Developmental Pathways to Offending: From Childhood Risk to Adult Crime

4 days ago

19 min read

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

Ukrainian Psychological Hub | October 11, 2026 | Psychology for the AI Era

Developmental pathways to offending are patterns through which unlawful or delinquent behavior begins, changes, persists, or declines between childhood and adult life. The field investigates how a person's development interacts with family environments, peers, education, neighborhoods, institutions, social opportunities, and changing circumstances. It asks why some early difficulties are followed by persistent involvement in offenses, why others are confined to adolescence, and why people with similar childhood experiences can develop in very different directions. Moffitt (1993) Moffitt (2018)

The central scientific finding is that risk accumulates across multiple domains while individual futures remain uncertain. A history of adversity, behavioral difficulties, or an early offense can raise an observed group-level probability without determining what any particular child will do. The 2024 umbrella review of youth-crime research identified several well-supported associations; it did not establish a deterministic pathway from a childhood characteristic to adult criminality. Ayano et al. (2024) Jolliffe et al. (2017)

What Is a Developmental Pathway to Offending?

A developmental pathway is a sequence of behavioral and contextual observations collected over time. Researchers can measure self-reported rule-breaking, aggressive behavior, police contact, juvenile adjudications, arrests, or convictions, but each measure captures a different event or institution. An offense against a person, a property offense, and a technical violation also differ in mechanisms and legal gravity. The endpoint chosen changes the apparent shape of a trajectory and how it should be interpreted. Farrington et al. (2009)

Four terms organize this research. Onset means the first observed occurrence within a specified measurement system. Persistence describes continued recorded involvement across observation periods. Escalation describes increasing frequency or seriousness, although these can diverge. Desistance refers to reduction or cessation of offending over time and may unfold gradually. An official first arrest is not necessarily a person's first unlawful act, and a period without an arrest is not necessarily evidence of complete behavioral change.

Psychological constructs and legal categories

Aggression, impulsivity, self-control, sensation seeking, and antisocial behavior are psychological constructs assessed through specified measures. Clinical conditions have diagnostic criteria and require clinical assessment. A risk factor is a statistical correlate or potential contributor, not proof of an act. Protective factors describe conditions associated with lower risks. A screening score represents the output of a particular instrument. A forensic opinion addresses a referral question with a documented method, while guilt and other legal findings belong to the legal decision-maker.

Terms such as juvenile delinquency, status offense, adult crime, criminal responsibility, competency, and capacity have legal meanings that vary by state and country. The age dividing juvenile and adult jurisdiction is established by law rather than developmental science. A developmental formulation may help explain context and guide services, but it cannot determine whether a statute was violated or whether an individual meets a legal standard.

The Age–Crime Curve and Why It Matters

Across many datasets, recorded offending rises through adolescence and falls as people move toward adulthood. The exact height, peak, and decline vary across eras, offenses, places, and measurement techniques. This age–crime curve describes a population distribution, not a universal experience of teenagers. It also conceals important differences: some people first show behavior early, some only during adolescence, others intermittently or after a later onset. Moffitt (1993) Moffitt (2018)

Researchers must avoid using the age–crime curve as a substitute for personal history. The same recorded offense committed by two youths can arise in different settings: a brief incident shaped by peers, an escalating pattern with multiple adverse exposures, or a behavior associated with a current crisis. The relevant scientific questions concern the timing, consistency, severity, setting, and opportunities for change.

The Main Developmental Models

Life-course-persistent antisocial behavior

Terrie Moffitt's developmental taxonomy proposed a relatively small group whose antisocial behavior begins early and remains visible in changing forms across life. In the theory, neurodevelopmental vulnerabilities can interact with adverse environments, conflict, restricted opportunities, and cumulative disadvantage. The same person may show disruptive childhood behavior, later delinquency, and subsequent adult law violations, but the particular manifestations vary. Life-course-persistent is a research pattern rather than an irreversible personal classification. Moffitt (1993) Moffitt (2018)

The theory highlights reciprocal processes. An early difficulty with self-regulation can contribute to difficult peer and school experiences, which may limit access to prosocial supports. Exclusion, repeated institutional contact, and exposure to offending opportunities can produce additional disadvantages. These interactions can continue over time, but each link requires empirical investigation in a particular population. The model does not justify labeling a child as destined for an adult criminal career. Fairchild et al. (2013)

