LOCAL POLICYMAKING IN AN AGE OF FEDERALISM: THE ROLE OF FEDERAL INCENTIVES IN MUNICIPAL HOMELESSNESS POLICY By Melanie Bowers A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for t he degree of Political Science Doctor of Philosophy 2015 ABSTRACT LOCAL POLICYMAKING IN AN AGE OF FEDERALISM: THE ROLE OF FEDERAL INCENTIVES IN MUNICIPAL HOMELESSNESS POLICY By Melanie Bowers Over the last 15 years there has been a radical ch ange in the way that cities approach homelessness. For decades, cities had dealt with the issue by using geographic isolation and criminalization. Although these practices have not entirely been eradicated, they are joined by efforts to provide compassiona te, comprehensive services with the goal of ending, not just ameliorating, homelessness. In general, policy change is slow but in the case of homelessness there has been a dramatic and fairly rapid shift in municipal policy, prompting the question of why c ities have changed their approach . Through historical analysis, the dissertation suggests that cities have changed their approach to homelessness in response to federal incentives to develop 10 - year plans to end homelessness. These plans require cities to engage in long - term, comprehensive planning with the goal of ending homelessness or chronic homelessness within the city over a ten year period. By develop ing a pl an, cities can increase their scores for competitive federal homelessness funding; however, doing so has been entirely voluntary and not all cities have responded to the incentive. The diversity in responses to the 10 - year plan incentive raise s two questions that are at the heart of the dissertation: 1. What encourages cities to respond to federa l incentives? 2. Do incentives to engage in relatively symbolic behaviors encourage cities to adopt the tangible policy changes that promote federal goals? I empirically evaluate these questions using a novel dataset of over 300 cities. In chapter 2 I sugg est responses to federal incentives result from the interaction between local attention, capacity , and need. I test this idea in Chapters 4 and 5. Chapter 4 explores what encourages cities to respond to federal incentives. It argues that inc entives are a f orm of coo perative federalism and as a result cities should respond to them when they receive a benefit from doing so ; th is benefit is most likely when the incentive helps meet local needs. As a result, cities should be most likely to respond to federal in centives by developing a 10 - year plan when doing so helps them meet a fiscal, political or problem - based need. Analysis 10 - year plan adoption finds that cities are most likely to develop plans when they are fiscally and legally dependent on higher level g overnments, when local ideology is consistent with compassionate care for the homeless, and when they face a more severe homelessness problem. Chapter 5 asks whether the act of developing a 10 - year plan increases the likelihood that cities adopt three app roaches shown to reduce homelessness: increasing supportive permanent housing, using conventional public housing services for the homeless, and adopting a housing first approach. The chapter suggests that planning can be a largely symbolic act and uses the symbolic policy literature to suggest that cities with increased long - term attention, more extensive capacity and increased oversight would be more likely to move beyond 10 - year plan development to engage in substantive policy. The results are mixed but o ne consistent finding emerges: cities that adopt a 10 - year plan are no more likely to engage in substantive practices than their counterparts without a plan. Despite the finding that 10 - year plans are largely symbolic documents, the dissertation suggests that federal incentives have been successful in that they have helped change the conversation around homelessness and increased attention to the issue. Further, it suggests opportunities to improve the effectiveness of federal homelessness incentives for t he future. Copyright by MELANIE BOWERS 2015 v To Gary, Sal, Santa and the many other homeless men and women who have taught me about kindness, love and humanity. To my mother who helped me understand that privilege comes with a responsibility t o others. And, to my husband who never let me give up on myself, even when I fought tooth and nail to do so . vi ACKNOWLEDGMENTS I would like to express my deepest gratitude to my committee , Laura Reese, Josh Sapotichne, Sarah Reckhow, and especially to my chair Ric Hula . For six years I have come to you for support, guidance and council, and for six years you have opened your doors to me with warmth and enthusiasm . You have encouraged me, challenged me, and at times given me a hard dose of reality. Throu gh it all I have known that you are rooting for me, willing me to succeed. That is a gift that not everyone has and for me it has been invaluable. I would like to thank my parents Trey Bowers and Karen Garritson for putting up with my being at university for 10 years. I know that it is not easy to have me half way across the country earning less than minimum wage, but in spite of the difficulty, you have been an amazing support system. I want to give thanks to both my sister Natalie Ward and my best frie nd Autumn Fischer . The two of you have helped keep me sane. I am so grateful that you have stuck around through estament to your character and friendship. Finally, to my husband. Thank you, thank you thank you. Under the best circumstances I am quit. Instead of saying that though, you have patiently waited on the other side, knowing that my craziness and sleepless nights would not last forever. You have been my rock, my safety net, and the most important person in my life. I never w ould have finished without you, that is ce rtain. I love you each day more th an the last and I am so grateful to have gone on this journey with you. vii TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ................................ ..... ix LIST OF FIGURES ................................ ................................ ................................ ................................ .... x CHAPTER 1. Introduction ................................ ................................ ................................ ........................ 1 Denver ................................ ................................ ................................ ................................ ...................... 2 Cleve land ................................ ................................ ................................ ................................ ................. 3 Los Angeles ................................ ................................ ................................ ................................ .............. 4 CHAPTER 2. Theory and Literature Review ................................ ................................ ....................... 11 F ederal Incentives ................................ ................................ ................................ ................................ . 12 Explaining Local Responsiveness to Federal Incentives ................................ ................................ .... 17 Responsiveness Typology ................................ ................................ ................................ .................... 21 Predicting Responsiveness: Attention, Capacity and Need ................................ ............................... 22 Attention ................................ ................................ ................................ ................................ .............. 23 Capacity ................................ ................................ ................................ ................................ .............. 26 Need ................................ ................................ ................................ ................................ .................... 28 Homelessness and Municipal Responsiveness ................................ ................................ .................... 32 CHAPTER 3. A History of Intergovernmental Dynamics in Homeless Po li cy ................................ ... 36 Indentured Servitude, Poor Houses and Outdoor Relief: The Early Years ................................ .... 36 Feder al Expansion: The Great Depression, WWII and Beyond ................................ ...................... 41 Basic Services through Local Response and Federal Intervention: The 1980s and 1990s ............. 43 CHAPTER 4: Explaining the Adoption of 10 Year Plans to End Homelessness ................................ 54 Planning and Prevention through Federal Incentives: 10 Year Plans to End Homelessness ........ 54 Why Adopt? Insights from Fiscal Federalism and Urban Politics ................................ ................... 62 Constituent Demand ................................ ................................ ................................ ............................ 67 Fiscal Solvency ................................ ................................ ................................ ................................ ... 67 Community Problems ................................ ................................ ................................ .......................... 69 Evaluating the Evidence: Data and Methods ................................ ................................ ..................... 70 Sample ................................ ................................ ................................ ................................ ................. 70 Specifying the Model ................................ ................................ ................................ ........................... 72 Evaluating the Evidence: What has Influenced Individual City Adoption? ................................ ... 78 Discussion and Conclusions ................................ ................................ ................................ ................. 84 CHAPTER 5. Symbol or Substance?: Assessing the Real Impacts of Federal Incentives to End Homelessn ess ................................ ................................ ................................ ................................ ............. 89 Predicting Symbolic and Substantive Policies ................................ ................................ .................... 91 Data and Methods ................................ ................................ ................................ ............................... 100 Independent Variables and Controls ................................ ................................ ................................ 103 Evaluating Expectations ................................ ................................ ................................ ..................... 107 viii Discussion and Conclusion ................................ ................................ ................................ ................. 116 CHAPTER 6. Conclusion ................................ ................................ ................................ ....................... 120 Implications of Findings ................................ ................................ ................................ ..................... 121 Limitations and Opportunities for I mprovement ................................ ................................ ............ 128 Future Research ................................ ................................ ................................ ................................ .. 130 A Final Thought ................................ ................................ ................................ ................................ .. 132 APPENDIX ................................ ................................ ................................ ................................ .............. 133 WORKS CITED ................................ ................................ ................................ ................................ ...... 143 ix LIST OF TABLES Table 2.1 . Responsive Action Typology 21 Table 3.1 . Contents of McKinney - Vento Homeless Relief Act o f 1987 47 T able 4.1. Cox Model Predicting w hen Cities Adopt 10 - Year Plans to End Homelessness 61 Table 4.2 . Descriptive Statistics for Sample Cities ... 7 2 Table 4.3 . Table 4.4. Logistic Reg ression Predicting 10 - Table 5.1 . Summary of Included Variables Table 5.2. Linear Regression: Predictors of Beds and Permanent Housing Avai lable to Homeless 109 Table 5.3. the Homeless. . 110 Table 5.4 . Logistic Regression : Predictors of Municipalities Joining and Actively Participating in 100,000 Ho 112 Table 5.5 . Predicted Probabilities for Binary Measures of Substantive ... ... 113 x LIST OF FIGURES Figure 4.1. Municipal Adoption of Plans to End Homelessness 58 Figure 4.2. Hazard Rate with and without Federal Incentives ... 61 Figure 4.3. Predicted Probability of Adoption Based on Percent of Revenue from Federal IGR 82 Figure 4.4. Predicted Probability of Adopti on Based on Functional Autonomy .. 82 Figure 4.5. Predicted Probability of Adoption Based on Homeless Population . 83 Figure 4.6. Predicted Probability of Adoption Based on Homeless Population Living in . 84 1 CHAPTER 1. Introduction Mayor John Hickenlooper incorporated thi s bold statement into a press release announcing the formation of the Denver Commission on Homelessness. The commission, he explained, would fulfill a campaign promise to develop and mobilize an advocacy coalition for the homeless. Their primary purpose wo uld be to develop a 10 - year plan to end homelessness, an act that would strengthen the city and answer a 2002 call from the US Department of Housing and Urban Development (HUD). Although the federal government has provided funding for homeless services sin ce the late 1980s, during the early 2000s President Bush and his administration stepped into a leadership role that would alter the way communities throughout the nation talk about and approach homelessness. Echoing efforts in the homeless advocacy communi ty, HUD called on cities to change their existing approaches to homelessness by developing long - term, comprehensive planning that incorporates compassionate care and has a goal of ending homelessness in 10 years. While a handful of cities like Denver quick ly responded to the call, many others needed extra encouragement and the federal government provided this in the form of a fiscal incentive to develop 10 - year plans. Municipal responses to federal efforts have been mixed, with incentives encouraging plan a doption but not always substantive action, and this variation provides important insight into federal - local relations generally and the role of federal incentives in local policymaking more specifically. Denver, Cleveland and Los Angeles help clarify the v - year plan incentives; vignettes of their stories follow. 2 Denver In pledging to develop a 10 - year plan Mayor Hickenlooper highlighted the seriousness of ly 80 percent increase in homelessness over the previous decade. He also expressed his faith that a strong coalition and carefully thought out - term attention to the issue have led to marked su - Year plan. What he had to say shed light on the way that city officials conceptualized their role in homelessnes s recognition that homelessness is an expensive and dangerous problem, one existing approach of providing emergency services was not helping. We presently spend roughly $70 million a year on back - end services that are clearly not addressing the root problems of homelessness. Our plan to end homelessness will cost less than $13 million annually, combining dramatic net savings with measurable results. We have chosen a model that is both compassionate and results - not about hand - outs, coddling, or expanding the welfare system. It is about creating opportunity, about helping people regain control of their lives, so that we can all regain control of our community, (Hickenlooper 2005, 5). in its myriad successe s. The city added nearly 3,000 housing opportunities for the homeless, created over 100 percent of the permanent housing it identified in the plan, prevented more than 6,000 families from becoming homeless, created nearly 7,000 employment and training oppo rtunities for the homeless (Current 2015) and reduced its homeless population by about 40 percent, including a nearly 50 percent reduction in family homelessness (HUD 2005, 2014). It also developed innovative policies to address quality of life issues pose d by large homeless 3 populations. For example, to minimize panhandling without criminalizing it, the city installed hundreds of parking meters downtown where residents can donate money to organizations that serve the homeless rather than giving directly to those begging. In many ways Denver is a poster child for homelessness prevention. Both former Mayor John Hickenlooper and current Mayor Michael Hancock have served entrepreneurial roles, developing new programs, forging regional partnerships, and keeping municipal government focused on homelessness issues. The city has a strong advocacy coalition of both governmental and non - governmental actors that has been able to consistently work together towards common adopted by other communities and a variety of cities have looked to Denver as a model for developing their own programs. Denver represents what is possible in local homelessness prevention and highlights how federal incentives can help national goals can b e translated to local action. However, its success is only one side of a diverse spectrum of municipal approaches to homelessness. Cleveland On the other side of the spectrum is Cleveland, OH. The city that has yet to respond to equests and incentives to develop a 10 - year plan to end homelessness, has no municipal employee in charge of homeless issues, and has reduced the homeless of the Northeast Ohio Coalition for the Homeless, Brian Davis, had this to say about conditions in the area. In Cleveland, there is a rising number of fair housing c omplaints combined with an inadequate supply of housing that meet basic requirements. We have talked about a funding source in the creation of a Local Housing Trust Fund, but it has not happened. There are waiting lists of 19,000 for public housing, 6,000 for voucher 4 programs, and 64,000 people applied for housing in 2011 when Section 8 was opened. 22,000 people are homeless and a growing number of homeless families are attempting to find shelter every night. There were 30 families sleeping in the overflo w shelter in Cleveland last week, because we did not have space. Also and unfortunately, Cleveland did not receive any of the state tax incentives to build housing in the competition announced last month, which means a year of not developing any affordabl e housing. We also see repeated cuts to housing and homeless programs with Sequestration and other budget austerity programs resulting in the closing of shelters, elimination of rental assistance, and reductions in staffing for housing and homeless progra ms, (Davis 2014) Cities like this one are more numerous than one might expect and make it difficult to imagine the federal government achieving its goal of ending homelessness through local action. Yet again, however, Denver and Cleveland are on two extre mes. The reality is that most cities fall somewhere between these two, often developing plans but struggling to carry them out. Communities in the middle represent the complexity of addressing homelessness, developing homeless policies, and relying on loca lities to carry out federal goals. No city better exemplifies the struggle to respond to federal incentives and homelessness than Los Angeles. Los Angeles With the second largest homeless population in the nation (HUD 2014) Los Angeles is acutely aware of its homelessness problem but has struggled to develop a comprehensive approach to solving it. Los Angeles, like Denver, was among those early cities who began creating 10 - year plans in 2003; unlike Denver, however, limited public support and a lack of pol icy entrepreneurship meant that by 2005 the plan was dead (Roberts 2012). In the mean - time the city continued a century - Row, an area of town that looks more like a third world slum than a modern US neighborhood. It also spent tens of millions of dollars annually arresting and prosecuting the homeless. This effort included the Safer City Initiative, which was a police crackdown on Skid Row that used broken 5 windows theory to justify prosecuti Los Angeles would accept the level of violence and crime that this area of the city has had to withstand over t program cost the city more than it was spending on all homeless services combined (Homes 2009) and while it did significantly reduce crime in the area, homeless advocates argue tha t it did homeless into areas without services and put destitute individuals into a criminal justice system they could not easily escape (Homes 2009; Jaffe 2009). I n 2010 the city did pass a 10 - year plan to end homelessness, but its ability to promote change is questionable. As a case in point, the city did not assess how much it was spending on homelessness, where the money was going, or how much was needed to bette r address the issue appears to be no consistent process across city departments for dealing with the homeless or with urther, the plan offered little insight into how it would be carried out or what measures would be in place to evaluate progress. Despite these challenges, Los Angeles has followed the national trend of reducing its homeless population over the last 10 ye ars. Today the city has roughly 34,000 homeless individuals, down from about 70,000 in 2005 (HUD 2005 2014). Although some argue that the population has simply been displaced, there have been increases in affordable housing and emergency beds. Further, in 2013 the city elected a new mayor, Eric Garcetti, who may well end homelessness in Los Angeles and he recently blocked city council legislation that made it 6 easie r for police to remove and destroy homeless belongings and temporary shelters (Holland 2015b). Conflict between the mayor, city council and the police over the proper approach to homelessness makes for a challenging policy environment that is made worse b y competing demands from public not - in - my backyard concerns and federal imperatives to prevent and end homelessness. Just weeks ago in June of 2015 the LAPD held community meetings with Los about the homeless in the community (Banks 2015). These meetings came only months after two separate instances where the LAPD killed unarmed homeless men (Bloom et al 2015; US 2015), and only weeks before they killed another (Blo om and Pascucci 2015). This police activity has occurred at the same time that the mayor has attempted to fight city council ordinances that increase the criminalization of the homeless, highlighting intragovernmental conflict. This conflict is unlikely to efforts are likely in vain. The council does not need his support to pass the laws and he does not have the authority to stop police from enforcing them. All in all homeless policy in Los Angeles is messy. While the city has adopted a 10 year plan and has made some strides to reduce the homeless population residing within city limits, compassionate policy is limited. The conflict between city officials has perhaps kept the most draconi an policies from being implemented but it has also inhibited a comprehensive approach to addressing the issue. Not in my backyard demands have seemingly played a much larger role than federal and advocacy efforts to move towards compassionate care. And ef forts to engage the Mayor as a policy entrepreneur have fallen short because he lacks significant authority. In 7 - level plan to end homelessness and lack of coordination are unlikely to help local offic ials meet the federal goal of ending homelessness It is in part this messiness that makes homelessness an interesting lens through which to study intergovernmental relations. Homelessness is an acutely local issue but it involves federal - local, federal - sta te, state - local, interlocal, and local - ngo relations that make for an incredibly complex and layered policy environment. One the most interesting intergovernmental relationships to emerge has developed over the last several decades between federal and loc al governments. In 2001 the federal government made a national goal of ending chronic homelessness in 10 years ( United 2010), a goal that it had no way of meeting on its own. It was because of the mismatch between the goal and federal capacity to achieve i t, that a unique federal - local relationship began. While in other areas of policy the federal government has used mandates to force lower - level governments to do their bidding, in this case they have used cooperative federalism in the form of incentives to alter municipal action. More specifically the federal government created financial incentives for cities to alter their approach to homelessness and engage in long - term, comprehensive planning with the goal of ending rather than ameliorating the problem. As the cases above demonstrate, localities have responded in vastly different ways to What is unclear however, is why cities have responded in the ways they have. What has lead Denver and Los Angeles to respond to the federal incentive by developing a 10 - year plan while Cleveland has not? Why has Denver not only developed a plan but also followed it up with substantive practices while Los Angeles has struggle d to do so? This dissertation takes up these questions by exploring the effects of federal 10 - year plan incentives on municipal behavior, the 8 predictors of municipal responses to incentives, and the characteristics that predict a symbolic or substantive re sponse using a dataset of over 300 cities. Chapter 2 develops a theoretical framework for investigating these questions. More specifically it lays out the argument developed in existing work that federal incentives can be powerful drivers of policy change . These incentives can be financial carrots or sticks as well as non - tangible policy statements. One of the places that the existing incentives literature can be strengthened is in investigating what encourages responsiveness to incentives and the remainde r of the chapter uses organizational and policy theories to suggest that responsiveness results from the interaction between municipal attention, capacity, and political, fiscal, and problem - based need. In other words, cities act based on whether they know about an incentive, whether they have the money and human resources to respond to that incentive, and whether the incentive meets an underlying political, fiscal or problem - based need. Chapter 3 lays the foundation for understanding how homelessness fits into the incentive framework by providing a historical analysis of US homelessness policy. The important takeaway from this chapter is that in the absence of federal intervention cities have a long and robust history of criminalizing and isolating the home 20 th century homelessness was addressed exclusively at the local level. Although the specifics changed over the years - for example the nation did away with indentured servitude - the general picture remained fair geographically isolate them in particular areas of the city, or remove them entirely from municipal boundaries. During what was deemed the homelessness crisis of the late 1970s and early 1980s the federal government got involved in homelessness, for the first time creating funding for homeless - specific programs and services. This effort significantly increased 9 emergency services and set the tone for the federal government to take a more pronounced role in homeless policy during the 2000s. intervention into homeless policy and specifically the incentives it has provided to develop 10 - year plans to end homelessness have radically changed the way that some cities think about the problem. The federal government committed the nation to ending homelessness in the early 2000s and proceeded to provide both policy - based and financial incentives to encourage cit ies to develop 10 year plans to end homelessness. Since that time, there has been a pronounced transformation in the ways that many localities deal with homelessness, transitioning focus away from criminalization and isolation towards prevention, case mana gement, and comprehensive service provision. Empirical evidence for this assertion comes in the form of an event history analysis that highlights the catalytic role federal incentives played in the timing of plan adoption. From there, the chapter explores the determinants of individual plan adoption. Working from the idea that cooperative relationships only emerge when both parties benefit, it suggests that when cities have political, fiscal, and problem - based needs that can be met by developing a plan they are more likely to respond to federal incentives. It finds evidence for this assertion through analysis of a novel dataset of over 300 cities. Chapter 5 acknowledges the limits of planning without action by investigating the determents of substantive hom elessness policy. Existing content analysis of over 200 10 - year plans has shown how hollow many plans are, missing vital components like funding sources, timelines, and accountability structures. Given that the federal government has pushed 10 year plans i n order to help achieve a very tangible and ambitious goal, these deficits have potentially serious consequences. Building on theories of symbolic policy the chapter suggests that attention 10 and capacity should be much more important in predicting substanti ve action. Specifically, it lays out the expectation that attention in the form of external monitoring and a policy community priorities will increase the likeli hood of adopting substantive practices. The chapter explores these expectations by analyzing the predictors of three types of substantive action that represent practices which have been identified as essential for eradicating homelessness: creating permane nt housing for the homeless, having the public housing authority engage with and prioritize the homeless, and adopting a housing first approach. There is mixed support for the expectations laid out above, but the more important finding is that having a 10 - year plan does not predict any of the substantive behaviors, suggesting that the plans are largely symbolic in nature. Chapter 6 concludes the dissertation by connecting the chapters with insights and policy implications. The most important points from th is chapter are threefold. First, while federal incentives have been extremely effective at promoting plan adoption, plans are insufficient to promote substantive practices. Second, there have been important gains made in coalition development within cities ; however, there needs to be a more concerted effort to clarify the simultaneous roles of local, nonprofit and state actors. Finally, there needs to be greater thought about the relationship between cooperative action and oversight. The chapter concludes b y suggesting that federal incentives have been very effective at instigating policy change, even if they are largely symbolic. This change has increased attention to homelessness as a problem and has encouraged communities to engage in discourse about the ways they deal with this complex issue. This, I contend, is an important end in and of itself, it is an outcome that makes the federal incentive to promote 10 - year planning a success. 