Housing and Child Well Being: Implications for Research, Policy, and Practice

Abstract

Inadequate housing and homelessness represent significant barriers to family stability and child development. An accumulating body of evidence documents the relatively high risk of family separation among families experiencing housing instability and homelessness, the extent of housing problems experienced by families involved in the child welfare system, and the disproportionately high rates of homelessness among youth aging out of foster care. Vulnerable youth and families interact frequently with various social service programs intended to mitigate multifaceted and multilevel risks, however, systems efforts and resources are rarely coordinated and results to date are mixed. We introduce 13 papers that are part of a burgeoning, increasingly sophisticated body of scholarship that inform coordinated responses to inadequate housing experienced by families involved in child welfare and related interventions. We note emergent themes and state a pressing need for research that accounts for ecological and contextual influences, examines the differential impact of housing and service interventions, identifies critical ingredients of effective housing and service interventions, and positions for scale-up. We distill findings into a set of observations and recommendations that align with best intentions to improve quality of life and promote well being among some of society’s most vulnerable individuals.

Introduction

The role of safe, stable housing circumstances as a determining factor in child health and well being is well established in theory and research (Bronfenbrenner, 1979; Fowler et al., 2015; Leventhal & Newman, 2010). Families experiencing unsustainable rent burden, residential mobility, inadequate and unsafe housing, and literal homelessness struggle to meet the physical and emotional needs of children and adolescents; yet, relatively little evidence guides scalable approaches to prevent and mitigate instability and promote healthy child development (Fowler and Schoeny, in press). This recognition led to our Call for Papers for a special issue of the American Journal of Community Psychology (AJCP) focused on housing and well being. We hoped the issue would address a series of important questions. How do housing policies and practices affect families with children? What social services, housing interventions, and program characteristics enable parents to focus their resources on ensuring the care, development, and education of their children? On balance, what are the costs and benefits of housing and related interventions? Do housing subsidies operate effectively as a means of averting family separations, and if so, how? What circumstances distinguish those youth who become homeless during the transition out of foster care and into adulthood? How can our systems and interventions shore up these vulnerable youth to avert intergenerational cycles of poverty, child welfare involvement, and homelessness? What service models and support characteristics are most helpful as families move into tentative stability through the provision of housing subsidies? What implications do research findings hold for programs to support family life and how can family scholarship guide effective policies? Most importantly, what tangible steps might help families to capitalize on their own capacities and those of the community to cope, adjust, be self sufficient, and (yes) thrive in spite of multiple challenges?

The intentionally ambitious questions proposed reflect a demand for evidence that informs practice and policy. Service providers and administrators seek information on how best to serve families and youth facing high economic press, housing problems, family strain in contexts polarized by race, class, and politics. A need exists for applied research that moves beyond documentation of negative outcomes associated with homelessness, as well as the overlap between homeless and other social services. We hoped that submissions would reflect a contemporary understanding of the ecology of child development that incorporates insights into mechanisms and potential leverage points for intervention. The issue aimed to do more than summarize the state of the science but also stimulate discussion, anticipate new perspectives, shape the emerging scholarly agenda, and influence policy and practice.

The response to the Call plus concurrent initiatives demonstrate the considerable work being done to close the knowledge to practice gap. The special issue received more than 50 submissions that reflected a range of approaches, methodologies, geographies, disciplines, and interventions. The 13 accepted articles incorporate strong conceptual and methodological features that inform theory and service provision in key domains of policy import, screening, and prevalence, youth homelessness and inadequate housing, and housing interventions in context of scale. Two additional journal issues published this year on homeless youth and families contribute further insights (Haskett, 2017; White & Rog, 2017). Moreover, a series of federal initiatives that rigorously test housing interventions targeting homeless families and youth promises to change the programmatic and policy landscape. Considerable momentum exists among researchers and decision makers to promote the well being of inadequately housed households.

Summaries of Contributions in Context

The special issue begins with a commentary that explores pathways for development of housing policy targeting child and family well being and provides examples to illustrate them (Samuels, 2017). The author reminds the reader that policy emerges from legislative, agency, and practice contexts, and from several pathways, including litigation, high-profile events, community-based practice innovations, and research evidence. He exhorts scholars to conduct policy analyses that make clear the implications of research for policy, and to recommend specific policy changes that make evident the foreseeable impact on programs, costs, and outcomes. Samuels (2017) includes a list of federal policies that are implicated by the research in this special issue.

