Bateman has similar concerns. Samuel Sinyangwe offers good reasons to worry about police in schools.

Self-Defense: As Eugene Volokh notes, sometimes "good guys with a gun" (including gals) really do stop mass murderers.

Red-Flag: If people threaten to commit acts of violence, as the Texas murderer did, people need to report them to the police, and police need to initiate action to take away their guns under red-flag laws.

Culture: Here is a shocking and uncomfortable statistic: "Black males ages 15 to 34 were over 20 times more likely to die by gun homicide than their white counterparts" in 2020. We can entertain two explanations for this: Systemic racism and pathologies within certain subcultures. I think both things are at play. One thing the "it's the guns" narrative ignores is that it very much matters who has the gun.

Fathers: They matter. Kids can grow up without fathers and turn into wonderful people, but life generally is easier and better in a stable, two-parent family (regardless of the gender of the parents).

Ari Armstrong's Web Log (Main) | Archives | Terms of Use