Examining Cross-System Outcomes in a Supportive Housing and Child Welfare Initiative

Background: Child welfare and homeless research acknowledge the intersection of homelessness/housing hardship and child welfare involvement. Homelessness can precipitate child maltreatment reports, threaten family preservation, and/or delay reunification. Entering the shelter system, whether fleeing from domestic violence or due to other circumstances, can heighten the risk of family breakup. This study is based on a five-year evaluation of a federally funded demonstration of supportive housing for homeless/unstably housed families with open child welfare cases in a small Midwestern city. The demonstration sought to bridge child welfare and housing systems to strengthen housing and family stability. Using a “housing first” approach, 100 families were quickly moved into housing, with service coordinators helping each family access concrete, therapeutic, and supportive services.

Methods: Longitudinal data were collected through multiple methods. Data sources include screening and intake information, child welfare and housing administrative data, and repeated family surveys and assessments including the North Carolina Family Assessment Scales. Eligible families had to meet all criteria for randomization into experimental (supportive housing) or control group (standard child welfare and community services): 1) living on the street, in emergency or transitional housing, or facing eviction within two weeks; 2) open child welfare case with a goal of family preservation or reunification; 3) a complex service need such as domestic violence, chronic mental illness, or substance misuse; 4) within 30% of the average median income; and 5) all children aged 12 or younger. As a cross-system initiative, outcomes were defined and measured for both housing stability and family preservation or reunification. For the current study, which focuses on the experimental group only, data were analyzed using chi-square tests of association, Analysis of Variance, and logistic and multinomial regression.    

Results: Using a four-category typology of cross-system outcomes, results indicated: 1) 60% of families achieved both housing and family preservation or reunification; 2) 20% achieved housing stability but families did not remain intact or reunify; 3) 5% of families achieved family integration but not stable housing; and 4) 15% had unstable housing and child welfare outcomes. Controlling for an initial goal of family preservation, length of time known to the child welfare agency, and baseline Parental Capacities (NCFAS subscale), factors that most strongly differentiated families with housing and family stability from those with instability in both areas were a substance misuse relapse and a criminal justice incident during the service period. Number of moves did not differentiate families with stable and unstable housing outcomes. Length of time in the program was comparable for families with stable housing, regardless of child welfare outcomes.

Implications:  Interventions involving multiple service systems have intersecting outcomes.  Findings suggest that supportive housing provided within a child welfare context helped the majority of families achieve both housing and family stability, but stabilizing housing did not guarantee family preservation or reunification. Policy and practice implications for improving supportive housing services with child welfare-involved families will be discussed.