Age and Crime: Why Offending Changes Across the Lifespan

4 days ago

16 min read

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

Ukrainian Psychological Hub · October 11, 2026 · Forensic & Legal Psychology

Age is one of the strongest population-level correlates of recorded criminal offending, yet an age–crime curve cannot tell us whether any particular person will commit a crime. Many U.S. datasets show offending rising through adolescence, reaching relatively high levels in the teenage years or early adulthood, and declining thereafter. The shape, timing, and steepness of that pattern depend on the offense, country, historical period, measurement method, and social environment. An important 2025 cross-national and historical review shows why the familiar sharp teenage peak should never be treated as a universal biological law.

The question “Why does crime change with age?” therefore calls for several levels of explanation: developmental changes in decision-making; relationships, responsibilities, and opportunities; the distribution of exposure to peers and institutions; histories of prior behavior; and the way police, courts, surveys, and researchers measure crime. These explanations operate together. Psychological development matters, and the available evidence also demonstrates substantial social and historical variation.

What Is the Age–Crime Curve?

The age–crime curve is a description of how an aggregate measure of criminal offending varies by age in a specified population and period. Depending on the study, that measure may be the number of arrests per 100,000 people, the proportion reporting at least one offense, the frequency of self-reported acts, convictions, or incidents known to police. Those outcomes answer different questions. The curve itself is a statistical pattern, rather than a diagnosis, a legal standard, or an explanation of an individual person’s conduct.

In an influential review, Farrington (1986) emphasized that the familiar aggregate peak can largely reflect changes in prevalence—the share of people who offend—rather than a uniform rise and fall in the number of offenses committed by every active offender. The U.S. National Institute of Justice describes a commonly observed increase from late childhood, with a peak roughly between ages 15 and 19 and a subsequent decline into the early twenties in certain Western datasets. That is a summary of particular research traditions, and its scope must be read alongside the NIJ discussion and the newer international literature.

A peak in the curve does not mean most teenagers are criminals

An elevated rate in one age group is compatible with most people in that group having no arrest and with considerable differences among people who do report offending. Some early theories treated adolescent rule-breaking as nearly universal. But Steffensmeier, Slepicka, and Schwartz (2025) reviewed cross-national self-report and historical evidence and found that the prevalence of teenage offending was often far below such universalizing projections. The size and position of the apparent peak must be established for a given dataset.

At What Age Is Crime Most Common?

There is no single scientifically valid peak age for all crime. In many contemporary U.S. arrest and self-report series, common street and property offending is disproportionately concentrated in adolescence and young adulthood. Violence may reach its highest levels later than property offending, while offenses that require specialized access, occupational position, sustained planning, or organizational opportunity may have different age profiles. The National Institute of Justice explicitly notes differences between violent and property-crime curves (NIJ, 2014).

One must also specify whether a graph shows the age of a suspect at arrest, the age when an alleged act occurred, the age at conviction, or the age at release from custody. Arrest and conviction are justice-system events influenced by detection, enforcement priorities, reporting, and case processing. They are imperfect proxies for acts of offending. A graph reporting fewer arrests among older adults does not establish that every kind of illegal conduct disappears with age.

How Does Offending Change Across the Lifespan?

Childhood: early difficulties, diverse pathways

Some behaviors relevant to later offending—such as persistent rule violations, impulsive conduct, exposure to violence, chronic family adversity, or difficulties at school—can be studied in childhood. These are observations and possible risk factors, not criminal diagnoses. Most children with any one of these characteristics will not follow the same developmental path, and childhood behavior should not be converted into a prediction of an adult criminal identity. Developmental theories emphasize interactions among temperament, cognitive regulation, caregiving, peers, neighborhood resources, and institutions (Moffitt, 1993).

Adolescence: changing social worlds

Adolescence often brings greater autonomy, more time in peer groups, new mobility, exposure to parties and public space, heightened importance of status, and experimentation with adult roles. Opportunities to violate rules can expand as external supervision changes. Emotional salience and sensitivity to immediate rewards may also affect some decisions, especially in socially charged situations. These features help explain risk-taking patterns, but risk-taking in a laboratory cannot be equated with criminal offending.

