The Cause of America’s Gun-Death Epidemic? It’s Guns

The prayers for Americans murdered by guns reads like a litany: Robbs Elementary, Tops Friendly Markets, Marjory Stoneman Douglas High, Orlando Nightclub, Sandy Hook Elementary, El Paso Walmart, Columbine High, and on and on. According to the nonprofit Gun Violence Archive, through June 12th, 2022, there’ve been 260 mass public shootings (four or more shot in a single incident) in the United States since January 1st—on pace for a record year. As of June 14th, a total of 19,723 Americans died by guns from all causes: homicide, suicide, and accident. That’s about 119 per day, or five per hour, shot dead by a firearm. By the time you finish reading this article, someone will likely have died by gunshot. Why? What is the cause of gun violence?

My hypothesis should be as uncontroversial as it is obvious: guns. But for the gun, none of these mass public murders would have happened, and only a handful of those who died by gun suicide would be dead.

To be sure, every act of gun violence has a unique deeper cause. The Buffalo shooter was a white supremacist. The El Paso Walmart murderer was a racist. The Sandy Hook gunman was mentally ill. The Tulsa, Oklahoma hospital slayer was a disgruntled patient seeking revenge on his surgeon. The professional gambler who massacred 60 concert goers from a Las Vegas casino resort hotel room was, according to the Las Vegas FBI office, seeking to “obtain some form of infamy.”

In the study of causation, this is known as the overdetermination problem: Just how many causes can an effect have? For gun violence, all of the above social factors (and more) have been proposed. How should we sort them out?

No human action has a single cause, yet a focus on overdetermination can obscure our focus on the most important causes, which we target through legislation, public policy, and changing norms. There is no shortage of moral outrage over guns on both sides of the political aisle, but determining the cause of gun violence must be pursued objectively. Let’s see if we can understand the phenomenon from a scientific perspective.

One theory that’s popular among politicians and the general public is that gun violence is linked to violent video games. But the vast majority of studies find no such causal connection. For example, a 2019 study of over 1,000 teens that measured youth aggressiveness against the video games they played, published in Royal Society Open Science, found no support for the hypothesis “that recent violent game play is linearly and positively related to [caregiver] assessments of aggressive behaviour.” Other researchers have pointed out that countries such as Japan have equally high rates of consumption of violent video games but almost no gun violence. And it should go without saying that, of the tens of millions of Americans who play violent video games, the vast majority never shoot anyone or commit any acts of violence.

The reason so many people are confused by the issue of causal inference in social science—including the issue of gun violence—is that we mistake proximate explanations with ultimate explanations. JFK was killed by two bullets from an Italian-made Mannlicher-Carcano bolt-action rifle fired from the sixth floor of the Texas School Book Depository building by Lee Harvey Oswald. That Oswald was a psychopathic hate-filled anti-Kennedy pro-Castro Communist in search of notoriety (i.e., the ultimate explanation) is secondary to what is important in determining the proximate cause of Kennedy’s death—Oswald’s gun.

Or consider another example, this one drawn from Judea Pearl’s 2020 book, The Book of Why: The New Science of Cause and Effect. When a death-row prisoner is shot dead by a firing squad, what is the cause of his death? The bullet from the gun that hit him, obviously. But no one holds the members of the firing squad morally culpable, especially since there are usually multiple shooters (so no one rifleman knows who actually killed the prisoner); and in any case, they were ordered to do so. So the causal chain moves up to the officer who gave the command to shoot, and further up the chain to the prison warden who ordered the procedure that day, and further still to the court case that determined the death penalty in the case, and finally to the legal regime that legalized capital punishment in the first place.

The reason why the causal analysis is different in these two cases—Oswald’s assassination of JFK versus a state-ordered firing squad—is that the former was a crime perpetrated voluntarily by one person, while the latter is an officially state-sanctioned killing.

Most perpetrators of gun violence fall into the former category, since no one has ordered them to kill anyone (except in the case of, say, drug-cartel foot soldiers, or a Mafioso execution). Unlike with officially sanctioned killings, the proximate cause of illegal gun violence is guns, full stop. All other explanations on offer—mental illness, racism, white supremacy, a culture of violence, raging hormones, maleness, and the like—are really attempts at finding an ultimate explanation. These are important, but they don’t inform policy decisions as readily as most of us would like when we demand that the government “do something.”

