Understanding the Emergence and Persistence of Mandated Collaboration: A Policy Feedback Perspective of the United States’s Model to Address Homelessness

Abstract

Collaboration is commonplace in contemporary public administration. In many instances, policy mandates collaboration between previously unconnected organizations for those organizations to obtain essential funding for public services, thus creating new administrative structures grounded in collaboration. There exists substantial research that focuses on the collaborative process and potential outcomes of these structures, yet their emergence and development is less understood. The Housing and Urban Development (HUD) continuum of care (CoC) model is one such collaborative structure that has been the dominant administrative service delivery system used to address homelessness in the United States since the early 1990s. A historical analysis reveals that policy feedback effects help explain the emergence and persistence of the CoC model from before its origin to its eventual codification in the Homeless Emergency Assistance and Rapid Transition to Housing (HEARTH) Act of 2009. A policy feedback perspective of the CoC model demonstrates how the interplay of policy, politics, and administration led to a mandate to collaborate to address a large-scale social problem.

Keywords

mandated collaboration collaborative governance policy feedback theory homelessness networks

Recent national estimates suggest that on any given night, 549,928 people were homeless in the United States (U.S. Department of Housing and Urban Development [HUD], 2016). While this presents as a staggering figure, the rate of homelessness has declined nearly 18% from counts in 2007 (647,258), and the capacity to house homeless and formly homeless people has increased by nearly 42% (HUD, 2016). In addition, impressive is the fact that this encouraging decline endured through the Great Recession of the late 2000s—what the International Monetary Fund has called the deepest global recession since World War II (International Monetary Fund, 2009).

The ability of a society to maintain a positive impact on a challenging social problem such as homelessness is a demanding task, even in times of economic prosperity.¹ To accomplish such a feat requires concentrated efforts from multiple partners across governments, nonprofits, private enterprises, and local community members. While demand from these entities to address homelessness may not come from normative concerns (Cronley, 2010), the financial costs of ignoring homelessness (e.g., increased medical hospitalizations, increased prison and jail stays) demand attention (Garrett, 2012). The U.S. federal government’s response to these concerns and costs of homelessness—as implemented largely by the U.S. Department of HUD—has supported an approach that requires the creation of local community collaboratives to obtain vital federal funding to combat homelessness locally. This approach is called the continuum of care (CoC) model and has been in practice since the early 1990s, but was not codified into legislative law until the passage of the Homeless Emergency Assistance and Rapid Transition to Housing (HEARTH) Act in 2009.

The CoC model is an example of mandated collaboration in action. Mandated collaboration is a concept that is commonly described as “. . . when bureaucratic or hierarchical mechanisms [CoC funding guidelines] are used by a third party [HUD] to bring separate organizations together [local governments and nonprofits] to pursue complex objectives [address local homelessness]” (McNamara, 2016, p. 68; Rodríguez, Langley, Béland, & Denis, 2007). The term “mandated collaboration” may appear contradictory—collaborative endeavors that involve multiple organizations and interests forming an interorganizational network to address a commonly shared problem are generally viewed as voluntary. At times, collaborative endeavors may initiate because of an underlying funding mandate that is tied to collaboration efforts, or collaboration is at least driven in part by the presence of the mandate (Provan & Lemaire, 2012). Yet, the presence of a mandate does not preclude these endeavors from being collaborative—there may still exist a level of reciprocal interdependence between network actors that exceeds levels found in instances of cooperation or coordination (Keast & Mandell, 2014).

There is substantial research on the collaborative process and outcomes of collaborative administrative structures (e.g., Ansell & Gash, 2008; Emerson, Nabatchi, & Balogh, 2012; McGuire, 2006; Provan & Lemaire, 2012), yet their emergence and persistence is less understood. The purpose of this study was to present a historical analysis of the CoC model—the collaborative administrative structure that serves as the federal government’s response to homelessness—to explore the emergence and persistence of mandated collaboration. The historical narrative evidences policy feedback effects (Béland, 2010; Mettler & SoRelle, 2014), which contribute to providing insight on how the model received legitimate codification in the HEARTH Act nearly 15 years after it was first conceptualized. Soss and Moynihan (2014) called for further research that utilizes policy feedback theory to explore the relationships between politics, policy, and administration. Adopting a policy feedback perspective when analyzing the CoC historical narrative provides insight into the development of the collaboratively based federal response to homelessness that has helped mitigate a complex social problem (i.e., homelessness). The current analysis demonstrates that negative policy feedback effects contributed to creating an environment that supported the emergence of the CoC model, and lock-in effects supported the use of the model in administrative implementation until its eventual codification (Pierson, 2000).

The remainder of the article is divided into four sections: The first section is a review of policy feedback theory to serve as a background for the subsequent historical narrative and analysis. The second section is the historical narrative, which includes a brief background of homelessness in the United States, and how the CoC model evolved out of federal legislation and administrative action in the 1980s and 1990s. The third section is the analysis of the historical narrative from a policy feedback perspective. The final section discusses the policy feedback effects of the CoC narrative, posits why the federal government’s response to homelessness may continue to support the CoC model, and discusses the implications of the current study for current and future research on mandated collaboration.

Policy Feedback Theory

Literature on policy feedback theory focuses on understanding the “feedback effects” of public policy. While not a new idea (e.g., Lowi, 1972; Schattschneider, 1935), the concept has gained noticeable attention from policy studies scholars within the political science literature over the last two decades, even earning a spot in the most recent edition of the seminal Theories of the Policy Process (Sabatier & Weible, 2014). The approach has proven useful in discovering policy feedback effects related to many diverse areas, including Social Security (Béland, 2010; Campbell, 2002), the G.I. Bill (Mettler, 2002), welfare reform (Soss & Schram, 2007), criminal justice (Soss, Fording, & Schram, 2011), agriculture (Daugbjerg, 2003), and performance management in public administration (Wichowsky & Moynihan, 2008). For example, Campbell (2002) demonstrated how the dependence on Social Security relates to the increased political participation level of low-income seniors. A similar effect was demonstrated by Mettler (2002) who found that beneficiaries of the education provisions of the G.I. Bill enhanced their levels of civic participation. Daugbjerg (2003) concluded that policy feedback effects are also important for understanding changes in policy.

