- One in four adolescents aged 13 to 17 experiences school bullying. That violence affects how they learn, how long they stay in school, and the opportunities available to them for the rest of their lives.
- The evidence shows that schools do more than improve learning. They can also reduce violence, build social-emotional skills, and change the life trajectories of millions of children and young people.
A teenager changes his route to avoid the corner where someone is waiting for him. Another student walks into class knowing that before recess, classmates will corner and humiliate him in front of his peers. Other children face violent punishment and discipline at home before and after school. For millions of students across Latin America and the Caribbean, violence isn't a one-off event — it's part of everyday life.
One in four adolescents aged 13 to 17 experiences school bullying, according to the latest UNICEF and the Pan American Health Organization(PAHO) report. The problem affects girls (25%) and boys (25.3%) at nearly identical rates, though boys report more physical bullying while girls report more psychological aggression and sexual harassment. Nearly half of all children and adolescents (46%) say they have suffered emotional abuse — the most common form of violent discipline in the region — while almost four in ten (38%) have experienced physical punishment. These numbers describe an everyday reality that shapes how much children learn, how they relate to others, and how far they get in school and in life.
At the Inter-American Development Bank (IDB), we have a long track record of helping countries across the region strengthen their education systems and build evidence on which policies truly improve children's and young people's lives. In recent years, that agenda has expanded to keep students safe at school and to understand how education itself can help prevent violence, build social-emotional skills, and create safer communities.
To mark International Youth Day, the IDB is launching Schools that Care, a blog series that brings together the evidence on what works to prevent violence starting in the classroom and how education policy can better protect children and young people. This first post opens a series that will bring together the latest evidence, experiences from Latin America and the Caribbean and beyond, and key questions governments should consider when putting proven approaches into practice.
This exposure to violence has direct, measurable consequences for how children do in school. Those who experience it miss more classes, learn less, and drop out more often — and when violence declines, their reading, math, and science scores rise. A child who learns less today becomes an adult with fewer opportunities and lower earnings tomorrow. School violence is estimated to cost the world close to $11 trillion in lost lifetime earnings. In Latin America and the Caribbean, the toll is even more concrete: the direct costs of crime and violence alone equal 3.44% of regional GDP every year — a figure equivalent to 78% of everything the region spends on public education, and nearly 1% in lost human capital. The real question, then, isn't whether prevention is worth investing in, but how much we're actually paying by not doing it.
To prevent violence and invest in effective solutions, we first need to understand that this violence rarely begins in the classroom. It takes shape across several environments at once — in homes where physical punishment still passes for discipline, in neighborhoods where the illegal economy offers belonging and income before any institution does, in digital spaces where bullying continues long after the school day ends, and on the routes children walk every day, where organized crime sometimes waits to recruit them.
The ecological model of violence, adopted by the United Nations (UN) and World Health Organization (WHO), explains violence not as an isolated act but as the result of a complex interaction among individual, relational, community, and societal factors. The violence that shows up in the classroom almost never starts there.
There's one institution that reaches nearly every child, every day, during the years when the brain and social-emotional skills are still taking shape — the window the evidence points to as the most decisive one for intervention. That institution is school. It's not the only place where violence happens, but it is the best place to get ahead of it.
Governments tend to address violence only after it has already escalated, relying on police, prosecutors, and justice systems. Intervening in schools makes it possible to step in much earlier, while behaviors are still forming and easier to shape. No other public service has the reach of an education system with near-universal primary coverage, or the sustained contact that comes from seeing the same child every day for years. That daily contact is what makes it possible to teach kids how to resolve conflict without violence, spot those at risk early, and connect families with other support services.
- First Window: Childhood. It's precisely during school-age years that the two windows with the highest returns on skills investment open up. Cunha and Heckman showed that the returns to investing in a person's development decline with age — and not in a straight line. A dollar invested in early childhood structurally outperforms that same dollar invested at any other point in life, because skills beget skills. A child who learns to self-regulate at age 6 learns more in primary school and faces a lower risk of becoming involved in violence later on.
- Second Window: Adolescence. More recently, research has shown that adolescence is the second window of opportunity — particularly for building the social-emotional skills that are key to prevention. The teenage brain shows significant plasticity, and interventions at this age can shift those skills with economic returns of between $5 and $10 for every dollar invested. That window doesn't stay open forever, and every year of inaction makes it more expensive to change a young person's trajectory.
