Impact Findings From a Randomized Controlled Trial of a Supportive Housing Demonstration Programme for Homeless, Child Welfare-Involved Families

A "Housing First" approach to homelessness, showing promise for single adults, might also be effective for families whose homelessness combined with other challenges bring them to the attention of the child welfare system. This paper reports impact findings from the evaluation of a randomized, controlled trial in San Francisco that tested the effectiveness of an intervention designed to provide permanent housing and supportive services to such families, with the goal of safely reducing the use of foster care. Compared with families in the control group (33 families with 60 children), we found no significant impact on the likelihood of out-of-home placement up to 2 years after randomization. For families whose children were already in foster care, children in the treatment group (43 families with 63 children) reunified faster than those in the control group (42 families with 68 children), although nearly always before being housed. There was no difference between groups in time to case closure or likelihood of subsequent child welfare involvement. We discuss these findings in the context of important implementation challenges that have implications for further testing this intervention in tight housing markets.