Skip to main content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Cityscape. Author manuscript; available in PMC 2018 Jan 9.
Published in final edited form as:
Cityscape. 2017; 19(3): 293–307.
PMCID: PMC5760190
NIHMSID: NIHMS927088
PMID: 29326756

Mismatch Between Homeless Families and the Homelessness Service System

Marybeth Shinn, Scott R. Brown, Brooke E. Spellman, Michelle Wood, Daniel Gubits, and Jill Khadduri
Author information Copyright and License information Disclaimer

Abstract

The enrollment phase of the Family Options Study provides information about the mismatch of the homeless service system and the needs and desires of families experiencing homelessness in 12 communities. One-fourth (25.8 percent) of the 2,490 families screened for the study after shelter stays of a week were deemed ineligible for one or more of the interventions at initial screening, with ineligibility highest for those screened for transitional housing programs (28.9 percent) and lower for short- and long-term rental subsidies (9.2 and 4.1 percent). Families given priority offers of housing and service interventions for which they appeared eligible faced additional screening by programs and made decisions about whether to enroll. Considering all stages of this process, families were least likely to be eligible for and subsequently choose to enroll (within 9 months) in transitional housing programs (32.5 percent of those initially screened) and most likely to be eligible for and subsequently lease up with long-term subsidies (73.4 percent) with short-term subsidies in between (51.0 percent). Homeless system interventions 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.

Introduction

The U.S. Department of Housing and Urban Development (HUD) requires communities that seek federal funding for homeless services to organize a Continuum of Care—a “community-based homeless assistance program planning network … intended to increase the efficiency and effectiveness of coordinated, community-based systems that provide housing and services” to people experiencing homelessness.1 The term “Continuum of Care” is used in a variety of contexts, mostly within health systems, to emphasize connections among components of care to assure that no patient is lost to followup (Kerber et al., 2007; Lefkovitz, 1995; McBryde-Foster and Allen, 2005). The idea that care will be available to all members of the community is generally implied, and HUD “defined the minimum planning requirements for a Continuum so that it coordinates and implements a system that meets the needs of the homeless population within its geographic area” (emphasis added).2 In this article, we use the term “homeless service system” because, in the context of homelessness, the term Continuum of Care is used to refer to the governance structure for the system as well as the system, the geography it covers, the grants that fund it, and sometimes a philosophy in which participants earn their way to increasing independence by successfully meeting system requirements.

This article uses data from the enrollment phase of the Family Options Study to examine how well the homeless service system in 12 sites around the United States meets the needs of families experiencing homelessness. The fact that more than half a million people in families experienced homelessness during the course of 2015 and have done so every year but 1 since 2008 (Solari et al., 2016) suggests at a minimum that the homeless service system faces resource constraints. However, our question is different. We examine the fit between homeless assistance programs and the families they intend to serve, with respect to programs’ capacity and willingness to accommodate families and families’ willingness to participate in programs they are offered. We also observe the reasons that each party sometimes rejects the other. More specifically, we catalogue the criteria that programs use to screen families for admission, the extent to which those criteria serve to exclude families living in homeless shelters, and the extent to which families enroll in programs after being given a priority offer of an open slot reserved for them. Enrollment reflects a combination of additional screening of families by programs, decisions by families as to whether programs were acceptable, and, in the case of subsidies for the private rental market, families’ ability to find landlords who accept the subsidies and lease up. Findings predate the advent of coordinated entry systems now used by many communities to allocate people experiencing homelessness among programs but suggest some of the challenges facing those systems, why it is sometimes difficult to get programs to accept their allocations, and why families often turn down programs.

The homeless service system in most communities consists of four types of programs—emergency shelter, transitional housing, short-term rental subsidies, and permanent supportive housing. Emergency shelters are typically the point of entry to the system and the place where we recruited families for this study. Because all study families had spent at least 7 days in an emergency shelter at the time of enrollment, we have limited information about the extent to which families were able to access this first step in the homeless service system. Permanent supportive housing—subsidized housing with supportive services—is intended to help people with disabilities to live independently in the community. Because its target clientele is limited to individuals and families with disabilities, the Family Options Study did not include it. However, in addition to transitional housing and short-term rental subsidies, the study did include long-term rental subsidies without supportive services largely in the form of housing choice vouchers (HCVs). Thus, the study also provides insights into a particularly vulnerable group of families’ eligibility for and ability to use HCVs.

In the remainder of this introduction, we describe the three interventions compared with each other and with usual care by the Family Options Study, focusing on their role in the homeless service system and the families they might be expected to serve.

