Modern family homelessness emerged in the United States in the 1970s when widespread urban renewal eliminated a great deal of low-income affordable housing. In the ensuing decades, many other structural factors have coalesced to increase social exclusion, reduce housing stability, and spur dramatic increases in family homelessness, including growing income and wealth disparities, gentrification, racism, restrictions on welfare support, large reductions in housing subsidies, and tax policies favoring the wealthy (Shinn & Khadduri, 2020). The cumulative and interactive effects of these influences, combined with the lack of consensus among the institutions and systems that deal with family homelessness suggests that it is a “wicked problem” (Rittel & Webber, 1973): it is defined differently by different stakeholders; its determinants are located in multiple sources, and it is unclear what might constitute a solution. Even with the $1.5 billion allocated by the American Recovery and Reinvestment Act in 2009,Footnote 1 family homelessness has, at best, declined only slightly in the past decade (Shinn & Khadduri, 2020). It remains an incalcitrant problem in the United States, yet it is woefully under-researched as a focus in its own right, and, unfortunately, is virtually completely ignored by behavior analysts and culturo-behavior scientists.Footnote 2

The complexity of the phenomenon and the need for scholarship and change provide the opportunity to introduce new approaches and strategies to the field. We suggest that culturo-behavior science (CBS) can be one of these new voices by employing the metacontingency within a behavioral community psychology (BCP) approach (cf., Suarez-Balcazar et al., 2019). Such a conceptual analysis can lead to the generation of research questions, intervention options, and advocacy actions at all levels of systems analysis.

We hone in on the current policies prioritized in the Housing First/Rapid Re-Housing best practices approach to reducing homelessness because modest changes in it have the potential to improve people’s lives now. Fawcett (1991) emphasizes that BCP “should contribute to fundamental social change” but that it should do so “at the level of change and timing likely to optimize beneficial outcomes” (p. 633). He alludes to a paramount ethical choice in conceptualizing what is the “optimal beneficial outcome” when he notes that “the challenge of effecting change at a broader community level suggests the importance of a strategy of seeking small wins. Small wins are those concrete outcomes of modest significance that attract allies and deter opponents” (p. 627). Even though small wins do not change larger inequitable systems, they can embody “fundamental change” if they “contribute to prevention of problems in living, capacity building and empowerment of people of marginal status” (Fawcett, 1991, p. 633). We agree that it is important to seek “small wins” that improve the present situation for people struggling with housing instability within unjust systems that are beyond their control.

A crucial foundation for any CBS analysis, however, lies in a thorough understanding of the target system-level phenomena. Thus, we will first examine family homelessness – its pervasiveness and impact on family members, the current policies that prioritize Housing First/Rapid Re-Housing (HF/RRH)Footnote 3 (Padgett et al., 2016; Tsemberis, 1999), and what these policies actually demand of families and service providers. We will then review the sparse research on housing and the stabilization of homeless families and effort to prevent repeat episodes of homelessness before offering a CBS analysis, based on behavioral community psychology principles, as a first step in developing an analytic framework for the research and advocacy activities that are absent at present.

The Phenomenon of Family Homelessness

Number of Unhoused Families with Children

As noted above, even defining the scope of family homelessness and counting the number of families with children experiencing homelessness is complicated; different stakeholders define family homelessness in different ways, leading to different numbers. The US Department of Housing and Urban Development (HUD), which is the nation’s major funder of homeless services, defines family homelessness in a more restrictive manner than other federal agencies; it excludes families who are couch-surfing, doubled-up, living in motels, hotels and campgrounds, or otherwise precariously housed, as well as those living in regions of the United States that do not have Continuums of Care that coordinate homeless services (Bassuk et al., 2020).Footnote 4 HUD counts the number of people literally experiencing homelessness during one night in January by mobilizing staff and volunteers throughout the different regional Continuums of Care for what it calls an annual Point-In-Time (PIT) count and adding those results to its Homeless Management Information System database for tracking the number of people in shelter and transitional programs on that night (Shinn & Khadduri, 2020). In January 2020, HUD (2021a) estimated that 53,739 families representing 171,575 people were homeless, 111,592 of which were children under 18 years of age.Footnote 5 According to HUD’s census, homeless families represent more than 30% of the homeless population, with half of those families residing in shelters (Bassuk et al., 2020; The National Center on Family Homelessness [NCFH], 2011; United States Interagency Council on Homelessness [USICH], 2018).

However, many scholars point out that HUD may significantly undercount the number of families experiencing homelessness due to its narrow definition and its lack of a standardized and monitored census-taking methodology. Bassuk et al. (2020), for example, assert that HUD’s “PIT numbers may be from 2.5–10.2 times lower” (p. 249) than the numbers of families actually experiencing homelessness. Further, HUD’s (2021a) identification of 111,592 homeless children under 18 years of age contrasts starkly with the count under the less restrictive US Department of Education’s (DOE) definition, which includes families at imminent risk of losing housing, families who do not live independently or who have moved frequently, and families who are considered homeless by other federal agencies (Bassuk et al., 2020; Evans et al., 2019). Using the DOE definition, during the 2017–2018 school year, more than 1.3 million school-aged children were homeless while attending school and 1.4 million children under the age of six experienced homelessness (DOE, 2020, 2021; USICH, 2018).Footnote 6

These discrepancies in the numbers make it difficult to determine the effectiveness of current responses to family homelessness. Though some data suggest that the numbers appear to be decreasing slowly (e.g., Shinn & Khadduri, 2020), advocates argue that the presence of any type of risk to housing insecurity requires an all-hands-on-deck effort to intervene on behalf of families (National Alliance to End Homelessness [NEAH], 2021).