Adolescence-limited antisocial behavior

A second trajectory in Moffitt's framework concentrates antisocial or offending behavior around adolescence and declines as adult opportunities and roles expand. Proposed explanations include peer imitation, autonomy seeking, immediate rewards, and a maturity gap between increasing independence and restricted adult status. This is a useful theoretical contrast with persistent pathways. It does not suggest that all adolescent misconduct is trivial, harmless, or part of normal development. Moffitt (1993) Moffitt (2018)

The category adolescence-limited is particularly susceptible to hindsight. A researcher can determine that someone's behavior was confined to adolescence only after observing later years. At the moment of an incident, a clinician, educator, or court cannot know whether the behavior will continue. A label inferred from one episode can easily overstate the evidence and produce unjustified confidence about future events.

Late-onset, intermittent, and additional trajectories

Later longitudinal work has documented more than two group patterns. The Cambridge Study in Delinquent Development followed a sample of South London males from childhood into middle adulthood and compared people with youth-only convictions, later-onset convictions, persistence across periods, and no observed convictions. The groups had differing later outcomes and early correlates, underscoring the importance of timing, duration, and study design. Farrington et al. (2009)

Some participants begin offending later, some move in and out of official records, and some change offense type or seriousness. These observations challenge a simple division between permanently high-risk children and temporary adolescent offenders. They also show why observation length matters: a person followed only until age 18 may be placed in a different category than the same person followed to age 40. Researchers should report how categories were built and how sensitive they are to statistical assumptions.

How distinct are the proposed types?

A systematic review examined early risk factors for life-course-persistent, adolescence-limited, and late-onset patterns in prospective studies. Of 55 studies with potential relevance, only a small subset supplied direct comparisons across the specified groups. The authors found little consistent evidence that early factors uniquely identify distinct trajectory types. Persistent groups tended to have more risk exposures or stronger exposure levels than less persistent groups, implying differences of degree more clearly than entirely separate childhood causes. Jolliffe et al. (2017)

The distinction matters for practice. A child's combination of risk markers may justify assessment of current needs and supportive services, but it does not reveal a hidden lifelong category. Developmental groups summarize past observations. They should guide testable explanations and opportunities for prevention, rather than define a person's identity.

Childhood Risks: Individual and Environmental Processes

Early self-regulation and behavioral difficulties

Researchers have studied persistent disruptive behavior, hyperactivity, impulsive responding, poor behavioral regulation, and difficulties in classroom functioning as early markers associated with later conduct problems. Their meaning depends on duration, severity, context, and impairment. A single episode of defiance or a tendency toward restlessness is not evidence of a future offense. Similar behavior may reflect development, disability, learning needs, family stress, environmental mismatch, or several processes at once. CDC (2024) Wasserman et al. (2003)

A neurodevelopmental difference cannot be equated with dangerousness or delinquency. ADHD, autism, learning conditions, and other clinical questions require their own differential assessment when relevant. Group-level correlations must be interpreted within appropriate diagnostic and nonclinical comparison groups. The primary purpose of assessment should be to clarify support needs and functioning, not to attach an enduring prediction of criminality.

Caregiving, parental supervision, and family relationships

Parent–child relationships, consistent caregiving, supervision, family conflict, neglect, and harsh or inconsistent discipline have been examined across decades of research. The 2024 umbrella review reported highly suggestive associations linking several family-level factors with youth offending. Yet these are pooled correlates, often measured through different informants and methods. They do not justify blaming one caregiver, assuming a particular family structure is criminogenic, or diagnosing a family from a statistical checklist. Ayano et al. (2024)

Caregiving is embedded within material and social conditions. Housing instability, lack of childcare, economic insecurity, neighborhood violence, discrimination, scarce treatment resources, and parental illness can shape what support is available. Preventive approaches should strengthen safety and reliable relationships while recognizing structural barriers. Treating poverty as proof of poor parenting or a direct cause of crime confuses demographic association, contextual disadvantage, and individualized behavior.