Interventions Studied

At the time that families enrolled in the study in 2010 and 2011, transitional housing was a mainstay of the homelessness assistance system, with more than 200,000 beds including both individuals and families nationwide—only slightly fewer than the number of beds in emergency shelters. The stock of transitional housing has decreased by one-fourth since then, but, with 75,599 transitional housing beds dedicated to families in 2016, the model remains an important part of the service system (Henry et al., 2016). Families receive a place to stay for up to 2 years, along with case management and an array of supportive services designed to address barriers to housing and to help them become self-sufficient. Most families experiencing homelessness are homeless only briefly and do not return to shelters (Culhane et al., 2007). Transitional housing programs evolved to help families who were unable to leave shelter on their own and needed more assistance to cope with trauma, overcome mental health or substance problems, develop job skills, and so on. Thus, one might assume that transitional housing would serve the neediest families and be attractive to families they were designed to serve. Across the multiple transitional housing programs in any community, one would expect most families to be accommodated, although programs often specialize in serving families with particular characteristics. To provide greater contrast between transitional housing and short-term rental subsidies in the study, only project-based or (to a lesser extent) scattered-site transitional housing programs were included; transition-in-place programs that allow for families to live in community-based rental housing and remain there at the conclusion of the program were not included.

Short-term rental subsidies became widely available on a temporary basis when Congress appropriated $1.5 billion for the Homelessness Prevention and Rapid Re-Housing Program (HPRP), as part of the American Recovery and Reinvestment Act (ARRA) of 2009,3 although the model has since grown to supply 56,589 beds for families in 2016 (Henry et al., 2016). Families received short-term and sometimes declining rental subsidies along with limited case management focused on finding housing and achieving economic self-sufficiency. Families could receive assistance for up to 18 months, although the programs to which families in this study were assigned lasted 6 to 8 months on average. The aim was to provide only enough assistance to help families become re-housed and enable them to pay the rent after the program ended, combined with light case management focused on housing and employment. HUD suggested but did not require communities to use HPRP funds for households that would be able to pay for housing on their own once the assistance ended, leading some communities to focus eligibility on households with stable employment histories or future prospects (Cunningham et al., 2015).

The third intervention—long-term rental subsidies without dedicated supportive services—is neither part of the homeless service system nor ordinarily available to families in shelters, unless they have previously secured places on waiting lists. Rental subsidies are not an entitlement, and only one in four households that are eligible for them by reason of income do not receive them (JCHS, 2016). Public housing agencies (PHAs) that agreed to allot HCVs to the study provided most subsidies in our study. Vouchers cover the difference between 30 percent of families’ income and a housing unit’s rent and utilities.

Past studies show that such subsidies are successful both in preventing homelessness (Wood, Turnham, and Mills, 2008) and ending it (Cragg and O’Flaherty, 1999; Shinn et al., 1998; Wong, Culhane, and Kuhn, 1997). Further, the experimental results of the Family Options Study show that long-term subsidies resolved homelessness, enhanced other forms of residential stability, and had positive radiating impacts for other aspects of life for both parents and children (Gubits et al., 2016, 2015). However, this study asks whether families who have been homeless in shelters at least a week meet minimum eligibility requirements for PHAs, which check for criminal backgrounds and exclude undocumented households and whether, when given vouchers, families are able to use them by leasing up private-market housing units. Past studies show that about one-third of families issued vouchers (most of whom are not experiencing homelessness) do not use them. Documented reasons for low lease-up rates include difficulty in locating landlords who will accept vouchers and getting units inspected and approved in the time allotted (Bacon, 2005; Basolo and Nguyen, 2005; Daniel, 2010; Edin, DeLuca, and Owens, 2012; Finkel and Buron, 2001; Gubits et al., 2009; Jacob and Ludwig, 2012; Kennedy and Finkel, 1994; Mills et al., 2006; Sard, 2001; Sterken, 2009), challenges one might expect to be exacerbated for families experiencing homelessness. Alternatively, such families might lease up at higher rates, given their strong motivation to achieve stable housing. Finkel and Buron (2001) observed higher lease-up rates among the lowest-income renters, who had the most to gain from using vouchers.

Methods

Families enrolled in the Family Options study from September 2010 to January 2012, with dates varying by community.4 We screened 2,490 families who had been in one of 57 homeless shelters for at least 7 days to determine their eligibility for programs that had slots available at that time. We screened families at times when programs representing at least two of the interventions—transitional housing, short-term subsidies, and long-term subsidies—had openings in their community. Near the end of the enrollment period in some communities this availability requirement was relaxed to one intervention, which could still be compared with usual care.

This article is based on five sources of data from the enrollment period and shortly thereafter. First, as part of the enrollment process, we monitored the availability of interventions by calling participating programs regularly to determine whether they had openings. Across all sites, 51 transitional housing programs, 18 long-term subsidy programs, and 28 short-term subsidy programs participated in the study.

