A couple of hours before a young man shot an AR-15 rifle at former President Donald Trump, killing one bystander and wounding others, I had been finishing a column about gun violence as a public health threat. It was an eerie coincidence but not an unlikely one: More than 100 people die from gunshots on an average day in the United States.
In June, the surgeon general declared gun violence a public health crisis. Data show it’s now the leading cause of death for American kids 17 and under. The Centers for Disease Control and Prevention reported that in 2022, there were 48,000 deaths from firearms, about 40% of which are homicides. Many more people were disabled or maimed. And yet many Americans believe owning a gun makes them safer. In fact, self-defense is the number one reason people give for owning a gun.
Like other public health crises, gun violence has been studied, and scientists have data pointing to ways the carnage can be reduced. But Congress has been slow to pass any laws that would meaningfully restrict gun violence. Although there are more gun safety laws at the state level, the Supreme Court and lower courts have rolled some of them back, sometimes pointing to data allegedly showing guns make people safer. Those data are being grossly misinterpreted.
In the precedent-reversing 2022 New York State Rifle &Pistol Association v. Bruen decision, which struck down longstanding restriction on who could carry a handgun in New York, justices cited an unpublished survey that seemed to show guns are used well over a million times a year in self-defense. That survey, by Georgetown University researcher William English, was paid for by the gun lobby, according to reporting by the New York Times, who also picked apart his research methods. English responded in a WSJ op-ed, arguing that he’d not hidden his funding sources — they were declared in all his published work.
But research funding is only part of the story. English’s research raised a different question: Are his estimates credible? What do they tell us about the overall impact of guns on public health? How do they line up with what other researchers have found?
Stanford University law professor John Donohue said in the 35 years he’s been doing gun research, he’s never seen any work by English that met “what I consider to be the relatively low standards it takes to get something published.”
Part of the reason English found so many defensive uses of guns is that he allowed survey respondents to define “self-defense” for themselves. When I asked him about the 1.67 million number, he said in only 300,000 of those cases was a shot likely fired. He estimated that in 852,000 times the gun was only brandished, and in about 518,000 times neither happens — e.g. someone may have said they had a gun to intimidate the other party.
But David Hemenway, a professor of public health at Harvard, says it can be a problem to define self-defense so broadly. Hemenway has also done surveys asking people about their defensive use of guns, and he says most are not defending themselves against a mugger or a rapist. They are more like the subject who told Hemenway he went and got his gun after arguing with a neighbor who threw a beer. Or the guy who said that the alarm at his business went off, so he went down to the site, saw people standing outside on the sidewalk, and shot the ground. Or two groups of young men who exchanged gunfire at a gas station at 3 a.m. Should any of these cases really be considered self-defense?
To look more directly at how guns may be used by innocent people to defend themselves from criminals, he and other researchers have looked to a dataset from the National Crime Victimization Survey, put together each year by the Census Bureau and the Department of Justice. Recently that’s included more than 200,000 people. They’re asked whether they were victims of a crime or attempted crime and how they responded.
The results, said Hemenway, show that in cases where a person was present during a crime attempt, only about 1% responded by using or brandishing a gun. Extrapolating to the population at large, the data suggest fewer than 100,000 incidents each year in which guns are used to defend against actual criminals.
What’s often missed in surveys such as the one English conducted, said Stanford’s Donohue, are the cases where something went horribly wrong. He’s thinking of a man responding to a break-in and shooting his 16-year-old son by mistake, or a man who used a gun to pursue someone who robbed him at an ATM, in the process shooting a 9-year-old girl on her way to a Valentine’s Day party.
Other researchers have tried to answer a different question — does having a gun with you make you safer? In 2017, 75% of gun owners told Pew Research they believed that it did.
But Donohue said he was impressed by a 2022 study by colleague David Studdert, comparing gun-owning households and gun-free households in comparable neighborhoods and showing the gun-owning households were twice as likely to die by homicide. And in that 2017 Pew survey, gun owners were three times as likely to have ever been shot as non-gun owners.
Back in 2009, Charles Branas, now a professor of epidemiology at Columbia University, was at the University of Pennsylvania and gathered data on shootings in Philadelphia between 2003 and 2006. He compared the victims to a control group matched for age, sex and race who had not been shot. His results, published in the American Journal of Public Health, showed the shooting victims were four times as likely as non-victims to be carrying a gun at the time.
I wrote about that study for the Philadelphia Inquirer, and quoted critics who pointed out that cause and effect are hard to tease out. The correlation might be explained by the gun carriers living in more dangerous situations than the unarmed group. Or it could be that gun owners felt braver, and avoided de-escalation because they felt protected by their firearm.
Suicides by guns have spiked over the last decade, and make up the majority of gun-related deaths. Experts say many of these are impulsive acts by people who might have had a chance to recover from their suicidal thoughts if a gun wasn’t easily available.
The cost of gun proliferation is clear in terms of homicides, suicides and accidents, but it’s harder to compute the other side of the balance — the benefits people derive from the ability to protect themselves. Gun researchers can’t conduct a randomized controlled trial, handing out guns to a subset of volunteers the way they would with a new medicine.
Instead, the pro-gun side is assuming that in these allegedly defensive gun uses, someone would have been hurt or killed if not for their gun. It’s the kind of assumption that’s fooled patients and doctors into believing in unnecessary or unhelpful treatment or prevention strategies. In this case, the most careful, thoughtful studies show how those assumptions are wrongheaded.
Neither Hemenway or Donohue have come across any studies that show more firearms make for a safer or healthier society — and no evidence that the proliferation of semiautomatic rifles like AR-15s make people safer.
Yet gun enthusiasts keep buying them. Nancy Lanza believed she needed an assault rifle to protect herself in the upscale neighborhood of Sandy Hook, Connecticut. She was later gunned down by her own troubled son before he went to an elementary school and killed 26 people.
This week, investigators are desperately searching for missed signs that would have identified the shooter in Butler, Pennsylvania, before his assassination attempt, but so far they’ve found nothing in the way of motives. Except of course that he appeared to be a loner, possibly troubled, and thanks to his parents’ gun collection, with easy access to a very dangerous weapon.