Demographics of Family Homelessness

The majority of homeless families are headed by single mothers – 77.6% in the 2017 PIT count – most of whom are between the ages of 18 and 30 with limited education and 2–3 children under six years of age. Families of color are overrepresented: in 2016, 78.3% of the families experiencing homelessness identified as non-white or white and Hispanic compared to 54.5% of homeless individuals (Bassuk et al., 2020; NCFH, 2011; USICH, 2018).

Homeless families often face challenges that are different from other homeless populations. Adverse childhood experiences (ACES) are common among homeless heads of households; over 92% of homeless mothers have experienced physical and/or sexual abuse either as children, from intimate partners, or both (USICH, 2018). Many homeless mothers lived outside their homes during periods of their childhood, often in foster care (USICH, 2018). In 2016, more than one in five adults in homeless families had a disability, a rate 2.6 times higher than housed adults in families and 1.5 times higher than housed adults in families that are living in poverty (USICH, 2018). Homeless mothers report higher levels of psychological distress and mental health issues than homeless men, including overwhelming feelings of anger, sadness, fear, and hopelessness (Nemiroff et al., 2011; NCFH, 2011). They have three times the rate of posttraumatic stress disorder and twice the rate of drug and alcohol dependence (NCFH, 2011). Fifty percent of homeless mothers have experienced a major depressive episode, have had suicidal thoughts, or have attempted suicide while homeless (Bassuk et al., 2020; USICH, 2018). In addition, over one third of homeless mothers have chronic physical health conditions, which also represents a rate higher than their housed peers (NCFH, 2011). Families experiencing homelessness interact with the child welfare system at higher rates than low-income but housed families with stable housing (USICH, 2018). Unstable housing also contributes to family separations (Shinn & Khadduri, 2020), to the placement of children in out-of-home care, and to delays in family reunifications (USICH, 2018).

These experiences take their toll on homeless children as well as their caretakers, exposing the majority to ACEs before, during, and/or after their episode(s) of homelessness. Many children witness and/or have experienced high rates of conflict and violence, including at least one serious violent event, which a quarter of the time is within their own family (Bassuk et al., 2020; NCFH, 2011). Compared to their housed peers, homeless children tend to be sick four times more often and go hungry twice as often (NCFH, 2011); they have three times more emotional, mental health and behavioral problems and are four times more likely to show delayed development and/or other disabilities (USICH, 2018). Homeless children also experience significant educational disruptions as a result of their housing instability, leading to poorer academic performance, greater learning disabilities, more repeated grades and suspensions, and higher drop-out rates, than their housed peers (Bassuk et al., 2020; NCFH, 2011).

Current Policies Addressing Family Homelessness

As noted earlier, current federal policies intended to prevent or reduce family homelessness have been grounded in the HEARTH Act and the adoption of HF/RRH as the best practice model. Prior to HF/RRH, family shelters were individually operated public and private agencies, often with requirements that homeless families complete short-term residential and treatment programs that addressed their barriers to stability before qualifying for assistance with housing. The HEARTH Act and HPRP reversed this practice and transformed service delivery by creating community-wide, coordinated homeless service systems (CoCs) that prioritized moving families into housing rapidly and then providing stability services once housed (Evans et al., 2019).

The guiding principle of HF/RRH is that housing itself is a precondition for resolving risks to housing stability such as mental health and substance abuse, incomplete education, and unemployment/underemployment. This approach grew out of the success of the Pathways to Housing demonstration project, pioneered by community psychologists (Tsemberis, 1999), in housing chronically homeless men by providing permanent, supportive housing subsidies coupled with voluntary individualized stability services. However, unlike the Pathways to Housing model, most of the current HF/RRH programs offer only temporary housing subsidies, particularly for families (and individuals) who have not experienced chronic homelessness (Burt et al., 2016; Evans et al., 2019). The HF/RRH model, with its goal to house homeless families as rapidly as possible, offers homeless families only short-term rental assistance and case management services focused primarily on locating and moving into housing (Jefferson et al., 2020; Shinn et al., 2018).Footnote 7

The Behavioral Experience of HF/RRH Service Delivery

To comprehend the impact of HF/RRH policies and practices on service delivery, let us step into the shoes of a homeless family. Based on the demographic, the head of the household is likely to be a mother who may not have finished high school (USICH, 2018), and if under 25 years of age, may not yet have developed fully matured executive functioning (Arain et al., 2013). The family likely has experienced multiple ACES, including interpersonal and/or family violence as well as mental health and physical wellness issues. Parents may have recently separated, further compounding the stress on the newly unhoused family. In addition, there is always the fear that the child welfare system will step in and remove the children.

In 2012, all CoCs were required to establish a Coordinated Entry System, resulting in the centralization of the intake process for access to homeless services (Evans et al., 2019; Shinn & Khadduri, 2020). Intake usually includes an in-person meeting with a specialist to identify each family’s needs and match them to a shelter program that can address their barriers. For homeless families, this generally requires that a mother, most often in her twenties, transport herself and her children to the intake agency, where she and her children will wait to be interviewed among all the other people seeking services, including those with serious mental illness, substance abuse issues, and those recently released from prison.

At the conclusion of the intake interview, the family is assigned, often without the mother’s input, to an HF/RRH shelter program. Moreover, due to the chronic shortage of shelter beds, the family may be sent to an overflow shelter for several days or weeks, before being able to move into their designated shelter program. The many overflow shelters are congregate living settings, do not provide supportive services, and usually operate for limited hours (typically only overnight) with strict curfews and clear-out times. This means that the family has to rise early every morning, pack up their belongings and make their way to the Coordinated Entry agency where they may be served breakfast. The mother must then deliver her children to school and/or day care, and report to work. If she is unemployed, she must figure out where to spend her days – maybe hanging out in nearby parks or, when the weather is cold, riding the buses all day, sometimes with her youngest children trailing behind her.

