The School-to-Prison Pipeline:
How the Education System Feeds the Carceral System

Before 1994, school discipline was handled within schools — by teachers, principals, and counselors. Suspensions and expulsions existed, but the decision to involve law enforcement in a school disciplinary matter was rare, discretionary, and reserved for serious criminal conduct. The concept of arresting a child for a school rule violation had no meaningful policy infrastructure.

The Gun-Free Schools Act of 1994 — signed by President Clinton as part of the broader “tough on crime” legislative architecture of that era — mandated that any school receiving federal funding must expel for one year any student found with a weapon on school grounds. This was the founding statute of zero-tolerance policy. The logic was explicit: remove discretion from administrators, standardize punishment, make consequences automatic regardless of context, intent, or circumstances.

Within two years, states and school districts had extended zero-tolerance policies far beyond weapons — to drugs, alcohol, fighting, and eventually to broad subjective categories like “threatening behavior,” “disruption,” and California’s “willful defiance.” Zero tolerance for a weapon became zero tolerance for a butter knife in a lunchbox. Zero tolerance for drugs became zero tolerance for an aspirin. The policy sold as a security measure became the mechanism for removing children from school at scale.

The racial data emerged immediately. Black students, who were 17% of the student population, were 32% of students suspended and 30% of students expelled in the first decade of zero-tolerance enforcement — and the gap was not explained by higher rates of actual misconduct. Studies that controlled for the type of infraction found that Black students received harsher punishments than white students for identical behavior. The mechanism was not explicit racial animus. It was subjective judgment, applied in a racially unequal institutional environment, producing racially unequal outcomes.

School Resource Officers — sworn police officers assigned to school buildings — existed in some districts before 1994, but the Columbine shooting in April 1999 produced a federal funding surge that placed officers in schools at scale for the first time. The 1999 COPS in Schools grant program provided $68 million to place officers in schools. By 2003, approximately 14,000 officers were stationed in American schools. By 2019, the number exceeded 50,000.

The presence of armed police officers inside school buildings had a specific and measurable effect: it converted school disciplinary matters into criminal matters. Before SROs, a student who threw something at a teacher received a detention or a trip to the principal’s office. After SROs, the same student could be handcuffed, arrested, and charged with assault — creating a juvenile criminal record at age twelve.

The data on SROs and racial disparity is unambiguous. A 2019 study in Educational Researcher found that schools with SROs reported 6 times higher arrest rates and 2.5 times higher referral-to-law-enforcement rates — with no corresponding improvement in school safety. The benefit the SRO program was designed to provide was not demonstrated. The harm it produced was.

Black students represent 31% of students referred to law enforcement and 36% of students arrested at school in a population that is 15% Black. The most common charges are disorderly conduct, disturbing the peace, and simple assault — statutory language applied to behavior that was, in previous decades, handled with a phone call to a parent.

In 2007, the Children’s Defense Fund published America’s Cradle to Prison Pipeline. The report documented a system of compounding disadvantages — poverty, inadequate healthcare, failing schools, zero-tolerance discipline, and mass incarceration — that funneled Black children from birth toward imprisonment. The key finding: a Black boy born in 2001 had a 1-in-3 lifetime probability of going to prison. For white boys the figure was 1 in 17.

The report documented the pipeline’s mechanics. Each link increased the probability of the next. Children in poverty have worse health outcomes. Children with health challenges have higher rates of learning disabilities. Children with learning disabilities are more likely to be flagged as “disruptive.” Disruptive children are more likely to be suspended. Suspended children fall behind academically. Children who fall behind are more likely to be held back a grade. Being held back one grade doubles the likelihood of dropping out. Dropping out multiplies the probability of incarceration.

The Obama administration’s response, in 2014, was the first federal policy acknowledgment that the pipeline existed. The Departments of Education and Justice issued a “Dear Colleague” letter to every school district in the country, warning that racially disparate discipline practices violated Title VI of the Civil Rights Act and making — for the first time — federal funding contingent on racial equity in school discipline.

Sources: U.S. DOE Civil Rights Data Collection; Government Accountability Office 2018; NAACP Legal Defense Fund

In December 2018, Secretary of Education Betsy DeVos rescinded the 2014 Obama-era “Dear Colleague” guidance on school discipline disparities. The stated rationale was that the guidance had created “undue pressure” on schools to reduce suspensions of Black students in ways that were not “in the best interests of students” — an argument that equity requirements had caused schools to under-discipline dangerous students in order to balance racial statistics.

This argument ignores the controlled studies showing that racial disparities persist after controlling for the type and severity of infraction. It assumes the disparity reflects differential behavior, not differential judgment. It takes the outcome of a biased system as evidence that the system is fair.

The practical effects were immediate. The Office for Civil Rights at the Department of Education — which processes Title VI complaints against school districts — deprioritized and closed dozens of pending investigations into discriminatory discipline. School districts that had reduced disciplinary disparities under federal scrutiny in some cases reversed course within two years of the rescission.

The window from 2018 to the present represents the longest period since 2000 without meaningful federal oversight of school discipline racial disparities — precisely the years in which COVID disruptions, remote learning loss, and return-to-school behavioral challenges produced documented spikes in suspension rates nationwide.

The pipeline has a demonstrated interruption: restorative justice. Restorative practices replace punitive, exclusionary discipline with facilitated processes that bring together the person who caused harm, the person harmed, and the surrounding community — to name what happened, hear its impact, and determine what repair looks like. A student who disrupted a class doesn’t get suspended. They sit in a circle with their teacher and classmates, speak and listen, and the relationship gets repaired. The behavior is addressed. The student stays in school.

The evidence is consistent and robust. In Oakland Unified School District, which implemented restorative justice at scale between 2010 and 2014, suspensions dropped 87% over four years. In Denver Public Schools, a restorative practices pilot reduced suspensions 68% and out-of-school time 46% with no increase in school violence. In every district where restorative justice was implemented with fidelity and adequate resources, suspensions fell, racial gaps narrowed, and academic outcomes improved — including for students who had previously been suspended repeatedly.

The resistance is not empirical — the data is not contested. It is political and institutional. School administrators cite training time. Some teachers’ unions have objected to the emotional labor it requires. Police associations have lobbied against it because it typically involves reducing SRO presence. In conservative-dominated state legislatures, restorative justice has been characterized as permissiveness toward dangerous students — a framing with no basis in the research record.

The pipeline has no single off-switch. It runs through federal law, state juvenile codes, district policy, and the individual judgment calls of teachers and administrators who have been trained — by institutional culture, not individual malice — to perceive Black children as more threatening than white children exhibiting identical behavior. Implicit bias studies show consistently that teachers rate identical disruptive behavior as more severe when the student is Black. This is the pipeline’s starting point.

The question is not whether the pipeline exists. The question is whether the people with the power to dismantle it — in Congress, in statehouses, in school board elections — will choose to do so when doing so requires redistributing resources away from policing and toward counseling, toward restorative practice facilitators, toward the mental health infrastructure that underfunded Black schools have never had. The evidence says what works. What is missing is the political will to fund it.