Second, we screened families to determine their eligibility for available programs, based on criteria the programs supplied. This process mimicked the usual, although less-structured, process for families attempting to find housing, where eligibility criteria for programs that have no openings also are irrelevant. Eligibility questions took this form, varying according to the specific program rule—

Some programs will only accept families in which the head of household is clean and sober and who can demonstrate at least 30 days of sobriety. Would you like to be considered for programs with this requirement? (Gubits et al., 2013)

Thus, we have data on the extent to which families were screened out of particular programs or an entire intervention, because they reported themselves ineligible for all programs that had openings in that intervention at the time of screening.

Third, we have data from the study’s intake interviews with the 2,307 families who passed the screening and were randomly assigned to an intervention.5 These data include information about family members living elsewhere and provide a window into the extent to which the 57 participating shelters were able to accommodate families in their entirety.

Fourth, we have multiple types of data about 2,282 families’ enrollment in the programs to which they were randomly assigned. Although we screened families based on program-specific questions, we did not verify family responses before making assignments to programs. Programs conducted their own intake processes and could reject families who did not meet their criteria. For example, we asked families about criminal convictions, but PHAs conducted criminal background checks; we asked about sobriety, but transitional housing programs might require drug testing. Families could also drop out of the process, declining to accept the programs they were offered. Information on enrollment was obtained by calling programs to determine whether they had openings (to the end of August 2012 only), information from the HUD Public and Indian Housing Information Center, or PIC, and the HUD Tenant Rental Assistance Certification System, or TRACS, about leasing up with long-term subsidies, along with tracking surveys and a 20-month followup survey with families. All these data sources were combined into a data set on program usage.6

Finally, we have data from qualitative interviews with a nonrandom subsample of 80 families from four sites about their reasons for taking up or declining intervention programs.

Results

We report first on the extent to which enrollment was constrained because programs representing an intervention did not having an opening at the time a family entered the study. Then we examine the percentage of families who were screened out of programs and entire interventions and the reasons for ineligibility. Finally, using both quantitative and qualitative data, we explore the fact that many families referred to interventions did not participate.

Effects of Availability

Access to interventions in the study communities was quite constrained by lack of availability of slots in participating programs. In all, 53.8 percent of the 2,490 families screened for the study lost access to at least one intervention, because no openings were available or expected within the next month in any program representing that intervention at the time of the screening. This figure does not include families who did not have access, because transitional housing and long-term subsidies were never available to the study at their site.

Effects of Ineligibility

Exhibit 1 shows the percentage of families who, by self-report, failed to meet specified eligibility criteria for at least one program that had an opening in the appropriate geographic locale, and the percentage who lost access to a program, or to an entire intervention, as a consequence. Numbers do not sum to 100 percent, because a family could fail more than one criterion. More than one-fourth (25.8 percent) of the 2,490 families screened for available interventions failed to pass the eligibility screening for at least one, including 183 families who were screened out of the study altogether.

Families were most likely to lose access to transitional housing programs—50.9 percent of those screened lost access to at least one transitional housing program, and 28.9 percent did not meet the criteria for any transitional housing program with an opening at the time they were screened. This percentage varied widely from site to site, with only 1 percent of families in Honolulu and 2 percent in Salt Lake City excluded from this intervention compared with 60 percent in Connecticut and 93 percent in Minneapolis. Because the transitional housing programs offered units in agency-controlled housing (either project based or scattered site), families needed to have appropriate size and composition for the available units—a large family could not fit into a studio, a mother with an infant would not be eligible for a two-bedroom unit, and families with fathers or teenage boys were often excluded. More than one-fifth of families failed to meet the employment or income requirements imposed by a transitional housing program. For example, the minimum income category includes programs that required monthly incomes of $500, $1,000, $1,200, and even $2,000. No family was excluded for having too much income. Questions about sobriety were frequently asked but rarely failed (2.6 percent). That is, families were happy to be assigned to programs that required them to verify their sobriety. Conversely, few programs required family heads to have a mental health issue or other disability or to be in recovery from substance use, but this requirement excluded somewhat more families (5.9 percent). Not having lived in the municipality prior to entering shelter or not having family, employment, or residential ties to the area excluded some families. Despite their mission to assist families who need additional services to escape homelessness, transitional housing programs were more likely than programs that did not provide psychosocial services to exclude families for lack of income, substance issues, poor credit history, and lack of education or work experience. Few families lost access because they declined to participate in mandatory services.

Families were less likely to lose access to short-term rental subsidies (13.3 percent of those screened lost access to at least one short-term rental subsidy program and 9.2 percent lost access to the entire intervention), but most frequently did because they did not meet minimum income requirements. A history of housing instability or lack of education and work experience also excluded a few. Site-to-site variation was less extreme than for transitional housing but bimodal. One-half the sites had ineligibility rates of 3 percent or less, and the other one-half between 10 and 20 percent.

