Gender and Criminal Offending: Pathways, Risk, and Evidence
Author: Ukrainian Psychological Hub · Published: October 11, 2026 · Editorial Policy
Author: Ukrainian Psychological Hub · Published October 11, 2026 · Editorial Policy
Gender and criminal offending examines how sex-related classifications, gendered social conditions, developmental histories, opportunities, institutions, and legal processes are associated with behavior prohibited by criminal law. Men account for a substantially larger share of recorded serious interpersonal violence in many criminal justice datasets. That observation is an important starting point for research, but an explanation requires separate attention to what people do, how offenses are discovered, who enters a particular dataset, and which experiences alter a person's opportunities and choices over time.
Research also shows considerable overlap within groups, multiple pathways to offending, and substantial variation across offenses and jurisdictions. Gender is a population characteristic or social context, never evidence that a particular person committed a crime. The foundational reviews by Steffensmeier and Allan (1996) and Kruttschnitt (2013) locate crime patterns in relationships among opportunity, social control, social position, institutions, and life circumstances. The wider explanation of criminal offending is developed in Why Do People Commit Crimes? What Psychology Actually Shows.
What does gender and criminal offending mean?
Criminal offending is conduct prohibited by criminal law in a particular jurisdiction. A self-reported unlawful act, an allegation, a police-recorded incident, an arrest, a charge, a conviction, a jail admission, and a return to custody represent different stages or outcomes. Research frequently measures one of these rather than the underlying total of offenses committed. Researchers and readers therefore need to identify which outcome is being counted before interpreting a gender difference.
Sex may be recorded using biological, birth-assigned, legal, or administrative categories. Gender can refer to identity, social roles, norms, institutional expectations, and social relations. Many official studies use binary male/female fields and label them gender without documenting whether the underlying variable measures sex assigned at birth, current identity, or an institutional classification. These categories should be named precisely. The article retains the original terminology when reporting research while identifying important measurement constraints.
A gendered offending pathway is a hypothesized sequence in which experiences, environments, relationships, developmental processes, and available opportunities become associated with unlawful behavior. A pathway is an explanatory framework, subject to testing and revision. It does not mean that people who share an experience will reach the same outcome. A risk factor is an observed characteristic associated with a greater probability of an outcome in a defined population. A protective factor is associated with lower risk or adaptive functioning in a defined context. Neither establishes criminal intent, causation, future behavior, or legal responsibility for one individual.
The evidence: what gender differences in crime statistics show
Official criminal justice statistics
Official records repeatedly show male overrepresentation in serious recorded violence. For example, the Bureau of Justice Statistics' Jail Inmates in 2024 report found that local U.S. jails held approximately 95,200 people recorded as female at midyear 2024, around 14% of the jail population. This is a reliable description of custody at a particular date. It is not a behavioral prevalence estimate: jail populations reflect policing, arrests, pretrial decisions, bail and release policies, sentencing, supervision, and length of stay. The same individual may enter custody more than once within a year.
The FBI's 2024 Reported Crimes in the Nation publication summarizes incidents submitted by participating agencies. Reported offenses do not capture every criminal event experienced in a community. Police practices, victims' decisions to report, local legal definitions, available investigative resources, and incident classification all influence observed rates.
There is consequently no single meaningful percentage for “how much crime men or women commit” without a specified year, country, age range, population, offense, and measurement source. The gap between groups is much more pronounced for some severe forms of violence than for some other offense categories. A comparison that pools homicide, assault, fraud, drug offenses, shoplifting, and supervision violations into a single measure conceals important mechanisms.
Victimization and self-report data
Victimization surveys ask residents about experiences that may never reach police records. Self-report surveys ask participants about their own behavior, including incidents that have not led to arrest. Prospective developmental studies can track individual behavior through changes in age, relationships, and environment. Each design supplies information that official records alone cannot.
Victimization reports may omit events outside their sampling or question design. Self-reports can be affected by recall, stigma, honesty, and sample attrition. Police and court files provide legally important events but depend on who encounters legal systems. Convergence among several sources strengthens a descriptive inference, and disagreement can reveal changes in reporting or enforcement.
