‘Gun Safety’ Isn’t the Issue

I’m trying to get out of the habit of writing columns that are responses to other people’s columns, but this Nicholas Kristof piece on gun policy—full of sloppiness bordering on intellectual dishonesty—needs a response.

Kristof’s argument is that we can use “gun safety measures” to reduce violent crime in the United States. But most guns sold in the U.S. market are not lacking in safety features—the problem is not “gun safety” but the fact that people point perfectly functional, well-designed firearms at other people and then pull the trigger for the purpose of killing them. “Gun safety” is not the issue. Murder is the issue. Suicide is, by the numbers, an even greater issue. 

But that’s a view-from-35,000-feet issue. Kristof’s nonsense is worth addressing point-by-point.

Kristof writes:

Just as some kids will always sneak cigarettes or people will inevitably drive drunk, some criminals will get firearms — but one lesson learned is that if we can’t eliminate a dangerous product, we can reduce the toll by regulating who gets access to it.

That can make a huge difference. Consider that American women age 50 or older commit fewer than 100 gun homicides in a typical year. In contrast, men 49 or younger typically kill more than 500 people each year just with their fists and feet; with guns, they kill more than 7,000 each year. In effect, firearms are safer with middle-aged women than fists are with young men.

We’re not going to restrict guns to women 50 or older, but we can try to keep firearms from people who are under 21 or who have a record of violent misdemeanors, alcohol abuse, domestic violence or some red flag that they may be a threat to themselves or others.

The law already prohibits retailers from selling firearms to people who have most of the “red flags” Kristof mentions: people convicted of felonies, people convicted of serious misdemeanors including nonviolent crimes (“any other crime for which the judge could imprison you for more than one year”), people who have not been convicted of such crimes but are under indictment for them, people convicted of misdemeanor domestic violence, etc. Federal law already restrict commercial sale of handguns—by far the most common sort of gun used in violent crimes—to people at least 21 years of age. We even prohibit gun sales to many classes of people that don’t make Kristof’s list—people with a dishonorable discharge from the military, people under restraining orders, illegal aliens, unlawful marijuana users, etc.* This is pretty typical stuff in this debate: Proposing policies that we already have.

Kristof goes on to cite federal regulations of fully automatic weapons as a model.

To buy a machine gun made before 1986, you need a background check, a clean record and $200 for a transfer tax — a process that can take several months to complete. Then you must report to the authorities if it is stolen and get approval if you move it to another state. To buy a machine gun made after 1986 is more complicated.

None of this is terribly onerous, but these hoops — and stiff enforcement of existing laws — are enough to keep machine guns in responsible hands. In a typical year, these registered machine guns are responsible for approximately zero suicides and zero homicides.

The background check you go through to buy a pre-1986 machine gun looks for the same things as the regular background check; all it does is add fingerprints and a photograph to the process, along with several months of waiting for paperwork to be processed. And it is true that firearms sold through this process rarely turn up in crimes: As with the case of New York concealed-carry permits, the people who have used legally owned fully automatic weapons to commit crimes historically have been—almost exclusively—police officers. But, here’s the thing: Firearms sold through the regular retail process are rarely used in crimes, either. From my soon-to-be-updated Dispatch guide to gun questions and gun policy:

The problem with the regulatory approach is that it targets licensed dealers and the people who do business with them, who represent a vanishingly small share of criminals: Less than 2 percent of the criminals in custody today who had a gun with them when they committed their crimes acquired that gun from a licensed gun dealer, and only 0.8 percent from one of those gun shows we hear about all the time. Most violent crimes in the United States are committed by people with prior arrests and convictions for serious and often violent crimes, and the majority of murders in our country are committed by people who would not be legally eligible to purchase a firearm under existing law.

But it isn’t regulation that keeps people from using weapons from that stock of legally available machine guns to shoot up Chicago street corners. It’s money.

Do you know what a pre-1986 AR-15 costs? About thirty grand. Nobody robs a 7-Eleven with a $30,000 AR-15 for the same reason nobody robs a 7-Eleven with a $30,000 Holland & Holland safari magnum. For one thing, you can’t stick it in the waistband of your pants, but—the main thing!—it costs $30,000. Most criminals favor less expensive firearms, because, as one might expect a New York Times columnist to know, most criminals are poor.

