In the fall of 2020, a prominent technology company in California contacted forensic psychologist J. Reid Meloy, PhD, a threat assessment researcher, because one of their software engineers had posted worrisome messages on a social media platform. The messages alluded to plans of a violent attack. The platform flagged the posts and alerted the FBI. As Meloy worked with the tech company’s threat assessment team on the case, they discovered that the man had been struggling with social isolation—especially during the COVID-19 pandemic—and he had started connecting with political extremist groups online. When police entered his home with a warrant just before the November 2020 election, they found illegally modified assault weapons and evidence of online searches pinpointing the locations of public figures.
“Ironically, the case never received any news coverage because nobody was murdered, but this demonstrates the power of paying attention to warning signs,” said Meloy, a clinical professor of psychiatry at the University of California, San Diego, and a consulting psychologist for the FBI. “More than half of public mass murderers leak information about their intentions to family members, friends, teachers, coworkers, on social media, or in other ways,” he said (Silver, J., et al., Aggression and Violent Behavioropens in new window, Vol. 38, 2018).
People on the pathway to violence usually give more extensive information about their plans to family members and close friends than to supervisors, teachers, or coworkers, but family and friends are less willing to report this information, said Meloy. This is because reporting has primarily led to punitive consequences for the person of concern, but psychology researchers are uncovering the benefits of using leaked information as an opportunity to learn more about an individual’s situation and offer social and psychological interventions.
“There is usually a long pathway to violence, and so many missed points of prevention,” said Jillian Peterson, PhD, an associate professor of criminology and criminal justice at Hamline University in St. Paul, Minnesota.
Eager to better understand the trajectory to violence, Peterson partnered with sociologist James Densley, PhD, to launch The Violence Projectopens in new window, a study of 168 mass shootings from 1966 to 2020. They defined mass shootings as incidents in which four or more victims were murdered with firearms in public locations, and 2017 to 2019 were the worst years on record for the number of shootings, with a combined total of 23. Aided by a $300,000 grant from the National Institute of Justice, they gathered a team to comb through attackers’ journals, manifestos, blog and social media posts, interview transcripts, and other sources. The team searched for information about more than 100 life-history variables, including mental illness, trauma, academic performance, and bullying. They also contacted 32 living mass shooters—a small number of the total because most had died by suicide during an attack or had been killed by police—and five agreed to participate in the study through letters or phone interviews from prison.
The team discovered that 42% of the shooters had experienced early childhood trauma and exposure to violence at a young age, such as physical or sexual abuse, having a parent commit suicide, bullying, or witnessing domestic violence. More than 80% had reached a crisis point in the hours, weeks, or months leading up to the incident. About 50% had been reprimanded, suspended, or fired from work shortly before committing the crime, and 1 in 4 had lost a romantic relationship before the attack. These crises had triggered noticeable changes in behavior in most shooters, such as increased agitation or isolation (The Violence Project: How to Stop A Mass Shooting Epidemicopens in new window, Abrams Press, 2021).
Meloy has studied the patterns of mass murderers for 25 years, and his research team has identified eight proximal warning behaviors suggesting that someone is on the pathway to violence. For example, “fixation” is increased preoccupation with a person or cause, accompanied by deterioration in social or occupational function. “Pathway” warning behavior is researching and planning an attack, and “identification” is identifying with previous mass shooters, such as when Adam Lanza compiled a spreadsheet with details of hundreds of mass violence incidents before he launched the Sandy Hook Elementary School attack in 2012 (Journal of Threat Assessment and Managementopens in new window, online first publication, 2021).
In the case of the software engineer in California, the social media company’s algorithms flagged the warning behavior. The man was transferred to a psychiatric facility, but Meloy was concerned about the transition from hospitalization back to the community. “This is often a weak point in the process when people can slip through the cracks,” he said. Meloy found a psychiatrist and a therapist who were both trained in forensics; they met with the man regularly, and the tech company checked in with him frequently. He received months of treatment as his criminal case slowly proceeded, and he disengaged from the political extremist groups. The court ultimately approved his decision to relocate to the Midwest to be near family. “He is a talented individual, and now he’s looking into further work and education,” Meloy said.
Paying attention to suicide risk
The Violence Project data also revealed that 72% of the shooters in the database were suicidal either before or at the time of the attack. This doesn’t surprise Karie Gibson, PsyD, a supervisory special agent for the FBI’s Behavioral Analysis Unit. She analyzes incidents of mass violence and works with law enforcement agencies on cases to prevent attacks. Perpetrators frequently experience suicidal ideation before a shooting because they feel helpless and hopeless, Gibson said.
They often have grievances against people or organizations, which become rationalizations to make others suffer before killing themselves, she explained. If someone is a grievance collector and becomes suicidal, it’s important to learn if they are thinking about or have gotten pleasure from thinking about hurting others.
“Sometimes psychologists do not probe deeply enough and they let the patient dictate the interaction,” Gibson said. “At the heart of it, these people are hurting and have unmet needs, and asking these questions will help mental health providers be more strategic about intervening.” If psychologists identify warning behaviors, they can share the safety concerns with the patient and explore the possibility of voluntary hospitalization. If a patient is a danger to themselves or others but does not agree to this, then psychologists are usually required to break patient confidentiality and pursue involuntary hospitalization, said Gibson.
