Education is among the most thoroughly documented social determinants of criminal involvement — a relationship that operates through multiple theoretically distinct pathways and that has been examined with sufficient methodological rigor to support confident causal claims rather than merely descriptive associations. The negative relationship between educational attainment and criminal involvement appears across every measurement instrument, every offense category, every demographic group, and every national context that researchers have examined, making it one of criminology’s most consistent empirical regularities. Understanding the mechanisms through which education reduces crime, the specific educational investments most effective for crime prevention, and the policy implications of the education-crime relationship for both education and criminal justice agencies is essential to designing policies capable of reducing crime through upstream investment rather than downstream enforcement.
Criminology and Public Policy engages the education-crime relationship not only because education is among the strongest predictors of individual criminal involvement but because educational institutions — schools, colleges, and the systems that govern them — are among the most consequential determinants of the social conditions that produce crime at the community level. The quality of schools in a neighborhood determines the social capital, peer networks, and legitimate opportunity structures available to its young people. Suspension, expulsion, and school discipline practices determine which students are excluded from educational institutions and the protective influences they provide. And the school-to-prison pipeline — the well-documented pathway through which exclusionary discipline practices, school-based law enforcement presence, and the prosecution of school disciplinary infractions funnel students into criminal justice contact — represents an institutional mechanism through which schools actively contribute to the criminal trajectories of the most disadvantaged students rather than protecting them from those trajectories.
Introduction
The theoretical frameworks that explain the education-crime relationship fall into several distinct accounts that have different implications for which educational policies most effectively reduce crime. The human capital account, drawing on labor economics, proposes that education reduces crime by increasing legitimate earnings — through the skills, credentials, and social capital that education provides — and thereby raising the opportunity cost of criminal activity. On this account, the most crime-preventive educational investments are those that most effectively increase earnings, including vocational training and credential programs that connect directly to labor market demand. The social control account, drawing on Hirschi’s bond theory, proposes that school attachment and commitment provide the prosocial bonds that inhibit criminal behavior independent of any labor market effect; on this account, dropout prevention and school engagement programs that build attachment to educational institutions reduce crime by strengthening informal social control even among students who will not complete high school. The developmental account, drawing on life-course criminology, proposes that schools’ effects on cognitive development, self-regulation, and social competence in early childhood and adolescence shape the personality characteristics and behavioral repertoires that determine criminal propensity across the life course, independent of any specific economic or social control mechanism.
These distinct theoretical accounts generate different predictions about the timing, targeting, and design of educational crime prevention investments. The human capital account suggests that the returns to crime prevention are highest for educational investments that maximize labor market earnings, concentrated in the secondary and post-secondary years when credentials are conferred. The social control account suggests that dropout prevention and school engagement are the critical mechanisms, concentrated in the middle school and high school years when dropout risk is highest and when detachment from school most strongly predicts delinquency onset. The developmental account suggests that the most durable crime prevention effects come from early childhood educational investments that shape cognitive and social-emotional development before criminal trajectories are established. The three accounts are not mutually exclusive — all three mechanisms may operate simultaneously — but their different policy implications mean that the mix of investments in early childhood, K-12, and post-secondary education matters for the crime prevention return on educational investment.
Research designs that can establish causal relationships between educational investments and crime are relatively abundant in this domain compared to many criminal justice research areas, primarily because compulsory schooling laws, funding formulas, and program eligibility thresholds create natural experiments in educational exposure that researchers have exploited with quasi-experimental methods. The evidence generated through these designs provides substantially stronger causal evidence than the observational correlations that characterize much of the education-crime literature, supporting policy conclusions with a degree of confidence that mere association cannot provide.
The Education-Crime Relationship: Causal Evidence
Compulsory Schooling Laws and Crime
The most direct causal evidence on the education-crime relationship comes from studies exploiting variation in compulsory schooling laws — state-level variations in the age at which students are legally required to remain in school — as natural experiments in educational attainment. If education reduces crime causally rather than merely correlating with it, then individuals whose educational attainment was increased by higher compulsory schooling requirements should show reduced criminal involvement compared to otherwise similar individuals who were not bound by those requirements. Research exploiting cross-state variation in compulsory schooling age thresholds has found consistent evidence that additional years of required schooling reduce adult crime, with effects concentrated in property crime and with magnitudes consistent with the human capital account — additional schooling raises earnings, reduces property crime motivation (Lochner & Moretti, 2004).
