Desistance from Crime: Why People Stop Offending

4 days ago

18 min read

Author: Ukrainian Psychological Hub · Published: October 11, 2026 · Editorial Policy

Desistance from crime is the process through which a person reduces and eventually ends involvement in criminal offending. Research treats desistance as a change unfolding across time: behavior may become less frequent, less serious, or stop, while relationships, opportunities, responsibilities, self-understanding, and social conditions change. A person with a history of offending can follow a different course at different life stages. Explaining that change is central to life-course criminology, correctional psychology, reentry policy, and public safety.

The fundamental question is why people who once committed offenses come to offend less or cease offending. Studies identify several recurring correlates—aging, changes in social bonds, a shift in identity, fewer opportunities to offend, and improved access to legitimate roles—but their effects vary by person and context. No single experience, personality feature, employment intervention, or legal sanction reliably explains every individual transition. The National Institute of Justice's comprehensive desistance review places those processes within developmental, social, institutional, and correctional research.

What Is Desistance from Crime?

In criminology, desistance refers to a declining propensity for criminal involvement over the life course, ordinarily visible through reduced or discontinued offending. It can describe gradual behavioral change, extended periods without known offending, or the eventual end of a criminal career. The term applies to observed behavior and its development over time; it is not a psychiatric diagnosis, clinical score, judicial finding, or legal declaration of rehabilitation.

A foundational distinction separates termination—the point at which criminal activity stops—from desistance—the process that leads to and sustains that change. That distinction comes from Laub and Sampson's review of desistance research. Someone who has not offended for six months may be early in an enduring transition or may resume offending later. A period of nonoffending is therefore an observation; its meaning depends on the time horizon, measurement method, and individual history.

Michael Rocque's NIJ review of defining and measuring desistance emphasizes the lack of one universally accepted operational definition. Some researchers focus on sustained abstinence from crime; others examine a declining frequency of offenses or change in the estimated underlying tendency to offend. These approaches answer related but different questions.

Primary, Secondary, and Tertiary Desistance

Desistance scholarship often distinguishes three overlapping dimensions. These are research concepts and organizing perspectives rather than formal legal categories or standardized stages that every person must pass through.

Primary desistance concerns observable periods of nonoffending or the early interruption of criminal behavior. A person may stop because circumstances, opportunities, priorities, or constraints have changed, even before their self-image has altered in a lasting way. Since future behavior is unknown at the time of observation, an initial interruption cannot independently prove permanent change.

Secondary desistance refers to a deeper transformation in how a person understands themselves, including movement away from an identity organized around offending. A person may begin to see themselves as a reliable parent, worker, neighbor, mentor, or community member. Identity change matters because future-oriented commitments can help organize conduct, but declarations of a new identity do not by themselves establish behavioral change.

Tertiary desistance refers to recognition and belonging in the wider social world. It draws attention to whether employers, families, institutions, and communities make it possible for a person to live as someone whose past conduct no longer defines every opportunity. Social recognition may support lasting change, although the precise causal contribution is difficult to isolate.

The three dimensions can influence one another. A new role may promote a different self-concept; changed conduct may restore trust; social acceptance can reinforce the practical value of a nonoffending life. Progress may be uneven, and these dimensions should be assessed through behavior and context rather than inferred from a formula.

How Desistance Differs from Recidivism, Rehabilitation, and Deterrence

Recidivism generally means a return to criminal behavior following a prior offense, sanction, or intervention. Administrative studies operationalize it using rearrest, reconviction, or reincarceration over a specified follow-up period. The National Institute of Justice's recidivism overview explains why these measures are related but distinct.

Desistance asks how offending changes across time and why a person may move away from it. A binary recidivism indicator may treat a person with one minor rearrest similarly to someone with numerous new serious convictions, while a desistance-oriented analysis may detect reduced frequency or severity. Conversely, a person can have no recorded rearrest despite continued undetected offending. Neither framework should erase the other.

Rehabilitation refers to changes associated with intervention, functioning, and reduced criminogenic needs. A correctional program may develop useful skills without yet showing a reduction in offending. Deterrence concerns how anticipated or experienced sanctions may affect decisions. Incapacitation reduces opportunities to offend in the community by confining someone. These mechanisms have different definitions and should never be counted automatically as desistance.

