Does Supportive Housing for Child Welfare Involved Families Help Keep Families Together?

Background and Purpose: When considering an allegation of abuse or neglect, child welfare service workers must decide whether the home environment is unsafe enough to necessitate the placement of a child in out-of-home care. Inadequate housing – and its impact on family wellbeing – influence that decision. Families experiencing homelessness or housing instability can struggle to provide a safe, stable environment for their children thus making it less likely a family stays intact. What if housing could help child welfare agencies improve child safety and permanence? In 2012, the Children’s Bureau funded a five-year demonstration that provided supportive housing to families in the child welfare system in five sites. The Urban Institute conducted the cross-site evaluation, a mixed-methods randomized controlled trial that included 807 families who were randomized to either receive supportive housing (treatment group, N = 377) or services-as-usual (control group, N = 430). This paper examines child welfare outcomes over 4.5 years and considers whether providing supportive housing to families impacted their involvement with the child welfare system.

Method: We drew upon child welfare administrative data that tracked each family’s child welfare events longitudinally from randomization to at least 54 months after. These data include information on allegations, removals, and placements for 98 percent of families in the treatment and control groups. We also look to referral data for these families. Referral data include information on housing status at referral, the caregiver’s child welfare history, and family challenges such as disability, mental health issues, substance abuse issues, past criminal justice involvement, domestic violence history, and any issues related to children’s health and development.

Results: Families receiving supportive housing showed more family stability with a lower rate of removal, fewer days spent in care, and a higher rate of reunification. Children from treatment families were less likely to be removed at 4.5 years after randomization (32 percent) compared to their control group counterparts (44 percent). Children in treatment families also spent about 18 percent more time at home than children in control group families. For children who were removed from home at randomization, those in treatment families were 1.3 times more likely to experience reunification (62 percent versus 47 percent). All outcomes varied between the five demonstration sites.

Conclusions and Implications: While supportive housing has shown some promise for improving child welfare outcomes early on, this is one of the first studies to examine supportive housing outcomes up to 4.5 years. Our longitudinal examination utilizing robust administrative data indicates that supportive housing provides stability for families that appears to decrease their long-term involvement with the child welfare system. Although child welfare outcomes varied by site, the housing subsidy and receipt of other supportive services appears to lead to lower removal rates, less time in care, and higher rates of reunification.