Families were least likely to fail the screening questions for long-term rental subsidies (7.0 percent lost access to a specific program and 4.1 percent to the intervention). The most common reasons were presuming they would fail the criminal background check, poor credit history, and lack of ties to the location. Unit-size restrictions applied to Connecticut, where one of the three providers used project-based vouchers with a resulting screenout rate of 10 percent; otherwise screenout rates ranged from 0 to 6 percent. Interestingly, very few families lost access to any of the interventions because of their immigration status.

We do not have information on the percentage of families who failed to get into the shelters where the study began; however, we can infer from the data that eligibility requirements similarly affected family members’ access to shelter. Of the 2,307 families who were ultimately randomly assigned to an intervention in the study, 10.1 percent had an adult partner living elsewhere and 23.9 percent had a minor child living elsewhere. Of the 80 enrolled families who participated in qualitative interviews, 7 reported separations from partners in shelter, 5 because of shelter rules regarding men and 2 for other reasons. Nine families separated from children on shelter entry or during a shelter stay so that the child could avoid shelter conditions. One family was separated from an adult child, and two others from minor children due to shelter rules, and in a third, three-generational family, the family split up with the grandmother taking one child and the mother the other so that both adults were eligible for a family shelter (Shinn, Gibbons-Benton, and Brown, 2015). It is possible that shelters screened families for other criteria such as immigration status or sobriety.

Effects of Program and Family Choice on Enrollment and Takeup

Based on the initial screening, 2,307 eligible families were enrolled in the study and randomly assigned to programs that had availability and for which the families appeared to be eligible. Referrals for the assigned interventions were provided to the family and communicated to the program, which reserved a slot for the family. Families were responsible for pursuing the referrals, and the programs conducted their normal intake process to verify family eligibility. Family preferences also affected program enrollment. Not all families pursued the referrals, in some cases because they suspected they would be found ineligible. It is difficult to distinguish exclusions by programs from families’ decisions not to pursue offers, not least because perceptions of programs and families are sometimes at odds. For example, during the time that study staff called short-term rental subsidy programs to ascertain availability of slots, program staff reported that only 9.3 percent of the referred families failed to meet eligibility requirements, but in qualitative data, 9 of 19 families assigned to short-term subsidies during this period reported that the program to which they were referred found them ineligible (Fisher et al., 2014). Families given priority offers of short-term rental subsidies or long-term HCVs also had to find landlords who would accept the vouchers for units that passed inspection and lease up.

Exhibit 2 shows the end results of the two stages of screening by programs (before and after priority offers were made) and of the choices made by families. Row A shows the number of families screened for each intervention, because slots were available, and Row B shows the percentage of families who passed initial screening prior to randomization. These same numbers are shown in or derived from exhibit 1. Row C shows the numbers of families randomly assigned to receive priority offers of each intervention, and Row D shows the percentage of families given priority offers who moved into units of the offered type within 9 months, based on program usage data. Although only a subset of the families screened received priority offers, the percentages in Rows B and D can be multiplied to give a percentage who passed both stages of screening, accepted programs, found units if required to do so, and moved in.

As shown in exhibit 2, only slightly more than one-half of the families given priority offers of transitional housing passed any additional screening by programs and chose to move in. In conjunction with the high screen-out rates prior to randomization, we estimate that only about a one-third of families who stayed at least a week in emergency shelter in the 12 jurisdictions both found transitional housing acceptable and were acceptable to transitional housing programs. Nearly three-fifths of families given priority offers of short-term rental subsidies passed additional screening and moved in, so we estimate that short-term subsidy programs and families were mutually acceptable about one-half of the time. More than three-fourths of families given priority offers of long-term subsidies passed any additional screening, found units if necessary, and moved in. Given the very low rate at which families were screened out of long-term subsidies, nearly three-fourths of families in shelter were acceptable to and made use of such subsidies. This figure includes some families given offers of public housing or project-based vouchers. Among families given priority offers of HCVs, where families had to find a landlord and unit, 82.3 percent passed PHA screening and leased up within by the 20-month followup (Solari and Khadduri, 2017).7 Even this figure underestimates the “success rate” among families to whom the PHA actually issued a voucher.

Qualitative Information Regarding Family Preferences

Qualitative interviews with 80 families (19 to 22 assigned to each of the three interventions or usual care across four sites) provide some insight into the reasons that families in the study may have turned down the interventions made available to them through the study in favor of staying in shelter or finding housing on their own (Fisher et al., 2014).

Location was a key issue that worked in favor of private-market leases and against project-based transitional housing. In the qualitative interviews, families indicated they wanted to live in familiar locations close to family, friends, jobs, transportation, and their children’s schools. Families offered short-term rental subsidies could live anywhere the subsidy enabled them to afford. Families offered HCVs were sometimes restricted for a year to the city or county served by the housing agency that issued the voucher but then could later move, or port, to another area of their choice. In contrast, families offered transitional housing, like those offered project-based subsidies, had to move into the project or unit to which they were assigned.

