Homelessness in the child welfare system: A randomized controlled trial to assess the impact of housing subsidies on foster care placements and costs

Abstract

Inadequate housing and homelessness among families represent a substantial challenge for child and adolescent well-being. Child welfare services confront housing that threatens placement into foster care with little resources and evidence to guide practice. The present study provides the first rigorous test of the Family Unification Program (FUP) – a federal program that offers housing subsidies for inadequately housed families under investigation for child maltreatment. A randomized controlled trial assesses program impact on foster care placement and costs.

The experiment referred intact child welfare-involved families whose inadequate housing threatened foster placement in Chicago, IL to FUP plus housing advocacy (n = 89 families with 257 children) or housing advocacy alone (n = 89 families with 257 children). Families were referred from 2011 to 2013, and administrative data recorded dates and costs of foster placements over a 3-year follow-up. Intent-to-treat analyses suggested families randomly assigned for FUP exhibited slower increases in rates of foster placement following housing intervention compared with families referred for housing advocacy alone. The program generates average savings of nearly $500 per family per year to the foster care system. Housing subsidies provide the foster care system small but significant benefits for keeping homeless families together. Findings inform the design of a coordinated child welfare response to housing insecurity.

Introduction

A growing body of research demonstrates connections between family homelessness and child maltreatment (Courtney, McMurtry, & Zinn, 2004; Cowal, Shinn, Weitzman, Stojanovic, & Labay, 2002; Fowler et al., 2013). National estimates suggest one-in-six of the nearly three million children investigated for abuse and neglect experience inadequate housing that threatens removal from the home (Fowler et al., 2013). Homelessness also delays reunification with parents among children already placed in foster care (Fowler et al., 2013). Charged with the responsibility of protecting children, the child welfare system struggles to address needs for safe and secure accommodations with long-term consequences for mental health and child development (Fowler, Henry, Schoeny, Taylor, & Chavira, 2014; Fowler, McGrath, & Henry, 2015).

Little evidence guides the child welfare response to inadequate housing that contributes to risk for out of home placement (Fowler, Taylor, & Rufa, 2011). A small body of research investigates the impact of housing subsidies on child welfare outcomes, including child abuse reports, substantiated reports, and child removal. Typically allocated by public housing authorities, these subsidies provide monthly financial allowances that ensure no more than 30% of household income goes toward rent. An randomized controlled trial of more than 2000 families entering homeless shelters shows promising effects of housing subsidies. The multiple-armed trial compared families referred for subsidies to homeless services as usual or temporary housing with supportive services. Families referred for subsidies report significantly better mental health and fewer parent-child separations 20 months later compared with services as usual (Gubits et al., 2016). However, differences in separations diminish by 37 months as more families referred for homeless services as usual access housing (Cunningham, Pergamit, Baum, & Luna, 2015). The study also demonstrates no treatment differences on reunifying children separated from parents at either follow-up (Cunningham, Burt et al., 2015; Gubits et al., 2016). The findings show some promise for using housing subsidies to protect children at risk for family separation due to homelessness, but it remains unknown whether similar benefits emerge among families already involved with the child welfare system.

The Family Unification Program (FUP) represents the largest effort to connect child welfare-involved families with housing vouchers that address instability. The federal initiative sponsored by the U.S. Department of Housing and Urban Development (HUD) incentivizes partnerships between local child welfare and public housing agencies. The child welfare system identifies families whose inadequate housing threatens foster placement or delays reunification for children already placed out-of-home, and the housing authority provides subsidies for families who meet eligibility for the Housing Voucher (Section 8) Program (Cunningham, Burt et al., 2015; U.S. Department of Housing & Urban Development, 2017). Families pay no more than 30% of income toward rent through the intervention in housing that meets minimal physical quality standards. Program theory assumes that safe and affordable accommodations reduce propensity for maltreatment, as well as subsequent need for expensive foster care services (Fowler & Schoeny, 2017). Enacted by Congress in 1992, FUP has allotted more than 47,000 vouchers to 386 communities across the United States based on a competitive application process (National Center for Housing and Child Welfare, 2011). Evidence for program impact remains space despite its long history.

An initial evaluation of FUP followed 995 child welfare-involved families in 31 cities that were allocated subsidies (Rog, Gilbert-Mongelli, & Lundy, 1998). The pretest-posttest single group design showed families who received FUP exhibited greater stability at a 12-month follow-up; 90% of families initially at risk for out-of-home placement remained intact, and approximately 85% of families with children already placed out of home were reunified. Furthermore, 85% of families receiving FUP subsidies remained stably housed over the 12-month follow-up. The lack of a comparison group limits the ability to rule out alternative explanations for program effects, including the natural stabilization of homeless families over the follow-up period. The demonstration may have actively promoted selection effects in that subsidies allocation was based, in part, on family ability to reunite with children quickly. Therefore, stabilization may be attributable to family motivation, not FUP. In addition, the evaluation did not assess impact on child safety or child well-being. The inability to isolate the influence of FUP limits the ability to draw strong conclusions on program effects. Lack of evidence impedes examination of program capacity locally and nationally.

A more recent study evaluated FUP using propensity score matching to derive a comparison group from which to compare child welfare outcomes (Pergamit, Cunningham, & Hanson, 2017). Using administrative records on families receiving child welfare services, the quasi-experiment compared families who received FUP with matched eligible families on a waitlist in Portland, OR (146 families) and San Diego, CA (238 families). Across an 18-month follow-up period, families who received FUP exhibited more rapid child welfare case closure and lower probability of re-reports for abuse and neglect. No significant differences existed on overall probabilities of preventing out-of-home placement or promoting return of children already placed across cities. Thus, among families who received subsidies to stay together, FUP reduced involvement with the child welfare system without diminishing overall risk for foster placement. Findings suggest smaller than expected impact of the program; however, findings are limited by the lack of a true comparison group and little information on cost savings associated with FUP.

The lack of evidence for FUP limits its potential utility. Estimates suggest FUP subsidies cost on average $14,000 per family per year, while foster placement expenditures would be $56,892 (Harburger & White, 2004). The promised savings in human suffering and social expenditures make FUP commonly recommended as a policy solution to homelessness in the child welfare system, despite little evidence of impact (Gubits et al., 2016; U.S. Department of Housing & Urban Development, 2011). However, Congress sporadically allocates funding to expand the program with no additions in the past five years. Sporadic funding impedes the ability of communities to scale up FUP, but the lack of rigorous evaluation questions the ability of FUP to keep families together as intended.

The study represents the first and only randomized controlled trial that examines the impact of FUP on keeping families together (Fowler & Chavira, 2014). The trial addresses policy- and practice-relevant questions regarding whether FUP reduces rates of child out of home placement and saves associated expenditures on foster care. The rigorous design randomly assigns intact families whose inadequate housing threatens child out of home placement. Families are assigned to FUP plus housing case management or housing case management alone. Child welfare records on foster placements and associated payments track outcomes 36 months following random assignment.

The study tests two primary hypotheses. First, families randomly assigned for FUP plus housing case management would exhibit significantly greater decline in the rate of foster care placement across 36-months following referral compared with families assigned to housing advocacy only. This tests a key assumption regarding the stated impact of FUP, and addresses limitations in current research from non-experimental studies with limited follow-up. Second, random assignment to FUP would relate with significant reductions in foster care expenditures on out-of-home placement costs. The hypothesis addresses a lack of policy-relevant information on FUP. Program theory assumes large cost savings from avoiding expensive non-kin foster placements; however, this has never been tested. The RCT with child welfare records provides the first rigorous test of the impact of FUP implemented within a large urban setting.