In a controlled experiment involving 306 participants, Gardner and Steinberg (2005) found stronger peer-related increases in risk-taking among adolescents and youths than among older adults. This supports a plausible developmental mechanism. It does not prove that peer presence caused any particular recorded offense, and the magnitude of laboratory effects should not be imported uncritically into policing or court decisions.

Emerging adulthood: transitions may be uneven

The transition through the late teens and twenties is often a period of changing schooling, employment, housing, partnerships, and peer networks. Some people reduce offending rapidly after adolescence; others do so gradually or after subsequent changes in circumstances. Some begin offending in adulthood or engage primarily in offenses available in later institutional roles. Sweeten, Piquero, and Steinberg (2013) examined explanations for age-related offending patterns and illustrate why changes in psychosocial functioning and social circumstances require empirical testing rather than assumption.

Middle adulthood: adaptation and changing opportunities

For many people, recorded offending becomes less prevalent across adulthood. The social processes can include changes in peer associations, work routines, intimate partnerships, community ties, daily activities, goals, and interpretations of prior experience. The classic longitudinal work of Laub and Sampson (1993) treated adult turning points and social bonds as possible components of change. A job or relationship is not automatically protective: its quality, stability, meaning, and connection to a person’s actual circumstances matter.

Later adulthood: lower average rates with specific exceptions

Many reoffending studies find lower rates of rearrest among people released at older ages. For example, a U.S. Sentencing Commission analysis of a federal release cohort reported substantial differences between the youngest and oldest release groups. These figures concern a selected population, with a particular definition of recidivism and follow-up period; they are not national estimates of all offenses, and rearrest is not the same as proven guilt. Older age does not establish individual harmlessness, just as younger age does not establish individual dangerousness.

Why Might the Age–Crime Relationship Change?

Development of judgment and self-regulation

As people move through adolescence and young adulthood, average patterns of future orientation, planning, emotion regulation, resistance to peer pressure, and the weighting of consequences can change. The National Research Council’s developmental review identifies these domains as relevant to juvenile justice. Such capacities vary widely at every age, develop along different timelines, and depend on the demands of the situation. Biological development contributes to the discussion, while a brain-based story alone cannot account for the distribution of crime across societies.

A direct challenge to neurobiological universality appears in Steffensmeier and Schwartz (2025), which evaluates international evidence against claims that a fixed imbalance between reward systems and cognitive-control systems produces the same criminal-age pattern everywhere. Developmental neuroscience should therefore inform a multilevel explanation, rather than supply a predetermined graph.

Peers, imitation, and exposure to situations

Friendship networks influence norms, opportunities, co-offending, and perceptions of consequences. At the same time, people select friends and settings partly on the basis of their own prior behavior. Observed peer similarity can reflect influence, selection, and shared environments. Longitudinal research is needed to disentangle these processes. Among adolescents adjudicated for serious offenses, Leverso, Bielby, and Hoelter (2015) found that some peer-related predictors differed by age, while impulsiveness remained associated with subsequent self-reported offending.

Opportunity and routine activities

A person’s changing daily routine alters where they spend time, whom they encounter, which rules apply, and what offenses are practically available. Changes in school attendance, leisure settings, travel, employment, family life, or access to money can influence observed patterns independently of changes in personality. Crime categories are not interchangeable: an offense requiring a workplace position has a different opportunity structure from adolescent peer-group vandalism. These differences help explain why the age–crime relationship varies by offense type.

Social bonds, identity, and future expectations

Stable commitments may make offending less attractive or less compatible with a person’s goals, especially when people experience meaningful belonging, legitimate opportunities, and a credible route forward. Personal narratives about becoming a different kind of person may accompany reduced offending, yet retrospective accounts can be influenced by memory and self-presentation. Adult change may result from accumulating experiences rather than a single dramatic turning point. Sampson and Laub (1992) integrate these life-course relationships in a foundational review.