11 CHAPTER 2. Theory and Literature Review Since the Great Depression f ederal intervention in municipal affairs has been an important and enduring dimension of intergovernmental relations. Nevertheless, we have a more limited understanding of federal - local interactions than one might expect. In part, this is because our cons titution recognizes only national and state governments as legitimate, independent entities. This places municipalities in a dependent position, existing at the discretion of state governments. This federal arrangement has led scholars to describe US citi es as having significant political power but little legal legitimacy (Miller 2002) and has encouraged researchers to focus heavily on federal - state and state - local interactions, paying less attention to federal - local ones (Davidson 2007, Shipan and Volden 2005). Despite the relatively limited academic focus on federal - local relations, the federal government has played a significant role in establishing municipal priorities, policies, and initiatives. Davidson (2007) explains the disconnect between scholar ship and reality, writing: reference to each other. Commentators portray parallel, largely disconnected worlds in which the federal government relates only to the s tates, and the states, in turn, hermetically encompass role in areas of This chapter reviews one dimension of federal - local relations, federal incentives to localities, and develops a framework for evaluating when and how cities respond to this policy 12 tool. Incentives - in the form of policy imperatives, fiscal resources, and federal inactivity - can be a significant driver of municipal action, but the degree to which cities respond to these incentives depends upon the interaction of inter nal conditions that are familiar to policy scholars. These include: attention, capacity and need. In the remainder of this chapter I explore existing literature on federal incentives, which is primarily based on state - federal interactions, I then discuss t he need to expand our understanding of incentives to local governments, and develop expectations about the conditions that facilitate municipal response. In the final section I explain why homelessness is an ideal case for investigating the role of federal incentives on local governments. This chapter serves as a general framework for understanding the determinants of municipal responsiveness to federal incentives. In Chapters 3, 4, and 5, I incorporate additional literature that extends the general framewo rk and provides more nuanced understandings of responsiveness. Federal Incentives The federal government faces substantial barriers to carrying out its policy goals because of constitutional limits to its authority and a lack of capacity to implement policies across 50 states and thousands of local governments. As a result, it must rely upon state and local governments to adopt its policy priorities and implement its initiatives. To ease the process, the federal government has used a variety of tools, most notably devolution, mandates, and incentives. Much has been written in recent years about the nationalization of policy and the conflictual federalism that has emerged as a byproduct (for example see Conlan 2006; Conlan and Dinan 2007; Conlan and Posn er 2012; Kincaid 1990; Posner 2007). In fact, many of the studies on local - federal interactions address topics concerning coercive federal behavior in the 13 form of mandates, preemptions and devolution (Eads 2007). This work points to cases like No Child Lef t Behind, environmental regulations and immigration policies that force cities to carry out federal goals, often at the expense of local priorities (for example see Coleman 2012; Eisinger 1998; Provine et al 2012; Shelly 2011; Steinzor 1996; Zimmerman 1992 ). Incentives, in contrast, are distinct from mandates and preemptions in that they are voluntary and this voluntary nature makes them far more reminiscent of cooperative than coercive federalism. Whereas coercive mechanisms require, incentives ask . The i mplications of this cooperative relationship for homelessness policy are discussed more fully in Chapter 4, so for now the most important points are that cooperative federalism, including incentives, relies upon voluntary relationships, requires both parti es to benefit from the relationship (Elazar 1964, 1987), and recognizes that both parties are necessary to achieve the desired outcome (Davis 1967; Wong and Peterson 1986). While devolution and mandates have received sizeable scholarly attention (see f or example Boles et. al. 2011, Cimitile et al 1997, Gullo 2004, Gullo and Kelly 1998, Lowi et. al. 2010 ), incentives have received less and the majority of work has focused on federal - state interactions. Nonetheless, this policy tool provides a powerful alternative to forcing the municipal hand. Mandates, devolution, and state - imposed institutional constraints either prevent cities from doing what they want to do or force them to do things they likely do not want to do. In contrast, incentives offer a wa y for higher levels of government to achieve their goals without directly interfering in municipal affairs, maintaining a semblance of municipal autonomy. In one of the first articles to expressly test the role of federal incentives on policy adoption Wel 14 eprive states strictly carrot/stick conceptualization has been expanded to include not just tangible, monetary incentives but also written or verbal policy s tatements and federal inactivity. The federal government frequently provides indications of its policy agenda and national goals through written and verbal statements, some of which include information that indicates what it wants lower - level governments t o do (Dubnick and Gitelson 1981, Hamilton and Wells 1990). These statements provide states with an incentive to alter their behavior because they serve as a cue about possible future federal actions such as new policies and altered funding opportunities (A llen et al 2004, Karch 2006). In addition, though it has not been studied, voluntarily agreeing to engage in behavior that the federal government wants has the potential to have positive political consequences for cities that range from future financial al locations to national government action on an issue is strong and clear, state leaders can use it to marshal support for the adoption of related policy pro In addition to monetary and policy incentives, federal inaction serves as a passive incentive for cities to act (Allen et al 2004; Riverstone - Newell 2014). In one of the few works that looks at federal - local incentives , Riverstone - Newell (2014) argues that federal inaction combined with an evolving intergovernmental context has spurred municipal activism in immigration, LGBT rights and foreign policy. She contends that when the federal government fails to act on issues that create significant challenges for localities, municipal governments have an incentive to pursue policy change and engage in activism they would be unlikely to engage in otherwise. 15 From this scholarship we can develop a working definition: federal in centives are a tool, either active or passive, that provides cities with an impetus to voluntarily alter their behavior in ways that are consistent with federal goals. Active incentives come in the form of tangible reward or punishment and in written and v erbal policy statements, while passive incentives come in the form of federal inaction on issues that significantly impact lower level governments. These types of incentives can signal a new area the federal government would like to see addressed or can al ter the way lower - level governments treat an existing issue. Further, active incentives have the potential to be a win - win for both federal and lower - level governments, increasing the likelihood that the federal government can achieve its policy goals, whi le giving lower - level governments access to political and fiscal resources. What we know about incentives comes out of the policy diffusion literature and largely neral diffusion trends. This literature indicates that federal incentives do influence state policy making (Krane 1993; Soss et al 2001; Volden 1999; Welch and Thompson 1980) by either making states dependent on the federal government to carry out programs that they could not otherwise afford or by making programs cheaper so that states can use resources that would typically go to the program to fund other priorities (Peterson, Rabe, Wong 1991 ). In this way incentives help 2006, 426). But importantly, not all incentives are created equal and amongst the different types influenc financial reward - the carrot - is more effective in promoting policy change than the stick (Welch 16 and Thompson 1980) and direct, unambiguous policy statements/actions are mo re likely to elicit a response than those that are less clear (Allen et al 2004). In addition, federal incentives do not simply encourage policy innovation among individual communities but actually increase the speed with which policies diffuse across the nation (Welch and Thompson 1980), increasing the overall impact of a given policy change. This may be because states are more positive in their responses to incentives than to mandates (Albritton 1989; Groga n 1999), demonstrating less resistance to the vo luntary nature of the policy tool. That said, at least in the case of renewable energy, mandates like renewable energy portfolio standards appear to be more likely to spur real policy change than economic incentives (Bird et al 2005; Horner et al 2013). In sum, incentives can be powerful tools to promote federal goals. They encourage states to engage in policies at a speed and fervor they would otherwise be unlikely to show. That said, there is still much we do not know. As noted in the working definition o f incentives developed earlier, responding to incentives is voluntary and lower level governments can choose not to do what the federal government wants. An important question that comes out of this voluntary nature is: What encourages lower - level governme nts to respond to incentives in ways that are consistent with federal goals? This question is at the heart of this dissertation, but even the federal - state literature offers limited insight into what makes individual communities respond to incentives. Ther e are two notable exceptions, however. First, scholars have found that responding to incentives can reflect political motivations as much as anything else. For example, in their study of the American Recovery and Reinvestment Act of 2009 Miller and Blandi ng (2012) find strong evidence that the speed with which governors certified their intent to respond to federal incentives reflected partisan 17 tendencies, such that Republican governors took significantly longer to respond to the incentive than Democratic o nes. They argue the economic and political climate made it untenable for governors to outright reject the incentive, but their study suggests that under different circumstances political mismatch between upper and lower levels of government could result in a lack of responsiveness on the part of lower governments. If this were the case, the study suggests that failing to respond to the incentive would serve as a political statement rather than as a reflection of the quality of the incentivized program or s tate need for resources. The second major finding about what makes communities likely to respond to incentives comes out of research on nuclear waste disposal. In two important works Sapat (2004) and Daley and Garand (2005) find strong evidence that state s develop specific nuclear waste disposal policies in response to problem severity. Daley and Garand (2005) expand this finding, emphasizing the importance of problem severity by showing that politics and ideology played very little role in policy adoption ; states developed policies regardless of the their ideological tendencies and existing environmental regulations when the problem was severe and they had the resources (state economic conditions significantly predicted adoption) to carry out legislation. In addition, they were more likely to adopt legislation when there was greater regional pressure to do so. This work suggests that tangible problems play a more important role than many political theories suggest and offers a focal point for this study Ex plaining Local Responsiveness to Federal Incentives The influence of federal programs on local behavior has been studied across numerous policy areas. For example, scholars have investigated how the Intermodal Surface Transportation Efficiency Act (ISTEA ) increased municipal authority over transportation and encouraged regional transportation planning (Goldman et al 2000), they have studied how the introduction of 18 2 011), as well as the influence of many other federal initiatives on local policy. Though these works typically do not discuss federal - local relationships in terms of incentives, they do investigate the multifaceted ways that cities respond to federal progr ams. Here, I try to bridge this local policy literature with the incentives literature. In terms of work that specifically discusses incentives to localities. There are two topics that have been investigated. First, the ways that federal incentives in the 1940s encouraged cities to return to municipal planning and spurred the modern urban planning profession (Hanchett 2007). Second, the role of state incentives on local smoking ordinances (Shipan and Volden 2005, 2006, 2008, 2014). This scholarship is rich in findings about local policy diffusion and does an excellent job of fleshing out the complexity of policy learning between states and Volden 2008, 853). In other words, there is evidence that cities respond to pressure from both higher level governments (state and federal) and from other localities by adopting policy innovation s. Second, both small and large cities are susceptible to what Shipan and Volden call Together, these findings lay the groundwork for the present research because they suggest that cities interact with their external environments and can be responsive to incentives from higher - level governments. Since the demise of comprehensive federal urban policy in the 1970s and 1980s, incentives have become an important tool f or the federal government to influence local priorities, affecting areas as diverse as e - government, construction code standards, and renewable energy 19 (Gamkha r and Pickerill 2012; GAO 2010 ). Incentives are an outgrowth of a changing federal - local relations hip, a relationship that began when states and localities asked the federal government to help them combat problems resulting from the Great Depression. Responding to rams that launched an intentional, long - term federal - urban relationship. "The evolution of an active and direct federal - city relationship since the 1930s has helped to overcome the institutional neglect of cities within the American federal system. It has also empowered cities as units of government by granting them political recognition in national policy and by functionally acknowledging their right to make claims on national priorities and resources," (Gunther 1990, 9) For the next 50 years, the federa l government actively engaged in efforts to address urban - urban policies 1 cities substantial resources. By the 1960s significant portions of municipal budgets were made up blems sought and facilitated a cooperative relationship with cities, treating urban problems as national ones (Aaron 1978; Barnes 2005; Ladd and Yinger 1991; Unger 1996). Starting in the 1970s with commitment to urban problems was fundamentally undermined. The Reagan administration made significant inroads towards dismantling compre hensive federal urban policy because of beliefs in a. smaller government; b. local self - sufficiency; c. the exit of federal government from 1 For example efforts to redevelop the urban core happened at the same time as highway expansion that led to suburbanization. 20 local affairs; and, d. the promotion and expansion of the market and market - based solutions (Dreier, 2004; Pierson 1 995). In his first year in office, Reagan cut spending on programs that assisted US cities by 23 percent and attempted, though largely failed, to make additional cuts in his subsequent years in office. These cuts sent a clear message to US cities - the fed eral government is no longer responsible for your success (Garner 1990). In addition to general cuts, the Reagan administration also altered the way that cities received funding, transitioning away from program - specific resources to block grants. Throughou t the 1990s and continuing today, the federal government has largely continued to diminish financial resources for cities (Martell and Greenwade 2012), but at the same time there has been a nationalization of policy priorities (Conlan and Posner 2012). The lack of a comprehensive federal - urban policy paired with national goals that require local implementation has led the federal government to create mandates, facilitate second - order devolution, and develop incentives to carry out its goals as well as devel oping numerous policies that have urban implications without being directly about cities (Sapotichne 2010). And so the question reemerges: in the absence of a mandate or forced devolution, what makes cities respond to incentives and comply with federal go als? For this project responsiveness is defined in terms of action, specifically those acts that alter municipal behavior in ways that are consistent with federal goals. The reality is that even in the face of an appealing financial reward, not all cities will respond to incentives. Some do not need inducement to engage in the desired behaviors and others may either be unable or unwilling to respond. Put another way, on one extreme cities can be self - driven, proactively addressing issues without federal i ncentives; on the other extreme, cities can intentionally choose not to act despite federal 21 incentives. Depending on the choice, federal incentives can facilitate behavior that would otherwise be unlikely to occur or have little impact on local policy. Res ponsiveness Typology How cities respond to mandates can be described using a two dimensional typology that accounts for available incentives and municipal responses. Though an oversimplification, this typology, depicted in Table 2. 1, provides theoretical leverage by placing cities in broad categories that identify common traits and characteristics that facilitate or inhibit responsiveness. Table 2. 1. Responsive Action Typology Federal Incentive Yes No Responsiv e Action Yes Responsive Proactive No Unresponsive Status Quo Responsive cities change behavior in ways that are consistent with the federal incentive. In contrast, an unresponsive city does not alter their behavior in the face of federal incenti ves. Proactive cities respond to a given issue even when federal incentives are lacking. Evidence from research on state - local learning suggest that proactive cities may actually motivate the federal government to create incentives to induce policy adoptio n across municipalities (Mossberger 1999; Shipan and Volden 2006). Status quo cities do not alter their behavior without incentives. 22 This dissertation focuses on responsive and unresponsive cities, setting up a framework to evaluate responsiveness. To ide ntify why cities fall into each category, I turn to the policy and federalism literatures, which offer three organizing concepts that are at the core of local responses to federal incentives: attention, capacity and need. Predicting Responsiveness: Attent ion, Capacity and Need In his seminal study, Mohr (1969) argued that public entities innovate as a result of and the availability of resources for overcoming widely identify policy innovation as occurring when a government adopts a new policy, regardless of whether that policy is well - established elsewhere (Walker 1969, Mintrom 1997). From this perspective being respo nsive to a federal incentive often requires municipal policy that facilitate response. However, to develop some level of testable hypotheses we must define these abstract concepts more concretely and mold them a bit to address the specific nature of incentives. questions necessary for understanding what encourages ci ties to respond to an incentive: Does the city know about the incentive and incentivized policy issue? Does the city face barriers to respond to the incentive theori 23 attention . As I will discuss momentarily, a city cannot respond - cannot innovate - if they are unaware of a policy issue or incentive. Regarding the other questions, the answer to whether the city has the resources to respond stems from their fiscal and organizational capacity and much of need . The answer to the final question regardin g the barriers to innovation stems from answers to the other three questions, since barriers to response often results from a lack of attention or knowledge, a lack of capacity, and a lack of need or desire. And so we can narrow the way that we think about the predictors of response to incentives to three theoretical concepts, each of which have substantial literatures of their own: attention, capacity and need. Attention At its core, responding to incentives, like engaging in policy innovation, is policy c hange. This change can be formal, such as adopting a new statute, or it can be informal, such as offering employees a new training opportunity. It is nonetheless change and the policy process literature shows that political attention is a necessary condit ion for this change. The argument is fairly straightforward: an issue must be on the public agenda before a policymaker can act on it but this is difficult because the government is made up of people who have numerous demands on their limited attention. A s a result, no individual can focus on all important issues at a given time and because of this, a catalyst must help transition an issue from a distant community condition or a shock to (Baumgartner and Jones 1993; Jones and Baumgartner 2005; Kingdon 1984). In thinking about the responsiveness typology, attention is a necessa ry but insufficient condition for a municipality to be responsive to federal incentives. Responsive cities must be 24 aware of both the problem in their community and the federal resources and incentives available to address it. Proactive cities need only und erstand the problem. In contrast, a lack of attention can be sufficient but is not necessary to generate unresponsive or status quo responses. Certainly if local officials are inattentive or unaware of a community condition or federal incentive they cannot act on them, but a city can understand the problem and incentive and still choose not to act. Attention can be generated in ways too numerous to fully explicate, but the literature suggests that there are three sources of attention that might be particul arly important in facilitating municipal responses to federal incentives: executives, local bureaucracy and advocacy coalitions. On a variety of issues the executive can act as a policy entrepreneur, pursuing and advocating for policy change on a given is sue (Kingdon 1984; Yates 1977). At the local level, strong mayors in particular have an agenda setting capacity that allows them to focus public and legislative attention on a problem, and in many cases direct their administration to pursue particular poli cy initiatives and resources. A mayor in a mayor - Fe iock 2014, 17). Further, scholars have found that mayor - council systems are more likely to adopt economic development policies, for example, because the political nature of the position makes them more responsive to constituents and interest groups than their council - manager counterparts (Lineberry and Fowler 1967, Feiock and Clingermayer 1986). Further, mayors have interaction with higher levels of government and a political incentive to foster strong relationships with federal government units (Cammisa 1995; Conlan 2010; Haider 1974), which may encourage attention on with federal goals. 25 I discuss capacity in more detail in the next section, these two functions are interrelated. attention to existing problems and awareness of federal incentives by increasing local capacity. Larger, more specialized bureaucracies are likely to be better equipped to follow a variety of local problems, be more knowledgeable about potential solutions, and have staff who actively follow federal goals, program changes and initiatives (Fleischmann and Green 1991, Newton 1982, Rich 1989 a , Rub problems and incentives. Finally, the extent to which there is a network of non - governmental agencies working on a given issue is likely to influence government attention and responsive ness to federal incentives. Advocacy Coalition Framework suggests that policy change occurs within subsystems of intergovernmental actors, the media, and experts. Within these subsystems, coalitions of individuals with similar belief systems promote parti cular policy approaches (Sabatier 1987; Sabatier and Jenkins - Smith 1999). Having strong advocacy coalitions with many non - governmental actors is likely to improve the chances that cities both address problems on their own and respond to federal incentives by focusing municipal attention and providing resources to respond. Strong advocacy coalitions can increase attention to a policy area by promoting policy learning across the public and government agencies (Bennett and Howlett 1992; Jenkins - Smith and Sabat ier 1994), participating in media and philanthropic activities that raise awareness, and formally advocating for a response (Wallack 1993; Sabatier and Jenkins - Smith 1999). Further, diverse, well - organized coalitions are likely to have non - governmental mem bers who are familiar with federal initiatives, increasing awareness and attention to emerging 26 incentives. Importantly, the nongovernmental entities do not simply increase attention but also pplementing government resources with research and human resources (Lipsky and Smith 1989; Milward and Provan 2000; Young 2000). Capacity respond. I adopt a definitio achieve what it wants to achieve. This ability is influenced by resources and organizational pres ence of greater capacity predisposes public - sector organizations to achieve various potentialities multiple dimensions of capacity (Bowman and Kearney 1988), but for my purposes I focus on words, whether a city has the money and staff to respond to an incentive. As with attention, capacity is necessary but insufficient for a response. Witho ut adequate human/administrative or financial resources a city is unable to respond; but, on the flip side, if they either do not pay attention or do not want to act, having adequate capacity is insufficient to elicit a response. The magnitude of administr ative and fiscal resources necessary for a response depends on the nature of the incentive itself. For example, federal financial incentives that stipulate matching local funds require cities to have greater fiscal resources than grants without a matching component. Similarly, incentives that require extensive specialized knowledge or have complex legal dimensions require cities to have greater human resources than those that are less complicated. 27 In the case of fiscal resources, I refer specifically to l ocal government budgets and available funds, which may be measured in a variety of ways based on the specific demands of the incentive. For example, in one case it might be appropriate to evaluate fiscal capacity using total revenue, in another it might be more fitting to use own - source revenue, and in yet another it might be best to examine revenue - to - debt ratios. The fiscal capacity to respond to incentives is less about where the money comes from and more about whether there is enough money. This is emph asized by a variety of findings that indicate, among other things, that cities with greater revenue (greater fiscal capacity) are more likely to attract federal grants (Hall 2008), pursue economic development incentives (Green and Fleischmann 1991, Rich 19 89 a , Rubin 1989), and engage in environmental regulation (Terman and Feiock 2014). In relation to the second dimension of capacity, human or administrative resources, economic development research suggests that cities with larger bureaucracies, increased expertise, and specialized agencies that focus on a given policy area are more likely to implement economic development policies (Fleischmann and Green 1991, Rich 1989 a , Rubin significance of economic development plan suggests that cities with limited staff expertise may face barriers to adopting and implementing a broad range of de been replicated in studies of federal grant seeking (Collins and Gerber 2006, 2008; Hall 2008; Manna and Ryan 2011) and in research on how effective local implementation of federal policies has been (Terman and Feiock 2014). The consistent finding that policy innovation and change require high levels of administrative capacity show that human resources are likely to strongly influence whether a city can respond to federal incentives. 28 Need Attention and capacity dictate whether a city is able to respond to federal incentives, but there are a variety of considerations that influence whether a city wants to respond. The most important of these considerations can be lumped under the general category of municipal nee d: does the incentive respond to an existing local need? Despite its importance, meeting local needs is neither sufficient nor necessary for a response. Without attention and capacity, having local needs that could be met by an incentive is insufficient to lead to responsiveness; conversely, so long as a city has attention and capacity it is not necessary that an incentive address a local need for a city to respond. Nonetheless, need is an important predictor of responsiveness because it provides an impetus for cities to voluntarily conform to federal goals and respond to federal incentives. While cities have a myriad of needs three stand out in the literature as directing significant portions of municipal behavior: meeting political demands (for example se e Callahan 2007; Miller and Miller 1991; Yang and Callahan 2007), maintaining fiscal solvency (for example see Hendrick 2004; Peck 2012; Skidmore and Scorsone 2011), and addressing local problems. When incentives help address municipal needs, cities are more likely to respond to incentives. Local governments face political pressure to meet constituent demands while working within the constraints of the intergovernmental system. To get reelected officials must behave in ways that are consistent with cons ideological beliefs and political priorities (Einstein and Kogan 2015; Tausanovitch and Warshaw 2014) . When incentives are consistent with local ideology they have the potential to help local gov ernments address the political need of meeting constituent demands. Alignment is also important between elite ideology and federal incentives, since recent work investigating the role 29 they rely upon their beliefs when making policy decisions about local service provision (Schumaker and Kelly 2013). An excellent example is the consequences of misalignment between federal incentives and elite ideology at the state level through the Affor dable Care Act. Nineteen governors have declined federal incentives to expand Medicaid because the policy is inconsistent with their ideological beliefs (17/19 of these governors are Republicans) (Where 2015). Thus, when an issue or incentive is aligned wi th mass and elite ideology it is likely to help cities meet their political needs, thereby increasing the likelihood of responsiveness. When incentives are misaligned with local ideology they are more likely to lead to unresponsive cities. environment provides its own incentive for municipal responsiveness. Like all organizations, cities have an inherent goal of self - preservation in addition to their mandate to serve a variety of community functions. These activities take money and unfortun ately many cities face a bleak reality when it comes to their fiscal condition. Faced with numerous financial challenges ranging from cuts in state and federal aid to declining tax bases, cities have two options: attract more resources or shrink municipal government and its functions. As a result, struggling cities have an incentive to pursue intergovernmental resources and thus, may be more likely to respond to federal incentives, particularly those that are financial in nature. That said, financial wellbe ing is not the only dimension of fiscal stability. Because of their position within the federal system, cities face varying degrees of dependency on higher level governments and this dependency creates a unique need to conform to higher level government de mands to maintain stability. As many cities struggle to maintain services and fiscal solvency, there has been increasing scholarly attention on how state constraints on municipalities influence their behavior. While 30 municipal autonomy has no universal def inition, the Advisory Commission on Intergovernmental Relations (1981, 1993a, 1993b) has identified four generally accepted dimensions: allowed, local referendums all owed), functional autonomy (broad functional home rule, limited functional home rule, service agreements allowed), fiscal autonomy (debt purpose, bond life, short - term borrowing, property tax limits), and administrative autonomy (purchasing standards, meri t system, collective bargaining, retirement, employee political Across this literature the dominant argument is that what happens, or more importantly does not happen, in cities can be traced to the legal limitations they face. While cities may want to address diverse problems and develop innovative solutions, they are often unable to do so because they lack authority to carry out policy or generate the revenue required to pay for it due to constraints like taxing and expen diture limitations (TELs) (for example see Anderson 2006; Dye and McGuire 1997; Mullins and Wallin 2004; Pagano and Hoene 2010, 2012). This work suggests that state - level constraints may mediate federal incentives aimed at changing municipal behavior. Cit ies with less autonomy and particularly those that face TELs may be more likely to respond to incentives first because they are accustomed to responding to higher level government demands and second because they have a limited ability to maintain fiscal so lvency without state and federal support. Similarly, states that rely heavily on intergovernmental transfers to meet budgetary needs are likely to be more responsive to federal incentives because they have an increased need to: 1. A ttract ou tside federal r esources and 2. D emonstrate cooperation to promote future financial support. The final dimension of municipal need, responding to tangible community problems, is one that deserves more attention in the public policy literature. In general, policy theories promote the idea that the policy process is predominantly political, allowing policymakers to 31 easily ignore existing problems if they do not have the proper political sway (for example see Baumgartner and Jones 1993; Jones and Baumgartner 2005). Problem severity often only enters policy theories as a catalyzing force that focuses political attention on an issue (Kingdon 1984), with little attention to the idea that governments develop policy because problems exist. At least in part, this reflects the nat ional context in which most policy theories have been developed. At the national level, attention is sparse and groups and problems vying for that attention are both ting problems is so pervasive that Kingdon (1984) suggests problems and solutions are actually distinct streams where solutions essentially wait for the politically opportune problem to present itself. While policy may originate from separate streams at the local level as well (Henstra 2010), the local context is distinct from the national one in that elected officials are immediately confronted with the problems their cities face (Robinson and Eller 2010). Whereas national congressional leaders must deci de things like federal housing policy without any direct knowledge of the conditions in the majority of cities, local leaders have a much narrower solutions. Further, residents - for example those waiting on public housing lists or protesting the construction of low income housing in their neighborhoods - have far easier access to elected officials and more diverse, lower cost ways to have their concerns heard ( see for example Adams 2004; Arnstein 1969; Dear 1992; Garrett et al 2011; Khan 2004; Rydin and Pennington 2000). In other words, cities have an objective universe of tangible problems of which both citizens and elected officials are likely to be aware. Ci ties have a need to address the problem for the 32 likely to be met with responsiveness. As problems become more severe, local need to address them is higher and thus responsiveness should also be more likely. To reiterate then, incentives are being used as a way for the federal government to guide municipal behavior, policies, and priorities without directly intervening. There is evidence that some cities respond to t hese incentives, but it is less clear what differentiates cities that respond from those that do not. Based on the scholarship of policy innovation and change, variation in municipal responses can be explained by interactions between municipal attention, c apacity and need. In other words, whether cities know about an incentive, whether they have the money and human resources to respond to that incentive, and whether the incentive meets an underlying political, fiscal or problem - based need that would motivat e the city to want to respond. Below I explain why homelessness is the ideal case to evaluate this framework. Homelessness and Municipal Responsiveness Homelessness policy in the US provides an excellent case to study municipal responses to federal inc entives for three reasons. First, there have been significant changes in the approach that many cities take to address homelessness, transitioning from more punitive measures to more compassionate ones. 2 Second, dominant urban theories predict that cities would not behave in the way that they have. And third, there are distinct federal incentives that have encouraged some - but certainly not all - cities to alter their behavior. Each of these three points is discussed extensively in the next three chapters, s o here I discuss them in limited detail with the goal of clarifying the logic for choosing homelessness. 