Restorative justice in schools is the proposition that when a child harms another — fights, steals, insults, disrupts — the appropriate response is not suspension but a structured conversation among the people affected, aimed at understanding the harm, repairing it, and reintegrating the child rather than excluding them. The concept traces to Maori and other Indigenous practices, was articulated for Western audiences by Howard Zehr in his 1990 book Changing Lenses, and entered American schools in earnest in the 2000s as districts began to confront the data on suspension: that Black students were suspended at roughly three times the rate of white students for comparable offenses, that suspension predicted dropout and arrest more reliably than it predicted improved behavior, and that the post-Columbine fortification of schools had produced a generation of children processed through metal detectors and zero-tolerance codes without becoming demonstrably safer.
The collective question restorative justice asks is unsettling. If suspension is, in effect, the school exiling the child to whoever or whatever is at home — and if home is often the unstable site that produced the school behavior — then suspension is the institution that knows the child best handing the problem to the institutions that know them least. It is the opposite of connection. The Connect Law, which holds that wellbeing is built in relationship, predicts that exclusion as a primary disciplinary tool will produce children who feel less attached to school, less invested in its norms, and more available to whatever community will accept them, which is often not the one the school would choose. Restorative justice is the institutional attempt to keep the relationship alive through the rupture rather than ending it.
The evidence is mixed and instructive. The RAND Corporation's 2018 evaluation of restorative practices in Pittsburgh Public Schools, conducted by Catherine Augustine and colleagues, randomly assigned twenty-two schools to receive training in the SaferSanerSchools program developed by the International Institute for Restorative Practices. Suspensions dropped substantially in treatment schools, the racial gap in suspensions narrowed, and teachers reported improved school climate. But academic outcomes did not improve, and in middle schools, math scores declined modestly. A 2019 evaluation of Oakland Unified's restorative justice implementation by Trevor Fronius and colleagues at WestEd found similar suspension reductions and similar academic ambiguity. The pattern across studies is consistent: when implemented well, restorative practices reduce exclusionary discipline and improve climate; they do not, by themselves, raise test scores; and poorly implemented programs produce the worst of both worlds — relaxed discipline without the relational infrastructure that was supposed to replace it.
The Humility Law sits behind the academic finding. Schools that adopted restorative justice as a slogan, sent staff to a two-day training, and announced the end of suspensions discovered what should have been obvious: replacing a punitive system with a relational system requires building the relational system first. Circles, conferences, and affective statements are not techniques you bolt onto a school that has spent decades teaching children that adults are agents of consequence rather than agents of repair. The actual work is years of staff development, schedule changes that permit unhurried conversations, a counselor-to-student ratio that allows follow-through, and a principal who refuses to use suspension as the easy release valve when the harder work is slow. Districts that funded only the training and not the infrastructure got the worst results.
Howard Zehr's original framing — that crime is a violation of people and relationships, not primarily a violation of rules — translates uncomfortably into a school context where the rules exist for reasons. A child who hits another child has violated both. A purely restorative response that addresses only the relational rupture and ignores the rule may produce a victim who feels their harm was minimized and a school community that no longer trusts the discipline system. The most credible restorative practitioners, including Dominique Morgan and others working in carceral and school contexts, are careful about this: restoration is not absolution, accountability is part of the process, and consequences may still follow. The shift is in the order of operations and the inclusion of the affected parties, not in the abandonment of all consequence.
The Revise Law applies. The first wave of restorative justice in American schools, roughly 2010 to 2018, produced the data above. The honest second wave is now examining what worked, what didn't, and what was oversold. Suspension reductions are real and matter. Climate improvements are real and matter. The academic gap is real and requires answer. The implementation infrastructure is non-negotiable. A school that wants to do this seriously should plan for five years of phased adoption, not five weeks of professional development. A district that wants to do this seriously should accept that the metric of success is not the suspension rate alone but the durability of relationships across years.