Do School Threats Really Predict Student Violence?

School violence is rare—but fear of it is not. Every year, thousands of students make statements that sound threatening: angry remarks, impulsive comments, dark jokes, or expressions of distress that raise alarms among teachers and administrators. The hard question for schools has always been the same: Which threats matter, and which do not?

A large new study published in the Journal of Threat Assessment and Management may provide the clearest answer yet. Using more than 14,000 threat assessment cases from nearly 1,700 Florida schools, researchers Jordan Kerere, Dewey Cornell, Jennifer Maeng, and Francis Huang examined what actually happens after a student is flagged for a threat. Their findings cut through common myths—and strongly support a structured, evidence-based approach to school safety.

Most Threats Do Not Become Attacks, but Some Do

The reassuring news comes first. In this statewide sample, almost 9 out of 10 threat assessments ended without any violent attack. That result mirrors earlier studies from Virginia and Florida and confirms what threat assessment advocates have long argued: Most student threats are expressions of emotion, not intent.

But the less comfortable finding is this: about 11 percent of cases did involve an attack, usually a physical fight. While serious injuries were rare—just 0.4 percent of cases—the overall attack rate is not trivial, especially when multiplied across thousands of schools. The real value of the study lies in explaining why some cases escalate while most do not.

The Critical Distinction: Transient vs. Substantive Threats

Behavioral threat assessment is built on a simple but powerful idea: Not every threat reflects a real danger. The Comprehensive School Threat Assessment Guidelines (CSTAG), now used in most U.S. public schools, require teams to distinguish between:

• Transient threats—statements made in anger, frustration, or joking, where there is no real intent to harm
• Substantive threats—statements or behaviors suggesting sustained intent or unclear intentions, requiring protective action

This distinction is not theoretical since, according to available data:

• Less than 10 percent of transient threats resulted in an attack
• About one-third of substantive threats did

When schools classified threats as serious or very serious substantive threats, the odds of an attack increased dramatically—by more than tenfold compared to non-threat cases.

In practical terms, this means well-trained school teams are not guessing. They are identifying real risk with meaningful accuracy.

Who Is More Likely to Be Involved in an Attack?

The study also examined student characteristics associated with higher odds of an attack following a threat assessment. Several factors stood out, though the effects were modest:

• High school students were more likely than younger students
• Students with individualized education plans (IEPs) showed a slightly higher risk
• Students receiving free or reduced-price meals were more likely to be involved
• Black and Hispanic students had somewhat higher odds than White students

These findings require careful interpretation. They do not suggest inherent risk tied to race or socioeconomic status. Instead, they likely reflect broader inequities—differences in exposure to stress, trauma, disciplinary practices, and access to support services.

Threat assessment is not about profiling. In fact, one of its goals is to reduce unnecessary exclusionary discipline by replacing zero-tolerance reactions with individualized intervention.

Education Essential Reads

Andragogy Doesn’t Explain Learning—It Explains Control

4 Signs Your Child Has a Good Teacher

Does Community Violence Spill Into Schools?

One unanswered question in earlier research was whether schools in high-crime communities face greater danger after a threat is made.

This study found a nuanced answer.

• Community violent crime rates did not predict whether an individual student would attack
• However, districts with higher juvenile violent crime arrest rates did show slightly higher overall school attack rates

In other words, community context matters at a systems level—but it does not determine the behavior of a specific student. That reinforces the importance of focusing on individual assessment, not assumptions based on neighborhood or background.

What This Means for School Safety

Several conclusions are hard to ignore:

1. Threat assessment works—but only when done properly. The transient/substantive distinction is not just a bureaucratic label. It is a validated risk indicator that helps schools allocate attention where it is most needed.
2. Overreaction is a real danger. Removing students from school for non-serious threats does not improve safety and may worsen outcomes. This study supports interventions over punishment.
3. Training and resources matter. Attack rates varied widely across districts—from zero to nearly half of cases—suggesting differences in implementation quality, staffing, and support systems.
4. Serious injuries are rare, but prevention still matters. The absence of shootings or fatalities in this large sample is encouraging. It also suggests that early identification and intervention may be working.

The Bottom Line

School threats are common. School attacks are not—but they are predictable.

This study shows that schools can identify higher-risk situations with reasonable accuracy, intervene early, and prevent most threats from escalating to violence. Going forward, the challenge lies not in inventing new models but in ensuring the consistent, fair, and adequate implementation of existing ones.

The uncomfortable truth is that no system can eliminate risk entirely. The hopeful truth is that evidence-based threat assessment remains one of the strongest tools schools have—not just to prevent violence, but to keep students connected to school rather than pushed out of it.