Why Do People Stop Offending? The Main Explanatory Frameworks

Aging and developmental change

The association between age and criminal offending is a durable population-level finding. Many longitudinal samples show declining offending as people move through adulthood, and most adolescents who offend do not continue on a persistent high-rate criminal trajectory. Changes in judgment, self-regulation, responsibilities, peer contexts, and opportunity can accompany this transition. The NIJ's overview of pathways among juveniles and adults discusses these patterns and their implications.

The familiar image of an identical age–crime curve across cultures is too simple. A 2025 Annual Review of Criminology analysis of international and historical variation documents differences in the timing and shape of age–crime patterns across societies. Aging is therefore an important correlate and part of a larger developmental explanation; reaching a particular birthday neither guarantees desistance nor establishes why it happened.

Social bonds and turning points

Life-course theory examines how social relations and institutions can redirect behavior. Stable work, supportive partnerships, caregiving, military service in some historical samples, and other structured commitments may provide routines, expectations, legitimate resources, and informal accountability. A turning point is meaningful when it changes daily life and relationships, not merely because a person acquires a new label.

Laub and Sampson identified work, marriage, aging, and identity transformation as recurrent correlates while also acknowledging the limitations of causal evidence. A supportive relationship may matter because it alters time use and attachments, while an abusive or destabilizing relationship could have the opposite effect. Marriage is neither necessary nor sufficient for desistance, and associations observed in older cohorts cannot simply be transferred to every contemporary population.

Cognitive transformation and identity

People can reinterpret their previous conduct, imagine an alternative future, and become more receptive to an opportunity for change. In Giordano, Cernkovich, and Rudolph's theory of cognitive transformation, readiness for change interacts with practical opportunities and the meanings people assign to them. The theory highlights agency while recognizing that social constraints affect which alternatives are attainable.

A longitudinal study by Copp and colleagues found that identity change was associated with declining offending, alongside changes in parental closeness and affiliation with delinquent peers. Its results illustrate why internal resolve and relational context belong in the same explanation. They do not show that an identity intervention alone causes crime to stop for a particular individual.

Opportunity, routines, and social environments

Offending opportunities can change as daily movements, leisure activities, housing, transportation, work schedules, or peer networks change. Someone who no longer spends time in a setting where theft, assault, or illegal markets were normalized may have fewer occasions to participate. Social environments also affect exposure to conflict, illicit opportunities, and reinforcement of lawbreaking.

Changes in routine can support desistance without a dramatic psychological transformation. In other cases, a new identity enables someone to avoid the places and situations in which offending previously occurred. These pathways can reinforce one another. The causal sequence depends on individual history and the offense context.

Biological and developmental influences

Research on brain development, stress regulation, and neuropsychological functioning can inform broad developmental understanding of offending trajectories. However, neurobiology does not provide a definitive individual test of whether someone has desisted or will offend again. The NIJ's biosocial review considers such influences within a wider social and correctional framework.

The strongest responsible interpretation is multilevel: biological development may influence capacities and responses, which interact with adversity, opportunity, relationships, economic conditions, supervision, and choices. Neural correlates and psychological traits do not establish criminal responsibility, legal culpability, or future behavior.

What Changes in a Person's Life Can Support Desistance?

Stable, meaningful employment

Employment can provide income, predictable routines, social contacts, recognition, and a legitimate role. Job quality, safety, sustainability, and compatibility with family obligations matter. A temporary low-wage job that leaves housing and basic needs insecure may create fewer durable resources than stable work with realistic advancement.

Employment is also vulnerable to selection effects: people already moving away from crime may be more likely to keep a job. A systematic review of noncustodial employment programs reported limited or modest effects on recidivism in its randomized studies. The U.S. Department of Labor's 2024 synthesis reviews more recent evaluations of employment-focused reentry programs. Helping people work has independent social value, but job placement alone must not be advertised as a universally proven crime-prevention treatment.

Family, parenting, and supportive relationships

Reliable relationships may create emotional support, prosocial expectations, supervision outside the justice system, and reasons to plan beyond the immediate moment. Parenting can change priorities and responsibility, but the ability to fulfill parental roles depends on resources, safe relationships, and realistic access to children. Family pressure, stigma, or ongoing conflict can complicate change.

Not every person has a spouse or supportive family. Friends, mentors, colleagues, community organizations, and chosen family may supply important forms of social connection. The Copp and colleagues longitudinal findings support attention to the combined role of identity, parental closeness, and peers rather than prescribing one family arrangement.