In fact, as I am obliged to repeat about once a week, semiautomatic rifles account for a tiny share of homicides in the United States—all rifles together account for about 2.6 percent of homicides, meaning only 364 of the 14,000 or so homicides the United States saw in 2021. Kristof cites this fact, too, but doesn’t take the obvious lesson from it. 

From my gun-policy guide:

Are AR-type rifles especially dangerous? Look at the FBI data: In 2019, there were almost 14,000 homicide victims in the United States. All rifles combined—not just AR-type rifles, but all rifles—accounted for only 364 of those incidents. Even if we assume that all of those rifles were AR-type rifles (and they weren’t), we’d still have four times as many murders with knives, more murders with blunt objects, more murders with fists and feet, etc. For crimes short of murder, rifles are even more scarce: Only 13 show up in robberies in the 2019 FBI statistics, three in burglaries, eight in narcotics cases, and two in domestic-violence cases—in a country of 345 million souls, these are not big numbers.

For as long as we have been keeping records, the most common sort of firearm used in violent crimes in the United States is whatever handgun happens to be in widest use at the time: Back when people were robbing stagecoaches, it was the cap-and-ball revolver; back when people wore fedoras, it was the snub-nosed .38; today, it’s the 9mm automatic. All of the murders carried out with, say, an AK-47-pattern rifle in the United States in any given year will amount to a rounding error, but people enjoy pronouncing the syllables “AK-47” because they sound so interesting and authoritative. In reality, exotic weapons are a poor fit for the needs of a typical career criminal in the United States. And even when it comes to mass shootings, many more lives are lost to handguns than to scary-looking black rifles with military-sounding names.

At this point, Kristof really goes off the rails, implying that it is more difficult to adopt a dog in Mississippi, or to vote, than it is to buy a firearm. He uses the following two graphics:

He lists the ID-requirement to vote, but not for buying a gun—and, of course, it is required. Not only is a photo-ID required, but, to buy a gun—unlike the case of voting in most jurisdictions—all of the information on the license must be up-to-date. Have you moved since your driver’s license was issued? You can’t buy a gun with that license. (The questionnaire asks for a Social Security number as well, though this is not, strictly speaking, required.) You can “mail or hand deliver” a voting application, as Kristof notes, but you cannot buy a firearm by mailing in your application. You can buy a gun on the Internet, but a licensed dealer will ship it only to another licensed dealer, at which point you’ll have to go through the regular background-check process. You can try to mail somebody a pistol, but that is a federal offense. 

As for the dog-adoption stuff, Kristof is citing the practices of one pet-adoption agency as though they were law—which they are not. In fact, Mississippi pet-adoption law is an example of—angels and ministers of grace defend us!—self-regulation. If you adopt a pet and fib to the pet-adoption agency about how big your yard is, you don’t go to federal prison. Lie on the Bureau of Alcohol, Tobacco, Firearms and Explosives form and you could. (If the ATF could be bothered to, you know, do its job.) Pet-adoption agencies do not subject you to a federal background check—the FBI is not involved at all! This does not demonstrate the point Kristof thinks it does. 

Presenting these situations as though it were in any meaningful way easier to buy a gun from a licensed retailer than it is to adopt a greyhound or vote in Biloxi is deeply and fundamentally dishonest. Kristof should be ashamed of writing this, and the New York Times should be ashamed of publishing it—it is journalistic malpractice.

Making things worse, Kristof offers this flourish:

Why should it be easier to pick up military-style weapons than to adopt a Chihuahua? And why do states that make it difficult to vote, with waiting periods and identification requirements, let almost anyone walk out of a gun shop with a bundle of military-style rifles?

The answer is: It isn’t. Gun shops have identification requirements, and they most certainly do not “let almost anyone walk out of a gun shop with a bundle of military-style rifles,” unless by “almost anyone” you mean “anyone who is not a felon or under felony indictment, who has not been convicted of a high-level misdemeanor or is under indictment for one, who hasn’t been dishonorably discharged from the military, who hasn’t been ruled mentally unsound by a court of law, who hasn’t been convicted of domestic violence or currently facing such charges, who isn’t the subject of a restraining order, who isn’t an illegal drug user, who isn’t an illegal alien, who isn’t buying the gun for someone else,” etc. I’m open to putting other people on the no-no list: Do we want to mandate a 700 Experian credit score and an eye exam? Fine, let’s have that conversation—but it is, again, fundamentally and obviously dishonest to pretend that we “let almost anyone walk out of a gun shop with a bundle of military-style rifles.”