Béland (2010) suggests that increased attention toward policy feedback effects began with historical institutionalists, who began to focus on how “policy once enacted, restructures subsequent political processes” (Skocpol, 1992, p. 58) and who grounded their work on feedback effects in the policy-learning work of Hugh Heclo (1974). The work of Heclo stressed “the ways in which state bureaucrats and policy experts draw lessons from existing policies to inform the decision-making process” (Béland, 2010, p. 570). Contemporary literature on policy feedback theory has continued to articulate the idea that “new policies create new politics” (Schattschneider, 1935). For example, Mettler and SoRelle (2014) suggest that policy feedback effects can stem from various parts of the policy process (e.g., design, implementation) and have the potential to “. . . shape the attitudes and behavior of political elites and mass publics, as well as to affect the evolution of policymaking institutions and interest groups,” which in turn effect policymaking processes (p. 152).

Current perspectives on policy feedback literature suggest that the approach has several implications and directions. Soss and Moynihan (2014) suggest conceptualizing the policy feedback literature into two streams—one which focuses on policy effects for political elites and mass publics (e.g., Pierson, 1993; Pierson, 2000; Skocpol, 1992), and a second that focuses on using feedback effects for causal (e.g., early policy constrains later policy) and constructivist (e.g., citizenship constructed through interactions with political institutions) explanations (e.g., Schneider & Ingram, 1993). Béland (2010) suggests that early policy feedback literature focused on effects related to state-building, interest groups, and lock-in effects (e.g., path dependence), while recent literature has branched into application of feedback effects to private institutions, linking policy feedback to actual political behavior and exploring symbolic policy effects. In addition, Mettler and SoRelle (2014) note that they have observed four major streams of inquiry in policy feedback literature: How policies shape the meaning of citizenship, affect the form of governance, influence the power of groups, and affect political agendas and policy problem definitions. While there is certainly overlap between these categorizations of policy feedback effects and directions in the policy feedback literature, such a record demonstrates the wide applicability of a policy feedback perspective. However, despite the wealth of insight the perspective may lend, research regarding the implications of policy feedback effects for public administration has largely been ignored by public administration scholars (Soss & Moynihan, 2014). This has occurred in spite of the fact that “. . . administration is a critical moment in the ongoing interplay of policy and politics” (p. 11).

When discussing the impact of policy feedback theory for use in public administration research, Soss and Moynihan (2014) stress the necessity to make two translations and subsequent assumptions that are based on the original policy feedback theory assertion that policies shape politics: Administrative actions may equate to policies (i.e., administration shapes politics), and administrations are inherent with politics (i.e., policies shape administration). It is by adopting these translations that feedback effects may be best discovered and understood in a public administration context (Soss & Moynihan, 2014).

Soss and Moynihan (2014) also suggest that historical analysis is a relevant method to explore the relationship between policy feedbacks and administration. Other scholars have suggested that historical analysis is a method of analysis that is highly applicable to administrative studies (Raadschelders, 2010), yet is frequently overlooked (Spicer, 2004). Historical analysis is a method by which past events are uncovered from historical records to understand macrolevel processes (Marshall & Rossman, 2016; Raadschelders, 2010). The macrolevel process under study here is the emergence and persistence of an instance of mandated collaboration.

The instance of mandated collaboration in the current study is the use of funding guidelines by the U.S. federal government (via HUD) that coerce local governments and nonprofits to collaborate to obtain necessary funding to address homelessness in their locality. The following section details the historical narrative of contemporary homelessness in the United States and how the federal government has responded. This account provides the context for understanding how policy feedback effects assist in accounting for the emergence and persistence of the HUD CoC model.

Homelessness and the Federal Government Response

The historical analysis is divided into two subsections: The first includes a brief background of homelessness in the United States that leads to a description of how the federal government first formally responded to this complex social problem during the 1980s. The second includes a detailed description of the emergence and persistence of the CoC model used by the federal government beginning in the early 1990s to address homelessness.

Homelessness in the United States and the Federal Response

Detailing the origins of homelessness in the United States and the federal government’s entire response to address homelessness through policy is well beyond the scope of this article (for detailed historical accounts, see DePastino, 2003; Hopper, 2003; Kusmer, 2002). However, a brief account is warranted.

Leginski (2007) notes that homelessness has been a part of American history from colonial times to contemporary times, and while the pervasiveness of homelessness has had lulls and waves it has never been completely absent. Not surprisingly, the country’s economic performance has been one of many underlying structural factors that contribute to homelessness (Leginski, 2007). Any presence or absence of governmental intervention and the public’s attitudes toward homeless persons has also shaped the face of homelessness. For example, prior to the 20th century, public-based relief for homelessness did not differ from relief that was directed toward addressing poverty (Leginski, 2007). As federal social welfare programs emerged in response to the Great Depression, so did public attention toward homelessness; however, no legislation was passed that directly addressed homelessness. Increased public focus on the problem of homelessness continued through Lyndon Johnson’s War on Poverty and Great Society initiatives, which included the formation of HUD in 1965 (Cronley, 2010). Although it gained increased attention, homelessness remained a major public issue and was exacerbated by additional social phenomenon, such as the poorly planned deinstitutionalization of the mentally ill population during the Civil Rights Movement of the 1960s (Lamb, 1984).

The 1980s are commonly seen as a period of change regarding the public perception of homelessness and the federal government’s response to the issue (e.g., Burt et al., 2002; Cronley, 2010; Leginski, 2007). The economic recession and federal social program cuts of this time naturally worsened the homelessness situation, and also changed the stereotypical demographic of the homeless population, which showed increased heterogeneity—“. . . fewer and fewer of the homeless resembled skid-row residents and more and more were precariously housed and economically vulnerable women, children, and families and minorities” (Cronley, 2010, p. 323). It became increasingly apparent that Americans from all walks of life were vulnerable to homelessness, and a focused public effort to address homelessness became a prominent political issue. The federal government’s initial response to the problem was modest, and was in the form of short-term emergency assistance (e.g., a one-night stay at a homeless shelter) administered through the Federal Emergency Management Agency (FEMA; Hoch, 2000). In 1984, HUD completed a study on the current status of the emergency shelter system within the United States, and concluded that the current efforts were merely warehousing the homeless population and were doing little to alleviate the actual problem (Hoch, 2000).