Once an HF/RRH shelter bed becomes available and the family moves into their assigned shelter program, they are connected with a case manager and begin the work required to access their HF/RRH housing subsidies (Jefferson et al., 2020). Most HF/RRH programs have limits to the amount of time a family can remain in a shelter to qualify for housing subsidies and services. These will differ by community, but HUD’s recommended benchmark is 45 days, based on data from only one study (Gubits, Bishop et al., 2018a). This is a target only slightly less ambitious than the 30-day recommendation by National Alliance to End Homelessness (NAEH, 2016a). Thus, from the moment of entry into the shelter, the family finds itself under the constant presence of looming deadlines, which can add yet another level of stress to families who may already be traumatized by their experiences and overburdened by their challenges. For example, families may not have access to their identification documents that are required to qualify for HF/RRH assistance (e.g., birth certificates, drivers’ licenses, state IDs, social security cards, etc.). Therefore, they must request them before they even start applying for their HF/RRH housing subsidies. They also must reconnect with their benefits (Medicaid, food stamps, childcare voucher) and enroll their children in schools and daycares. Consequently, it is not unusual for families to spend their first few weeks in the HF/RRH shelter organizing and attending appointments at different county offices, the social security administration, doctor’s offices (for their children’s vaccinations and/or vaccination records), and at schools and daycares. It may take several weeks before they receive the needed documents; meanwhile, the days are ticking by and both the family and their case manager may become increasingly anxious about their HF/RRH deadlines.Footnote 8

After the required documents finally have been received and their children have been settled into schools and daycares near the HF/RRH shelter, the focus turns to locating housing, which may necessitate meeting with yet another service provider – a housing locator. Applying for HF/RRH subsidies requires filling out still more forms and undergoing further assessments before the family can begin the often-discouraging search for housing. Based on each family’s demographics, HPRP will set parameters on the type of house that will qualify the family for subsidies, such as the maximum rent, the number of bedrooms, etc. HF/RRH programs also require that landlords agree to certain conditions, such as accepting a fair market value for rent and passing a HUD-based inspection criteria. Additional impediments to locating housing include lack of affordable options, especially in areas with good school systems; landlords’ hesitancy to rent to families with short-term housing subsidies, especially families with young children, past histories of eviction and/or criminal backgrounds; and the requirement that landlords meet HUD deadlines, including tight timelines for the submission of paperwork and the completion of HUD-required-repairs. It is not unusual for families to find housing, only to have it rejected upon inspection, either because the housing did not meet HUD standards or because the landlord cannot or will not make the required repairs in the required time frame (Jefferson et al., 2020). When this happens, the family has to start the disheartening process of searching for housing all over again, still under the same approaching deadlines.

Once housing has been located and approved for subsidies, most HF/RRH programs give families a very short time to move out of the shelter. The reason is twofold: (1) rental subsidies are being expended so the family should be living in their rent-subsidized housing; and (2) the family’s shelter bed should be freed for another homeless family waiting in overflow shelter. However, the list of tasks that the family must accomplish in order to move is long. For example, they must change the utilities accounts into their names; family belongings must be packed, frequently in trash bags, and transported to their new house, often via several trips on a bus with their children in tow; and a fridge, stove, furniture, beds, chairs, tables, and other necessities must be located. Most families must apply for vouchers to purchase their furnishings, but this cannot be done in advance of securing housing as voucher eligibility often is based on identifying a specific address. All of these tasks take time, effort, and a high degree of logistical organization for both the head of household and the case manager. Consequently, many families move into empty or near empty housing, sometimes with nothing more than a hot plate, an ice chest, and cots.

After moving into their housing, the family will need to stock their kitchen with food – which may require multiple trips, with children, to and from the closest grocery store, often on buses due to the food-deserts in low-income neighborhoods. They must switch enrollment of their children to local schools, also under deadline, and update their ongoing support systems, which may require multiple visits to county offices, mental health clinics, substance abuse programs, doctors’ offices, and nearby child daycares. They must figure out new bus routes and learn about the resources available in their new communities.

Thus, the family’s journey from being unhoused to being home is complex and filled with stressful deadlines and discouraging setbacks. By and large, homeless service systems are not trauma-informed nor are they sensitive to the serious economic, societal and health challenges experienced by members of homeless families (cf. Shinn & Khadduri, 2020). It is not surprising, given the enormous number of tasks that each family has to accomplish under great time constraints and difficult circumstances, that most homeless services tend to be reactive, crisis-oriented and short-term; they provide few education or modeling opportunities to enhance behavioral and cognitive capacities that may help homeless families achieve better housing and economic stability (cf. Jefferson et al., 2020). These deadlines also foster interactions between shelter service providers and homeless families that are fraught with tension and urgency. Moreover, as will be discussed in the next section, there is little evidence that HF/RRH short-term housing subsidies and limited case management services advance long-term stability for homeless families (Gubits et al., 2016; Gubits, Shinn et al., 2018b). On the contrary, growing evidence suggests that parents struggle to maintain their housing stability after the conclusion of HF/RRH assistance, due to such predictable challenges as unemployment and low-wage employment; child care needs; physical and mental health problems; lack of social supports; and unrealized behavioral and cognitive capacities (Bassuk et al., 2020; Evans et al., 2019; Jefferson et al., 2020; Kahneman, 2011; Shinn & Khadduri, 2020). Today’s Covid environment creates even greater challenges.

