Friendship and bullying prevention

The anti-bullying industry emerged in earnest following high-profile school violence events in the 1990s and 2000s, drawing on earlier Norwegian research by Dan Olweus. By the 2010s, most American states had anti-bullying legislation, most school districts had anti-bullying policies, and the market for anti-bullying curricula, assemblies, and training programs had become substantial. The outcome data is discouraging. A 2013 meta-analysis by Ferguson, San Miguel, Kilburn, and Sanchez examined anti-bullying programs across 57 studies and found effect sizes close to zero for most interventions. A Cochrane Review (Vreeman and Carroll, 2007) found that social skills and curriculum-based interventions produced modest effects at best and that most commonly used programs produced no significant change in bullying prevalence. This is not an argument against intervention; it is an argument about which interventions work and why. Programs that target individual behavior without addressing the social ecology that produces that behavior are treating a systemic problem with an individualist solution. The results predict themselves.

Bullying is not primarily a dyadic phenomenon — one bad actor targeting one victim. It is a social-ecological event: it occurs in a social environment, is witnessed by a social audience, and is sustained or terminated by the responses of that audience. Salmivalli's participant role framework, developed in the 1990s and extensively validated, identifies the social architecture of bullying incidents: the bully, the target, assistants (who actively join), reinforcers (who laugh and encourage), outsiders (who disengage), and defenders (who intervene on behalf of the target). Research consistently finds that defenders are present in a large proportion of bullying incidents but rarely act. The key variable is not the presence or absence of defenders but whether defenders feel sufficient social security and moral license to act. Interventions that focus on the bully-target dyad while ignoring the bystander audience are addressing roughly 20 percent of the social dynamic. Effective bullying prevention must change bystander norms — what bystanders believe is socially expected of them and what they believe the social consequences of defense will be.

The single most robust predictor of reduced bullying victimization in the research literature is the presence of mutual friendships. Children with at least one genuine reciprocal friendship — not just acquaintance but mutual selection — are significantly less likely to be bullied, and when they are targeted, the bullying is less likely to persist and less likely to produce lasting psychological harm. The mechanism is multiple: friends provide direct defense in real time, friends create social visibility that makes targeting riskier for perpetrators, friends provide psychological resources (belonging, self-worth, social information) that reduce the target's vulnerability to the self-concept damage that bullying is designed to inflict. Hodges, Boivin, Vitaro, and Bukowski (1999) documented this in a longitudinal study showing that children with friends who were victimized at one time point were significantly less likely to be victimized at follow-up than children without friends who were similarly targeted. Friendship is not just socially valuable; it is physiologically and psychologically protective. Schools that fail to build conditions for genuine friendship formation are failing to deploy one of the most effective prevention mechanisms available.

The dominant cultural narrative of bullying frames the bully as a confident, socially dominant aggressor — someone who bullies from a position of social power. The research is more complicated. While some bullies do hold high social status, a significant portion are themselves socially marginalized children using aggression as a strategy for status acquisition they cannot achieve through other means. Dodge's social information processing model explains the bullying-aggression pattern as a product of hostile attribution bias — the tendency to interpret ambiguous social cues as hostile — which is itself a product of early relational adversity. Children who have not experienced reliable, responsive relationships tend to read social ambiguity as threat and respond with preemptive aggression. This does not excuse the harm bullying causes, but it reframes the intervention target. Punishing a child whose aggression is driven by relational deficit and hostile attributional patterns does not address the deficit or correct the patterns. Building genuine relationship competency — through instruction, through reliable adult relationships, through peer connection — addresses both. The bully who becomes capable of genuine friendship is the bully who no longer needs to bully.

Online bullying is not a separate phenomenon from playground bullying; it is the extension of the same social-ecological dynamics into a digital medium that removes some friction while adding some amplification. The research on cyberbullying consistently finds that it is closely correlated with in-person bullying — the same social networks, the same participant roles, the same underlying disconnection patterns. What digital media adds is persistence (screenshots last), audience scale (a humiliation can be witnessed by hundreds rather than dozens), and availability (the school day no longer provides a temporal boundary to the bullying experience). The response to cyberbullying that focuses on platform policy and monitoring without addressing the underlying social ecology is equally misplaced as the response to playground bullying that focuses on punishment without addressing the social climate. Children who are genuinely connected — who have real friendships with real relational capital — are more resilient to cyberbullying both as potential targets and as potential bystanders who might intervene. Digital citizenship programs are useful but insufficient unless they are accompanied by the relational infrastructure that makes digital courage possible.

