Yesterday’s Washington Post has a nice story under the self-explanatory headline: Most mass shooters aren’t mentally ill. So why push better treatment as the answer? Drawing on work by Columbia University forensic psychiatrist Michael Stone, Post reporter Michael Rosenwald notes that a surprisingly small fraction of mass shooters suffer from severe mental illnesses such as schizophrenia or psychosis. Yet politicians often call for “better mental health care” in response to mass shootings or other atrocities.

This focus on mental illness does bring some benefits. It might bring desperately-needed resources to our beleaguered mental health system and to valuable bench science and clinical research. Public safety concerns have occasioned valuable discussion of ways we might improve inpatient psychiatric care, for example by altering Medicaid policies that block reimbursement for much residential care. Congress is also focused on some valuable measures to make it harder for dangerous or suicidal people to buy guns or to maintain access to dangerous weapons. We could certainly use additional resources for prohibited-possessor databases that may have produced missed opportunities to prevent shootings at Virginia Tech and Charleston, SC.

Although civil libertarians rightly worry, it is always important to consider whether we have struck the right balance in matters such as outpatient commitment that are pertinent to a subgroup of people living with mental illness. Interventions such as forensic assertive community treatment can be helpful in this area.

Acknowledging all that, conflating dangerousness with mental illness can bring profound harms and provide a misleading guide to public policy. Discussing mental illness in the context of mass shootings only intensifies the accompanying stigma. This ironically causes more people to hesitate before seeking help from the mental health system. It is also a fairly transparent effort to divert public attention from policies such as universal background checks that might be helpful in reducing gun violence.

Framing the violence conversation in terms of severe mental illness also feeds into false hopes that we can specifically identify the most dangerous people. Unfortunately, this is rarely possible until after an atrocity has occurred. In so many cases, a mass shooter has left various red flags that seem painfully obvious in hindsight, but where no truly actionable information was previously available–particularly in a society that trusts virtually everyone to own a gun who is not a convicted felon or is one of the very tiny minority who have become legally enmeshed in the institutional system of psychiatric care. The Planned Parenthood shooter, for example, was an eccentric loner who scared his neighbors. I know of nothing in his prior record that would have barred him from purchasing lethal weapons.

Ironically, stereotypes about the severely mentally ill also lead us astray when we do have specific warning signs that someone is dangerous. Elliot Rodger went on a shooting spree in Isla Vista, California. I don’t believe police should have predicted this, but they did have specific warnings. As I noted in the Washington Post, Rodger’s parents and other people were sufficiently frightened that they contacted police. Several officers went to Rodger’s house and conducted a short interview with him. And they left–determining that Rodger was not an imminent threat to himself or others.

It’s easy to second-guess. The officers didn’t scan California’s Dealer’s Record of Sale (DROS) database, which would have identified Rodger’s scary handgun and ammunition purchases. They never viewed various Youtube videos Rodger had posted, which had rightly alarmed his parents. The real problem was that the officers were probably looking for the wrong things. They were looking for someone out of touch with reality, and that wasn’t him. Rodger was perfectly capable of holding a lucid conversation at his front door. He wasn’t schizophrenic or psychotic. He was just incredibly angry with other people.

A more sensible conversation at the nexus of violence and mental illness must start in a different place.

First, we need better tools to keep potentially dangerous people away from lethal weapons. Most of these men and women cannot and should not be involuntarily committed. Yet they show other signs of dangerousness. A better background check system might help here. So would innovations such as gun violence restraining orders. We might also look more carefully at gun purchases made by young adults who may have life difficulties, but who have had less time to accumulate a legal and psychiatric track record. Some groups of people on disability who require the help of a representative payee might also merit heightened scrutiny for firearm purchases. We might also restrict certain weapons such as AR-15s that appear especially connected with mass homicides. This wouldn’t appreciably reduce homicides, but it would help.

More important, we should bolster our mental health safety-net, particularly components at the nexus of criminal justice and substance use. Dangerous people with mental health issues often experience co-occurring substance use disorders. Efforts such as Hawaii HOPE and Dakota 24/7 can be very helpful with this population.

Medicaid expansion is another essential tool. People at-risk of violence frequently require health coverage. Places such as Cook County do an admirable job signing people up, including most people passing through the Cook County Jail. That allows individuals access to key services, and it supports the entire ecosystem of mental health care. We are just learning how to use this new tool to its best advantage.

We have a long way to go. I recently spent an evening in the basement of Cook County Jail. A woman was discharged who was experienced sad mental health and substance use issues, along with various other life challenges. She was a nonviolent offender, one of the thousands of urban rabble who pass through our jails every day. A local social service agency arranged for her to have an appointment at a local hospital early the next day. She would need to take a couple of busses to get there. I’m not sure anyone can or will follow up if she didn’t show up. I fear that her next mental health system encounter will happen right there in the jail, next time she’s arrested for petty thievery, drug possession, public intoxication, or some similar crime.

The boring blocking-and-tackling of our mental health system needs to work better for this woman, and for so many others. It’s the hard work that must be done. Of course, tt’s tempting to rally support for such efforts with claims that these might prevent mass shootings or other scary violence. Most dangerous people don’t experience severe mental illness, and most people living with severe mental illness aren’t dangerous. So most of this political rhetoric is empty bluster. Whatever its short-term utility, the right answer to the wrong question often causes unexpected harm.

Our ideas can save democracy... But we need your help! Donate Now!

Harold Pollack is the Helen Ross Professor at the School of Social Service Administration at the University of Chicago.