A Latent Class Analysis of Child Welfare System Involvement Prior to Commercial Sexual Exploitation

Background and Purpose: Behavioral factors that contribute to commercial sexual exploitation of children (CSE) are well documented, but the impact of decisions made within the child welfare system (CWS) remains largely untested. The purpose of this exploratory study is to identify subgroups of children within California’s population of young people with child welfare system involvement affected by CSE risk/victimization based upon characteristics tied to their cumulative CWS involvement.

Methods: Latent class analysis was used to identify unobservable groups within a population of 6,399 children that had CSE victimization or risk identified by the CWS after the age 15. LCA model fit was examined using the Akaike information criterion (AIC) and Bayesian information criterion (BIC). The final model included seven observable binary measures: incomplete maltreatment investigation; maltreatment reports in multiple counties; CWS diversion following substantiated maltreatment; initial CWS case open 24 months or longer; initial case concluded with family reunification or family preservation; placement with relatives; and placement in congregate care.

Results: A 5 class model was identified. Over half (56.8%) of the study population was assigned to the “Low Involvement” class, characterized by minimal involvement in the CWS at the investigation, case, and placement stages. Under a tenth (8.1%) of the study population was classified as “Highly Mobile,” who had the highest probability of being reported to the CWS in multiple counties and having one or more investigations closed as incomplete. The “CWS Reentry” class comprised just under one-fifth (18.0%) of the study population, all of whom had multiple cases before the age of 15. About one in ten (9.4%) youth in the study population were assigned to the “Relative Care” class, who had lengthy, stable CWS involvement. The remaining 7.8% were assigned to the “Congregate Care” group based on their predominant placement histories.

Conclusions and Implications: Results from this latent class analysis document the approximate size of California’s population of youth who have been exposed to child protective services with known CSE risk or victimization and reveal marked heterogeneity in terms of the timing, duration, continuity and setting of CWS intervention they experienced. Children classified as Low Involvement and Relative Care (who comprised two-thirds of the study population) would likely have benefitted from family based CSE prevention services, yet few evidence-based interventions have been evaluated for use with CSE-impacted families. This is particularly relevant for prevention work with the families of Black children, a quarter of whom were assigned to the Relative Care class. This analysis refines the empirical knowledge base on pathways into the commercial sex industry during childhood by identifying multiple patterns of CWS involvement among CSE-impacted youth.