Gibson also teaches mental health providers about the timing of different stages in the pathway to violence. More than 62% of attackers are in the research-and-planning stage for 1 month to 2 years, and 54% are in the preparation phase—when an individual acquires supplies for the attack—for 7 days or less, according to a 2018 studyopens in new window published by the FBI. “If someone is in the preparation phase, it’s critical to respond quickly,” Gibson said.
Gibson’s FBI unit receives three to five referrals per week from law enforcement agencies throughout the country, and she works with probation officers, mental health providers, court-ordered psychologists, and other stakeholders to assist the person of concern. “We have to share information with one another to understand what is really going on and how to help the individual get to higher ground,” she said.
Making communities safer
According to The Violence Project, 31% of the mass shootings in its database occurred in workplaces—followed by retail sites (16.7%), restaurants (13.7%), and K–12 schools or universities (11.9%). The vast majority of workplace shooters were insiders in the organization.
Psychologist Stephen White, PhD, president of Work Trauma Services, a threat management consulting firm in San Francisco, integrates research findings on risk factors for violence with years of case experience to help organizations reduce the likelihood of a workplace tragedy. White has helped hundreds of companies throughout the country and globally develop multidisciplinary threat assessment teams that are trained to respond when employees share concerns about a coworker. He also consults with these teams—which may include representatives from human resources, security, and the legal department—when they need an expert for a more difficult case. “Behaviors such as making threats or intimidating people are not reliable predictors in themselves of a serious threat,” said White. “But they must be taken seriously.”
White teaches teams how to talk to people who have been reported by coworkers. “Explain that there is a concern about harm, and that you want to hear their version of the story,” he said. “These individuals usually feel humiliated and marginalized, and they have often experienced significant losses or rejections.” In most cases, these conversations reveal that the person who made a threat does not actually pose a threat. Even when employees do not seem serious about hurting others, they may still benefit from having the opportunity to express their frustrations or grievances, White said.
One of the most common mistakes organizations make is hastily firing someone for threatening behavior. This may be viewed as yet another insult or intolerable loss by the employee. Instead, termination can be postponed while the threat assessment team gathers information about the situation. For employees who are interested in accessing mental health services, White encourages companies to consider offering paid leave or a severance package with extended COBRA benefits. “Many times, the employee no longer wants to work at the organization,” he said. “This allows them to find a respectful way out of the dilemma and get support.”
Schools are another common target for shootings, and forensic psychologist Dewey Cornell, PhD, is at the forefront of the effort to establish effective threat assessment programs in K–12 environments. Cornell, a professor of education at the University of Virginia, developed a five-step model geared to help schools support at-risk students rather than automatically suspend or transfer youth who have made a threat. “Schools need to recognize that concerning behavior usually means a student is having a problem that he or she can’t resolve, so we need to take a problem-solving approach,” he said.
In Cornell’s Comprehensive School Threat Assessment Guidelines (CSTAG), he teaches teams how to gather information from witnesses and the student of concern to discern the severity of the threat. Only in the most serious situations—less than 10% of cases—the team should notify law enforcement and ask a school psychologist to conduct a mental health assessment. He also teaches teams how to develop a plan for a safe return to school for the student, which could include counseling, a check-in process with administrators throughout the day, and community-based treatment.
Cornell’s team has trained staff and administrators at more than 1,000 schools throughout Virginia to use the model, and the program reduced school suspensions by 49% and school transfers by 80%. Students in schools using the model were 4 times more likely to receive counseling services (School Psychology Reviewopens in new window, Vol. 41, No. 1, 2012). A recent study of schools that used the program from 2013 to 2019 showed that all threats were resolved without serious injury.
This approach to threat assessment also reduces racial disparities related to discipline. Cornell and his colleagues found no disparities among Black, Latinx, and White students in out-of-school suspensions, school transfers, or legal actions in schools using any type of threat assessment (School Psychology Reviewopens in new window, Vol. 47, No. 2, 2018).
In 2020, the state of Florida adopted the program, and Cornell’s team provided trainers for 4,000 schools. While the growing interest in the model is encouraging, his work in the Virginia schools has shown that maintaining a program year after year is far harder than starting a new one. “Institutions do not change easily,” he said. “There were districts who took the model to heart by committing time and resources to train new staff and organize team meetings, but this was not the case in every school.”
Although the willingness to change varies, administrators and teachers have shared with Cornell how much the program has helped them navigate difficult situations. One team in Virginia worked with a high school student who made a threat because he felt depressed, alienated, and bullied. The team facilitated mental health counseling and regular parent communication, and they placed him in a special education program with a high staff-to-student ratio. When the student later heard about a school shooting on the news, he thanked members of the team for helping him avoid that kind of notoriety. In 2020, he successfully graduated from high school.
“These kinds of experiences tell me we’re on the right track,” said Cornell. “It’s a reminder that there is a tremendous need to establish standards for training and practice so that all schools can achieve the beneficial outcomes that threat assessment can offer.”
