Lihtc and Homelessness: How Low-Income Housing Tax-Credit Allocation in the 2010s Suggests a Need to Streamline Regional Funding

Is the Low-Income Housing Tax Credit (LIHTC), a federal housing production program, being used today to address homelessness? A key previous study (Jackson & Kawano, 2015) analyzed 1990-2000 decennial census data to indicate that the siting of LIHTC projects reduces homelessness at the county level over a 10-year period. However, little research has used the modern HUD-mandated Point-In-Time (PIT) homelessness count data at the Continuum of Care (CoC; representing distinct geographic areas of homeless services funding) level to explore this relationship in a modern decade. As all state tax credit-awarding agencies now offer incentives to use LIHTC for supportive housing, these incentives suggest a broader adoption of LIHTC by nonprofit developers to address homelessness. This paper examines the relationship between CoC-level PIT homelessness counts and LIHTC units awarded to those CoCs from 2010-2020.

The current study utilizes panel data on 392 CoCs in 2010-2020, linking homelessness data from the HUDuser PIT count database with LIHTC allocation data (crosswalk from Byrne et al., 2021). This paper reverses previous intuition about the direction of the relationship between LIHTC affordable housing and homelessness, seeking to understand the degree to which homelessness is driving where LIHTC units are sited. Using OLS linear regression, I assess the association between homelessness (measured as CoC-level PIT count) and LIHTC unit allocation in matched CoCs. Anticipating lag in the responsiveness of LIHTC unit allocation, I test associations of homelessness counts with LIHTC unit counts 1-5 years later.

Preliminary results indicate a statistically significant correlation (p<.01) between homelessness counts and LIHTC units at the CoC level. Exploring time windows of responsiveness, the highest beta estimate significant relationship is between homelessness count and 1-year-lagged LIHTC units allocated, compared to 3-year and 5-year-lagged unit counts. Adjusting for LIHTC units allocated the previous year, the relationship between homelessness count and 1-year-lagged LIHTC unit count remains statistically significant. Concretely, for every 20 people experiencing homelessness in each CoC, we would expect an additional unit of LIHTC housing to be allocated the following year (p < .01). Additional results will follow incorporating robust covariates from the Census American Community Survey to control for other factors influencing LIHTC allocation.

Building permanent supportive housing with the low-income housing tax credit is an expensive and slow-moving process involving subsidies from multiple levels of government and extensive community engagement. The LIHTC program was designed to allow the market to produce affordable housing and not to directly address homelessness. At the same time, this data suggests that it is increasingly allocated to build housing for people experiencing homelessness today. If LIHTC is becoming a de facto program to address homelessness, HUD can take steps to integrate its state and regional programming. As it stands, HUD funds states to allocate LIHTC units based on total state population, while funding Continuums of Care to allocate homelessness spending based on regional homelessness counts. Aligning these funding pathways at the regional level would speed the production and lower costs associated with funding permanent supportive housing with the Low-Income Housing Tax Credit.