The Facebook killer and why we can't always predict gun violence
© Getty

Shortly after the cold blooded slaying by Steve Stephens, residents across at least five states — Ohio, Pennsylvania, New York, Indiana and Michigan — were advised to be on high alert for a man armed and dangerous. The gunman killed a 74-year-old man at random on Easter Sunday in Cleveland and posted it on Facebook.

When sightings were called in to the police, some places went on lockdown. Today Pennsylvania authorities confirmed that the killer had taken his own life. This has left many people with questions.

Some may say Stephens is a bad man with a gun inside our borders. The real culprit of gun violence is all of us who refuse to acknowledge that we have an urgent, complex public health problem in our country. Gun violence requires further systematic scientific study and interventions nationally across legal, community, and health systems in our country.


When random shootings or mass killings like this happen, the public understandably looks for comfort and control. Part of that is thinking if we could predict these murderous rampages beforehand, the shooters could be stopped and violence prevented. This desire makes sense but the answer is less comforting.

On his now deactivated Facebook page, Stephens blamed the shootings on his mother and his former girlfriend. He also noted he lost all his money gambling at casinos during the past year. But this doesn’t hold up.

Plenty of people have difficult relationships with their mothers and don’t go out and intentionally harm people because of it. Many couples break up and the jilted don’t commit heinous violence. Many people have significant financial hardships including those brought on by risk-taking behaviors and we don’t go out on random shooting sprees.

This then begs the question, “Why would someone react like this and what are some of the signs that a person displays that would lead them to going on a killing spree?”

As a psychologist who studies trauma and treats its survivors, I’d love nothing more than to help make the world a safer, saner place. But the prediction of violent behavior is an incredibly difficult task for mental health professionals and one that should not be done based primarily on clinical experience and expertise. There are comprehensive statistical models that include person and environmental predictors of violence. These models can improve our prediction of violence but even these are not 100 percent infallible and some dangers will always go undetected.

Predictive models of potential violence involve looking at a person’s stressful life events (e.g., occupational, academic, financial, family or relationship problems), typical coping behaviors and emotional reactions (e.g., intimidating, belligerent, confrontational, easily agitated) to difficult situations, history of violent, reckless or antisocial behavior, threatening statements about killing/harming self or others, and references to or preoccupation with vengeance and violence.

From what little the public knows now, Stephens’ violence was not easily apparent and thus not clearly preventable. For example, according to police, though Stephens had many motor vehicle violations, he had no criminal records. According to his former girlfriend, he was nice to her and her children.

Apparently Stephens reliably held a job as a case manager at a children’s behavioral health agency. Yes, Stephens had pictures and videos on his Facebook account indicating that he had visited a gun range and commenting on being a good shooter.

But many Americans have guns and don’t use them in senseless killings. Also on his former Facebook page, Stephens said he had reached his “breaking point” and just “snapped.” But plenty of us have felt at wit’s end and don’t engage in injurious behaviors to self or other.

Over the past couple years, prominent psychiatrists have taken pen to paper to communicate to the public how most homicides are committed by people without mental illness. This is hard for some to believe and probably not very reassuring. But the truth is drug and alcohol abuse are far more powerful risk factors for violence. However, we still need to study gun violence to understand it's roots and causes.

Joan Cook is a psychologist and associate professor at Yale University who researches traumatic stress and clinically treats combat veterans, interpersonal violence survivors and people who escaped the former World Trade Center towers on 9/11.

The views expressed by contributors are their own and not the views of The Hill.