In response to the HUD study and increased public pressure, the federal government passed the Stewart B. McKinney Homeless Assistance Act of 1987 (later renamed to the McKinney-Vento Act in 2000), which was the first federal law directly addressing the problem of homelessness (Burt et al., 2002). The act provided funding for a host of homeless assistance programs that were to be administered across various federal departments (e.g., healthcare assistance for the homeless through the Department of Health and Human Services). The basic intention was “. . . to offer a federal incentive to local communities to develop comprehensive homeless assistance systems . . . as being able to offer the array of services and skills necessary to address the needs of all types of homeless people” (p. xi). However, Fuchs and McAllister (1996) note that

. . . rather than a comprehensive policy, the McKinney Act evolved as a set of diverse spending programs, including emergency shelter, transitional and permanent housing, food, primary health care services, mental health service, alcohol and drug treatment, and education and job training. (p. 8)

During the early years of the act, individual homeless service agencies applied directly to the federal governments’ various departments to receive funding monies for the various programs. There was no requirement to indicate that these agencies were involved with any regional planning or service coordination with other homeless service agencies in their area (Burt et al., 2002). While the McKinney Act demonstrated the federal government’s newfound commitment to addressing homelessness, the fragmented approach to administering the programs it created was found to be cumbersome to service providers (Fuchs & McAllister, 1996) and did not fully support the ideal local community “system” that HUD had desired (Burt et al., 2002).

In addition to initiating the process of federal funding for various homeless assistance programs beyond emergency shelters, the McKinney Act established the U.S. Interagency Council on the Homeless² (USICH). The USICH was established as a collaborative independent agency comprised of representative leaders from several diverse federal agencies³ that operated within the executive branch. They were charged with reviewing and coordinating the federal response to homelessness (USICH, 2012). Despite a decision by Congress in 1993 not to reauthorize the USICH (2012), “. . . the Clinton administration decided to continue the work of USICH by establishing it as a working group of the White House Domestic Policy Council” (p. 3). In the same year, President Clinton issued an executive order that mandated the council to develop within 9 months’ time a “Federal Plan to Break the Cycle of Homelessness” that requested of the council recommendations to improve the current disorganized system (USICH, 1994). Unbeknown at the time, the USICH and the plan they devised would be critical to initiating the move toward a new form of organizing that would typify the federal response to homelessness that would be continued for 15 years until it was eventually codified into law.

The CoC Approach to Addressing Homelessness

Burt and colleagues (2002) note that while obtaining reliable data on the homeless population is difficult (and was notably difficult in the 1980s and 1990s due to inferior data systems), the data available suggest that the U.S. homeless population grew from estimated ranges of 250,000-350,000 in 1984 to 500,000-600,000 in 1987, and continued to grow to an estimated 640,000-840,000 in 1996. While McKinney had initiated a much needed federal government response directly focused on addressing homelessness, it had fallen short of being able to efficiently and effectively address the contemporary homeless population. This was attributed to a lack of coordination at the federal government level in terms of the application process for receiving homeless assistance funds from service providers and a lack of comprehensive planning at the local government level in terms of the actual homeless service delivery systems (Burt et al., 2002). A change in the administrative structures that McKinney created was needed.

The largest change was based in part on the USICH’s response to President Clinton’s 1993 executive mandate to create a unified federal plan to address homelessness. As part of the response, the USICH (1994) issued the report titled Priority Home! The Federal Plan to Break the Cycle of Homelessness, within which the council recommended that the administration take a “full-scale attack on homelessness, focusing public and private sector energies to make a real difference” (p. 4). The report itself was a massive collaborative effort that tapped into the knowledge of local government officials, homeless service providers, and homeless individuals to better understand the state of contemporary homelessness in the United States. According to the USICH (1994), a necessary direction in the “attack” would require the federal government adopting a “continuum of care” approach to combatting homelessness. The CoC approach was initially envisioned as making an array of services available to the homeless to meet the diverse needs of this complex population. The four basic components of the CoC array of services were conceptualized as (a) prevention, as supplied through community outreach and assessment; (b) emergency shelter as an alternative to the street; (c) short-term transitional housing accompanied with supportive services (e.g., employment services); and (d) long-term permanent housing to stop the cycle of homelessness (Fuchs & McAllister, 1996). The idea of a CoC service delivery system was aligned with the original intent of the McKinney Act to create “comprehensive homeless assistance systems” but which was never truly achieved (Burt et al., 2002, p. xi).

In addition to the CoC approach, the USICH (1994) indicated in the report that the federal government needed to remove direct contracting from the homeless assistance program application process. Thus, they suggested that numerous individual homeless service agencies throughout the country should no longer be required to apply individually to obtain federal homeless assistance funding. Instead, the USICH observed value in holding local governments responsible for understanding the needs and resources in their communities, thus making them key in coordinating a comprehensive CoC system in their locale and being the driver in the funding application process. The USICH also noted that while the local government was key for coordination, nonprofits entities were key in delivering the actual homeless assistance services.

The federal administration responded to the report by substantially morphing the competitive application and funding process for the housing programs under the McKinney Act in a direction that supported a CoC approach. These changes were enacted without any change being made to the actual underlying statute (Berg, 2013). Beginning in 1996, HUD required applicants to submit community-based applications rather than individual program-based applications to receive funding for McKinney-based homeless assistance programs. HUD also required the applicants to demonstrate the need for homeless assistance in their locale as well as the quality of the CoC they were implementing (Berg, 2013). Applications began to be competitively judged against one another on how well they fit the CoC approach and demonstrated collaborative planning (Burt et al., 2002). These changes required organizations in local communities (e.g., nonprofits, state and local government) to band together as a whole community (i.e., coordinate and collaborate regional planning) to receive McKinney funding that would be used to address, and ultimately end, homelessness in their locale. Burt and colleagues (2002) describe the fundamental reasoning behind shifting to such an approach:

The hope has been that by requiring communities to come together to submit a single comprehensive application, HUD could stimulate them to move toward greater structure and a more strategic vision of their programs and services for homeless people. This increased system structure and rationality was expected to improve services for homeless people and increase the chances that their needs would be met. (p. xii)

In addition to HUD’s substantial change to the funding and application structure, Berg (2013) notes that there were several other policy changes enacted by HUD and Congress (none of which directly altered McKinney) beginning in the mid 1990s that emphasized a CoC approach to addressing homelessness: Communities were empowered by HUD through the allowance of broad discretion to communities regarding how they themselves scored and prioritized funding for the homeless assistance programs that constituted their CoC; Congress enacted a requirement that 30% of the overall competitive McKinney funding was to be used for permanent housing, a standard which HUD imposed upon each CoC applicant; and HUD increased attention toward improving CoC data collection and management by requiring CoCs to conduct homeless person point-in-time counts and to implement CoC-wide homeless management information systems (HMISs).

Adoption of a CoC approach spurred by the Priority Home! report greatly expanded the federal government’s response to homelessness as compared with the prior response during the post McKinney period prior to 1996. The recommendations of the Priority Home! report and the way HUD interpreted and modified the current system resulted in a new system where collaboration in service delivery and planning was key in a competitive funding environment.