2 Compassionate policy seeks to eradicate homelessness using holistic practices rather than criminalizing or geographically isolating the homeless. 33 Though they come to their conclusions through different arguments, power, regime, and social choice theories all suggest that the policies a city purs ues, and thus the types of problems a city responds to, depends upon the policy area under consideration. They predict that cities will focus on development and land policy, paying little attention to social issues and especially not engaging in social ser vice provision or, by extension, compassionate homeless care (Banfield 1961; Dahl 1961; Domhoff 1978; Hunter 1953; Mills 1956; Polsby 1980; Peterson 1981; Stone 1989, 1993, 2005; Tiebout 1956). Although many cities belie these theoretical expectations by p roviding homeless services, the theories highlight the singular nature of this type of municipal behavior. The reality is outside of external incentives, cities do not have many reasons to develop compassionate policy and the fact that they do presents an interesting puzzle. There are a handful of reasons to expect cities not to behave in compassionate ways. First, the homeless are the epitome of a powerless group, lacking leadership, organization, and both financial and political resources to advocate for benefits and policy. Second, to the extent that they have a voice, it is through service organizations that may or may not be politically active, but typically have fewer resources than traditional business interests . Third, homelessness stems from compl ex issues, virtually none of which local governments have direct control or even marginal influence over. Numerous studies have identified the sources of homelessness as being both systemic and personal (for example see Anderson 1961; Snow and Anderson 199 3; Phillips 2014), ranging from economic conditions like unemployment to individual issues like mental illness. With extremely limited control over both the macro and micro issues that perpetuate homelessness, municipalities have limited incentives to legi slate on and respond to the issue. Finally, homelessness is the epitome of a redistributive good and from a social choice perspective, unless cities conceptualize addressing homelessness as a precondition 34 to development, we would expect cities to defer to private organizations and higher levels of government to address the issue. In sum, urban politics theories are very useful in conceptualizing why a city would be unresponsive - homelessness is simply not a policy area that municipal governments are likely to handle - but they provide a limited understanding of why in reality we see widespread compassionate homelessness policy. While for safety and economic reasons it may be important to isolate the homeless in a single part of the city, keeping them out of wealthier neighborhoods and central business districts, this can be done without developing a comprehensive approach to ending homelessness. In fact, for nearly two hundred years this is precisely what cities did, using existing zoning laws, police, and busing to either remove or isolate this population (Mitchell 1997; Wagner and Gilman 2012). This is important because what allows us to make the leap from not responding to responding is understanding that on their own most cities will not respond to the p roblem with compassionate care; instead, what has facilitated this type of response, at least in terms of planning, are federal incentives to do so. Although homelessness is an acutely local issue, over the last thirty years the federal government has bee n a dominant player in both creating the problem as we know it today, and in generating policy solutions. This case is made through historic evidence in the next chapter, but the extremely short explanation is that in the 1980s federal economic, mental hea lth and housing policy created a dramatic increase in homelessness that reached crisis levels. In response, the federal government developed and funded a program for cities to address the issue, taking a leadership role in identifying the best ways to go a bout doing this. Importantly, federal involvement has been channeled through local governments and service providers - the feds are not directly engaged in addressing homelessness. At the same time, they have also refrained from 35 mandating a response, instea d relying on voluntary compliance with incentives to carry out their policy goals. Cities have responded to these incentives in different ways, with some actively pursuing federal resources and making homelessness a priority and others devoting little ener gy to attracting resources and confronting the issue. For these reasons, homelessness presents a useful case for understanding both the power of federal incentives in municipal behavior and the drivers of municipal responsiveness to these incentives. In t he next chapter I outline the history of federal involvement in homelessness policy, making the case that localities diverged from punitive practices when the federal government provided incentives to do so. In Chapters 4 and 5 I explore the individual det erminants of municipal responsiveness and the predictors of substantive and symbolic policy development. In these chapters I expand upon the attention - capacity - need framework outlined above by incorporating additional insights from scholarship on federalis m and symbolic policy. 36 CHAPTER 3. A History of Intergovernmental Dynamics in Homeless Po li cy know it in the US today has developed over the last forty years. As this modern, age - old problem has evolved, public sentiments and beliefs about homelessness have shifted reflecting both changing economic conditions and altered social constructions of the homeless. Public policy has vacillated between punitive measures, effor ts to meet basic needs, and the promotion of rehabilitation and healing. What has been constant in this process is the fragmented and intergovernmental nature of responses to homelessness. As is discussed in Chapters 4 and 5, this reality continues in cont emporary municipal responses to homelessness. What follows here are important historical dimensions of homelessness and homeless policy in the US; it is intentionally brief, as historians and social scientists have developed excellent, detailed accounts o f the history of homelessness elsewhere (for example see Howard 2013; Kusmer 2003). What follows is useful context for understanding the late 20th/early 21st century developments in federal and local policies that are the subject of the remainder of the di ssertation. Indentured Servitude, Poor Houses and Outdoor Relief: The Early Years Homelessness and poverty go hand in hand, and prior to the early 20th century the United States handled these problems as its colonial ancestor Great Britain had, with limited charitable relief, indentured servitude, and eventually, the poor house (Mitchell 2011). In fact, the US should be local and minimalist, providing only what was needed for survival. This meant that in the 18th and early 19th century when an individual fell on hard times and could no longer support themselves they had three options: 1. 37 appeal to their church and family for charity; 2. put themselves up for auction as an indentured authorize ad hoc public assistance that would provide tax money to either cover basic necessities in their existing homes (outdoor relief) or place them with someone the city contracted with to care for the poor (Hansan 2011). These arrangements often left the poor in abusive or exploitative situations and became increasingly untenable as t he ranks of the destitute grew with urbanization and the industrial revolution (Wagner and Gilman 2012). Industrialization spurred many people to leave their rural roots and start new lives in rapidly growing cities. Although for many this move meant new o pportunities, much of the new employment was cyclical and not everyone found good fortune in urban America; because of this, industrialization also meant increasing concentrations of both the able - bodied poor and those made destitute by age, illness or dis ability (Fischer 2011; Monkkonen 1984). Beginning in the mid - 19th century and continuing until the mid 20th century, local governments responded to intense poverty by developing poor houses, which theoretically were a more efficient and cost - effective way to deal with those who would otherwise be homeless. Those picked up by police for begging and vagrancy or those found to be facing long term, extreme poverty and deemed unable to care for themselves were frequently sent to these tax - funded residential ins titutions where they were provided with basic amenities (Rothman 1987). At this time, destitution was seen as a personal shortcoming and a form of deviance; as such, poor houses were designed to be unpleasant places that provided an existence no one would choose unless they had no other options, while at the same time indoctrinating the able bodied into a strong work ethic. To do this, they required extremely arduous work in exchange for meager accommodations, a combination that made conditions so undesirab le that poorhouses filled with those who literally 38 had nowhere else to go: widows, orphans, the elderly, infirm, and disabled. On the whole, poor houses housed able - bodied individuals primarily during winter months when work was scarce (Katz 1986). For man y reasons, ranging from corruption to abysmal living conditions, poorhouses were deemed a failure in achie ving their primary goals of 1. M - those who could support thems elves but chose not to - and 2. P roviding a refuge for - those who were truly helpless and in need of state aid. Yet, for all of their shortcomings, poor houses were successful in doing one thing: they limited the size and thman 1987). They also kept issues of homelessness and poverty at the local level, which allowed each community to provide assistance in the way it deemed fit and helped them limit service access to members of the community. Though poor houses provided a r esponse to the geographically stable poor - those who were members of a community - they did not provide a solution for the wandering poor, a group that was often made up of Civil War veterans, former slaves, and migratory workers who were essential to the i ndustrial economy (Rothman 1987). Compounding this problem was the fact that even during the early days there was severe distrust and social pressure against those who were considered vagabonds or wanderers. Local governments often required - and many still do require - proof of residency to access services and communities made clear distinctions between neighbors and strangers (Hopper et al 1985; Mitchell 2011). This distinction necessarily identified neighbors as deserving or potentially deserving of assist ance, while strangers were refused help and often sent back on the road. Rather than recognizing homelessness as an outgrowth of the structural problems associated with capitalist economies, such as the depression 39 of the 1870s, tramps, as those who wandere d from place to place begging were called, were seen as morally deficient (Cresswell 2001; DePastino 2003, Hopper 2003). Cities enforced, often separating the worthy from the creation of some of the most iconic locales of urban homelessness - skid row in the inner city and hobo jungles on the periphery. These locations offered space where the homeless could escape the heavy hand of the law (Monkknonen 1984; Schweik 2009). By the late 1890s, yet another depression had wreaked havoc on the US economy and a new type of homelessness had emerged. The hobo, as used in common parlance, was a man who traveled throug hout the country in search of work. These men rode the rails, moving from place to place to work as short term employees in agriculture and industry. In addition to the deviant construction associated with being transient, the hobo population was seen as particularly problematic because it was full of highly politicized radicals who railed against the wage sy stem (Hopper et al 1985; Hopper 2003). As a result, although they filled an essential role in the economy, local governments found it particularly important to keep them away from mainstream society and encouraged their isolation through the development of main stems, which were relatively isolated areas of the city that catered to hobos through housing, saloons, single occupancy lodging, missions, temp agencies and other businesses. While many of these areas would eventually deteriorate into skid rows, in their heyday Hobos would spend the money they earned from short - term employment in these areas, using hobo jungles (encampments near railroads that al lowed for easy mobility) when they ran out of money and were again looking for work. Police and city officials typically ignored these places, allowing them to flourish and giving the homeless population a way to exist outside of the public eye (Hopper et al 1985; 40 Hopper 2013). It is important to reiterate that because government policies were designed to help the deserving poor, which included women, children, and the elderly, throughout most of the 20th century the homeless population consisted primarily of working - aged men, with only a (DePastino 2003; Mitchell 2011). responses to homeless men, missions and other charitable organizations emerged to fill the service gap (Mitchell 2011). The first missions began operating in the late 1890s with the goal of providing both spiritual and physical sustenance to the homeless. Locating in the heart of m ain the complex, public - private partnerships that would characterize future municipal responses to homelessness. In essence, these private organizations stepped in when government and most of society would not, creating a precedent that would have long - lasting implications: until the end of the 20th century government played an extremely limited role in serving the homeless and even today when all levels of govern ment play some role in homeless services and prevention, charitable organizations are the primary actor in service delivery. It was only with the Great Depression and its concomitant rise in homelessness that the Federal government began to alter its respo nse to homelessness and poverty by providing for human welfare for the first time. The New Deal introduced numerous policies aimed at improving the structural drivers of poverty and, by proxy, homelessness and established a norm of federal intervention tha t laid the groundwork for its current incentivizing role in municipal responses to homelessness. 41 Federal Expansion: The Great Depression, WWII and Beyond The Great Depression devastated the American population and led to widespread homelessness at rates previously unexperienced in the US (Wagner and Gilman 2012). The combination of global economic failures and the dust bowl caused many to lose their homes and businesses and encouraged increasing numbers to migrate across the country, though even with the se moves permanent housing remained elusive for many. The need for financial stability led many men to adopt a hobo lifestyle, allowing them to have the extreme mobility necessary to take advantage of limited, geographically - dispersed employment but adding to the numbers of desperate poverty and swelling homeless populations w ere more than cities and local missions could cope with; as a result they sought help from higher levels of government. Though cities looked to their states for assistance, they were met with insufficient responses and as a result they increasingly petitio ned the federal government for help. Unlike in earlier times when extreme poverty and homelessness were seen as the result of moral shortcomings, politicians and indi viduals nor local authorities had control over (Kusmer 2003; Miller 1991). This shift in the construction of the problem and target population was important because it allowed cities to successfully argue that since municipalities lacked the authority, aut onomy and influence to address the root causes of extreme poverty and had insufficient resources to provide services to the massive numbers of poor and homeless, only the federal government was capable of improving the situation. 42 The federal government res ponded with the New Deal and in it created a variety of lize the housing market, these programs focused on broad social, monetary and economic policies that addressed the structural drivers of poverty and situational homelessness (for excellent accounts of the New Deal see Brinkley 2011; Heclo et al 1985; Schl esinger 2003). This legislation did not address homelessness directly, but instead worked from the premise that improving economic conditions would allow those who were homeless through no fault of their own to pull themselves out. Put another way, while t he New Deal helped stabilize the economy, which reduced new instances of homelessness and allowed many of the newly homeless to return to housing, it did not provide resources for homeless services or homelessness prevention and it offered little for those who were chronically rather than situationally homeless, instead leaving this to localities and charitable organizations. eliminated much of the sympathy for the hom eless that had been present during the Depression. When poverty and homelessness were widespread and a structural cause was easily identifiable, many Americans stopped viewing the homeless as morally degenerate and instead saw them as victims of systemic f ailure. As those who were temporarily homeless returned to stability and the most visible causes of situational homelessness were under control, attitudes towards the homeless reverted to ones of contempt and disdain. The homeless population itself was dom inated by groups we would now call chronically homeless but were often called 43 deficient and antisocial, while homelessness was largely co nsidered either a choice or a reflection of poor decisions (Baxter and Hopper 1981; Wagner2005). Because of this, localities returned to policies of geographically isolating and criminalizing the homeless and the services, as meager as they were, were prov ided only through charitable organizations. This state of homelessness continued until the late 1970s/early 1980s when the public and both local and Federal governments were confronted with a level of homelessness that was higher than that of the Depressi on era and had its roots in economic fluctuations and Federal policy. Basic Services through Local Response and Federal Intervention: The 1980s and 1990s From the mid - 1960s to the early 1980s the Federal government made a variety of policy decisions that would have drastic consequences for both vulnerable populations and municipalities throughout the nation. The Community Mental Health Act of 1963 shifted federal resources away from state mental hospitals towards community mental health centers and began the process of de institutionalizing the mentally ill. Though fewer than half of the envisioned community centers were ever built, by the 1980s 90 percent of beds in state hospitals had been eliminated and significant portions of the mentally ill population found themselves without a place to live or the treatment they needed to function. Meanwhile, urban renewal and neighborhood gentrification eliminated substantial portions of existing market - based affordable housing without adequately replacing it or prov iding sufficient public alternatives (Marcus 2006; Mitchell 2011). By the 1980s the poor were already in a precarious position, but the Reagan Revolution and its policies to drastically reduce the social safety net and funding for subsidized housing amplif ied the problem, ensuring that demand for low income housing would far exceed 44 supply (Dreier 2004; Pierson 1995). In combination with growing income inequality, unemployment, the return of drug addicted and psychologically - damaged Vietnam Veterans, the crack cocaine epidemic, and the global economic recession of 1982 - 1983, these policies would 2012; Why 2009). consciousness in the mid - young and old, male and female, white and black. For the first time since the Great Depression, the homeless became a visible mass and homelessness became a condition marked more by a lack of housing than a lack of home (Marcus 2006; Mitchell 2011). Alth ough accurate counts of the homeless population at this time are unavailable, the Clinton Administration estimated that between 5 and 9.5 million Americans were homeless at some point during the second half of the 1980s (Interagency 1994). The magnitude an d visibility of the homeless crisis, in combination with protests and lawsuits organized by both local and national homeless organizations like the Community for Creative Non - Violence (CCNV) and the National Coalition for the Homeless 3 , brought enough atte ntion to the issue that neither the federal nor local governments could ignore the growing problem (Hombs 2001 ). At the local level, charitable and faith - based organizations took the lead in providing a nationwide response to the crisis, increasing the num ber of shelters from 1,900 in 1984 to 5,400 in 1988 and leading similar increases in food banks, soup kitchens and other services for the homeless (Couzens 1997). In contrast to the outpouring of charitable aid, local governments 3 See Hombs 2001 for an extensive chronological account of the organizations and events that addressed homelessness and pushed government to respond. 45 were far less generous; wh ile a handful of cities responded by increasing financial resources for soup kitchens and shelters, many - arguably even most - used policing, zoning, busing and local Not In My Backyard (NIMBY) protests to keep the homeless out of their communities (Ellicks on 1996; Foscarinis 1996; Link et al 1995; Lyon Callo 2001; Salinger 2006). At the federal level, it was not so easy to move the problem out of sight. The federal government and the American people saw the homeless population growing not just in isolated communities, but across the nation; because of this, public demand for federal intervention increased (Foscarinis 1991; Mitchell 2011; Wagner and Gilman 2006; Wolch and Akita 1989). In December of 1982 the US home less for the first time since the Great Depression. Listening to testimony from service providers, local officials and the homeless themselves; this hearing set the stage for elected legislation addressing the homeless, 4 the McKinney - Vento Homeless Assistance Act of 1987. 5 Congress argued that: The Nation faces an immediate and unprecedented crisis due to the lack of shelter for a record increase in homelessness, states, units of local government, and private voluntary organizations have been unable to clear responsibility and an existing capacity to fulfill a more effective and responsible role to meet the basic human needs and to engender respect for the human dignity of the homeless, (42USC11301). Although the McKinney - Vento Act was inconsistent with the neoliberal politics of the day, and in spite of the fact that President Reagan believed homelessness should be addressed at the local 4 Additional legislation has been passed to address the needs o f particular sub - populations, for example the Runaway and Homeless Youth Act aimed at protecting street youth and preventing their exploitation (About the 2014; Alliance 2015) 5 This legislation was originally called the Stewart B. McKinney Homeless Assi stance Act but was renamed in 2000 by the Clinton administration after a major supporter of the legislation, Representative Bruce Vento, died. 46 level, significant pressure from advocacy groups and many revisions to the legislation led him to reluctantly sign the bill, providing $497 million for homeless servic es and prevention, a figure that would rise to over $1 billion by the early 1990s (McKinney - Vento 2006). The original legislation and its subsequent amendments significantly altered the approach localities and service organizations have taken to address h omelessness. A brief explanation of - 77) is provided in Table 3. 1. For our purposes the most important are providing a definition of homelessness that created a clearer target group, and providing financial resources fo r homeless services. Providing for basic needs, offering case management and developing programs to address the diverse dimensions of homelessness has become known as compassionate care. By developing new programs tied to monetary resources, the Federal go vernment made it much easier for cities to provide services. These resources would eventually encourage many cities to depart from their legacy of punitive responses to the homeless and move towards a more compassionate approach with many transitioning awa y from criminalization and isolation and towards service provision. However, the full incentivizing power of financial resources would not be felt until the beginning of the 20 th century; nonetheless this legislation laid the groundwork for the radical dep arture in policy approaches that cities throughout the nation have taken in the 21 st century. 47 Tab l e 3. 1 Con t en t s o f M cK i nney - Ven t o Ho m e l es s Re li e f Ac t o f 198 7 T itl e I D e fi ne s ho m e l essnes s includes: 1. An individual who lacks a fixed, regular and adequate nighttime residence; and 2. An individual who has a primary nighttime residence that is a. A supervised publicly or privately operated shelter designed to provide temporary living accommodations (including welfare hotels, congregate shelters, and transitional housing for the mentally ill) b. An institution that provides a temporary reside nce for individuals intended to institutionalized; or c. A public or private place not designed for, or ordinarily used as, a regular sleeping accommodation for human - 77 Title 1, Sec 103) T itl e II Es t ab lis he s t h e I n t e r agenc y Counc i l o n t h e H o m e l e s s n i ndependen t en tit y w it h i n t h e Execu ti v e B r anc h co m pose d o f t h e head s o f 1 5 f ede r al agenc i es , ( M cK i nney - Ven t o 2006 , 2) . 6 Th e Counc i l ha s see n m i xe d suppo rt : Congress refused to reauthorize it i n 199 3 and as a result it was c o n s o li da t e d i n t o t h e W h it e H ou s e D o m es ti c Po li c y Counc i l. It was r ees t ab li she d as an independent entity i n 2001 unde r t h e Bu s h e ff o rt s t o co m ba t ho m e l essness . T itl e III Au t ho ri ze s t h e E m e r genc y Foo d an d She lt e r P r og r a m ad m i n i s t e r e d b y FE M A Titl e I V Au t ho ri ze s f ou r e m e r genc y she lt e r an d tr ans iti ona l hous i n g p r og r a m s ad m i n ist e r e d b y HU D : E m e r genc y She lt e r G r an t Prog r a m , Suppo rt i v e H ous i n g D e m ons tr a ti o n P r ogra m , Supp l e m en t a l Ass i s t anc e f o r Fac iliti e s t o Ass i s t t h e Ho m e l ess , an d Sec ti o n 8 S i ng l e Roo m O ccupa nc y M ode r a t e Rehab ilit a ti on . I n 201 0 t he s e p r og r a m s w e r e con s o li da t e d i n t o a s i ng l e co m pe titi v e g r an t p r og r a m ca ll e d t h e Con ti nuu m o f Ca r e P r og r a m . 6 Toda y t h e f o ll ow i n g 1 9 a g en ci e s/ a g en c y head s a r e i n v o lv e d i n t h e Coun cil : U . S . Depa rt m en t s o f A gric u lt u r e , Com m e rc e , De f en s e , E du c a ti on , E ne r g y , Hea lt h an d Huma n S e rvic e s , Ho m e l an d S e c u rity , Hou si n g a n d U r b a n D ev e l op m en t , I n t e ri o r , J u stic e , L abo r , T r an s po rt a ti on , Ve t e r an s A ff a irs , Co r po r a ti o n f o r Na ti ona l an d Com m un it y S e rvic e , Gene r a l S e rvic e s Ad m i n istr a ti on , O ffi c e o f M an a ge m en t an d Bud g e t , S o ci a l S e c u rit y Ad m i n istr a ti on , Un it e d St a t e s Po st a l S e r v ic e , an d t h e W h it e Hou s e O ffic e o f Fa it h - b as e d an d Com m un it y I n iti a tiv e s ( Abou t USICH 2014 ). 48 Tab l e 3. 1 T itl e V Requ ir e s f ede r a l agenc i e s t o g i v e s t a t es , l oca liti e s an d se r v i c e p r ov i de r s acce s s t o s u r p l u s, f ede r a ll y - o w ne d p r ope rt y t ha t ca n b e use d t o he l p t h e ho m e l ess . T itl e V I Au t ho ri ze s f ou r D epa rt m en t o f Hea lt h an d Hu m a n Se r v i ce s p r og r a m s t ha t p r ov i d e hea lt h ca r e an d m en t a l hea lt h se r v i ce s t o t h e ho m e l ess : Co mm un it y M en t a l H ea lt h Se r v i ce s B l oc k G r an t , H ea lt h Ca r e f o r t he H o m e l ess , t w o de m ons tr a ti o n p r o j ec t s f o r m en t a l hea lt h an d subs t anc e abuse . A 199 0 a m end m en t c r ea t e d t h e She lt e r P l u s Ca r e P r og r a m , des i gne d t o p r ov i d e hou si n g f o r t ho s e w it h add i c ti on , A I D S , m en t a l ill nes s and d i sab i li t i es . T itl e V II Au t ho ri ze s educa ti on , j o b tr a i n i n g an d se r v i c e g r an t p r og r a m s: Ad m i n i s t e r e d b y t h e D epa rt m en t o f Educa ti on , t h e D epa rt m en t o f Labo r an d t h e D epa rt m en t o f H ea lt h an d H u m a n Se r v i ce s r espec ti ve l y : Adu l t Educa ti o n f o r t h e Ho m e l ess , Educa ti o n f o r Ho m e l es s Ch il d r e n an d Y ou t h Prog r a m , J o b T r a i n i n g f o r t h e Ho m e l es s De m ons tr a ti o n Prog r a m , E m e r genc y Co mm un it y Se r v i ce s Ho m e l es s G r an t Prog r a m . Th e J o b T r a i n i n g p r og r a m w a s de f unde d i n 199 5 an d t h e Adu l t Educa ti o n f o r t h e H o m e l es s Prog r a m an d E m ergenc y Co mm un it y Se r v i ce s Ho m e l es s G r an t Prog r am w e r e de f unde d i n 199 8 ( M cK i nne y Ven t o 2006 ) . T itl e V II I G i ve s t h e ho m e l es s acces s t o f oo d s t a m p s b y a m end i n g t h e f oo d s t a m p p r og r a m an d expand i n g t h e D epa rt m en t o f Ag ri cu lt u r e ad m i n i s t e r ed Te m po r a r y E m e r genc y Foo d Ass i s t anc e Prog r a m . T itl e I X Ex t end s t h e Ve t e r an s J o b T r a i n i n g Act . In particular, the legislation created a pool of fiscal resources to address homelessness that were to be predominantly allocated through competitive grants. To this day 90 percent of available funds are a llocated competitively, making it possible for municipalities to pursue resources if they want to provide services to the homeless, but in no way requiring them to do so. Importantly, in the first years of the program the same pool of resources was made av ailable to states, localities and non - profit organizations, meaning that when the legislation was initially implemented there was a complex web of organizations that were often competing for the same 49 resources to help the same people without any requiremen t of inter agency communication. In combination with the fact that the programs developed under the McKinney - Vento Act were housed in 15 different federal agencies, the stage was set for frustration and eventually reform. In the mid - 1990s this reform would come out of the Clinton Administration, which sought to increase interagency cooperation at both the local and federal levels by creating a more structured and strategic service network. More specifically, since 1995 HUD has required communities to submit a single application for funding through an entity called a Continuum of Care (CoC). The CoC is several things. From an organizational perspective it is a single entity tasked with coordinating services and service providers in a HUD - designated area and can be a lo cal unit of government, a quasi - governmental entity like a housing authority, an established nonprofit like United Way, or a nonprofit developed specifically to be the CoC (it cannot be a for profit entity). From a service perspective, the CoC is a network of organizations that ideally provides the full range of services to homeless individuals including emergency shelter, transitional and permanent housing, supportive services, and outreach, assessment and prevention. From a funding perspective, the CoC is for funds, prioritizes requests based on the needs of the entire CoC area, submits a single application for CoC programs , and coordinates the dispersal of received funds. By transitioning to a CoC structure, the federal government enforced coordination, cooperation and networking, which now extends to the tracking of homeless individuals and their service use by requiring o rganizations to use a single CoC - selected Homeless Management Information System (HMIS) (Hambrick and Rog 2000; Sparks 2012). 