Lochner and Moretti’s foundational research estimated that a one-year increase in average educational attainment would reduce incarceration rates by approximately 4% for white men and approximately 5% for Black men, translating to substantial reductions in incarceration costs that substantially exceed the educational investment required to produce them (Lochner & Moretti, 2004). Subsequent research has extended and refined these estimates using different compulsory schooling instruments, different crime outcomes, and different demographic groups, generally finding positive effects of education on crime reduction that are consistent across methodological approaches even as the specific effect magnitudes vary.
Research on school dropout and crime has complemented the compulsory schooling law studies by documenting the timing and magnitude of the dropout-to-crime pathway. Dropout substantially elevates criminal risk, with the elevation concentrated in the months immediately following leaving school — consistent with the social control account that school’s crime-suppressing effect operates through current attachment and engagement rather than only through its eventual effect on adult earnings. The immediate timing of the crime increase after dropout suggests that dropout prevention programs that keep at-risk students in school longer reduce crime through the social control pathway without necessarily improving labor market outcomes — an important practical implication for the design of dropout prevention as crime prevention.
School Quality and Neighborhood Crime
The quality of schools serving a neighborhood matters not only for the individual educational outcomes of the students they serve but for the neighborhood-level social organization that school quality affects. Research documenting the relationship between school quality — measured by teacher quality, school resources, and educational achievement — and neighborhood crime rates has found that higher-quality schools reduce neighborhood crime beyond the individual-level effects on school attenders, through mechanisms that include reducing concentrated educational disadvantage and building the institutional capacity for community social organization that high-quality schools support. Research by Patrick Sharkey and colleagues has found that community-level investments in educational quality reduce violent crime in the neighborhoods they serve at effect sizes that compare favorably to policing investments of equivalent cost.
Research by David Kirk and Robert Sampson documenting the relationship between neighborhood collective efficacy and school engagement established that school quality effects on crime operate partly through the community social organization that high-quality schools help create — schools serve not only the students they directly educate but as anchors of community social organization that build the collective efficacy that suppresses crime independently of individual educational attainment (Kirk & Sampson, 2013). This community-level school quality effect has implications for school closure decisions — particularly the closure of under-enrolled schools in declining urban communities, which has been associated in research with increased crime rates in affected neighborhoods, as the social organization anchor that the school provided is removed alongside its educational function. The Center for Research on Education Outcomes at Stanford University and the Urban Institute‘s education research program have both documented community effects of school closures that extend beyond the direct educational outcomes of affected students to the broader neighborhood social organization consequences that criminological research predicts.
Teacher Quality and Classroom Environment
Research on teacher quality and its effects on student outcomes — including criminal involvement — has established that teacher effectiveness, measured through value-added assessment of student learning growth, is among the strongest within-school determinants of student outcomes. Raj Chetty, John Friedman, and Jonah Rockoff’s landmark study found that replacing a teacher in the bottom 5% of the effectiveness distribution with an average teacher would reduce the probability of teenage pregnancy, increase college attendance rates, and increase lifetime earnings substantially — effects consistent with the human capital mechanism through which education reduces crime (Chetty et al., 2014). While the study did not directly measure criminal outcomes, the mediating outcomes it documented — educational attainment, employment, and economic stability — are among the strongest proxies for criminal involvement in the developmental criminology literature, and the magnitudes of the effects suggest that teacher quality differences translate into meaningful differences in downstream criminal justice outcomes.
The racial and socioeconomic distribution of teacher quality — more experienced, higher-performing teachers are concentrated in higher-income schools while lower-income schools disproportionately serve students with less experienced and less effective teachers — amplifies the educational inequality that predicts criminal involvement. Research documenting that effective teachers disproportionately transfer away from low-income schools early in their careers has motivated school funding reform advocates to argue for compensation structures and working condition improvements that would retain effective teachers in the schools where their crime prevention benefits would be largest. The Learning Policy Institute and the Albert Shanker Institute have both produced research documenting the maldistribution of teacher quality and its consequences for the educational inequality that criminological research consistently identifies as a driver of criminal involvement.