Peer networks and community belonging

Reduced involvement with peers who facilitate illegal behavior may lower exposure to offending opportunities and pressures. Positive peer and community networks can offer practical assistance and recognition of change. A person may simultaneously need to leave one social environment and gain access to another, which can be particularly difficult following incarceration or long-term exclusion.

This is why commands to change one's associates have limited practical meaning without attention to accessible housing, jobs, transport, and trustworthy community institutions. Sustainable change often depends on building viable alternatives.

Housing and everyday stability

Safe and stable housing makes employment, appointments, treatment, education, and family contact easier to sustain. Housing instability can generate repeated crises, conflict, exhaustion, and exposure to settings that undermine progress. It is a practical reentry issue and a possible contributor to offending trajectories, although its isolated causal effect varies by setting and study design.

The NIJ's reentry practice review identifies housing, employment, cognitive and behavioral skills, substance use, and health as domains requiring individualized attention. Better housing outcomes also matter in their own right; they should be measured separately from criminal justice outcomes.

Substance use, treatment, and health

For individuals whose offending is intertwined with harmful substance use, effective treatment and recovery support may help reduce exposure to certain risks and strengthen stability. The relationship is heterogeneous: many people with substance use disorders never commit crimes, while crime can arise through mechanisms unrelated to substance use. A clinical diagnosis is not an assessment of guilt, dangerousness, or criminal intent.

Mental and physical health care, including continuity after release, can improve functioning and access to support. Treatment should address a person's assessed needs; there is no evidentiary basis to treat psychiatric symptoms or diagnoses as interchangeable with criminal propensity.

Agency, hope, and practical possibility

The intention to lead a different life can shape decisions and persistence, particularly when the person sees plausible future roles and receives concrete support. Personal agency is exercised within material constraints. Legal employment barriers, housing exclusions, debt, discrimination, transportation problems, and fragmented services can make nonoffending goals hard to enact.

Hope becomes actionable when aspirations can be linked to obtainable opportunities. Researchers therefore examine both subjective motivation and the social structures in which a person tries to change. Neither dimension alone is a complete causal explanation.

Does Everyone Desist in the Same Way?

Desistance trajectories vary across offending histories, ages, offense types, socioeconomic situations, and jurisdictions. Some people disengage rapidly; others show a gradual decline in frequency or seriousness, interrupted by setbacks or periods of instability. Many people stop without participating in a named rehabilitation program.

An adolescent with episodic delinquency, an adult involved in repeated acquisitive offenses, and a person with a long history of serious violence should not be assumed to share a uniform developmental or intervention pathway. Patterns in samples of one population cannot establish the likely outcome for an individual from another population.

Sex and gender, caregiving arrangements, economic opportunity, exposure to discrimination, and institutional responses may alter pathways. In their 2002 study of serious adolescent offending and later change, Giordano and colleagues found that job stability and marital attachment did not have the same strong associations assumed by some earlier accounts. The finding is a reason to investigate heterogeneity, not a universal rule for women or men.

Juvenile Desistance and the Transition to Adulthood

Adolescence is a period of major developmental and social change. In many longitudinal studies, involvement in lawbreaking peaks around adolescence or young adulthood and subsequently decreases. Early offending does not establish an inevitable lifelong criminal career. Developmental pathways nevertheless vary widely, and a minority of individuals persist with high rates of offending.

For youth, family relationships, schooling, peers, neighborhood conditions, access to support, substance use, developmental functioning, and contact with justice institutions can all be relevant. Youth-specific assessments should account for age, maturity, changing social environments, and the shorter period over which behavior has been observed. The NIJ's synthesis of juvenile and adult desistance illustrates why youth policy should recognize developmental change.

A young person's previous conduct cannot by itself establish their permanent character or future conduct. Equally, aggregate age-related declines cannot rule out serious risk in a particular situation. Both mistakes substitute a statistical stereotype for individualized, context-sensitive evaluation.

Desistance after Prison and during Reentry

Release from custody can reorganize nearly every aspect of daily life. A person may have to secure identification, housing, transportation, income, health care, contact with children, and compliance with legal conditions in a short period. These demands affect the ability to establish sustainable routines and supportive relationships.

Incarceration also can interrupt labor-market participation, family connections, and health care. The NIJ desistance volume reviews institutional effects and the limited evidence that confinement itself generates lasting desistance among people who previously engaged in persistent offending. A period without community offending while confined is not automatically equivalent to a durable post-release transition.