Other factors influenced decisions. Concern about time limits of short-term subsidies, along with uncertainty about how long they would last, engendered considerable anxiety for families. Some families felt that they could not generate the requisite income fast enough to retain the housing units after the program ended. Families sometimes felt that transitional housing was not a good environment in which to raise children, and several families had left early or planned to do so in search of better environments. Maintaining family integrity was important but sometimes competed with housing opportunities. One respondent turned down an offer of transitional housing because her partner was not allowed, but another respondent accepted a long-term subsidy despite the fact that her partner’s criminal conviction meant that he could not be housed with her and her children.

Families clearly preferred long-term subsidies and often expressed palpable relief at having their own place. Families encountered some of the barriers to using subsidies that are in the literature, but they persisted. For example, four families encountered landlords who did not accept their vouchers but ultimately found other landlords who did.

Discussion

The enrollment phase of the Family Options Study revealed a mismatch between the availability and targeting of homeless system resources and the characteristics and preferences of families in shelter who are intended to use those resources to return to permanent housing. Policymakers should consider the ways in which the mismatch affects communities’ ability to assist homeless families.

It is important to understand our results in the context in which they were obtained. Participants in our study had access to resources not usually available to families experiencing homelessness. With the cooperation of PHAs, the study made long-term rental subsidies, typically in the form of HCVs, available to the experiment, and ARRA made short-term subsidies more available than they had been previously.

Despite these resources, one-half (53.8 percent) of the families who were willing to participate in the study lost access to at least one intervention due to availability. Access to transitional housing, the intervention of the three that was typically available for homeless families at the time of the study, was particularly constrained; 26.2 percent of families in communities with project-based transitional housing could not even be screened for it, because no units were expected to become available at the time of recruitment for the study. In the case of rental subsidies, availability meant the availability of a subsidy, not of a unit. Families still needed to find units, sometimes with help from programs.

Availability of transitional housing was constrained by turnover—a new family cannot move in until the previous family has vacated—which is not inherently aligned with the seasonal surges that have been documented for families in shelter (Colburn, 2017). Many times in the enrollment process, transitional housing programs had families but no openings, and at other times, the programs had open units but no families to screen for them. Further, given the constraints of project-based programs, at times the physical size of the available units did not align with the size of the eligible families. For rental subsidies, family size may have entered into the process later if large families had trouble finding units that would accommodate them.

As policymakers urge central intake of families and individuals into homeless services systems and require communities to establish coordinated assessment systems, localities must consider the impact of limited availability on their triage processes. For instance, if a community has an assessment process that dictates referral to a specific kind of program, it must have a plan B—either another referral option or a way to create flexible capacity within the system. Providing transitional housing in units rented on the open market could alleviate some of the problems related to the project-based nature of transitional housing but would not address the overall resource constraint.

The study transpired during the great recession, when economic factors rather than psychosocial factors probably played an especially large role in shelter entry. Given this likely bias, it is of special concern that only 71.1 percent of families passed initial screening criteria for transitional housing and that programs found additional families ineligible after the study made priority offers. Transitional housing, which is the most service-intensive intervention, would be expected to accommodate families with both economic and psychosocial needs, but that was not the case. It is understandable that particular programs select families that they think will benefit most from those programs, but policymakers should examine the extent to which service-intensive programs are excluding families with the greatest needs for services.

The finding that less than three-fifths of families given priority offers enrolled in either transitional housing or short-term rental subsidy programs should also give pause to policymakers. The Family Options Study enrolled families who had spent at least a week in emergency shelters, who presumably had few options, yet many of them nevertheless found the offered programs unattractive.

Across shelters, and the transitional housing and long-term subsidy interventions, programs frequently required the separation of families who wanted to stay together—most often because men were not permitted in congregate programs, but also because some family members were excluded from housing based on their criminal backgrounds. It is reasonable that congregate facilities like shelters and project-based transitional housing programs exclude men if they cannot guarantee some level of privacy to families. In the qualitative interviews, we found families who feared for the safety of their daughters in programs that admitted men and older boys, as well as families who lamented forced separation. Again, a mismatch exists between the service system and family needs if families must choose between housing and family integrity. Scattered-site transitional housing models may have some advantage over project-based transitional housing models here. Policy-makers should consider whether rules excluding some family members from long-term subsidy programs do more harm than good.

Conclusion

This study documents that the homeless service system, despite being augmented for this study, does not fully accommodate the characteristics and desires of the families it is designed to serve. Without that augmentation by long-term rental subsidies and ARRA funds, it would fall further short of meeting family needs. HUD is gradually switching resources from transitional housing to short-term rental subsidies and urging communities to reduce entry criteria for programs under the rubric of housing first. This study suggests that such moves will reduce but not eliminate the mismatch between family characteristics and homeless service programs.

