More than two decades ago, in 2003, Johanna Wald and Daniel Losen argued that suspensions and expulsions were funneling children—disproportionately black and Latino boys and students with disabilities—out of schools and into the justice system. They called it the “school-to-prison pipeline.” The phrase stuck, changing how schools discipline children.
Associating discipline with racism and diminished life prospects for minority youth, school districts increasingly responded by reducing suspensions and replacing them with mental health-oriented approaches like “restorative justice,” social-emotional learning, and “trauma-informed” practices. In Los Angeles, for example, the school board voted in 2013 to end suspensions for “willful defiance” after years of advocacy from CADRE, a Los Angeles parent organization, and the ACLU of Southern California. Similar reforms spread nationwide as educators and advocates viewed exclusionary discipline as the problem.
Finally, a reason to check your email.
Sign up for our free newsletter today.
The Obama administration soon reinforced the premise. In 2014, the Departments of Education and Justice sent every school district a “Dear Colleague” letter warning that they could face federal civil rights investigations if their suspension rates differed by race—even when discipline policies were race-neutral and applied consistently. Education Secretary Arne Duncan argued that the disparities stemmed from differences in adult training and policy, not in children’s behavior. “It is adult behavior that needs to change,” he said. The federal government thus embraced the idea that disparities in discipline reflect problems with the punishment itself, rather than differences in behavior.
This view overlooks a basic behavioral principle: that consequences shape what children do next. Some consequences make a behavior less likely—that’s punishment. Others make it more likely—what B. F. Skinner called reinforcement. It doesn’t matter whether the reward is getting something good or merely escaping something bad.
Much classroom misbehavior is motivated by the desire to escape. The student is trying to get out of something he finds unpleasant, whether it’s work he can’t do, a subject he’s failing, or an audience he’d rather not face. For that student, a suspension acts as a reward—it gets him exactly what he wants, reinforcing the behavior. The harm is what the days outside of the classroom bring––disengagement and the growing sense that school is not where he belongs. Other misbehavior is motivated by desire for attention; in those cases, pulling the student aside for a one-on-one discussion is unlikely to be an effective deterrent.
The restorative-justice practices that have replaced suspension, such as mediation circles or discussion with a teacher, distribute the same kinds of unintentional rewards. A child escaping a math worksheet doesn’t care whether he gets a suspension slip or is placed in a restorative circle; either way, the worksheet is gone. Likewise, a child who seeks adult attention doesn’t care whether it’s punitive or therapeutic; either way, he got it.
Schools often––unintentionally––deepen the reward of attention through the use of mental-health programs. A standard response to repeated misbehavior is a visit to the school counselor or school psychologist. There the misbehaving student receives a caring adult’s undivided, one-on-one time. Students who need adult attention should get it, but not as the payoff for acting out.
Escape works the same way in reverse. The trip to the counselor’s office ends the math lesson just as surely as a suspension does. For the student acting out to avoid work that he cannot do, an effective response must ensure that the work does not go away. The assignment should wait for him––in the office, after school—so that the misbehavior no longer gets him out of it. And because the skill he lacks is usually the real problem, the lasting fix is help with the work itself, which will give the student more reason to want to stay in the classroom.
Research supports this. A meta-analysis of 69 studies found that changing behavior depended less on whether an intervention is called punitive or restorative than on whether it changes the reinforcement maintaining the behavior. A response reduces a bad behavior only if it breaks the connection between the acting out and its payoff—in other words, make sure that the student no longer gets the desired reward. Do that consistently and the behavior changes. Skip that step, and the behavior won’t change.
Discipline tied to the function of behavior is the only approach that keeps a student in school engaged—and engagement is what keeps him out of the justice system.
The anti-punishment reforms of recent decades made schools less responsive to student behavior and its motivations, and non-responsive schools became the most reliable feeders to the justice system. New York City went further than any district in the nation, beginning restorative justice programs in 2015, expanding them citywide in 2019, and spending an estimated $97 million over the decade. When critics warned that pulling back on punishment would make schools less safe, then-Schools Chancellor Richard Carranza dismissed the idea: “I say that’s a false narrative.”
Not really. Despite the investment in non-disciplinary approaches, violence in schools has not declined even when disciplinary acts are reported less often. New York City’s district schools issued 35,963 disciplinary removals and suspensions during the 2024–25 school year. Black students accounted for 38 percent and Hispanic students 42 percent of these actions. Black and Hispanic students are also chronically absent at around 42 percent, above the citywide average. That same year, 163 students received suspensions lasting longer than four weeks. A child kept out of school that long is not being taught that behavior has consequences. He is being taught that school is somewhere he no longer goes. That’s a recipe for disengagement—and it is ultimately disengagement, not suspensions, that paves the road to the justice system.
Even the study most often cited as proof of the “school-to-prison pipeline”––which followed Charlotte–Mecklenburg students reassigned to stricter middle schools––found that those assigned to these stricter schools were more likely to be arrested as adults but also completed less schooling. The true damage was the lost education. Longitudinal research has found that student disengagement, measured through attendance, academic performance, and school behavior, predicts not only dropping out but also later delinquency, official offending, and problem substance use.
None of this means that the answer is to stop suspending kids. A student who poses a danger to others, or whose behavior can be deterred by the right consequences, sometimes must be removed. What matters most is that the response matches what’s driving the behavior, discourages it––and keeps the student engaged in school. The surest way to keep a student out of prison has never been a particular punishment, or the absence of one. It is a school that teaches rules, expectations, and how to belong.