Reentry planning can start before release and continue afterward. Effective coordination may connect correctional services with community agencies while protecting continuity of support. The aims should include safety, lawful conduct, health, stability, and realistic access to social participation.

What Can Rehabilitation and Supervision Actually Achieve?

Psychological and behavioral interventions

Programs informed by cognitive-behavioral principles may address problem-solving, decision-making, harmful behavioral patterns, self-management, and situations associated with reoffending. Motivational approaches may help people articulate reasons for change, and interventions for substance use or other clinical conditions may improve health and functioning when indicated. Program effectiveness depends on whom it serves, implementation quality, engagement, and measured outcomes.

A desistance-focused system uses evidence-based interventions while acknowledging that an improved questionnaire score or completion certificate is a process indicator, not proof of a lower crime rate. Claims about intervention effects must rest on trials and valid comparative evaluations, not on the theory alone.

Employment and reentry programs

A meta-analysis of nine randomized reentry program evaluations for adult men found pooled effects on rearrest, reconviction, and reincarceration that were statistically uncertain; confidence intervals included no effect. The finding does not mean supportive services lack value. It means that broad claims of reduced recidivism cannot be established from those trials.

A systematic review and meta-analysis of interventions to improve employment after prison examines employment outcomes specifically. Work outcomes, personal stability, and recidivism are separate endpoints. Improving one does not automatically establish improvement in another.

Probation, parole, and community supervision

Supervision can combine accountability, monitoring, referrals, and support. Its effects depend on the conditions, sanctions, resources, and population involved. Excessively complex requirements may increase technical violations without demonstrating new criminal behavior; inadequate support may leave practical obstacles unresolved.

A desistance-oriented approach asks what helps a person sustain lawful living while maintaining proportionate public-safety safeguards. As NIJ's review of desistance-focused practice observes, the field continues to debate how concepts such as reduced frequency, reduced severity, and sustained nonoffending can be operationalized in everyday correctional decisions.

How Do Researchers Measure Desistance?

Desistance is difficult to measure precisely because it unfolds over time, criminal behavior is incompletely observed, and different studies employ different definitions. A robust study specifies who is followed, how the starting point is determined, which offenses count, what observation sources are used, and how long participants remain at risk of being observed.

Administrative records may include arrest, conviction, and incarceration. Each is affected by policing, prosecution, reporting, and legal procedure. Arrest is not proof of guilt; conviction is a legal outcome; reincarceration can result from technical supervision violations. A finding about one outcome cannot silently be generalized to all criminal activity.

Self-report surveys can capture unrecorded conduct and changes in frequency, though they are subject to recall errors, disclosure concerns, and attrition. Victimization data and official records provide different perspectives. Combining sources may improve interpretation without removing all biases.

Longitudinal research can study individual trajectories, rates of offending, escalation or de-escalation, time to a subsequent event, and sustained periods of nonoffending. A brief follow-up may mistake temporary interruption for long-term change. Longer follow-up creates other problems, including changes in surveillance intensity, life circumstances, and the availability of records.

The NIJ's methodological review recommends clarity about definitions and measurement. Researchers should report observation windows and attrition, compare alternative operationalizations, and avoid treating a single binary indicator as the entire process.

Evidence, Association, Causation, and Individual Prediction

Research can establish that characteristics such as employment stability or changed peer associations correlate with declining offending. It does not follow that assigning employment or separating peers would produce the same change for everyone. People select into relationships and programs; their motivation and circumstances may have changed before the observed event.

Randomized studies can strengthen causal inference about the specific intervention tested but may have limited generalizability to other programs or populations. Natural experiments and careful longitudinal modeling can add evidence, yet unmeasured confounding and changing circumstances remain important.

A risk factor is a statistical correlate or contributor under specified conditions. A protective factor is associated with reduced likelihood or resilience in a particular context. A screening result or psychometric score summarizes data according to an instrument's design. A forensic opinion is a professional interpretation addressing a defined referral question. None of these is a judicial finding or a reliable prophecy of individual future behavior.

If a risk assessment is used in supervision or sentencing, validity must be checked for the relevant population, setting, time horizon, and outcome. Accuracy should be evaluated through discrimination, calibration, false-positive and false-negative rates, subgroup performance, and consequences of use. A prediction about group-level reoffending rates is different from certainty about one person. Legal admissibility and permitted uses vary by jurisdiction.

