The school-to-prison pipeline is a structural feature of US education

The claim

Zero-tolerance school discipline policies, combined with the post-Columbine expansion of school resource officers (SROs) and campus police, have created a documented pathway from school suspension to dropout to incarceration that falls disproportionately on Black, Hispanic, Native American, and disabled students. The claim is structural: the pipeline is not the accidental byproduct of well-intentioned policies applied unevenly, but an emergent property of institutional design decisions that systematically criminalize normal adolescent behavior in under-resourced schools serving low-income communities of color. Proponents of this structural view argue that the pattern is too consistent, too nationally uniform, and too robust to control variables to be explained by individual conduct differences. The architecture of punishment — who writes referrals, for what offenses, at what rates, with what downstream consequences — encodes racial and economic disparity at every step.

The mechanism

The pipeline operates through a sequence of institutional decisions, each of which has a measurable disparity and a compounding effect on the next stage. The mechanism has four links.

Link 1: Suspension and expulsion. The Gun-Free Schools Act of 1994 mandated one-year expulsions for weapons possession and triggered a national wave of zero-tolerance policies that expanded far beyond the original statutory scope. By the early 2000s, districts across the country had adopted zero-tolerance rules for drug and alcohol possession, threats, and — critically — subjective categories like “defiance” and “disrespect.” The DOE Civil Rights Data Collection documents that these policies are not applied uniformly. In 2017-18, Black students received out-of-school suspensions at 3.6 times the rate of white students. Black girls were suspended at higher rates than students of any other gender and race except Black boys. Students with disabilities were more than twice as likely to be suspended as their non-disabled peers.

Link 2: Dropout amplification. Suspension removes students from instructional time, disrupts academic progress, and signals to students that they do not belong in school. Each suspension event is associated with substantially elevated dropout probability. Losen and Gillespie’s (2012) analysis of CRDC data estimated that 3.5 million students lost approximately 18 million days of instruction in a single school year due to out-of-school suspensions. Fabelo et al.’s (2011) Texas longitudinal study — tracking approximately one million middle school students from 6th grade through graduation — found that students who had been suspended or expelled were three times more likely to have contact with the juvenile justice system in the following year, even controlling for prior juvenile contact, poverty, and disability.

Link 3: Criminal justice entry. The expansion of SROs into schools since the late 1990s brought police discretion into disciplinary contexts previously handled by administrators. Conduct that would formerly have resulted in detention or in-school suspension — hallway fights, disruptive behavior, truancy — became eligible for arrest and criminal charge. The 2015-16 CRDC found that approximately 1.7 million students were enrolled in schools that had a police officer but no counselor. Schools with SROs referred students to law enforcement at substantially higher rates, including for infractions that carry no mandatory reporting requirement.

Link 4: Incarceration feedback. Dropout is the single strongest predictor of adult incarceration that the education system produces. Losen et al. document that one in ten young male dropouts is incarcerated on any given day — a rate that dwarfs the incarceration rate for high school graduates. The mechanism is partially direct (criminal records limit employment and housing, increasing reoffending risk) and partially indirect (dropping out removes young people from structured time, adult supervision, and future economic opportunity). The racial gap in dropout rates, driven partly by the discipline disparity in link 1, propagates into the racial gap in incarceration in link 4.

The evidence

DOE Civil Rights Data Collection: consistent, national, large-scale

The CRDC is a mandatory census of all US public schools, administered every two years since 1968, with expanded scope since 2011. Its discipline data are not a sample — they are the universe of public school discipline events, reported at the district and school level. The racial disparities it documents are therefore not subject to sampling error. The 2017-18 CRDC found that Black students represented 15% of enrollment but 38% of students receiving one or more out-of-school suspensions. This pattern holds across regions, school types, and poverty levels. High-poverty majority-white districts suspend at lower rates than low-poverty majority-Black districts — a finding that cannot be explained by poverty alone and that directly implicates race as an independent determinant of disciplinary exposure.

The disability intersection is equally stark. Students with disabilities — particularly those with emotional and behavioral disorders — are removed from school at rates that expose them to the pipeline with even higher frequency than race alone predicts. Disabled Black students face a compounding disparity: the 2017-18 CRDC reports that Black students with disabilities were removed from school at the highest rate of any demographic group analyzed.

Fabelo et al. (2011): Texas million-student longitudinal cohort

The Council of State Governments Justice Center study tracked all students enrolled in Texas middle schools in the 2000-01 school year through graduation or dropout, merging school administrative records with Texas Youth Commission (juvenile justice) and Texas Department of Criminal Justice (adult prison) records. The dataset comprises approximately one million students over a seven-year follow-up period. Key findings: 60% of Texas public school students received at least one in-school or out-of-school suspension or expulsion over a six-year middle and high school period. Having been disciplinarily removed even once nearly tripled the odds of juvenile justice contact in the following year. Black students were disproportionately disciplined for discretionary infractions (those that allowed administrative judgment) but not for mandatory ones (those with required reporting thresholds), confirming that discretionary enforcement drives the racial gap, not underlying behavioral differences.

