Association for Public Policy Analysis & Management

As of January 2016, 22% of the total homeless population was experiencing what is known as “chronic” homelessness, meaning they were homeless for an extended time period, often due to complex health and behavioral disorder issues (HUD 2016). Ending chronic homelessness has been a major policy priority for both ethical and financial reasons. People who are chronically homeless suffer from untreated mental illness, substance abuse disorders, and a variety of medical conditions; their mortality rates are 3 or more times higher than the general population (Hwang, 2000). At the same time, a chronically homeless person will consume an estimated $30,000-50,000 in public expenses per year due to their heavy involvement in public health care, shelters, and criminal justice systems (U.S. Interagency Council on Homelessness 2015)

Chronic homelessness is what is known as a “wicked problem” and a poster child for the collaboration imperative (Rittel & Webber 1973; Kettl 2006). Many attempts have been made at reducing chronic homelessness but the fragmentation of the social service system has led to ineffectiveness (Burt 2009). Collaboration between government entities and nonprofit providers, and across fragmented policy areas, is thought to help manage complexity, break down jurisdictional boundaries, and reinforce mutual accountability. To meet these goals HUD mandates a collaborative governance process in every US region: the Continuum of Care (CoC) process.

RESEARCH QUESTIONS: 1) To what degree are particular structural arrangements in collaborative governance networks associated with greater effectiveness? We address this by looking to see if participant-led networks are associated with a greater reduction in chronic homelessness over time than networks that are government-led or professionally managed. 2) What pathways contribute to a relationship between structure and effectiveness? In this case, we evaluate both engagement factors and financial factors as potential mediators between structure and effectiveness.

DATA: We leverage publicly available HUD administrative data and a national survey of CoCs (N=312, 75% response rate) carried out by the authors. To explore both direct and indirect effects of governance structure, we conducted path analysis using structural equation modeling.

FINDINGS: Controlling for a variety of regional and CoC-network attributes, CoC structure impacted the chronically homeless population trend between 2009 and 2014, but this effect is fully mediated by greater provider influence. Having a participant-led structure was indirectly associated with 7.3% decrease in the chronically homelessness, with a direct effect of increased provider influence over the CoC. On the contrary, government-led CoCs indirectly contributed to 5.8% increase in chronic homelessness because providers in those networks had less influence in regional service planning. In other words, reductions in chronic homelessness are associated most strongly with CoCs that meaningfully engage providers; that is more likely to be true in those CoCs with a provider-led structure.

CONCLUSIONS: Collaborative governance only meets its goals at solving wicked problems like chronic homelessness when it is truly collaborative and providers are given meaningful ways to engage with and influence the process. “Collaboration” can easily become synonymous with “another meeting;” to avoid this, structures should be chosen that maximize collaborative engagement.