The fact that the vast majority of families passed screening for long-term subsidy programs, found units, and leased up at higher rates than in other studies of poor families has additional implications for policy. Families’ lease-up success and their clear preference for the HCV program, along with evidence from the Family Options Study of the salutary effects of subsidies across domains of housing, family preservation, self-sufficiency, and adult and child well-being suggests that long-term subsidies are an important tool to end family homelessness (Gubits et al., 2016, 2015). The Housing Opportunity Through Modernization Act of 20168 gives PHAs greater flexibility to use project-based vouchers to house families experiencing homelessness, particularly in areas where tenant-based vouchers may be more difficult to use (Fischer, 2016). Other provisions streamlining unit inspection rules may also help vulnerable families who are offered vouchers to move into housing more quickly. Expansion of the subsidy program and deeper targeting of subsidies (to needier families) might avert or put an early end to much family homelessness (Early, 2004).

​

Exhibit 1

Percent of Families Who Failed To Satisfy Specified Eligibility Criteria for at Least One Program, by Intervention Type

Question TypeIntervention Type (%)
Transitional Housing (n = 1,700)Short-Term Subsidies (n = 2,067)Long-Term Subsidies (n = 1,745)
Unit size26.6NA1.1
Minimum income or employment20.97.4NA
Family composition18.1NANA
Disability requirement5.9NANA
Geographic location2.83.41.7
Required sobriety or treatment2.60.5NA
Credit history2.50.31.8
Able to pay some rent1.80.0NA
Education or work experience1.60.7NA
Lack of criminal conviction1.40.32.5
Mandatory participation in services or activities0.5NANA
Health screening0.20.0NA
Citizenship or documentation of status0.20.30.3
Domestic violence0.1NANA
Maximum income0.0NANA
Lack of housing authority arrears0.00.00.4
Recent drop in incomeNA0.5NA
Housing stability historyNA0.8NA
Percent lost at least one program50.913.37.0
Percent lost intervention28.99.24.1

NA = no program of this type asked question.

Notes: N = 2,490 families screened. Questions were asked only for programs with openings. Numbers differ from those in Gubits et al. (2013) because they reflect all families who were screened, not only families eligible for the study, with the exception of Honolulu, where information on reasons for exclusion from the long-term subsidy condition is not available.

Exhibit 2

Percentage of Families Who Survived Two Stages of Program Enrollment, by Intervention Type

Transitional HousingShort-Term SubsidyLong-Term Subsidy
A. Number screened for available slot1,7002,0671,745
B. Percent passed initial screening71.190.895.9
C. Number given priority offer368569599
D. Percent passed any additional screening, accepted offer, and moved in within 9 months45.756.176.5
E. Final percent who moved in (B * D)32.551.073.4

Notes: Rows A and B include all 2,490 families initially screened for the study. Information on eligibility screening for Honolulu long-term subsidy providers is not available. Row C reflects 2,282 families enrolled in the study at baseline, and Row D shows the percentage that moved in within 9 months, based on the program use and living situation data.

Acknowledgments

This study was supported by U.S. Department of Housing and Urban Development contract C-CHI-00943, Task Orders T-0001 and T-0003, to Abt Associates and the National Institute of Child and Human Development 5R01HD066082 to Vanderbilt University.

Footnotes

1“Homeless Emergency Assistance and Rapid Transition to Housing: Continuum of Care Program: Interim Final Rule,” 24 CFR Part 578. Federal Register 77 (147) July 31, 2012. https://www.hudexchange.info/resources/documents/CoCProgramInterimRule.pdf.

2“Homeless Emergency Assistance and Rapid Transition to Housing: Continuum of Care Program: Interim Final Rule,” 24 CFR Part 578. Federal Register 77 (147) July 31, 2012. https://www.hudexchange.info/resources/documents/CoCProgramInterimRule.pdf.

3Pub. L. 111–5. 123 Stat. 115. February 17, 2009.

4Participating sites were Alameda County, California; Atlanta, Georgia; Baltimore, Maryland; Boston, Massachusetts; Denver, Colorado; Honolulu, Hawaii; Kansas City, Missouri; Louisville, Kentucky; Minneapolis-Hennepin County, Minnesota; Phoenix-Maricopa County, Arizona; Salt Lake City, Utah; and the combined site of Bridgeport and New Haven, Connecticut.

5This number includes 25 families later deemed not to have met the entry criterion of having a child age 15 or under and so were excluded from followup.

6Note that for this article, we include all 2,282 families eligible for followup, whereas Solari and Khadduri (2017) examine enrollment for only those families reinterviewed at 20 months.

7The cited figure from Solari and Khadduri (2017) includes only families given priority offers of HCVs who were reinterviewed at 20 months and look at lease ups during the full 20-month period. Analyses for this article include all families given priority offers of long-term rental subsidies and examine leaseups during the first 9 months.

