Assessing the Effectiveness of the U.S. Youth Homelessness Demonstration Program

Background and Study Aim: Youth homelessness remains a critical issue in the U.S., with an estimated 4.2 million youth experiencing homelessness in 2023 – a 29% increase in the point-in-time (PIT) count from the previous year. Research links youth homelessness to adverse outcomes, including mental health issues, substance use, and school underperformance. Despite attention from the research and practice community, the Youth Homelessness Demonstration Program (YHDP) is the only national effort solely focused on addressing youth homelessness. From 2016 to 2023, the U.S. Department of Housing and Urban Development (HUD) awarded funding to 124 Continuums of Care (CoCs) to develop community-wide strategies to prevent and end youth homelessness. While YHDP represents a major federal investment, existing evaluations are limited in scope (e.g. focused on a single year of funded CoCs).. The current study addresses this knowledge using a multilevel negative binomial regression to compare youth homeless trends in the 56 CoCs that implemented YHDP between 2016 and 2019 and the 261 CoCs that never received YHDP funding. The study examines whether YHDP implementation resulted in a significant reduction in the annual count of homeless youth within funded CoCs compared to non-funded CoCs. Findings will provide evidence on YHDP’s impact to inform future policy decisions.

Method: Data and Sample: This study constructs a panel dataset using annual PIT count data from 2013 to 2024 on youth aged 24 and under experiencing homelessness across 317 CoCs. The sample includes 56 CoCs that received YHDP grants between 2016 and 2019, and 261 CoCs that did not receive YHDP funding.

Measures: The dependent variable is the annual count of youth experiencing homelessness. The independent variable is a binary indicator of YHDP funding We controlled for calendar year, the CoC type as defined by HUD (largely rural, suburban, major city, or other urban areas), and the type of homelessness count conducted (sheltered-only count vs. sheltered and unsheltered count). Robust standard errors were clustered by CoC to account for within-unit correlation over time.

Analysis: Multilevel negative binomial regression was conducted to examine the effect of YHDP funding on youth homelessness.

Results: YHDP implementation did not have a statistically significant effect on youth homelessness trends in funded CoCs compared to non-funded CoCs. Although overall youth homelessness declined across CoCs from 2013 to 2024, this trend was independent of YHDP implementation. Notably, CoCs that received YHDP funding had consistently higher baseline rates of youth homelessness than those that never received funding.

Conclusion and Implications: These counterintuitive findings raise questions about the effectiveness of YHDP. One possible explanation is that CoCs receiving YHDP funding were already experiencing upward trends in youth homelessness, and the intervention was insufficient to reverse them. Alternatively, YHDP may have improved CoC’s capacity to identify and count homeless youth, potentially offsetting the decrease or even contributing to an increase when compared to the pre-implementation period. Given the large variation in how YHDP was implemented by different CoCs, future research should consider in-depth qualitative studies to explore implementation differences that potentially influenced program outcomes.