System Factors and Homelessness Among Youth Transitioning to Adulthood from Foster Care

Background: Multiple studies have documented that youth transitioning to adulthood from foster care face disproportionately high rates of homelessness compared to their peers not in foster care (Berzin et al., 2011; Morton et al., 2018). And while research has identified a variety of individual characteristics of youth transitioning to adulthood from care that are associated with homelessness (Prince et al., 2019), little attention has been paid to how characteristics of the foster care system influence this population’s likelihood of homelessness. This study explores predictors of self-reported homelessness in early adulthood among transition-age foster youth, focusing on characteristics of the system and the context of service delivery. Three questions are of particular interest:

1. Does remaining in extended foster care reduce the likelihood of homelessness among transition age youth?
2. Does the living arrangement type in which a youth resides (foster home, kinship foster home, therapeutic foster care, supervised independent living, and transitional supportive housing) influence the likelihood of homelessness?
3. Does specialization of case management in a county to focus on provision of services to young adults influence the likelihood of homelessness?

Method: The study used data from the California Youth Transitions to Adulthood Study (CalYOUTH), including a longitudinal study of youth in care (n = 595) who were interviewed at ages 17 and 19, surveys of caseworkers who served transition-age youth in care (n = 235), and child welfare administrative data from California. CalYOUTH survey data captured the timing of youths’ self-reported experience of homelessness, and a wide range of measures of individual characteristics and functioning at baseline. Caseworker surveys measured aspects of the county-level services delivery context. And administrative data were used to measure the youths’ maltreatment and foster care histories, including time in extended foster care and placement type. Predictors of homelessness between 17 and 19 years of age were estimated using Cox proportional hazard regression models, including time-invariant covariates measured at baseline and a time-varying measure of youths’ living arrangements.

Results: Approximately 18% of the youth reported having experienced homelessness during the study period, a rate significantly higher than that of their non-foster care peers (5%). Youth-level predictors of homelessness included race, poor health, physical assault, and identifying as nonheterosexual. Remaining in foster care into adulthood as opposed to leaving care reduced the risk of homelessness; the estimated odds of youth remaining in care becoming homeless were one-third to one-half those of youths who left care during the study period. With respect to living arrangement, placement in transitional supportive housing was associated with the greatest reduction in the risk of homelessness. County-level specialization of caseworkers to focus on transition-age youth was also associated with a reduction in the risk of homelessness.

Conclusion and Implications: Homelessness remains a challenge for many young people aging out of foster care. Study findings add to evidence of the potential of extended foster care to reduce homelessness. Study findings also support the development of treatment-oriented placement options and case management models appropriate for young adults.