The promise of ‘restorative justice’ starts to falter under rigorous research
Studies in Pittsburgh and Maine show difficulty of implementing trendy alternative to traditional discipline
In 2014, the Obama Administration jolted the education world with a report detailing unfair and racist school discipline practices across the country. Sixteen percent of all black students were being suspended, more than three times the rate of white students. Even preschoolers were being suspended at alarming rates. Other scholars produced research showing that the kind of zero-tolerance discipline then in vogue was hurting students’ long-term academic prospects and feeding the school-to-prison pipeline.
Desperate for alternatives, many educators rapidly embraced something called “restorative justice,” a set of practices that originated in the criminal justice sector in the 1970s to help victims and offenders reconcile or come to some sort of agreement on reparations. In education settings, there are various ways to implement this alternative to suspensions and expulsions but it generally involves having kids sit in a circle and discuss their conflicts with the help of an adult mediator.
Early research seemed promising. Developers of “restorative justice” programs described how “talking it out” yielded benefits, such as a reduction in bullying and fighting, lower suspension rates and fewer missed days of school. Sometimes, early adopters even claimed that student achievement improved. But the studies tended to be small, and they tracked only what happened to students who participated in the program without comparing them to similar students who didn’t participate. No studies could prove that the restorative justice programs were causing any of the positive changes that the advocates had noticed.
By late 2016 even proponents openly worried that schools had moved too quickly. Samuel Song, co-director the National Network of Restorative School Researchers and an associate professor at the University of Nevada, Las Vegas, penned a damning article entitled, “The Cart Before the Horse,” in which he described the “void” in the research literature and a lack of understanding of the basics, such as what the day-to-day role of school psychologists should be.
At last, more sophisticated research has been commissioned, and the results are starting to trickle in. For proponents of restorative justice, the first two studies are not especially promising with both failing to show clear benefits for these non-punitive approaches to student discipline. Academic achievement fell for some students who were exposed to restorative justice compared to students at schools who were disciplined as usual. Implementation problems were common.
Both studies were conducted by RAND Corporation, a research firm, which randomly assigned schools in the city of Pittsburgh and the state of Maine to try restorative justice practices. The Pittsburgh study was commissioned by National Institute for Justice as part of its Comprehensive School Safety Initiative. (The institute is a research, development and evaluation agency of the U.S. Department of Justice .) The Maine study was supported by the National Institute of Child Health and Human Development, a unit of the U.S. Department of Health and Human Services.
In the Pittsburgh study, published in December 2018, suspension rates fell at the 22 schools that tried restorative justice. But suspension rates also fell at 22 comparison schools in the city that didn’t adopt restorative justice, echoing the dramatic decline in suspensions across the nation. During the second year of the Pittsburgh experiment, 12.6 percent of the kids at the restorative justice schools had been suspended during the 2016-17 school year, compared with 14.6 percent of the students at traditional discipline schools. (Teachers in the treatment schools were trained in restorative justice techniques and encouraged to talk with students instead of punishing them but suspensions were still a discipline option.)
The academic performance of middle schoolers actually worsened at schools that tried restorative justice. Math test scores deteriorated for black students in particular.
The number of student arrests was similar at both treatment and control schools. That suggests the restorative justice experiment wasn’t doing much to alleviate the school-to-prison pipeline.
In surveys, teachers at the schools that tried restorative justice said that their school climate improved. But students reported that teachers struggled more to manage classroom behavior. I wondered if disruptive behavior in the classroom might have detracted from learning time, or perhaps even worthwhile and productive restorative justice conversations ate away at precious instructional minutes. Either way, it could potentially explain why some kids’ performance suffered.
The Maine study, published online in March 2019 by the Journal of Youth and Adolescence, found no difference in school climate between middle schools that tried restorative justice and those that didn’t. It didn’t look at suspension rates or academic outcomes. But the fact that bullying and other school climate measures didn’t budge is another sign that restorative justice programs aren’t a slam dunk solution to addressing student misbehavior.
The biggest insight from the Maine study was how hard it is for schools to implement restorative justice even after days of teacher training, monthly consultations and visits by coaches. Students’ survey answers revealed that they didn’t experience very much restorative justice in their day-to-day classrooms even after two years of effort. Restorative justice also requires a high degree of student buy-in. Students cannot be forced to talk about their grievances face-to-face with their classroom enemies. It’s a voluntary process and not every kid wants to talk.
“Conceptually and theoretically, restorative practices should work,” said Francis Huang, an associate professor at the University of Missouri’s College of Education who is conducting another study of restorative justice in schools in Brooklyn. “But it’s harder.”
“It’s not like a suspension where you send a kid out, and you’re done with it,” he said. “With restorative practices, it’s a process. You actually have to sit down, get them in a circle and talk it through. It’s tough. That’s why you have coaches and coordinators. You need some training. You can’t just react.”
Meanwhile, as the Maine study was underway, elements of restorative justice had entered the zeitgeist. Talking to kids instead of punishing them right away was becoming more popular among teachers. Even in the control group of schools, which were supposed to be meting out discipline as usual, teachers were borrowing bits and pieces of restorative practices without any formal training and trying them out in classrooms.
That left researchers with quite a strange situation: kids in the treatment group hadn’t experienced as much restorative justice as they imagined while kids in the control group experienced more than expected. Perhaps, in the case of the Maine study, it’s not correct to say that restorative justice ideas don’t work but that restorative justice programs aren’t particularly effective or necessary.
One bright side for restorative justice was that the more that Maine students reported that they personally experienced elements of restorative justice, such as discussing problems in circles, the more that student felt connected to his or her peers and the less cyber-bullying he or she experienced. This is probably obvious, but there seems to be some benefit from facilitating discussions between students at school and not rushing to punish every infraction.
Huang says it’s far “too early” to say whether restorative justice works or doesn’t work in education. “All we have are these two studies now,” he said. In addition to Huang’s Brooklyn study, another study is underway by Center for Court Innovation, also in Brooklyn.
But researchers are already confronting some of the same problems in Brooklyn that the Maine researchers found: control group schools are doing some elements on their own while treatment schools can struggle to adopt the full soup-to-nuts program. That will make it hard to show scientific evidence for restorative justice. It’s like comparing the effectiveness of flossing between a person who agreed to floss every day but doesn’t and a person who didn’t commit to flossing but is doing it anyway.
Not only is restorative justice a challenge for schools to implement, it’s also a tricky thing for researchers to study because too many students and teachers make decisions that are beyond a researcher’s control.
This story about restorative justice in schools was written by Jill Barshay and produced by The Hechinger Report, a nonprofit, independent news organization focused on inequality and innovation in education. Sign up for the Hechinger newsletter.