The Family Unification Program (FUP) provides permanent housing vouchers to families involved in the child welfare system who are homeless or unstably housed. This study used a randomized controlled trial (RCT) design in six sites to evaluate the impact of FUP on the probability that children would be placed in foster care and the probability that children in foster care would be reunified with their families. The study also examined how FUP was implemented in each of the six sites and collected perceptions of the program from parents, program staff, and administrators.

Why This Matters

Although child abuse and neglect remain highly prevalent and serious threats to children’s well-being, the evidence base for interventions that effectively meet the needs of the children, young people, and families served by our nation’s child welfare agencies is extremely limited. This shortage of evidence severely constrains agencies’ abilities, at all levels of government, to implement evidence-based programs and practices.

As part of a larger project aimed at increasing the number of evidence-supported interventions for the child welfare population, the Supporting Evidence Building in Child Welfare project designed a rigorous evaluation of FUP when used to provide housing assistance to families in the child welfare system who are homeless or unstably housed with a permanent housing voucher. Prior evaluations of FUP showed some positive impacts on child welfare outcomes.

What We Found

• For families with children at home at the time of randomization, we did not find a statistically significant impact of FUP on the probability that a child was placed in foster care.
• For families with children in out-of-home care at the time of randomization, at the child level, although the rate of reunification was 7 percentage points higher in the treatment group than the control group, the differences were not statistically significant. At the family level, we found that families randomized to treatment were 11 percentage points more likely to have all children returned home from foster care than families randomized to the control group.
• We found no statistically significant impacts on case closure or new substantiated allegations at the child or family level.
• Only 67 percent of families referred to FUP actually entered housing using a FUP voucher. However, for those who entered housing, the program had high housing stability rates, with 87 percent of families who signed a lease maintaining their voucher for two years.
• How the program was implemented and the impact results varied widely by site.
• This study strongly suggests that having dedicated staff to help families navigate the housing authority process and housing search helps families enter housing with a FUP voucher.

How We Did It

We recruited six study sites who were awarded FUP vouchers through the Department of Housing and Urban Development’s 2017/2018 and 2019 Notices of Funding Availability (NOFAs): Bucks County, Pennsylvania; Chicago, Illinois; King County/Seattle, Washington; Orange County, California; Phoenix, Arizona; and Santa Clara County, California.

The impact study used an RCT design in which families referred to FUP were randomly assigned to one of two groups: one was offered FUP (treatment group), and the other received the usual services in the community (control group).

We accessed administrative data from child welfare agencies to measure the impact of FUP on child welfare outcomes. We analyzed outcomes using regression models that controlled for site, baseline differences between the treatment and control groups, and whether the children were in foster care. We ran models at both the child and family levels.

The implementation study examined key process-related information, including how the families progressed from referral to FUP through entering housing with a FUP voucher. We collected data on program engagement and housing entry as well as services provided to families to help them enter and maintain housing. We also conducted interviews with staff and families.