Association for Public Policy Analysis & Management

Childhood homelessness is a growing social problem in the United States and has consistently been found to threaten children’s behavioral and educational outcomes. However, recent research has found notable variation in the extent to which homelessness affects these outcomes. First, sizeable subgroups of children appear to be faring well in both domains during shelter stays. Second, the few longitudinal studies of families that experience homelessness suggest a general pattern of recovery after an initial shelter stay. Housing policy supports through the homeless service system or federal housing assistance programs could be one explanation for recovery, but research is scant on whether and how these interventions affect child outcomes. Better understanding of the mechanisms that explain variation in how childhood homelessness and our policy responses to it affect children’s well-being is needed to identify how to support healthy developmental outcomes.

The present study draws on 80 semi-structured interviews with mothers participating in an experimental study of housing interventions for homeless families. All families were initially living in emergency shelters and were randomly assigned to one of three housing interventions—short-term rental subsidy (n=20), permanent housing subsidy (n=20), or transitional housing (n=20)—or to a usual care condition (n=20) of working with shelter staff to locate housing. Interviews were conducted an average of 7 months after shelter entry in four study sites. By this time, the majority of families (n=45) were living in their own housing, and few were living in emergency shelters (n=9). Our study examines three questions: 1) How do children’s educational and behavioral functioning change after a shelter stay? 2) To what extent are these changes are driven by housing and non-housing factors? 3) How do changes in functioning and drivers of these changes differ across study intervention conditions?

For behavioral and educational outcomes, children recovering after faring worse in shelter and doing similarly well during and after a shelter stay were the most common patterns. Bouncing back was most common for behavioral outcomes and consistently doing well was most common for educational outcomes. A small group of children had problems emerge only after moving out of shelter, often due to difficulties with concurrent school changes.

Mothers primarily perceived changes in the quality of their housing environments as explaining changes in their children’s behavior, with behavioral problems that had emerged in shelters dissipating when they moved back into their own housing. A smaller group of families that had moved frequently after exiting shelter were more likely to report ongoing problems attributed to a lack of stability in their children’s lives, though some of these parents were able to mitigate these challenges by maintaining strong routines across multiple living situations and consciously trying to keep moves within the same school district.

Preliminary results suggest that permanent housing subsidies reduce child behavior problems by enabling families to sustain independent housing and reducing parental stress. Some families offered rapid rehousing were quickly housed but in very poor-quality apartments that increased parental distress and negatively affected children’s well-being.