The policy commentary is followed by two brief reports on the prevalence of housing instability and homelessness among families being investigated for child maltreatment (Farrell, Dibble, Randall & Britner, 2017) and couch surfing among young people (Curry et al., 2017). Farrell, Dibble et al. (2017) report the findings of a mixed methods population study of housing instability and homelessness conducted in Connecticut. Using a brief (three-question) tool, child welfare investigators scaled housing history, current housing arrangement, and housing condition along a Likert scale from “asset” to “severe risk.” Among nearly 7000 families, approximately 5% demonstrate severe to very severe housing problems and one-third demonstrate moderate housing risk. When only families with substantiated maltreatment allegations are considered, the proportion with severe housing concerns rises to 21%. Investigators find the tool effective, feasible, and useful for “applying a housing lens” early in child welfare involvement. In fact, Connecticut is executing plans to expand these housing procedures statewide. The screening instrument provides child welfare systems an easy way to monitor and address housing issues that threaten family separation.

Using data from a nationally representative household telephone survey (over 13,000 adults with youth ages 13– 25 in their households or themselves ages 18–25), Curry and her colleagues in the Voices of Youth Count study found couch surfing to be relatively common, particularly among the older age group (Curry et al., 2017). Among households with 13- to 17-year-olds and 18- to 25-year-olds, 4.0% and 20.5% (respectively) couch surfed in the preceding 12 months. The story may lie in the subgroups; important social, economic, racial/ethnic, sexuality/gender-based identity, and educational circumstances distinguish youth who “only” couch surfed from those who experienced literal homelessness. These initial prevalence estimates require replication and verification going forward and the field eagerly anticipates additional evidence in support of planning for youth transitions.

The next three papers leverage unique datasets to understand youth homelessness and inadequate housing in the transition to adulthood. Linked administrative records provide population-level and prospective data to explore the intersection between child welfare involvement and homelessness (Putnam-Hornstein, Lery, Hoonhout, & Curry, 2017; Shah et al., 2017). Ethnographic data collection embedded within a representative and longitudinal survey of low-income families illuminates how young Latina mothers navigate housing instability (Elliott, Shuey, Zaika, Mims & Leventhal, 2017). Collectively, the studies show the prevalence of homelessness among vulnerable youth and identify key risk and protective processes.

Shah et al. (2017) identify key risk factors for homelessness among youth aging out of foster care from integrated administrative data. An extensive array of service records were merged for all youth aging out of foster care in the State of Washington between 2010 and 2012. Findings suggested that within 12 months of exit, 28% of youth experience homelessness. Prevalence is comparable to a national prospective survey of adolescents involved in the child welfare system that estimate approximately 27% reported some housing problems within a 12-month period (Fowler, Marcal, Zhang, Landsverk & Day, 2017). Shah also demonstrates the potential utility of screening on key domains for early identification of risk.

Putnam-Hornstein and colleagues (2017) use administrative data to further probe the connection between child welfare and subsequent homelessness. In contrast to Shah’s study focusing on youth aging out of foster care, the paper examines the lifetime frequency of child welfare involvement among adolescents seeking homeless services. The study estimates approximately half of youth aged 17–24 seeking homeless services in San Francisco between 2011 and 2014 had at least one prior child welfare report in the state of California, while differences existed by demographic and child welfare experiences. Estimates coincide with prior retrospective surveys of homeless adults and further emphasize potential opportunities for intervention due to the intersection between service systems.

Elliott et al. (2017) conducted a qualitative study of the experiences of 15 Latina adolescent mothers. Grounded theory and narrative analyses shed light on two types of housing instability: “horizontal” moves across family homes and “vertical” moves between living with family and independent living. Whereas residing even intermittently with family is perceived as fundamental to the pursuit of independent housing (due to material and emotional support), this comes at the cost of high conflict and family instability. This research illuminates how family support networks, even when they are characterized by conflict and instability, can enable moves toward independence and prevent forms of homelessness. It further underscores policy limitations including age restrictions on housing access and a dearth of culturally sensitive housing support services.

We turn next to a series of papers that evaluate the effects of housing interventions targeting different homeless populations (Newman & Holupka, 2017; Shinn, Brown & Gubits, in press; Pergamit, Cunningham & Hanson, 2017; Fowler & Schoeny, 2017; Rog, Henderson, Lun, Greer & Ellis, 2017). The studies use experimental or rigorous quasi-experimental designs to test the impact of permanent and permanent supportive housing models on child and family well being. Taken together with the extant literature, the collection of papers shows mixed outcomes. Housing interventions reduce family homelessness that is related with reductions in family separations (Fowler, Farrell, Marcal, Chung & Hovmand, 2017; Pergamit et al., 2017; Rog et al., 2017; Shinn et al., 2017) and provide some benefit for child well being (Newman & Holupka, 2017; Gubits et al., 2016). Yet, risk for child maltreatment and child welfare involvement remains (Gubits et al., 2016; Fowler & Schoeny, 2017; Pergamit et al., 2017; Rog et al., 2017). The complex needs of homeless families likely require multicomponent interventions of as-yet unknown duration, while the role of prevention remains unclear.