The KiVa anti-bullying program, developed by Christina Salmivalli and colleagues at the University of Turku and implemented nationally in Finland, is the most rigorously evaluated whole-school bullying prevention program in existence. Its core innovation is the focus on bystanders rather than perpetrators. KiVa works by changing the social norms of the bystander group — making defense of targets socially expected and socially rewarded, and making passive reinforcement of bullying socially costly. It includes classroom curriculum, school rules, and parent components, but the active mechanism is bystander norm change. Randomized controlled trials in Finland found reductions in bullying victimization of approximately 20–30 percent, with the effects concentrated in schools where bystander norm change was most successfully implemented. KiVa's success is instructive: it worked not by targeting bad actors but by changing the social climate in which bullying is evaluated by witnesses. This is a friendship-infrastructure approach — it builds the relational norms that make genuine connection more likely and predatory disconnection more costly.

Restorative justice practices in schools address bullying by treating it as a relational harm requiring relational repair rather than a rule violation requiring punishment. The restorative approach brings together the people involved in a harm — including bystanders — in a structured conversation focused on understanding the impact, identifying what needs to change, and establishing agreements for the future. Research on restorative approaches to bullying in schools has found reductions in repeat offenses, improvements in school climate, and — crucially — better long-term relational outcomes between parties than punitive approaches. The mechanism is direct: restorative conversations require participants to take perspective, articulate impact, and negotiate repair. These are core friendship competencies practiced under conditions of real stakes. Punitive approaches teach children that harm produces exclusion; restorative approaches teach children that harm produces the obligation of repair. The latter is the norm that adult friendship requires.

Bullying is not randomly distributed across social categories. Children who are visibly different from peer group norms — by weight, disability, racial minority status, gender nonconformity, sexual orientation, or socioeconomic disadvantage — face higher rates of bullying. The dynamics here are not simply individual meanness; they are collective norm enforcement. Children learn from their social environment which differences are acceptable and which are to be punished. Schools that reflect and reinforce social hierarchies of difference through their own structure — ability tracking, dress codes that police gender expression, discipline disparities by race — create the social context in which targeted bullying of difference becomes logical. Anti-bullying programs that do not address the institutional transmission of these norms address the expression while leaving the source intact. Friendship across difference — genuine, voluntary, reciprocal connection between children who occupy different social categories — is both the indicator and the mechanism of a social climate resistant to targeted bullying. Schools that produce cross-group friendships are schools that are reducing the conditions under which targeted bullying makes social sense.

The relationship between childhood bullying experience and adult social behavior is not limited to the perpetrators. The dynamics of workplace bullying, intimate partner coercion, political intimidation, and institutional abuse of power follow similar social-ecological patterns to school bullying: perpetrators seeking status through dominance, targets isolated by social disconnection, bystanders paralyzed by risk calculations, and the whole system sustained by social norms that make challenge costly. Schools that address bullying as a childhood problem without connecting it to the relational patterns it produces in adults are treating a developmental episode rather than a social formation. The child who learns that bystander passivity is the safe response carries that lesson into adulthood. The child who learns that genuine friendship includes defense of the vulnerable carries that lesson too. Bullying prevention is not just about safer childhoods; it is about what kind of adults schools are producing for the collective life that follows.

Teachers are the most powerful bystanders in schools. When a teacher witnesses peer exclusion, mocking, or low-level cruelty and does nothing, they are functioning as a reinforcer in Salmivalli's framework — their inaction communicates that the behavior is within acceptable norms. When teachers intervene — not necessarily dramatically, but consistently, naming what they observed and making clear it is not acceptable — they shift bystander norms for the entire class. Research on teacher bystander behavior in bullying situations finds significant variation: some teachers intervene in a large proportion of witnessed incidents, others in almost none. The variation is not primarily explained by awareness of the problem (teachers generally know bullying occurs) but by beliefs about whether intervention is their role and whether it is effective. Teacher training that addresses bystander efficacy — building teachers' confidence and skill in addressing relational harm as it occurs — is a high-leverage intervention. The teacher who treats every witnessed moment of cruelty as an opportunity for relational norm-setting is doing more sustained bullying prevention than any assembly program.