Social context and historical change

Schools, labor markets, family structures, housing patterns, transportation, policing, youth culture, alcohol and drug markets, and technology differ substantially across historical periods and countries. These arrangements change the age at which people encounter different opportunities and constraints. Research published in 2025 found postadolescent peaks and broader adult distributions in many non-U.S. settings (Steffensmeier et al., 2025). The implication is direct: there is an empirically important relationship between age and crime, but there is no universally fixed numerical age at which it must peak.

What Do the Major Developmental Theories Explain?

Moffitt’s developmental taxonomy

Moffitt (1993) proposed a distinction between an adolescence-limited pathway, in which antisocial behavior is concentrated around adolescence, and a life-course-persistent pathway characterized by earlier and more sustained difficulties. The model connects trajectories to different developmental and environmental processes. The theory was influential because it asked how an aggregate adolescent increase could coexist with stability among a smaller group. Moffitt’s 2018 review revisits evidence for the framework.

These categories are research concepts, not medical diagnoses or legally recognized offender types. Researchers differ over how many developmental trajectory groups the data actually support, and estimates change with samples, measures, follow-up periods, and statistical modeling. An individual who offends in adolescence cannot be reliably classified as permanently persistent from the fact of youth offending alone.

Life-course social control and turning points

Sampson and Laub’s approach emphasizes continuity and change in social ties, institutions, and experiences across the life course. Adult opportunities and attachments can redirect trajectories, even when earlier histories were serious. Their longitudinal work provides an important counterweight to any assumption that early offending fixes an immutable adult future (Laub & Sampson, 1993). Causal interpretation remains challenging because people who obtain stable work or relationships may already differ in relevant ways from those who do not.

Age-graded opportunities and broader societal explanations

Other explanations focus on how societies distribute autonomy, responsibilities, labor, inequality, surveillance, and opportunities across ages. This approach predicts variation in age profiles across social contexts, historical periods, and crime types. The 2025 Annual Review of Criminology synthesis places such variation at the center of the evidence. It also cautions against treating the U.S. adolescent peak as the necessary pattern for every society.

Age, Onset, Persistence, and Desistance Are Different Questions

Age of onset is the age at which a person first engages in an observed or recorded behavior. Persistence concerns continuation across periods. Desistance refers to the process through which offending diminishes or ceases; studies operationalize it with different observation windows and outcomes. Recidivism generally refers to a new measured justice-system event after a prior sanction or intervention, but definitions vary between rearrest, reconviction, and reincarceration. A population age–crime curve cannot resolve these individual career questions.

A prospective study of 1,119 serious male adolescent offenders identified heterogeneous self-reported trajectories during the three years after court involvement and found that background characteristics did not cleanly separate some persistent and desisting groups (Mulvey and colleagues’ Pathways to Desistance analysis). This illustrates both the value of following individuals over time and the limits of labels that imply a certain future. The detailed mechanisms of desistance, rehabilitation, and recidivism prediction belong to their own forensic research questions.

How Do Researchers Measure Age and Crime?

Arrests and official records

Official police and court records have known dates, legal classifications, and administrative coverage, but they capture behavior that reached the attention of justice institutions. Reporting decisions, detection, discretionary enforcement, case processing, institutional practices, and neighborhood conditions affect counts. Some offense categories have low reporting or detection rates. An arrest is an allegation or procedural event, not proof that the person committed the offense. Conviction records answer a different question again.

Self-report surveys and longitudinal studies

Self-report research can capture acts absent from official records, including minor offenses, but it is affected by recall, honesty, questionnaire wording, survey participation, and the breadth of behaviors included. Longitudinal panels track changes within individuals and can assess whether a rise or decline reflects behavioral changes rather than different people entering a cross-sectional sample. Attrition, incarceration, mortality, migration, and time spent in the community require explicit consideration.