Table 1. Education Policy Interventions and Crime: Evidence Summary
| Intervention | Mechanism | Study Design | Crime Outcome | Effect Direction | Key Finding |
|---|---|---|---|---|---|
| Compulsory schooling age increases | Human capital; social control | Quasi-experimental (law variation) | Adult incarceration; property crime | Positive | Each additional year of required schooling reduces incarceration probability by ~4%; property crime effects larger than violent crime |
| High-quality early childhood education (Perry; Abecedarian) | Developmental (cognitive, self-regulation) | RCT with long-term follow-up | Adult arrest; incarceration | Positive | Large long-term crime reductions; returns concentrated in highest-risk children |
| Dropout prevention / graduation incentives | Social control; human capital | Quasi-experimental | Juvenile delinquency; adult crime | Positive | Keeping students in school reduces crime immediately; effects through both current social control and future earnings |
| School-based cognitive-behavioral programs (PATHS, LST) | Developmental (social-emotional skills) | Multiple RCTs | Aggression; delinquency | Positive | Consistent modest effects; strongest for high-risk students with skilled implementation |
| GED / adult literacy programs for incarcerated individuals | Human capital | Quasi-experimental; RCT (some evaluations) | Recidivism | Positive | Moderate recidivism reductions; job placement and wages improve when combined with vocational training |
| School discipline reform (restorative practices) | Social control; school engagement | Pre-post; comparison group | Suspension rates; delinquency | Positive | Suspension reductions achieved; crime reduction evidence less definitive |
| College in prison programs | Human capital; cognitive | Quasi-experimental; observational | Recidivism | Strongly positive | College education in prison produces some of the largest recidivism reductions in the correctional literature |
The School-to-Prison Pipeline
Exclusionary Discipline and Criminal Justice Contact
The school-to-prison pipeline — the institutional pathway through which school discipline practices, law enforcement presence in schools, and the prosecution of school-based infractions channel students into criminal justice contact — is one of the most consequential intersections of education and criminal justice policy, with effects concentrated among Black students, students with disabilities, and students in low-income communities. Research documenting the pipeline has consistently found that school suspension and expulsion substantially elevate subsequent criminal justice contact, that schools in low-income and majority-minority communities use exclusionary discipline at rates far higher than comparable schools in more affluent communities, and that the presence of school resource officers — law enforcement stationed in schools — is associated with increased arrests of students for behaviors that were previously handled internally through school discipline processes.
Exclusionary discipline — out-of-school suspension and expulsion — elevates criminal justice contact through multiple mechanisms simultaneously. Suspended students are unsupervised during school hours, increasing their exposure to criminal opportunity and criminal peers. Suspension signals to students that they are not valued members of the school community, weakening the school attachment that social control theory identifies as the primary mechanism through which school protects against delinquency. Repeated suspension increases dropout risk, removing students from the protective influence of school entirely. And the racial disparities in suspension — Black students are suspended at approximately three times the rate of white students for similar conduct — contribute to the racial disparities in criminal justice contact that begin in school and compound throughout the justice system.
The American Psychological Association‘s zero-tolerance policy review concluded that zero-tolerance discipline approaches — mandatory suspensions and expulsions for specified conduct regardless of circumstances — do not improve school safety and are associated with worse student outcomes including increased dropout and criminal involvement (APA, 2008). Research by John Hemphill and colleagues using longitudinal data from Australian schools found that suspension substantially elevated self-reported delinquency in the years following suspension among suspended students relative to comparable non-suspended students, providing quasi-experimental evidence of the causal effect of suspension on crime rather than only the observed correlation (Hemphill et al., 2006).
