ProPublica is a nonprofit newsroom that investigates abuses of power. Sign up for Dispatches, a newsletter that spotlights wrongdoing around the country, to receive our stories in your inbox every week.
After a former student killed six people last year at the private Covenant School in Nashville, Tennessee, state leaders have been looking for ways to make schools safer. Their focus so far has been to ramp up penalties against current students who make mass threats against schools.
Months after the killings, legislators passed a law requiring students who make such threats to be expelled for a year (unless a school superintendent decides otherwise) and allowing schools not to enroll them afterward. This year, the legislature passed bills that make the offense a felony and that revoke driving privileges for a year.
But a large body of research shows these zero-tolerance measures are not the most effective way to prevent violence in schools. In fact, some experts say those measures can counteract what they consider a crucial tool for protecting students as well as the larger community: threat assessments. When carried out correctly, threat assessments sort out behavior intended to cause real physical harm from simply disruptive acts and provide troubled students with the help they need.
The Secret Service pioneered threat assessments to help identify viable threats against public officials. After the Columbine High School massacre in 1999, a team of University of Virginia professors began adapting Secret Service and FBI threat-assessment recommendations for use in schools. They relied on reports from the two agencies showing that school shooters typically expressed their intentions well before acting violently, but that those statements and the underlying potential for harm were seldom thoroughly investigated.
In the course of their research, they learned that educators were concerned about overreacting to students who did not pose a serious threat.
“We also know that students frequently make threatening statements just in the routine course of their day. We have to be very careful that we don’t confuse the two,” said psychologist Dewey Cornell, who led the University of Virginia research and continues to study threat assessments. “And so we need a systematic process to sort out serious threats from threats that are not serious.”
We spoke to Cornell about how schools are handling threat assessments and his concerns about their overreliance on harsh discipline. More than two decades after he began his research, a growing number of states, including Tennessee, require school districts to adopt threat-assessment policies. But Cornell worries that too many are not properly carrying them out.
How Threat Assessments Work
Threat assessments are intended to be an alternative to zero tolerance, giving school leaders a way to resolve problems before they escalate to violence and allowing them discretion over whether and how to discipline students.
Cornell helped create a process to help school administrators carry out threat assessments, starting by interviewing anyone involved with the threat to assess the risk of serious injury. His research shows that the majority of the time, the assessment reveals there was no threat or no serious threat. The threat-assessment team should warn the intended victims of any major threats it finds, take necessary precautions to protect them and seek ways to resolve conflict.
One of the most important parts of the process requires the threat-assessment team to refer students to mental health services they may need, no matter the level of threat. And the process states that law enforcement involvement and harsh disciplinary penalties should be reserved for the most serious cases.
Cornell’s initial research in Florida and Virginia shows that, when done well, threat assessments reduce expulsions and keep more students in school. In 2013, Virginia was the first state to mandate that public schools adopt threat-assessment teams. By 2018, nearly 80% of schools in the state reported at least one threat-assessment case; about three-fourths of cases resulted in students being referred for counseling, mental health treatment or psychological assessments. Most students were not expelled, placed in an alternative school or juvenile detention, or hospitalized as a result of a threat assessment.
A 2024 study of Florida public schools conducted over three years found that the implementation of threat assessment had been “widely, but not uniformly, successful,” with a third of cases resulting in a student being referred for mental health services and just 2% resulting in an expulsion.
“We’re not just looking for the needle in the haystack, that rare student, the one in a million students who’s going to actually shoot someone,” Cornell said. “We’re actually dealing with thousands and thousands of kids who maybe are angry or upset and so they say something threatening. Maybe they’re being bullied. We have an opportunity to intervene and work with them long before there’s any issue, if there is any issue at all.”
Threat Assessments and Zero Tolerance Don’t Work Together
According to Cornell, zero-tolerance policies and threat assessments are “antithetical” approaches to handling school safety. Tennessee legislators, however, passed a law mandating that school districts conduct threat assessments less than two weeks after they passed a law requiring expulsion for mass threats.
Zero tolerance requires school officials to automatically punish students who act out, no matter the circumstances. Threat assessment, on the other hand, requires officials to consider the context and motivation of the behavior before deciding how to respond.
There is no research showing that zero-tolerance policies make schools safer, according to a review of available evidence by the American Psychological Association. In fact, such policies can harm Black students and students with disabilities, who are more likely to be suspended or expelled from schools with zero-tolerance discipline policies and, by extension, more likely to end up in the criminal justice system, studies show.
“So we have a disciplinary practice that research tells us does not work, yet we keep doing more and more of it,” Cornell said. “It’s the educational equivalent of bloodletting. The medical field for years would bleed people, and when they didn’t get better, they concluded they didn’t bleed them enough.”
After Tennessee lawmakers made “threats of mass violence on school property” a zero-tolerance offense last year, they received numerous complaints about students being arrested and disciplined even when they clearly didn’t pose a serious threat. At a recent education committee hearing, a lawmaker referred to a case in which a middle schooler threatened to fly a plane into the school. “I don’t know too many 12-year-olds that either A, have access to an aircraft, or B, know how to fly it,” he said. Tennessee has not released statewide numbers on expulsions for threats of mass violence.
Lawmakers are now considering a bill that would require a threat assessment to be completed and the threat to be deemed valid before an expulsion. Cornell said that limiting expulsions to valid threats still could pose safety risks. “If I am really concerned that a child is dangerous, I don’t want to turn them loose in the community without supervision,” he said. “I want to try to reach them and work with them and convince them that there is a better way to deal with whatever problem or concern they have.”
When Threat Assessment Goes Wrong
Civil and disability rights advocates argue that threat assessments can do more harm than good and point to examples of school officials referring students with disabilities and students of color for threat assessments more often. In Albuquerque Public Schools, for example, children with disabilities and Black children made up a disproportionately high percentage of those referred for threat assessments, according to a Searchlight New Mexico report. A study of four Colorado school districts found that Black students, Native American students, male students and students with disabilities were overrepresented in the threat-assessment data. No national study exists showing how schools are implementing threat assessments; Cornell has conducted studies in Virginia and Florida and is working on a national one with funding from the federal Department of Justice.
A Texas Observer investigation into threat assessments showed the vast majority of districts in the state failed to properly implement them. Only a small percentage of districts provided students with needed mental health support and other services as required by state law. In Tennessee, school officials are required to include law enforcement in their assessment of whether a student poses a threat — but including mental health professionals is optional.
“If you’re installing threat assessment in a school that doesn’t have a school psychologist, doesn’t have a school social worker and has one counselor for every thousand students, you’re gonna have a problem. It’s like putting new tires on a car with a busted engine,” Cornell said. “We have a school system that is strained and stretched to the limits.”
But Cornell said the reports of schools failing to properly carry out threat assessments shouldn’t serve to indict the entire idea. His research in Florida shows that Black students experienced slightly higher rates of out-of-school suspensions and expulsions than white students — a much smaller disparity than the national average. Students with disabilities did not receive harsher discipline or legal action than other students after a threat assessment.
“States seem to be willing to spend millions of dollars on security hardware, but almost nothing on training and coaching,” he said. “The result is many schools are implementing threat assessment without adequate training, without the time allotted to them to carry out the procedures that are required. So they end up cutting corners.”
Aliyya Swaby is a reporter in ProPublica’s South unit covering children, families and social inequality.
