WASHINGTON — Roughly 22 million Americans — 8.9 percent of the adult population — have impulsive anger issues and easy access to guns. 3.7 million of these angry gun owners routinely carry their guns in public. And very few of them are subject to current mental health-based gun ownership restrictions.

Those are the key findings of a new study by researchers from Harvard, Columbia and Duke University. “Anger,” in this study, doesn’t simply mean garden-variety irritation. It means explosive, uncontrollable rage, as measured by responses to the National Comorbidity Survey Replication in the early 2000s. It is “impulsive, out of control, destructive, harmful,” said lead author Jeffrey Swanson of Duke University. “You and I might shout. These individuals break and smash things and get into physical fights, punch someone in the nose.”

Angry people with guns are typically young or middle-aged men, according to Swanson’s research. They’re likely to be married, and to live in suburban areas. In a recent op-ed, Swanson and a co-author point to Craig Stephen Hicks, a North Carolina man who “had frightened neighbors with his rages and had a cache of fourteen firearms” and who shot three Muslim students earlier this year, as a quintessential example of an enraged gun owner.

“To have gun violence you need two things: a gun and a dangerous person,” Swanson said. “We can’t broadly limit legal access to guns, so we have to focus on the dangerous people.” Taken at face value this isn’t a controversial claim. After all, guns don’t kill people, people kill people, as gun rights advocates are fond of saying.

But in practice we haven’t done a great job of identifying these dangerous people. After high-profile national tragedies like Sandy Hook or Aurora, the conversation quickly turns to limiting gun access for the seriously mentally ill — people with schizophrenia or severe bipolar disorder, for instance. This is palatable to gun rights advocates because it suggests that controlling the mentally ill, rather than controlling firearms, is the solution to gun violence.

But serious mental illness is only associated with a fraction of all violent crime in America. Swanson said the best available research shows that if you were to wave a magic wand and cure all serious mental illness in the United States, you’d only decrease violent crime by about 4 or 5 percent. Keeping guns out of the hands of these folks is a common-sense first step, but it won’t take a serious bite out of our gun crime problem.

To do that, we need to have a better sense of which kinds of people make for the most high-risk gun owners. And that’s where the new research by Swanson and his colleagues come in. In addition to the startling findings about the share of the overall population with both gun access and serious anger issues, the researchers also found that people with lots of guns — six or more — are more likely to carry their guns in public and to have a history of anger issues. And people with more than 11 were significantly more likely to say that they lose their temper and get into fights than members of any other gun ownership group.

It’s important to note here that gun owners as a whole aren’t any more likely to suffer anger issues than non-gun owners. And the vast majority of these people never have, and never will commit a gun crime. So you can’t simply ask someone if they have anger issues and take their guns away if they say “yes.” But there are other things policymakers can do to limit gun access among the most high-risk members of this population.

Federal law already limits gun access for individuals convicted of a felony, and for people with misdemeanor domestic violence convictions. Swanson and his colleagues suggest taking that one step further and placing additional misdemeanors on the restriction list: assault, brandishing a weapon or making open threats, and especially DUI, given the well-documented nexus between problematic alcohol use and gun violence. Other research has shown that people with a single prior misdemeanor conviction are “nearly five times as likely as those with no prior criminal history to be charged with new offenses involving firearms or violence.”

Again, the argument goes like this: if people, not guns, kill people, then it only makes sense to limit gun access among the most dangerous people. Swanson’s research suggests it’s time to move the conversation past the low-hanging fruit of serious mental illness, and start asking which other types of behavior might reasonably disqualify a person from owning a gun.

Ingraham writes about politics, drug policy and all things data. He previously worked at the Brookings Institution and the Pew Research Center.