Arrests are not identical to changes in violent behavior
A well-known problem is interpreting increases in recorded female assault arrests as direct proof of increases in female violent conduct. A study triangulating arrest, conviction, incarceration, and victim-report trends identified discrepancies between changes in police records and other indicators. Historical enforcement and recording changes helped explain some of the apparent narrowing of the gender gap. Related work comparing sources for girls' violence trends illustrates how the selected indicator changes the story.
These studies address specific historical data and should not automatically settle contemporary trends. Their enduring methodological lesson is that a claim about behavior requires evidence measuring behavior. A change in arrests may reflect changed conduct, police attention, legal rules, reporting practices, or a combination of these influences.
Aggression research contributes one piece of the puzzle
In a meta-analytic review of sex differences in real-world aggression, Archer found average differences in forms of aggression, particularly direct physical aggression. Aggression is a psychological and behavioral construct, and an aggressive act may be lawful, unlawful, minor, severe, defensive, or coercive depending on facts and jurisdiction. Psychological measures of aggression cannot be substituted for estimates of homicide, conviction, or criminal recidivism. The distinction also prevents researchers from interpreting average behavioral differences as a diagnosis or individual prediction.
Why might pathways into criminal offending differ?
Opportunities, institutions, and social organization
Opportunities for unlawful behavior differ with work arrangements, financial resources, local markets, peer networks, family roles, mobility, age, and exposure to policing. A person with access to a particular workplace, online environment, social group, or illicit market encounters different opportunities and constraints than someone in another setting. Gender may help structure such access, yet opportunity remains context-dependent and subject to change.
Social expectations about dominance, toughness, status, retaliation, or public risk-taking may shape behavior in some peer networks. These expectations can help explain particular conflicts or offending environments. They are not characteristics of every man and do not demonstrate that masculinity causes crime. Most men are never involved in serious violent offending, and men who do offend have widely varying histories and circumstances. Gender and crime research is strongest when it specifies a setting, hypothesizes a mechanism, and evaluates the mechanism against alternatives.
Developmental adversity and exposure to violence
Adversity can affect emotional regulation, trust, housing, educational continuity, social support, and access to safe opportunities. Exposure to family violence, inconsistent care, poverty, or victimization can be associated with later behavioral problems in some populations. The systematic review of female pathways to antisocial behavior brings together evidence about multiple individual and contextual vulnerabilities and emphasizes relationships among levels of influence.
The association between adversity and offending requires careful interpretation. Most people exposed to trauma do not commit serious crimes, and many individuals who offend have no single identifiable causal experience. Family adversity may be correlated with neighborhood conditions, service access, substance-related problems, and other influences. Selection into correctional samples can exaggerate how common particular histories appear among all people with similar experiences.
Relationships, coercion, and survival pressures
Studies of some justice-involved women document histories involving coercive intimate relationships, exploitative environments, economic precarity, social isolation, or partner-linked involvement in illicit activity. These patterns help investigators and service providers ask more precise questions about life circumstances. They should not be treated as a standardized biography of women who offend. Women can exercise agency, participate in planning, and commit serious offenses; men can also be coerced or exploited.
Research on gendered pathways to crime explores how relationships, substance use, and social conditions may combine for particular groups. A pathway account is separate from the legal question whether coercion, necessity, duress, intent, or any defense has been established. Such standards are jurisdiction-specific and require evidence about the actual event.
Substance use, mental health, and accumulated disadvantage
Substance-related problems may be linked to offending through multiple routes: impaired judgment during some episodes, involvement in illicit markets, economic pressures, social settings, and difficulties maintaining work or housing. Substance use can also arise alongside adversity without causing a criminal offense. A substance-use disorder or another psychiatric diagnosis does not establish dangerousness or a motive, and cannot by itself explain a particular act.
Justice-involved people may face barriers to appropriate clinical care, especially when releases or transfers interrupt continuity. These issues affect rehabilitation planning and public health. Their presence in a correctional sample should not be used to equate mental illness with criminality. The English Hub's separate clinical article Schizophrenia and Violence: What Research Really Shows addresses the evidence on one diagnosis and violent outcomes under its established scientific ownership.