The CoC model continued to mature and was implemented across the country but was not made legislative law until 2009 with the passage of the HEARTH Act, an amendment and reauthroization of McKinney. As suggested by Berg (2013), the HEARTH Act “. . . codifies the continuum of care as it had evolved over the previous decade” (p. 319). By 2009, the CoC approach had been in process for 13 years,⁴ and the HEARTH Act did little to change the fundamentals of the approach, given that “HUD’s homelessness programs were widely regarded as well run and accomplishing a good purpose” (p. 319). As noted by Berg, the HEARTH Act did include some elements that affected the CoC approach by requiring CoCs to focus on ending homelessness and not simply managing it: Selection criteria for the CoC grants were more visibly detailed than it had previously been, and included an increased focus on rewarding CoCs that could demonstrate positive outcomes and increased collaboration; incentives to use evidence-based programs as part of the CoC were put into place; funding for the administration of a CoC, which was previously not funded, was made available; additional administrative funding was made available to CoCs who elected to centralize their grant-receiving and disbursing functions across all CoC programs to a single entity; and CoCs that reached a high-performance mark were allowed to use competitive grant funding for a wider population (e.g., at-risk-homeless).

Leginski (2007) notes that the CoC concept has had a flexible nature as it has evolved over the years since its beginning. In its conceptual early stages as part of the Priority Home! report, the USICH referred to the CoC concept as the specific array of services that were required to meet the diverse needs of the contemporary homeless population in the United States (e.g., families, people with addiction and mental health issues). Due to other recommendations that were made in the report (e.g., shifting coordination responsibility to the local level) and the manner in which HUD interpreted and implemented them (e.g., changing the competitive application process), the CoC concept has also come to refer to the infrastructure that makes up the collaborative network of local partners that plan, manage, and deliver the CoC array of services (Leginski 2007). Thus, an understanding of the CoC concept occurs at two levels—one, as a local, comprehensive homeless service network that is intended to work together to meet individualized client needs given the diverse needs of the homeless population, and two, as a collaborative community planning network that is charged with coordinating the service system and “. . . to be a mechanism for generating community-wide discourse and interaction around the problem” (Fuchs & McAllister, 1996, p. 9).

The narrative of the CoC model as described above is rich with evidence of policy feedback effects. These feedback effects contributed to a situation of sequential increasing returns that constricted policy decisions to those that did not veer from the CoC path (Pierson, 2000). The discussion in the following section emphasizes the feedback effects that are noticeable in the narrative, illustrating the “dialogic relationship” between policy and public administration (Soss & Moynihan, 2014, p. 321).

A Policy Feedback Perspective of the Emergence of the CoC Model

Negative policy feedback effects of the McKinney Act of 1987 help to explain, in part, how and when the CoC model emerged (Weaver, 2010). The first U.S. federal government response to homelessness through the passage of McKinney was a product of the rising prevalence of homelessness and the changing face of those who were experiencing homelessness during the 1980s (Burt et al., 2002; Cronley, 2010; Leginski, 2007). This newly framed social problem created a political climate that demanded a response from the federal government. While McKinney was diverse in its approach to addressing homelessness (e.g., establishing funding for food banks, shelters, healthcare for the homeless), it was viewed as a failed attempt to construct a desired collaborative approach to addressing the problem (Burt et al., 2002; Fuchs & McAllister, 1996). The change in the then current approach and emergence of the CoC model was in part a consequence of the McKinney Act and the fragmented administrative structures that supported it.

Two key reports demonstrated the shortcomings of the administrative structures produced by McKinney and set the path for foundational change in the federal government’s response to homelessness: the 1994 USICH Priority Home! report and HUD’s Office of Policy Development and Research 1995 report to Congress on the McKinney programs (U.S. Department of HUD, 1995). The HUD report echoed the claims of the earlier USICH report, such that successfully addressing homelessness requires locally controlled coordination of service delivery between homelessness assistance programs and locally controlled collaboration in homelessness planning. Essential to the production of these reports—and, in turn, the emergence of the CoC model—was the decision of the newly elected President Clinton to continue the work of the USICH (2012) despite its loss of Congressional funding in 1993. Both reports were spearheaded with the support of the then HUD Secretary Henry Cisneros. The recommendations in the reports to shift the direction of McKinney programs may not have had the impact they did without Cisneros at the helm of HUD, and serving as the Chairperson of the USICH. His personal buy-in to addressing homelessness is made clear in a quote of his that tops the Executive Summary of Priority Home! “. . . I have designated addressing homelessness my number one priority” (USICH, 1994, p. 1). Both Clinton and Cisneros were key actors in bringing substantial attention and subsequent administrative action to the homelessness policy subsystem by constructing a dominant coalition where shifting power to localities was a core belief (Jenkins-Smith, Nohrstedt, Weible, & Sabatier, 2014). While opposition existed (e.g., Congress cutting funding for USICH), the lack of an apparent and powerful minority coalition may be in part due to the lack of political interest in constructing an opposing coalition that did not support the new changing face of the homeless population. In addition, the dominant coalition’s support of devolving power from the national to the subnational level of government to improve efficiency and effectiveness of services was in line with the leading trend in devolution governing of the 1990s (Cho & Wright, 2004).

A pivotal recommendation in USICH’s Priority Home! was that the federal government move toward supporting a homeless service system that adopts a CoC model based on a local, interorganizational network approach. As head of HUD, Cisneros assisted with making this recommendation a core belief of the federal approach homelessness policy (Jenkins-Smith et al., 2014). HUD embraced and implemented the recommendation as two conceptualizations—as a homeless service delivery network based on an array of services and as a collaborative community planning network (Leginski, 2007). Although the CoC model had top-down buy-in from the Clinton administration, its acceptance and implementation should not be perceived as only a top-down mandate. The acceptance of the CoC model may also be attributed in part to the collaborative way the recommendation came to fruition through the Priority Home! initiative. In the early stages of the initiative, the USICH (1994) leaders developed an intensive nationwide outreach and consultation plan that resulted in feedback from

. . . more than 14,000 representatives of state and local governments, not-for-profit providers of services and housing, advocates for homeless people, economic and community development leaders, educators and social service professionals, and currently and formerly homeless individuals and families. (p. 11)

Obtaining feedback from these numerous constituents to create “a truly representative policy on homelessness” (USICH, 1994, p. 12) involved holding public forums to discuss the federal government’s response to homelessness; collecting responses from a mailed questionnaire—accompanied with a letter from Cisneros—that was delivered to more than 12,000 organizations and individuals; and interviewing 400 individuals living in homeless shelters and transitional housing programs across 10 cities to ensure that the voice of the homeless was involved in planning.