Adverse childhood experiences, abuse, and neglect

Adverse childhood experiences can include maltreatment, exposure to violence, and other potentially traumatic events. Such exposures have been linked to certain youth-crime measures, but the association does not mean that trauma produces criminality. Exposure may be connected to schooling, physical health, emotional functioning, peer relationships, institutional encounters, and other mechanisms that vary among children. A trauma history is primarily information about something a child experienced and the support they may need. Ayano et al. (2024) CDC (2024)

Protective relationships and accessible services can alter developmental conditions. Researchers must also account for retrospective recall, overlapping forms of adversity, shared family or neighborhood factors, and differences in how offending is identified. For a comprehensive account of a victim-centered clinical and developmental subject, the Hub's Childhood Exposure to Domestic Violence: Effects, Risk, and Resilience retains ownership of that broader topic. The present article deals specifically with forensic-developmental associations.

School experiences and educational opportunity

Poor academic engagement, learning difficulties, chronic absenteeism, exclusionary discipline, and school disconnection may correlate with later justice involvement. Their causes vary and frequently overlap: educational support may be inadequate, a young person may experience bullying, family instability may disrupt attendance, or repeated disciplinary responses may strain belonging. A supportive school environment appeared as a protective factor in the umbrella review, making institutional context central to interpretation. Ayano et al. (2024)

School-based prevention can include timely learning assessments, supportive relationships, predictable environments, fair discipline, responses to victimization, and access to extracurricular opportunities. Such supports can be justified by educational and developmental needs even when no crime risk is established. Screening children solely for future criminality can impose stigma and unnecessary surveillance if predictions are imprecise.

Communities, neighborhood conditions, and opportunities

Community violence, housing insecurity, concentrated deprivation, limited safe activities, and reduced employment opportunities may influence both exposure and opportunities for unlawful behavior. Measured crime and arrest statistics can also reflect differences in enforcement and reporting across communities. The CDC's public-health model treats youth violence as a product of interacting individual, relationship, community, and societal conditions, with protective conditions operating at each level. CDC (2024)

A neighborhood description is not an assessment of a resident. A postcode, income indicator, or family background should not be interpreted as personal culpability. At most, such variables help researchers study how social environments are distributed and whether changes in resources, housing, education, and safety can improve outcomes.

Adolescence: Peers, Opportunity, and Decision-Making

Peer influence and peer selection

Peers become especially important contexts for learning, belonging, and social reward during adolescence. Association with delinquent peers and experiences of peer pressure are linked to youth offending in research syntheses. These relationships can reflect direct influence, selection of similar peers, shared opportunities, or common environmental causes. One adolescent's friendship circle is not in itself evidence of offending, and observational studies rarely isolate a single causal mechanism. Ayano et al. (2024)

Peer relationships can be protective as well as risky. Positive friendships, school belonging, supervised activities, and caring adults may create alternatives to risky settings. In an early-onset Pittsburgh cohort, positive peer relationships and family involvement were promotive correlates of desistance. The study's selected sample of boys and its follow-up design limit generalization to all adolescents. van Domburgh et al. (2009)

Reward sensitivity and social context

Adolescent decision-making can be affected by immediate rewards, emotionally charged situations, peer observation, and still-developing self-regulatory capacities. Steinberg's social neuroscience model explains how social and motivational processes may temporarily increase risk-taking in some contexts. These developmental processes vary considerably, and they do not imply that all adolescents lack understanding or are unable to make reasoned decisions. Steinberg (2008)

Developmental findings should inform age-appropriate support and legally authorized responses. The National Research Council's consensus report emphasized developmental science, accountability, fairness, rehabilitation, and safeguards within juvenile justice. Research on maturation informs policy deliberation; it does not itself establish a legal age of responsibility, invalidate all adolescent choices, or determine the proper sentence in an individual case. National Research Council (2013)

Substance use and other co-occurring difficulties

Substance use is a well-documented statistical correlate of youth crime and may affect immediate judgment, peer opportunities, or the environments in which offending occurs. These relationships are reciprocal and context dependent; a substance-use diagnosis and a criminal act are separate matters. Likewise, clinical symptoms can be relevant to individual care without providing a general explanation for illegal behavior. Ayano et al. (2024)

Mental illness should never serve as shorthand for criminality. Psychiatric diagnoses require diagnostic assessment, and risk assessment requires its own properly defined outcome and evidence. A separate clinical owner in the English Hub, Schizophrenia and Violence: What Research Really Shows, addresses one often-stigmatized intersection without collapsing a diagnosis into an individual prediction.

