LAUSR

Abstract Background Supportive housing is a promising intervention for insecurely housed families in the child welfare system. Objectives We examined factors associated with housing and child welfare outcomes to understand whom supportive housing benefits the most. Participants and setting Analyses included 500 insecurely housed families involved in child protective services across five U.S. sites. Researchers randomized families to receive supportive housing or the usual services offered to families in local child welfare systems. Methods We examined predictors of homelessness and child welfare outcomes with logistic regression, survival analysis, and receiver operating characteristic curves. We also examined supportive housing’s effect across risk levels for each outcome and tested theoretical moderators of intervention effects. Results Consistent with findings from the larger project of which this was a part, supportive housing improved housing outcomes for families, especially those most at risk form homelessness. Its effects on child welfare outcomes varied by study site more than by risk level. Family characteristics and experiences that predicted housing and child welfare outcomes also varied by site and added limited accuracy to predictive models. Theoretical moderators did not influence supportive housing’s effects. Conclusions Supportive housing increases housing security and may influence child welfare outcomes in some contexts. Family characteristics and experiences marginally improve targeting of supportive housing to insecurely housed families involved with CPS.