The scale of racial disparities in school discipline is both well-documented and persistent despite decades of awareness. The Civil Rights Data Collection maintained by the U.S. Department of Education’s Office for Civil Rights documents that Black students are suspended at 3.5 times the rate of white students, that students with disabilities are suspended at twice the rate of students without disabilities, and that these disparities hold across school district income levels, suggesting they reflect racial bias in discipline decision-making rather than only differences in student behavior. Research by Brea Perry and Edward Morris using large administrative datasets from a single large school district found that Black boys were suspended for subjective conduct violations — defiance, disrespect, disruption — at higher rates than white boys for the same conduct categories, with the gap attributable not to differences in conduct frequency but to differences in teacher perception and response — findings consistent with implicit bias research demonstrating that identical conduct is perceived differently depending on the race of the student exhibiting it.
Restorative Practices as an Alternative
Restorative practices in schools — approaches that address disciplinary infractions through structured processes of accountability, repair, and community reconciliation rather than exclusion — have been adopted by school districts seeking to reduce suspension rates without sacrificing school safety or clear accountability for harmful behavior. Research on restorative practices in schools has found consistent reductions in suspension and expulsion rates following adoption of restorative approaches, with the most rigorous studies — including a large-scale randomized trial in Pittsburgh public schools — finding reductions in office discipline referrals and suspensions without equivalent increases in crime or school safety incidents (Augustine et al., 2018). The crime prevention implications of restorative discipline reform are direct: reducing suspension reduces exposure to the criminal opportunity and peer networks that suspension creates, maintains school attachment that suppresses delinquency, and reduces dropout risk that would further elevate criminal involvement.
The Pittsburgh Public Schools evaluation of restorative practices — a school-randomized trial assigning schools to receive restorative practices training and support or continue with standard discipline practices — represents one of the most rigorous evaluations of an education policy intervention with criminal justice implications, finding significant reductions in suspensions and improvements in school climate in treatment schools relative to control schools. The study’s findings, replicated in observational research across multiple school districts that have implemented restorative approaches, have contributed to growing adoption of restorative practices among American school districts seeking to reduce racial disparities in discipline and the school-to-prison pipeline they feed.
Post-Secondary Education and Reentry
College-in-Prison Programs
The evidence on post-secondary education The research base on college-in-prison programs has been substantially strengthened by Pell Grant restoration, which has enabled rigorous evaluations of program effectiveness that were difficult to conduct when programs were small and self-selected. The RAND Corporation’s systematic review and meta-analysis of correctional education programs remains the most comprehensive synthesis available, finding that incarcerated individuals who participate in academic or vocational education programs are 43% less likely to return to prison than comparable non-participants — an effect size that translates to a return on investment of $5 for every $1 invested when reincarceration cost savings are calculated (Davis et al., 2014). The effect is consistent across education levels, program types, and incarceration settings, suggesting that the crime reduction benefit of education in prison is robust rather than dependent on specific program features that might not generalize to real-world implementation. The restoration of Pell Grant eligibility in 2023 is projected to expand college-in-prison program enrollment by tens of thousands of students annually, creating both an opportunity for continued rigorous evaluation and a meaningful expansion of the educational crime prevention investment that the evidence supports.
for incarcerated individuals is among the most consistently positive in the correctional rehabilitation literature — stronger in effect size and consistency than many other rehabilitative programming approaches. Studies using quasi-experimental designs to compare incarcerated individuals who participated in college programs with comparable individuals who did not have consistently found recidivism reductions of 40–60% in some estimates, with the most rigorous studies continuing to find substantial effects even after controlling for selection into participation (Davis et al., 2014). The mechanisms are multiple: college education increases legitimate earnings capacity, reducing economic motivation for crime after release; it builds cognitive skills, problem-solving capacity, and future orientation that reduce impulsive criminal decision-making; and it provides educational credentials that open legitimate employment pathways previously unavailable.
The Pell Grant restriction that eliminated federal financial aid eligibility for incarcerated students from 1994 to 2023 — enacted as part of the Violent Crime Control and Law Enforcement Act despite the absence of evidence that college programs reduced criminal involvement (their effectiveness had not yet been established) — was reversed by the Second Chance Pell Experimental Sites Initiative and subsequently by the restoration of Pell Grant eligibility in 2023 through the FAFSA Simplification Act. The evidence base that accumulated during the Pell restriction period, through state and philanthropically funded college programs that operated without federal support, documents the crime prevention return on post-secondary educational investment with sufficient strength to make the policy case for the restored eligibility on both equity and public safety grounds.