The substantial amount of bottom-up feedback that went into the report may be seen in part as bottom-up pressure for HUD to abandon the current ineffective administrative structure developed by McKinney and adopt a CoC model. For example, USICH (1994) noted in the report that the largest number of recommendations that came from their widespread survey of “people who understand homelessness the best” were related to simplifying the homeless assistance programs (e.g., consolidate the funding and application process) and to improve local coordination (pp. 59-60). USICH also noted in the report that “. . . providers have also voiced serious concern about the fragmented nature of the McKinney assistance programs” (p. 47). This combination of support from federal-level decision makers and local, on-the-ground homelessness professionals demonstrates the impact of the negative policy feedback effects of the McKinney Act that prompted a policy shift—the emergence of the CoC model and the collaborative interorganizational networks the model embodies.

The decision made by HUD in 1996 to alter their McKinney administrative structure to one that fully supported the CoC model (i.e., requiring communities to apply for McKinney funds through a single CoC application and by rewarding localities for demonstrating collaboration) illustrates a lock-in effect—the federal government’s response to homelessness is one that supports mandated collaboration (Béland, 2010). This may be attributed in large part to the nature of the decision itself, such that it created hundreds of new local administrative structures (i.e., local CoCs). These structures enjoyed substantial autonomy in organizing and planning to meet local demands; however, the federal government required demonstrating local collaboration to obtain McKinney funding (Burt et al., 2002).

The state-building power of the decision to support creation of numerous CoCs across the nation—currently 400 according to the most recent count (HUD Exchange, 2017)—created substantial buy-in for the CoC model from local constituents. This state-building feedback effect—described by Skocpol (1992, p. 58) as how “. . . policies transform or expand the capacities of the state”—assists in explaining the persistence of the CoC model. This effect is evidenced by Fuchs and McAllister (1996) through an evaluation they conducted of the CoC model in its early stages of implementation. As part of the evaluation they conducted “. . . case studies of nine representative sites [CoCs] across the country” (p. 42) to understand how CoCs were being implemented and how the CoC model was being perceived by constituents. A primary finding across all cases was that the CoC model significantly changed the planning process that communities undertook to address homelessness. It involved bringing together a larger base of stakeholders than had previously been involved to engage in a collaborative process. As Fuch and McAllister state, “The concept of ‘community participation’ has expanded, bringing together a broad-based group of public and private stakeholders” (p. 42), which also included a chance for smaller neighborhood-based organizations to become involved and obtain funding support; thus increasing the buy-in base for the CoC model. Fuchs and McAllister note several examples of this local buy-in, including that of a service provider who was part of Detroit’s CoC who acknowledged that “being part of the collaboration [CoC] is like becoming part of a family” (p. 43). An additional example includes how the Las Vegas CoC experienced a substantial increase in collaborative participation—prior to 1993, only two agencies provided homeless assistance, while after CoC implementation the county and more than 50 nonprofits have come together to form a comprehensive response to homelessness (Fuchs & McAllister, 1996).

Fuchs and McAllister (1996) also noted that the incentivized and competitive application and community planning process that focused on rewarding collaboration helped break down “turf” barriers. For example, they noted that several large nonprofit service providers that were under the jurisdiction of the Houston CoC had decided to submit applications by themselves; however after receiving no grant funds from HUD for their applications, they recognized the benefits of collaboration and joined their local CoC. In addition, a Denver CoC member indicated that the level of collaboration pre- and post-CoC implementation was like “night and day” (p. 46). Another example is the case of the New York CoC: they indicated that bringing together a large number of stakeholders was difficult at first, but with the leadership of the city’s Department of Homeless Services the CoC became an institutionalized planning process (Fuchs & McAllister, 1996). It is clear that the buy-in from these newly formed collaborative governance networks (i.e., CoCs) was solid. As further evidence, Fuchs and McAllister (1996) state that the “recommendations for a continued emphasis on collaborative planning, and on comprehensive program efforts, echoed across sites,” including that the CoC model appealed to local stakeholders because of the “. . . flexibility it offers local communities to prioritize their own needs and to develop locally-driven solutions” (p. 49). The manner in which HUD maintained respect for local autonomy and authority contributed to creating an environment that was in strong support of the CoC model.

The large-scale evaluations of the CoC model conducted by Fuchs and McAllister (1996) and Burt and colleagues (2002) both serve as evidence that the policy feedback effects from the late 1980s and early 1990s were strong enough to support the CoC model until its eventual codification in the 2009 HEARTH Act. Fuchs and McAllister, who explored how initial CoC implementation was progressing, presented a clear and simply stated conclusion to their evaluation: “Overall, the study suggests that the Continuum of Care policies implemented by HUD have had a positive impact on communities across the nation” (p. 7). The findings of Fuchs and McAllister are echoed by Burt and colleagues, who studied CoCs further along in the implementation process. They deliberatively studied high-performing CoCs—“22 of the 25 [CoCs] were in the top 10 percent of FY 2000 applicants” (Burt et al., 2002, p. x)—to answer questions regarding how these CoCs were organized and how they operated. Based on their analysis, they reached a conclusion that illustrates the value that has been placed in using an administrative mechanism based on collaboration to address homelessness (i.e., the CoC model):

We conclude that HUD’s adoption in 1996 of a community-wide approach to distributing its competitive homeless assistance dollars has moved communities further in the direction of broad planning and program development than would have happened without the CoC approach. The ensuing networks of programs and services have been able to offer more support to homeless people, with more cohesion, than would otherwise have been possible . . . The mechanism that has brought communities together to accomplish this result should be maintained and strengthened. (p. xxi)

Discussion

Homelessness is a vastly complicated social problem that has always been and will continue to be a part of American history (Leginski, 2007). The present study’s scope is limited by its purpose to bring to light one aspect of the homelessness policy subsystem—the macrolevel process of the emergence and persistence of the CoC model. This model is a form of mandated collaboration that has become the federal government’s primary response to addressing homelessness. The current study adopted a policy feedback perspective to analyze the historical narrative of the CoC model to highlight how multiple feedback effects contributed to the emergence and persistence of this instance of mandated collaboration, which was eventually codified with the passage of the HEARTH Act. One of the most evident feedback effects is the negative feedback effect spurred by the original McKinney Act. In addition, HUD’s administrative implementation of the recommendations of the Priority Home! report (e.g., require local collaboration to receive funding) produced a large state-building effect: It resulted in the creation of hundreds of local CoCs across the country that became interorganizational networks that were mandated to collaborate to receive important federal funding to address homelessness in their locality. The decision to implement the CoC model following the Priority Home! recommendations may be viewed as a lock-in effect: Hundreds of CoCs began to form, and these new administrative structures were perceived as beneficial to addressing homelessness as is evidenced in multiple reports on the CoC model (Burt et al., 2002; Fuchs & McAllister, 1996). Policies have proven to be powerful catalysts for group formation and continuance, particularly surrounding the resources such policies provide (Mettler & SoRelle, 2014). Given the large amount of evidence regarding positive feedback effects related to the continuation of the CoC model, it is not surprising that the HEARTH Act eventually codified the CoC model nearly 15 years after its origination.