One underappreciated mechanism by which friendship prevents bullying is social visibility. Children with friends are seen. Their presence is noticed; their absence is registered. When something happens to them, someone knows and is motivated to respond. Isolated children — children without genuine friendship connections — are socially invisible in a specific way: the harm done to them registers in no one's relational accounting. Sociometric mapping of school peer networks consistently finds that isolated children (those receiving few or no peer nominations as friends) are clustered at the highest risk for sustained victimization. This is not a character deficit of the isolated children; it is a structural condition. Schools that track academic performance at the individual level but have no systematic awareness of who is relationally isolated — who has no friends, who is at the edge of every social network — are blind to one of the most significant risk factors in their student population. Systematic sociometric monitoring, used diagnostically and non-punitively, would give schools the information they need to target relational support before harm escalates.

What does a school climate resistant to bullying actually look like, and how is it built? The research converges on several features: high adult-to-student relational investment (teachers know students by name, notice changes, express genuine care); strong belonging norms (students feel they matter to the school community, not only to their immediate friend group); clear and consistently enforced community standards about how people treat each other; visible cross-group friendship across social categories; conflict handled restoratively and publicly enough that students see repair modeled; and bystander culture in which defending the vulnerable is socially expected and socially rewarded. None of these features can be produced by a single program or a designated anti-bullying coordinator. They are properties of the entire relational ecology of the school, produced by the cumulative effect of hundreds of daily decisions about how adults treat each other, how conflict is handled, who is included, what is noticed and named. The school that has these features is, in effect, a school that has taken Law 3 seriously — treating connection as the ground of collective life rather than as an optional enrichment. The anti-bullying outcome is a side effect of the friendship infrastructure.

---

1. Olweus, Dan. Bullying at School: What We Know and What We Can Do. Oxford: Blackwell, 1993.

2. Salmivalli, Christina. "Bullying and the Peer Group: A Review." Aggression and Violent Behavior 15, no. 2 (2010): 112–120.

3. Ferguson, Christopher J., Claudia San Miguel, John C. Kilburn Jr., and Patricia Sanchez. "The Effectiveness of School-Based Anti-Bullying Programs: A Meta-Analytic Review." Criminal Justice Review 32, no. 4 (2007): 401–414.

4. Vreeman, Rachel C., and Aaron E. Carroll. "A Systematic Review of School-Based Interventions to Prevent Bullying." Archives of Pediatrics and Adolescent Medicine 161, no. 1 (2007): 78–88.

5. Hodges, Ernest V.E., Michel Boivin, Frank Vitaro, and William M. Bukowski. "The Power of Friendship: Protection Against an Escalating Cycle of Peer Victimization." Developmental Psychology 35, no. 1 (1999): 94–101.

6. Dodge, Kenneth A., and John D. Coie. "Social-Information-Processing Factors in Reactive and Proactive Aggression in Children's Peer Groups." Journal of Personality and Social Psychology 53, no. 6 (1987): 1146–1158.

7. Kärnä, Antti, Marinus Voeten, Todd D. Little, Elisa Poskiparta, Anne Kaljonen, and Christina Salmivalli. "A Large-Scale Evaluation of the KiVa Antibullying Program: Grades 4–6." Child Development 82, no. 1 (2011): 311–330.

8. Thorsborne, Margaret, and Peta Blood. Implementing Restorative Practices in Schools: A Practical Guide to Transforming School Communities. London: Jessica Kingsley Publishers, 2013.

9. Espelage, Dorothy L., and Susan M. Swearer. Bullying in American Schools: A Social-Ecological Perspective on Prevention and Intervention. Mahwah, NJ: Lawrence Erlbaum Associates, 2004.

10. Polanin, Joshua R., Dorothy L. Espelage, and Therese D. Pigott. "A Meta-Analysis of School-Based Bullying Prevention Programs' Effects on Bystander Intervention Behavior." School Psychology Review 41, no. 1 (2012): 47–65.

11. Holt, Melissa K., Glori Kaufman Kantor, and David Finkelhor. "Parent/Child Concordance About Bullying Involvement and Family Characteristics Related to Bullying and Peer Victimization." Journal of School Violence 8, no. 1 (2009): 42–63.

12. Ladd, Gary W., and Becky J. Kochenderfer-Ladd. "Identifying Victims of Peer Aggression from Early to Middle Childhood: Analysis of Cross-Informant Data for Concordance, Prevalence of Victimization, Characteristics of Identified Victims, and Investigations of Victimization." Psychological Assessment 14, no. 1 (2002): 74–96.