Biological and developmental influences
Research examines neurodevelopment, inherited characteristics, hormones, stress responses, executive functions, and age-related processes. These may contribute to hypotheses about behavior when studied alongside the environment. Even where average differences are observed, findings rarely identify a single cause of an individual's offending. Biology and social experience can influence one another across development, and the same observable behavior can arise under different circumstances. For a focused examination of biological evidence and its interpretive limits, see Genes, Brain, and Criminal Behavior: What Biological Evidence Can and Cannot Show.
Claims that testosterone, a brain scan, a genetic marker, or a single personality score explains “male crime” exceed what such measures establish. Evaluators should distinguish laboratory associations, clinical symptoms, prospective predictors, criminal-record outcomes, and evidence that actually concerns the offense under investigation. No biological test is a shortcut to guilt, criminal responsibility, or a prediction about one person.
Age, onset, persistence, and desistance
Offending is often studied across the life course because age changes opportunities, relationships, legal exposure, and developmental circumstances. Some people engage in unlawful behavior mainly during adolescence, some have a later onset, some show recurring incidents, and others cease involvement as circumstances change. A trajectory summarizes a pattern over observations; it is not a permanent identity. The neighboring developmental analysis is Age and Crime: Why Offending Changes Across the Lifespan.
A particularly important update is the 2026 systematic review of female offending trajectories by Coles and colleagues. The review identified 26 studies from seven countries, encompassing approximately 92,023 participants with possible cohort overlap. Across studies, identified patterns included early-onset desistance, adolescent-limited offending, later onset, sporadic involvement, and persistent trajectories. Persistent female offending patterns were documented, challenging any assumption that persistent offending is a phenomenon exclusive to men.
The review also identifies limitations, including differences in sample design, follow-up, model specification, and international coverage. Many statistical trajectory models assign people to latent classes because the model seeks patterns in data. Class boundaries can shift with the number of observations, offense definition, and estimation procedure. These classifications are useful for research and program design, yet they should not be mistaken for natural kinds or treated as individual destiny.
Adolescence and gendered institutional contact
Adolescent offending may arise with peer conflict, social exclusion, unstable housing, substance exposure, and opportunities for rule breaking. Some episodes are temporary, and others signal recurrent needs requiring proportionate services. Girls and boys may also experience different family, school, and policing responses to similar conduct. Researchers should examine institutional pathways as carefully as the underlying behavior.
Desistance and protective conditions
Desistance describes a reduction or ending of offending over time. Potentially supportive circumstances include stable housing, safe relationships, employment, education, access to treatment where indicated, reduced exposure to coercive environments, and meaningful roles in community life. These variables can be correlated with improvement without establishing which one caused it. Effective services must assess circumstances at the individual level and evaluate change across an appropriate follow-up period.
Gender and different types of offenses
Serious violence, coercive relationships, sexual offenses, financial misconduct, cyber-enabled wrongdoing, drug-market offenses, and property crimes involve different opportunities and mechanisms. A general finding about the distribution of physical violence cannot automatically explain fraud. Similarly, an arrest-based difference in property offending may partly reflect which acts are detected and prosecuted.
Serious interpersonal violence
Men are disproportionately represented in serious recorded interpersonal violence across many settings, but the base rates, severity definitions, and contexts vary. Reliable analysis should disaggregate homicide, threats, robbery, assault causing injury, repeat violence, and violence during criminal-market disputes. Victim sex, weapon involvement, incident location, and multiple-perpetrator events can further alter interpretation.
Intimate-partner violence and coercive control
Counts of discrete physical acts, repeated intimidation, threats, coercive control, injuries, sexual violence, and homicide measure different dimensions of intimate-partner harm. Simple comparisons of “who used violence” can be misleading if severity and sustained coercion are omitted. The relevant evidence includes context, chronology, documentation, and each person's safety. A forensic assessment must not infer responsibility solely from gender. The English Hub's Childhood Exposure to Domestic Violence: Effects, Risk, and Resilience examines childhood psychological consequences separately from this article's focus on offending patterns.
Financial, occupational, drug-related, and digital offenses
Financial opportunity differs with occupational roles, access to accounts or assets, institutional control, and investigative resources. Drug-law enforcement is strongly affected by local regulation, market structure, and policing priorities. Digital offending may depend on access, technical capabilities, peer groups, and platform design. Evidence about any of these categories must be sought in offense-specific datasets, rather than mechanically extrapolated from violence statistics.