8Pub. L. 114–201, § 1 et seq., 130 Stat. 782.

Contributor Information

Marybeth Shinn, Vanderbilt University.

Scott R. Brown, Vanderbilt University.

Brooke E. Spellman, Abt Associates.

Michelle Wood, Abt Associates.

Daniel Gubits, Abt Associates.

Jill Khadduri, Abt Associates.

References

  • Bacon Laura. Godinez v. Sullivan-Lackey: Creating a Meaningful Choice for Housing Choice Voucher Holders. DePaul Law Review. 2005;55(4):1273–1308. [Google Scholar]
  • Basolo Victoria, Nguyen Mai Thi. Does Mobility Matter? The Neighborhood Conditions of Housing Voucher Holders by Race and Ethnicity. Housing Policy Debate. 2005;16(3–4):297–324. doi: 10.1080/10511482.2005.9521546. [CrossRef] [Google Scholar]
  • Colburn Gregg. Seasonal Variation in Family Homeless Shelter Usage. Housing Policy Debate. 2017;27(1):80–97. doi: 10.1080/10511482.2016.1158200. [CrossRef] [Google Scholar]
  • Cragg Michael, O’Flaherty Brendan. Do Homeless Shelter Conditions Determine Shelter Population? The Case of Dinkins Deluge. Journal of Urban Economics. 1999;46(3):377–414. [Google Scholar]
  • Culhane Dennis P, Metraux Stephen, Park Jung Min, Schretzman Maryann, Valente Jesse. Testing a Typology of Family Homelessness Based on Patterns of Public Shelter Utilization in Four U.S. Jurisdictions: Implications for Policy and Program Planning. Housing Policy Debate. 2007;18(1):1–28. [Google Scholar]
  • Cunningham Mary, Burt Martha R, Scott Molly, Buron Larry, Klerman Jacob, Fiore Nicole, Stillman Lindsey. Prevention Programs Funded by the Homelessness Prevention and Rapid ReHousing Program. Washington, DC: U.S. Department of Housing and Urban Development; 2015. https://www.huduser.gov/portal/publications/affhsg/HPRP-homeless-prevention-report.html. [Google Scholar]
  • Daniel Tamica H. Bringing Real Choice to the Housing Choice Voucher Program: Addressing Voucher Discrimination Under the Federal Fair Housing Act. The Georgetown Law Journal. 2010;98(3):769–794. [Google Scholar]
  • Early Dirk W. The Determinants of Homelessness and the Targeting of Housing Assistance. Journal of Urban Economics. 2004;55(1):195–214. doi: 10.1016/j.jue.2003.09.005. [CrossRef] [Google Scholar]
  • Edin Kathryn, DeLuca Stefanie, Owens Ann. Constrained Compliance: Solving the Puzzle of MTO’s Lease-Up Rates and Why Mobility Matters. Cityscape. 2012;14(2):181–194. [Google Scholar]
  • Finkel Meryl, Buron Larry. Report prepared for the US Department of Housing and Urban Development. Cambridge, MA: Abt Associates; 2001. Study on Section 8 Vouchers Success Rates: Volume 1 Quantitative Study of Success Rates in Metropolitan Areas. [Google Scholar]
  • Fischer Will. Housing Bill Unanimously Passed by House Would Build on Effectiveness of Rental Assistance. 2016 http://www.cbpp.org/research/housing/bipartisan-housing-bill-would-build-on-effectiveness-of-rental-assistance.
  • Fisher Benjamin W, Mayberry Lindsay S, Shinn Marybeth, Khadduri Jill. Leaving Homelessness Behind: Housing Decisions Among Families Exiting Shelter. Housing Policy Debate. 2014;24(2):364–386. doi: 10.1080/10511482.2013.852603. [PMC free article] [PubMed] [CrossRef] [Google Scholar]
  • Gubits Daniel, Khadduri Jill, Turnham Jennifer. Working paper. Cambridge, MA: Joint Center for Housing Studies of Harvard University; 2009. Housing Patterns of Low Income Families With Children: Further Analysis of Data From the Study of the Effects of Housing Vouchers on Welfare Families. [Google Scholar]
  • Gubits Daniel, Shinn Marybeth, Bell Stephen, Wood Michelle, Dastrup Samuel, Solari Claudia D, Brown Scott, Brown Steven, Dunton Lauren, Lin Winston, McInnis Debi, Rodriguez Jason, Savidge Galen, Spellman Brooke E. Family Options Study: Short-Term Impacts of Housing and Services Interventions for Homeless Families. Washington, DC: U.S. Department of Housing and Urban Development; 2015. http://www.huduser.org/portal//portal/sites/default/files/pdf/FamilyOptionsStudy_final.pdf. [Google Scholar]