50 While the CoC system made it easier for local communities to provide compassionate services like shelters, day rooms and work - trai ning programs, the 1990s saw decreasing public support for this type of response. Scholars describe the 1990s as a time of compassion fatigue in frustration over thei r continued presence (Foscarinis 1996; Link et al 1995; Salinger 2006). Two things amplified this reality. First, the US economy was better than ever before, with record low unemployment and growing wealth; this economic reality made the structural drivers of homelessness less apparent and again directed public sentiment to a notion of the homeless as lazy, immoral and full of personal shortcomings; in other words, the homeless largely returned to the position of the undeserving poor in the eyes of the publ ic (Wagner and Gilman 2006). Second, compassion for the homeless generally transitioned into compassion for homeless families as the face of homelessness again changed. While individual homelessness continued to be rampant, individuals were joined by incre asing numbers of families with children who encouraged a shift in the way many people thought about homelessness. Increasingly government and service providers recognized that there was a distinction between chronically homeless 7 individuals who were often mentally ill or drug addicted and the families who found themselves temporarily and situationally homeless. These families were not viewed as underserving in the same way that individuals were and many local organizations found support to develop family - s pecific services like emergency family housing, emergency child care, 7 an unaccompanied homeless individual with a disabling condition who has been continuously homeless for a year or more, or (2) an unaccompanied individual with a disabling condition who has had at least four episodes of homelessness 51 compassion and new services, communities increasingly viewed homeless individuals as problematic a nd used existing vagrancy and zoning laws as well as new legislation to criminalize everything from bathing in public to camping within city limits (Amster 2003; Illegal 2004; Mitchell 1997). Broken windows policing in major US cities amplified the effects of these types of laws. As a criminal justice theory, broken windows theory puts a premium on punishing small, nuisance infractions with the idea that doing so will reduce more serious crime. As a relatively powerless group that consistently breaks minor laws, the homeless created an ideal population on which to carry out these principles. Unfortunately, this group is also less able to pay tickets and fines and to show up for court dates, pushing many homeless into a cycle of crime, punishment and recidivi sm that is not easily escaped (Amster 2003; Foscarinis, et. al. 1999; Hodulik 2001). By making it exceedingly difficult to be homeless, many have argued these policies and policing techniques have made the very state of homelessness illegal in US cities. In addition to pushing the homeless into the criminal justice system, many communities - especially suburbs and smaller cities with limited resources - bussed their homeless out of the area, literally providing a one way ticket to anywhere - but - here. This practice continues today and highlights the inherent tensions that are created when limited resources meet substantial needs and when the imperative to provide compassionate care faces local desires to keep the proble m out of the community. In 2009 New York City Mayor Bloomberg famously defended get to the . 52 Despite widespread return to criminalization and isolation, during the late 1980s and early 1990s, some local governments attracted new resources and supported new services; however, their approaches tended to be narrow in both vision and influence. In large part this was because both the federal programs designed to address homelessness and the charitable organizations working with the homeless on the ground focused on ameliorating immediate needs - giving food to the hungry, clothes to the cold, a bed for the night to those who lacked one - rather than addressing the causes of homelessness. It would take years for there to be a comprehensive approach with significant municipal buy - in. Homeless advocacy organizations played an important role in this trans ition, speaking out against criminalization and geographic isolation, devoting significant resources to studying the costs of illegality, and developing alternative programs that were effective at reducing rather than hiding homelessness. In combination wi th federal leadership and incentive creation, these advocacy efforts paved the way for ambitious plans to end homelessness, comprehensive care, and population - specific programing. anning led many cities to adopt broad - based plans to end homelessness. The transition from basic service provision and/or criminalizing and isolating the homeless was not automatic but was the consequence of federal incentives. Cities throughout the natio n have adopted these federally - motivated changes but not all municipalities have been willing to engage in comprehensive planning to carry out federal goals. Chapter 4 investigates this variation, examining the characteristics that predict whether cities h the 10 year plan to end homelessness. After making the case that the federal government has directed local homelessness policy through incentives, empirical analyses demonstrate that 53 variation in plan a doption is driven by differences in the benefits cities derive from cooperating with the federal goals. 54 CHAPTER 4: Explaining the Adoption of 10 Year Plans to End Homelessness Chapter 3 discussed the key dimensions of US policy that have made the last d ecade and a half of homelessness intervention possible, namely the introduction of the McKinney - Vento Act and the financial resources it made available. It also alluded to the fact that the federal government has been responsible for a fairly radical shift in the way that cities and nonprofits approach homelessness. The chapter that follows makes the case that the federal government has pushed and incentivized cities to develop comprehensive planning and compassionate care. By doing so, the federal governme nt has engaged municipalities in cooperative federalism. The introduced financial incent ives to promote adoption. Despite the fact that over 300 municipalities have adopted these plans, many cities have not and by doing so have chosen not to engage in a cooperative relationship with the federal government. After briefly discussing 10 - Year Pla ns, I use a novel dataset of over 300 cities to show the importance of federal incentives in promoting plan adoption. I then explore why some cities have developed 10 year plans while others have not, arguing that what has driven variation is the extent to which individual cities derive fiscal, political and problem - based benefits from cooperation. Planning and Prevention through Federal Incentives: 10 Year Plans to End Homelessness During the 1990s the federal government commissioned several reports aime d at Leshner 1992; US 1994), but large - scale, comprehensive planning did not begin in earnest until the early 2000s. In 2000, after years of research into best practices, the National Alliance to End H omelessness introduced a framework 55 ( National 2000). It called for accurate data on the homeless, the acknowledgement of different homeless subpopulations and their di stinct needs - particularly the chronically homeless - the development of better prevention services and more accountable poverty programs, the provision of additional low income and transitional housing, and the development of infrastructure to better preve nt crisis poverty. The plan was revolutionary in that it challenged the existing service framework and its tendency to manage rather than eliminate homelessness, but it needed broad buy - in to be effective. It would get this from the national government, wh ich saw the ambitious (Martinez 2002). At the end of 2001 the Bush Administration embraced the 10 Year Plan and its focus on comprehensive planning, committing the nation to end chronic homelessness in 10 years ( United 2010). As part of this effort the administration reinvigorated the United States Interagency Council on Homelessness (USICH), which had been developed under the McKinney - Vento Act but had largely become defunct, and appointed Philip Mangano to head the council. At the same time, Bush appointed heads of the Department of Housing and Urban Development who saw homelessness as a pressing issue, first Mel Martinez and later Alphonso Jackson. In a 2002 a ddress to the National Alliance to End Homelessness Martinez outlined the federal A year ago, I talked to you about the Bush Administration's commitment to ser iously committing this Administration and this nation to meeting the challenge of homelessne ss, Homeless Assistance Act became law... A decade and a half later, despite the substantial 56 investment of resources called for by McKinney - Vento, homelessness remains. The d ollars Washington has devoted to solutions have made us more effective in managing homelessness, but no more effective in preventing individuals from becoming homelessness. The ass umption was that intervention was more important than prevention. So even as federal programs ended homelessness for thousands of men, women, and families, thousands refilled the empty beds. Until now, government was misdiagnosing the condition and prescri bed inadequate medicine. This Administration is taking a holistic approach to homelessness. We are looking at the entire picture - prevention and intervention - not just the individual pieces, (Martinez 2002). Though it had identified and committed the na tion to a holistic plan for ending homelessness, the Bush Administration understood that the federal government could not carry out or drive the plan on the ground; for that, it would need state and local buy - in. In 2002 the USICH challenged the 100 larges t cities to develop their own 10 year Plans to End Homelessness and in 2003 it encouraged all cities, not just the 100 largest, to develop similar plans (Fact 2010). The Bush - based incentive cities had to adopt a holistic, compassionate response to homelessness. Although plan development was and continues to be re financial incentives that would be tied to plan development. These policy incentives were an important catalyst for early adopters, who pointed to federal goals as key reasons for development in press releases, media coverage and the 10 - Year plans thems elves. For example, when New York City Commissioner of Homeless Services Lisa Gibbs announced plan development in November of 2003, the New York Times wrote that chronic homelessnes s in the city comes from Washington, where President Bush's Interagency Council on Homelessness has the goal of ending the problem nationwide in a decade. The council has been encouraging localities throughout the country to create such plans, and many oth er cities have started similar planning processes, (Kaufmann 2003, 3). 57 In other places, municipalities appear to have misunderstood the policy incentive and the voluntary nature of plan development, believing instead that the federal government had mandat ed compliance. For example in December of 2003 the El Paso Times reported that - year strategy to end long - term homelessness to meet new funding requirements by the U.S. Departme requiring 8 cities to develop plans to end chronic Pasadena Star News reported that The U.S. Department of Housing and Urban Development has mandated that every continuum of homeless care providers in the country develop a 10 - year strategic plan to end chronic homelessness - 1). Whether responding to stated federal priorities or believing that fundin g was linked to plan development, when they were not monetary in nature. In fact, 10 percent of cities that would eventually adopt a plan did so between 2002 and 2005, responding simply to encouragement from the federal government. Despite the power of the policy incentive for some cities, most did not immediately feder al government created a financial incentive by linking plan creation to CoC competitive grants that the nation began to see widespread plan adoption. consider stated that an application will receive a higher score if it demonstrates, among - homelessness. In th e detailed presentation of the various criteria, the NOFA explained - year strategy 8 Inflection in quotations is added. 58 - year plans in the c By doing this, HUD ostensibly maintained municipal and organizational autonomy by keeping 10 - year plan creation voluntary, but made non - adoption potentially costly for communities. This approach proved to be fairly effective, with over 50 percent of cities that have adopted a plan cities have developed 10 Year Plans to End Homelessness (Ci ty 2011), about 90 percent of them adopting plans after HUD linked plan development to CoC resources. Figure 4. 1 depicts this trend, showing plan adoption over time. Among the cities that adopted a plan, the majority did so between 2006 and 2008, suggestin was a major catalyst in plan diffusion . 0 10 20 30 40 50 60 70 80 90 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 NUMBER OF CITIES ADOPTING 10 YEAR PLANS YEAR Figure 4.1. Municipal Adoption of Plans to End Homelessness New Municipal Plans to End Homelessness 59 To confirm the anecdotal evidence that federal incentives have driven 10 - year plan adoption I conduct an event history analysis on a sample of over 300 c ontinuum of care cities. Specifically, I use a Cox Proportional Hazards model with robust standard errors clustered by time. 9 I use the same sample and data for both of the analyses in this chapter, but for parsimony I save a full discussion of the sample and variables for the next section, providing only an overview here. Since the underlying question in this analysis is the factors driving the timing of 10 - year plan adoption, I use a dependent variable that measures when cities adopt a plan. Cities are c oded 1 in the year they adopt and 0 in all other years; these data come from the US Interagency Council on Homelessness ( United 2011), the National Alliance to End Homelessness ( National 2010 b ), and independent internet searches. The primary independent va riable is the introduction of federal incentives to adopt a 10 - Year Plan and is coded using a time - based dummy variable. The Federal government introduced either verbal or resource incentives to adopt a plan in 2001, 2003, 2004, 2006 and 2009 10 , these years are coded 1 while the remainder of years - 2000, 2002, 2005, 2007, 2008, 2010, 2011, 2012, and 2013 are coded 0. I control for eight municipal characteristics (their measurement and data sources are discussed in greater detail in the next section). First, I control for basic demographic information including population, population density, the percent of non - white residents, and the percent of people living in poverty. Second, I control for problem severity including both the size of the homeless population and the percent change in the size of the homeless population from the previous year. Third, I control for municipal involvement in homelessness, housing, and community development, including whether the city was the primary applicant for CoC grant 9 By using robust standard errors, I account for serial correlation in time that results from temporal dependence and time - varying covariates (Buckley and Westerland 2004; Rogers 1993). 10 For a full explanation of these incentives please refer to Chapter 3. 60 funds and the percent of total spending used on housing and community development. Finally, I control for policy learning/mimicking by accounting for the number of cities that adopted a 10 - year plan in the previous year. Results from the event history model mir ror the anecdotal evidence that federal incentives have pushed cities to adopt 10 - Year plans. Table 4.1 highlights these results, which show that in years when the federal government introduced either a policy or fiscal incentive, cities were 370 percent more likely to develop a plan. To see what this looks like graphically, Figure 4. 2 depicts the hazard rate - or the likelihood that a city adopts a plan over time when all covariates a re held at their means. Figure 4.2 shows the importance of the federal go vernment in promoting plan adoption: holding all else constant, if the federal government had introduced an incentive in a given year, the likelihood that a municipality would have adopted a plan ranges from 45 and 80 percent. In contrast, if the federal g overnment did not introduce an incentive in those same years, the likelihood of adoption falls to below 10 percent. This provides strong evidence that federal incentives have driven policy development at the local level. In addition to federal incentives, the event history analysis shows the importance of problem severity; however, this finding is more readily understood in the next analysis and so I withhold treatment of the topic until the next section. 61 Table 4. 1. Cox Model Predicting when Cities Ad opt 10 - Year Plans to End Homelessness Observations= 3421 Time at Risk= 3442 Subjects= 296 Adoptions= 189 Log Pseudolikelihood= - 1049.7 Prob>Chi2= 0.00 Covariates Hazard Ratios Robust SE Federal Incentives 3.70* 2. 31 Total Homeless 1.00** 0.00 Percent Change Homeless 0.90 0.16 City is Consolidated Organization 1.10 0.13 Housing & Community Dev. as percent of Expenditures 0.05 0.08 Cities Adopting in Previous Year 0.99 0.00 Total Population 1.00 0.00 Percen t White 0.98 0.34 Percent Living in Poverty 2.65 1.23 **sig>=.01 *sig>=.05, two - tailed test 62 Despite the role that the financial incentive had in spurring cities to adopt 10 - year plans, the 300+ cities that have developed a plan are only a fraction of the possible adopters. There are over 19,000 municipalities in the United States and nearly (Number 2015) 1,000 of them have populations over 25,000 (Area 2010). This means that there are ma ny cities, not just tiny towns, that have chosen not to develop long - term plans to end homelessness; as a consequence, cities like Kansas City, Yonkers, NY and Charleston, SC have not responded to federal incentives. Further, the cities that have adopted plans are diverse. So much so that even though homele 15,000 people have developed plans. Likewise, although social service provision is largely thought of as a liberal prerogative, liberal and conservative communities a like have responded to the federal government and developed a 10 - Year Plan. This diversity suggests that there is no simple answer to the question of why some cities have adopted plans and others have not. Yet, answering this question is important in unde rstanding not just the types of communities that are actively working on homelessness issues, but also in understanding the ways that cities respond to federal incentives and the characteristics that influence whether the Federal government can be a change - maker in the local context. Fortunately, important insights can be gleaned from the federalism and urban politics literatures, which together suggest that incentives reflect a form of cooperative federalism that requires both parties to benefit. Below, I argue that variation in adoption reflects differences in the benefits that cities derive from cooperation. Why Adopt? Insights from Fiscal Federalism and Urban Politics Chapter 2 set up a general framework for investigating federal - local incentives; th e following section provides additional nuance by applying insights from cooperative federalism to 63 10 - year plan incentives, suggesting that whether individual cities respond to incentives reflects variation in the benefits they derive from cooperation. In cooperative federalism both parties - in this case the federal government and municipalities - come to the table with the understanding that they will engage in a mutually beneficial act - be it a program, grant or policy (Elazar 1964, 1987). In these relati onships the local government retains a degree of autonomy to dictate the terms of the arrangement, the federal government is free to pursue its ends, and both parties are dependent upon one another for the desired outcome (Davis 1967; Wong and Peterson 198 6). In the case of 10 - year plans, cities retain control over the planning process and plan contents; and, the federal government benefits by being able to pursue its goal of significantly reducing the US homeless population, which it cannot do without loc al cooperation. In contrast to this clear federal benefit, what local governments get is less obvious than it may initially appear. In the previous section we saw that when the federal government linked plan development to competitive grant evaluation citi es began develo ping plans en mass e , suggesting that the financial advantage was cities primary motivation. While this is most certainly part of the benefit, the fiscal federalism literature a 10 year plan was this straightforward, many more - perhaps even most - cities would respond to the incentive and develop a 10 year plan. This has not happened. To understand this argument more clearly, let us briefly digress and talk about the implicati ons of fiscal federalism. There is a deeply held belief that cities chase money wherever they can get it, be that from business, economic development or, as is the case in this dissertation, grants in aid. In fact there is an old adage with an uncertain or igin that goes something like this: if the federal government was giving out money to be infected with 64 tuberculosis, cities would line up in droves. This belief is bolstered by a couple of hard truths and a few theories of bureaucratic behavior. First, cit ies are highly dependent on intergovernmental transfers for their financial security; the average US city receives nearly 40 percent of its revenue from state and federal grants (Martell and Greewade 2012). Second, many cities face substantial state - impose d limitations on their ability to tax and generate own - source revenue. In hard economic times like those of the last half decade, these limitations - particularly in the face of things like declining property values - have left cities vulnerable to financial crisis and insolvency, making them more dependent on fees and state and federal resources ( Brunori et al. 2008 ; Mullins and Wallin 2004 ; P agano and Hoene 2010 ). And third, state and federal transfers to cities have declined and become more r estrictive over time even as the need for them has increased (Martell and Greewade 2012). These facts set up an environment where cities need to pursue money where it is available. When we add in theories of bureaucratic behavior the formula is set for de sperate - even shameless - money chasing. Bureaucratic theory suggests that bureaucrats have an incentive to grow their organizations and entrench them in the government system (Dunleavy 2014; Etzioni - Halevy 2013). To do this they will maximize budgets and pursue resources whether they need them or not, leading to inefficiencies and more importantly to administrative units focused on bringing in the green (Niskanen 1974). Further, the flypaper effect described in the public finance literature suggests that g bureaucratic tendencies, when a city receives additional state and federal grants it increases r esult from a commensurate increase in the tax base. In other words, these grants do not necessarily free up resources to be used in other places and do not necessarily encourage 65 substitution away from tax revenue; instead they just grow the bureaucracy/pol icy area (Courant et al 1978; Hamilton 1983). Combining these ideas with the aforementioned facts leads to the belief that cities are primarily motivated by financial resources; because of this, when funding changes, cities pursue new resources or make cha nges necessary to attract resources. Although there has been recognition that cities have varying capacity to pursue resources (Porter et al 1973; Rich 1989 b ; Stein 1979), the idea of the money - chasing city is alive. Yet, if money was the sole or even the primary motivating benefit associated with developing a 10 - year plan we would expect most cities to develop them, particularly because doing so is a fairly low - cost endeavor. Earlier in the dissertation I laid out three basic conceptual prerequisites to responsiveness: attention, capacity and need. Th e corresponding questions are: D oes a city know about the is sue to which it must respond?; D oes it have the res ources to be able to respond?; A nd, does it need to respond? These concepts can be used to under stand adoption, but the question is which dimension best explains variation in 10 - year plan adoption. I suggest the answer is that differences in the degree to which plan development addresses local needs drives variation. Cities must receive a benefit to engage in cooperative behavior and meeting municipal needs is a clear way to provide that benefit. Although attention and capacity are still necessary to respond to an incentive, there are several reasons to expect them to be less important in the decisio n to develop a plan than they are for something like carrying out substantive changes. In terms of attention, the incentives to develop a 10 year plan to end homelessness have largely come from the Department of Housing and Community Development, a federa l agency that the vast majority of cities have ongoing relationships with. At a minimum, all cities over 50,000 (and many smaller communities as well) 66 receive Community Development Block Grants, a vital resource for housing and community development. HUD i s also involved in most types of low income housing, dimensions of economic development and environmental restoration, mortgage lending, and fair housing enforcement. Further, HUD has a regional office in every state and many major cities. In other words, regardless of the size of the community, cities have consistent interaction with HUD and are likely to be familiar with funding opportunities, changes in regulations and new initiatives. Thus, in this case, being aware of the incentive and having attention focused on HUD is less likely to be a barrier to cooperation than other federal policy initiatives. In terms of capacity, plan development is a relatively low - resource activity. In general, cities convene planning boards that include representatives from local government, service organizations, local businesses and lay community members. The group then engages in some level of research about best practices and community need and from there develops a plan that is brought before local government for approv al. While time consuming, the planning process requires limited financial resources and spreads the burden of human resources throughout the community; as a result, it is unlikely that a community that wanted to develop a plan would be unable to do so beca use of a lack of resources. In contrast to attention and capacity, the degree to which municipalities get their needs met by creating a 10 - year plan varies considerably by community and can help explain the benefit they receive from the cooperative relat ionship. As described in Chapter 2, I focus on three dimensions of need that are particularly pressing for municipalities: meeting constituent demands, maintaining fiscal solvency, and addressing local problems. It is my expectation that cities are most li kely to respond to federal incentives and develop a 10 - year plan when the benefit of this cooperative action would mean that one of these needs is met. 67 Constituent Demand Cities are likely to derive benefits from 10 - year plan development when doing so w ould meet constituent demands. This is likely to happen when there is general support for redistributive policy and community demand for municipal social service provision. Urban public choice theory suggests that cities have a limited interest in develop ing redistributive policies because their primary focus on economic growth and development leads them to focus on policies that attract and retain business and wealthier citizens (Peterson 1981). To the extent that some cities do engage in more extensive s ocial service provision, regime theory suggests that this is because of the makeup of the governing regime and population, which tends to be more liberal and socially minded (Stone 1993, 2005). Together these theories suggest that when we see cities engagi ng in social policy it is because there is a liberal population with a high tolerance for taxation and redistributive policies. This means that cities are likely to benefit politically from developing a 10 - year plan when 1. T he population is politically li beral and/or 2. T here is a community norm of high social service provision. Thus my first and second hypotheses are that H 1 : Cities with more liberal populations will be more likely to develop 10 - year plans. H 2 : Cities with higher spending on social serv ices will be more likely to develop a 10 - year plan. Fiscal Solvency discussed earlier, doing so can prove challenging. Cities rely heavily on state and federal 68 government s for both their financial resources and legal legitimacy and this dependent role makes them more vulnerable to changes in both available resources and in state and federal goals. Although in the face of declining intergovernmental resources we have seen many cities innovate and rely more heavily on local taxes, fees and special districts, this has been inadequate in some places and impossible in others because of state - imposed restrictions on local fiscal autonomy (Brunori et al 2008; Forester and Spindle r 1990; Martell and Teske 2007; Mullins and Wallin 2004; Springer et al 2009). As a result, we have also seen budget and service cuts, layoffs, threats of and even actual bankruptcy, and requests for bailouts (Peck 2012; Peck 2014; Pew 2012). These actions reveal the degree to which some cities need state and federal resources and the ease with which cuts in aid can undermine local stability. Because local governments are acutely aware of both their need to maintain fiscal stability and the role that state and federal aid plays in doing so, I expect cities to recognize the fiscal benefits of cooperation, including both the direct benefit of increasing the likelihood of federal aid and the indirect benefit of promoting federal goals in ways that may lead to f avor and/or future resources. Though all cities may recognize these benefits, not all cities have the same level of need, as some are more stable and independent than others; thus, I expect cities that are more dependent on intergovernmental aid to be mor e likely to cooperate with the federal government by adopting 10 - year plans. Similarly, I expect that cities with less autonomy to tax and spend, carry out desired policies, and pursue local priorities, will be more likely to comply with federal requests f or two reasons. First because they are more accustomed to doing what higher levels of government want and second because they are, by design, more fiscally dependent upon state and federal governments. This leads to my third and fourth hypotheses. 69 H 3 : Cit ies with higher dependence on federal aid will be more likely to develop 10 - year plans. H 4 : Cities with less autonomy will be more likely to develop a 10 - year plan. Community Problems ocal problems. In recent years localities have innovated on issues ranging from climate change mitigation to smoking cessation (for example see Krause 2011; Shipan and Volden 2012), revealing the complexity of issues on the local agenda. While some of thes e efforts have occurred in cities that have limited issues, many cities act when problems change dramatically or reach a critical threshold (for example see Hopkins 2010, Sharp et al 2010). In other words, as in all levels of government, cities have limite d time and energy and thus are more likely to act in situations where there is a tangible problem. In cities with substantial homeless populations, cooperating with the federal government to develop a 10 year plan has the likely benefit of addressing a com munity need: responding to the homelessness problem. In communities where the homeless population is limited, cities may feel less need to confront the problem and thus see a more a limited benefit in complying with federal requests. As a result, I expect cities with larger homeless populations to be more likely to cooperate with the federal government to develop 10 - year plans. Further, since it is estimated that up to 62 percent of the US population has no emergency savings (Holland 2015) and up to 1/3 of US households are one paycheck away from homelessness (Smith 2011), cities with larger low income populations are likely to recognize the possibility of future homelessness as a significant issue, be more likely to recognize the benefit of addressing this issue through cooperation and thus be more likely to develop a 10 - year plan. This leads to my final two hypotheses. 70 H 5 : Cities with larger homeless populations will be more likely to develop 10 - year plans. H 6 : Cities with larger populations of low incom e individuals will be more likely to develop a 10 - year plan. Evaluating the Evidence: Data and Methods To evaluate these hypotheses I use logistic regression with robust standard errors clustered by state. In this case, the robust standard errors help t o account for correlated errors that result from cities that function in the same state environment with the same/similar degrees of dependence, restriction, taxation and aid. Sample Both the analysis that follows and the previous event history analysis use the same (CoC). I use CoC cities for two reasons. First, from a pragmatic perspective, systematic homelessness data are gathered and federal resources are all ocated based on HUD - designated Continuums of Care so it is very difficult to evaluate the influences of things like the homeless population in non - CoC communities because the data are simply not available. Second, from a theoretical perspective, CoC cities are connected to the national homelessness network, which increases both the likelihood that they are aware of federal incentives and the possible benefits they would receive from responding favorably to these incentives. Given this reality, the fact that many CoC cities have not developed a 10 - year plan provides a particularly interesting case for understanding the complex determinants of municipal responsiveness to federal incentives. 71 CoCs have several geographic scopes; those included in the sample hav e a central city/town that can be identified. CoCs are often structured around a central city, for example Denver CoC or Baltimore CoC. Some CoCs include a county where the central city has been identified, for example Battle Creek/Calhoun County CoC, in these cases the identified city is included (i.e. Battle Creek). Other CoCs include a county where a primary community has not been identified but where one exists, for example Montgomery County, MD CoC where Rockville is the central community in the coun ty. In these cases, I use the largest community in the county (frequently the county seat), since homeless populations are more concentrated in larger communities and smaller communities have a long history of sending their homeless to larger ones (Paisner 1994; Rahimian, Wolch and Koegel 1992; Simon 1991). Rural CoCs in which no primary community can be identified have been excluded 11 . Appendix A lists all included CoCs and the primary city used in the sample. This sample provides substantial diversity on key measures that have been shown to be important in the fiscal federalism and urb an politics literatures. Table 4.2 pr ovides descriptive statistics for some key characteristics. 