Prevalence, frequency, and crime type

Prevalence means the proportion of people with at least one specified event during a period. Frequency refers to the number of events per person or active offender. Severity and offense mix are further dimensions. A decline in the percentage of people reporting a crime can coexist with a stable frequency among a smaller active group. Farrington’s 1986 analysis makes this distinction central to interpreting age–crime relationships.

Age, period, and birth-cohort effects

An apparent age difference may reflect growing older, conditions prevailing during the calendar period, differences between birth cohorts, or interactions among them. For example, a cohort exposed to different schooling, policing, economic conditions, or technology may show distinctive patterns at particular ages. Age–period–cohort models require strong assumptions and careful interpretation. A published analysis of U.S. robbery, aggravated assault, and homicide arrest patterns from 1960 to 2014 reports meaningful cohort differences and changes within cohorts. Its findings are about the analyzed arrest statistics, not a universal psychological age effect.

Does the Age–Crime Curve Differ by Gender, Offense, and Society?

Yes. Studies frequently report different overall rates and offense mixes among men and women, along with changes in the shape of age patterns. Those differences reflect intertwined social exposure, opportunity, reporting, legal processing, and development. They cannot be reduced to one psychological trait or biology-based explanation. Societal expectations about gender and age may also influence where individuals spend time and which behaviors become visible to authorities. Detailed gender-specific offending explanations require their own evidence base.

Cross-national comparisons are equally important. The 2025 historical and international review finds considerable variation in when curves peak and how much recorded or reported crime occurs in adulthood. Differences in legal definitions, record systems, data quality, sample composition, and reporting behavior complicate simple comparisons. Even within one society, property, violence, fraud, cybercrime, and regulatory offenses may follow distinct age distributions.

Does Criminal Behavior Automatically Stop as People Get Older?

No. A declining average is compatible with continued offending by some people, onset later in life, temporary interruption, and changes in offense types. The statistical curve cannot supply a personal timetable for desistance. It also cannot decide whether a person has changed, whether a particular intervention worked, or whether they will reoffend. These are separate empirical and forensic questions that demand individual evidence and contextual assessment.

The U.S. Sentencing Commission’s 2017 federal recidivism report found rearrest during eight years of follow-up among 13.4% of people aged 65 or older at release, compared with 67.6% of those younger than 21. The sample consisted of federal offenders released or placed on probation in 2005; its definition of recidivism was tied to new recorded events, especially rearrest. These numbers should never be applied as a personalized probability to someone with a different history, jurisdiction, offense type, follow-up duration, or exposure to the community.

Can Age Predict Whether an Individual Will Commit a Crime?

Age may help estimate baseline rates within a specified group, but prediction requires a defined outcome, population, time horizon, and validated method. A factor correlated with offending is not a cause of an offense, and a higher group rate does not establish that a particular person committed a crime or is dangerous. The base rate matters: when the outcome is rare, even a seemingly strong statistical classifier may produce many false positives. False negatives are also possible. Age alone has poor conceptual coverage of individual context.

A defensible evaluation must ask whether a model discriminates between outcomes, whether predicted probabilities are calibrated to observed outcomes, whether it works for people like the person being assessed, and whether errors differ across groups or settings. Predictive performance from one cohort cannot simply be transferred to another historical period, jurisdiction, gender, or offense category. No age-based score proves guilt, establishes a diagnosis, determines competency or criminal responsibility, or replaces the applicable legal standard.

Forensic psychologists must explain the limits of their evidence, and courts determine legal findings under the law of the jurisdiction. The APA Specialty Guidelines for Forensic Psychology provide professional guidance on competence, evidence, and communication in legal contexts; they are professional guidelines rather than a uniform criminal law.

What Does the Research Mean for Juvenile Justice?

A developmental approach supports careful consideration of adolescents’ evolving capacities, susceptibility to context, and substantial potential for change. The National Research Council’s Reforming Juvenile Justice (2013) recommends aligning juvenile justice practices with developmental science, fair procedures, and evidence-informed responses. Youthful behavior requires assessment of immediate safety and needs without presuming that adolescent offending predicts a permanent adult offending pattern.