From Adolescence to Adulthood: Persistence, Turning Points, and Desistance

Accumulated risk and reciprocal processes

Persistent patterns may arise when disadvantages reinforce one another. School disengagement can reduce qualifications, justice-system involvement may interrupt education and employment, unstable housing can complicate treatment access, and high-conflict networks may perpetuate offending opportunities. The relationship can run in both directions: offending can worsen circumstances, which then alter the probability of further offending. This developmental account is a set of mechanisms to investigate rather than a description that must fit every case.

The longitudinal review by Jolliffe and colleagues suggests that longer observed criminal careers often involve more numerous or stronger earlier risk markers. However, the same broad markers also occur in less persistent groups. Accumulation should therefore trigger questions about preventable barriers and current support rather than justify assigning a predetermined criminal career to a young person. Jolliffe et al. (2017)

Turning points and gradual change

Desistance may follow improved relationships, education, stable accommodation, employment, changes in social networks, personal commitments, or reductions in substance-related difficulties. Some changes unfold gradually without one decisive event. Reduced offending frequency, complete cessation, improved social functioning, and cessation of official justice contact are different outcomes that studies should measure separately.

In a study of 310 boys from the Pittsburgh Youth Study with offending before age 12, researchers distinguished serious and moderately serious persisters, desisters, and intermittent participants. Family involvement and positive peer relationships were associated with more favorable trajectories. The findings are informative within this selected sample; a retrospective assignment to one category is not an individually precise prediction. van Domburgh et al. (2009)

Why apparently late onset is difficult to interpret

Adult-onset recorded offending may be genuinely new, may reflect a delayed change in social opportunities, or may involve earlier behavior that was not recorded. Conviction records depend on reporting, enforcement, prosecution, and time observed. Comparisons between youth-only and adult-onset groups in the Cambridge cohort therefore illuminate patterns in official data without identifying a universal onset mechanism for adults. Farrington et al. (2009)

How Strong Is the Scientific Evidence?

What umbrella reviews can establish

Ayano and colleagues' 2024 umbrella review brought together earlier systematic reviews and meta-analyses, identifying 58 candidate factors and 11 with highly suggestive or suggestive evidence of association. Factors with reported support included prior offending, substance use, family supervision, maltreatment, peer pressure, and supportive school settings. These grades characterize the review's epidemiologic evidence, not the validity of an individual prediction or proof of causality. Ayano et al. (2024)

The pooled findings combine heterogeneous populations and outcomes; some estimates draw on overlapping primary studies, retrospective measures, or domains subject to confounding. A large odds ratio does not reveal a person's absolute probability. The review's evidence can support prevention priorities, but conclusions must remain tied to the particular population, comparator, time horizon, and measure used.

What prospective cohort studies can establish

Prospective studies measure some exposures before outcomes, allowing a more credible examination of developmental timing than retrospective narratives alone. They can identify patterns of continuity and change and show how early variables relate to later observed events. However, prospective observation alone cannot distinguish every causal mechanism, especially when family, neighborhood, peer, and institutional effects overlap or influence each other. Jolliffe et al. (2017) Farrington et al. (2009)

The age at first recruitment limits what can be learned about early onset. Participants lost to follow-up may differ from those retained. Self-reports can reveal behavior missed by official records but have recall and disclosure limitations. Arrest and conviction data are more formally documented but depend on criminal law and justice-system selection. Trajectory groups also vary with the statistical model chosen.

Why trajectory labels can be misleading

Statistical trajectory models often divide continuous variation into a small number of classes for description. Group membership may change if different time points, offenses, measures, or model settings are used. Labels such as 'high chronic' or 'adolescent-limited' are interpretive summaries of sample patterns, not discrete natural categories proven to exist within every population. Fairchild et al. (2013) Jolliffe et al. (2017)

This is particularly important in forensic contexts. An apparent fit between a child and a statistical group does not establish a future course. A responsible interpretation reports uncertainties, protective resources, context, and alternative explanations and avoids using an offender category as a clinical diagnosis.

Can Childhood Risk Predict Adult Criminal Behavior?