Research on Housing First Approaches to Stabilizing Families in Homes

HF programs can be categorized as either primary or secondary prevention of homelessness. The former is intervention before the problem phenomenon ever occurs (i.e., homelessness prevention), either by intervening to change the structural inequities that are ultimately responsible for the phenomenon or to alter current practices to improve their capacity to avert the problem even while the systemic injustices persist. Secondary prevention is intervention as soon as feasible that is aimed at avoiding a return of the problem (i.e., prevent additional episodes of homelessness).Footnote 9 Unfortunately, the research on both types of prevention of family homelessness is sparse, often uncontrolled, and largely inconclusive (Evans et al., 2019), in contrast to the attention that has been directed to homeless individuals with substance use and mental health problems (cf. Levy, 2011; Levy & Havens, 2013; Padgett et al., 2016; Tsemberis, 2010, 2016). Because HF/RRH is designed to prevent additional episodes of homelessness, it can be conceptualized as a secondary prevention approach; we will review that limited research first, then summarize the few investigations of primary prevention programs in an effort to generalize the data base as much as possible.

Secondary Prevention Programs for Family Homelessness

Preventing re-entry into homelessness by families through HF/RRH programs has been the focus of only a handful of investigations. Evans et al. (2019) summarize four uncontrolled studies that suggest temporary housing vouchers provide a modest reduction in return-to-shelter within two years, with the homelessness re-entry rate ranging from 6–26%. The only experimental study to date is the Family Options Study that compared re-entry to homelessness from four intervention approaches: HF/RRH, transitional housing, “usual care,” and permanent housing subsidy programs (Gubits et al., 2016). HF/RRH included temporary rent assistance and time-limited services; transitional housing involved placement in agency-controlled housing along with provision of intensive services; “usual care” left families to seek services on their own; and the permanent housing subsidy typically was a Section 8 voucher.

The three-year Family Options data suggest that HF/RRH is comparable to transitional housing and “usual care” on measures of housing stability and family, parent, and child adjustment outcomes; its main advantage is an estimated cost savings of 10% compared to “usual care,” which in turn was estimated to cost 4% less than transitional housing (Evans et al., 2019; Gubits et al., 2016; Shinn & Khadduri, 2020). The most effective but by far the most expensive option was the provision of permanent subsidies; families receiving this support evidenced significantly fewer repeat episodes of homelessness, fewer child problems, less parental distress, and more freedom to leave violent relationships (Gubits, Shinn et al., 2018b; Shinn et al., 2018; Shinn & Khadduri, 2020). The higher cost occurs not only because the subsidies are ongoing but also because families participate in this type of assistance program at a very high rate. Research has shown that families accept other types of interventions much less frequently; in fact, most families did not respond to the offer of services in HF/RRH and transitional housing programs (Evans et al., 2019; Gubits, Shinn et al., 2018b).

Vaclavik et al. (2018) found the impact of HF/RRH may be enhanced when at-risk parents receive housing subsidies combined with a variety of financial and housing stabilization services. This intervention resulted in an 18.8% return to shelter rate for at least one member of the family, with an average re-entry time of 16 months.Footnote 10 It was most successful with families who had fewer children or who had received funds for rent and moving expenses. However, because the HF/RRH assistance is temporary, Vaclavik et al. argue that families need a range of services for parents to, for example, gain and maintain a job. As HF/RRH is practiced now, Shinn and Khadduri (2020) conclude that the data fail “to show that rapid re-housing or transitional housing is any better than usual care at ending homelessness…none of these interventions have the sort of radiating benefits that long-term subsidies provide for families” (p. 118).

Primary Prevention Programs for Family Homelessness

Controlled investigations of primary prevention programs designed to avert family homelessness are limited to two studies that evaluated Homebase (HB), a community-based initiative in New York City. This program connected at-risk families who came to their neighborhood HB facility for services addressing their specific housing vulnerabilities, such as conflicts with landlords, legal issues, financial pressures, mental health and substance abuse concerns, childcare problems, and employment. Goodman et al. (2016) found that the program reduced shelter entry by only 5–11%, perhaps because, as in the Family Options Study, these services were accessed infrequently or only in times of crisis. These findings reinforce data from an earlier, smaller study by Rolston et al. (2013) that showed HB reduced shelter entry by 6.5%, though the decrease in the length of shelter stays was a more robust 70%.

Homebases’ sparse participation rate and limited impact on shelter entry raises the question of what factors prompt at-risk families to seek out support programs that might avert homelessness. Vaclavik et al. (2018) reported data from an uncontrolled study of a primary prevention program in Indianapolis for families “at imminent risk” of homelessness. Families were offered a “menu” of housing and financial services that were utilized significantly. Among participants, only 10.9% of families had a member enter a homeless shelter with an average entry time of 22 months after program completion.Footnote 11 Prevention was compromised when the family included younger children or a military veteran. The program was most successful when families received rental and utility financial support, included older adults, were involved in the prevention program for a longer time period, and did not require legal services.