One method of controlling for such confounding variables is what is called the Randomized Controlled Trial (RCT), in which individuals are randomly assigned either to the experimental treatment group that receives X, the placebo group that receives a fake X, and the control group that receives no X. The effects on some dependent variable, Y, are then compared. The problem with RCTs is that while they’re great if you’re trying to study if, say, Tylenol helps alleviate the effects of arthritis, they can’t be conducted (for ethical and/or practical reasons) in regard to a lot of problems we care about, such as the health effects of smoking among teenagers. (Imagine the public reaction if researchers recruited high-school students, and instructed some of them to smoke cigarettes for a year, while handing out fake cigarettes to others, and paying a third group not to smoke anything.)

In those cases, what we do instead is study so-called natural experiments—in which (to take the smoking example) we compare the health outcomes of people who smoke with the health outcomes of those who don’t, while doing our best to control for other factors (such as diet and exercise).

When it comes to natural experiments involving guns, the most cited study in support of gun control came out of Australia following the 1996 Port Arthur firearm massacre, in which a 28-year-old man killed 35 people and wounded another 23 with an AR-15 assault rifle—the civilian variant of a weapon designed by the US military to rapidly kill as many people as possible. Following this crime, Australian state governments agreed to ban semi-automatic and pump-action shotguns and rifles through a set of gun-law reforms.

A 2006 follow-up study measured “changes in trends of total firearm death rates, mass fatal shooting incidents, rates of firearm homicide, suicide and unintentional firearm deaths, and of total homicides and suicides per 100,000 population.” It showed that in the 18 years before the gun-law reforms were implemented, there were 13 mass shootings in Australia; while in the decade following those reforms, there were none. The authors concluded:

Australia’s 1996 gun-law reforms were followed by more than a decade free of fatal mass shootings, and accelerated declines in firearm deaths, particularly suicides. Total homicide rates followed the same pattern. Removing large numbers of rapid-firing firearms from civilians may be an effective way of reducing mass shootings, firearm homicides, and firearm suicides.

The researchers also found “no evidence of substitution effect for suicides or homicides,” and “the rates per 100,000 of total firearm deaths, firearm homicides, and firearm suicides all at least doubled their existing rates of decline after the revised gun laws.”

Another case study is Austria, which in 1997 passed gun-control legislation in line with a European Council directive on controlling the acquisition and possession of weapons. That country restricted firearm purchases, upped the minimum age to purchase a firearm to 21, included an extensive background check and three-day cooling-off period, required buyers to indicate a reason for the gun purchase, added psychological testing for obtaining semi-automatic firearms and repeating guns, and required that buyers have safe firearm storage. A decade later, the British Journal of Psychiatry published an article entitled “Firearm Legislation Reform in the European Union: Impact on Firearm Availability,” in which University of Vienna Professor Nestor D. Kapusta and fellow researchers concluded that—as the figures reproduced below indicate—“the introduction of restrictive firearm legislation effectively reduced the rates of firearm suicide and homicide.”

Austrian firearm data reported by Kapusta et al. in 2007. Left: Firearm license rate vs. time, before and after passage of 1997 firearm legislation. Center: Firearm-related suicide rate vs. time. Right: Firearm-related homicide rate vs. time.

Let’s look at some additional examples of natural experiments. The figures below show (a) the relationship between gun ownership in different US states and gun deaths per capita, with a best-fit regression line showing that more guns are associated with more deaths; and (b) a cross-nation comparison showing the relationship between per capita guns and per capita gun deaths, with a best-fit regression line revealing, once again, that more guns correlates with more deaths (with the United States being an extreme outlier in regard to both variables).

Top: Gun deaths per capita vs. gun ownership per capita for different US states. Bottom: Gun deaths per capita vs. guns per capita for various nations.

In a comprehensive 2011 study reported in the Journal of Trauma and Acute Care Surgery, researchers compared the United States to 25 other high-income countries in regard to homicide, suicide, and unintentional gun deaths among five- to 14-year-old children. The results were reported as a series of ratios, expressing the US rate as a multiple of the average for 25 other high-income countries.

For non-gun-related homicides, the ratio was 1.7 to one, indicating that per-capita homicide deaths among US five- to 14-year-old children were found to run about 70 percent higher than the average for other wealthy nations. But for gun-related homicides, the ratio was an astonishing 13.2 to one.