11 There is increasing recognition that rural homelessness is a serious problem. In 2009 HUD changed its funding structure to allow rural communities to access homelessne ss resources; however strictly rural CoCs are not included because of the difficulty associated with disaggregating data on rural homelessness into individual community statistics. 72 Table 4. 2. Descriptive Statistics for Sample Cities Mean Minimum Maximum Total Population 236,245 14,381 8,384,734 Population Per Square Mile 3,916 12 27,012 Size of Homeless Population 1,726 7 101,815 % Non - White 37% 4.3% 99% % Poverty 29% 3% 52% Median Household Income $40,149 $19,278 $118,363 Average Revenue Per Capita (1996 - 2006)* $2,760 $262 $15,600 Averag e % of Revenue from Federal IGR (1996 - 2006)* ** 4.40% 0.04% 29% Average Spending on Housing and Community Development Per Capita (1996 - 2006)* $101.90 $0.01 $1,382 *In 2012 Dollars **IGR: Intergovernmental Revenue Specifying the Mode l To assess variation in the adoption of 10 year plans I use a binary variable that accounts for whether a city has adopted a plan at any point from 2000 - 2014. This variable is coded 1 if a city has adopted a 10 year plan and 0 if it has not. Though I exp ect political, fiscal, and problem based need to drive variation in adoption, I include measures of attention and capacity as well. Since logistic regression cannot account for time - varying data, in cases where data vary over time, I average the data to pr ovide a single estimate. A summary of included variables ar e included in Table 4. 3 . 73 Table 4. 3. Summary of Included Variables Concept Variable Source Measurement Response to Federal Incentive Municipal 1 0 - Year Plan Adoption National Alliance to End Homelessness; Interagency Council on Homelessness; Independent Internet Search Dummy (1=Yes) Bureaucratic Attention City is Member of CoC HUD's Provider Portal Dummy (1=Yes) Bureaucratic Attention City i s Consolidated Applicant HUD's Provider Portal Dummy (1=Yes) Advocacy Attention Number of Organizations involved in a city's CoC HUD's Provider Portal Continuous Executive Attention Mayor - Council Form of Government ICMA and Independent Internet Search es Dummy (1=Mayor ) Capacity: Human Resources Number of Municipal Employees Census of Governments State and Local Employment Survey Continuous Capacity: Human Resources Number of Housing and Community Development Employees Census of Governments State and Local Employment Survey Continuous Capacity: Fiscal Resources Revenue Per Capita State and Local Finance Survey Continuous Political Need: Incentive's Consistency with Local Priorities Housing and Community Development Spending a s Percent of Total Spending State and Local Finance Survey Continuous Political Need: Local Ideology City's Democratic Vote Share in Presidential Elections Dave Leip's Atlas of US Presidential Elections Continuous Fiscal Need: Fiscal Dependence % of total revenue from federal intergovernmental revenue State and Local Finance Survey Continuous Fiscal Need: Functional Autonomy Index of the autonomy states' grant localities Local Autonom y Index (Wolman et al 2010) Ordinal (1 - 6, low to high) Current Problem Based Need Size of Homeless Population HUD's Homeless Populations Reports Continuous 74 Table 4. . Concept Variable Source Measurement Future Problem - Based Ne ed % of Population in Poverty US Census Continuous Control: Population Population Size US Census Continuous Control: Community Education Level % of Population with a Bachelor's Degree US Census Continuous Control: Racial Makeup % of Population that is White US Census Continuous I measure three types of attention: bureaucratic, advocacy, and executive. Bureaucratic the housing authority is a member of the C applicant. 12 13 and is coded 1 if the housing authority is a member or if the city is the consolidated applicant and 0 otherwise. I use the num ber of organizations in a CoC as a proxy for advocacy attention. The types of organizations involved in a CoC vary by community, but typically include some combination of service providers, advocacy/policy organizations, governmental agencies and religious organizations. All member organizations are involved in homelessness in some way and as a result, I expect that increasing the number of organizations in an area also increases the number of eyes on a problem, the number of resources drawing attention to the issue, and the prominence 12 The CoC consolidated applicant is a local government or not for profit o rganization that prioritizes funding requests from all of the organizations in a CoC area and creates a single application for funding. It then administers this funding and is responsible for tracking program and outcome data. 13 site. As of 5/15/15 this information could be accessed through https://www.hudexchange.info/grantees/ 75 proxy for executive attention. I do this for two reasons. On one hand, the federal government has called upon mayors to tackle homele ssness and in many cities they have been entrepreneurial in their pursuit of solutions for homelessness (Mayors 2014; Mayors 2015). On the other, there is a long - standing belief that city managers are better equipped to address the real problems that citie s face because their attention is not diverted by political considerations (Hays 1964) and that these managers can be entrepreneurial when problems present themselves (Teske and Schneider 1994) (though there is increasing recognition that the administrator /politician dichotomy is not as pure as the original progressive reformers might have hoped (Svara 1999). If public administrators are indeed better equipped to deal with tangible problems, we would expect council/manager forms of government to be more lik ely to develop 10 year plans than their mayor/council counterparts whereas if mayoral entrepreneurship is helping overcome the politicized nature of the position, mayor - council forms would be more likely to adopt. This information comes from the Internatio database of local governments (ICMA 1996, 2001, 2006, 2011) and independent searches of municipal websites. Cities are coded 1 if they have a mayor/council system and 0 otherwise. In terms of resources, the f oundational question is whether cities have the staff and money to respond to federal incentives. Human resources are measured as the number of municipal employees and the number of municipal housing and community development employees. This information co Employment Survey, which is given to all cities in years ending in 2 and 7 and is given to major cities (including most sample cities) in intercensual years. For cities where intercensual data are n ot available, I use linear interpolation calculated from the values in years ending in 2 and 7 to provide estimates for the interim years. COGs data are currently available through 2011 and 76 linear extrapolation was used to provide estimates for years 2012 - 2014. To account for fiscal capacity I use revenue per capita and housing and community development spending per capita. 1999 2011. All cities self - report these da ta every five years in the full Census of Governments; most major cities also report these figures in the annual intercensual survey. I also account for the three dimensions of need that I expect to drive variance: political, fiscal and problem. I identi fy political need in two ways: consistency with existing policy priorities and consistency with local ideology. To the first point, I include spending on housing and community development as a percentage of total spending. Cities that spend a higher percen tage of their budgets on housing and community development are assumed to have policy values/priorities that are consistent with policy and action on homelessness. These data come 2011, w ith years 2012 - 2014 linearly extrapolated. To the second point, I include the Democratic percent of the vote share in presidential elections in 2000, 2004, 2008, and 2012. These data are gathered at the unty. In cases where a primary county does of US Presidential Elections (Dave 2014). For fiscal need I account for fiscal dependence and functional autonomy. The degree to which cities are financially dependent on the federal government is measured by the percent of total revenue made up by federal intergovernmental transfers. These data come from the Census 2011, with years 2012 - 2014 linearly extrapolated. For legal autonomy I use the local autonomy index developed by Wolman et al 77 are measured by state (so all localities in a single state receive the same score). Here the index is coded from one to six, with values of one representing lower levels of autonomy and values of six repres enting higher levels of autonomy. Finally, in terms of the problem, the clearest measure of problem - based need is the size to accurately record the size of the hom eless population in their area every two years (though most CoCs hold an annual or even biannual count). This data provides our most accurate estimate of the homeless population, but there are two challenges with the data. The first is that systematic effo rts to count and record the homeless across US cities only began in 2005. To deal with this, I use the average yearly change in the homeless population to linearly extrapolate the homeless population from 2000 - 2004. The second challenge is that our estimat es of the homeless late January volunteers count all those individuals sleeping on the streets, in homeless encampments, and in homeless service centers or shelter s. Though these counts underestimate the size of the homeless population by missing all those who are able to find temporary extensive study of the best methods for co unting the homeless and service providers have worked to make the counts as accurate as possible. The real problem then, is that CoCs often include multiple communities, not just the central city, and the homeless count reflects the size of the homeless po pulation across the entire CoC area. In other words, in many communities we do not have a count of the homeless population residing in the city itself but instead have an estimate that includes surrounding communities. 78 Despite this challenge, I treat CoC counts as if they belong exclusively to the central city in the COC for two reasons. First, from a pragmatic perspective, there are no other estimates of the homeless population that would offer more accurate information, and second there is ample evidenc e that the homeless tend to be concentrated in urban areas, with smaller communities often directing their homeless to larger communities to access services (Paisner 1994; Rahimian, Wolch and Koegel 1992; Simon 1991). This reality likely tempers any proble ms that arise from attributing the full homeless population to the primary city in the CoC. Data thus come from the homeless figures reported to HUD by individual CoCs (COC 2014) and include both sheltered and unsheltered homeless from 2005 to 2014, with extrapolated data from 2000 to 2004. In the event history analysis I include both total homeless counts and the percent change in the homeless population over the previous year, while in the logistic regression I include only the absolute size of the homel ess population. In addition, I include the percent of individuals living in poverty as a measure of the low income population. These data come from the 2000 decennial Census for the years 2000 - 2004, from intercensual estimates for 2005 - 2009 and the 2010 De cennial Census for 2010 - 2014. In addition to these measures I control for population, the that is white. These measures also come from the 2000 decennial Census for the years 2000 - 2004, from intercensual estimates for 2005 - 2009 and the 2010 Decennial Census for 2010 - 2014. Evaluating the Evidence: What has Influenced Individual City Adoption? The results from the logistic regression, detailed in Table 4.4 , larg ely support the theoretical expectations laid out above and highlight the importance of fiscal, political, and problem - based need as a driver of plan adoption. Increased bureaucratic and advocacy attention to homelessness and 10 - year plan incentives - as me asured by the involvement of the local 79 housing authority in the CoC, the city as the primary CoC applicant, and the size of the CoC advocacy community - are insignificant. Similarly, administrative and fiscal capacity - as measured by overall municipal emplo yment and revenue, and housing and community development - specific employment and revenue - are insignificant. Interestingly, the only non - need - based measure that significantly predicts plan adoption is the form of government. Based on predicted probabilitie s, when all else is held constant, cities with council - manager forms of government are 15 percent more likely to adopt a plan than their mayor - council counterparts ((74 percent probability of adoption vs. 59 percent probability of adoption). This finding p rovides some support for the argument that city managers are better equipped to address tangible problems because of their relatively a - political position. It also suggests that even though many mayors have taken a leadership position in the fight against homelessness, this leadership is not only unnecessary for plan development but does not even make their cities more likely to adopt a 10 - year plan. Beyond the form of government, the other predictive characteristics fall into the need dimension of responsi veness; I discuss these results in terms of predicted probabilities below. 80 Table 4. 4. Logistic Regression Predicting 10 - Year Plan Adoption N= 255 Prob>Chi2=0.00 Log Likelihood= - 90.49 Pseudo R 2 =.291 Covariate Robust SE Housing Authority in CoC 0.34 0.43 City is Consolidated Applicant - 0.43 0.36 Number CoC Orgs/Capita - 2369.82 1324.50 Total Municipal Employment 0.00 0.01 Municipal Housing & Comm unity Dev Employment - 0.01 0.01 Total Revenue/Capita 0.0 1 0.18 Democratic Population 1.65** 0.47 Housing as Percent Total Expenditures - 3.09 2.83 Federal IGR/Capita 4.17** 1.47 Functional Autonomy - 0.77** 0.21 Homeless Population 0.01** 0.00 Percent Poverty 10.31** 2.80 Mayor - 1.06** 0.34 Population 0 .00 0.00 Percent White 3.46** 1.31 Percent Bachelors 3.91 2.50 Constant - 5.63** 1.42 **sig>=.01 *sig>=.05, two tailed test In Hypotheses 1 and 2, I predicted that constituent demands would influence plan adoption, arguing that more liberal po pulations and communities that had revealed policy priorities consistent with increased action on housing issues would be more likely to adopt a 10 year plan. Although policy priorities - as measured by relative spending on housing and community development - was insignificant, political ideology proved to be a significant predictor of plan adoption. Democratic cities are 30 percent more likely to have adopted a 10 - year plan than their Republican counterparts ((75 percent probability of adoption vs. 45 percen t probability of adoption), lending strong support to H 1 and the argument that one of the benefits political needs. 81 In Hypotheses 3 and 4 I predicted that ci ties that are more dependent on the federal government for revenue and those that have less autonomy to tax and spend would be more likely to develop a 10 year plan because their fiscal needs encourage cooperative action and compliance with federal request s. The results, depicted in Figures 3 and 4 provide strong support for these hypotheses. In Figure 4. 3 we see a 32 percent increase in the probability of adoption as we move from cities with the lowest percent of revenue from the federal government to the highest. More dramatically, in Figure 4. 4 we see the effect that municipal autonomy - and the lack thereof - has for plan adoption. Cities in states with the highest level of municipal autonomy are sixty percent less likely to adopt a plan than those cities in states that give the least autonomy. Together, these results reveal that dependence and constraints from higher level governments, whether legislative or revenue based, create unique benefits for cooperation and provide cities with a distinct motivation to respond to federal incentives. 0.58 0.66 0.73 0.79 0.84 0.88 0.91 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1% 5% 10% 15% 20% 25% 30% PR(X) PERCENT REVENUE FROM FEDERAL TRANSFERS Figure 4.3. Predicted Probability of Adoption Based on Percent of Revenue from Federal IGR 82 Finally, Hypotheses 5 and 6 develop expectations about how problem severity influences plan adoption, suggesting that as homelessness and poverty increase in a community, so too will the probability of adoption. The results, depicted in Figures 5 and 6 provide strong evidence for both of these hypotheses. Figure 4. 5 shows that when the homeless population is very small, there is a 50 - 50 chance that a city will adopt a 10 - year plan but that as the population increases , the probability of adoption also increases. Communities with at least 1000 homeless individuals are more likely than not to adopt a plan and those with a population of 10,000 or more have a 99 percent chance of adoption. As a point of reference, in 2012 Chapel Hill, NC had a homeless population of 156, Kansas City had a homeless population of about 1,800, and Los Angeles had a homeless population of over 42,000. Similarly, Figure 4. 6 shows that moving from cities with the lowest poverty to those with the highest, increases the probability of adoption by over 40 percent. 0.84 0.72 0.59 0.44 0.31 0.21 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1 2 3 4 5 6 PR(X) FUNCTIONAL AUTONOMY SCALE (1=LOW AUTONOMY 6=HIGH AUTONOMY) Figure 4.4. Predicted Probability of Adoption Based on Functional Autonomy 83 0.53 0.64 0.73 0.82 0.88 0.92 0.95 0.97 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 100 1000 2000 3000 4000 5000 6000 7000 PR(X) SIZE OF HOMELESS POPULATION Figure 4.5. Predicted Probability of adoption Based on Homeless Population 0.38 0.43 0.52 0.63 0.72 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 5% 10% 20% 30% 40% 50% PR(X) PERCENT LIVING IN POVERTY Figure 4.6. Predicted Probability of Adoption Based on Population Living in Poverty 84 Discussion and Conclusions It has been over four decades since the nation abandoned comprehensive urban policy; yet, the federal government is perhaps as active today as it has eve r been in shaping local policies disaggregated into a mix of pol icies and programs that tackle functionally isolated urban that address everything from water quality to drug prevention and in doing so layer federal goals in to most aspects of urban governance. In recent years, scholars studying these efforts have framed federal - local relations in terms of coercive actions, suggesting that localities are increasingly forced to engage in and pay for programs and policies that r eflect higher - level goals rather than local priorities. However, the case of homeless policy generally and 10 - year plans specifically highlights the fact that cooperative federalism not only still exists but can be extremely effective in promoting federal goals while helping local governments meet their needs. In the preceding analysis we saw both the anecdotal and empirical evidence that the federal government spurred widespread adoption of 10 - year plans by linking plan creation to financial incentives. Cooperative federalism suggests that both parties must receive benefits from cooperation and it is often assumed that the most effective benefit higher levels of government can offer is financial resources. Yet, we also saw that cities have not just blind ly chased federal money; they have been most likely to respond to the federal government when it helped them address more fundamental needs, which include but are not limited to fiscal concerns. This reality reflects the dynamic environment in which cities operate, one where they must respond not simply to fiscal demands and political pressures but also to tangible problems. 85 In this chapter I have argued that variation in adoption reflects differences in the benefits cities receive from cooperation; in ot her words, a significant part of what determines whether a city adopts a plan is the degree to which cooperation helps it meet its fiscal, political and problem - based needs. While the fiscal federalism literature often portrays cities as driven by money fo revenue a city has - not how wealthy or poor it is - but rather the degree to which the municipality depends upon federal resources and possesses autonomy to generate new revenu e and policies. Thus, cities are not just pursuing money blindly, those cities that function in more dependent environments are more likely to respond to federal incentives, suggesting that rather than, for example, being motivated simply by a desire to gr ow their bureaucracy, cities respond to the constraints they face and recognize the benefits that cooperation has in light of their dependent position. The significance of this distinction is made more evident when we compare cities like Holland, MI and W aco, TX to cities like Yakima, WA and Provo, UT. All of these cities are majority - Republican and have small homeless populations (from 150 - 450 people). Alone, these facts suggest that the cities would be unlikely to develop plans, since there is little po litical or problem - based pressure to do so. Yet Holland and Waco receive large portions of their budgets from federal sources - 17 percent and 30 percent respectively - and have responded to federal incentives by developing 10 - year plans. In contrast Yakima and Provo depend very little on federal resources, which make up less than four percent and less than one percent of their municipal budgets respectively; without fiscal, political or problem based needs, these cities have not developed a 10 - year plan. Th is suggests that while fiscal dependence and limited autonomy make cities more vulnerable to fiscal instability, they also make them more responsive to the 86 goals, incentives and programs that higher level governments create and can help state and federal g overnments overcome local ideological pressures that might otherwise impede their goals. Importantly, while fiscal dependence can help municipalities develop plans in places where it would not appear to be politically tenable, the local political enviro nment is still a significant part of the story. Though policy priorities - as reflected in housing and community development spending - are insignificant, local political ideology is not only statistically significant but is also substantively important. Onl y 23 percent of the adopting cities in the sample have a Republican majority; the rest are Democratic cities that many would expect to be of policies is enough tha t small Democratic cities like Scranton, PA, Montgomery, AL and Stamford, CT, with small homeless populations (250 - 350 individuals) and low dependence on federal resources , have nonetheless adopted 10 - year plans. Thus, the political benefits that cities de rive from cooperation appear to be very meaningful and are an important predictor of how readily cities adopt federal incentives. While one conclusion is that federal incentives may be more successful when they have bipartisan support, it is essential to remember that 10 - year plans came out of a conservative, Republican administration. So, it is perhaps not enough to have national bipartisan support if the of 10 - year plans, the Bush administration framed adoption in terms of compassionate conservatism, an effort to transition the party into more social policy, but that message does not appear to have resonated with many conservative localities. To gain wides pread local buy - in, the federal government might do well to frame its incentives in multiple ways that appeal to different 87 ideological audiences. For example, there are strong fiscal arguments for dealing with homelessness that might be more effective in s purring plan adoption in conservative areas. I have suggested that the final type of benefit cities derive from cooperation is meeting problem - based needs. This is something that is not discussed frequently in the policy, urban politics and federalism literatures, which focus more heavily on the politics, attention and economic drivers of policy development. Nonetheless, as I discussed in Chapter 2 and again earlier in this chapter, local governments are distinct from state and federal units in that th ey come face to face with the problems their communities deal with on a daily basis. This means that the problems themselves are likely to drive policy choices in a pronounced way. As expected, this is precisely what we see. Cities with large homeless pop ulations are far more likely to develop plans than cities with small populations - even when those cities have Republican majorities. On average, cities without a 10 - year plan have a homeless population of 660 people and 50 percent of communities have a hom eless population under 400 people. In contrast, the average city with a 10 - year plan has a homeless population of 2236 and 50 percent have a population over 1020. When the local problem is severe, it is enough for many cities to overcome their political te ndencies, so that strongly Republican cities with substantial homeless populations like San Antonio, TX, Riverside, CA, Omaha, NE and Greenville, SC have adopted a 10 - year plan in spite of the fact that it is inconsistent with a traditional conservative po licy agenda. Further, we saw that cities are influenced not simply by the immediate problem but also by the prospect of future problems - as measured by the percent of individuals living in poverty. This suggests that cities are not only aware of the curre nt conditions but are also making strategic choices based on the possibility of future conditions. When cities anticipate that they will be able 88 to address an underlying problem by responding to a federal incentive, they are more likely to engage in this t ype of cooperative action. Throughout this chapter I have focused on the benefits that localities derive from cooperation, treating federal benefits as a given. In the next chapter I challenge this a bit. Previously, I suggested that the benefit the feder al government gets from engaging in a cooperative relationship and providing an incentive for 10 year plan development is that it can achieve its goal of reducing homelessness throughout the nation. Further, I explained that this goal is not possible witho ut local buy - in. The key policy question for federal actors is whether widespread plan adoption actually helps reduce homelessness. Obviously there is a big difference between developing a planning tool and actually carrying out the substantive policy change that is necessary to reduce the homeless population. While cities might be motivated to develop a low - cost planning tool for ideological reasons or to improve their chances of receiving funding, these motivations are unlikely to be sufficient to fac ilitate real change. In the next chapter I investigate what helps cities transition from simply having a planning tool to implementing policies and practices in ways that might actually reduce homelessness, arguing that capacity and attention drive real po licy changes. 89 CHAPTER 5. Symbol or Substance ? : Assessing the R eal Impacts of Federal Incentives to End Homelessness In 2009 the National Alliance to End Homelessness began to assess the viability of 10 - year plans as drivers of real change. While the id ea of a comprehensive planning tool to end homelessness continues to be one of the most visionary changes in homelessness policy, the can only happen if the p implementation, setting numeric outcomes, identifying a funding source, and setting a cl ear 234 plans developed up to June 2009. For those hoping that the plans would be implemented with ease, the results are disheartening. Less than 25 percent h ad a timeline, only 15 percent had identified an implementing body, fewer than 20 communities had included numeric outcomes, and only eight communities, less than 5 percent of those analyzed, had identified a funding source. The lack of specificity and li mited foresight about implementation suggest that in many communities plans may serve as symbolic statements rather than substantive policy. not deliver what they appea r to deliver; they allocate no tangible advantages and disadvantages. common at all levels of government and in priva te organizations (Anderson 2015 ). They serve to legitimize organizations (Berrone 2009), appease particular interests (Calavita 1993; Rodrigue et al 2013), and provide a sense of wellbeing to the masses (Edelman 1964). What they do not typically do is lead to real change. In the case of 10 - Year Plans , this is problematic because the 90 federal government has encouraged localities to adopt the plans as a way to promote the substantive policy change necessary to fulfill the national goal of ending homelessness. At the same time, advocates contend that 10 - y ear plans have helped many communities make real strides towards ending homelessness. Whether it has been the plans themselves or the increased attention that planning brings, there is truth in the contention that homelessness has improved since 10 - year pl ans began being adopted. While it continues to be a serious problem throughout the nation, many communities are making progress in tackling homelessness. In fact, during the first few days of 2015, New Orleans announced that it had housed all of its homele ss veterans in permanent accommodations. Similarly, both Phoenix and Salt Lake housed all of their chronically homeless veterans between the end of 2013 and 2014 (Lerner 2015). These stories of success indicate that real change is happening in some places ; and, in combination with the significant deficiencies that the majority of 10 - year plans have, suggest that long - term planning has been symbolic in some communities and substantive in others. Deciphering which communities have gone beyond developing a sy mbolic planning tool to implementing substantive policies is the next step in understanding how federal homelessness incentives have worked. This chapter takes up this question, exploring the conditions under which communities have engaged in substantive policy. The chapter first discusses the literature on symbolic policy, drawing heavily from both the social sciences and organizational business theory, to develop hypotheses about what encourages cities to engage in substantive practices. It lays out expe ctations about the role of external monitoring, the importance of local capacity and the significance of policy entrepreneurs in promoting substantive action. Further, it connects to the general theory laid out in Chapter 2 by suggesting that capacity and attention are more important 91 for carrying out substantive practices than for developing a 10 - Year Plan. It then assesses the predictors of three types of substantive practices: providing adequate beds for the homeless, giving the homeless priority in publ ic housing, and actively engaging in the 100,000 Homes Campaign. The heart of the question is twofold. First, whether the federal push to develop 10 - year plans (including plan incentives) has actually influenced substantive practices. And second, beyond 1 0 - year plans, what local characteristics are associated with substantive approaches. Regression analysis shows that 10 - year plan adoption is almost wholly unassociated with substantive policy, suggesting that plans are largely symbolic. Beyond this consist ent finding, the results are mixed, showing that what predicts substantive policy varies considerably across the practices explored. Predicting Symbolic and Substantive Policies Symbolic policies are a mainstay of modern American politics, providing a way for government to respond to problems but do so in a way that does little more than set ideas to paper. In organizational theory, symbolic policies are conceived of as a decoupling of policies that conform to social pressures from the real daily polici es and practices that govern organizational operations (Meyer and Rowan 1977). While this practice can be frustrating and even seem nefarious, there are very real reasons that governments and other types of organizations engage in this practice. In his se minal work on the topic, Edelman (1964) echoed Laswell (1930) in suggesting that symbolic policies give government a way to either placate or arouse emotion in the masses. They allow government to publically confront issues and in so doing offer a sense of well - being that allows concerned parties to turn their attention to something else. However, as Lasswell (1986) argued, 92 It should not be hastily assumed that because a particular set of controversies passes out of the public mind that the implied problem s were solved in any fundamental sense. Quite often a solution is a magical solution which changes nothing in the conditions affecting the tension level of the community, and which merely permits the community to distract its attention to another set of eq ually irrelevant symbols. The number of statutes which pass the legislature, or the number of decrees which are handed down by the executive, but which change nothing in the permanent practices of society, is a rough index of the role of magic in The incentive to placate the public stems from two equally important motivations. First, regarding policies that are specific to group interests, government officials have an incentive to minimize dis cord in a way that leads to a peaceful society. In part this can be done by using symbolic policies to appease groups that demand resources but are satisfied with nontangible promises that reassure the group. By lulling some groups into quiescence through symbolic acts, the government has available tangible resources for those groups that have power and are not so easily pacified (Edelman 1964). Second, regarding issues that have broader public concern, the government has a primary interest in maintaining legitimacy and as a result must at least appear to respond to issues that the public cares about. When those concerns are inconsistent with what government officials want to do they are likely to develop and publicize policies that appear to address the is sue but which in widely publicized administrative activities can most confidently be expected to convey a sense of well - being to the onlooker because they sugges t vigorous activity while in fact signifying have used the general idea of legitimacy to discuss the ways in which organizations use symbolic policies to manage their public image (Ashforth and Gibbs 1990; Bansal and Kistruck 2006; stakeholder relationships...or even manipulate stakeholder perceptions