Response, Not Prevalence: What Administrative Bullying-Discipline Data Reveal

Schools are complex ecosystems where the safety of a child depends on a web of factors: the behavior of peers, the stability of the classroom, the support of families, and the broader conditions of the neighborhood. For decades, educators and policymakers have relied on official records to track bullying, treating these documents as a direct window into how often students are being hurt. The logic seems sound: if a student is bullied, a teacher sees it, a report is filed, and a punishment is recorded. By counting these records, officials believe they can measure the prevalence of bullying and judge whether a school is safe. However, this assumption overlooks a critical gap. A record in a file is not the same as an event in a hallway; it is the result of a specific chain of actions where an incident must be noticed, labeled, and processed by the administration before it ever appears in the data.

A team of researchers set out to examine what these official records actually reveal about the reality of bullying in schools. They looked at thirteen years of data from public schools in Massachusetts, covering nearly 1,840 schools and over 23,000 school years. Their goal was to understand the difference between the bullying that happens and the bullying that gets written down. They asked whether the numbers on a spreadsheet reflected the true amount of bullying students experienced, or if they merely reflected how schools responded to it. By comparing these official discipline records with independent surveys where students reported their own experiences, the researchers uncovered a startling disconnect. The data suggested that the official numbers are not a measure of how much bullying occurs, but rather a measure of how schools handle the incidents they choose to address.

The study began by looking at the sheer volume of bullying-related discipline. The researchers found that these records were surprisingly rare. Across the entire state, 81 percent of school years showed no recorded incidents of bullying-related discipline at all. The median school recorded zero incidents every single year. Furthermore, the number of recorded cases dropped significantly over time. Between 2013 and 2025, the statewide rate of students disciplined for bullying fell by 44 percent. At first glance, this might look like a success story, suggesting that schools had become much safer and that bullying was disappearing. However, when the researchers compared these falling numbers with data from a separate statewide survey where students were asked directly if they had been bullied, the picture changed completely. While the official discipline records were going down, the percentage of students reporting that they had been bullied was actually going up. In 2018, about 25 percent of students reported being victimized; by 2024, that number had risen to 30 percent. The two trends moved in opposite directions, diverging sharply except during the pandemic year of 2021, when schools were closed and in-person contact was minimal.

This divergence suggests that the decline in official records does not mean bullying has vanished. Instead, it indicates that fewer incidents are making it into the formal disciplinary system. The researchers found that the official records capture only a small fraction of the bullying that is reported. For every student who was disciplined for bullying, there were roughly eleven reports of bullying that did not result in formal punishment. This gap highlights that the administrative data is a measure of institutional response, not a direct count of student behavior. A zero in the record does not necessarily mean no bullying happened; it may simply mean the incident was handled informally, was not reported, or was not classified as a disciplinary offense.

The study also investigated what factors predicted whether a school would have any bullying records at all, and what predicted the severity of those records. The researchers used a two-step approach to separate these questions. They found that the decision to record any bullying at all was driven mostly by the size of the school and its general disciplinary climate. Larger schools were much more likely to have at least one recorded incident, likely because they have more students and more opportunities for incidents to be noticed and reported. Once a school had crossed that threshold and recorded at least one incident, the number of cases was influenced by different factors. In schools that did have records, the severity of the bullying discipline was linked to lower academic achievement and a stricter overall disciplinary climate. Schools with lower test scores and more frequent disciplinary actions in general tended to have more bullying-related discipline.

Surprisingly, the researchers found that the community surrounding the school did not independently predict these outcomes. Common assumptions suggest that schools in poorer neighborhoods or areas with higher crime rates should see more bullying. However, after accounting for the school's own size, climate, and academic performance, the socioeconomic conditions of the neighborhood and local crime rates did not explain the differences in bullying records. The only community factor that mattered was whether the school was located in a city versus a suburb. City schools were more likely to have records, but the data suggested this was because more bullying was entering the reporting system in cities, not because the schools were punishing similar incidents more harshly. The number of reports per incident was similar in cities and suburbs.

The researchers also looked at how changes within a single school over time affected the records. They found that when a school's behavioral climate changed, the bullying records changed with it. If a school saw an increase in chronic absenteeism or a shift in its overall disciplinary tone, the number of bullying incidents recorded tended to shift as well. This suggests that the internal environment of the school—the way adults respond to behavior and the general atmosphere of the classroom—plays a larger role in the records than the specific mix of students attending the school. The data did not show that changes in the student population, such as the percentage of low-income students or students with disabilities, drove the changes in bullying records.

One of the most significant findings concerned the groups of students who were most affected. The student surveys showed that bullying was reported most frequently by English learners, Black students, low-income students, and students with disabilities. Yet, the official administrative records did not reflect these disparities in the same way. Schools serving more non-White students were not more likely to have bullying records; in some cases, they were less likely to have them. This mismatch raises important questions about equity. It suggests that the official records may be missing the bullying that happens to the most vulnerable students, perhaps because these incidents are less likely to be detected, reported, or classified as formal disciplinary offenses.

The researchers concluded that administrative bullying records should not be used as a standalone measure of school safety or bullying prevalence. These records are filtered indicators that tell us more about how a school responds to behavior than about how much bullying is actually occurring. A decline in the number of recorded incidents does not prove that bullying has decreased; it may simply mean that schools are handling incidents differently, or that fewer are being reported. To truly understand school climate, officials need to look beyond the discipline files. They must pair these records with student surveys, anonymous reporting systems, and qualitative information from school staff. Only by combining these different sources can educators get a clear picture of the safety of their schools and ensure that the students who need help are actually receiving it. The study serves as a reminder that what gets measured is not always what is happening, and that the silence of a file folder can sometimes be just as loud as the noise of a crisis.