Criminalization and custodial decisions
Differences in laws, arrest practices, pretrial detention, supervision, and sentence length shape the composition of custody populations. Researchers must distinguish the proportion of people arrested, the proportion admitted, and the proportion held on a census date. These outcomes can move differently. Comparisons across countries require consistent definitions because acts classified as criminal and penalties imposed can vary greatly.
Are women who offend psychologically different from men who offend?
Some justice-involved samples reveal differences in the distribution of prior victimization, relational difficulties, parental obligations, substance-related needs, and economic instability. Such findings can inform screening and service planning. They do not demonstrate that every woman has a trauma history, that men lack these experiences, or that gender determines criminal intent.
Within-group variation is substantial. Two women who have committed the same legally defined offense may differ profoundly in motives, conduct, coercive circumstances, earlier convictions, substance use, health needs, and opportunities for desistance. The same is true for two men. Individualized explanations require individualized evidence.
Personality constructs, antisocial behavior, aggression, psychiatric diagnosis, and offending overlap imperfectly. A clinical diagnosis reflects criteria relevant to mental health; a standardized assessment score describes measurement on an instrument; a risk opinion concerns a probabilistic future outcome; a forensic opinion addresses an instructed question; a court finding concerns legally established matters. These concepts should be connected only where appropriate evidence and the applicable legal standard permit the connection.
Transgender and nonbinary people: measurement and justice-system exposure
Binary administrative sex records frequently fail to describe transgender and nonbinary people's gender identities or the context in which classifications were assigned. A prison dataset may represent institutional placement, legal records, or data-entry rules rather than contemporary gender identity. Analyses that merge these measures can produce incorrect comparisons.
Scholarship on the transgender criminal legal system nexus examines criminalization, policing, exposure to victimization, and institutional treatment. Differences in contact with the justice system can arise from differential policing, insecurity, discrimination, or reporting practices. Such evidence cannot establish an inherent propensity to offend. Privacy, informed recording practices, and personal safety should be incorporated into assessment procedures.
Researchers need transparent definitions, representative sampling, comparable outcome measures, and sufficient statistical precision before drawing group comparisons. In many datasets those conditions are unmet. A gender identity is neither a risk score nor evidence of a crime.
Gender and recidivism risk assessment
A forensic risk assessment concerns a specific future outcome in a stated population and timeframe. General rearrest, violent reconviction, institutional aggression, and supervision failure are different outcomes. Assessments may use past behavior, changing needs, social conditions, and structured instruments. A validated score can support decision-making under specified conditions, but it cannot prove an offense, establish a legal fact, or yield certainty about future conduct.
Predictive validity and calibration
Discrimination evaluates how well a method tends to rank people who experience a specified outcome above people who do not. Area under the receiver operating characteristic curve is one commonly reported measure. Calibration asks a different question: whether assigned probabilities correspond to observed outcome rates. A tool can rank people moderately well while overstating or understating their absolute risk.
Base rates are crucial. A classification with apparently respectable sensitivity and specificity can produce many false positives when the outcome is uncommon. False positives classify people as higher risk although the defined outcome does not occur; false negatives fail to identify some outcomes that later occur. Changing the threshold affects both types of error, the people affected, and the practical consequences of the classification.
Do instruments work equally well for women?
A 2015 systematic review of predictive validity in female offenders evaluated 15 studies examining nine tools, identifying variations in accuracy and study quality. The fact that a tool predicts outcomes in one validation sample cannot establish equivalent calibration in every jurisdiction, historical period, or subgroup.
A 2024 forensic psychiatric study of 525 discharged women compared multiple risk instruments with later outcomes. Some showed moderate-to-good discrimination, while limited sensitivity meant that scores were unsuitable as the sole basis for discharge decisions. Its population consisted of women discharged from specialized forensic psychiatric facilities and must not be treated as representative of all women, people with mental health diagnoses, or female offenders.
A broader preregistered meta-analysis comparing structured tools with unstructured judgments reported higher average predictive validity for structured tools across directly comparative studies. That does not imply accurate individual prediction, error-free use, or freedom from fairness concerns. Differences in selection, validation samples, measurement, and decisions remain consequential.