Thus, the research indicates that primary as well as secondary prevention programs for family homelessness are more effective when they include financial support of some kind in addition to longer-term housing subsidies. These few studies of primary prevention programs also suggest that financial assistance will be most impactful when it is coupled with additional time-sensitive supportive services that link families, especially young heads of households with young children, to community resources such as landlord–household mediation, short-term financial assistance, case management, and legal assistance.Footnote 12

Implications of the Limited Data

These housing stability findings should be considered in the context of research that shows poverty itself has a negative impact on psychological functioning. Shinn and Khadduri (2020), in fact, argue that the central cause of homelessness are systemic inequities in income, wealth, societal exclusion, and social justice that result in a lack of affordable housing for those living in “deep poverty.” Footnote 13,Footnote 14 Because poverty-related stresses, uncertainties, and aversive consequences may challenge coping abilities (Mani et al., 2013; Shilbach et al., 2016),Footnote 15 permanent subsidies may derive part of their benefit by reducing the cognitive and behavioral demands attendant to managing persistent financial shortfalls. In contrast, the offer of temporary time-regulated services in HF/RRH programs may result in increasing the ongoing overstimulation rather than functioning as a positive inducement. Thus, Bassuk et al. (2020) emphasize that focusing on housing affordability to prevent family homelessness is a narrow interpretation of the Family Options Study because a wide range of services in addition to housing subsidies are needed to achieve housing stability.

Shinn and Cohen (2019) also argue that “policy and practice…[must] ensure that supportive services intended to help people maintain stable housing [be] easily accessible and targeted to address specific needs of people in the community” (Implications for Policy section, bullet 2). The NAEH concurs: “rapid re-housing...does not provide all the services that families need...it provides families what they need to exit homelessness, but does not provide everything that families need [to stay housed]...[M]ore services are needed to support families than homeless programs can provide on their own” (2016b, para. 2; NAEH, 2017). Of course, services that are available must be voluntary and based on consumer-informed desires.

Despite the consensus that there is a “need for future research on...interventions that identify unique service needs among families who are experiencing housing instability or homelessness” (Vaclavik et al., 2018, p. 591; see also Bassuk & Rosenberg, 1988; Shinn et al., 2013), such investigations of HF/RRH have not been undertaken. Instead, HF/RRH has been characterized by a mismatch between family needs and homeless service initiatives (Greer et al., 2016; Shinn et al., 2017) resulting in “homeless system interventions [that] systematically screen out families with housing and employment barriers, despite the presumption that these families are the families who need interventions in order to achieve housing and economic stability” (Shinn et al., 2017, p. 293).

Thus, while HF/RRH is the current best practice secondary prevention model, the data indicate that this approach has a modest impact on reducing family homelessness. We suggest that CBS analyses of HF/RRH have the potential to improve their outcomes for unhoused and at-risk families.

Culturo-Behavior Science Analysis Via Behavioral Community Psychology

Culturo-behavior science is emerging as a leading way for behavior scientists to engage in social and cultural systems analyses (Cihon & Mattaini, 2020). An early forerunner of this approach was Behavioral Community Psychology (BCP; cf. Jason et al., 2021; Watson-Thompson et al., 2020), which combines behavior analysis’ experimental rigor and focus on behavior, environmental contingencies, and effective intervention with community psychology’s collaborative research strategy and emphasis on systems analysis, prevention of social problems, and advocacy for societal-level change (Suarez-Balcazar et al., 2019). BCP research is community-informed and quantitatively evaluated; intervention tools include behavior analysis, systems analysis, behavioral capacity training, contingency modification, and advocacy for structural change. Importantly, both of BCP’s germinal disciplines view the environment as the source of missing or ineffective behaviors and reject explanations that attribute the problem to personal failure.Footnote 16

BCP research and intervention embrace a set of core values, including establishing community and consumer collaborations and addressing functional relationships that are important to the stakeholders; conducting research that is practical, reflects community and consumer concerns, and uses appropriate measures to capture the dynamic interaction between behavior and environment; developing and disseminating sustainable interventions that are owned by and maximize the benefits for the community and consumers; communicating successfully with community stakeholders; and fostering important social change (cf. Fawcett, 1991). BCP strives to enhance social justice through community empowerment, political activism, and the promotion of collective wellness (Suarez-Balcazar et al., 2019) – all of which comport well with the CBS approach to social change.

Addressing Family Homelessness through Behavioral Community Psychology

As noted earlier, the BCP approach views family homelessness, and indeed all homelessness, as a product of environmental and systemic variables at several levels of analysis – not as the outcome of personal deficiency or disorder. It recognizes that family homelessness in the United States is embedded in an exceptionally complex sociocultural and economic context, as befits a “wicked problem.” BCP promotes research and advocacy for change at all levels of society utilizing behavioral and systems analyses that elucidate functional relationships affecting housing stability outcomes.Footnote 17 One systems analytic tool that has proven increasingly useful in CBS and could find similar utility in BCP is the metacontingency (Glenn, 2004; Glenn & Malott, 2004; Houmanfar et al., 2010).Footnote 18

The Metacontingency as a CBS Analytic Tool

The metacontingency has emerged as a leading CBS mechanism to address structural social problems because it offers the breadth and complexity to elucidate cultural, sociological, and ecological factors that shape the behaviors of multiple persons interacting in groups. Houmanfar et al.’ (2010) “elaborated” metacontingency is especially useful. It specifies five elements: cultural milieu, interlocking behaviors, aggregate product, impact on societal practices, and feedback to the cultural milieu.

The cultural milieu includes historical, sociological, psychological, economic, and geographical factors, all of which may function as antecedent stimuli that affect both motivation and behavior. The milieu factors include structural and institutional rules and policies that specify the contingent consequences that select the interlocking behaviors of the participants in the system, such as the interactions between individuals in a service agency. These interlocking behaviors result in aggregate products, which are the outputs of the organization or system. Ideally, aggregate products then impact consumer or societal practices that relate to or use the products. Finally, feedback from modified societal practices to policy-makers can prompt new rules that meet cultural practice demands and yet remain consistent with the values of the cultural milieu (Houmanfar et al., 2009).