Predictive tools may rank groups by observed later event rates, but individual predictions face base-rate limits, measurement error, changing environments, and differences between training and application populations. A useful evaluation reports discrimination, calibration, sensitivity, specificity, false positives, false negatives, and performance in relevant demographic or geographic subgroups. Validation should specify the outcome: any arrest, violent reoffending, conviction, or another endpoint.

Consider a hypothetical and purely mathematical example. In a sample of 1,000 people where 5% later meet a defined outcome, a test with 70% sensitivity and 80% specificity would correctly flag about 35 eventual cases and incorrectly flag about 190 noncases, while missing roughly 15 cases. Fewer than one in six flagged people would experience the outcome in this example. These numbers do not describe an actual offender assessment instrument; they illustrate the dependence of positive predictive value on the base rate.

A higher score is not evidence of guilt, criminal responsibility, or certainty about an individual act. A forecast may be inaccurate even when it distinguishes groups reasonably well. When tools are used in legal settings, error costs and fairness matter alongside predictive performance. The intended population, collection date, follow-up period, and local conditions must be examined.

Population, culture, and fairness

Several historically influential cohorts involved boys or men from limited communities and decades. Their conclusions cannot be transferred automatically to girls, gender-diverse youth, different countries, or changed justice systems. Official contact may also differ because of disparities in enforcement, legal representation, institutional exclusion, and socioeconomic opportunity. Tools using school exclusion, arrest history, or neighborhood markers may reproduce existing inequities unless tested carefully. Farrington et al. (2009) National Research Council (2013)

Cultural competence involves evaluating measurement equivalence and whether a score has comparable error rates and calibration within the actual population. Researchers and forensic professionals should disclose missing evidence rather than presume that a risk model trained in one jurisdiction applies universally.

Prevention, Intervention, and Justice-System Responses

Support before formal justice contact

Early intervention can improve developmental conditions by supporting caregiver relationships, effective schooling, learning accommodations, treatment when clinically indicated, community safety, and access to stable resources. A child's need for assistance should be established through functioning, distress, and contextual circumstances, not an unsupported assumption of future criminality. Federal developmental-delinquency guidance has long emphasized both risk accumulation and the importance of protective factors. Wasserman et al. (2003) CDC (2024)

Programs should be assessed against tangible outcomes such as well-being, sustained school participation, later offending under a clear definition, and improvements in family functioning. Improvements on a short questionnaire are not automatically reductions in adult crime. Effects may vary by local services, intervention quality, age, and baseline needs.

How strong are family and community interventions?

A Campbell systematic review of multisystemic therapy reported favorable outcomes for some U.S. arrest/conviction and placement measures but weaker or absent comparable outcomes in other countries, with methodological concerns. The evidence does not establish universal effectiveness across justice systems or clinical problems. Comparing a specialized program with a poorly resourced usual-care service is different from comparing it with an established, effective alternative. Campbell MST review

A more recent meta-analysis of multisystemic therapy and functional family therapy found limited evidence of superiority to usual care for delinquency and antisocial behavior overall. Multisystemic therapy reduced time in out-of-home care, while some favorable functional family therapy estimates relied on poor-quality evidence. Intervention choice should reflect actual needs, acceptability, available services, and evaluation quality rather than marketing claims or a program name. Hunkin et al. (2025)

Diversion and the effects of formal processing

Justice-system responses can themselves become part of the developmental environment. A Campbell review of 29 controlled trials involving 7,304 youths found that formal juvenile processing did not improve subsequent delinquency outcomes and was associated on average with worse outcomes compared with alternatives such as diversion or release. The comparisons do not show that every diversion program succeeds or that legal accountability should disappear; they identify a policy-relevant risk of unnecessary formal processing. Petrosino et al. (2010)

Developmentally appropriate responses should address safety, legal rights, accountability, support, and proportionality together. Whether diversion is authorized or suitable depends on the offense, local law, procedural safeguards, and individual circumstances. Long-term studies are essential because short-term reductions in court contact do not automatically prove lasting behavioral change.

What Developmental Evidence Means in Forensic Practice

Forensic developmental assessment should begin with a precise referral question: for example, explaining a documented history, identifying service needs, or evaluating a construct under a specified legal framework. A clinician should identify the information sources and distinguish firsthand observations, collateral reports, historical records, psychological testing, inferences, and findings made by legal authorities. The purpose is transparent formulation rather than an impressionistic narrative.