Skiba et al. (2011) and Smolkowski et al. (2016): the discretion mechanism

Russell Skiba and colleagues’ work on the referral-to-suspension pipeline consistently identifies a core pattern: the racial disparity in discipline is concentrated in subjective infraction categories — defiance, disrespect, disruption — rather than in objective categories like weapons possession or assault. Smolkowski et al.’s (2016) analysis of elementary school discipline in a Pacific Northwest sample found that Black students were 2.19 times more likely to be referred for subjective infractions than white students with similar prior behavioral records. This is the mechanism: not that Black students commit more punishable acts, but that the same acts are more likely to be noticed, interpreted as infractions, and referred to the office. Implicit bias research (Okonofua & Eberhardt, 2015) extends this finding into experimental conditions, demonstrating that teachers shown identical student behavior attributed to a Black student versus a white student recommended harsher discipline for the Black student.

Zero-tolerance origins and expansion: Gun-Free Schools Act 1994 to post-Columbine SRO surge

Zero-tolerance as a federal mandate began with the Gun-Free Schools Act of 1994 (GFSA), which conditioned Title IV federal education funding on states adopting one-year expulsion policies for firearms possession. Congress framed the law narrowly, but states and districts used it as authority to expand zero-tolerance policies to drugs, alcohol, and behavioral infractions far removed from the weapons context that motivated the original legislation. The 1999 Columbine High School shooting triggered a second, larger wave of SRO deployment and security hardening. Federal grants through the COPS in Schools program funded 6,000 additional school police officers between 1999 and 2005. The number of schools with SROs grew from roughly 10% of schools in 1975 to over 40% by 2015, with higher rates in high-poverty, majority-minority schools than in low-poverty, majority-white schools — the inverse of what a risk-based deployment model would produce.

Restorative justice as a natural experiment

The strongest evidence that the pipeline is structural rather than behavioral comes from districts that dismantled the architecture of zero tolerance and replaced it with restorative practices. Denver Public Schools provides the clearest case: between 2003 and 2012, the district progressively replaced zero-tolerance suspensions and expulsions with restorative circles, conflict mediation, and behavioral support. Expulsions fell by 67%; suspensions fell substantially; school-based arrests dropped. Importantly, there was no documented increase in school violence or reported safety incidents — the core claim of zero-tolerance advocates, that harsh discipline is necessary for safety, was not supported. Oakland Unified School District produced similar findings over a comparable period. These are not ideal randomized experiments, but they are the best available evidence on what happens when the structural mechanism is altered: outcomes change in the direction the structural theory predicts.

Cross-national comparison

The United States is an extreme outlier in youth incarceration among peer nations. The US incarcerates young people at approximately 5 to 7 times the rate of Canada, the UK, Germany, Finland, and the Netherlands. This gap is not primarily attributable to higher rates of youth crime: comparative victimization surveys do not support a corresponding behavioral difference of that magnitude. What differs structurally is the architecture of response to youth misconduct. Finland has no equivalent of zero-tolerance expulsion: the Education Act requires that schools make “every effort” to keep students enrolled and specifies that disciplinary exclusion cannot exceed three months and must be paired with a plan for reintegration. Germany’s school system does not route disciplinary problems through law enforcement. Canada’s province-level codes prohibit indefinite expulsion and require re-enrollment plans. The Netherlands prioritizes attendance support over punitive exclusion. In each comparator country, the school-to-incarceration pathway that US data document does not exist at comparable scale, and the pathway’s absence is explicable by the structural difference in disciplinary architecture.

Who benefits

The expansion of school security infrastructure has created a substantial commercial market. Security technology vendors — including companies selling metal detectors, surveillance systems, and SRO equipment — have a direct financial interest in zero-tolerance disciplinary frameworks that require physical security infrastructure. The Student Safety and Security market was estimated at over $3.5 billion annually by the mid-2010s. Industry associations have lobbied against restorative justice initiatives in multiple states on the grounds that they reduce demand for security hardware.

Private prison and prison services companies (historically GEO Group and CoreCivic) operate facilities that house adjudicated youth and benefit from higher juvenile justice system referral rates. Both companies have documented lobbying activity on sentencing and criminal justice policy, and both operate juvenile facilities under government contracts where occupancy rates affect revenue. The corporate interest in high incarceration rates maps directly onto the pipeline’s output.

Political coalitions that frame crime as an individual moral failing — including think tanks funded by Koch-network donors such as the Heritage Foundation and certain state Policy Network affiliates — benefit from pipeline rhetoric that locates the causes of youth incarceration in student behavior rather than in disciplinary policy. This framing resists redistribution of school resources toward counselors, social workers, and restorative programs, and supports punitive criminal justice policies that generate continued demand for incarceration infrastructure.