Social Inequality, Stigma, and the Possibility of Change

A person's ability to pursue desistance depends partly on whether lawful opportunities are accessible. Restrictive licensing rules, criminal-record screening, unstable housing markets, exclusion from education, and persistent labeling can complicate reentry. Their influence will vary with law, local institutions, resources, and offense history.

Recognition of change does not require minimizing past harm or disregarding victims' interests. Accountability, restitution where appropriate, fair procedure, and practical reintegration can coexist. A constructive public-safety framework considers both the harm of further offending and the conditions under which sustained nonoffending becomes possible.

Race, ethnicity, neighborhood, disability, and socioeconomic circumstances can affect surveillance and recorded outcomes. Analysts should distinguish differences in detected events from differences in underlying offending and examine whether measurements and interventions are fair for the population in which they are applied.

Is Desistance Permanent? Setbacks and Long-Term Change

A person may stop offending for years, experience a new arrest, or have an isolated offense amid a much longer reduction in criminal activity. Researchers choose different thresholds for calling these patterns desistance. A single event changes the evidence about a trajectory but does not erase all earlier change in frequency, severity, or personal functioning.

The concept of relapse should be used carefully because crime is a legal-behavioral category rather than a single clinical condition. Persistence, interruption, and renewed offending are patterns of conduct; they are not diagnoses. Where substance-use recovery and offending change overlap, each requires its own measures.

Long-term desistance is best understood through repeated observations and contextual evidence. The practical question is how to extend stable nonoffending and reduce future harm, while retaining accurate records of setbacks and uncertainty.

What Desistance Means for Criminal Justice Policy

First, justice systems can distinguish reduced offending from a narrow all-or-nothing recidivism metric. Reporting frequency, severity, and timing alongside rearrest or reconviction can reveal meaningful behavioral change without hiding harmful events.

Second, correctional responses can connect accountability to supports that make lawful living realistic, especially during reentry. Housing, employment, health care, treatment where appropriate, family contact, and community participation are relevant domains, but each intervention requires evaluation.

Third, policy should assess both immediate public-safety needs and longer-term consequences. An intervention that increases arrests through intensified monitoring may appear to worsen recidivism even if underlying behavior is unchanged. Conversely, fewer recorded offenses may reflect differences in detection. Interpretation demands attention to the measurement process.

Finally, the NIJ's desistance research and policy volume argues for better integration of research into correctional practice and coordinated work among justice agencies, local services, families, and communities. Effective systems measure their actual results, review unintended effects, and adjust interventions to evidence.

Common Myths About Why People Stop Committing Crimes

Myth: People either are criminals or are not

Offending is behavior observed over time, and the same person's conduct can change substantially. Longitudinal patterns do not support treating a past offense as a fixed lifetime identity. A criminal conviction has a defined legal meaning; it does not automatically specify a person's psychological future.

Myth: Everyone grows out of crime at the same age

Aggregate offending often declines with adulthood, but individual and cross-national trajectories vary. The 2025 international review of the age–crime curve shows why a single supposedly universal timetable is misleading.

Myth: Marriage or getting a job automatically ends offending

Stable relationships and employment can support change, but the evidence includes selection effects and inconsistent intervention findings. The quality of the relationship or job and the surrounding circumstances matter.

Myth: An absence of arrests proves desistance

Arrests reflect detection and procedural decisions. Self-reported offending, conviction records, observation periods, and longitudinal context can yield different conclusions.

Myth: Prison alone produces lasting change

Confinement constrains opportunities for some forms of offending while it lasts. Its longer-term consequences include disrupted relationships and economic opportunity; sustained post-release desistance requires separate evidence.

Myth: A psychological test can certify that someone will never reoffend

No assessment establishes future behavior with certainty. Forensic use demands population-specific validity, calibrated interpretation, fairness analysis, and respect for legal standards.

Frequently Asked Questions

What does desistance from crime mean in simple terms?

It means a person is moving away from committing offenses, often by offending less frequently, less seriously, or ultimately stopping. Researchers study the behavioral pattern and the changes in life circumstances and motivation that accompany it.

Why do most people stop committing crimes?

There is no single explanation. Developmental change, shifts in opportunities and peers, stronger prosocial relationships, improved stability, identity change, and individual decisions may interact. Their importance varies by history and context.