It appears likely that policy feedback effects from the HEARTH Act will continue to support a federal response to homelessness that continues to support mandated collaboration in the form of the CoC model. For example, as a result of the HEARTH Act HUD has enacted detailed regulations (specified in the CoC Program Interim Final Rule, 77 Fed. Reg. 147) that continue to encourage collaborative efforts conducted by CoCs. When CoCs meet certain collaboratively based criteria, they are awarded with the status of “high-performing CoC” and are granted certain rewards (e.g., additional flexibility with use of funding). HUD’s encouragement to the CoCs to strive to increase their collaborative efforts will likely extend the buy-in base for the CoC model to other community partners across the country; involving more constituent buy-in makes it more difficult to abandon using the model. However, it is possible that the additional administrative regulations imposed through the HEARTH Act may prove detrimental to the continuation of the CoC model. For example, a recent survey of the lead contacts for local CoCs across the country revealed that there is some backlash to the HEARTH Act (Jarpe, Mosley, Ray, & Reed, 2015). Many survey respondents expressed their concern with the increased administrative burden that is being placed on the CoC as an administrative entity. These individuals cite the lack of resources to adequately carry out the administrative duties of the CoC to meet the new regulations. While HEARTH made it possible to receive funding for CoC administrative duties, respondents felt that this funding is not adequate, and that it is in competition with funding for housing programs, which is counterproductive.

The findings from this survey and the present study suggest that a certain level of mandated collaboration (i.e., the CoC model prior to the HEARTH Act) may be acceptable to and even embraced by those being coerced to collaborate; however, there comes a point where too much mandated collaboration (i.e., the CoC model after the HEARTH Act) may result in resistance. Additional regulations and administrative duties placed on CoCs by HUD may be perceived as an attack on the rather substantial collaborative autonomy CoCs were historically afforded (Burt et al., 2002). The concept of autonomy is key to collaborative processes—organizations within a collaborative endeavor must balance organizational interests and goals with the interests and goals of the collaborative (Thomson, Perry, & Miller, 2007). This conceptualization of autonomy is important to organizations within instances of both emergent (i.e., voluntary) and mandated collaboration. The analysis of the CoC model suggests that there is an additional level of autonomy worthy of consideration in mandated collaboration—autonomy between the organization(s) giving the mandate (e.g., funders) and those being mandated to collaborate (e.g., nonprofit service providers). Mandates that constrict the ability of a collaborative to govern and organize the collaborative as they see fit to best meet the needs of their local context will be perceived as an attack on their autonomy. It is likely that a long-standing collaborative with well-developed ties would be more resistant to increasingly constricting mandates than a newly formed collaborative (Berry et al., 2004; Provan & Milward, 2001). Regarding the implementation of the CoC model, HUD provided no strict guidelines or preplanned configurations concerning how CoCs ought to be organized or governed. For example, some CoCs have formal leadership while others do not (Burt et al., 2002), and some collaborative homeless networks were even organized before HUD implemented the CoC model (e.g., Delaware County Homeless Services Coalition, 2014). If HUD were to implement guidelines that forced all CoCs into mandated modes of network governance (Provan & Kenis, 2008), CoCs that do not fit into the mandated mode would most likely resist. This is an important lesson for administrative entities that mandate collaboration as well as an area of research that demands more attention.

The present study also brought forward the often-overlooked macrolevel context of mandated collaboration (McNamara, 2016). Doing so laid the groundwork for future microlevel research of CoCs as mandated collaborative interorganizational networks within the same policy subsystem that likely possess substantial variation. All CoCs have the common, overarching goal of addressing homelessness (Provan & Lemaire, 2012); yet, there are hundreds of CoCs spread out across the United States operating in diverse geographic and political contexts. They are a rich source for detailed comparative analyses to explore interorganizational network governance, leadership, and structure (Popp, Milward, MacKean, Casebeer, & Lindstrom, 2014). Such work may demonstrate how different CoC arrangements contribute to outcomes, whether positive or negative, in different contexts (Provan & Kenis, 2008). In addition, as CoCs had substantial autonomy in their early development (Burt et al., 2002) individual CoCs’ evolutional history may inform current gaps in interorganizational network knowledge regarding how certain types of networks evolve and, in turn, thrive or fail (Popp et al., 2014).

The substantial autonomy that was afforded CoCs during their development puts into question where mandated collaboration belongs on a cooperation–coordination–collaboration continuum (Keast & Mandell, 2014; McNamara, 2016). The CoC model is indeed one of mandated collaboration, yet elements of the CoCs (e.g., key personnel, interorganizational design) were not dictated through the mandate (McNamara, 2016). This highlights the importance of understanding that interorganizational collaborative arrangements can come together in a variety of ways and have no presumed optimal arrangement even within the same policy subsystem (Nylen, 2007; O’Toole, 1993). The CoC model of mandated collaboration spurred the creation of numerous collaborative arrangements that likely span from loose cooperation to intense collaboration. Yet, all these different collaborative arrangements have the underlying mandate to collaborate to receive federal funding. This presents an instance where the primary role of the government in collaboration may be understood as the federal government incentivizing emergent collaboration to occur between local policy implementers (Provan & Lemaire, 2012). The collaboration is emergent, in that partners come together voluntarily—they are not required to participate and could opt to replace the potential federal funding received from collaboration through other mechanisms (e.g., seek philanthropic funding).