The geographic concentration of school discipline in under-resourced schools also benefits affluent suburban districts indirectly: the pipeline functions as a sorting mechanism that removes behaviorally challenging students from schools serving high-income families, whose property values are partly capitalized on school quality and perceived safety.

The counter

The most serious counterargument is the school safety case. Zero-tolerance proponents argue that disruptive students impose significant costs on the educational experience of their classmates, that early decisive intervention prevents escalation to more serious misconduct, and that SROs provide genuine safety benefits including faster response to genuine emergencies. This view has some empirical support: order and safety in school are prerequisites for effective teaching, and high-violence school environments depress achievement for all students, not just those disciplined. The strongest form of the structural critique does not deny this — it argues that zero tolerance achieves order through exclusion at a population cost that exceeds its individual benefit, and that restorative alternatives achieve comparable order without the exclusionary disparity.

There is also a genuine debate about the direction of causation between discipline and dropout. Students who are heading toward dropout may accumulate more suspensions because their disengagement manifests as disciplinary incidents, rather than suspensions causing dropout. The Texas longitudinal study attempts to address this through controls and sequencing, but the identification is imperfect by the standards of a randomized experiment. The restorative justice evidence, which changes the institutional mechanism and observes downstream outcomes, provides the cleanest test available of the causal hypothesis, but those studies have methodological limitations — districts that adopt restorative practices may simultaneously adopt other supportive interventions, making it difficult to isolate the effect.

The racial discipline gap is genuinely contested in its proximate cause. Some researchers (e.g., Morgan et al., 2015, controversially) argued that Black students with behavioral disorders may be under-identified for special education relative to white students with similar profiles, and that once disability status is correctly assigned, some of the raw racial gap shrinks. This claim was widely criticized on methodological and empirical grounds (Sullivan & Bal, 2013; Skiba et al., 2016), and the bulk of the literature does not support it as a primary explanation. But it represents a legitimate area of ongoing methodological debate.

These counterarguments inform better policy design rather than undermining the structural diagnosis. The evidence that discretionary enforcement drives the racial gap more than objective behavioral differences, that restorative alternatives reduce suspensions without increasing violence, and that cross-national peer nations maintain safe schools without zero-tolerance architecture, is collectively sufficient to support the structural claim.

References

Fabelo, T., Thompson, M. D., Plotkin, M., Carmichael, D., Marchbanks, M. P., & Booth, E. A. (2011). Breaking schools’ rules: A statewide study of how school discipline relates to students’ success and juvenile justice involvement. Council of State Governments Justice Center. https://csgjusticecenter.org/publications/breaking-schools-rules/

Losen, D. J., & Gillespie, J. (2012). Opportunities suspended: The disparate impact of disciplinary exclusion from school. UCLA Civil Rights Project. https://civilrightsproject.ucla.edu/resources/projects/center-for-civil-rights-remedies/school-to-prison-folder/federal-reports/upcoming-ccrr-research

Skiba, R. J., Horner, R. H., Chung, C. G., Rausch, M. K., May, S. L., & Tobin, T. (2011). Race is not neutral: A national investigation of African American and Latino disproportionality in school discipline. School Psychology Review, 40(1), 85–107. https://doi.org/10.1080/02796015.2011.12087730

Smolkowski, K., Girvan, E. J., McIntosh, K., Nese, R. N. T., & Horner, R. H. (2016). Vulnerable decision points for disproportionate office discipline referrals: Comparisons of discipline for African American and White elementary school students. Behavioral Disorders, 41(4), 178–195. https://doi.org/10.17988/0198-7429-41.4.178

Okonofua, J. A., & Eberhardt, J. L. (2015). Two strikes: Race and the disciplining of young students. Psychological Science, 26(5), 617–624. https://doi.org/10.1177/0956797615570365

U.S. Department of Education Office for Civil Rights. (2020). Civil Rights Data Collection: School Climate and Safety (2017-18 school year). U.S. Department of Education. https://ocrdata.ed.gov/

Nance, J. P. (2016). Students, police, and the school-to-prison pipeline. Washington University Law Review, 93(4), 919–987.

Hirschfield, P. J. (2008). Preparing for prison? The criminalization of school discipline in the USA. Theoretical Criminology, 12(1), 79–101. https://doi.org/10.1177/1362480607085795

Morgan, P. L., Farkas, G., Hillemeier, M. M., Mattison, R., Maczuga, S., Li, H., & Cook, M. (2015). Minorities are disproportionately underrepresented in special education: Longitudinal evidence across five disability conditions. Educational Researcher, 44(5), 278–292. https://doi.org/10.3102/0013189X15591157

Fronius, T., Darling-Hammond, S., Persson, H., Guckenburg, S., Hurley, N., & Petrosino, A. (2019). Restorative justice in U.S. schools: An updated research review. WestEd Justice and Prevention Research Center. https://www.wested.org/resources/restorative-justice-in-u-s-schools-updated-research-review/