Vocational Education and Employment Pathways
Vocational and career-technical education — programs that provide work-relevant skills and credentials alongside or instead of academic coursework — represent the educational investment most directly connected to the human capital mechanism through which education reduces property crime. Research on vocational education’s effects on crime has generally found positive effects, with the most rigorous evidence coming from European countries where the transition from vocational tracks to employment is more institutionalized than in the American context. American research on vocational education’s crime consequences has been more limited, partly because the American secondary school system has been historically ambivalent about vocational tracks in ways that have limited both program quality and research opportunities.
The Job Corps program — the largest federally funded vocational education and training program in the United States, serving approximately 100,000 low-income youth annually — has been evaluated in the most rigorous randomized trial of any employment and training program for at-risk youth, finding that participation reduced arrest rates and improved earnings for older youth (17–20 years old) but showed weaker effects for younger participants (Schochet et al., 2008). The crime reduction effects for older youth were larger in the first two years post-program than in longer follow-ups, consistent with the human capital account that employment stability, rather than training alone, sustains the crime reduction benefits of vocational investment. The Job Corps evaluation remains one of the most methodologically rigorous evaluations of a major federal anti-poverty program and provides the strongest experimental evidence available on vocational training’s effects on criminal involvement.
Chronic Absenteeism and School Disconnection
Research on chronic absenteeism — missing 10% or more of school days — has documented that persistent school disconnection elevates delinquency and criminal involvement through the same mechanisms that dropout elevates criminal risk, but beginning earlier and among students who are nominally enrolled in school. The Everyone Graduates Center at Johns Hopkins University has documented that chronic absenteeism is among the strongest early warning indicators of eventual dropout and that it concentrates in the same low-income and predominantly minority communities where crime rates are highest — creating a spatial overlap between school disconnection and crime concentration that is not coincidental but reflects common causes in concentrated disadvantage.
Interventions targeting chronic absenteeism — including school-based mentoring, attendance-focused case management, positive behavioral interventions and supports (PBIS), and family engagement programs — have been evaluated with increasing rigor in recent years, with findings generally supporting the effectiveness of intensive, relationship-based interventions for the highest-risk chronically absent students. The crime prevention case for chronic absenteeism intervention parallels the dropout prevention case: keeping students physically present in school maintains the social control, peer monitoring, and institutional connection that suppresses delinquency, even before the question of academic learning and eventual educational attainment is addressed. The Every Student Succeeds Act‘s requirement that states track chronic absenteeism as an accountability indicator has elevated the policy attention to absenteeism as an early warning system and created incentives for school systems to develop effective attendance intervention programs.
Adult Education and Reentry for Criminal Justice Populations
The education needs of individuals with criminal justice involvement extend well beyond the incarcerated population to encompass the far larger population under community supervision — the approximately 3.7 million individuals on probation and 840,000 on parole — and the millions of formerly incarcerated individuals in the community whose criminal records create labor market barriers that education and credential attainment can partially address. Research on educational participation among probationers and parolees consistently finds that educational enrollment is associated with reduced recidivism, though the causal direction is complicated by the possibility that individuals with the motivation and stability to pursue education would reduce recidivism regardless of whether they enrolled. Natural experiments exploiting variation in community supervision conditions — including education requirements as a supervision condition — provide somewhat stronger causal evidence, generally finding positive effects of required educational participation on both employment outcomes and recidivism.
The Workforce Innovation and Opportunity Act (WIOA) — the primary federal legislation governing adult education and workforce development — explicitly includes justice-involved individuals as a priority population for Title II Adult Education and Title I workforce development services, creating a statutory framework for the integration of education and employment services into reentry planning that the evidence base clearly supports. The challenge is implementation: the institutional separation of criminal justice agencies — which control supervision conditions and case management — from workforce development agencies — which provide educational and employment services — creates coordination barriers that the statutory framework does not automatically overcome. Building the cross-agency coordination mechanisms that effective reentry education requires is among the most important implementation challenges in the evidence-based reentry field.