The present study demonstrated that a form of mandated collaboration that allows sufficient local autonomy is a potentially powerful and successful administrative development (Siddiki, Carboni, Koski, & Sadiq, 2015; Soss & Moynihan, 2014). Prior to the Great Recession, Leginski (2007) noted that “the consistent structural variable in America’s homelessness history is economic performance” (pp. 1-2). HUD’s implementation of the CoC model beginning in the 1990s resulted in the creation of numerous collaborative governance networks that were well established by the time the Great Recession hit. While making a direct causal link between the CoC model and homelessness rates was not the purpose of this study, it seems likely that these networks contributed at least in part to mitigating the impact of the Great Recession on homelessness, which places Leginski’s (2007) claim into question. The CoC model did not mandate strict rules for organizing collaboration, but instead allowed local autonomy and provided substantial incentive (i.e., McKinney funds) to organize (Rodríguez et al., 2007). Further detailed research is needed to understand how the CoC model may have mitigated increases of homelessness during the Great Recession, and how local autonomy and aspects of CoC maturity may be related. One possible direction would be to explore the level of maturity of CoCs’ HMIS—a requirement of HUD for CoCs to create data collection, management, and reporting systems utilizing local client-level data. HMISs are intended to serve as shared knowledge warehouses for CoCs and are expected to lead to data-driven interorganizational decision making, provided that the collaborative avoids “big data” pitfalls (Lavertu, 2016). A mature and sophisticated HMIS may positively affect a CoC’s ability to manage homelessness in their region. In addition, it is likely that some mature CoCs fared better than others during the Great Recession based on their established ability to successfully apply for McKinney funding.

In conclusion, this study has contributed to advancing research on mandated collaboration in a variety of ways. It has also answered the calls of public administration researchers to utilize a policy feedback perspective to demonstrate the interconnectedness of politics, policy, and administration (Soss & Moynihan, 2014), and calls to further the use of historical analysis in public administration research (Raadschelders, 2010; Spicer, 2004). By utilizing these analytical approaches, this study endeavored to contribute to “knowledge for practice” by providing a rigorous illustration of how mandated collaboration has been used as an administrative structure to address a complex public problem (Lynn, 1996; Thomson et al., 2007).

Footnotes

Acknowledgements

The author thanks the individuals who attended the presentation and provided useful feedback and support.

Author’s Note

An earlier version of this manuscript was presented at the 2016 Midwest Public Affairs Conference.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biography

Joseph A. Hafer is a doctoral candidate in public administration at the School of Public Affairs, Penn State Harrisburg. His research interests include collaborative governance networks in local governance contexts, public administration theory, and behavioral public administration. His work has appeared in Administrative Theory & Praxis.

References

American Recovery and Reinvestment Act of 2009, 16 U.S.C. § 2601, 42 U.S.C. § 13201, 42 U.S.C. § 15801.

Ansell

Gash

(2008). Collaborative governance in theory and practice. Journal of Public Administration Research and Theory, 18, 543-571.

Béland

(2010). Reconsidering policy feedback: How policies affect politics. Administration & Society, 42, 568-590.

Berg

(2013). The HEARTH Act. Cityscape: A Journal of Policy Development and Research, 15, 317-324.

Berry

F. S.

Brower

R. S.

Choi

S. O.

Goa

W. X.

Jang

Kwon

Word

(2004). Three traditions of network research: What the public management research agenda can learn from other research communities. Public Administration Review, 64, 539-552.

Burt

M. R.

Pollack

Sosland

Mikelson

K. S.

Drapa

Greenwalt

. . . Smith

(2002). Evaluation of continuums of care for homeless people: Final report. Washington, DC: The Urban Institute.

Campbell

A. L.

(2002). Self-interest, social security, and the distinctive participation patterns of senior citizens. American Political Science Review, 96, 565-574.

Cho

C. L.

Wright

D. S.

(2004). The devolution revolution in intergovernmental relations in the 1990s: Changes in cooperative and coercive state–national relations as perceived by state administrators. Journal of Public Administration Research and Theory, 14, 447-468.

Cronley

(2010). Unraveling the social construction of homelessness. Journal of Human Behavior in the Social Environment, 20, 319-333.

10.

Daugbjerg

(2003). Policy feedback and paradigm shift in EU agricultural policy: The effects of the MacSharry reform on future reform. Journal of European Public Policy, 10, 421-437.

11.

Delaware County Homeless Services Coalition. (2014). Homeless services coalition of Delaware county. Retrieved from http://www.delcohsa.org/adultfamily/HSC_Brochure_sept_14.pdf

12.

DePastino

(2003). Citizen hobo: How a century of homelessness shaped America. Chicago, IL: The University of Chicago Press.

13.

Emerson

Nabatchi

Balogh

(2012). An integrative framework for collaborative governance. Journal of Public Administration Research and Theory, 22, 1-29.

14.

Fuchs

E. R.

McAllister

(1996). The continuum of care: A report on the new federal policy to address homelessness. Washington, DC: Barnard-Columbia Center for Urban Policy. Retrieved from http://catalog.hathitrust.org/Record/003163820

15.

Garrett

D. G.

(2012). The business case for ending homelessness: Having a home improves health, reduces healthcare utilization and costs. American Health & Drug Benefits, 5(1), 17-19.

16.

Heclo

(1974). Modern social politics in Britain and Sweden: From relief to income maintenance. New Haven, CT: Yale University Press.

17.

Hoch

(2000). Sheltering the homeless in the US: Social improvement and the continuum of care. Housing Studies, 15, 865-876.

18.

Homeless Emergency Assistance and Rapid Transition to Housing: Continuum of Care Program; Interim Final Rule, 77 Fed. Reg. 147 (to be codified at C.F.R. Part 578)

19.

Hopper

(2003). Reckoning with homelessness. Ithaca, NY: Cornell University Press.

20.

HUD Exchange. (2017). Retrieved from https://www.hudexchange.info/

21.

International Monetary Fund. (2009). World economic outlook. Retrieved from http://www.imf.org/external/pubs/ft/weo/2009/01/pdf/text.pdf

22.

Jarpe

Mosley

J. E.

Ray

Reed

(2015). Homeless continuums of care: A report of research findings. School of Social Service Administration, University of Chicago. Retrieved from http://www.shpa-il.org/assets/CoC_Resources/coc%20uc%20survey%20report%209%2030%2015.pdf

23.

Jenkins-Smith

H. C.

Nohrstedt

Weible

C. M.

Sabatier

P. A.

(2014). The advocacy coalition framework: Foundations, evolution, and ongoing research. In Sabatier

P. A.

Weible

C. M.

(Eds.), Theories of the policy process (pp. 183-223). Boulder, CO: Westview Press.

24.

Keast

Mandell

(2014). The collaborative push: Moving beyond rhetoric and gaining evidence. Journal of Management & Governance, 18, 9-28.

25.

Kusmer

(2002). Down and out, on the road: The homeless in American history. New York, NY: Oxford University Press.

26.

Lamb

R. H.

(1984). Deinstitutionalization and the homeless mentally ill. Psychiatric Services, 35, 899-907.

27.