Important questions include when behavior first appeared, whether it was consistent across settings, what was happening around each change, how peer and family circumstances evolved, whether disability or trauma-related needs require separate assessment, and which current protective factors exist. Risk or screening instruments, if used, require evidence for the population and purpose and must be interpreted with uncertainty and competing explanations.

No developmental trajectory proves that a defendant committed an offense. A psychological diagnosis cannot settle guilt. A risk score cannot substitute for evidence, and an offense history alone cannot determine competency, legal capacity, or criminal responsibility. The relevant courts and statutory authorities decide legal questions within their jurisdictions.

Limitations, Ethics, and Misconceptions

Childhood risk is not destiny

Many children exposed to adversity or displaying persistent behavioral difficulties do not become adult offenders, and some early offenders later stop. The strongest developmental explanations account for change, overlap, and differences among individuals. Research patterns should create opportunities for support rather than fixed categories that determine educational, clinical, or legal treatment. Jolliffe et al. (2017)

A diagnosis is not a crime explanation

Neurodevelopmental, mood, trauma-related, psychotic, and substance-use conditions each require proper diagnostic and clinical interpretation. Their presence alone does not explain any specific criminal act. It is equally inappropriate to assume that behavior found unlawful must have arisen from mental illness.

Peer influence does not remove agency

Peer context can change immediate incentives and opportunities, especially during adolescence, but its influence varies. Developmental science describes conditions affecting decisions and possible prevention strategies. Whether an individual is legally accountable is decided according to the jurisdiction's applicable standards and the evidence in the case. Steinberg (2008) National Research Council (2013)

A statistical association is not a causal intervention target

Associations in a review may identify useful hypotheses, but changing a measured correlate does not guarantee that offending will decrease. Intervention studies need valid comparisons, defined outcomes, attention to harms, and follow-up. Research should also report when a promising theory fails to produce practical benefit. Ayano et al. (2024) Hunkin et al. (2025)

Official records are not the complete history of behavior

Arrest statistics and convictions are products of reporting, enforcement, procedural decisions, and substantive law. Comparisons of official outcomes should account for these conditions, especially when assessing apparent racial, regional, or historical differences. The OJJDP briefing book is an authoritative source for American youth arrest trends but not a census of every delinquent act. OJJDP statistics

Frequently Asked Questions

What are the two best-known developmental pathways to offending?

Moffitt's influential model contrasts a life-course-persistent pattern starting early and extending through adulthood with a more adolescence-limited pattern. Subsequent research also examines later onset, intermittent trajectories, and variation in seriousness. The categories are theoretical or statistical, not psychiatric diagnoses. Moffitt (1993) Moffitt (2018)

Does childhood aggression predict adult crime?

Persistent childhood aggression can correlate with later offending in studied groups, especially alongside other adversity or difficulties. Its ability to predict an individual's adult behavior remains limited, and many children showing aggression never offend as adults. The measurement, age, context, and protective factors matter. Jolliffe et al. (2017) CDC (2024)

Can adverse childhood experiences lead to crime?

Some adversity measures are associated with higher rates of certain offending outcomes. This relationship may involve several mechanisms and confounding influences. Experiencing trauma does not imply that someone is criminal or dangerous; appropriate support follows from the person's experiences and needs. Ayano et al. (2024)

Why do crime rates commonly rise during adolescence?

The age–crime pattern has been linked to peer environments, opportunities, reward sensitivity, and developing regulation, with variation by offense and setting. It does not mean every teenager offends or that adolescent decisions are identical across contexts. Moffitt (2018) Steinberg (2008)

Can an early-onset offender desist?

Yes. Longitudinal work documents cessation or reduction even among people with early recorded offending. Developmental supports, family involvement, and prosocial peer relationships may matter, though no one variable guarantees a favorable outcome. van Domburgh et al. (2009)

Is life-course-persistent offending a diagnosis?

No. It describes a pattern of behavior across observation periods in developmental research. Clinical diagnoses and legal determinations require distinct criteria and evaluations.

What is the difference between developmental risk and recidivism risk?