Research on the Harlem Children’s Zone (HCZ) — Geoffrey Canada’s ambitious cradle-to-college initiative that provides comprehensive educational and social services to all children in a 97-block area of Central Harlem — offers evidence on the crime prevention benefits of integrated, place-based educational investment that goes beyond individual program evaluations to examine what comprehensive educational ecosystem investment produces. Studies by Will Dobbie and Roland Fryer found that HCZ Promise Academy charter schools produced large improvements in educational achievement, particularly for Black students, and that these improvements were associated with downstream reductions in teen pregnancy, criminal involvement, and poverty in early adulthood — effects that randomized lottery-based analyses attributed to the educational intervention rather than to selection of motivated families into the program (Dobbie & Fryer, 2011). The HCZ model’s emphasis on both educational quality and comprehensive wraparound services — addressing the health, housing, and family support needs that affect children’s ability to benefit from educational investment — provides a template for place-based educational crime prevention that individual program evaluations understate by measuring only the effects of isolated program elements.
Conclusion
The education-crime relationship is among the most thoroughly documented causal relationships in applied criminological research, supported by natural experiments, longitudinal cohort studies, and randomized trials that together establish that education reduces crime through multiple mechanisms operating across the full developmental span. High-quality early childhood education, effective K-12 schooling and dropout prevention, restorative discipline practices that keep at-risk students connected to school, and post-secondary education for incarcerated individuals all reduce criminal involvement with effect sizes that compare favorably to equivalently funded criminal justice interventions.
The policy implication is not that education alone can solve crime — the education-crime relationship is one of multiple pathways through which social conditions produce criminal involvement, and educational investment without complementary investment in housing, employment, and health will not address the full causal architecture of crime risk. It is that educational investment should be understood and evaluated as crime prevention investment, not only as human capital investment, and that criminal justice agencies, school systems, and education policy advocates should develop the collaborative frameworks needed to design and fund educational interventions that serve both the educational and crime prevention purposes that the evidence demonstrates they can achieve.
The school-to-prison pipeline represents the inverse of this argument: when educational institutions contribute to criminal involvement through exclusionary discipline practices, criminalized school environments, and the effective expulsion of the most disadvantaged students from the protective influence of schooling, they undermine rather than advance crime prevention. Reforming school discipline practices, removing school-based criminal justice contacts that funnel students into the justice system for conduct better handled through educational responses, and investing in school environments that maintain engagement of the highest-risk students are all evidence-based crime prevention strategies whose implementation requires leadership from school systems rather than criminal justice agencies.
References
- American Psychological Association Zero Tolerance Task Force. (2008). Are zero tolerance policies effective in the schools? American Psychologist, 63(9), 852–862. https://doi.org/10.1037/0003-066X.63.9.852
- Augustine, C. H., Engberg, J., Grimm, G. E., Lee, E., Wang, E. L., Christianson, K., & Joseph, A. A. (2018). Can restorative practices improve school climate and curb suspensions? RAND Corporation. https://doi.org/10.7249/RR2101
- Davis, L. M., Bozick, R., Steele, J. L., Saunders, J., & Miles, J. N. V. (2014). Evaluating the effectiveness of correctional education: A meta-analysis of programs that provide education to incarcerated adults. RAND Corporation. https://doi.org/10.7249/RR564
- Hemphill, S. A., Toumbourou, J. W., Herrenkohl, T. I., McMorris, B. J., & Catalano, R. F. (2006). The effect of school suspensions and arrests on subsequent adolescent antisocial behavior in Australia and the United States. Journal of Adolescent Health, 39(5), 736–744. https://doi.org/10.1016/j.jadohealth.2006.05.010
- Lochner, L., & Moretti, E. (2004). The effect of education on crime: Evidence from prison inmates, arrests, and self-reports. American Economic Review, 94(1), 155–189. https://doi.org/10.1257/000282804322970751
- Welsh, B. C., & Farrington, D. P. (2008). Effects of improved street lighting on crime. Campbell Systematic Reviews, 4(1), 1–51. https://doi.org/10.4073/csr.2008.13