In practice, this can motivate developmentally appropriate interviewing, clear explanations of legal processes, attention to education and family circumstances, proportionate responses, and opportunities for sustained community connection. The research does not generate one universal minimum age of responsibility, sentence, or jurisdictional rule. Statutory definitions of juvenile status, transfer to adult court, sentencing, competency, and rehabilitation differ between U.S. states and other countries. Legal outcomes must be determined from applicable law, not from the shape of a research graph.

What Does the Research Mean for Adult Corrections and Reentry?

Correctional and reentry decisions often depend on changes in relevant needs and opportunities over time. Age can provide context, but a complete assessment considers prior conduct, current behavior, circumstances, supervision conditions, intervention history, housing, work, health, support networks, and the precise outcome being evaluated. Some of these variables change with time, while others do not. Risk assessment should be updated when circumstances meaningfully change.

Age-related differences among federal release cohorts, such as those documented by the U.S. Sentencing Commission in 2022, support careful analysis of the population actually studied. They do not mean a person’s release decision follows mechanically from age. Humane and evidence-informed policy attends to individual circumstances, proportionality, rehabilitation, public safety, and law.

Common Myths About Age and Crime

“Every crime peaks at age 18.”

Different offenses, places, periods, and measures yield different peaks. Recent international findings explicitly contradict a globally invariant adolescent maximum (Steffensmeier et al., 2025).

“Teenagers are naturally criminals because their brains are unfinished.”

Development affects behavior in context, and neuroscientific accounts do not convert a developmental stage into criminal identity. Criminal behavior depends on legal definitions, opportunities, experiences, relationships, institutions, and choices. Most developmental variability cannot be reduced to a single brain maturation threshold.

“An early offender will inevitably become a lifelong offender.”

Longitudinal studies show variation in onset, persistence, and change. Developmental categories are imperfect research summaries, and no childhood or adolescent label determines a fixed future (Moffitt, 2018).

“Older people never reoffend.”

Older release cohorts tend to have lower average rearrest rates, but some older people do reoffend. The Sentencing Commission’s federal data demonstrate lower rates, not zero rates.

“Arrest statistics measure everything people actually do.”

Arrest rates depend on reporting, policing, and administrative recording; self-report surveys and longitudinal studies add information but have their own errors. Research findings should specify what was measured.

Frequently Asked Questions

Why does crime often increase during adolescence?

Adolescence frequently brings greater autonomy, peer exposure, new opportunities, and developmental changes in how immediate rewards and consequences are weighed. Those factors help explain some observed patterns, while cultural institutions and enforcement systems influence how the pattern appears in a given dataset.

At what age do most people stop committing crimes?

There is no universal stopping age. In many populations, the proportion engaging in specified offenses declines after adolescence or early adulthood. Individual desistance unfolds differently and cannot be inferred from a single age threshold.

Is the age–crime curve the same in every country?

No. The 2025 cross-national review documents later peaks and broader adult age distributions outside the U.S. in many datasets. Both real social differences and measurement practices matter.

Does the curve prove that adolescent brains cause crime?

No. Developmental features may contribute to some forms of risk-taking, but the curve is descriptive and has multiple competing explanations. Cross-cultural variation challenges a universal one-cause account.

Is the age–crime curve about arrests or actual offending?

It depends on the study. Researchers use arrests, convictions, self-report surveys, or other outcomes. Always read the definition, denominator, jurisdiction, and observation period before comparing curves.

Are age and recidivism the same topic?

No. Age is a characteristic or time coordinate. Recidivism is a new measured justice-system event after a prior event or sanction. Their association depends on the population, offense and definition of recidivism.

Can age alone be used to sentence or release someone?

Age may be legally relevant under some jurisdiction-specific rules and may inform an individualized assessment, but a population crime curve is not a sentencing rule, criminal responsibility standard, or validated stand-alone release instrument. Courts and agencies must follow applicable law and procedure.

Does having a mental health diagnosis explain criminal offending?

