Bridging Data Gaps to Illuminate and Eliminate Violence Against Children in School

Schools should provide a safe and nurturing environment where children are able to learn and thrive. But, in a terrifying perversion of the role school should play in a child’s life, children are often harmed and abused in and around schools by teachers, other school staff, and by fellow students.

Now let’s say you’re an education policy adviser in a country. Imagine that you have championed this issue in your country and gained an audience with the minister of education. When you explain how serious a problem violence in schools is, she is aghast and wants to take action! She would probably ask a series of follow-up questions to guide that action. How many children were sexually abused by teachers last year? How many had physical injuries resulting from corporal punishment by teachers? How linked are youths' experience of violence with dropout rates in the country? She points out that her predecessor implemented an initiative to reduce violence in schools: have rates of violence fallen in the wake of that initiative?

In a new CGD working paper, we assess whether you could answer the minister's questions.

Specifically, we examine the state of data on school-related violence in low- and middle-income countries (LMICs) and reveal alarming gaps in our knowledge---despite the fact that many surveys include at least some questions about school-related violence. The findings underscore the urgent need for comprehensive data to inform effective policy making to make schools safer.

We constructed a comprehensive database of surveys to provide a full picture of the state of the data on violence in schools

We reviewed nationally representative international surveys conducted between 2013 and 2023 for questions about violence in school during childhood and adolescence, encompassing violence occurring at or around school and perpetrated by teachers, students, or others. We complemented this by randomly selecting 30 percent of low- and middle-income countries and searching for nationally representative surveys on violence in schools in those countries. The resulting dataset (which you can read all the details of here) provides information about each international and national survey, including country coverage, survey year, target population, and question content.

We then set out to analyse whether the available data would help policymakers to take action against school-related violence

Policymakers need to know more than the level of violence in order to take action against it. We propose six questions that an education minister might ask about violence in and around schools. These include how many children were harmed by their teacher last year, whether violence in schools is getting better or worse over time, and does violence in schools lead to higher school dropout rates? These questions are not comprehensive—there are lots of aspects of violence in school that a policymaker might care about. But exploring whether the available data will let us answer these questions gives an illustration of whether policymakers would have the kinds of data they need to plan new policies and interventions. In most cases, they wouldn’t (see figure below).

Note: The figure does not show the answer to the question, but rather the low percentage—in most cases—of countries where the question could even be answered in response to a policymaker query.

Six things we learn about data on violence in schools from this analysis

  1. There is very limited data on sexual violence in schools. International surveys primarily focus on physical and psychological violence, largely ignoring sexual violence. Only 17 percent of countries have data on sexual violence perpetrated by school staff and no LMIC country can measure sexual harassment on the way to or from school. This gap in data is a major barrier to addressing this rampant and egregious abuse of children.

  2. Most LMICs cannot measure whether school violence is getting better or worse over time. Monitoring trends in school violence is essential to gauge the effectiveness of interventions to make schools safer. Yet our paper reveals that only 25 percent of low and lower-middle-income and half of upper-middle-income countries have conducted repeated surveys to track changes in school violence over time. And where such surveys exist, they predominantly cover physical and psychological bullying, sidelining other forms of violence.

  3. The available data do not provide enough information on the consequences of violence. Only a small fraction of countries have data on injuries resulting from violence. We did not find any surveys that allow us to understand whether suffering violence in school is directly connected to poor academic performance or higher dropout rates. (Some international surveys would allow examining a correlation of past school violence and dropout rates, but without any direct attribution.)

  4. Data on violence among vulnerable groups and younger children is lacking. None of the existing international surveys provide data to explore school violence statistics among LGBTQ+ children and children with disabilities. And only one international survey—the ERCE covering countries in Latin America and the Caribbean—collects data on violence suffered by children below nine years old in school.

  5. The surveys that do cover violence in schools are too different to allow meaningful comparisons between them. The specific questions about violence in school vary significantly across surveys, in terms of how the questions are formed and the options for responses. This lack of standardisation renders cross-survey comparisons difficult—both for a given country over time and for countries across a region. To provide actionable data on whether violence is being curbed, we need a standardised module on violence in schools.

  6. There are big disparities in data availability across regions. Latin America and the Caribbean and East Asia and the Pacific have more robust data collection efforts compared with Sub-Saharan Africa, in both international and national surveys. But even in those regions, few countries have the data to answer a Minister’s simple, follow-up questions.

Time to get serious about measuring violence in school

Data on school violence aren’t academic; they are the foundation for effective action. There is an urgent need for better data collection efforts focused on school-related violence. We propose that standardised measures of violence in school be developed, with consistent definitions, question framing, victimisation timeframes, and frequency measures, enabling meaningful comparisons over time and across countries, alongside dedicated modules within existing surveys that measure violence in schools. In particular, we need to start measuring the experiences of younger children and specific vulnerable groups, including children with disabilities and LGBTQ+ children, all while making sure we protect children in the process...

All our efforts to get more children in school and learning will be worth little if they aren't safe. Our new study shows that data limitations hamper effective policy-making. This needs to change.


CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.

Image credit for social media/web: rod / adobe stock