We should never presume that anyone can predict a specific atrocity. After a mass shooting, the red flags look so much more obvious than they possible could have before. Elliot Rodger’s shooting spree in Isla Vista, California probably could not have been predicted or prevented. It’s presumptuous to claim otherwise.
Rodger’s parents were sufficiently concerned about him to call the police. So were others. Four deputies, a UC Santa Barbara police officer and a dispatcher in training interviewed Rodger. After a short conversation, they determined Rodger was not an imminent threat to himself or others. And so they left. This welfare check didn't include a scan of California’s Dealer’s Record of Sale (DROS) database, which would have identified Rodger’s pattern of handgun and ammunition purchases. The officers never viewed the Youtube videos that had so alarmed Rodger’s parents.
I wonder whether the officers might also have been looking for the wrong warning signs, the wrong sort of person, at Rodger’s door. Some dangerous people satisfy our stereotypes regarding people who are obviously crazy. Rodger didn’t. Many dangerous people don’t. They become violent when they drink too much. Others have difficulty with explosive anger or raise domestic violence concerns. Or they display what my University of Chicago colleague Matt Epperson calls “veiled hostility” towards specific individuals or maybe towards groups of other people.
A lifetime of social difficulties and mental illness set the stage for Rodger’s crimes. Yet his immediate homicidal motives were rooted in envy and hatred for others. One of the more chilling insights of psychiatric research is how easily dangerous people can calmly deflect the usual questions, concealing their malice towards others without appearing dangerously ill.
For this reason and others, identifying perpetrators of violence will never be easy. I asked several leading mental health experts about best-practice when concerns are raised about a specific person. None minimized the challenge. Many suggested that police and mental health authorities could still do better through evidence-informed strategies.
Rutgers University mental health services researcher Beth Angell investigates clinical, public policy, and ethical challenges associated with involuntary treatment of mentally ill persons. As she describes things:
"These shootings teach us that those who screen psychological state need to go beyond assessment of psychosis, as it seems that in many of these cases the perpetrator was not out of touch with reality, but displayed features of psychopathy (callousness toward others, etc) and narcissism (grandiosity mixed with perceived outrage at others' indifference or rejection)."
Angell elaborates on the potential value of a model called Crisis Intervention Teams [CIT] and related systems that allow police to effectively collaborate with mental health professionals:
"In New Jersey… police could have called the designated emergency screening center (mandated by state law) that serves the local area, which would have sent out a trained clinician who might have been more effective in determining Rodger's psychological state, or perhaps could have convinced him to reveal more information about his ideation and plans to harm others in the near future."
Paul Appelbaum is Dollard Professor of Psychiatry, Medicine, and Law at Columbia University. He offered a similar take:
"Most police officers have frequent contact with people with mental illness, but have minimal training in recognizing the symptoms and assessing when they should be taken to a clinic or emergency room for further evaluation. Perhaps the most useful intervention … is to provide police officers with some good, basic mental health knowledge. They don't need to be trained to make diagnoses, much less to assess dangerousness (which is difficult even for psychiatrists)--merely to know when there are enough signs that something serious is wrong that further professional evaluation is warranted."
Police face a particular challenge given the lack of consensus regarding what to do when we have good reason to be concerned about someone, who does not meet the stringent standards of immediate dangerousness required to lock someone up. The proper burden of proof for involuntary inpatient commitment is—and should remain—quite high. We need less extreme options to address serious concerns without removing someone’s basic freedoms.
Involuntary outpatient commitment offers one potential path. Under this approach, a court can require an individual to participate in mental health services if there are specific risks that he will become dangerous to himself or to others if he does not remain engaged in care. For ethical and other reasons, there are no definitive randomized trials of involuntary outpatient commitment. Promising results from well-designed quasi-experimental studies do underscore the value of legal leverage to engage patients in treatment, particularly patients with histories of dangerous non-adherence to required psychiatric treatment or who display other risk-factors such as the combination of severe mental illness and substance abuse, or a history of violence.
Angell notes that it’s possible but hardly a sure thing that improved assessment might have helped at Isla Vista:
"It is difficult to assess whether [Rodger] could have been committed to outpatient treatment. (He clearly did not meet standards for inpatient treatment.)…. It certainly seemed that many who knew him were not surprised to learn of his actions… and this suggests that there needs to be effort to determine how to 'connect the dots' when an individual is showing signs of instability, and in this case, again - an aggravated blend of entitlement and aggrieved rejection in relation to others."
This composite picture might possibly have been used to support involuntary treatment in the absence of unambiguous evidence that Rodger posed an imminent danger to himself or others.
Whom do we trust with a gun?
We should also face discomfiting questions about whom we trust to purchase and possess lethal weapons. After all, the one unambiguous piece of evidence was that Rodger had legally purchased handguns and a lot of ammunition.
Jeffrey Swanson, professor of psychiatry and behavioral science at Duke University emailed:
"Police might have done more to find out about access to firearms, just given the family's concern about Rodger’s emotional state. There's no reason that police responding to people in crisis couldn't routinely address gun risk--talk about it, try to remove guns in various ways--instead of focusing on trying to predict when exactly somebody is going to be violent; that's very difficult even for experienced psychiatrists."
Swanson is now planning to study a training intervention for CIT police officers to routinely inquire about guns in mental health crisis calls. When guns are present, officers might use de-escalation skills to temporarily remove weapons from individuals at-risk of violence or suicide. If one happens to be in a state such as Indiana that has a preemptive "dangerous person" gun seizure law, police can remove firearms without a warrant, pending a judicial hearing, even if the person with mental illness is not imminently dangerous at the time and wouldn't meet criteria for involuntary commitment.
The Consortium for Risk-Based Firearms Policy has issued many recommendations in this area. One recommendation concerns the idea of a gun violence restraining order (GVRO) to restrict access to weapons among individuals who might pose a temporary danger to themselves or others. As Swanson and collaborators describe it, the main idea is to "create a new restraining order process to allow family members and intimate partners to petition the court to authorize removal of firearms, and to prohibit firearms purchase and possession temporarily based on a credible risk."
Writing in the Los Angeles Times, Renee Binder, president-elect of the American Psychiatric Association, wonders whether such a restraining order might potentially have been helpful, too. Given the concerns expressed by Rodger’s parents, a judge would have had the authority to temporarily seize his weapons, examine the relevant evidence, and perhaps connect some of dots that were missed in the original investigation. It’s possible—though no sure thing—that a judge might have temporarily prohibited Rodger from possessing firearms if it were determined that he posed a credible (though perhaps not immediate) risk.
I also wonder about something else. Like so many perpetrators of violence, Rodger was young: 22 years old. FBI data indicate that 45 percent of identified murderers are younger than age 25. We might examine the minimum age at which people can easily purchase guns. Many young adults are experiencing (often for the first time) serious mental health or social difficulties, and thus pose special risks to themselves and others. They also have shorter paper trails. Many young adult perpetrators have serious difficulties, but have yet to display legally actionable warning signs. Rental car companies apply extra scrutiny to drivers under the age of 25. Perhaps there’s a lesson there for gun policy.
Whom should we trust to buy or posses firearms? Right now, we trust almost everyone who hasn’t been convicted of a felony and hasn’t been involuntarily committed or otherwise deemed mentally incompetent. This binary approach works well for most gun owners. It doesn't work so well for a subset of troubled young men in the grey zones of gun policy. We need further options.