Examining predictors of recidivism for youth at first contact with the juvenile justice system

Featured Article | Law and Human Behavior | 2022, Vol. 46, No. 2, 140-153

Peer, Substance Use, and Race-Related Factors Associated with Recidivism among First-Time Justice-Involved Youth

Evan D. Holloway, Johanna B. Folk, Catalina Ordorica, and Marina Tolou-Shams Department of Psychiatry & Behavioral Sciences, University of California, San Francisco

Division of Infant, Child, & Adolescent Psychiatry, Zuckerberg San Francisco General Hospital

Objectives: Peer deviancy and substance-related consequences are dynamic criminogenic needs associated with increased risk of recidivism for justice-involved youth. Most prior research in this area, however, is based on samples of primarily male youth charged with delinquent offenses. Because identification of dynamic criminogenic needs is essential to delinquency risk reduction efforts, the purpose of this study was to examine the role of peer deviancy and substance-related consequences in a sample of youth at first contact with the juvenile justice system, with relatively equal representation of males and females and youth charged with delinquent and status offenses.

Hypotheses: We hypothesized that higher levels of peer deviancy and more severe alcohol- and cannabis-related consequences would predict recidivism. We also hypothesized that Black and Brown youth would be more likely to recidivate than non-Latinx White participants.

Method: First-time justice-involved youth (N = 401) aged 12 to 18 and their caregivers reported independent variables at baseline (demographic, legal, psychiatric, and peer factors). Official records of recidivism (i.e., number of new charges 2 years later) was the dependent variable for nested multivariate negative binomial regression models.

Results: Peer deviancy reported by caregivers, but not by youth, predicted recidivism 2 years later. Consequences related to alcohol, but not cannabis, increased recidivism risk. Finally, participants who were younger, male, charged with a delinquent offense, and Black, multiracial, and/or Latinx were more likely to recidivate than non-Latinx White participants after controlling for covariates. Conclusions: Results highlight the influence of institutionalized racism on later court involvement for youth of color at first court contact, regardless of individual risk. Deviant peers and consequences of alcohol are salient intervention targets for this population.

peer deviancy, substance-related consequences, juvenile justice; disproportionate minoritized contact; institutionalized racism

“Despite decreasing crime rates, arrests, and commitments to correctional facilities over the past few decades, 842,300 youth were involved with the juvenile justice system in the United States in 2018. Of those, approximately 744,500 youth were charged with a delinquent offense and 97,800 were charged with a status offense that was petitioned by a juvenile court in 2017. Regardless of the offense type, justice-involved youth are at substantial risk for future court involvement. Although predicting delinquent acts is challenging, given the interaction of myriad individual and contextual risk factors, a wealth of empirical evidence has pointed to eight risk factors for delinquency, commonly referred to as criminogenic needs. In this study, we examined the association between two dynamic criminogenic needs (i.e., substance-related consequences and peer deviancy) and the number of new juvenile charges over a 2-year period in a sample of first-time justice- involved youth” (p. 140-141).

“The first aim of this prospective study was to examine whether two dynamic criminogenic needs (i.e., peer deviancy and substance-related consequences) predicted recidivism for a sample of first-time justice-involved youth. As hypothesized, peer deviancy and alcohol-related consequences were found to be salient predictors of future court involvement for this sample with relatively equal representation of participants who identified as male and female as well as those charged with status and delinquent offenses; these relationships were statistically significant after accounting for the impact of demographic, psychiatric, and historical static risk factors. Bivariate analyses showed that cannabis-related consequences, but not alcohol-related consequences, predicted recidivism. Results of multivariate analyses, however, suggested the opposite: Cannabis-related consequences no longer predicted recidivism, whereas alcohol-related consequences did. The second aim was to examine whether youth self-and/or caregiver-reported deviancy in the youth’s peer group predicted recidivism. Results of bivariate analyses showed that whereas youth-reported peer measures did not predict recidivism, caregivers’ reports of their youth’s deviant and prosocial peers predicted recidivism in the expected direction. After controlling for demographic, legal, psychiatric, and historical static risk factors, only caregivers’ reports of their youth’s deviant peers predicted more new charges during the 2-year follow-up period. The third aim was to examine whether Black and Brown youth were more likely to recidivate compared to White non-Latinx youth, after controlling for covariates and independent variables associated with recidivism. Consistent with our hypothesis, Black and multiracial non-Latinx and Latinx youth were more likely to recidivate than White non-Latinx youth” (p. 149).