and/or decis 93 (Rodrigue et al 2013, 111). This is particularly common when stakeholder demands are inconsistent with internal goals (Oliver 1991; Scott 1995; Westphal and Zajack 2001). In terms of homelessness policy, these motivations - minimizing dissent and re source demands and maintaining legitimacy/managing image - provide significant leverage in understanding why a community might treat a 10 - year plan as a symbolic document. Cities face both top down and bottom up pressures to deal with homelessness. From the top the federal government wants localities to end homelessness and from the bottom local publics want cities to order to appease federal and public demands for action, maintain legitimacy as an effective and re sponsive government, and encourage an image of the city as a good place to live and do business, city officials have an incentive to develop policies that, at least on their face, address the issue. Developing a long term plan is a low cost way to be respo nsive without requiring substantive change. By developing a symbolic plan, cities are able to publically proclaim their commitment to ending homelessness and offer the public comfort that their community is standing on high moral ground. One of the more i moral nature (Gusfield 1963) and by serving the neediest among the population, cities can use 10 - c ities throughout the country homeless organizations and in rare cases the homeless themselves, have increasingly demanded that cities do more than criminalize and isolate the homeless (Williams 2005), something that not all cities want to do. A plan provid es something that cities 94 can point to as a response to these demands without actually having to provide material resources. In addition to these more cynical views, there is a nobler reason governments may develop symbolic policies. Many times government o fficials want to improve conditions and pass laws they believe will help achieve this end. It is only later that they find there are insufficient resources to effectively carry out the policy. Whether it is a lack of money, technology, human resources or s ustained attention, these policies can languish and become little more than a symbol of good intentions. Despite increases in federal monies, cities and service organizations alike continue to struggle to find adequate financial and human resources to figh t homelessness. It is very possible that some communities have a real desire to end homelessness and have developed a 10 - year plan with this in mind but have faced resource and attention barriers in actually developing the substantive changes that would he lp achieve the goal. This is not to say that all communities have adopted a strictly symbolic approach. Even Edelman (1964) acknowledged that policies typically have both symbolic and substantive components and as was established above, a variety of commun ities have made strides in confronting homelessness. It is however to say that cities have many reasons to develop symbolic policies; why they go beyond this to develop substantive practices is somewhat less clear. Answering why cities engage in substant ive practices is in fact an effort to understand the conditions under which cities provide more than a surface level response to the federal - year plan incentives. With a goal of ending homelessness, the federal government clearly did not in tend plans to be strictly symbolic, yet what is required for funding advantages is just a document. There is no oversight about plan content or the practices that are developed in light of the plan. Instead, to make progress towards their goals, the federa l 95 government has relied upon localities to take heart of the underlying mission and engage in the substantive practices likely to reduce homelessness on their own. Understanding the conditions that promote substantive action gets at the predictors of meani ngful responses to federal incentives. The theoretical framework laid out in Chapter 2 suggests that local attention, capacity and most basic response - devel oping a plan - local need drove responsiveness, with attention and capacity playing a very limited role. In contrast, for a substantive response that includes tangible practices, cities need prolonged attention and substantial resources. In fact, the degree to which they need a substantive response is somewhat irrelevant if they cannot muster the attention and resources to carry out tangible practices. The same understanding of capacity that was used in Chapter 4 - human and fiscal resources - is maintained he re; however, the concept of attention is expanded to include external oversight. The organizational and policy literatures suggest that there are at least three major predictors of whether organizations decouple policy from practice. Two of these deal wit h attention in the form of oversight and policy entrepreneurship; another - consistency between a - has already begun to be discussed. This literature leads to several hypotheses about whether cities treat 10 - year plans as symbolic statements or move beyond them to engage in substantive policies. These hypotheses, as well as one relating to the importance of human and fiscal capacity, are detailed below. Research on the use of ethics codes in executive decision ma king suggest that third - party monitoring reduces the likelihood that companies treat their codes of conduct as symbolic documents (King et al 2005). Monitoring and oversight are forms of attention to local behavior. 96 As it relates to homelessness policy, we might expect cities to engage in substantive policies when there is third - party attention/monitoring of their activities. Unlike industry however, municipal governments do not necessarily have a clear auditor; instead, third - party oversight comes in a var iety of forms, most notably state and federal governments, watchdog agencies and the public or affected group. requiring cities to report the size of the homeless population , gaps in service provision, and plans for future action. All communities receiving federal funding are subject to this annual monitoring. However, HUD officials know only what localities tell them and, as in business, limited knowledge of the situation on the ground may limit the effectiveness of this type of oversight (Boiral 2003; Swift et al 2000). As a result, monitoring from state governments, advocacy networks, and the public may be necessary to promote tangible action. Regarding state oversight, th e assumption is that the more heavily involved a state is in addressing homelessness issues the more likely they are to pay attention to what local governments are doing to combat homelessness. In combination with the fact that cities are dependent on stat es, this oversight is likely to promote substantive action. As a result, hypothesis one is as follows: H 1 : Cities in states that are more heavily involved in homeless issues will be more likely to engage in substantive policy than their counterparts in states that are less involved. In terms of advocacy networks, community organizations can serve as watch dogs, calling attention to undesirable municipal (in)activity. For example, the National Law Center on 97 Homelessness and Poverty regularly identifies t attention to policies that criminalize and isolate the homeless (Homes 2009; No 2013). On the flip side, these same organizations can be essential partners for localities in developing and carrying out substantive policies. The combination of oversight, advocacy and service provision suggests that strong advocacy networks would encourage substantive policy, leading to the second hypothesis: H 2 : Cities with large advocacy networks will be more likely to have substa ntive policy than those without large networks. Finally, regarding public oversight, the courts provide powerful recourse for publics disheartened by local policies (or lack thereof). For their part, lawsuits can provide a clear signal to cities that they are being monitored and that what they are doing is unacceptable, at least to some people. Even more than on other issues, lawsuits about homelessness are likely to primarily serve a signaling role because they are rarely won by the plaintiff and conseque ntly infrequently require municipalities to alter their behavior (Homes 2009; No 2013b). Nonetheless, they are a reminder to the city that someone is paying attention and as a result may induce more substantive policy. As a result, hypothesis three is as f ollows: H 3 : Cities that have faced litigation on homeless - related issues will be more likely to engage in substantive practices than those that have not faced such litigation. In addition to oversight, work on local environmental policy suggests that cit ies with a clear policy entrepreneur are more likely to engage in substantive policy than those without an entrepreneur. More specifically, Krause (2011) shows that after agreeing to reduce greenhouse gas emissions and joining climate - protection networks, cities are more likely to actually engage 98 in tangible efforts to reduce emissions when they have a dedicated municipal employee working on the issue. This individual can help maintain attention, advocate for policies and generally keep things from falling through the cracks. As was discussed in Chapter 4, mayors have served as primary policy entrepreneurs on homeless issues, focusing municipal governments on compassionate, long term approaches to ending homelessness. Although mayoral cities were actually le ss likely than council - manager cities to adopt a 10 - year plan, the sustained attention and work required to carry out substantive policy is likely to require an entrepreneur in a way that developing a plan does not. This leads to the fourth hypothesis: H 4 : Cities with a mayor - council form of government will be more likely to engage in substantive practices than those with other forms of government. In addition to oversight and entrepreneurship, much of the literature on symbolic policy focuses on distingu ishing between those policies that are intended to affect change and those that are designed to legitimize government and appease constituent demands. Work on waste - removal (Gregory 1999), criminal justice (Gusfield 1963), immigration (Calavita 1993), and environmental policy (Seis 1993) all suggest that a singular determinant of substantive action is that the policy aligns with internal goals. Conversely, when policies indicate outcomes that are ore likely to adopt symbolic policies. For example, Calavita (1993) argues that major US immigration laws are largely symbolic because the federal government faces conflict between competing demands. On one hand, they want to promote economic growth and th at growth is dependent on immigrant labor; As it applies to homelessness, this would suggest that communities may adopt a 10 year plan as a way to legitimize themselves in the eyes of both higher level governments and the public, but fall 99 short of substantive policy because the plan is inconsistent with local priorities. Hypothesis five follows from this idea: H 5 : Cities will be more likely to engage in substa ntive practices when they are consistent with local goals and priorities. Finally, beyond the three characteristics identified from the literature, local capacity is likely to be a much greater predictor of substantive practices than it was in plan adopti on. Implementing real change requires substantial resources. There has to be money and space to create new housing for the homeless, staff to coordinate not just services but funding sources, communication between government and community organizations; th e list goes on and on. As a case in point, a 2012 HUD survey found that over 40 percent of local housing authorities identified a lack of human resources as a barrier to working with the homeless (PHA 2012). Further, putting a homeless resident into transi tional or permanent housing, while significantly cheaper than using emergency services and the criminal justice system, still costs tens of thousands of dollars annually (Larimer et al 2009). Many of the resources to house the homeless are channeled to cha ritable organizations from private donations and foundations or come from state and federal resources, but even if we assume that cities need not put a penny into service provision, they still need municipal employees to develop and carry out policies and coordinate financial resources. The State of Florida estimates that it takes 5.5 full - time equivalent employees and nearly $130,000 simply to meet the prerequisites necessary to apply for federal funding (Council 2013). With many cities under financial str ain, this may be a tall order. Thus hypothesis six follows: H 6: Cities with greater fiscal and human resources will be more likely to engage in substantive practices than their counterparts with fewer resources. 100 Together these hypotheses suggest tha t in order for a community to engage in substantive practices, they need to have internal goals that are consistent with using a compassionate approach to reduce homelessness. Further, external monitors and policy entrepreneurs need to pay attention to mun icipal practices and help keep municipalities focused on homelessness. Finally, the community needs substantial human and fiscal resources to be able to carry out the practices. Below, I test these hypotheses on three types of substantive action: having be ds and permanent housing for the homeless, having a public housing authority that engages with the homeless, and joining the 100,000 Homes Campaign. Data and Methods Homeless advocates have identified three practices that are essential for reducing the homeless population: developing permanent supportive housing ( Tsemberis 2010; Tsemberis et al 2004 ; US 2013 ), and adopting a housing first model (Housing 2015; Tsemberis et al 2004). To assess the driv ers of substantive behavior the following analysis uses measures of these practices as dependent variables. Having a sufficient number of beds to accommodate the homeless population is the first step in getting people off the streets. The second is putti ng them into long - term housing; yet, communities throughout the country have sighted a lack of permanent and transitional housing as a key barrier to addressing homelessness. As a result, one of the most important substantive behaviors a community can enga decent, safe, affordable, community - based housing with flexible, voluntary support services designed to help the individual or family stay housed and live a more productive life in the communit 101 the number of beds available to the homeless relative to the size of th e homeless population housing and is measured as the percent change in permanent beds available in the city from 2005 to 2014. Both of these housing measures co Reports, available through the HUD Exchange. 14 an d vouchering programs, food stamps, SSI, SSDI, general assistance, TANF, job training, 2002, xv). The agency has further identified the singular importance of ha ving public housing authorities (PHA) actively working on homeless issues, since many homeless individuals need below - market housing if they are to stay off the streets (PHA 2013). Public housing agencies are so important that in 2012 HUD conducted a surve y of all 3,988 PHAs to assess their involvement in homeless issues. PHA Homeless Preferences Web Census Survey (PHA 2012) had an 80 percent response rate with 3,210 cities completing the survey. Three of the measures from this survey assess involvement in progressively more resource - intensive activities and represent the extent to which localities are using their mainstream public housing services to deal with homel essness. These measures are 1. R eviewing ngaging t he homeless in any way and, 3. G iving preference to the homeless in public housing decisions. Each of these is a dummy variable where one indicates that the city engages in the practice and zero indicates they do not. 14 Available at https://www.hudexchange.info/manage - a - program/coc - housing - inventory - count - reports/ . 102 Finally, housing first is a relatively new app roach to addressing homelessness, but it is one that advocates and scholars have identified as a key to keeping formerly homeless individuals from returning to the streets. The approach works from the premise that in order for homeless individuals to deal with the issues that have led to their homelessness they must first have stable housing. This model disrupts previous approaches that sought to implement things like addiction counseling prior to providing someone with housing. In July 2010 the homeless ad vocacy organization Community Solutions launched the 100,000 Homes Campaign with goal was for community partners to put 100,000 homeless individuals into permanent housing using a housing - first approach. By joining, communities were given access to resources, three major goals of getting individuals into housing, knowing those on the street by name, and tracking progress in order to make data - driven decisions. Over 100 communities signed on to the Although it is an imperfect measur e, joining the campaign is used as a proxy for adopting a housing - first approach and is measured using a dummy variable where one indicates a city has joined. Further, since joining requires few resources and can itself be symbolic, I assess active members hip in the campaign. Communities are considered active when they have reported a non - zero value for the percent of the homeless population they have housed through the program. Communities that both joined and housed homeless individuals through the progra m are coded one while all others are coded zero. These data come from the 100,000 Homes 103 These measures provide variation along two important dimensions: the spectrum of possible substantive behaviors and the actors necessary to carry out substantive practices. Variation in actors is particularly important since what is necessary to get a city to act alone is likely different than what is required for cities to work in partnerships. In the practices that have been cho sen, local governments can only act independently in making decisions about how responsive their PHAs will be to the homeless. They cannot provide beds for the homeless or develop permanent supportive housing alone; instead they must work in partnership wi th community organizations. Similarly, though they can commit to taking a housing - first approach and joining the 100,000 Homes Campaign, they must rely on community organizations and coalition members to help them achieve active membership. Independent Va riables and Controls The independent variables account for the attention - capacity - need framework set up in Chapter 2. For this reason, most of the variables are identical to those used in the empirical analysis in Chapter 4. Chapter 4 provides a detailed measurement information and justification for variable inclusion; as a result, the discussion is not repeated here. Table 5. 1 provides a summary of the variables used in the analysis that follows. Four variables account for attention: the number of CoC or ganizations in a community, whether the public housing authority is part of the CoC (1=Yes), whether the city is the consolidated applicant for the CoC (1=Yes), and whether the city has a mayor - council form of government (1=Yes). As in Chapter 4, form of g overnment serves as a proxy for executive policy entrepreneurship, since mayors have served this role in cities across the country. In addition to these variables, two new measures account for state and public oversight. As was discussed above I use state involvement in homeless issues as a proxy for their role in external 104 monitoring and homelessness - related litigation as a proxy for public oversight. State involvement is measured with an additive scale of the following dummy variables: developing a state 1 0 - year plan to end homelessness, having a state - level homelessness committee, having a gubernatorial homeless initiative, having a state interagency council on homelessness, and developing a state homeless bill of rights. These variables come from a combin ation of independent internet searches and data from the National Survey of Programs and Services for Homeless Families: The Red White and Blue Book (National 2015). States that are the most active in homeless issues, engaging in all practices, have a valu e of five while those that are the least active, score of .8. Litigation data come from the National Law Center on Homelessness and Poverty, which provides an notated lists of all homelessness - related litigation in the country ( National 2009; N ational 2013 a ). This is a dummy variable where one indicates that a city has been involved in litigation and zero indicates that it has never been involved in litigation. Capacity is accounted for with four measures: total revenue per capita, housing and community development spending as a percent of total spending, total municipal employment and housing and community development employment. As in Chapter 4, housing and co mmunity development spending also serves as a proxy for local policy priorities, which when paired with local ideology helps account for the consistency between 10 - year plan goals and local priorities. 105 Table 5. 1. Summary of Included Variables Concept Variable Source Measurement Substantive Action: City has Adequate Beds for Homeless Number of Beds Available for the Homeless Relative to the Size of the Homeless Population HUD's Housing Inventory Count Repo rt and Homeless Population Report Continuous Substantive Action: Developing Permanent Supportive Housing Percent Change in Permanent Housing from 2005 - 2014 HUD's Housing Inventory Count Report Continuous Substantive Action: Engaging Mainstream Services PHA Reviews a City's 10 - Year Plan PHA Homelessness Preferences Web Survey Dummy (1=Yes) Substantive Action: Engaging Mainstream Services PHA Engages Homeless PHA Homelessness Preferences Web Survey Dummy (1=Yes) Substantive Action: Engaging Mainstream Services PHA Gives Preference to Homeless PHA Homelessness Preferences Web Survey Dummy (1=Yes) Substantive Action: Housing First Model City Signed on to and Actively Participates in 100,000 Homes Campaign 100,000 Homes Campaign Dummy (1=Signed; Act ive Participation) Bureaucratic Attention City Member of CoC HUD's Provider Portal Dummy (1=Yes) Bureaucratic Attention City is Consolidated Applicant HUD's Provider Portal Dummy (1=Yes) Executive Attention Mayor - Council Form of Government ICMA and Independent Internet Searches Dummy (1=Mayor - Council) Advocacy Attention and Organizational Oversight Number of Organizations involved in a city's CoC HUD's Provider Portal Continuous State Oversight Degree to which State is Involved in Homelessness I ssues Additive Scale of State Involvement Independent internet searches Ordinal (1 - 5, low - high) 106 Table 5.1 Concept Variable Source Measurement Community Oversight Has city been involved in homeless based litigation National Law Center on H omelessness and Poverty Dummy (1=Yes) Capacity: Human Resources Number of Municipal Employees Census of Governments State and Local Employment Survey Continuous Capacity: Human Resources Number of Housing and Community Development Employees Census of Gov ernments State and Local Employment Survey Continuous Capacity: Fiscal Resources Revenue Per Capita State and Local Finance Survey Continuous Fiscal Resources and Political Need: Incentive's Consistency with Local Policy Priorities Housing and Community Development Spending as Percent of Total Spending State and Local Finance Survey Continuous Political Need: Local Ideology City's Democratic Vote Share in Presidential Elections Dave Leip's Atlas of US Pre sidential Elections Continuous Fiscal Need: Fiscal Dependence Percent of total revenue coming from federal intergovernmental revenue State and Local Finance Survey Continuous Fiscal Need: Functional Autonomy Index of the autonom y states' grant localities Local Autonom y Index (Wolman et al 2010) Ordinal (1 - 6, low to high) Current Problem Based Need Size of Homeless Population HUD's Homeless Populations Reports Continuous Future Problem - Based Need Percent of Population Living i n Poverty US Census Continuous Control: Population Population Size US Census Continuous Control: Local Education Percent of Population with a Bachelor's Degree US Census Continuous Control: Racial Makeup Percent of Population that is White US Censu s Continuous 107 Finally, need is represented by five measures. Political need is measured by democratic vote share; fiscal need is measured by federal intergovernmental revenue per capita and functional autonomy (measured from low autonomy to high autonomy o n a scale of 1 - 6); problem - based need is measured as the size of the homeless population and the percent of the white) and education level (percent with BA) ar e controlled for. Together these variables allow for hypothesis evaluation. The analysis below uses linear regression to evaluate the ratio of beds to homeless individuals and the change in permanent supportive housing, and uses logistic regression to eva luate PHA behavior and involvement in the 100,000 Homes Campaign. All analyses use robust standard errors clustered by state to account for correlated errors. Evaluating Expectations Regression results for the number of beds/homeless individual and change in permanent housing from 2005 - 2014 are detailed in Table 5. 2, those for PHA behaviors are detailed in Table 5. 3, and those for 100,000 Homes participation are detailed in Table 5. 4. Results from the logistic regressions are discussed in terms of predicte d probabilities that are detailed in Table 5 .5 . 108 Table 5. 2. Linear Regression: Predictors of Beds and Permanent Housing Available to the Homeless Beds/Homeless % Change in Permanent Bed N= 255 F (19,39)= 18.97 N= 255 F (19,39)= 12.43 R= 0.315 Prob>F=0.00 R= 0.078 Prob>F=0.00 Covariate Robust SE Robust SE City has 10 - Year Plan to End Homelessness 0.250* 0.124 0.051 0.425 Housing Authority in CoC - 0.170 0.167 0.072 0.334 City is Consolidated Applicant 0.135 0.127 - 0.413 0.479 Number CoC Organizations 0.009** 0.003 - 0.028* 0.013 Total Municipal Employment - 0.000 0.000 - 0.000 0.000 Municipal Housing & Comm Dev Employment - 0.000 0.001 0.001 0.001 Total Revenue per Capita - 0.069 0.063 0.225 0.172 Housing as Percent Total Expenditures 1.254 1.672 0.793 1.971 Federal Intergovernmental Revenue/Capita 0.655 0.895 - 2.219 2.083 Homeless Population - 0.001** 0.000 - 0.000 0.000 Functional Autonomy 0.012 0.093 0.11 0.425 Mayor 0.256* 0.117 - 0.180 0.38 City Involv ed in Homeless - Based Litigation 0.125 0.149 - 0.049 0.30 State's Homeless Climate - 0.104 0.067 0.017 0.08 Percent Democrat 0.015 0.179 - 0.365 0.419 Percent Poverty 0.409 1.286 0.841 3.018 Population 0.000000675* 0.00 0.0000022* 0.00 Percent White 0 .63 0.436 0.911 1.222 Percent Bachelors - 1.911* 0.926 - 2.781 2.505 Homeless Population x Revenue 0.001* 0.000 Constant 2.026** 0.813 1.345 1.662 **sig>=.01 *sig>=.05, two tailed test 109 Table 5. 3. Logisti c Regression: Predictors of Municipal Housing Authorities' Treatment of the Homeless PHA Reviewed 10 - Year Plan PHA Engages the Homeless N= 255 Wald Chi 2 =303.2 N= 255 Wald Chi 2 =61.21 Prob>Chi 2 =0.00 Psuedo R 2 = 0.208 Prob>Chi 2 =0.00 Psuedo R 2 = 0.114 Covariate Robust SE Robust SE City has 10 - Year Plan to End Homelessness 0.490 0.355 0.143 0.332 Housing Authority in CoC 2.025** 0.468 - 0.401 0.509 City is Consolidated Applicant 1.367** 0.552 0.653 0.454 Number CoC Organizations 0.009** 0.003 - 0.025** 0.0 09 Total Municipal Employment 0.000 0.000 - 0.000 0.000 Municipal Housing & Comm Dev Employment - 0.001 0.002 0.001 0.001 Total Revenue per Capita - 0.013 0.257 0.141 0.177 Housing as Percent Total Expenditures 13.161* 6.365 7.45* 3.305 Federal Intergov ernmental Revenue/Capita - 4.787 3.062 0.337 3.021 Homeless Population - 0.000 0.000 0.001 0.001 Functional Autonomy 0.009 0.266 - 0.126 0.174 Mayor 0.388 0.457 0.196 0.370 City Involved in Homeless - Based Litigation 0.055 0.486 0.486 0.431 State's Home less Climate 0.178 0.132 - 0.204* 0.084 Percent Democrat 0.399 0.468 0.731 0.448 Percent Poverty - 4.205 4.181 7.294 4.385 Population 0.000 0.000 0.000 0.000 Percent White 0.141 1.348 2.822* 1.254 Percent Bachelors - 3.502 2.964 2.623 4.385 Constant - 2.787 2.357 - 4.563* 2.378 **sig>=.01 *sig>=.05, two tailed test 110 Table 5. 3 PHA Gives Preference to Homeless N= 255 Wald Chi 2 =54.53 Prob>Chi 2 =0.00 Psuedo R 2 = 0.113 Covariate Robust SE City has 10 - Year Plan to End Homelessness - 0.162 0.37 Housing Authority in CoC - 0.499 0.341 City is Consolidated Applicant 0.256 0.417 Number CoC Organizations - 0.029** 0.011 Total Municipal Employment 0.000 0.000 Municipal Housing & Co mm Dev Employment - 0.001 0.002 Total Revenue per Capita 0.130 0.171 Housing as Percent Total Expenditures 7.709* 3.427 Federal Intergovernmental Revenue/Capita - 1.928 2.615 Homeless Population - 0.000 0.000 Functional Autonomy 0.725 0.891 Mayor 0.66 6* 0.317 City Involved in Homeless - Based Litigation 0.929* 0.045 State's Homeless Climate - 0.281** 0.100 Percent Democrat 0.695 0.462 Percent Poverty 3.101 4.831 Population 0.00 0.00 Percent White 2.132 1.278 Percent Bachelors - 0.100 3.691 Con stant - 2.787 2.357 **sig>=.01 *sig>=.05, two tailed test 111 Table 5.4. Logistic Regression: Predictors of Municipalities Joining and Actively Participating in 100,000 Homes Campaign Join 100,000 Homes Campaign Active in 100,000 Homes Campaign N= 255 Wald Chi 2 =109.63 N= 255 Wald Chi 2 =109.63 Prob>Chi 2 =0.00 Psuedo R 2 = 0.322 Prob>Chi 2 =0.00 Psuedo R 2 = 0.322 Covariate Robust SE Robust SE City has 10 - Year Plan to End Homelessness 0.043 0.49 - 0.085 0.777 Housing Authority in CoC 0.100 0.473 - 0.177 0.469 City is Consolidated Applicant - 0.309 0.527 - 0.577 0.502 Number CoC Organizations 0.017 0.014 - 0.009 0.015 To tal Municipal Employment 0.001* 0.000 0.001* 0.000 Municipal Housing & Comm Dev Employment - 0.001 0.002 0.003 0.001 Total Revenue per Capita 0.081 0.238 0.164 0.274 Housing as Percent Total Expenditures 1.755 4.894 6.486 4.138 Federal Intergovernmenta l Revenue/Capita 3.494 2.932 - 0.621 3.150 Homeless Population 0.001** 0.000 0.005* 0.000 Functional Autonomy - 0.247 0.296 - 0.316 0.308 Mayor - 0.692 0.618 - 0.058 0.368 City Involved in Homeless - Based Litigation 0.126 0.688 1.208* 0.622 State's Homele ss Climate - 0.072 0.165 0.064 0.182 Percent Democrat - 0.233 0.564 - 0.379 0.740 Percent Poverty 7.231 4.776 9.736** 3.200 Population 0.000 0.000 0.000 0.000 Percent White 0.614 1.813 1.535 1.618 Percent Bachelors 7.724* 3.262 2.529 4.730 Constant - 5.834* 2.602 - 6.721 1.850 **sig>=.01 *sig>=.05, two tailed test 112 Table 5. 5. Predicted Probabilities for Binary Measures of Substantive Practices Predictor Min/No Mean Max /Yes Probability Housing Authority Reviews 10 - Year Plan Municipal Housing Expenditures 0.49 0.61 0.98 City is the CoC Consolidated Applicant 0.54 0.77 Number of CoC Organizations 0.47 0.57 0.80 Housing Authority is Part of the CoC 0.29 0.68 Probability Housing Authority Engages the Hom eless Municipal Housing Expenditures 0.53 0.60 0.98 Number of CoC Organizations 0.70 0.60 0.22 Percent of Municipality that is White 0.27 0.62 0.78 State Homelessness Environment 0.69 0.59 0.53 Probability PHA Gives Preference to Homeless M unicipal Housing Expenditures 0.38 0.46 0.96 Number of CoC Organizations 0.60 0.48 0.11 State Homelessness Environment 0.61 0.45 0.31 Municipality has Mayor - Council Form of Government 0.37 0.51 City has Been Involved in Homelessness Litigation 0.42 0.61 Probability Municipality Joins 100,000 Homes Campaign Municipal Employees 0.28 0.42 0.84 Size of Homeless Population 0.28 0.42 0.99 0.31 0.46 0.80 Probability Municipality is Active in 100,000 Homes Campaign Municipal Employees 0.1 0.15 0.9 Size of Homeless Population 0.07 0.14 0.99 City has Been Involved in Homelessness Litigation 0.12 0.29 113 The results provide mixed support for the hypotheses outlined above. Hypothesis 1 suggests that communities in states that are more involved in homelessness are more likely to develop substantive policies because of increased monitoring from state government. State involvement in homelessness is significant only for the more resource - intensive public housing practices - engaging the homeless and giving the homeless housing preferences. However, its effect is negative, indicating that cities in more involved states are actually less likely to engage in these practices. These cities are 16 percent less likely to have PHAs that engage the homeless and 19 percent less likely to have PHAs that give preference to the homeless than their counterparts in states that are less involved in homelessness. Hypothesis 2 suggests that cities with larger CoC netw orks are more likely to engage in substantive practices because of increased oversight, attention and advocacy. Though the size of the CoC network is significant for five of the seven practices, its influence is mixed. On the positive side, larger CoC netw orks are correlated with higher ratios of beds to homeless individuals, increasing the ratio by 9/10 of one percent for every additional organization. They also increase the likelihood that cities review their 10 - year plan, with the largest networks being 33 percent more likely to do so than the smallest ones. On the negative side however, cities with the largest networks are roughly 50 percent less likely to both engage the homeless and give them preference to public housing. Further, cities with the large st networks have significantly lower rates of change in permanent housing, with a nearly three percent reduction for each additional organization. This last result however should be taken with a grain of salt since communities with large CoC networks are l ikely to have had more permanent housing available to begin with and thus have less room for change. 