The pattern was similar for suicide. In the case of non-gun-related suicides, the ratio was found to be 1.3 to one, indicating that per-capita suicide deaths among US five- to 14-year-old children were found to run about 30 percent higher than the average for other comparable nations. Where the suicides are gun-related, on the other hand, the ratio is 7.8 to one. And in the category of unintentional gun deaths, the ratio was found to be 10.3 to one.

In other words, for rates of non-gun-related homicides and suicides, there’s a significant but not startling difference between the United States and other Western nations. But the ratio shifts dramatically, almost by a full order of magnitude, once someone picks up a gun. And the reason isn’t complicated: Guns are far less forgiving than other methods of attempted homicide and suicide. When a couple of drunken guys get into fisticuffs at a bar, it mostly results in bruised bodies and egos, but add a gun into the mix and someone is likely going to the morgue while the other is headed to prison.

When people attempt suicide, they don’t always want to kill themselves. According to my friend and colleague Dr. Ralph Lewis, a psychiatrist who treats people in crisis, many say, “I don’t know what came over me. I don’t know what I was thinking.” An overdose of medications or a botched attempt at slit wrists may grant someone a second chance at life. With guns, that is much less likely.

As well, consider the fact that, homicide excepted, crime rates in the United States are comparable to those in other Western countries that have few guns‚ including rates for car theft, burglary, robbery, sexual assault, aggravated assault, and adolescent fighting. It’s US homicide rates that are a category of their own—because of guns.

If it isn’t clear by now that the primary cause of gun violence is guns, and that curbing their availability and capacity can attenuate the resulting carnage (though such measures cannot eliminate it entirely), let me offer a few additional observations, starting with the popular meme among gun-rights activists that if guns are outlawed, only outlaws will have guns.

This slogan simply isn’t true. Gun-control legislation does not mean outlawing guns any more than the licensing and regulation of automobiles means that only outlaws will have cars. The 1934 National Firearms Act, regulating the manufacture and sale of machine guns (and backed by the National Rifle Association at time), did not result in only outlaws having machine guns. Where is today’s George “Machine Gun” Kelly? He’s not around, because machine guns are regulated and restricted. This is the kind of rational gun control that even today’s more militant NRA can (or at least should) get behind.

Then there’s NRA Executive Vice President Wayne LaPierre’s oft-quoted proclamation, “The only thing that stops a bad guy with a gun is a good guy with a gun.” The May 24th Uvalde, Texas massacre of 19 children and two adults put the lie to this trope, as well-armed good guys spent over an hour restraining parents desperate to rescue their children while the mass murderer continued his slaughter inside the elementary school. More generally, the NRA’s solution to crime and violence—of arming everyone and hoping the good guys out-gun the bad guys—contravenes our understanding that, except in rare and exceptional circumstances, designated law-enforcement officials have a monopoly on the use of force. In essence, the NRA is presuming America to be an effectively lawless society in which might makes right, so let’s all arm ourselves to the teeth.

If the “good guy with a gun” is an appropriately armed and professionally trained member of the police or military who routinely practices his craft at shooting ranges and in simulation drills, then yes, the deployment of such officers is one among many factors that has contributed to the decline of peacetime violence over the centuries. But if “good guys” is taken to mean armed private citizens with little to no gun training, this could not be more mistaken. A brief scan of YouTube videos under the search string “gun mishaps and negligent discharges” (or any similar search) will provide hours of darkly entertaining—and sometimes unspeakably tragic—gun accidents due almost entirely to human error.

This week, there is cautious optimism in Washington as 10 Republican senators broke ranks with their party by working with Democrats toward modest legislative steps to prevent school shootings—including mandating additional scrutiny of young gun buyers. Even if the legislation passes, no one is pretending that it comes close to comprising the sort of comprehensive gun-control reform that America requires. But at the very least, it shows that even some conservative politicians are beginning to wake up to the fact that America’s status-quo policy is tragically misguided.

This is a companion discussion topic for the original entry at

Rape is an epidemic in the US, too. I think we ought to figure this one out. Here’s a few ideas:

  1. Register every penis in the US.
  2. Embed in every penis a microchip that allows the government to track it.
  3. Use microchip to prevent use of penis for sexual activity unless owner has been through 250 hours of training.
  4. Allow people to “red flag” any penis that they think is a danger to the community. Said microchip will then be activated. Owner will regain control of penis in about 25 years after bureaucrats decide it is no longer a threat.

better yet - just cut them all off at birth…


A key graphic in this paper, which shows a strong association at the state level between gun ownership and gun deaths, combines suicides and homicides. Rajiv Sethi’s analysis (Gun Violence - by Rajiv Sethi - Imperfect Information) assesses these associations separately and shows that the correlation between gun ownership and gun deaths is carried almost entirely by suicides. If so, reducing gun ownership may indeed reduce gun deaths by reducing suicides (certainly a desirable outcome) but have less effect on homicides, which are more frequently carried out with lost or stolen guns and guns brought in from other states.