(Set aside the tomfoolery of “military-style rifles” as a construction for the moment, even though the rifles you can buy in most gun shops are not the kind of rifles actually issued by militaries, which generally choose “select-fire” rifles.)

Kristof notes that age is a factor here, and suggests that reforms take that into consideration:

This may be more politically feasible than some other gun safety measures. Wyoming is one of the most gun-friendly states in America, but it establishes a minimum age of 21 to buy a handgun.

This is kind of a funny one: Almost every state in the country has a law prohibiting a gun dealer from selling a handgun to anyone under 21, and all these laws together add up to nothing—because federal law already prohibits gun retailers, all of whom are federally licensed, from selling handguns to people under 21. What varies is possession laws: Can an 18-year-old legally possess a pistol given to him by someone else, one inherited from his granddad, one he is just using during handgun-hunting season, etc.? That gets a little weird at times: Texas had sought to defend its law prohibiting those under 21 from carrying a handgun and recently gave up the effort. (A Trump-appointed federal judge ruled that Texas’ “constitutional carry” rule applies an unconstitutional age limit.) It is not clear whether the state legislature will revisit the issue. In reality, even certain rifle parts can’t be sold to those under 21, because they could, in theory, be used to assemble a handgun. 

Kristof holds up Massachusetts as an example of good policy:

In effect, Massachusetts applies to firearms the sort of system that we routinely use in registering vehicles and licensing drivers to save lives from traffic deaths. Gun registration unfortunately evokes among some gun owners alarm about jackbooted thugs coming to confiscate firearms, which is another reason to work to lower the temperature of the gun policy debate.

Massachusetts is kind of an interesting case. As Kristof notes, it has relatively strict gun laws and relatively low gun crime. The neighboring states of New Hampshire and Vermont have very libertarian gun laws—and even lower rates of gun crime. Vermont has been a “constitutional carry” state for as long as there has been a constitution and enjoys quite low rates of violent crime. The people who imagine themselves to be social engineers who can guide the course of 345 million souls have trouble dealing with the fact that public policy, as such, often is a minor factor. There are lots of reasons you don’t worry too much about getting murdered in Vermont in spite of its liberal gun laws. Mobile, Ala., is one of the most dangerous cities in the country when it comes to violent crime; El Paso, Texas, is one of the safest—and the two states have similar, relatively loose gun laws.

Kristof plays the hits: “We should reassure gun owners that we’re not going to come after their deer rifles or bird guns,” he writes, ignoring the fact that AR-style rifles are almost certainly the single most common hunting rifle in the United States; he writes about hollowpoint bullets as being especially dangerous without noting that hollowpoints often are chosen precisely because they are in some ways less dangerous than non-expanding bullets, for example, being less likely to pass through the intended target and hit someone else; etc. He then descends into imbecilic fan service, suggesting that we apply shocking “graphic” warning labels to ammunition: “ One proposed ammunition label has a photo of a bloody face and states that a gun increases the risk of someone in a home being killed.” If there is one thing we know about guns and ammunition, it is that the scarier they seem, the better they will sell: The example of Death brand cigarettes, with its skull-and-crossbones packaging, leaps to mind. 

And, predictably, he focuses on gun stores and their customers (who commit very few crimes) while ignoring the actual correlations: The majority of U.S. murderers already have a prior felony conviction at the time of the killing for which they are charged. The vast majority of all violent crimes in the United States are committed by people with prior criminal records. The shooters are the problem and always have been. 

Violent crime in the United States is largely the work of habitual criminals. The people who can buy guns in gun shops have clean criminal records. The overlap between the people the gun-controllers seek to regulate and the people who commit the violent crimes is very, very small. Pretending that this is not so is foolish and dishonest. 

*Correction, January 31: This article stated incorrectly that people who received an “other than honorable” discharge from the military are prohibited from buying guns.