Schematically, the 5-factor metacontingency can be laid out as follows, with each element’s functional impact on the next element represented by arrows:

  • Cultural milieu → Interlocking behaviors → Aggregate products → Impact on societal practices → Feedback to institutional and cultural decision-makers → Cultural milieu

Metacontingency Analysis of HF/RRH Policy

An examination of the current HF/RRH family homelessness metacontingency reveals contextual factors that are both functional and amenable to change – to achieving “small wins” that may improve housing stability outcomes. It suggests feasible points where BCP interventions may be beneficial, and where changes can be implemented. The goal in this case is to modify the metacontingency so that it selects interlocking behaviors of clients and service providers that improve the aggregate product – rapid housing stability – while exerting a positive influence on cultural and social practices.

The Functional Impact on Families of the Current HF/RRH Metacontingency

We begin our metacontingency analysis with the cultural milieu – the laws, policies, and regulations that shape the HF/RRH experience. The current HF/RRH metacontingency prompts and reinforces interlocking behaviors of clients and service providers that are overwhelmingly focused on the immediate goal of securing housing. It requires that housing occur as quickly as possible regardless of the families’ demographics and composition, context in which they live, psychosocial and health problems, skill competencies and challenges, and the availability of services and resources to address their issues. Most centrally, as discussed earlier, the HF/RRH timeline presents a more imposing burden for families, especially those with young children, compared to individuals. As we saw with the description of the HF/RRH system’s demand on families experiencing homelessness, the metacontingency introduces a barrage of high-alert discriminative stimuli in the context of numerous already ongoing high intensity stressors to which family members must respond. Because the institutional milieu is part of the larger cultural context, service providers also find themselves subject to numerous high intensity discriminative stimuli, often in the form of rules that signal the approach of aversive contingencies. Thus, the metacontingency establishes the conditions for strong delay discounting, selecting interlocking crisis-solving behaviors associated with getting housed instead of selecting interlocking behaviors that increase the likelihood of staying housed.Footnote 19

The interlocking behaviors selected by the HF/RRH metacontingency produce an aggregate product – rapid housing stability – that is comparable to both transitional housing and to “usual care” outcomes, albeit with a 10% cost saving compared to being housed in a transitional shelter (cf. Gubits, Shinn et al., 2018b). However, HF/RRH as currently practiced has failed to spur institutional, social, or cultural demands for change, partially because its constituents are not given a voice in its application and partially because programs such as HF/RRH address an uncomfortable and minimally visible social problem. As Shinn and Khadduri (2020) conclude, the existence of homelessness in a wealthy society “is a choice, not a choice by people sleeping on the streets but a choice by the rest of us to look the other way” (p. 176). Thus, there has been an absence of feedback to decision-makers, resulting in the continuation of a static system that primarily selects interlocking crisis resolution behaviors that produce rapid housing more than longer-term stable housing.

Modifying the HF/RRH Metacontingency for Families

We examine four potential points of intervention in the metacontingency, discussing them in the order we hypothesize changes are likely to be most feasible: the broad cultural milieu (Intervention Point 1), interlocking behaviors of clients and service providers (Intervention Point 2), disseminating aggregate product data in community, popular media, and scientific outlets (Intervention Point 3), and finally, providing feedback to decision makers (Intervention Point 4).

Intervention Point 1: Advocating with Cultural Milieu Participants for Collaborative Research

We intervene first with cultural milieu and institutional factors because reducing family homelessness is not a social priority (cf. Shinn & Khadduri, 2020). There has been little if any feedback to legislators and regulators regarding the actual functional impact and behavioral products of the HF/RRH metacontingency. We suggest that efforts to modify the HF/RRH metacontingency begin with advocacy strategies directed toward the participants who enact, implement, and respond to societal expectations and demands. Change initiatives, of course, will include persons who have experienced homelessness as well as decision makers. Given the dearth of research on family homelessness, advocacy efforts will need to commence with theoretically and empirically sound analyses of current practices.

Researching Metacontingency Changes

Three promising candidates for changes in the current HF/RRH family homelessness metacontingency that might improve housing stability outcomes are (a) increasing the elasticity of the allowed time frame to secure rapid housing, (b) permitting house-sharing by unrelated families, and (c) offering a range of opportunities designed to prevent future episodes of homelessness through empowerment and “capacity building” (Fawcett, 1991). Each of these changes has the potential to make the HF/RRH metacontingency more family-friendly with a focus on establishing a “home,” not just locating a house.

The first modification addresses the crux of the issue: the current HF/RRH metacontingency does not recognize that moving from homelessness into housing is considerably more complex for families with young children compared to a single person, even with a behavioral health problem. We are not suggesting a return to previous practices, but that the HF/RRH time frames for housing location be relaxed to some degree to permit elasticity based on each family’s unique needs. The impact of a timing modification can be evaluated directly. As described earlier, the failure of a family to secure housing in the specified time frame often results in the loss of HF/RRH subsidies. An elastic time frame without this fear may result in improved housing stability, as clients and service providers are able to organize the move more deliberately. The housing search process can become one characterized by positive reinforcement rather than by the negative reinforcement that is the current motivator.