Does desistance always happen gradually?

No. Some individuals show abrupt behavioral interruptions, while others reduce frequency over years. Observing a sudden stop does not tell researchers whether the underlying transition will last.

What is the difference between primary and secondary desistance?

Primary desistance concerns the observable interruption or absence of offending. Secondary desistance emphasizes an accompanying change in personal identity. Both are research concepts rather than diagnoses or legal verdicts.

What is tertiary desistance?

Tertiary desistance emphasizes recognition, social acceptance, and belonging: whether others allow a person to occupy valued roles beyond their past involvement in crime.

What is the difference between desistance and recidivism?

Desistance focuses on the process of reducing or ending offending over time. Recidivism focuses on return to offending after a prior event, often measured by rearrest, reconviction, or reincarceration over a fixed period.

Does getting older mean someone becomes safer?

Age is associated with changes in offending at the population level. It cannot establish the behavior or risk of a specific person. Any forensic assessment must consider the actual referral question and relevant evidence.

Can therapy or rehabilitation promote desistance?

Some interventions may address relevant behavioral or clinical needs, and supportive services can facilitate change. Whether a program reduces reoffending must be shown through appropriate evaluation rather than inferred from participation or improved wellbeing alone.

Can someone desist without prison or professional treatment?

Yes. Desistance may occur through developmental, relational, social, and practical changes without a named intervention. The causal mechanisms still require study, and there are no guarantees for an individual case.

How long must someone avoid offending before they have desisted?

There is no universally accepted threshold. Researchers specify a follow-up period and the measure of offending. An extended period without recorded offenses offers stronger descriptive evidence than a brief interval but never proves certainty about the future.

Is desistance a legal status or clinical diagnosis?

No. It is a criminological and psychological research concept. Courts make legal findings under applicable jurisdictional standards, and clinicians diagnose conditions under their professional frameworks. Neither activity is interchangeable with identifying desistance.

What is the most important takeaway for reentry services?

Support should be individualized, coordinated, and measured on several outcomes. A lawful, stable life may involve housing, work, relationships, health care, and changed routines; empirical evaluation must determine whether a given service actually reduces offending.

References

Beaudry, G., et al. (2021). Effectiveness of psychological interventions in prison to reduce recidivism. The Lancet Psychiatry. DOI00170-X).

Berghuis, M. (2018). Reentry programs for adult male offender recidivism and reintegration: A systematic review and meta-analysis. International Journal of Offender Therapy and Comparative Criminology, 62, 4655–4676. DOI.

Boisvert, D. L. (2021). Biosocial factors and their influence on desistance. National Institute of Justice. Official publication.

Connell, C., et al. (2023). Effectiveness of interventions to improve employment for people released from prison: Systematic review and meta-analysis. Health & Justice, 11, 17. DOI.

Copp, J. E., Giordano, P. C., Longmore, M. A., & Manning, W. D. (2020). Desistance from crime during the transition to adulthood: The influence of parents, peers, and shifts in identity. Journal of Research in Crime and Delinquency. DOI.

Giordano, P. C., Cernkovich, S. A., & Rudolph, J. L. (2002). Gender, crime, and desistance: Toward a theory of cognitive transformation. American Journal of Sociology, 107, 990–1064. DOI.

Laub, J. H., & Sampson, R. J. (2001). Understanding desistance from crime. Crime and Justice, 28, 1–69. DOI.

National Institute of Justice. (2021). But what does it mean? Defining, measuring, and analyzing desistance from crime in criminal justice. Official publication.

National Institute of Justice. (2021). Desistance from crime: Implications for research, policy, and practice. Official research collection.

National Institute of Justice. (2021). Desistance-focused criminal justice practice. Official publication.

National Institute of Justice. (2023). Five things about reentry. Official guidance.

National Institute of Justice. (n.d.). Recidivism. Official overview.

Steffensmeier, D., Slepicka, J., & Schwartz, J. (2025). International and historical variation in the age–crime curve. Annual Review of Criminology, 8, 239–268. DOI.

U.S. Department of Labor. (2024). Charting the path to employment after incarceration: A research synthesis of employment-focused reentry programs. Official research synthesis.

Visher, C. A., Winterfield, L., & Coggeshall, M. B. (2006). Systematic review of non-custodial employment programs: Impact on recidivism rates of ex-offenders. Campbell Systematic Reviews. DOI.

Psychological Encyclopedia