“A number of demographic characteristics also predicted recidivism at the multivariate level. Younger participants were more likely to recidivate than older participants and males were more likely to recidivate than females, consistent with meta-analytic findings. Also consistent with the extant literature, Latinx, Black non-Latinx, and multiracial participants were significantly more likely to recidivate compared to White non-Latinx participants, after controlling for other demographics, historical legal risk factors, and two dynamic criminogenic needs (i.e., alcohol-related consequences and peer deviancy). This finding is likely reflective of racial bias documented in studies of school disciplinary practices and referrals to juvenile court; police perceptions of criminality; juvenile court staff diversion decisions; prosecutorial discretion; juvenile court judges’ disposition decisions, including sentences of life without parole; and racist media descriptions of adult Black defendants increasing the likelihood of execution. Thus, there is ample evidence that racial bias permeates most, if not all, aspects of the justice system and accounts for a large portion of the variance in continued involvement in the juvenile justice system as well as at initial court contact for youth, per our study, regardless of their individual-level risk” (p. 149-150).

“Results of this study can inform risk assessment and case planning decisions with first- time justice-involved youth. The relatively equal representation in terms of gender identity and charge severity improves the generalizability of these findings to both female- and male-identified youth and those charged with both status and delinquent offenses” (p. 150).

“Best practices in risk assessment are to evaluate both historical static risk factors and dynamic criminogenic needs because they are associated with recidivism risk. In the current study, the results of multivariate analyses showed that both historical legal risk factors and dynamic criminogenic needs predicted recidivism during the 2-year follow-up period. Because this sample consisted of first-time justice-involved youth, past involvement with the justice system was not a static risk factor for any participants. Among legal factors, only being charged with a delinquent offense predicted recidivism. In contrast, past-4-month self- reported delinquency, caregiver-reported externalizing behavior (i.e., aggression, hyperactivity, and conduct problems), a history of out-of-home placement, and psychiatric hospitalizations did not. Thus, these findings indicate that being charged with a delinquent offense was the only salient legal risk factor for recidivism” (p. 150).

“Both dynamic criminogenic needs predicted recidivism, although these findings are nuanced when compared with the extant literature. Only alcohol-related consequences predicted recidivism—not consequences related to cannabis, which is inconsistent with meta-analytic results showing that drug-related consequences increased recidivism risk for justice-involved youth regardless of gender, whereas alcohol-related consequences predicted recidivism only for males; this inconsistency may be due to the relatively equal representation of males and females in our sample, which provided our study with greater statistical power than others to examine the interaction between gender and the consequences of different types of substances (i.e., alcohol & cannabis) in predicting recidivism... Our findings are also consistent with prior findings showing that peer influence may have different impacts on the use of alcohol and cannabis, perhaps due to engagement in delinquent acts together or to gene–social group interactions” (p. 150).

“Although our data were not collected in the context of a risk assessment procedure, the results highlight the importance of collecting data from both youth and their caregivers to identify recidivism risk and targets for intervention when youth first encounter the court, especially when identifying whether peer deviancy is a salient dynamic criminogenic need…Collecting information from both youth and caregivers is also likely to yield more information about specific responsivity factors, such as mental health needs, that should be targeted for intervention. Though not directly tied to recidivism risk, some evidence suggests that receipt of services for specific responsivity factors can reduce recidivism risk, potentially via increasing the number of dynamic criminogenic needs that are targeted for intervention” (p. 150-151).

“Notably, a greater number of prosocial peers was associated with less recidivism risk at the bivariate level but was no longer significant after controlling for covariates, including the impact of deviant peers. Meta-analytic findings suggest that prosocial peers are protective against recidivism and deviant peers increase recidivism risk for both male and female justice- involved youth. However, the current findings suggest that the protective effect of prosocial peers may be washed out after controlling for demographics, historical legal risk, and deviant peers. In this sample of justice-involved youth, recidivism risk 2 years after first contact with the court was more closely associated with caregiver-reported deviant peers than caregiver reported-prosocial peers” (p. 151).

“Although identifying the exact mechanism that increased the likelihood of ongoing court involvement for Black, multiracial, and/or Latinx youth participants in this sample is beyond the scope of this study, a review of the literature provides some guidance. For example, youth of color are more likely to have police contact because of increased surveillance in their neighborhoods prescribed by so-called proactive policing and the War on Crime in the 1980s. Further, results of longitudinal research with Black and brown ninth- grade boys in New York City suggested that the frequency of police stops was associated with increased self-report delinquency 12 months later; this relationship was partially mediated by psychological distress. The authors also found that the direct relationship between police stops and delinquent behaviors 6 months later interacted with age, with younger participants impacted to a greater extent. In other words, police stops appear to be iatrogenic, at least for boys of color in an urban area” (p. 150).

https://concept.paloaltou.edu/course/AAFP-Assessment-of-Risk-for-Violence-in-Juveniles?hsLang=en

https://concept.paloaltou.edu/course/Juvenile-Assessment-Issues-and-Evaluation-of-Juvenile-Waiver-to-Adult-Court