System Pathways and Recidivism: How Does the Timing of System Involvement Predict Risk?

Background and Purpose

Many youth involved in the juvenile legal system (JLS) also experience the child welfare system, have diagnosable mental health needs, and receive special education services for emotional disturbance (ED). However, most research has not examined the confluence of multi-system involvement and related needs. Nor has research demonstrated, over time, how these system pathways impact risk for recidivism and limit the development of holistic services to address youths’ needs. Furthermore, how does the timing of their pathways predict JLS involvement?

Methods

Data are from a large Midwestern metropolitan area longitudinal study of services and outcomes for youth reported for maltreatment. Youth under the age of 12 when sampled (1993-2010) were matched by birth year and resident city/county with same-age youth in families receiving AFDC at baseline. A survival analysis with a continuation ratio approach modeled recidivism, uniquely using time varying covariates for all system involvement indicators and incorporating family and individual characteristics.

Results

The sample included 8,283 youth: 50% male (n=4,114), 79% Black (n=6,520), and 21% White (n=1,763). Youth with one petition: 27% (n=2,223), two petitions: 57% (n=1,275), three petitions: 60% (n=766), and four petitions: 69% (n=531). Results indicated pathways varied by race and gender. Males had increased risk for all four petitions, with 60% higher risk of a first petition than females. Compared to Black youth, White youth had approximately 30% lower risk of the first two petitions.

Social and economic context increased risk for petitions more for females than males: neighborhood poverty increased risk by 0.8% for females’ third petition; receiving public assistance increased risk by 5% for females’ fourth petition; and for all youth each month in foster care decreased the risk of initial JLS petition by 0.6%, but for females only, each month spent in foster care increased risk for a fourth petition by 3.6%. The risk of JLS involvement due to neglect and physical abuse varied across petitions, with neglect having the highest increase of risk (109%) for females’ third petition. History of sexual abuse had a strong influence for females during later petitions, increasing risk by 294% for the third and 132% for the fourth petition. Mental health diagnoses were generally not significant predictors of juvenile legal system petitions. A special education diagnosis of ED increased the hazard for initial JLS involvement by approximately 180% in the overall sample and remained significant for males for the second and third petitions.

Conclusions and Implications

Differing pathways, by gender, to and through the JLS further support the need to 1) understand how gender drives engagement with the JLS and 2) tailor interventions and enact supportive policies. In addition, due to the long-standing, entrenched, and persistent nature of racial disparities and biases in each system, efforts must continue to acknowledge and eliminate them. The time before the first petition was most crucial, thus; understanding the relationship between systems and needs can inform how systems should collaborate and respond. Future studies should continue these efforts with a more racially, ethnically, and geographically diverse sample.