114 Hypothesis 3 suggests that cities that have experienced homeless - based litigation will be more likely to have adopted substantive practices because these cases send signals to cities that their actions are being monitored. This hypothesis is supported in the case of PHAs giving the homeless preference and cities actively participating in the 100,000 Homes Campaign. Having been part of litigation increases t he likelihood that the PHA gives the homeless preferences by 19 percent and increases the likelihood of actively participating in the 100,000 Homes Campaign by 17 percent. Hypothesis 4 suggests that cities with council - mayor forms of government will be mo re hypothesis is supported for PHAs giving the homeless preferences and beds per homeless individual. Cities with mayor - council governments are 14 percent more likely to have their PHAs give preference to the homeless and on average increase the ratio of beds to homeless by 25 percent. Hypothesis 5 suggests that cities will be more likely to engage in substantive practices when the goals of the 10 year plan are consiste nt with local priorities. Local priorities were measured using the democratic vote share and spending on housing and economic development. Local ideology does not predict substantive action of any kind. In contrast, increased spending on housing and econom ic development increases the likelihood that PHAs engage in all three substantive practices. Communities with the highest housing and community development expenditures are 40 percent more likely to review 10 - year plans, 45 percent more likely to engage th e homeless and nearly 60 percent more likely to give preference to the homeless than communities with the lowest expenditures. 115 Finally, Hypothesis 6 suggests that communities with greater fiscal and human capacity will be more likely to engage in substant ive practices. There is some support for this hypothesis. Housing and economic development spending reflects not just community priorities but also fiscal capacity and, as was just established, increased spending also increases the likelihood that PHAs eng age in substantive behavior. In addition, cities with the most municipal employees are 56 percent more likely to have joined the 100,000 Homes Campaign and 80 percent more likely to be active in it than cities with the smallest municipal workforce. Importa ntly, 10 - year plan adoption and problem severity play limited roles in predicting substantive behavior. Plan adoption is significant only in predicting the ratio of beds to homeless individuals. Cities with 10 year plans have about 25 percent more beds rel ative to the size of the homeless population than those without a plan. The size of the homeless population has mixed effects and influences only 100,000 Homes participation and the ratio of beds available to the homeless population. Cities with the larges t populations are about 70 percent more likely to join the 100,000 homes campaign and over 90 percent more likely to actively participate in the campaign than cities with the smallest homeless populations. In contrast, having large homeless populations dec reases the ratio of beds to homeless by about two percent for every 1,000 additional homeless individuals, suggesting that communities have a harder time accommodating large populations. Interestingly, however, the interaction between the size of the homel ess population and revenue per capita suggests that wealthier communities are better able to serve large homeless populations, mitigating the negative effect that large homeless populations have in poorer areas. Wealthier communities with large homeless po pulations can actually increase the ratio of beds to homeless. 116 Discussion and Conclusion When the federal government began promoting 10 - year plans nearly 15 years ago, officials assumed that planning would lead cities to change the ways they dealt wit h homelessness. In some places, major changes have taken place, with cities developing new permanent housing, increasing outreach and prevention efforts, engaging public housing authorities, developing coalitions, and much more. However, this chapter has d emonstrated that 10 - year plan adoption has not driven these changes. Compared to cities without plans, communities that have adopted 10 - year plans have created no more permanent housing and are no more likely to have PHAs involved in homelessness or to hav e signed on to the 100,000 Homes Campaign. In combination with past content analysis showing the hollow nature of the majority of 10 - year plans (Shifting 2009), these findings suggest that 10 - year plans have largely been symbolic, perhaps giving homage to noble goals , but most certainly serving to legitimize local governments in the eyes of constituents and the federal government as well as helping them attract federal funding. Yet, the expectations about the drivers of symbolic policy derived from the lite rature fall short of explaining when cities move beyond this symbolic document to engage in substantive action. In fact, beyond the insignificance of ten year plans, there are few sweeping statements to be made about the drivers of substantive action. Th e exception to this, is that as expected, capacity and attention appear to be an important determinant of substantive policies, with political, fiscal and problem - based need taking a back seat. Nonetheless, the cynical view of the findings presented above is that some things matter some of the time. However, there are a couple of lessons to be taken from these results. First, what encourages cities to engage in a particular practice depends on whether it requires independent action or coalitional behavior. 117 Most importantly, local capacity is almost exclusively influential for those practices that cities engage in independently and play little role in promoting activities that rely on partnerships with community organizations. Cities with more resources are b etter equipped to act independently and carry out federal imperatives to engage mainstream services in the fight against homelessness. Given this, the federal government would do well to provide additional resources and/or technical assistance to help empo wer local governments to serve as leaders in this policy area. Second, the CoC network plays a more complicated role than one might expect. While service providers can be policy advocates and draw additional attention to problems, the results suggest that this is not always the case. Large CoC networks help cities provide sufficient beds for the homeless, presumably because there are more non - governmental resources being devoted to developing housing solutions. They also encourage public housing authorities to engage in the do have some advocacy dimensions, perhaps increasing contact with local government or petitioning local government participation. But at t he same time, large networks make cities less likely to engage in the more meaningful practices of developing permanent housing, and having their housing authorities engage with and give preference to the homeless. These findings suggest that large CoC net works may help one another meet the needs of the homeless and may encourage local government to engage in low cost activities but may actually make local government feel their participation is less necessary. Put another way, when there are many service pr oviders and advocates working on homeless issues, local governments may feel it is less pressing for them to get involved in meaningful ways. This suggests that there needs be 118 additional work in educating local governments on their role in homeless advocac y networks and engaging them as meaningful partners. Finally, the results suggest both the complexity and limitations of assessing oversight. The expectations developed above worked from the assumption that when states increase their role in homelessness they also increase monitoring of local actions; however, the results suggest this assumption may be incorrect. As with CoC networks, cities in active states are far less likely to have the ir PHA s engage with and prioritize the homeless. This suggests that rather than experiencing a sense of obligation or responsibility to meet state expectations, cities may well see an active state as an opportunity to defer action rather than engage in their own practices. The combination of this result and the finding tha t large CoC networks decrease the likelihood of independent municipal behavior is troubling. The federal government and advocates alike recognize that if there is any chance of ending or even significantly reducing the homeless population, actors at all le vels must pull their weight. This means that local governments need to recognize their role and fulfill their functions even when other strong actors are working on homelessness. In other words, local governments need to overcome the tendency to act in cri sis and instead focus on sustaining attention and substantive practices. Finally, to this end, one of the more promising results is the role mayor - council governments play in improving bed to homeless ratios and increasing the likelihood of PHAs giving ho meless preferences. Form of government is an exceptionally crude proxy for policy entrepreneurship and is one that future iterations should improve upon. Nonetheless, given the evidence that cities need a push to move beyond symbolic measures, these result s offer hope that mayoral entrepreneurship may be what is needed to focus and sustain attention in ways that lead to meaningful action. 119 The next and final chapter explores these and other implications in more depth. In it, I discuss whether federal incen tives to develop 10 - year plans have been successful in light of the fact that they have largely been symbolic. I also draw connections between the findings in Chapters 4 and 5, evaluate the usefulness of the theoretical framework laid out in Chapter 2, dis cuss the limitations of the study, and present recommendations for future work. 120 CHAPTER 6. Conclusion On June 4, 2014 the Obama administration announced the Mayors Challenge to End Veteran Homelessness. I n a written statement urging cities to join the effort , t he administration discussed the complex and inter - dependent relationship between federal and local efforts to end homelessness, writing: something that requires the co ntinued leadership and support from mayors and other local elected officials. By publicly committing to ending Veteran homelessness in 2015, enabling them to build momentum to reach the finish line. Moreover, we have seen the incredibly powerful impact that national recognition has made on the efforts in communities like Phoenix and Salt Lake City, thanks to the leadership and support of their mayors toward the goal of ending Veteran homelessness (Mayors 2014b). Though its focus is narrower, the new, voluntary program mimics HUDs earlier calls for cities to develop 10 - year plans to end homelessness. Both programs rely on cooperative federalism where cities voluntarily carry out nationally - defined goals and in turn receive a variety of benefits ranging from monetary awards to political legitimacy . In both instances, numerous cities, including those that have limited homeless problems, have participated; over 300 cities develop ed 10 - year plans (City 2011) and over 500 have signed on to the Mayors Challenge (About 2014). Additionally, both programs are built upon relatively symbolic a cts that have minimal oversight, in the first case developing long - term pla nning efforts and in t he second signing on to a non - binding challenge. In both cases, the federal government has tangible goals that require cities to move beyond these symbolic actions and implement substantive policies if the goal is to be achieved. And, in both cases the fed eral government has provided both policy - based and fiscal incentives to encourage cooperation: 10 - 121 scores for competitive Continuum of Care Funding, while signing on to the challenge makes cities eligible for increase d resources for veteran - specific homeless services (About 2014). 15 Together, these two initiatives suggest a pattern: the federal government appears to be relying on cooperation and voluntary action to spur change and innovation in homelessness policy. At least for the moment, federal officials seem to be eschewing the mandates and regulations they have used to force change in other policy areas. This pattern of federal behavior makes the findings in the previous chapters all the more relevant, as they sugg est the power and limitations of this approach and indicate opportunities for improvements in future programs. Implications of Findings The first and most important conclusion to be drawn from Chapters 4 and 5 is that federal incentives have been extrem ely effective at promoting plan adoption, but plan adoption is not a significant predictor of the substantive policies shown to reduce homelessness. While over 300 cities developed 10 - year plans in response to federal incentives, these cities are no more l ikely to develop permanent supportive housing, use their public housing authorities to provide outreach and services to the homeless, or adopt a housing - first model. This finding calls into question whether the federal government has incentivized the right type of municipal behavior, whether symbolic acts are likely to help achieve tangible goals and ultimately whether the approach of promoting 10 - year plans has been successful at moving towards the federal goal of ending homelessness. 15 T homeless veteran health care in 2014 (About 2014). 122 These questions are complex. On its face, an incentive that fails to promote substantive policy change may seem like a failure; but, in the case of 10 - year plans this view is overly simplistic. While 10 - year plans are not directly correlated with the substantive policies tes ted here, the process has changed the wider conversation about homelessness. Not just in the numerous communities that have developed plans but in cities throughout the nation conversations about homelessness now include discussions of comprehensive care a nd prevention efforts, things that were on the periphery prior to the 10 - year plan movement. In fact compassionate care has become the industry norm, rather than a novel idea. It is perhaps for this reason that from 2005 to 2014 the homeless population dec lined by over 175,000 people (from The decline in homelessness has happened in spite of the most serious recession since the Great Depression, including record foreclosure rates and high unemployment. Even at the height of the foreclosure crisis, homelessness remained relatively constant or even decreased in many communities ( ). In part this success reflects changes in the ways that both cities and the federal government talk and think ab out homelessness. Because governments were already focused on homelessness and were talking about it in terms of compassionate care and homelessness prevention, they could respond to the foreclosure crisis with large - scale homelessness prevention and rehou sing efforts. Further, planning has helped sustain national attention on the issue. While there is still a large homeless population in the country, even the most ardent supporters acknowledge that the goal of ending homelessness was never really achievab le. Instead, 10 - year plan proponents argue - Year Plans] successfully focused this country's limited 123 sustain attention over time . Denver Mayor John Hickenlooper explained it this way: "No one is more disappointed [that the city did not end homelessness in 10 years] than I am -- or all the other mayors, right?...There were 280 cities across the United States that committed to 10 - yea r -- and we always knew that we weren't going to end homelessness. Right? That was a marketing effort to get everyone's attention to say, 'Alright, Let's really work on this," (De Yoanna 2015). This marketing effort has promoted a common problem def inition and has helped keep homelessness on the agenda for over 14 years. It has also provided journalists and the public with a way to evaluate and reflect on local homelessness policy. There are articles in newspapers across the country titled some varia - while it is possible that this type of attention to homeless policy would have happened without 10 - year plans, it seems unlikely. Further, efforts like the Mayors Challenge reflect the ways that thi s energy and attention have been harnessed towards narrower, more achievable goals and And so to return to the question of whether federal incentives were successful in spite of the fact that they promoted largely symbolic plans, I believe the answer is yes. Incentives encouraged plan diffusion and while local plans have largely been weak in terms of their content, they have promoted intangibles that are difficult to mea sure and even more difficult to place a value on. The 10 - year plan movement changed the conversation around homelessness, focused attention and provided opportunities to promote more achievable goals. These are big things, important things. Nonetheless, th e fact that not all communities responded to the incentive and that the plans did not lead to tangible practices offers important insight into the incentive process. 124 The second conclusion is that different characteristics predict symbolic and substantive responses to federal incentives. The theoretical framework developed in Chapter 2 combined major insights from the policy literature to suggest that variation in attention, capacity, and need explain differences in municipal responses to federal incentives . This framework is useful, but is not as nuanced as it should be, since the two types of responses are predicted by different aspects of the framework. The largely symbolic act of developing 10 - year plans requires limited attention and capacity and is ins tead driven by need. More specifically, cities are influenced by the need to develop policy that is consistent with local ideology, respond to more severe problems, and be responsive to higher level governments that influence fiscal stability and legal leg itimacy. In contrast, substantive practices require sustained attention and both fiscal and human resources. Need, be it political or problem - based, plays a much more limited role in whether cities engage in substantive behaviors. These findings have four implications. First and most importantly, given that incentives can change municipal behavior and that substantive practices are a reflection of attention and capacity, the federal government should incentivize behaviors that encourage long - term attention and resources. Although 10 - year plans had a financial incentive that in theory could increase resources available to a community, the behavior that was being incentivized - developing a 10 - year plan - largely tapped into the need for legitimization and allo wed many of these plans to essentially be symbolic. In contrast, the federal government could incentivize behaviors that would increase the resources and attention local governments devote to the issue. For example, HUD might provide matching funds to fina nce a full time staff member focused on homelessness issues. By doing this, they would increase local human resources available to tackle the issue and also increase the likelihood that local government was focused on homelessness over the long 125 term. The p ower of increased resources and t he importance of these types of incentives in pushing cities to reorient their approach to homelessness is highlighted in introductory remarks made by Salt Lake City Mayor Ralph Becker during the announcement of the Mayors C hallenge. He explained that The big difference between what we have been doing to meet the goal of eliminating and the great success for our valued veterans has been the additional resources the VA and HUD brought to the ta ble to join us in a singular effort to end veterans homelessness. They brought increased dollars increased staffing capacity and the same valued approach to partnerships that we had developed in Salt ( First 2014 ). This statement suggests that the federal government would be able to move closer to meeting its goals if it incentivized actions that increased capacity and local attention. Second, the finding that cities with large non - profit coalitions and stronger state involvement in homelessne ss are less likely to engage their housing authorities in meaningful ways suggests that significant work needs to be done to clarify the role of municipalities in homelessness policy. Although HUD and other federal entities have repeatedly expressed the im portance of local governments in the fight against homelessness, they have provided limited was that it introduced the continuum of care and encouraged coalit ion development. There are many resources for developing these coalitions but they have no required or even defined role for local government. At least in part the variation that we see in the way that cities have responded to federal imperatives to develo p 10 - year plans and end homelessness reflects the fact that cities do not necessarily know where they fit into the equation. While some have stepped into leadership roles, others are perfectly willing to defer to state entities and service providers rather than actively fighting the problem. Given the importance that HUD has placed on engaging mainstream services, developing compassionate policy and having all influential actors involved 126 in the process, the lack of a clearly defined role for local governmen t is a major problem. Though diversity in the way that localities behave is a hallmark of the federal system, a national goal requires national guidance and as the federal government tries to end homelessness nationally it should provide additional guidanc e about the steps that local governments can(should) take to achieve this goal. Third, the fact that municipalities that are heavily dependent on state and federal governments are more likely to develop a plan but no more likely to carry out substantive p olicy suggests that municipal dependence may force a response but that it does not necessarily bring the country any closer to its goals. Cities may well feel pressure to respond to federal incentives because they are dependent units, but this pressure app ears to encourage cities to do the bare minimum rather than motivating them to engage in substantive practices. How fiscal and legal dependence influences policy and the extent to which it inhibits meaningful engagement with state and federal governments a nd incentives should be a topic of future research. Finally, the fact that form of government has varied influences presents an interesting opportunity for additional inquiry. While mayor - council cities were more likely to give preference to the homeless in public housing decisions and had higher ratios of beds to homeless individuals, they were less likely to have developed a 10 year plan to end homelessness. Given the evidence that mayors have played an entrepreneurial role, this finding is curious. Is it that council - manager governments are more strategic, pursuing low cost activities with potentially significant payoffs but not pursuing more meaningful practices? Are mayors truly focused on the substantive changes required to meet the goal and less foc used on the symbolic act of developing a 10 - year plan? What is driving this peculiar difference? This puzzle should be investigated in case study work. 127 The third and final conclusion is that federal incentives have been a powerful way to change norms and encourage plan adoption while maintaining local sovereignty; but , the voluntary nature of both plan adoption and content has meant that it is difficult for the federal government to monitor whether plans are thorough and lead to substantive practice. Thou gh t he heart of cooperation is its voluntary nature, there needs to be additional thought about the tradeoffs that exist between voluntary cooperation and third party monitoring. External oversight has been shown to be important in promoting substantive pr actice s with in private organizations, but voluntary, cooperative behavior makes monitoring difficult . Nonetheless, if HUD had provided some o versight of plan content and had linked incentives to evaluations of how thorough and meaningful the plans are, the y may have been more successful in promoting the substantive practice s that lead to goal achievement. Moving forward, if the federal government continues to rely on cooperative initiatives that can be largely symbolic they may do well to establish some st andards of engagement and opportunities for oversight. In sum, if we are to use the case of homelessness to better understand the incentive process more generally, there are three takeaways. First, municipalities do respond to federal incentives and th ey can be quite effective in altering municipal behavior. However, there is no guarantee that this behavior will be meaningful and so incentives should be structured in ways that limit opportunities for purely symbolic responses. Second, even well - meaning cities with significant need cannot respond to incentives with meaningful action if they are short on capacity or attention; as a consequence, the most successful incentives are likely to be those that encourage localities to increase the resources and att ention they devote to an issue. And third, incentives without oversight are limited in their ability to produce meaningful change. As a result, policy makers need to balance the effort to promote a voluntary, cooperative relationship 128 with the reality that in order to achieve tangible goals there needs to be some level of external monitoring. Limitations and Opportunities for Improvement This study has a handful of limitations that point to opportunities for future improvements. First, existing literature suggests that policy entrepreneurship and external oversight may be key dimensions of municipal behavior on homelessness. However, in this study both of these concepts are measured in crude ways. As a result, the study only hints that these characteristic s may be important. Future work should develop stronger measures that specifically capture the phenomena. For example, in the case of entrepreneurship studies should evaluate whether a mayor or other executive has actually taken a leadership role in homele ssness by evaluating things like state of the city speeches, newspaper articles, and campaign literature. Further, this work should identify whether there are municipal employees specifically devoted to homelessness issues in order to account for policy le adership within the bureaucracy. In the case of oversight, future work should develop more precise indictors of oversight at all levels. For example, scholars might contact state officials to evaluate their relationship with municipalities and the extent t o which they provide oversight on homelessness issues. Scholars might also evaluate public hearings and newspaper articles to get a better sense of public oversight. Finally, it would be useful to interview officials within cities themselves to understand the degree to which they recognize, experience, and respond to external oversight. Second, most of the empirics within the dissertation use cross - sectional analysis to evaluate a fluid and evolving response process. While in some instances this was stra tegic - for example in the second analysis in Chapter 4 I was truly interested in whether cities had ever adopted a plan and was not interested in change over time - in the analyses in Chapter 5 this was 129 largely a reflection of data limitations. For example at a single point in time and no additional data is available. Nonetheless, the fact that cities continue to alter their policies and practices regarding homelessness highlights two things. First, 10 year plans and federal incentives are likely to have both short - term and long - term impacts that may be different and should be evaluated separately. And second, future work should try to piece together data that would allow for longitudinal analysis that might shed ligh t on the ways that municipal policy and responses to incentives change over time. Third, the study does not include county participation in homelessness issues. This exclusion was done in an effort to establish a consistent sample and create an achievabl e project, but it also leads to a limited view of local responses to federal incentives and homelessness policy. In some areas of the US counties play an important role in service provision. Unless the county is part of a consolidated city - county governmen t, I do not account for this activity. Since counties may be engaged in practices that influence the cities within them, this is a major limitation. Future work should better account for the unique role that counties play in this process. Forth, the stud y takes a large - N approach that offers the advantage of generalizability but is limited in its ability to provide a nuanced understanding of local motivations, influences and practices. Important questions could be clarified through a case study approach. For example, interviewing local officials would help clarify their justifications for responding (or not responding) to federal incentives, the barriers they identify in pursuing substantive policy, the way they view the city in relation to other entities working on homelessness, and much more. Incorporating interviews with CoC organizations would allow for a more nuanced understanding of the partnerships between service providers and government organizations, the extent to which 130 CoC organizations view thei r role as advocates and activists, and the degree to which they believe they can influence local approaches. Finally, digging in to original documents, budgets, and meeting minutes would help to clarify intentions, ability and willingness to devout additio nal resources to homelessness, and general approaches to the issue. Future work should incorporate case studies of cities that represent the diversity of responses identified in this dissertation. Future Research In addition to the improvements outlin ed above, there are several interesting avenues for additional research. First, d espite the fact that many cities have responded to federal incentives by adopting more compassionate responses to homelessness, many others have continued to criminalize the h omeless. In fact, many of the criminalization has intensified in the las t decade ( Homes 2009 ). The reality is that these paths - compassionate care on the one hand and criminalization on the other - are not mu tually exclusive . n some cities there is both a push to end homelessness with more effective, comprehensive services and a willingness to criminalize those who have not yet been housed. This duality reflects the challenges of both a complicated, emotionally - charged policy issue and the complex intergovernmental system that has developed to address it. Cities are at once faced with pressure from the public to address public safety concerns, economic de velopment imperatives, and NIMBY issues , and with federal incentives to shift municipal approaches in ways that facilitate federal goals to end homelessness through compassionate care . This reality creates an excellent opportunity for research that examine s how cities balance competing demands from constituents and higher level governments, how policy develops in these complex and often strained environments, and how local governments legitimize their decisions to allow constituent or federal interests to t ake precedence. 131 Second, the Mayors Challenge to End Veteran Homelessness, described at the beginning of the chapter, offers a unique opportunity to understand the relationship between policy entrepreneurs and federal incentives. More specifically, the c hallenge is designed to activate local leadership and push mayors into entrepreneurial roles. Because of this intentional effort and the number of cities participating, the challenge offers a good test of whether entrepreneurship is a key to getting cities to carry out federal goals. Further, because the challenge is largely homelessness, it offers an important test of the power and limitations of entrepreneursh ip. Does activation of a local entrepreneur lead to substantive change or is its strength simply in its ability to get cities on board with a symbolic effort? Finally, it provides an additional test of the relationship between federal incentives and substa ntive change at the local level. Future work should investigate whether activating local entrepreneurs is an effective way to achieve policy change, the extent to which this encourages cities to respond to corollary incentives, and the power and limitation s of this approach for achieving federal goals. Finally, this dissertation makes inroads in understanding how federal incentives influenced 10 - year plan development and how plan development influenced substantive policy practices. It does not however, ad dress whether 10 year plans or the substantive practices identified have actually reduced the homeless population. While we know that homelessness has declined throughout the nation and can theorize that this outcome is the result of planning and compassio nate care, there is little empirical evidence that this is the case. Future work should evaluate how planning, policy and practice have influenced the problem itself. 132 A Final Thought We have come dangerously close to accepting the homeless situation as a p federal incentives have helped dispel this belief. No longer are American cities looking at their homeless with the idea that the best they can do is p rovide a bed to sleep in or a bowl of soup to eat. Today over 300 cities have begun the discussion of how to end homelessness and many of these have developed the policies that will bring the nation closer to this goal. 10 - year plans to end homelessness ha ve certainly not drawn a line under an age old problem but they have made it possible to dream of a future where real help is available and where those who find themselves in times of crisis are not met with jail cells or bus tickets out of the city. There is power in this hope and there is power in the idea that national goals can be translated into local action if only we are strategic in the ways that we encourage and incentivize change. 