Fairness, jurisdiction, and practical safeguards
Assessment quality requires attention to validation setting, event base rate, subgroup discrimination, calibration, cultural and demographic fairness, false-positive and false-negative costs, and consequences of the decision. An instrument validated among adult men entering prison or women in a forensic hospital requires renewed examination before use in a different setting. The National Institute of Justice's Guidelines for Post-Sentencing Risk Assessment emphasize fairness, efficacy, transparency, and responsible communication.
Gender-aware assessment means testing the relevance of particular information and considering validated differences when evidence supports them. It does not mean using stereotyped assumptions as a substitute for evidence. Risk assessments should be tied to a purpose such as identifying service needs or planning supervision, and their limits must be explicitly reported.
Rehabilitation, gender-responsive interventions, and reentry
Gender-responsive rehabilitation adapts services to empirically supported needs and circumstances that may be especially salient for a particular group. Needs can include substance-related care, safe housing, victimization-related symptoms, childcare arrangements, economic resources, education, relationship safety, and continuity of medical services. Individual need should determine selection; a program's gendered branding is no guarantee of effectiveness.
What the intervention evidence shows
A 2016 meta-analysis by Gobeil, Blanchette, and Stewart synthesized 37 studies and almost 22,000 justice-involved women. Intervention participation was associated with better community outcomes. Across the complete set of studies, gender-informed and gender-neutral programs had comparable effects. In a higher-quality subset, gender-informed programs were more favorable for recidivism reduction. This is meaningful support for carefully designed, evaluated tailoring, with remaining methodological uncertainty.
The 2022 systematic review of offense-specific interventions for women identified important evidence gaps and uneven evaluation quality. A 2025 review of gender-responsive treatments considers promising programs and unresolved implementation issues. The official National Institute of Justice CrimeSolutions evidence profile provides a further structured appraisal of this field.
Program evaluation must define the outcome before declaring success: rearrest, reconviction, return to custody, technical violations, symptoms, employment, housing stability, or victim safety. Improvements in one outcome should not be represented as proven improvements in all others. Duration of follow-up, allocation procedures, comparable control groups, and implementation quality all affect interpretation.
Trauma-informed services and personal accountability
Trauma-informed assessment involves understanding documented experiences and responding to current needs while preserving appropriate boundaries. Adverse experiences may illuminate vulnerabilities, treatment requirements, or relevant contexts. They do not determine criminal responsibility or guarantee that rehabilitation will succeed. Ethical programs attend both to the rights and needs of justice-involved participants and to public and victim safety.
Women and reentry
The National Institute of Justice overview of women and reentry discusses barriers and potential supports after release. Housing insecurity, interrupted healthcare, disrupted family relationships, limited income, and exposure to unsafe environments can be decisive practical concerns. Some barriers vary by gender or caregiving role; others are shared across justice-involved populations. Reentry plans should address demonstrated needs with measurable follow-up rather than infer need from a sex category alone.
Forensic, courtroom, and public-policy implications
The American Psychological Association's Specialty Guidelines for Forensic Psychology provide a professional framework for competence, evidence-based practices, appropriate methods, accuracy, and respect for affected persons. The American Psychology-Law Society maintains relevant professional guidance information. Applying gender research responsibly requires identifying which facts concern an individual's behavior and which are statistical findings about populations.
Criminal responsibility, competence to stand trial, legal capacity, sentencing, and admissibility are defined by law. Definitions and procedural standards differ across U.S. jurisdictions and other countries. Psychological findings may inform some legal questions when appropriate, but neither sex nor gender nor an assessment score establishes a legal finding.
Evaluators should state how gender-related data were obtained, whether research reflects the relevant population, what alternatives were considered, and what uncertainty remains. Gender stereotypes can influence interpretations of credibility, fear, aggression, care, vulnerability, and responsibility. Such influences merit direct examination in research and procedural design. An actual legal decision must rely on the admissible evidence and governing law.
For public policy, accurate gender analysis requires audits of how incidents are recorded, arrests occur, detention is imposed, services are allocated, and outcomes are evaluated. Policymakers can use population evidence to design fair prevention and rehabilitation systems while preserving the individualized protections of legal procedure.
Common myths and misconceptions
Does being male mean a person is violent or criminal?
No. Average differences in population-level rates say nothing definitive about the behavior of a particular man. Most men do not commit serious violent offenses. Valid behavioral investigation requires evidence about actual actions and circumstances.