I wish American gun nuts would just be honest about it: they value their individual ‘right’ to own guns above their ‘right’ to live in a more peaceful society. In particular they value their assault rifles more than they value their children. All supposed arguments advanced by the gun types are clearly motivated and are comparable to the way the tobacco industry defended their products. They were lying and they knew it.


I assume your Canadian or Australian.


Let’s outlaw guns. After all, we outlawed guns and that is how we won the war on drugs. Nobody uses drugs anymore and nobody’s life is ruined by drugs.

Here’s another logical idea: let’s do away with cars. By doing so we will have saved the lives of tens of thousands of people who are killed by cars.

Just a thought.


Yes, Canadian. Tho we are very much like Americans in most things, the gun insanity is not so bad here. One can quite easily obtain a hunting rifle or even a handgun but it’s a bit more difficult to buy an assault rifle. Main thing is that you need a FAC first, which is easy enough to get – mine is expired but I had no trouble getting it. A very few nods in the direction of common sense would make a huge difference, but again, when it comes to guns, Americans have no common sense.

But not a good one. Almost nobody suggests outlawing guns, however many agree that being able to buy an AR-15 when you aren’t old enough to buy a beer is very strange.


If you believe that legally purchasing a handgun in Canada is “quite easy”, then you are dramatically misinformed.


All the talk about causality, makes my head spin. Variously:

  • It was two bullets, that caused JFK’s death, but also,
  • it was crazed Oswald, but
  • when a firing squad executes a sentence, it isn’t those firing who are the cause, because
  • in that case there was a judicial system involved, but
  • the title of the article says the cause of U.S. gun deaths is neither bullets nor individuals, but rather, guns.

jeff.brookings says that a key graph in the article, conflates suicide death with other homicide death. Could be, sounds reasonable. I don’t have time to research that.

Defensive Gun Usage is another one that’s hard to get a good handle on. Some sources that seem pretty believable, seem to indicate that it’s a lot more prevalent that one would gather from reading media reports. Apparently, fairly often, just mentioning or brandishing the gun brings a bad or threatening incident, to an immediate stop. This is obviously a highly contentious area and many sources I found seem to be indulging in “motivated research”. I found one interesting reference to CDC surveys carried out in the 1990s, results of which were never published.

If I had to summarize the original article I’d say “Scattering a few hundred million guns around a country increases the frequency of gun violence”. When put that way it seems kind of obvious.

More important to me is, just what are we supposed to do now?

There are a lot of guns physically existing in the U.S.A. - getting rid of all semiautomatics, I don’t know if we can even do it. There might have been a time, but these days, there is a “wrong kind” of gun fetishist. And there are a lot of them.


Lots of guns in Israel, very little criminal-related gun violence (at least among the Jews, the police have effectively stopped policing in Arab towns and hence therer there’s a problem with violence, including guns, even though a much smaller percentage of Arabs serve in the army and thus there’s less access to guns). Gun violence in the USA is a cultural thing.

I think the situation is best summed up with Homer Simpson’s line, when told there’s a five-day waiting period to purchase a gun: “Five days? But I’m mad NOW!” Homer Simpson in Gun Shop - YouTube


Not sure why you insist on insulting us and expect us to listen. True we all have our flaws. Canadas first female prime minister Justina just came out and stated without blinking that Canadians do not have a right to self-defense. We are nothing like you. We have right to free speech, you don’t. Your speech is compelled by the state. This is not to say that there aren’t rationale adjustments we should adapt relative to guns. Raising the age to 21 is one that Florida placed years ago and has not received a court challenge. Waiting periods, mandatory training and yearly recertification on rifles all reasonable. Of course, true to Michael Shermas progressive Bonafide’s barely a mention of the daily carnage in inner cities where black men blow each other’s brains out and absolutely nothing proposed will make a difference for them. How about securing schools. We just sent 50 billion to Europe to protect their kids, why not ours. You see it’s easier for you and Michael to refer to millions of law-abiding Americans as nuts. Michael can’t help himself as he is safe and secure along with other CA costal elites like Paul and Nancy Pelosi with lots of time on his hands to extrapolate his world view into perato charts. There certainly is a need to address this issue but referring to millions of Americans as nuts is a non-starter,


True…additionally what the Israelis did in the 70 after a massive school shooting that killed 35 was secure their schools. Almost no discussion around that. Try getting into a government building without a process. We can’t do that for schools?


