Guns, Mental Illness & Public Policy

Share This:

“It is not a gun problem. it is a Muslim problem, a mental health problem, another kind of problem.” Wrong! It is a gun problem. People are being increasingly massacred by people using guns especially rapid fire guns.

Identifying specific individuals who are a  gun violence risk is a fools task. Three items below address why.

1. SF Chronicle Nov. 6, 2017 -Trump wrong to blame mass killings on mental illness rather than guns,

Exerts follow:

President Trump on Monday attributed the slaughter of 26 people in a Texas church — the nation’s third mass killing in five weeks — to “a mental health problem,” saying it wasn’t a “guns situation.”

“He’s wrong on two counts,” said Michael Stone, professor of clinical psychiatry at Columbia University and author of “The Anatomy of Evil,” who has studied 360 of the most notorious mass murders of the past century. “It is a gun issue. And there are very few mass murderers who are certifiably crazy.”

…Trump “is not the only person who jumps to that conclusion. It is a popular misconception that people who commit mass shootings must be crazy,” said Liza Gold, a forensic psychiatrist who teaches at Georgetown University and edited “Gun Violence and Mental Illness.”

“Most gun violence — 98 percent — is not attributable to people with mental illness,” Gold said Monday.

2. SF Chronicle Nov 6, 2017 – As the killing continues, time to say ‘Enough!

This editorial helps to put the problem & policy issue in perspective. Exerts follow:

Shootings, both the headline-grabbing and the sadly routine kind, happen for infinite reasons — and no reason. In a small town outside San Antonio, the mass murderer appears to have been motivated by a grudge. In Orlando, Fla., last year, another claimed allegiance to the Islamic State. In Blacksburg, Va., a decade ago, still another showed symptoms of severe mental illness. And in Las Vegas a month ago, the shooter left such a paucity of clues as to his motive that authorities are puzzling over it to this day.

The multitude of potential reasons speaks to the futility of the question from a policy standpoint. Targeting the mentally ill at best affects a tiny fraction of crimes and an incalculable number of people who have never contemplated such a thing.

Gun deaths can be made far rarer by rationally restricting access to weapons. According to U.N. data, an American is about six times more likely than a Canadian to die in a gun homicide, 18 times more likely than an Australian, and 35 times more likely than an Englishman.

{After the Sutherland Springs Shooting] Trump weakly offered that the killing “would have been much worse” if an armed bystander hadn’t intervened, as if we should take comfort that this shooting didn’t break the top four. He asserted that it was “too soon to get into” questions of gun policy.

We’ve heard that refrain after Las Vegas, Orlando, Sandy Hook, Virginia Tech, Aurora and too many other massacres. We’ve heard enough thoughts and prayers. The 1994 assault weapon ban expired after a decade, with Congress too timid to act, because it’s never “the right time.”

If not now, when? We’ve seen enough horror inflicted on everyone from schoolchildren to worshipers in a small-town church.

We know what these weapons can do, how ubiquitous they have become, and how easily they can end up in the wrong hands.

3. The Likelihood of Identifying High Risk Individuals — Consider The Following

Trump and the Congress keep offering non-action and non-solutions such as “Extreme Vetting” and of all things Prayers. The absurdity of a “prayers policy” is obvious. Screening groups to identify high risk individuals is less obvious but also absurd. To see this it takes a bit of math — I know Trump and Republicans are not into into science stuff but consider the following:

Suppose you have some method that allows you to identify individuals as potential gun violence threats and the method is 99% accurate.

Also suppose that for a target population (e.g., criminals, people with mental health issues, Muslims, the general population) your pre-screening assessment is that about 10 in 1,000,000 individuals are a threat for gun violence and should be dealt with in some fashion – a policy.

After applying “the assessment method” the questions are: (1) what is the chance that serious threats are identified and (2) what is the chance that non-treats are mistaken to be threats. Bayes’ Theorem provides a rigorous method for computing these likelihoods.

Specifically given my hypothetical assumption that 10 in 1,000,000 are a threat. In a group of one million there is actually only 10 persons who are are a serious threat. Of these 10 people, assuming that the “assessment method” is 99% accurate, 9.9 will on average be identified as threats. For the remaining population (non-threats) 9,999.9 will also be identified as threats. Thus the method will identify 10,010 individuals as threats but the chance of actually being a treat given that the method identified you as a threat is 10/10,010 or 0.0001 or .01 percent.

This example demonstrates that for a group of people where individuals with a proclivity to commit gun violence are rare, a policy to identify the high risk individuals even if the vetting is 99% accurate, is problematical — in my example only 1 in 1,000 of the individuals  identified as high risk are actually high risk.

Maybe gun control makes sense. Identifying the specific individuals who are threats is a fools task. Limiting access to certain kinds of weapons is not.