Are women who offend always acting under coercion?
No. Coercive or exploitative relationships are documented in some pathways and may be important to assess. They are neither universal experiences nor automatic legal defenses. An individual's agency, circumstances, and the applicable elements of an offense must be established from evidence.
Does an increase in arrests prove greater female violence?
No. Arrests are affected by behavior, reporting, enforcement, and classification. Behavioral trends require corroboration across several measurement systems.
Do women lack persistent offending trajectories?
No. The 2026 systematic review documents heterogeneous trajectories, including persistent patterns, alongside sporadic and limited-duration trajectories. A statistical trajectory is still not a fixed individual identity.
Can a forensic tool determine whether a person will reoffend?
No tool provides certainty. Structured methods can improve probabilistic discrimination in relevant samples but require calibration, attention to error rates, and careful use. Scores do not establish guilt or a future legal fact.
Does a mental health diagnosis explain criminal conduct?
A diagnosis describes clinical phenomena and may indicate treatment needs. It does not by itself establish the cause of a crime, its legal elements, or a person's dangerousness.
Frequently asked questions
Why is the gender gap especially large in serious recorded violence?
Research identifies interacting processes involving social settings, opportunities, behavioral development, peer environments, and institutions. Severity and offense definitions matter, and no single explanation covers every form of violence.
What is the difference between a gender difference and a gendered pathway?
A gender difference is an observed contrast between classified groups on a defined outcome. A gendered pathway proposes how experiences and processes contribute to an outcome across time. A contrast by itself does not identify its cause.
How common is female criminal offending?
Any estimate must specify the offense, year, jurisdiction, ages, and whether the data concern behavior, arrests, convictions, or custody. The 14% share of U.S. local jail inmates recorded as female in midyear 2024 is a custody statistic rather than an estimate of all offending.
Are women becoming as violent as men?
Historical arrest trends cannot settle that question on their own. Valid current comparisons need matched behavioral outcomes and consistent definitions. Some earlier claims of convergence reflected differences in police recording rather than parallel changes in serious violence.
Is trauma the primary explanation for women's offending?
Trauma can be relevant in specific pathways but does not supply a comprehensive cause, prediction, or legal defense. Multiple combinations of circumstances and opportunities are associated with offending.
Are male-designed risk assessment instruments appropriate for women?
Their use requires validation for the outcome and setting. Several instruments demonstrate useful predictive performance in particular female samples, but error rates, calibration, clinical context, and fairness remain important.
Do gender-responsive rehabilitation programs work?
Some evaluated programs show beneficial outcomes, with a higher-quality subset of meta-analytic evidence favoring gender-informed interventions. Benefits depend on program content, population, implementation, and outcome definition.
Can sex or gender establish criminal responsibility or legal competence?
No. These are jurisdiction-dependent legal findings based on applicable standards and case-specific evidence. Group-level research cannot substitute for the legal inquiry.
How should transgender and nonbinary populations be represented in offending research?
Researchers should transparently define sex and gender measures, protect confidentiality, and avoid drawing inferences across incompatible administrative datasets. Unequal legal-system contact cannot establish intrinsic criminal propensity.
What makes gender-and-crime research scientifically strong?
It identifies the offense and outcome, distinguishes behavior from justice-system contact, uses comparable groups and longitudinal designs where relevant, triangulates independent sources, reports uncertainty, and tests alternative explanations.
Conclusion
The evidence on gender and criminal offending demonstrates substantial sex- and gender-classified differences in some recorded criminal outcomes, particularly serious violence, alongside heterogeneous individual pathways and institutional measurement effects. Explanatory research should integrate development, exposure, social conditions, opportunity, and justice-system processes. Effective application requires clearly defined outcomes, sound instruments, reliable evaluation, and respect for the difference between group statistics and individual evidence.
Gender-aware forensic psychology is strongest when it produces better questions, more appropriately tailored services, transparent uncertainty, and fair legal processes. Every clinical opinion, assessment result, risk estimate, and legal finding must retain its own meaning and evidentiary standard.
Related Articles
References
American Psychological Association. (2013). Specialty guidelines for forensic psychology. American Psychologist, 68, 7–19. https://doi.org/10.1037/a0029889.
