Social and emotional learning: what the latest evidence says about preventing violence  

Social and emotional learning (SEL) programmes help children develop skills such as emotional regulation, self-awareness and relationship-building. Increasingly used in schools, alternative provision and youth services, these approaches are often seen as an important part of violence prevention and behaviour improvement strategies. We recently reviewed the latest evidence to understand whether universal and targeted SEL programmes reduce violence and improve outcomes for children and young people. 

At the Youth Endowment Fund, we regularly update our Toolkit to ensure that it reflects the most up-to-date available evidence on approaches that affect children and young people’s risk of becoming involved in violence. We’ve just updated an old toolkit strand that combined a wide range of ‘social skills’ programmes under one label, to share more nuanced policy and practice recommendations with you. This strand, originally published in 2021 as social skills training, is now Social and Emotional Learning (SEL). 

Evidence from this toolkit strand helped inform YEF’s Education Practice Guidance   for school, college and alternative provision leaders and Education Systems Guidance providing recommendations for education policymakers to prevent children’s involvement in violence in England and Wales.  

Why? Seven out of 10 children who admit to perpetrating violence say that they did so because they were provoked. We know that supporting children to develop their social skills, think before they act, understand others’ perspectives and manage aggression can reduce their risk of being involved in violence.
 
“YEF Education, Children and Violence – Practice Guidance”

What is Social and Emotional Learning? 

In short, SEL helps children understand and manage their emotions, build positive relationships, and think before they act – skills that can lower their risk of violence, antisocial behaviour and offending. It can be delivered universally, to all children, or as a targeted offer for those who need more support. 

Why we revisited the strand  

As a Toolkit team, one of our responsibilities is making sure the evidence is not only accurate, but useful for decision-makers. Over time, we recognised that the social skills training strand in the toolkit was no longer capturing the full breadth of approaches being used in practice, nor the increasing nuance emerging from the research evidence. In particular, the strand grouped together both universal and targeted social and emotional learning under a single label, despite clear differences between these approaches. Universal SEL programmes are typically embedded within the education curriculum and delivered to all children in primary and secondary schools, while targeted programmes focus on children and young people at increased risk of, or already experiencing, social and emotional difficulties, including involvement in crime, violence or antisocial behaviour. 

As the evidence base developed, feedback from our audiences increasingly asked whether these two approaches show different impacts. This highlighted an important gap in our previous single-strand structure, so we commissioned two new evidence reviews, explicitly separating universal and targeted SEL. This allowed us to reflect the growing evidence more accurately and provide clearer, more relevant guidance for the different policymakers and decision-makers across education, children’s services, youth justice and the youth sector. 

What we found when we revisited the evidence? 

The evidence found universal social and emotional learning to have a moderate impact on violence among children and young people. Findings suggest these reduce violence by around 25%, an estimate we have very high confidence in. The evidence also points to a larger reduction in crime and offending, of around 45%, though we have low confidence in this figure, as it is based on just four studies. Beyond violence and offending, universal programmes were linked to better mental health, less bullying and fewer behavioural difficulties.  

Review of evidence on effectiveness of targeted social and emotional learning programmes found a high impact on violence, with programmes reducing violence by around 43%. However, we have low confidence in this impact rating, as it is based on just nine studies, all conducted in the US. A reduction in crime and offending of around 31% was also found, but with very low confidence, based on only three studies. Targeted programmes were also linked to improvements in school engagement, reduced behavioural difficulties, and better relationships. 

How can this be implemented well? 

The implementation findings also echoed what we’ve heard from practitioners. Leaders often tell us that social and emotional learning is not a single intervention but something that needs to be embedded consistently over time. The evidence supports that view: high-quality implementation, staff training and sustained delivery appear just as important as the programme itself. 

Both reviews pointed to a shared foundation for good delivery: programmes work best when they are delivered in line with the SAFE principles: 

S = Sequenced in a coordinated and connected way to support skill development, breaking down more complicated skills into smaller steps 

A =Active, to enable children to practice and rehearse new skills 

F = Focused, with specific sessions allocated to social and emotional learning and, 

E = Explicit, clearly defining and targeting skills, and explicitly teaching social and emotional skills to children. 

Investing in structured training, coaching and ongoing support for whoever delivers the programme also came through consistently across both strands. 

Beyond these shared principles, good implementation looks slightly different depending on how a programme is delivered. Where social and emotional learning is delivered to whole classes, protecting sufficient, realistic time in the timetable and monitoring delivery closely helps keep it consistent across classes and year groups; teaching self-regulation skills before more complex relationship-building skills, and designing sessions to be inclusive and accessible to every pupil, also supports stronger outcomes. Where delivery is more individualised – working with smaller groups or one-to-one – tailoring sessions to the specific needs of the children involved matters most, alongside combining structured skill instruction with approaches that help build strong, trusting relationships between the child and whoever is delivering the programme, and adapting activities to a child’s stage of development. 

Whatever the confidence rating, the underlying message is simple: every child benefits from being taught social and emotional skills, so a universal offer should be in place for all of them. On top of that, children showing behavioural difficulties, facing high absence, suspension or exclusion, or already in contact with the youth justice system need a tailored offer too. Even with lower confidence in the targeted evidence base right now, both remain good value investments in preventing violence. 

Every child benefits from being taught social and emotional skills, so a universal offer should be in place for all of them. On top of that, children showing behavioural difficulties, facing high absence, suspension or exclusion, or already in contact with the youth justice system need a tailored offer too. Even with lower confidence in the targeted evidence base right now, both remain good value investments in preventing violence. 

To see the full summaries, visit the Social and Emotional Learning-Universal and Social and Emotional Learning-Targeted pages on the Toolkit.