Newman and Holupka (2017) combine longitudinal data from the Panel Study of Income Dynamics, U.S. Census, American Community Survey, and U.S. Department of Housing and Urban Development administrative data to examine the nil effects of housing support noted in earlier research. Among more than 400 children studied, roughly half come from families who experienced housing assistance and the remainder, while eligible, do not. Propensity score matching and instrumental variables analysis uncover no overall effects for housing assistance on adolescent cognitive, behavior, and health outcomes. However, differences emerge in subgroups; assisted housing gives an “added boost” to children with the highest cognitive performance and fewest behavior problems and had opposite effects for children in the lowest quantile. Heterogeneity in effects signals the complexity of housing interventions for vulnerable families.

Shinn et al. (2017) present promising findings from the largest and most elaborate study of family homelessness to date — the Family Options Study. Using a randomized controlled trial with four intervention arms implemented across 12 communities, the study tests housing first approaches for families entering. Nearly 2,000 families entering homeless shelters were randomly assigned for permanent housing, rapid rehousing that provides time-limited rental assistance, traditional transitional housing with extensive case management, and homeless services as usual. Families were surveyed on average 20 and 37 months after randomization. Shinn et al. (2017) show that referral for permanent housing relates with significantly lower rates of child separation from families at 20 months. Moreover, mediation analyses show that risk reductions relate with improvements in housing stability and no changes in parenting risks. Findings provide strong support for housing first approaches that emphasize connection with safe and stable housing to end homelessness, which may also prevent subsequent child welfare involvement. It should be noted that effects diminish at 37-month follow-up (Gubits et al., 2016); yet, the timely and policy-relevant study is especially important given studies that show smaller effects of housing interventions for families already involved in the child welfare system.

Two studies in the issue evaluate the Family Unification Program – a federal initiative that provides permanent housing for families involved in the child welfare system. Fowler and Schoeny (2017) employ a randomized controlled trial embedded within the child welfare system in Chicago, IL to test the impact of referral for FUP compared with housing case management on keeping families together. A disconcerting pattern emerges across caregiver-reported child maltreatment assessed at five time points over a 30-month period; rates remain “stubbornly high” with no treatment differences. Pergamit et al. (2017) examine FUP in Portland, OR and San Diego, CA using a waitlist comparison with propensity score matching. The intervention shows little impact on preventing family separations and mixed effects on reunification. Moreover, FUP relates with faster closure of child welfare cases, which could be counter indicated given evidence in Chicago (Fowler & Schoeny, 2017). Together, the studies question whether permanent housing alone sufficiently supports families.

Rog et al. (2017) tested more intensive interventions that pair housing with supportive services across two evaluations in Washington state. In the first study, a one-group pretest–posttest design shows families entering homeless shelters who receive rapid rehousing with supportive services show no change in removal of children from home over 18 months, while 25% of families reunify with children already placed out of home. The second study incorporates comparison groups using propensity score matching with administrative data. Findings show families with children placed out of home who receive housing plus supportive services were more likely to reunify within 12 months compared to similar families receiving homeless services as usual, and reunification rates were similar to families with permanent housing. These results underscore the potential impact of intensive housing interventions on family unification.

The final papers of the special issue explore implementation and dissemination of interventions for homeless families and youth. Research shows the complexities involved in promoting well being among homeless families and youth. Inherently, the intensive interventions needed to mitigate recalcitrant problems are complex with multiple components, while opportunities for early intervention and prevention require considerable coordination across multiple service systems. The scarcity of resources requires careful consideration of the necessary components for delivery of cross-system interventions with fidelity. The scalability of various approaches should be considered early in the design of approaches.

D’Andrade and colleagues provide a nuanced portrait of a supportive housing program (Linkages) for child welfare-involved families (D’Andrade et al., 2017). Through four interlaced first person narratives — the statewide program director, a child welfare services coordinator, a caseworker, and a parent participant – we come to understand perspectives both convergent and distinct. Perhaps the most striking motif is an articulation of the “magnitude of the challenge” facing child welfare-involved families living in poverty and experiencing housing problems. Stakeholders note several program components considered critical to success: flexible approaches (e.g., definitions of homelessness; resource allocation, such as providing families with funds for security deposits and utility arrears); multilayered approaches that embed multiple programs; time-intensive efforts from caseworkers, agency leaders, and clients; advocacy; and a common vision among collaborating agencies.