American Psychology-Law Society. (2026). Specialty guidelines for forensic psychology: Current extension. Professional guidance.
Archer, J. (2004). Sex differences in aggression in real-world settings: A meta-analytic review. Review of General Psychology, 8, 291–322. https://doi.org/10.1037/1089-2680.8.4.291.
Bureau of Justice Statistics. (2026). Jail inmates in 2024: Statistical tables. Official report.
Coles, C., Wickersham, A., Lankester, M., Jolliffe, D., Blackwood, N., & Dickson, H. (2026). Female offending trajectories: A systematic review. Journal of Developmental and Life-Course Criminology, 12, Article 8. https://doi.org/10.1007/s40865-026-00295-y.
Dudeck, M., et al. (2024). Evaluation of whether commonly used risk assessment tools are applicable to women in forensic psychiatric institutions. Comprehensive Psychiatry, 135, 152528. https://doi.org/10.1016/j.comppsych.2024.152528.
Federal Bureau of Investigation. (2025). FBI releases 2024 Reported Crimes in the Nation statistics. Official release.
Galway, R., Swales, M. A., & Wane, J. (2022). The efficacy of offence-specific interventions in reducing risk and recidivism in women: A systematic review. The Journal of Forensic Psychiatry & Psychology, 33, 291–322. https://doi.org/10.1080/14789949.2022.2053185
Geraghty, K. A., & Woodhams, J. (2015). The predictive validity of risk assessment tools for female offenders: A systematic review. Aggression and Violent Behavior. https://doi.org/10.1016/j.avb.2015.01.002.
Gobeil, R., Blanchette, K., & Stewart, L. (2016). A meta-analytic review of correctional interventions for women offenders: Gender-neutral versus gender-informed approaches. Criminal Justice and Behavior, 43, 301–322. https://doi.org/10.1177/0093854815621100.
Javdani, S., Sadeh, N., & Verona, E. (2011). Expanding our lens: Female pathways to antisocial behavior in adolescence and adulthood. Clinical Psychology Review, 31, 1324–1348. https://doi.org/10.1016/j.cpr.2011.09.002
Jenness, V., & Rowland, A. (2024). The structure and operation of the transgender criminal legal system nexus in the United States. Annual Review of Criminology, 7, 283–309. https://doi.org/10.1146/annurev-criminol-022222-040947
Kruttschnitt, C. (2013). Gender and crime. Annual Review of Sociology, 39, 291–308. https://doi.org/10.1146/annurev-soc-071312-145605.
National Institute of Justice CrimeSolutions. (n.d.). Gender-specific programming for incarcerated females. Evidence review.
National Institute of Justice. (2021). Guidelines for post-sentencing risk assessment. Official publication.
National Institute of Justice. (2024). Five things to know about women and reentry. Official publication.
Salisbury, E. J., & Crawford, A. (2025). Gender-responsive treatment to improve outcomes for women and girls in correctional settings: Foundations, limitations and innovations. Health & Justice, 13, 11. https://doi.org/10.1186/s40352-025-00316-w
Schwartz, J., Steffensmeier, D., et al. (2009). Assessing trends in women's violence via data triangulation: Arrests, convictions, incarcerations, and victim reports. Peer-reviewed study.
Steffensmeier, D., & Allan, E. (1996). Gender and crime: Toward a gendered theory of female offending. Annual Review of Sociology, 22, 459–487. https://doi.org/10.1146/annurev.soc.22.1.459.
Steffensmeier, D., Schwartz, J., Zhong, H., & Ackerman, J. (2005). An assessment of recent trends in girls' violence using diverse longitudinal sources. Criminology, 43, 355–406. https://doi.org/10.1111/j.0011-1348.2005.00011.x.
Viljoen, J. L., Goossens, I., Monjazeb, S., & Cochrane, D. M. (2024). Are risk assessment tools more accurate than unstructured judgments in predicting violent, any, and sexual offending? A meta-analysis of direct comparison studies. Behavioral Sciences & the Law, 43, 75–113. https://doi.org/10.1002/bsl.2698
Yingling, J. (2016). Gendered pathways to crime. In The Handbook of Measurement Issues in Criminology and Criminal Justice. https://doi.org/10.1002/9781118868799.ch9.