A second proposed change at Intervention Point 1 builds on Vaclavik et al.’s (2018) findings that prevention efforts are enhanced by having housing stability and financial services available and utilized once families are housed. We emphasize that the BCP perspective – which combines the environmental determinism of behaviorism (Skinner, 1971) with the systemic determinism of community psychology (Jason et al., 2019) – rejects an interpretation of Vaclavik’s data that suggests homelessness results from personal deficiencies or a psychological disorder that services can rehabilitate. On the contrary, the data suggest only that the previous learning experiences of unhoused and at-risk parents in coping with inequitable, discriminatory systems may have left room for the refinement of counter-control strategies and tactics. Therefore, it is reasonable to hypothesize that housing stability may be improved if parents are offered opportunities to enhance response repertoires that may result in acquiring more reinforcement even within the current social, economic, and cultural context.

BCP research is needed to determine whether such learning opportunities will enhance stability among currently unhoused or at-risk families. The demographic and mental health risk factors reviewed earlier offer a slim basis from which to hypothesize which competencies may increase housing stability. However, the challenges at-risk families face suggest that housing stability might be improved when parents are able to impact their environment more effectively to produce reinforcement. CBS research can examine whether the aggregate product is enhanced by exposure to skill sets called “System 2 thinking” (Kahneman, 2011),Footnote 20 stress management, parenting, self-management, problem-solving, decision-making, social, communication, assertiveness, job search and interviewing, literacy, and financial management (cf. Bassuk et al., 2020; Goodman et al., 2016; Shinn et al., 2017).

After research has identified which behavioral and cognitive response repertoires typically are most helpful for achieving housing stability, behavior analyses (Kanfer & Saslow, 1969) can clarify where an individual’s learning opportunities lie and offer them appropriate services and programs. However, individualized pre-move behavior analytic assessments of each family’s capacity status, collaborative plans for enhancing stability, and navigation of services are possible only if the time frame for securing family housing is somewhat elastic and flexible in terms of each family’s needs.Footnote 21 Whether outcomes improve with behavioral momentum toward utilizing “stability-enhancing” services prior to moving into rapid housing is another area ripe for BCP research.

A third modification at Intervention Point 1 might be to develop assessment and support structures for unrelated families to choose to be housed together.Footnote 22 The potential advantages of co-housing include cost-sharing, social support, increased childcare resources, and improved work attendance. Psychological gains also may accrue if a single parent feels a greater sense of community integration and belonging (Nemiroff et al., 2011) or if one or both is a positive model for the other. Research is needed to identify the family characteristics and behavioral repertoires most important for positive co-housing outcomes (e.g., complementary skills and compatible reinforcers, motivating operations, and values). Child factors, such as the number, ages, and behavior assets and problems may also be important. A behavioral analysis can assess appropriateness for shared housing; while only a small minority of at-risk parents may be good candidates, careful matching may avert the interpersonal and landlord conflicts found with current “doubling up” practices (Bush & Shinn, 2019; Shinn et al., 2013). An additional support might be the creation of an Ombudsperson to mediate between co-tenants and between co-tenants and landlords.Footnote 23

Involving Stakeholders in Research on the Metacontingency

Culturo-behavior scientists, as discussed earlier, focus on identifying situationally effective outcomes instead of seeking absolutistic or ideal ones (cf. Fawcett, 1991; Hayes et al., 2012). From the BCP perspective, identifying goals and service needs that are both pragmatic and socially valid requires participation from the onset of the research project by all stakeholders, including persons who have experienced homelessness, in line with the current guideline of “nothing for me without me” (Charmelus, 2017). Examples of inclusive researcher collaboration in the homelessness field include participating in grassroots organizations, engaging with clients in legislative advocacy, creating consumer task forces, hosting focus groups with providers and clients, and incorporating perspectives on affordable housing by all stakeholders such as clients, residents, landlords, and politicians (Charmelus, 2017). Unfortunately, this collaborative research ideal is achieved too infrequently due to the intensive time and effort it entails from researchers (Burgess & Bachelder, 2020), the ambivalence and wariness of historically marginalized people to join mainstream initiatives (Charmelus, 2017), and the significant number of resources and amount of leadership support that is needed (Ni She & Harrision, 2021).Footnote 24

Collaboration with decision-makers is necessary to implement research projects. In fact, Brown et al. (2020) recommend that “researchers first identify…policy stakeholders and financial and political pressures that may impact intervention…[Such] collaboration with policymakers may shape research aims to better reflect their priorities, thereby facilitating stronger investment in the project” (p. 209–210). While it is debatable whether policy makers should be the first to be brought into the collaborative endeavor, their early involvement in research design is essential if they are to understand the actual functional impact of the current family homelessness metacontingency and be willing to explore potential modifications to it. With buy-in from the decision-makers, the research findings from demonstration projects can be drivers of policy, guiding them to modify other concerns with the HF/RRH family homelessness metacontingency and the interlocking behaviors it selects (cf. Burgess & Bachelder, 2020). Note that this research focus on the HF/RRH metacontingency stands in contrast to the Homelessness Research Institute’s (2020) emphasis on evaluation of diversion programs, HF variables, and doubling up of unscreened families.

Intervention Point 2: Supporting the New Interlocking Behaviors of Clients and Service Providers to Improve the Aggregate Product of Rapid Housing Stability

A family homelessness metacontingency modified as suggested at Intervention Point 1 would be more likely to select interlocking behaviors of clients and service providers that go beyond securing rapid housing to those associated with becoming stabilized in a home and preventing further episodes of homelessness. Specifically, service providers could use behavior analyses (Kanfer & Saslow, 1969) to clarify an individual’s strengths and challenges as well as to identify environmental influences on their responses.

