
New guidance from DHS to agencies on protecting federal employees in the workplace cautions that “it is difficult to predict violent behavior” and cautions against profiling employees.
It noted that studies in the field commonly list certain characteristics and behaviors as early warning signs, including those who make veiled threats, are in repeated conflicts with co-workers or supervisors, and who make statements indicating desperation over personal problems, among others. However, it said that “it is seldom (if ever) advisable to rely” on them.
“Profiles often suggest people with certain characteristics, such as loners and men in their forties, are potentially violent,” it says. “This kind of categorization will not help predict violence, and it can lead to unfair and destructive stereotyping. The same can be said of reliance on early warning signs that include descriptions of problem situations such as someone who is in therapy, had a death in the family, suffers from mental illness, or is facing a reduction in force.”
It says that experiencing such stresses is common and that “all but a very few people weather these storms without resorting to violence. Supervisors should, of course, be trained to deal with the kinds of difficulties mentioned here, such as bereavement or mental illness. However, this training should focus on supporting the employee in the workplace, and not in the context of, or the potential for, workplace violence.”
The guidance, issued just before the two mass shooting last weekend not involving federal workplaces, further says that some personal characteristics may lessen the risk of someone becoming violent.
“For example, a disruptive employee with a strong desire and commitment to complete their federal career and retire may present less actual risk of further violence than another employee who has no expectation of reaching retirement. A subject with strong family connections and no wish to disappoint others similarly may pose less of a risk. Sometimes strong religious conviction will mitigate violence,” it says.
It concludes that “One cannot tally the factors, arrive at a “score,” and then render from that score a probability that violence will occur.”