Behind the Data: How We Analysed Violence Against Teachers in U.S. Schools

While public discourse often centres around students’ safety and well-being—and scientific literature has surmountable studies on peer bullying—we set out to explore a largely overlooked dimension of school violence: when teachers become the victims.

At Sunstone Institute, this research area is part of a broader effort to understand education system resilience on a global scale. Examining school violence—particularly against teachers—offers insight into the conditions that undermine stability, as well as the knowledge or data gaps in how harm is recognised, measured, and addressed. Drawing on federal data, national surveys, and peer-reviewed studies, we set out to explore the school environments teachers are working within, and the prevalence of violence against them.

Violence against teachers is a global concern, but few countries provide the depth of national data that the United States does. Initially, we considered a comparative study across four countries—Sierra Leone, Estonia, U.S. and Nepal—as they all represent different stages of economic development and approaches to educational reform. However, the U.S. provided the most comprehensive and accessible data, allowing for a more granular analysis.

While bullying between students and school crime are well-documented, we chose to focus specifically on violence directed at teachers by students. Our analysis centred on a range of violent and aggressive behaviours, including verbal threats, intimidation, non-contact aggression (e.g., throwing objects), physical assault, cyberbullying, and sexual harassment.

For this topic, we set out to better understand how often and in what ways teachers experience violence in school settings—with a particular focus on the student as the aggressor. Our secondary focus included teacher violence against students, specifically corporal punishment and educator sexual misconduct, to present a fuller picture of the dynamics of power and harm within schools.

To capture the complexity of violence against teachers, otherwise known as ‘teacher victimisation’, we relied on a mixed-methods approach using five main data sources:

1. Global Context and Meta-Analysis

The term teacher victimisation is not used consistently across the academic literature. To capture the widest possible range of relevant research, we searched for a broad set of terms that reflect varying conceptualisations of the issue. This also allowed us to build a more complete picture of how researchers across contexts are approaching the topic.

Here is a list of terms we searched for on Google Scholar and Google search engine: learner-on-teacher bullying, teacher victimisation, student-on-teacher violence, violence against teachers, student-to-teacher victimisation, student violence against teachers, violence directed at teachers, violence against educators, aggression against educators, student violence against teachers, abuse of teachers, student violence against educators, student-generated violence directed against teachers, teacher-directed violence, and student-perpetrated violence against teachers.

Though our analysis focused on U.S. data, we reviewed international literature to frame the U.S. within a global context. A meta-analysis on student violence against teachers by Longobardi et al.’s (2019) pooled 24 studies globally and found that “the prevalence of any type of teacher-reported violence victimization within two years or less ranged from 20% to 75%”. 

In addition, a number of U.S. states allow corporal punishment in schools, which is unusual compared to many other high-income countries. This raised questions about its possible link to violence against teachers. Unfortunately, we were unable to explore the relationship due to insufficient data on a granular level.

2. National Center for Education Statistics (NCES)

Since 1867, NCES has been the federal statistical agency responsible for collecting, analysing, and reporting data on the condition of U.S. education, from early childhood to adult education, offering a broad view of the factors shaping school life.

NCES include surveys such as the School Pulse Panel (SPP), which provided crucial context on staffing shortages, behavioural disruptions, and educator well-being. In February 2024, SPP reported 82% of schools reported difficulty filling teaching positions, while 95% of school leaders cited student mental health and 92% cited staff mental health as top concerns.

NCES also tracks incidents of violence against teachers. Table 228.70, for example, reports the number and percentage of public school teachers threatened with injury or physically attacked by students in the previous 12 months, covering data from 1993–94 through 2020–21.

These findings helped us understand not only the prevalence of teacher victimisation, but also the pressures shaping the school climate.

3. American Psychological Association (APA)

In 2019, the American Psychological Association (APA) convened a Task Force on Violence Against Educators and School Personnel to examine the experiences of teachers and school staff facing violence and aggression. Led by psychologist Prof. Susan McMahon of DePaul University, the research team surveyed more than 14,966 educators, including 9,370 teachers, before and during COVID-19 restrictions, and 11,952 educators, including 8,153 teachers, after restrictions were lifted.

We initially reviewed data collected by the APA during and after the COVID-19 pandemic, but excluded it due to lack of access to raw data and inconsistencies in derived prevalence rates. The study on the prevalence of multiple forms of aggression from students, parents, and colleagues during pre-COVID times remained the most comprehensive and reliable for our analysis.

4. BMC studies (Moon et al.)

Led by Professor Byongook Moon, these studies included a longitudinal project in Texas (2016–2019) and later an expansion across 18 states and more than 4,000 teachers (2022-23). The studies offered insights into seven types of victimisation, including verbal abuse, theft/vandalism, in-person bullying, cyberbullying, sexual harassment, non-contact aggression, and physical assault.

The 2022 study sample was larger and more geographically diverse, confirming earlier prevalence rates. These studies were crucial in understanding victimisation severity, and its emotional and professional consequences, offering a rare look at the mental health toll, as well as trust towards students and quitting intentions.

The data from each year were normalised to per 1000 teachers, to analyse all the studies collectively.

5. Civil Rights Data Collection (CRDC)

The CRDC, maintained by the U.S. Department of Education, defines corporal punishment as paddling, spanking, or other forms of physical discipline. It has tracked its frequency and demographic breakdowns since 2000. While corporal punishment has declined significantly, disparities remain among Black and American Indian/Alaska Native students, as well as disabled and male students. The practice remains concentrated in southern states, including Mississippi, Texas, Arkansas, and Alabama.

In 2021, the CRDC began disaggregating data on rape and sexual assault in schools by whether the perpetrator was a student or staff member. Of the 19,271 reported cases that year, 313 (about 2%) involved school personnel. An additional 1,319 allegations of staff-perpetrated assault included cases where individuals resigned, retired, were reassigned, or remained under investigation.

Our goal was to assess prevalence of each violence or victimisation type and how each one impacts teacher’s well-being and job performance, rather than identify longitudinal trends, as this was not possible with the existing data. Because each dataset used different methodologies, sample populations, and timeframes, direct comparison or aggregation was not always possible.

●  APA and BMC studies used self-reported survey data, introducing potential bias or underreporting. Sunstone Institute also reached out to the researchers behind the APA and BMC studies to gain further insight into teacher victimisation, but received no response after several email attempts.

●  NCES rely on institutional reporting, and does not include more forms of violence other than ‘physically attacked’ and ‘threatened with injury’.

●  Victimisation scale that explores different types of violence and aggression varies across datasets, further complicating cross-study comparison.

Despite these limitations, the available data paint a consistent picture: violence against teachers is not isolated, and the systemic conditions that allow it to persist remain largely unaddressed.