A diagnosis is a clinical classification, not a finding of criminal conduct or a prediction of violence. When mental health factors are relevant to an individual legal matter, their role must be evaluated using appropriate clinical, behavioral, contextual, and legal evidence. No diagnosis should be substituted for such analysis.

References

American Psychological Association. (2013). Specialty guidelines for forensic psychology. https://www.apa.org/practice/guidelines/forensic-psychology.html

Farrington, D. P. (1986). Age and crime. Crime and Justice, 7, 189–250. https://doi.org/10.1086/449114

Gardner, M., & Steinberg, L. (2005). Peer influence on risk taking, risk preference, and risky decision making in adolescence and adulthood. Developmental Psychology, 41(4), 625–635. https://doi.org/10.1037/0012-1649.41.4.625

Laub, J. H., & Sampson, R. J. (1993). Turning points in the life course: Why change matters to the study of crime. Criminology, 31(3), 301–325. https://doi.org/10.1111/j.1745-9125.1993.tb01132.x

Leverso, J., Bielby, W., & Hoelter, L. F. (2015). Back on the streets: Maturation and risk factors for recidivism among serious juvenile offenders. Journal of Adolescence, 41, 67–75. https://doi.org/10.1016/j.adolescence.2015.02.008

Lu, Y., & Luo, L. (2021). Cohort variation in U.S. violent crime patterns, 1960–2014: An age–period–cohort analysis. Journal of Quantitative Criminology. https://doi.org/10.1007/s10940-020-09477-3

Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100(4), 674–701. https://doi.org/10.1037/0033-295X.100.4.674

Moffitt, T. E. (2018). Male antisocial behaviour in adolescence and beyond. Nature Human Behaviour, 2, 177–186. https://doi.org/10.1038/s41562-018-0309-4

Mulvey, E. P., Steinberg, L., Piquero, A. R., Besana, M., Fagan, J., Schubert, C., & Cauffman, E. (2010). Trajectories of desistance and continuity in antisocial behavior following court adjudication among serious adolescent offenders. Development and Psychopathology, 22(2), 453–475. https://doi.org/10.1017/S0954579410000179

National Institute of Justice. (2014). From youth justice involvement to young adult offending. https://nij.ojp.gov/topics/articles/youth-justice-involvement-young-adult-offending

National Research Council. (2013). Reforming juvenile justice: A developmental approach. National Academies Press. https://doi.org/10.17226/14685

Sampson, R. J., & Laub, J. H. (1992). Crime and deviance in the life course. Annual Review of Sociology, 18, 63–84. https://doi.org/10.1146/annurev.so.18.080192.000431

Shulman, E. P., Steinberg, L., & Piquero, A. R. (2013). The age-crime curve in adolescence and early adulthood is not due to age differences in economic status. Journal of Youth and Adolescence, 42, 848–860. https://doi.org/10.1007/s10964-013-9950-4

Steffensmeier, D., & Schwartz, J. (2025). Rethinking the age-crime curve: Neurobiological claims, international evidence, and sociocultural alternatives. Journal of Criminal Justice, 99, 102476. https://doi.org/10.1016/j.jcrimjus.2025.102476

Steffensmeier, D., Slepicka, J., & Schwartz, J. (2025). International and historical variation in the age–crime curve. Annual Review of Criminology, 8, 239–268. https://doi.org/10.1146/annurev-criminol-111523-122451

Sweeten, G., Piquero, A. R., & Steinberg, L. (2013). Age and the explanation of crime, revisited. Journal of Youth and Adolescence, 42, 921–938. https://doi.org/10.1007/s10964-013-9926-4

United States Sentencing Commission. (2017). The effects of aging on recidivism among federal offenders. https://www.ussc.gov/research/research-reports/effects-aging-recidivism-among-federal-offenders

United States Sentencing Commission. (2022). Recidivism of federal violent offenders released in 2010. https://www.ussc.gov/research/research-reports/recidivism-federal-violent-offenders-released-2010

Psychological Encyclopedia