Developmental studies examine associations between earlier life conditions and later offending, often in community cohorts. Recidivism research studies subsequent recorded offending within a defined justice-involved population and follow-up period. The base rates, selection processes, and relevant prediction tools differ.

Does a risk score establish that someone will commit a crime?

No. Even validated scores have errors, depend on event prevalence and population, and do not provide individualized certainty. They also do not establish guilt for a past alleged offense.

Which interventions work best?

Evidence varies by population and comparison services. Some multisystemic or family-focused interventions show benefits on certain outcomes, but modern reviews caution against assuming universal reductions in delinquency. Some alternatives to formal juvenile processing have better results than routine processing in the studied samples. Hunkin et al. (2025) Petrosino et al. (2010)

Can developmental psychology determine criminal responsibility?

Psychology can describe developmental mechanisms, needs, and relevant functional capacities in an authorized assessment. Criminal responsibility, adjudication, sentencing, and the admissibility of testimony are legal questions that vary by jurisdiction.

Conclusion

Developmental pathways to offending describe changing relationships among behavior, social opportunity, and life conditions across childhood, adolescence, and adulthood. Research supports multiple associated risk and protective domains, typical age-linked patterns, meaningful differences in duration and seriousness, and the possibility of change. It also documents weak points in individualized long-range prediction. Scientific and forensic integrity require attention to evidence quality, developmental context, human rights, and a consistent separation between psychological inference and legal findings. Ayano et al. (2024) Jolliffe et al. (2017) National Research Council (2013)

References

Ayano, G., et al. (2024). Risk and protective factors of youth crime: An umbrella review of systematic reviews and meta-analyses. Clinical Psychology Review, 113, 102479. Ayano et al. (2024)

Campbell Collaboration. (2021). Multisystemic therapy for social, emotional, and behavioural problems in youth aged 10 to 17: An updated systematic review. Campbell MST review

Centers for Disease Control and Prevention. (2024). Risk and protective factors: Youth violence prevention. CDC (2024)

Fairchild, G., van Goozen, S. H. M., Calder, A. J., & Goodyer, I. M. (2013). Research review: Evaluating and reformulating the developmental taxonomic theory of antisocial behaviour. Fairchild et al. (2013)

Farrington, D. P., Ttofi, M. M., & Coid, J. W. (2009). Development of adolescence-limited, late-onset, and persistent offenders from age 8 to age 48. Aggressive Behavior, 35, 150–163. Farrington et al. (2009)

Hunkin, H., et al. (2025). Systematic review and meta-analysis: Multisystemic therapy and functional family therapy targeting antisocial behavior in adolescence. Journal of the American Academy of Child & Adolescent Psychiatry, 64, 427–446. Hunkin et al. (2025)

Jolliffe, D., Farrington, D. P., Piquero, A. R., Loeber, R., & Hill, K. G. (2017). Systematic review of early risk factors for life-course-persistent, adolescence-limited, and late-onset offenders in prospective longitudinal studies. Aggression and Violent Behavior. Jolliffe et al. (2017)

Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674–701. Moffitt (1993)

Moffitt, T. E. (2018). Male antisocial behaviour in adolescence and beyond. Nature Human Behaviour, 2, 177–186. Moffitt (2018)

National Research Council. (2013). Reforming juvenile justice: A developmental approach. National Academies Press. National Research Council (2013)

Office of Juvenile Justice and Delinquency Prevention. (n.d.). Statistical Briefing Book: Law enforcement and youth crime. OJJDP statistics

Petrosino, A., Turpin-Petrosino, C., & Guckenburg, S. (2010). Formal system processing of juveniles: Effects on delinquency. Campbell Systematic Reviews, 6, 1–88. Petrosino et al. (2010)

Steinberg, L. (2008). A social neuroscience perspective on adolescent risk-taking. Developmental Review, 28, 78–106. Steinberg (2008)

van Domburgh, L., Loeber, R., Bezemer, D., Stallings, R., & Stouthamer-Loeber, M. (2009). Childhood predictors of desistance and level of persistence in offending in early onset offenders. Journal of Abnormal Child Psychology, 37, 967–980. van Domburgh et al. (2009)

Wasserman, G. A., et al. (2003). Risk and protective factors of child delinquency. Office of Juvenile Justice and Delinquency Prevention. Wasserman et al. (2003)

Psychological Encyclopedia