Fowler, Farrell et al. (2017) reviews emerging evidence on housing interventions for child welfare-involved families and youth through the lens of scale-up. Evidence of various housing policy initiatives considers the extent to which interventions sustainably reduce family separations that accompany housing instability and homelessness. Incorporating indicators of success beyond child welfare outcome (including cost, reach, capacity, and fit), the review identifies a number of systemic constraints. Limits include the reliance on finite quantities of rental assistance, the need for cross-systems collaboration, and the challenges of sustaining interventions once external funding is expended. Furthermore, both the relatively equal distribution of resources across a broad range of families and the leveling of resources toward high need families may mask a capacity to affect population-level well being. Recommendations provide short- and longer-term actions for service providers and policy makers to promote housing and family stability that supports well being.

Conclusions and Implications

The current state of knowledge in housing and child welfare — a mixed set of findings, a number of observational studies, and some controlled trials — provides general programmatic guidance but does not illuminate specific program strategies and components or account for contextual influences (policy, practice, geography, region) on process and outcome. The evidence plus a number of ongoing initiatives promise to inform understanding of the effective housing as a platform for well being; yet, the combined results are unlikely to offer a singular, definitive “answer” (Farrell, Dibble et al., 2017). Several factors combine to contribute to programmatic diversity across housing interventions for families and youth: structural variations in service systems within and across jurisdictions, differing levels of support and resource across types of housing assistance, the layered nature of housing and social interventions, uneven availability of adjunctive evidence-based interventions, large variation in the sequencing of program services, and shortages of housing vouchers. This complexity does not preclude effective policy and practice, but rather reflects the reality of achieving sustainable promotion of housing stability and well being.

In closing, this special issue brings to light a few themes that are worth highlighting. First, some notes about prevalence are in order. Housing problems remain especially prevalent among child welfare-involved families, and screening represents an important tool to improve systematic responses. Although couch surfing appears to be relatively common among adolescents, a subset of youth experience more enduring housing instability and literal homelessness, and again, child welfare involvement signals elevated risk. Importantly, homelessness and maltreatment history among youth are gendered phenomena that vary as well by race and ethnicity, and this must be considered in the design of interventions.

Second, compared with the state of the evidence on family housing assistance, the literature on transition-age youth is immature. Much work is needed to inform, test, and sustain interventions for youth. The nascent state of the science offers opportunity to consider scale-up early in development and the work on behalf of young people might ultimately be more effective if it meaningfully includes youth input, and engages policymakers around cross-systems solutions with braided funding and embedded evaluation.

Third, the “fishbowl” hypothesis must be considered carefully for housing interventions targeting child welfare-involved families. A growing number of studies involving families under investigation for child abuse and neglect paint a potentially troubling picture of recalcitrant child welfare problems. Several authors attribute the failure to achieve significant effects on family separation to relatively high levels of surveillance among homeless and child welfare-involved families (Fowler, Taylor & Rufa, 2011; Farrell, Randall et al., 2017). The fishbowl effect is a plausible argument, but one that becomes less convincing as additional studies across diverse settings show nil or modest effects of housing interventions on maltreatment. Families referred for scarce housing interventions often demonstrate significant housing and related problems that require very intensive, multicomponent interventions of as-yet unknown duration. The safety and well being of children must remain of central concern in cross-system interventions.

Fourth, we observe a growing inclusion of administrative data in evaluations of effectiveness. This is a positive development with additional potential, and it also behooves the field to work closely with policymakers and public administrators to ensure that administrative data capture meaningful processes and outcomes that relate to and enable measurement of well being. Administrative data hold untapped potential to understand the course, sequence, and nature of cross-systems involvement among vulnerable families.

A fifth observation pertains to the arena of housing assistance for families. The field needs to devote additional attention to scaling family assets and needs and matching program components and intensity accordingly. It is important to identify leverage points that optimize use of scarce resources. This may be done by focusing on the delivery of services for subgroups at particular risk, as well as selective prevention initiatives that consider broad application of less intensive interventions. Research designs and analytic procedures need to be capable of discerning differential effects based on factors such as history and demography.

We invite readers to review each of these contributions, recognizing the expanding perspectives and rigorous methodologies that contribute to advancing knowledge of housing and well being. We greatly appreciate Jack Tebes’ confidence in our capacity to curate this special issue, his careful stewardship of the journal, and his thoughtful guidance and encouragement as the issue took shape. Thank you to authors and reviewers whose contributions have taught us much. We extend a challenge to you, the readers of AJCP, to critically examine the perspectives, methods, and findings of these papers, and to apply them in your work. Leveraging these findings can meaningfully impact scholarship, policy, and practice to the benefit of our most vulnerable families and children.

Highlights.

• Inadequate housing and homelessness threaten family stability and child development.

• A growing body of research links inadequate housing and homelessness with the child welfare system.

• At-risk youth and families interact with various service systems, which are rarely coordinated.

• These 13 papers inform coordinated responses to inadequate housing in the child welfare system.