Early intervention as delinquency prevention: Evidence from the Chicago Longitudinal Study

This study explored the main effect association between a large scale federally funded preschool intervention and lower rates of juvenile delinquency. Child, family, peer, and school level mediators of the link between intervention and delinquency, and predictors of delinquency were also considered. This is one of the first studies to examine the processes by which early intervention leads to reductions in juvenile delinquency.;This study sample is youth of the Chicago Longitudinal Study, an investigation of the scholastic and social development of minority youth growing up in high poverty neighborhoods (CLS; Reynolds, Mavrogenes, Bezruczko & Hagemann, 1996; Smokowski, Reynolds & Bezrucko, 1999). The CLS began their investigation with the 1985 Chicago Public School kindergarten cohort and included all 989 students who were enrolled in the twenty Child-Parent Center (CPC) sites as well as a matched comparison of 550 children who received alternative preschool services available to the community at the time. The current study draws from 1,406 of the original 1,539 students who participated in the CLS and resided in Chicago at age 10 or older.;Findings suggested that early intervention was associated with lower overall rates of juvenile crime, fewer arrests to the juvenile court, lower rates of violent arrests, and lower rates of drug arrests. This pattern was replicated regardless of the covariate specification. Participation in the CPC program was associated with a 41.5% decrease in delinquency over the comparison group, controlling for covariates including CPC follow on participation, race/ethnicity, gender, and CPC program sites. This can also be interpreted as a mean rate of delinquency of 13.7% for the intervention group and a 23.4% mean rate of delinquency for the comparison group. Mediation findings indicated that family and school level mediators best accounted for percent reduction of the main effect. Full mediation models accounted for between 17% and 38% of the main effect relation. Only the outcome of drug arrests was fully mediated. The remaining outcomes were partially mediated. Prediction findings indicated that special education, school mobility, and peer social skills were the best predictors of later delinquency. Other predictors included gender and risk status. These findings have implications on the financial and social costs associated with crime, but also have important implications for social work practice and research.