Abstract: Predicting Housing Placement and Time to Housing: Triage, Prioritization, and Race and Gender Effects (Society for Social Work and Research 26th Annual Conference - Social Work Science for Racial, Social, and Political Justice)

435P Predicting Housing Placement and Time to Housing: Triage, Prioritization, and Race and Gender Effects

Saturday, January 15, 2022
Marquis BR Salon 6, ML 2 (Marriott Marquis Washington, DC)
* noted as presenting author
Courtney Cronley, PhD, Associate Professor, The University of Tennesee, Knoxville, Knoxville, TN
David Patterson, Ph.D., Professor and Director, University of Tennessee, Knoxville, Knoxville, TN
Amanda Fackler, MSW, LCSW, Graduate Research Assistant, University of Tennessee, Knoxville, TN
R. Chris Smith, MSSW, LCSW, Knox-HMIS, University of Tennessee
Background and Purpose: The federally-mandated Coordinated Entry system leads many Continua of Care (CoC) to rely on vulnerability assessments to triage and allocate scarce housing, particularly as homelessness continues to increase (Benavides & Nukpezah, 2020). The current study builds on prior investigations of reliability (Brown et al., 2018) and bias (Cronley, 2020) in triage assessments, and extends that work to test how vulnerability and housing prioritization scores predict housing placement and time to housing, in conjunction with intersectional race and gender effects.

Methods: The study used HMIS data (n=1,238) from a southeastern CoC, and reflects national demographics (Henry et al., 2020) in terms of race (30% Black) and mean age (43, SD=13), but with proportionately more females (46%). Vulnerability was assessed using the VI-SPDAT (Community Solutions & OrgCode, 2015), and prioritization scores were based on locally-derived guidelines. Bivariate tests examined mean differences in VI-SPDAT and prioritization between those housed and not housed, and odds of being housed among Black Females, Black Males, White Females, and White Males. Using binomial logistic regression, being housed was regressed on the VI-SPDAT and prioritization, dummy coded variables for RaceXGender and household type (family, unaccompanied adult, or unaccompanied youth), age, chronic homelessness, domestic violence, veteran, and disability. Cox regression was used to test time to housing, among those housed.

Results: Being housed was associated with higher vulnerability and lower prioritization, as was Black Female, Black Male, and in a family. Black Females scored significantly lower in vulnerability and housing prioritization compared to White Females. Families scored as significantly less vulnerable and lower in priority compared to unaccompanied adults and youth. Black Females had higher odds of being in families. In the multivariate model (2(df=12)=147.743, p<.001), housing odds dropped 26% for each 1-point decrease in vulnerability. Black Females’ housing odds were 63% higher, and Black Males’ odds were 51% higher, compared to White Males. Housing odds were 62% lower for youth, and 60% lower for adults, compared to families. Veterans’ odds were 270% higher. Prioritization and other controls were non-significant. In the Cox regression (2(df=12)=144.156, p<.001), lower prioritization, families compared to youth, being a veteran, and fleeing domestic violence significantly predicted faster time to housing.

Conclusions and Implications: Results support evidence of intersectional bias in triage assessments, with Black Females scoring as less vulnerable and lower in priority. Homelessness among Black Females may be more strongly related to structural factors, such as employment and education discrimination, racism, and over-incarceration of Black Men, much of which is not captured in vulnerability assessments.

However, the current study finds complicated relationships among race, gender, housing placement, time to housing, and triage decisions. Those assessed as less vulnerable, Black Females and Males, are being housed at higher odds. Furthermore, lower priority predicts faster time to housing. Findings may be explained partially by insufficient supply of dedicated permanent supportive housing beds, and overrepresentation of Black Females in families. Future research may help to explain how local affordable housing stock influences triage decisions.