Conventional wisdom among the prevention fraternity is to the effect that the US leads the way in the development and rigorous evaluation of effective programs – except in the case of bullying and victimization reduction, which tends to be associated more closely with Norway and the reputation of Dan Olweus. 

The Olweus program, considered “model” by Blueprints for Violence Prevention, and therefore satisfying the highest evidential standards, targets bullies, victims and bystanders in the classroom, school and community.

If you want to reduce victimization in your school, numerous evaluations indicate that it is a pretty safe bet. 

But a systematic review for the Campbell Collaboration by David Farrington and his colleague, Maria Ttofi, at the Institute of Criminology at Cambridge University suggest that the Olweus approach is far from being the only effective possibility. 

They identified over 600 reports from 30 countries of attempts to prevent or reduce bullying in school. By applying strict experimental evaluation criteria they whittled the list down to 44, from 16 countries as diverse as Czechoslovakia, New Zealand, South Africa, England, the US and Germany.  

The good news, according to Farrington and Ttofi, is that another 18 programs – some of them adaptations of the Olweus model – demonstrated positive effects. Taken as a whole, those qualifying interventions succeeded in reducing rates of bullying and victimization by 20% in the schools where they were introduced.  

Alarmingly, on the other side of the coin, another 17 of the 44 were as certainly ineffective, and one was associated with increased rates of bullying and victimization. (The effectiveness of the other seven was broadly positive but not as clearly demonstrated.)

[See also Bullying programs don't work? Shoot the messenger]

The message is clear: with the right intervention, it is possible to reduce bullying in schools, but there is a high risk of equivocal or harmful results. It has to be done right 

Farrington and Ttofi went beyond this general conclusion to try to distinguish between effective and ineffective program elements.

They categorized the programs included in their review according to their active ingredients, which ranged widely between whole-school anti-bullying policies, individualized work with bullies and victims, school tribunals or “bully courts” and virtual reality computer games to raise awareness. 

The most positive effects were observed for those programs that adopted a whole-school approach, firm discipline and classroom management practices, improved playground supervision and the provision of information about bullying and victimization for parents were associated with beneficial impact.

It added up to a spectrum of activity broadly in line with the components of the Olweus model. 

Programs that put their energies into peer mediation and mentoring were most frequently associated with deteriorating rates of bullying and victimization.

Here, Farrington and Ttofi conclude that it is not enough to rely on a single program, no matter how firmly it might be based on effective components. They say it is vital for service design to make use of variations and adaptation that improve knowledge about the active – and counterproductive – aspects of interventions in order to extend the range of what is available to schools. 

See: Farrington D and Ttofi M (2009) School-based programs to reduce bullying and victimization. Campbell Systematic Reviews.