Association for Public Policy Analysis & Management

This paper asks if the number of Continuums of Care (CoCs) – the U.S. Department of Housing and Urban Development’s (HUD) homeless service coordination jurisdictions – in a region affects outputs and outcomes related to housing services and homelessness. To do so, it creates the theoretical setting and conducts empirical tests for geographic competition and collaboration between CoCs. Beginning in 1994, HUD required communities to come together to create CoCs to coordinate homeless services and annually apply for federal funding together.

There are currently about 380 CoCs with various sizes and formats. This large variation in the shape and size of CoCs creates a unique opportunity to study the effects of federal funding for an area and how individual service providers compete and collaborate with each other, as well as CoCs competing and collaborating with each other. Past studies and HUD have seen CoCs as independent, although this ignores the fact that CoCs interact with each other and the actions of one likely affect the situation of another. Being treated independently for annual funding may however cause CoCs to focus on quick solutions within their boundaries, perhaps leading to unintended consequences for both their own and neighbors’ outcomes.

After creating a theory on how outcomes in one CoC depend on actions of others, it uses CoC level data from 2007-2016 to empirically test the effects. First, it uses a pooled cross-sectional model of the fifty largest metropolitan statistical areas’ (MSA) CoCs, aggregating CoC data to the MSA level where the main explanatory variable is the number of CoCs in the MSA. If having more CoCs improves outcomes, it may suggest competitive benefits. Alternatively, if having fewer CoCs improves outcomes, it may suggest regionalism may be appropriate for homeless services. Second, to achieve some causal evidence that the number of CoCs in a geographic area changes outcomes, it creates a synthetic control difference-in-difference, using the case of Connecticut. In 2011, Connecticut merged its twelve CoCs into only two, creating an opportunity to test the effects of consolidation on the performance of CoCs. This paper therefore provides both descriptive and causal evidence toward the relationship between the number of CoCs in a geographic area and outcomes. It contributes to the knowledge of CoCs and homelessness, as well as regionalism and inter-governmental partnerships. Additionally, results can help assist CoCs and HUD when considering consolidation of CoCs.