For decades, the leading theory held that education's main channel for reducing violence and future involvement in crime was human capital formation: more schooling opens up better job prospects and raises the opportunity cost of committing a crime.
But after a detailed review of robust causal evidence on the relationship between different education policies and crime, we found that while human capital accumulation — years of schooling and its quality — is an important channel, it's far from the only one that matters. Out of an exhaustive literature review covering 493 studies on the impact of education on crime and violence, we selected 78 that met a robust evidence standards. Years of schooling emerged as an explanatory mechanism in 33% of these studies, nearly matched by what the literature refers to as “incapacitation”—the protective effect of young people spending more time in structured, supervised educational settings and therefore having less exposure to situations where violence may occur— at 36%. But the development of social-emotional skills stood out as the most effective tool for preventing violence, cited in 37% of studies as the reason education interventions succeed in reducing violence.
That pattern is even sharper among extracurricular programs with behavioral components, where 78% of studies point to social-emotional skills as the main mechanism. These interventions can change impulsive decision-making patterns and reduce automatic reactions to provocation."
What protects a young person isn't just a calculation about future earnings — it's having learned to self-regulate and pause to think before reacting. In fact, when Heckman, Pinto, and Savelyev broke down the effects of the Perry Preschool program, they attributed roughly two-thirds of its benefits to improvements in non-cognitive skills like self-regulation, planning, and persistence — and showed that this is also the margin that's most malleable and fades the least over time.
The evidence for what works, program by program, confirms that these interventions are most effective when tailored to age.
- In early childhood, Perry Preschool delivered returns that various estimates place between $7 and $17 for every dollar invested, with most of that return coming from savings tied to reduced crime in adulthood. Closer to home, at that same stage, the Irie Classroom Toolbox was tested through a cluster-randomized trial across 76 preschools in Jamaica (229 teachers, 865 children) and effectively reduced physical and psychological violence by teachers against children, with benefits still holding a year later.
- In primary school, the Good School Toolkit, developed by Raising Voices — a whole-school approach tested experimentally in Uganda — reduced violence by teachers against students. Finland's antibullying program KiVa reduced bullying and peer victimization in a randomized trial spanning more than 8,000 students across 78 Finnish schools.
- In secondary school, during adolescence, Becoming a Man cut arrests for violent crime by 45% to 50% across two randomized evaluations in Chicago, using cognitive behavioral therapy delivered during the school day. And in the Caribbean — although outside the classroom — the IDB's evaluation of Project REASON in Trinidad and Tobago found that a community-based violence-interruption model reduced violent crime in the target area by 45.1% relative to a comparison area, and cut gunshot-wound hospital admissions at the nearest hospital by about 38.7%. Unlike the previous examples, this isn't a school program but a community one — and it makes clear that the level of violence in the surrounding environment can either amplify or limit how well a school-based intervention performs.
School Can't Solve This Alone. Coordinating the Response is Essential
Violence against children takes root at several levels at once — at home, in the neighborhood, and on the way to school. School alone can't contain a risk that largely originates outside its walls, which is why prevention has to work at those same levels.
The INSPIRE framework, developed by the WHO together with UNICEF and other partners, organizes this logic into seven evidence-based strategies — among them, law enforcement, support for parents and caregivers, household economic strengthening, and life-skills education. Its core premise is that these strategies work better combined than in isolation, because no single sector can deliver the full package on its own.
The point isn't to place the whole solution on schools' shoulders, but to recognize their unique position. No other public service sees the same child every single day. That contact lets schools identify risks early on and connect families with the appropriate social protection, health, and justice institutions. Schools don't replace that response; they make sure it arrives in time.
Making that work requires the education sector to partner closely with security and justice, social protection, gender and diversity, health, and local governments — sharing goals, information, and responsibilities. That's the multisectoral approach the IDB is promoting across the region.
Violence doesn't disappear simply because of more security, more laws, or tougher sentences. It can also be prevented much earlier — when a child learns to resolve conflict without violence, when a school catches a risk in time, or when a community works together to protect its students.
That's the focus of the Schools that Care series, in which the IDB brings together the latest evidence and global and regional experience with the questions governments need to ask to put it into practice. The series shows how education can become one of the most cost-effective social investments a country can make to prevent violence and expand opportunity for the next generation.
We invite you to follow the series on the IDB Education Blog to explore other contributions and more evidence on how education can help prevent violence.