Gender Differences in Recidivism: Varying Risk and Protective Factors

Background and Purpose: While the overall rates of juvenile crime have dropped, the proportion of females entering the juvenile justice system has increased. Studies find that risk and protective factors that affect delinquency and recidivism vary by gender. Risk factors identified as strong predictors for male delinquency appear less predictable for females. Victimization in childhood, mental health issues, family-related problems, delinquent peers or intimate partners, and drug use were found as strong risk factors for female delinquency. Due to the lack of gender-specific risk and needs assessments, research on risk factors for female delinquency still remains insufficient. In an effort to contribute to the literature with regard to gender differences in predictors of recidivism, the purpose of this study is two-fold: to examine the risk and protective factors related to recidivism and to see whether it varies by gender.    

Methods: Data were extracted the Los Angeles juvenile justice data system.  Inclusion criteria for the sample were youth placed under regular supervision and assessed with the Risk and Resiliency Checkup in 2008 (N= 2,825). Risk and protective factors were measured with RRC (α=.95). There were 6 risk and protective (for a total of 12) domains: environment, education, family, peer, substance use, and individual. Each domain had 5 items. Recidivism was coded yes (1) for youth who were rearrested or no (0) for youth who were not rearrested between March 2008 and September 2009. Survival months to recidivism were computed for the duration between the two consecutive arrests right before and after the RRC completion date. Two separate statistical tests were conducted with Hierarchical Cox Proportional Hazards Regression of survival analysis.   

Results: Males recidivated more than females. Males were significantly younger at first arrest, had more arrests, and were rearrested earlier than females. Females had lower risk factor scores in three domains: family, peer, and substance use. Females had higher protective factor scores in four domains: environment, education, peer, and individual. Hierarchical CPHR revealed that age at first arrests, number of arrests, and environment were significant risk factors for both males and females. Specific to females, education was a protective factor and family was a risk factor. Minority status, specifically being Black or Hispanic, education and substance use domains were significant risk factors for males, whereas environment was a protective factor.

Conclusions and Implications: With respect to recidivism, the study found that family and substance use are gender-specific risk factors.  While education was a risk factor for males, it was a protective factor for females. For females, the significance of family, education, and delinquent environment domains underscore the importance of relationships and attachment with caregivers and adults. The environment served as both a protective and risk factor for males, showing that males are influenced by both pro-social and anti-social environments. With respect to practice implications, these different predictors for recidivism point to the need for gender distinctive prevention and intervention activities.