Association for Public Policy Analysis & Management

The U.S. Department of Housing and Urban Development (HUD) in making its funding decisions for homeless services does not properly account for how one homeless service district may affect others. Since 1994, HUD requires communities to coordinate homeless services and annually apply for federal funding through Continuums of Care (CoCs). However, while CoCs coordinate services within their boundaries, HUD’s treating CoCs’ outcomes as within their own control ignores the possibility of cross-CoC externalities, likely leading to inefficient allocation of funds. Additionally, by its encouraging the merging of CoCs, HUD implies fewer CoCs improves outcomes, although HUD and scholars have done little analysis measuring its effect.

Given the policy relevance for homeless services, I thus ask how the number of CoCs in a geographic region affects homelessness in that region, applying theory on regionalism and inter-jurisdictional collaboration. I further ask if merging leads to a decrease in homelessness and increase in operational outcomes. I first create a theoretical model of how the number of CoCs in a region affects externality and coordination costs, using literature and theories from public economics, public management, and urban studies.

I then use panel data from 2007-2018 to conduct a dynamic event study at the CoC level. I find contrary to intent, merging CoCs did not significantly decrease homelessness relative to CoCs that did not merge. Additionally, chronic homelessness, a severe form of homelessness, increased by an average of 43%. Merging also led to a short-term decrease in permanent supportive housing offered, although it increased participation in Homeless Management Information Systems, a sign of increased coordination among service providers. Lastly, I rule out the increase in chronic homelessness and lack of decrease in total homelessness coming from neighboring CoCs or CoCs in the rest of the state. Analysis on nearby CoCs adds to our understanding of the mechanism on inter-jurisdictional effects on homelessness. I therefore provide causal evidence that fewer CoCs may not lead to improved outcomes as HUD suggests.

My findings not only contribute to scholars’ understanding of the economics of homelessness and CoCs but also inter-jurisdictional effects in public service provision. As federal policy related to homeless services has been understudied, I provide causal evidence toward a specific policy of how the number of CoCs in an area affects homelessness. My findings also suggest a regionalist approach to homeless services may lead to lower aggregate performance.