Lavertu

(2016). We all need help: “Big data” and the mismeasure of public administration. Public Administration Review, 76, 864-872.

28.

Leginski

(2007). Historical and contextual influences on the U.S. response to contemporary homelessness. In Dennis

Locke

Khadduri

(Eds.), Toward understanding homelessness: The 2007 National Symposium (pp. 1-1–1-36). Department of Health and Human Services & U.S. Department of Housing and Urban Development.

29.

Lowi

T. J.

(1972). Four systems of policy, politics, and choice. Public Administration Review, 32, 298-310.

30.

Lynn

L. E.

Jr. (1996). Public management as art, science, and profession. Chatham, NJ: Chatham House.

31.

Marshall

Rossman

G. B.

(2016). Designing qualitative research (6th ed.). Thousand Oaks, CA: SAGE.

32.

McGuire

(2006). Collaborative public management: Assessing what we know and how we know it. Public Administration Review, 66, 33-43.

33.

McNamara

M. W.

(2016). Unraveling the characteristics of mandated collaboration. In Morris

J. C.

Miller-Stevens

(Eds.), Advancing collaboration theory: Models, typologies, and evidence (pp. 65-85). New York, NY: Routledge.

34.

Mettler

(2002). Brining the state back into civic engagement: Policy feedback effects of the G.I. Bill for World War II veterans. American Political Science Review, 96, 351-365.

35.

Mettler

SoRelle

(2014). Policy feedback theory. In Sabatier

P. A.

Weible

C. M.

(Eds.), Theories of the policy process (pp. 151-181). Boulder, CO: Westview Press.

36.

Nylen

(2007). Interagency collaboration in human services: Impact of formalization and intensity on effectiveness. Public Administration, 85, 143-166.

37.

O’Toole

L. J

. (1993). Interorganizational policy studies: Lessons drawn from implementation research. Journal of Public Administration Research and Theory, 3(2), 232-251.

38.

Pierson

(1993). When effect becomes cause: Policy feedback and political change. World Politics, 45, 595-628.

39.

Pierson

(2000). Increasing returns, path dependence, and the study of politics. American Political Science Review, 94, 251-267.

40.

Popp

J. K.

Milward

H. B.

MacKean

Casebeer

Lindstrom

R. L.

(2014). Inter-organizational networks: A review of the literature to inform practice. IBM Center for the Business of Government. Retrieved from http://www.businessofgovernment.org/report/inter-organizational-networks-review-literature-inform-practice

41.

Provan

K. G.

Kenis

(2008). Modes of network governance: Structure, management, and effectiveness. Journal of Public Administration Research and Theory, 18, 229-252.

42.

Provan

K. G.

Lemaire

R. H.

(2012). Core concepts and key ideas for understanding public sector organizational networks: Using research to inform scholarship and practice. Public Administration Review, 72, 638-648.

43.

Provan

K. G.

Milward

H. B.

(2001). Do networks really work? A framework for evaluating public-sector organizational networks. Public Administration Review, 61, 414-423.

44.

Raadschelders

J. C.

(2010). Is American public administration detached from historical context? On the nature of time and the need to understand it in government and its study. The American Review of Public Administration, 40, 235-260.

45.

Rodríguez

Langley

Béland

Denis

J. L.

(2007). Governance, power, and mandated collaboration in an interorganizational network. Administration & Society, 39, 150-193.

46.

Sabatier

P. S.

Weible

C. M.

(2014). Theories of the policy process (3rd ed.). Boulder, CO: Westview Press.

47.

Schattschneider

E. E.

(1935). Politics, pressure, and the tariff. New York, NY: Prentice Hall.

48.

Schneider

Ingram

(1993). Social construction of target populations: Implications for politics and policy. American Political Science Review, 87, 334-347.

49.

Siddiki

S. N.

Carboni

J. L.

Koski

Sadiq

A. A.

(2015). How policy rules shape the structure and performance of collaborative governance arrangements. Public Administration Review, 75, 536-547.

50.

Skocpol

(1992). Protecting soldiers and mothers: The political origins of social policy in the United States. Cambridge, MA: Belknap Press of Harvard University Press.

51.

Soss

Fording

R. C.

Schram

(2011). Disciplining the poor: Neoliberal paternalism and the persistent power of race. Chicago, IL: University of Chicago Press.

52.

Soss

Moynihan

D. P.

(2014). Policy feedback and the politics of administration. Public Administration Review, 74, 320-332.

53.

Soss

Schram

(2007). A public transformed? Welfare reform as policy feedback. In Soss

Hacker

J. S.

Mettler

(Eds.), Remaking America: Democracy and public policy in an age of inequality (pp. 99-118). New York, NY: Russell Sage Foundation.

54.

Spicer

M. W.

(2004). Public administration, the history of ideas, and the reinventing government movement. Public Administration Review, 64, 353-362.

55.

Stewart, B. McKinney Homeless Assistance Act of 1987, 42 U.S.C. § 11301.

56.

Thomson

A. M.

Perry

J. L.

Miller

T. K.

(2007). Conceptualizing and measuring collaboration. Journal of Public Administration Research and Theory, 19, 23-56.

57.

U.S. Department of Housing and Urban Development. (1995). Review of Stewart B. McKinney homeless programs administered by HUD: Report to Congress. Washington, DC: Author.

58.

U.S. Department of Housing and Urban Development. (2010). HUD Press Release No. 10-260. Retrieved from https://archives.hud.gov/news/2010/pr10-260.cfm

59.

U.S. Department of Housing and Urban Development. (2016). The 2016 Annual Homeless Assessment Report (AHAR) to Congress. Retrieved from https://www.hudexchange.info/resources/documents/2016-AHAR-Part-1.pdf

60.

U.S. Interagency Council on Homelessness. (1994). Priority home! The federal plan to break the cycle of homelessness. Washington, DC: U.S. Department of Housing and Urban Development. Retrieved from http://catalog.hathitrust.org/Record/002887286

61.

U.S. Interagency Council on Homelessness. (2012). United States interagency council on homelessness historical overview. Retrieved from http://usich.gov/resources/uploads/asset_library/RPT_USICH_History_final__2012.pdf

62.

U.S. Interagency Council on Homelessness. (2015). About USICH. Retrieved from http://usich.gov/about_us/

63.

Weaver

(2010). Paths and forks or chutes and ladders? Negative feedbacks and policy regime change. Journal of Public Policy, 30, 137-162.

64.

Wichowsky

Moynihan

D. P.

(2008). Measuring how administration shapes citizenship: A policy feedback perspective on performance management. Public Administration Review, 68, 908-920.