  • Gubits Daniel, Shinn Marybeth, Wood Michelle, Bell Stephen, Dastrup Samuel, Solari Claudia D, Brown Scott R, McInnis Debi, McCall Tom, Kattel Utsav. Family Options Study: 3-Year Impacts of Housing and Services Interventions for Homeless Families. Washington, DC: U.S. Department Of Housing and Urban Development; 2016. https://www.huduser.gov/portal/sites/default/files/pdf/Family-Options-Study-Full-Report.pdf. [Google Scholar]
  • Gubits Daniel, Spellman Brooke, Dunton Lauren, Brown Scott, Wood Michelle. Interim Report: Family Options Study. Washington, DC: U.S. Department of Housing and Urban Development; 2013. https://www.huduser.gov/publications/pdf/HUD_503_Family_Options_Study_Interim_Report_v2.pdf. [Google Scholar]
  • Henry Meghan, Watt Rian, Rosenthal Lily, Shivji Azim. The 2016 Annual Homeless Assessment Report (AHAR) to Congress. Part 1: Point-in-Time Estimates of Homelessness. Washington, DC: U.S. Department of Housing and Urban Development, Office of Community Planning and Development; 2016. [Google Scholar]
  • Jacob Brian A, Ludwig Jens. The Effects of Housing Assistance on Labor Supply: Evidence From a Voucher Lottery. American Economic Review. 2012;102(1):272–304. [Google Scholar]
  • Joint Center for Housing Studies of Harvard University (JCHS) State of the Nation’s Housing 2016. Cambridge, MA: Joint Center for Housing Studies of Harvard University; 2016. [Google Scholar]
  • Kennedy Stephen, Finkel Meryl. Report prepared for the US Department of Housing and Urban Development. Cambridge, MA: Abt Associates; 1994. Section 8 Rental Voucher and Rental Certificate Utilization Study: Final Report. [Google Scholar]
  • Kerber Kate J, de Graft-Johnson Joseph E, Bhutta Zulfiqar A, Okong Pius, Starrs Ann, Lawn Joy E. Continuum of Care for Maternal, Newborn, and Child Health: From Slogan to Service Delivery. The Lancet. 2007;370(9595):1358–1369. [PubMed] [Google Scholar]
  • Lefkovitz Paul M. The Continuum of Care in a General Hospital Setting. General Hospital Psychiatry. 1995;17(4):260–267. [PubMed] [Google Scholar]
  • McBryde-Foster Merry, Allen Toni. The Continuum of Care: A Concept Development Study. Journal of Advanced Nursing. 2005;50(6):624–632. [PubMed] [Google Scholar]
  • Mills Gregory, Gubits Daniel, Orr Larry, Long David, Feins Judie, Kaul Bulbul, Wood Michelle. Report prepared for the US Department of Housing and Urban Development. Cambridge, MA: Abt Associates; 2006. Effects of Housing Vouchers on Welfare Families: Final Report. [Google Scholar]
  • Sard Barbara. Housing Vouchers Should Be a Major Component of Future Housing Policy for the Lowest Income Families. Cityscape. 2001;5(2):89–110. [Google Scholar]
  • Shinn Marybeth, Gibbons-Benton Jessica, Brown Scott R. Poverty, Homelessness, and Family Break-Up. Child Welfare. 2015;94(1):105–122. [PMC free article] [PubMed] [Google Scholar]
  • Shinn Marybeth, Weitzman Beth, Stojanovic Daniela, Knickman James, Jimenez Lucila, Duchon Lisa, James Susan, Krantz David. Predictors of Homelessness Among Families in New York City: From Shelter Request to Housing Stability. American Journal of Public Health. 1998;88(11):1651–1657. [PMC free article] [PubMed] [Google Scholar]
  • Solari Claudia D, Khadduri Jill. Family Options Study: How Homeless Families Use Housing Choice Vouchers. Cityscape. 2017;19(3):387–412. [Google Scholar]
  • Solari Claudia D, Morris Sean, Shivji Azim, deSouza Tanya. The 2015 Annual Homeless Assessment Report (AHAR) to Congress. Part 2: Estimates of Homelessness in the United States. Washington, DC: U.S. Department of Housing and Urban Development, Office of Community Planning and Development; 2016. [Google Scholar]
  • Sterken Krista. A Different Type of Housing Crisis: Allocating Costs Fairly and Encouraging Landlord Participation in Section 8. Columbia Journal of Law and Social Problems. 2009;43(2):214–243. [Google Scholar]
  • Wong Yin-Ling Irene, Culhane Dennis P, Kuhn Randal. Predictors of Exit and Reentry Among Family Shelter Users in New York City. Social Service Review. 1997;71(3):441–462. [Google Scholar]
  • Wood Michelle, Turnham Jennifer, Mills Gregory. Housing Affordability and Family Well-Being: Results From the Housing Voucher Evaluation. Housing Policy Debate. 2008;19(8):367–412. [Google Scholar]
Feedback