The behavior analysis’ firm empirical base for client assessment (Kazdin, 2013) stands in contrast to “existing [HF] assessment tools [that] do not have a strong evidence base and are limited in their ability to select the best interventions for families and individuals or to predict which families would be the most successful in different interventions” (Shinn & Khadduri, 2020, p. 109). Further, unlike current assessments that “screen out” candidates (e.g., approximately 50% of applicants to HF/RRH programs; Shinn et al., 2017), the behavior analysis is intended to “screen in” all candidates by identifying the services and resources each needs to prevent further episodes of homelessness.Footnote 25

Moreover, because research suggests that clients access few of the homelessness prevention services available to them (Goodman et al., 2016), the behavioral analysis can serve not only as an important tool for developing client-centered homelessness prevention plans but also for increasing participation in these plans.Footnote 26 An essential element is the inclusion of clients themselves through collaborative planning. The provision of antecedent supports that reduce response effort and increase positive reinforcers that enhance the worth of the effort also may improve participation, especially if resources exist to individualize the supportive stimuli. For instance, unhoused parents who participate in stability enhancing sessions could receive vouchers for round trip transit or parking and for child-care in the home from a list of approved neighborhood providers. They could also receive reinforcement at a training site via appealing refreshments and the opportunity to select one of several tangible objects currently available in the “program store” (e.g., lottery ticket, t-shirt, points toward a ticket to sports or entertainment events, vouchers for family restaurants and items necessary to establish a home such as small appliances, dishes, furniture, bedding, etc.). The costs of these programs and related environmental supports are likely to be substantial, though Bassuk et al. (2020) insist that “services matter and must be integrated with housing solutions that are effective and not driven by limited budgets” (p. 258–259). Fortunately, a new source of funding is available now: HUD’s allocation of $5 billion from the American Rescue Plan to reduce homelessness that includes money for “fund[ing] supportive services” (Eaton, 2021, para. 6). In addition, costs can be offset partially by the 10% savings achieved by HF/RRH efforts compared to transitional housing programs. This cost reduction in service provision is a component of the current aggregate product, to which we now turn.

Intervention Point 3: Evaluating the New Aggregate Product and Disseminating Findings

The spectrum of potential benefits that may result from changes in the HF/RRH metacontingency proposed at Intervention Points 1 and 2 comport well with the BCP emphasis on prevention. Research is necessary to confirm whether the modified metacontingency selects interlocking behaviors that improve the aggregate product – in this case, gains not only in increased housing stability but also perhaps in other measures related to societal health, such as family members’ physical and mental wellness, employment, income, and child, school, and legal stability, etc. Greater housing stability also may increase psychological integration, or a sense of belonging in a community, by providing more opportunities to socialize with neighbors, participate in community events, and build on the greater sense of integration that single homeless women with children already feel compared to peers without children (Nemiroff et al., 2011).

In addition, to affect sustainable change in the cultural milieu, data supporting improved aggregate products must be disseminated to the public in understandable ways that highlight their benefits (Fawcett, 1991). This may entail activities such as making school and community presentations, writing op-ed columns, and hosting podcasts and discussions on social platforms and on Internet blogs. A welcome outcome of this activism, for example, would be a reduction in the opposition many people express to locating supportive housing in one’s immediate community (“NIMBY-ism” or Not in My Back Yard; HUD, 2021b). All of these raise further questions worthy of serious investigation.

Intervention Point 4: Advocating for Policy Change

A key thrust of the BCP approach is to bring the data and their implications to the attention of decision-makers (Fawcett, 1991), ideally in collaboration with clients (Charmlus, 2017). This advocacy is intended to influence them directly and to build momentum among the general public to extend empirically supported changes. These alterations in the cultural milieu bring us back to where we began at Intervention Point 1 with culturo-behavior scientists adopting a research-advocacy role. Now, however, inclusion of homeless families and the rallying of the general public to influence decision-makers represents the evolution of a cultural practice – one that is more responsive, empathetic, and effective. Advocating for such “small wins,” of course, does not obviate the need for addressing the systemic injustices that lead to homelessness for families (and individuals) – it may even highlight the urgency in fostering the fundamental structural changes that US society at present chooses to ignore (cf. Shinn & Khadduri, 2020).

Conclusion: Lots of Research and Advocacy Opportunities for Behavioral Community Psychologists

We have suggested that CBS can utilize the BCP framework to improve HF/RRH secondary prevention efforts aimed at the seemingly intractable “wicked problem” of family homelessness. Our analysis indicates that the HF/RRH metacontingency selects crisis-energized housing-location behaviors by both clients and service providers to the detriment of the longer-term goal of family housing stability outcome measures. If research confirms that a modified HF/RRH metacontingency produces gains in housing stability, while still adhering to the core HF/RRH philosophical tenets, it would represent a “small win” – but an important one to the actors in the system.

These proposed changes raise numerous research questions regarding the factors associated with HF/RRH practices and outcomes. However, while research is a defining characteristic of CBS, the BCP approach goes a step further by also embracing advocacy as a core strategy to address systemic variables that produce vexing social problems (Fawcett, 1991; Jason et al., 2021). In the case of family homelessness, advocacy to change the large-scale system inequities that ultimately are responsible for the problem can coexist with advocacy that employs sound data and scientific theory to help decision makers understand the actual impact of the currently operative HF/RRH metacontingency. This may lead to improved housing stability for families who are struggling with housing insecurity now and for homeless prevention in the future. That is, because research and advocating for changes in the HF/RRH metacontingency dovetails with advocacy efforts targeting systemic inequities such as inadequate housing subsidies, insufficient affordable housing options, and social and economic policies that fail to address the root causes of modern homelessness. BCP offers culturo-behavior scientists a new and promising way to contribute meaningfully on multiple levels and in a range of ways toward resolving the “wickedness” of the contemporary family homelessness.