It is admittedly getting much harder under Justin.

Sure. Same in Switzerland. But the US is not Israel or Switzerland, as you say, it is a cultural thing thus America needs rules that will work there. Anyway it seems 10 GOP Senators are disobeying their NRA masters are are ready to cut a deal with the Rats.

Because in this case the insults are justified and given the carnage down there it would be in your own interest to get sane. We have our insanities up here too.

We now have the opposite problem to you, we’re going woke and I don’t support that either.

That too.

Exactly. And background checks and a few other no-brainer initiatives. So much could be done with so little impact on your ‘rights’.

Probably true. If both sides would start talking sense then facts like that one should be admitted.

Possibly entirely sane except when it comes to this foundational issue. Even then, they say most gun owners support reasonable legislation, it’s just the gun lobby that resists. What? 80% of NRA members support background checks IIRC. Even the ‘gun nuts’ aren’t really the nuts, it’s the lobby that’s nutty.

I’d say it’s a starter even if not a finisher. You guys do have this cultural fixation, it keeps you from thinking clearly.

You could indeed do it. Make every school a fortress. But you shouldn’t have to.


The school in Texas was secured. But a door was left open. What to do about that?

1 Like

At the end of the day it will be hard to turn schools into fortresses won’t it? Nice idea tho, shootout training for teachers, lockdown drills, the school armory, teaching the kids that at any moment of any day they should be prepared for war. Only in America.

1 Like

Do whatever you want, some person will prop a door to “help a friend”.

I liked this article. Sometimes, it is just that simple (guns cause gun deaths), and you have to call a spade a spade. I’m saying that as someone who respects people’s preference to own guns, though I personally don’t want to have much to do with them. I mean, it’s a democracy. I wish people had different ideas, but they don’t.

Really like (for the most part) what RayAndrews said above: “I wish American gun nuts would just be honest about it: they value their individual ‘right’ to own guns above their ‘right’ to live in a more peaceful society.”

Change gun nuts to something less pejorative (gun enthusiasts, or something), and throw in the right to self defense, and I think this is it in a nutshell. Yes, we have a lot of guns and accompanying violence and deaths in the US. But not enough to make it a daily-life kind of concern. People make a cost-benefit analysis, and for a lot of them the result is, let me have my gun. My own answer is different, and I really don’t see that people’s lives would be worse without AR-15s, but, like I said above, it’s a democracy. Bet I’d feel different if someone I knew was killed by gunfire though.


Guns kill people
Cars kill people
Drugs kill people
Radiation kills people
Knives kill people (look at the knife violence in Britain)
Clubs kill people (Cain and Able?)
Ropes kill people (lynching anyone?)

No shit. Of course guns kill people.

So when you have Countries with a lot of guns like Israel, Switzerland, etc. why don’t those guns kill people?

Why is there, typically, more gun violence in states with the strictest gun laws and least number of guns per capita? (including handguns here (also not talking suicides (but guns are a rounding error compared to opioid overdose right now)))

Why didn’t ubiquitous guns throughout US history predating the ‘scary’ AR but just as capable with Semi autos in the early 1900s (Or hell, the fully auto tommy guns that people could legally buy) kill people? (at least to the same degree?) What makes guns, right now, kill more people than guns did back then? The availability of the AR only occurred in the early 2000s…but other weapons were just as capable. Why didn’t they kill people? (to be fair, ARs are actually a tiny fraction of all gun crime but they ARE scary!)

Bigger question in a causation vs. correlation argument is what happened in the mid to late 90s, that in conjunction with the lifting of the AR ban amongst other things, suddenly started causing guns to kill people while gun laws have gotten substantially more strict over the same time frame? (and ARs make up a tiny volume of gun violence)

But guns also prevent tyranny. Stalin killed 40 million unarmed civilians. Mao killed 60 million unarmed civilians.


This is laughable coming from California where your Soros DAs won’t prosecute crimes committed by felons in possession of a gun.