133 APPENDIX 134 APPENDIX A: Continuum of Care and Primary Cities Continuum of Care Primary City State Anchorage CoC Anchorage AK Mobile City & County/Baldwin County CoC Mobile City AL Birmingham/Jefferson, St. Clair, Shelby Counties CoC Birmingham AL Huntsville/North Alabama CoC Huntsville AL Montgomery City & Coun ty CoC Montgomery AL Tuscaloosa City & County CoC Tuscaloosa AL Fayetteville/Northwest Arkansas CoC Fayetteville AR Little Rock/Central Arkansas CoC Little Rock AR Phoenix/Mesa/Maricopa County Regional CoC Phoenix AZ Tucson/Pima County CoC Tucson AZ Yuba City, Marysville/Sutter, Yuba Counties CoC Yuba City CA Long Beach CoC Long Beach CA Oxnard CoC Oxnard CA Pasadena CoC Pasadena CA San Francisco CoC San Francisco CA Bakersfield/Kern County CoC Bakersfield CA Chico/Paradise/Butte County Coc Chic o CA Daly/San Mateo County CoC Daly City CA Napa City & County CoC Napa City CA Oakland/Alameda County CoC Oakland CA Redding/Shasta County CoC Redding CA Richmond/Contra Costa County CoC Richmond CA Riverside City & County CoC Riverside CA San Bern ardino City & County CoC San Bernardino CA San Buena Ventura/Ventura County CoC San Buena Ventura CA San Jose/Santa Clara City & County CoC San Jose CA San Luis Obispo CoC San Luis Obispo CA Santa Ana/Anaheim/Orange County CoC Santa Ana CA Santa Maria /Santa Barbara County CoC Santa Maria CA Watsonville/Santa Cruz City & County CoC Santa Cruz CA Davis/Woodland/Yolo County CoC Davis CA Fresno/Madera County CoC Fresno CA Glendale CoC Glendale CA Los Angeles City & County CoC Los Angeles CA Merced Ci ty & County CoC Merced CA Roseville/Placer County CoC Roseville CA Sacramento City & County CoC Sacramento CA Salinas/Monterey County CoC Salinas CA San Diego CoC San Diego CA 135 Santa Rosa/Petaluma/Sonoma County CoC Santa Rosa CA Stockton/San Joaquin C ounty CoC Stockton CA Turlock/Modesto/Stanislaus County CoC Turlock CA Vallejo/Solano County CoC Vallejo CA Visalia, Kings, Tulare Counties CoC Visalia CA Colorado Springs/El Paso County CoC Colorado Springs CO Metropolitan Denver Homeless Initiative Denver CO Bridgeport/Stratford/Fairfield CoC Bridgeport CT Norwich/New London City & County CoC Norwich CT New Haven CoC New Haven CT Bristol CoC Bristol CT Waterbury CoC Waterbury CT Danbury CoC Danbury CT Hartford CoC Hartford CT Middletown/Middl esex County CoC Middletown CT New Britain CoC New Britain CT Norwalk/Fairfield County CoC Norwalk CT Stamford/Greenwich CoC Stamford CT District of Columbia CoC District of Columbia DC Delaware Statewide CoC Wilmington DE Ft. Lauderdale/Broward Count y CoC Ft. Lauderdale FL Ft. Myers/Cape Coral/Lee County CoC Ft. Myers FL Miami/Dade County CoC Miami FL Palm Bay/Melbourne/Brevard County CoC Palm Bay FL West Palm Beach/Palm Beach County CoC West Palm Beach FL Daytona Beach/Daytona/Volusia, Flagler C ounties CoC Daytona Beach FL Fort Pierce/St. Lucie, Indian River, Martin Counties CoC Fort Pierce FL Gainesville/Alachua, Putnam Counties CoC Gainesville FL Jacksonville - Duval, Clay Counties CoC Jacksonville FL Lakeland CoC Lakeland FL Monroe County C oC Key West FL Ocala/Marion County CoC Ocala FL Orlando/Orange, Osceloa, Seminole Counties CoC Orlando FL Pensacola/Escambia, Santa Rosa County CoC Pensacola FL Sarasota/Bradenton/Manatee, Sarasota Counties CoC Sarasota FL St. Petersburg/Clearwater/L argo/Pinellas County CoC St. Petersburg FL Tallahassee/Leon County CoC Tallahassee FL Tampa/Hillsborough County CoC Tampa FL Atlanta/DeKalb, Fulton Counties CoC Atlanta GA Augusta CoC Augusta GA Athens/Clarke County CoC Athens GA Columbus - Muscogee/R ussell County CoC Columbus GA 136 Marietta/Cobb County CoC Marietta GA Savannah/Chatham County CoC Savannah GA Albany GA Honolulu CoC Honolulu HI Des Moines/Polk County CoC Des Moines IA Sioux City/Dakota County CoC Sioux City IA Boise/Ada County CoC Boise ID Evanston CoC Evanston IL Rockford/Winnebago, Boone Counties CoC Rockford IL Aurora/Elgin/Kane County CoC Aurora IL Dupage County CoC Naperville IL East Saint Louis/Belleville/Saint Clair County CoC East Saint Louis IL Waukegan/North Chicago/ Lake County CoC Waukegan IL Bloomington/Central Illinois CoC Bloomington IL Champaign/Urbana/Rantoul/Champaign County CoC Champaign IL Chicago CoC Chicago IL Decatur/Macon County CoC Decatur IL DeKalb City & County CoC DeKalb IL Joliet/Bolingbrook/Wi ll County CoC Joliet IL Peoria/Perkin/Fulton, Peoria, Tazewell, Woodford CoC Peoria IL Rock Island/Moline/Northwestern Illinois CoC Rock Island IL Springfield/Sangamon County CoC Springfield IL Indianapolis CoC Indianapolis IN South Bend/Mishawaka/St. Joseph County CoC South Bend IN Wichita/Sedgwick County CoC Wichita KS Topeka/Shawnee County CoC Topeka KS Kansas City/Wyandotte County CoC Kansas City KS Lawrence/Douglas County CoC Lawrence KS Overland Park/Shawnee/Johnson County CoC Overland Park KS Kentucky Balance of State CoC KY KY Lexington/Fayette County CoC Lexington KY Louisville/Jefferson County CoC Louisville KY Lake Charles/Southwestern Louisiana CoC Lake Charles LA Alexandria/Central Louisiana CoC Alexandria LA Baton Rouge CoC Bat on Rouge LA Houma - Terrebonne CoC Houma LA Lafayette/Acadiana CoC Lafayette LA Monroe/Northeast Louisiana CoC Monroe LA New Orleans/Jefferson Parish CoC New Orleans LA Shreveport/Bossier/Northwest CoC Shreveport LA Slidell/Livingston/Southeast Louisia na CoC Slidell LA Boston CoC Boston MA Brookline/Newton CoC Brookline MA 137 Cambridge CoC Cambridge MA Fall River CoC Fall River MA Gloucester/Haverhill/Salem/Essex County CoC Gloucester MA Lowell CoC Lowell MA New Bedford CoC New Bedford MA Quincy/We ymouth CoC Quincy MA Somerville CoC Somerville MA Springfield CoC Springfield MA Brockton/Plymouth City & County CoC Brockton MA Cape Cod Islands CoC Cape Cod Islands MA Framingham/Waltham CoC Framingham MA Holyoke/Franklin, Hampden, Hamshire Counti es CoC Holyoke MA Lawrence CoC Lawrence MA Lynn CoC Lynn MA Malden/Medford CoC Malden MA Pittsfield/Berkshire County CoC Pittsfield MA Worcester City & County CoC Worcester MA Baltimore City CoC Baltimore MD Howard County CoC Ellicott City MD Mont gomery County CoC Rockville MD Annapolis/Anne Arundel County CoC Annapolis MD Portland CoC Portland ME Bangor/Penobscot County CoC Bangor ME Grand Rapids/Wyoming/Kent County CoC Grand Rapids MI Lansing/East Lansing/Ingham County CoC Lansing MI Ann Ar bor/Washtenaw County CoC Ann Arbor MI Pontiac/Royal Oak/Oakland County CoC Pontiac MI Battle Creek/Calhoun County CoC Battle Creek MI Dearborn/Dearborn Heights/Westland/Wayne County CoC Dearborn MI Detroit CoC Detroit MI Flint/Genesee County CoC Flin t MI Grand Traverse, Antrim, Leelanau Counties CoC Grand Traverse MI Holland/Ottawa County CoC Holland MI Jackson City & County CoC Jackson MI Marquette, Alger Counties CoC Marquette MI Monroe County CoC Monroe City MI Norton Shores/Muskegon City & C ounty CoC Muskegon MI Portage/Kalamazoo City & County CoC Kalamazoo MI Saginaw City & County CoC Saginaw MI St. Clair Shores/Warren/Macomb County CoC St. Clair Shores MI Dakota County CoC Hastings MN Duluth/Saint Louis County CoC Duluth MN 138 Minneapoli s/Hennepin County CoC Minneapolis MN Saint Paul/Ramsey County CoC Saint Paul MN Scott, Carver Counties CoC Shakopee MN Coon Rapis/Anoka County CoC Coon Rapids MN Moorehead/West Central Minnesota CoC Moorehead MN Rochester/Southeast Minnesota CoC Roche ster MN St. Cloud/Central Minnesota CoC St. Cloud MN Springfield/Greene, Christian, Webster Counties Springfield MO St. Joseph/Andrew, Buchanan, DeKalb Counties CoC St. Joseph MO St. Louis City CoC St. Louis MO Columbia/Boone County CoC Columbia heigh ts MO Joplin/Jasper, Newton Counties CoC Joplin MO Kansas City/Independence/Lee's Summit/Jackson County CoC Kansas City MO St. Charles, Lincoln, Warren Counties CoC St. Charles MO St. Louis County CoC MO Gulf Port/Gulf Coast Regional CoC Gulf Port M S Jackson/Rankin, Madison Counties CoC Jackson MS Montana Statewide CoC Billings MT Montana Statewide CoC Missoula MT Asheville/Buncombe County CoC Asheville NC Durham City & County CoC Durham NC Winston - Salem/Forsyth County CoC Winston - Salem NC Cha pel Hill/Orange County CoC Chapel Hill NC Fayetteville/Cumberland County CoC Fayetteville NC Charlotte/Mecklenburg County CoC Charlotte NC Gastonia/Cleveland, Gaston, Lincoln Counties CoC Gastonia NC Greensboro/High Point CoC Greensboro NC Greenville/ Pitt County CoC Greenville NC Raleigh/Wake County CoC Raleigh NC Wilmington/Brunswick, New Hanover, Pender Counties CoC Wilmington NC North Dakota Statewide CoC Bismark/Mandan ND North Dakota Statewide CoC Fargo ND Omaha/Council Bluffs CoC Omaha NE L incoln CoC Lincoln NE Manchester CoC Manchester NH Nashua/Hillsborough County CoC Nashua NH Trenton/Mercer County CoC Trenton NJ Atlantic City & County CoC Atlantic City NJ Bergen County CoC Hackensack NJ Burlington County CoC Evesham Township NJ C umberland County CoC Vineland NJ 139 Elizabeth/Union County CoC Elizabeth NJ Gloucester County CoC Washington Township NJ Jersey City/Hudson/Bayonne County CoC Jersey City NJ Lakewood Township/Ocean County CoC Lakewood Township NJ Monmouth County CoC Midd letown Township NJ Morris County CoC Parsippany - Troy Hills NJ New Brunswick/Middlesex County CoC New Brunswick NJ Newark/Essex County CoC Newark NJ Paterson/Passaic County CoC Paterson NJ Camden City & County CoC Camden NJ Albuquerque CoC Albuquerq ue NM Reno/Sparks/Washoe County CoC Reno NV Las Vegas/Clark County CoC Las Vegas NV Utica/Romeo/Oneida County CoC Utica NY New York City CoC New York NY Rochester/Irondequoit/Greece/Monroe County CoC Rochester NY Syracuse/Onondaga County CoC Syracuse NY Nassau County CoC Long Beach NY Poughkeepsie/Dutchess County CoC Poughkeepsie NY Rockland County CoC New City NY Yonkers/Mount Vernon/New Rochelle/Westchester County CoC Yonkers NY Albany City & County CoC Albany NY Binghamton/Union Town/Broome County CoC Binghamton NY Buffalo/Erie County CoC Buffalo NY Elmira/Chemung County CoC Elmira NY Glen Falls/Saratoga Springs/Saratoga County CoC Saratoga Springs NY Islip/Babylon/Huntington/Suffolk County CoC Islip NY Ithaca/Tompkins County CoC Ithaca NY Jamestown/Dunkirk/Chautaugua County CoC Jamestown NY Jefferson, Lewis, St. Lawrence Counties CoC Watertown NY Newburgh/Middletown/Orange County CoC Newburgh NY Niagra Falls/Niagara County CoC Niagra Falls NY Schenectady City & County CoC Schenectad y NY Troy/Rensselaer County CoC Troy NY Youngstown/Mahoning County CoC Youngstown OH Cleveland/Cuyahoga County CoC Cleveland OH Toledo/Lucas County CoC Toledo OH Cincinnati/Hamilton County Coc Cincinnati OH Columbus/Franklin County CoC Columbus OH D ayton/Kettering/Montgomery County CoC Dayton OH Norman/Cleveland County CoC Norman OK Oklahoma City CoC Oklahoma City OK 140 Tulsa City & County/Broken Arrow CoC Tulsa OK Portland/Gresham/Multnomah County CoC Portland OR Eugene/Springfield/Lane County Co C Eugene OR Hillsboro/Beaverton/Washington County CoC Hillsboro OR Clackamas County CoC Lake Oswego OR Medford/Ashland/Jackson County CoC Medford OR Salem/Marion, Polk Counties CoC Salem OR Allentown/Northeast Pennsylvania CoC Allentown PA Philadelph ia CoC Philadelphia PA Lower Marion/Norristown/Abington/Montgomery County CoC Norristown PA Pittsburgh/McKeesport/Penn Hills/Allegheny County CoC Pittsburgh PA Upper Darby/Chester/Haverford/Delaware County CoC Chester PA Erie City & County CoC Erie PA Harrisburg/Dauphin County CoC Harrisburg PA Lancaster City & County CoC Lancaster PA Reading/Berks County CoC Reading PA Scranton/Lackawanna County CoC Scranton PA Wilkes - Barre/Hazleton/Luzerne County CoC Wilkes - Barre PA Columbia/Midlands CoC Columbi a SC Charleston/Low Country CoC Charleston SC Florence City & County/Pee Dee CoC Florence SC Greenville/Anderson/Spartanburg Upstate CoC Greenville SC Myrtle Beach/Sumter City & County CoC Myrtle Beach SC South Dakota Statewide CoC Sioux Falls and Min nehaha County SD Knoxville/Knox County CoC Knoxville TN Nashville/Davidson County CoC Nashville TN Chattanooga/Southeast Tennessee CoC Chattanooga TN Jackson/West Tennessee CoC Jackson TN Memphis/Shelby County CoC Memphis TN Morristown/Blount, Sevier , Campbell, Cocke Counties CoC Morristown TN Murfreesboro/Rutherford County CoC Murfreesboro TN Amarillo CoC Amarillo TX Longview/Marshall Area CoC Longview TX Houston/Harris County CoC Houston TX San Antonio/Bexar County CoC San Antonio TX Austin/Tr avis County CoC Austin TX Beaumont/Port Arthur/South East Texas CoC Beaumont TX Bryan/College Station/Brazos Valley CoC College Station TX Corpus Christi/Nueces County CoC Corpus Christi TX 141 Dallas City & County/Irving CoC Dallas TX Denton City & Count y CoC Denton TX El Paso City & County CoC El Paso TX Fort Worth/Arlington/Tarrant County CoC Fort Worth TX Galveston/Gulf Coast CoC Galveston TX Victoria/Dewitt, Lavaca, Conzales Counties CoC Victoria TX Waco/McLennan County CoC Waco TX Wichita Falls /Wise, Palo Pinto, Wichita, Archer Counties CoC Wichita Falls TX Salt Lake City & County CoC Salt Lake City UT Provo/Mountainland CoC Provo UT Fredericksburg/Spotsylvania, Stafford Counties CoC Fredericksburg VA Newport News/Hampton/Virginia Peninsula CoC Newport News VA Roanoke City & County/Salem CoC Roanoke VA Virginia Beach CoC Virginia Beach VA Arlington County CoC Arlington VA Fairfax County CoC Fairfax VA Loudoun County CoC Leesburg VA Prince William County CoC Manassas VA Alexandria CoC Alexandria VA Charlottesville CoC Charlottesville VA Chesapeake CoC Chesapeake VA Danville/Martinsville CoC Danville VA Harrisonburg/Rockingham County CoC Harrisonburg VA Lynchburg CoC Lynchburg VA Norfolk CoC Norfolk VA Petersburg CoC Petersburg VA Portsmouth CoC Portsmouth VA Richmond/Henrico, Chesterfield, Hanover Counties CoC Richmond VA Staunton/Waynesboro/Augusta, Highland Counties CoC Staunton VA Suffolk CoC Suffolk VA Winchester/Shenandoah, Frederick, Warren Counties CoC Winchester VA Burlington/Chittenden County CoC Burlington VT City of Spokane CoC Spokane WA Everett/Snohomish County CoC Everett WA Seattle/King County CoC Seattle WA Tacoma/Lakewood/Pierce County CoC Tacoma WA Yakima City & County CoC Yakima WA Vancouver/Clark C ounty CoC Vancouver WA Madison/Dane County CoC Madison WI Milwaukee City & City CoC Milwaukee WI 142 Racine City & County CoC Racine WI Charleston/Kanawha, Putnam, Boone, Clay Counties CoC Charleston WV Huntington/Cabell, Wayne Counties CoC Huntington WV Wheeling/Weirton Area CoC Wheeling WV Wyoming Statewide CoC Cheyenne WY 143 WORKS CITED 144 WORKS CITED 100,000 Homes Campaign. 2015. Monthly Impact Map. http://100khomes.org/see - the - impact (May 5, 2015). Aaron, Henry J. 1978. Politics and the Professors: The Great Society in Perspective . Washington, DC: Brookings Institution Press. Public Administration Review 64(1): 43 - 54. Advisory Bo ard Committee. 2015. Where the States Stand on Medicaid Expansion: 28 States, D.C. Expanding Medicaid. http://www.advisory.com/daily - briefing/resources/primers/me dicaidmap_5/11/2015 (February 11, 2015). Advisory Commission on Intergovernmental Relations. 1981. Measuring Local Discretionary Authority . Washington, DC: US Advisory Commission on Intergovernmental Relations. Advisory Commission on Intergovernmental Relations. 1993a. Local Government Autonomy: Needs for State Constitutional, Statutory and Judicial Clarification . Washington, DC: US Advisory Commission on Intergovernmental Relations. Advisory Commission on Intergovernmental Relations. 1993b. State La ws Governing Local Government Structure and Administration . Washington, DC: US Advisory Commission on Intergovernmental Relations. Albritton, Robert B. 1989. "Impacts of intergovernmental financial incentives on state welfare policymaking and interstate equity." Publius: The Journal of Federalism 19(2), 127 - 142. Allen, Mahally D., Carrie Pettus, and Donald P. Haider - Local: Specifying the Conditions for National Government Influence on State Policy - State Politic s and Policy Quarterly 4(3): 318 - 344. Social Justice 30(1): 195 - 221. Anderson, James. 2015. Public Policymaking . Stamford, CT: Cengage Learning. Anderson, N National Tax Journal 59(3): 685 - 694. 145 Anderson, Nels. 1961. The Hobo: The Sociology of the Homeless Man . Chicago: University of Chicago Press. Journal of the American Institute of Planners 35(4): 216 - 224. - Edge of Organizational Organizational Science 1(2): 177 - 194. Banfield, Edward, C. (1961). Politica l influence . Piscataway, NJ: Transaction Publishers. Los Angeles Times, 30 June 2015. http://www.latimes.com/local/california/la - me - 0630 - banks - valley - homeless - 20150630 - column.html (July 1, 2015). Journal of Business Ethics 67(2): 165 - 180. Urban Affairs Review 40(5): 575 - 589. Baumgartner, Frank. R. and Bryan D. Jones. 1993. Agendas and Instability in American Politics . Chicago: University of Chicago Press. Baxter, Ellen, and Kim Hopper. 1981. Private Lives/Public Spaces: Homeless Adults on the Streets of New York City . New York: Community Service Society. f Learning: Reconciling Theories Policy Sciences 25(3): 275 - 294. - National Alliance to End Homelessness, Programs and Policy Divi sion. http://nlihc.org/sites/default/files/2014AG - 98.pdf (September 25, 2014). Substantive Action 778. Bird, Lori, Mark Bolinger, Troy Gagliano, Ryan Wiser, Matthew Brown, and Brian Parsons. E nergy Policy 33(11): 1397 - 1407. 146 KTLA 5 News , 13 July 2015. http://ktla.com/2015/07/13/police - open - fire - on - man - allegedly - armed - with - knife - in - venice - lapd/ (July 14, 2015). KTLA 5 News , 6 May 2015. http://ktla.com/2015/05/06/lapd - officers - open - fire - on - transient - near - venice - beach/ (July 10, 2015). Organizational Science 14(6): 720 - 737. Boles, Myde, Julia A. Diley, Clyde Dent, Miriam R. Elman, Susan C. Duncan, and Donna B. nd Practices in Washington after an Preventing Chronic Disease 8(6): A129. - City Room: Blogging from the Five Burroughs. Ne w York Times , 29 July 2009. http://cityroom.blogs.nytimes.com/2009/07/29/mayor - defends - one - way - tickets - for - homeless/?_r=0 (February 12, 2015). B Western Political Quarterly 41(2): 341 - 362. Brinkley, Alan. 2011. The End of Reform: New Deal Liberalism in Recession and War . New York: Vintage. British Broadca British Broadcasting Corporation , 2 March 2015. http://www.bbc.com/world - us - canada - 31688942 (July 10, 2015). Brunori, Urban and Regional Policy and Its Effects: Volume 1, ed. Margery Austin Turner, Howard Wial, and Harold Wolman. Wa shington, DC: Brookings Institution Press. 109 - 155. State Politics and Policy Quarterly 4(1): 94 - 112. University of Wisconsin - Madison Institute for Research on Poverty Focus 8(2): 20 - 23. 147 Calavita, Kitty. 1993. Making Law: The State, the Law and Structural Contradictions , ed. William J. Charnbliss and Marjorie S. Zatz. Bloomington, IN: Indiana University Press. 229 - 260. Public Administration Review 67(2): 290 - 301. Cammisa, Anne Marie. 1995. Governments as Interest Groups: Intergovernmental Lobbying and the Federal System. Santa Barb ara: Praeger. Public Administration Review 57(1): 63 - 74. National Association of Counties Research Brief 2(1). http://celdf.org/downloads/Home%20Rule%20State%20o r%20Dillons %20Rule% 20State.pdf (August 11, 2014). - Specific Devolution of Law and Policy 34(2): 159 - 190. Collins, Brian K., and Brian J. Gerber. Journal of Public Administration Research and Theory 16(4): 613 - 632. Granted? Managing for Social Equity in Journal of Public Administration Review 68(6): 1128 - 1141. - Century Anniversary of the Commission on Intergov Public Administration Review 66(5): 663 - 676. Publius: The Journal of Federalism 37(3): 279 - 303. Conlan, Tim othy J. 2010. From New Federalism to Devolution: Twenty - Five Years of Intergovernmental Reform. Washington, DC: Brookings Institution Press. 148 The Oxford Handbook o n State and Local Finance , ed. Robert D. Ebel and John E. Petersen. Oxford: Oxford University Press. 83 - 104. Council on Homelessness. 2013. State of Florida Department of Children and Families . http://www.dcf.state.fl.us/programs/homelessness/docs/2013CouncilReport.pdf (June 14, 2015). Effects of Intergovernmental Grants: Or Wh University of Michigan Institute of Public Policy Studies. Journal of Social Distress and the Homeless 6 (4), 275 - 282. American Journal of Political Science 54(4): 906 - 920. Cresswell, Tim. 2001. The Tramp in America . Edinburgh: Reaktion Books. Current State . 2015. http://denversroadhome.org/state.php?id_cat=1 (June 25, 2015). Dahl, Robert A. 1961. "The behavioral approach in political scie nce: Epitaph for a monument to a successful protest." American Political Science Review 55(04), 763 - 772. Internal Pressure in State Environmental Policymaking, 19 89 - American Politics Research 33(5): 615 - 644. - Local Collaboration in an Era of Virginia Law Review 93(4): 959 - 1013. Northeast Ohio Coalition for the Homeless. http://www.neoch.org/cleveland - homeless - blog/2014/6/25/an - open - letter - to - city - and - county - government.html (July 1, 2015). American Federalism in Perspective , ed. Aaron Widalvsky. Boston: Little Brown. Journal of the American Planning Association 58(3): 288 - 300. Homelessness . http://www.va.go v/homeless/about_the_initiative.asp (September 20, 2014). 149 DePastino, Todd. 2003. Citizen Hobo: How a Century of Homelessness Shaped America . Chicago: University of Chicago Press. - Year Push to End Home Colorado Public Radio , 24 June 2015. http://www.cpr.org/news/story/whats - next - after - denvers - 10 - year - push - end - homelessness (July 20, 2015). Dom hoff, G. William. 1978. Who Really Rules? New Haven and Community Power Re - Examined . New Brunswick: Transaction Books. Shelterforce Online 135(1). The Nationalization of State Government, ed. Jerome J. Hanus. Lexington, MA: Lexington Books:39. Dunleavy, Patrick. 2014. Democracy, Bureaucracy and Public Choice: Economic Approaches in Political Science . New York: Rout ledge. Journal of Public Economics 66(4): 469 - 487. The Oxford Handbook of State and Local Government Finance , ed. Robert D. Ebel and John E. Petersen. Oxford: Oxford University Press. 198 - 212. Edelman, Murray Jacob. 1964. The Symbolic Use of Politics . Urbana, IL: University of Illinois Press. Einstein, Kat Urban Affairs Review . Urban Affairs Review 33(3): 308 - 325. - State Collaboration in the Nineteenth - Political Science Quarterly 79(2): 248 - 281. Elazar, Daniel J. 1987. Exploring Federalism . Tuscaloosa, AL: University of Alabama Press. Ellickson, Robert C. 19 Skid Rows, and Public - Yale Law Journal 105: 1165 - 1248. 150 Etzioni - Halvey, Eva. 2013. Bureaucracy and Democracy: Routledge Library Editions: Political Science Volume 7 . New Yo rk: Routledge. Policy Studies Journal 15(34): 211 - 229. Downtown Congregations to End Homelessness. http://dceh.org/the - history - of - homelessness - in - america - 1640s - to - present/ (August 10, 2014). Fleischmann American Review of Public Administration 21(1): 1 - 15. Explaining Differences Western Political Quarterly 45(3): 677 - 699. State and Local Government Review 22(2): 73 - 83. American Pscyhologist 46(11): 1232 - 1238. Yale Law and Policy Review 14(1): 1 - 63. Foscarinis, Maria, Kelly Cunningham - - Out Georgetown Journal on Poverty Law and Policy 6(2): 145 - 164. Frieden, Bernard J., and Marshall Kaplan. 1975. The Politics of Neglect: Urban Aid from Model Cities to Revenue Sharing . Cambridge: MIT Press. - 2012: A Fend for Yourself and Activist Form of Bott om - Publius: The Journal of Federalism 42(3): 357 - 386. Public Administration Review 41(6): 649 - 658. 151 - Democracy Writ Sma ll: The Impact of the Information, Communication and Society 14(2): 177 - 197. Journal of Urban Affairs 33(3): 267 - 287. Berkeley Planning Journal 14(1): 46 - 75. ent: A Urban Affairs Review 27(1): 145 - 154. Grogan, Collen M. 1999. "The Influence of Federal Mandates on State Medicaid and AFDC Decision - Making." Publius: The Journal of Federalism 29, 1 - 30. National Tax Journal 57(3): 559 - 570. - Year Public Admini stration Review 58(5): 379 - 387. Gunther, John, J. 1990. Federal - City Relations in the United States: The Role of the Mayors in Federal Aid to Cities . University of Delaware Press. Gusfield, Joseph R. 1963. Symbolic Crusade: Status Politics and the Americ an Temperance Movement. Urbana, IL: University of Illinois Press. Haider, Donald H. 1974. When Governments Come to Washington: Governors, Mayors, and Intergovernmental Lobbying . New York: The Free Press. for Federal Grant - American Review of Public Administration 38(4): 463 - 479. Policy Studies Journal 28(2): 353 - 364. Journal of Public Economics 22(3): 347 - 361. Hamilton, Christopher, and Donald T. Wells. 1990. Federalism, Power, and Political Economy . Englewood Cliffs, NJ: Prentice Hall. 152 Hanch - Journal of the American Planning Association 60(2): 197 - 208. The Social Welfare History Project. http://www.socialwelfarehistory.com/poor - relief - early - amer/ (August 10, 2014). Pacific Northwest Qu arterly 55(4): 157 - 169. Urban Affairs Review 40(1): 78 - 114. Multiple Streams Analysis of Canadian Public Administration 53(2): 241 - 258. http://denversroadho me.org/files/StateOfTheCityAddress2005.pdf (June 20, 2015). Journal of Criminal Law and Criminology 91(4): 1073 - 1100. Holl Los Angeles Times, 16 April 2015. http://www.latimes.com/local/lanow/la - me - ln - homeless - cao - report - 20150416 - story.html (June 30, 2015). Los Angeles Times, 30 June 2015. http://www.latimes.com/local/lanow/la - me - ln - garcetti - homeless - 20150630 - story.html (July 1, 2015). CNBC , 7 January 2015. http://www.cnbc.com/2015/01/07/60 - percent - of - americans - cant - cover - unexpected - expenses.html (March 20, 2015). Hombs, Mary Ellen. 2001. American homelessness: A reference handbook . ABC - CLIO. Hona Public Administration Review 41(5): 575 - 580. American Politi cal Science Review 104(1): 40 - 60. 1 53 Hopper, Kim. 2003. Reckoning with Homelessness . Ithaca, NY: Cornell University Press. Urban Anthropology and Studies of Cultural Systems and World Economic Development 14(1/3): 183 - 236. Environmental Res earch Letters 8(4): 1 - 7. Howard, Ella. 2013. Homeless: Poverty and Place in Urban America . Philadelphia: University of Pennsylvania Press. the Problem of Waste in Criminal Justice Policy Review 10(2): 257 - 290. - Based Policy - Making: Canadian Public Administration 52(2): 153 - 175. Hunter, Floyd. 1953. Community p ower structure: A study of decision makers . Chapel Hill: University of North Carolina Press. Institute for Children, Poverty, and Homelessness. 2015. National Survey of Programs and Services for Homeless Families: The Red, White, and Blue Book . http://www.redwhiteandbluebook.org/ (May 20, 2015). National Public Radio , 21 April 2009. http://www.npr.org/templates/story/story.php?storyId=103289221 (June 30, 2015). Jenkins - Journal of Public Policy 14(2): 175 - 203. Jones, Bryan. D., & Baumg artner, Frank R. 2005. The Politics of Attention: How Government Prioritizes Problems . Chicago: University of Chicago Press. American Politics Research 34(4): 403 - 426. Katz, Michael. 1986. In the Shadow of the Poorhouse: A Social History of Welfare in America . New York: Basic Books. - New York Times , 22 November 2003, sec. B. 154 l Politics of Renewable Energy: Project Planning, Siting Conflicts and - Order Devolution and the State Pol itics and Policy Quarterly 10(4): 341 - 367. Annals of the American Academy of Political and Social Sciences 509: 139 - 152. tegic Use of Academy of Management Journal 48(6): 1091 - 1106. Kingdon, John. W. 1984. Agendas, Alternatives, and Public Policies . Boston: Little, Brown. Krane, Da PS: Political Science & Politics 26(02), 186 - 190. of Journal of Urban Affairs 33(1): 45 - 60. Environment and Planning C: Government and P olicy 29(1): 46 - 62. Kusmer, Kenneth L. 2003. Down and Out, On the Road: The Homeless in American History . Oxford: Oxford University Press. Ladd, Helen F., & Yinger, John. 1991. America's Ailing Cities: Fiscal Health and the Design of Urban Policy . Balt imore: John Hopkins University Press. Larimer, Mary E., Daniel K. Malone, Michelle D. Garner, David C. Atkins, Bonnie Burlingham, Heather S. Lonczak, Kenneth Tanzer, Joshua Ginzler, Seema L. Clifasefi, William G. Hobson, re and Public Service Use and Costs Before and After Journal of the American Medical Association 301(13): 1349 - 1357. Lasswell, Harold D. [1930] 1986. Psychopathology and Politics . Chicago: University of Chicago Press. Leip, David. 2014. Dave Leip's Atlas of U.S. Presidential Elections . http://uselectionatlas.org (October 22, 2014). 155 Think Progress, 8 Jan uary 2015. http://thinkprogress.org/economy/2015/01/08/3609181/nola - ends - vet - homelessness/ (June 12, 2015). Leshner, Alan I. 1992. Outcasts on Main Street: Rep ort of the Federal Task Force on Homelessness on Severe Mental Illness . Washington, DC: US Department of Health and Human Services in cooperation with the Interagency Council on the Homeless and the Task Force on Homelessness and Severe Mental Illness. Li American Political Science Review 61(3): 701 - 716. (Affordable H https://votesmart.org/public - statement/63019/keynote - speech - to - the - chamb er - of - commerce - of - hawaii - affordable - housinghomelessness#.VbfYiPlVikp (July 22, 2015). Link, Bruce G., Sharon Schwartz, Robert Moore, Jo Phelan, Elmer Struening, Ann Stueve, and less People: American Journal of Community Psychology 3(4): 533 - 555. Political Science Quarterly 104(4): 625 - 648. Lowi, Theodore J., Benjamin Ginsberg, Kenneth A. Shepsle, and Stephen Ansolabehere. 2010. American Government: Power and Purpose . 11 th ed. New York: W.W. Norton & Co. 59 - 60. Lyon - ing Sense of NIMBY Poverty, Power and Community City and Society 13(2): 183 - 209. Publius: The Journal of Federalism 41(3):522 - 546. Marcus, Anthony. 2006. Where Have All the Homeless Gone? The Making and Unmaking of a Crisis . New York: Berghahn Books. Pasadena Star News , 26 December 2013, sec. News. 156 The Oxford Handbook on State and Local Finance, ed. Robert D. Ebel and John E. Petersen. Oxford: Oxford University Press. 176 - 197. Public Administration Review 67(4): 673 - 687. Departm ent of Housing and Urban Development Archives http://archives.hud.gov/remarks/martinez/speeches/endhomelessness.cfm , (March 26, 2015). http://www.den vergov/org/mayor/MayorsOffice/Newsroom/tabid/442244/mid/504977/ newsid504977/5650/Mayor - Hickenlooper - Creates - Commission - on - Homelessness/Default.aspx (June 19, 2015). re as American Journal of Sociology 83(2): 340 - 363. Miller, David Y. 2002. The Regional Governing of Metropolitan America . Boulder, CO: Westview Press. anship and State Politics and Policy Quarterly 12(1): 58 - 74. Miller, Henry. 1991. On the Fringe: The Dispossessed in America . Lexington, MA: Lexington Books. Miller, Thomas I., and Michelle A. Public Administration Review 51(6): 503 - 514. Mills, C. Wright. 1956. The Power Elite . Oxford: Oxford University Press. Milward, H. Brinton, and Keith G. Pr Journal of Public Administration Research and Theory 10(2): 359 - 380. American Journal of Political Science 41(3): 738 - 770. Mitchel - Antipode 29(3): 303 - 335. 157 American Political Science Review 63 ( 01), 111 - 126. Monkkonen, Eric H. 1984. Willing to Work: Tramps in America, 1790 - 1935 . Lincoln, NE: University of Nebraska Press. - Federal Diffusion and Policy Learning: From Enterprise Zones Publius: The Journal of Federalism 29(3): 31 - 50. . Public Budgeting and Finance 24(4): 2 - 15. National Alliance to End Homelessness. 2000. A Plan: Not a Dream How to End Homelessness in Ten Years. http://b.3cdn.net/naeh/b970364c18809d1e0c_aum6bnzb4.pdf (March 5, 2014). National Alliance to End Homelessness. 2009. A Shifting Focus: Plans to End Homelessness. http://b.3cdn.net/naeh/8fb05553670fae3f64_9km6ibuf7.pdf (July 11, 2014). National Alliance to End Homelessness. 2010b. Ten Year Pla n Database . http://www.endhomelessness.org/library/entry/ten - year - plan - database (August 15, 2012). National Alliance to End Homelessness. 2014. Mayors Challenge to End Vet erans Homelessness Map. http://www.endhomelessness.org/library/mayors - challenge - to - end - veteran - homelessness - map (May 16, 2015). National Alliance to E nd Homelessness. 2015. Housing First . http://www.endhomelessness.org/ pages/housing_first (June 30, 2015). National Alliance to End Homelessness. 2015. Legislative Updates . http://www.endhomelessness.org/pages/legislative_updates (February 5, 2015). National Coalition for the Homeless. 2004. Illegal to Be Homeless: The Criminalization of Homelessness in the Unite d States . http://www.nationalhomeless.org/publications/ crimreport2004/report.pdf (May 5, 2014). - NC H Fact Sheet #18 . National Coalition for the Homeless. 2009. Why Are People Homeless? http://www.nationalhomeless.org/factsheets/why.html (August 10, 2014). 158 National Law Center on Home lessness and Poverty. 2009. The Criminalization of Homelessness in U.S. Cities. http://www.nationalhomeless.org/publications/crimreport/CrimzReport_2009.pdf (May 1 5, 2014). National Law Center on Homelessness and Poverty. 2013a. No Safe Place: Advocacy Manual . http://www.nlchp.org/documents/No_Safe_Place (May 15, 2014). National Law Center on Homelessn ess and Poverty. 2013b. No Safe Place: The Criminalization of Homelessness in U.S. Cities . http://www.nlchp.org/documents/No_Safe_Place (May 15, 2014). National League of Cities. 2015. Number o f Municipal Governments and Population Distribution . http://www.nlc.org/build - skills - and - networks/r esources/cities - 101/city - structures/number - of - municipal - governments - and - population - distribution (May 1, 2015). National Low Income Housing Coalition. 2007. The Continuum of Care Plan and the Ten Year Plan to End Chronic Homelessness: Potential Tools in the Affordable Housing Preservation Toolbox . http://www.somervillema.gov/sites/default/files/2007cocmanual.pdf (March 28, 2015). Beautiful? Is Big Really So Ugly? Size, Political Studies 30(2): 190 - 206. Niskanen, William A. 1974. Bureaucracy and Representative Government . New Brunswick: Transaction Publishers. Urban Problems and Community Development, ed. Ronald F. Ferguson and William T. Dickens. Washington, DC: Brookings Institution Press. 77 - 137. Office of Managemen t and Budget. Office of Information and Regulatory Affairs. 2012. 2012 Report to Congress on the Benefits and Costs of Federal Regulations and Unfunded Mandates on State, Local, and Tribal Entities. https://www.whitehouse.gov/sites/default/files/omb/inforeg/ 2012_cb/2012_cost_benefit_report.pdf (August 8, 2014). Academ y of Management Journal 16(1): 145 - 179. - Housing Policy Debate 11(2): 489 - 520. The Property Tax and Local Autonomy, ed. Michael E. Bell, David Brunori, and Joan M. Youngman. Cambridge, MA: Lincoln Institute of Land Policy. 243 - 284. 159 A Temple Law Review 67(2): 1259 - 1306. City 16(6): 626 - 655. unicipal Bankruptcy and the Crises of Cambridge Journal of Regions, Economy and Society 7(1): 17 - 44. Peterson, Paul E. 1981. City Limits . Chicago: University of Chicago Press. Peterson, Paul, Barry Rabe, and Kenneth Wong. Programs. Education Policy Implementation, eds. Allan Odden. Albany, NY: SUNY Press. Pew Charitable Trusts. 2012. The Local Squeeze . Washington, DC: Pew Charitable Trusts. Journal of Poverty 19(1): 1 - 19. Pierson, Paul. 1995. Dismantling the Welfare State? Reagan, Thatcher and the Politics of Retrenchment. Cambridge: Cambridge University Press. Innovation and the Role of Collaborative Planning in Environmental Policy and Governance 24(6): 377 - 390. Polsby, Nelson. W. 1980. Community power and political theory . New Haven, CT: Yale University Press. Porter, David. 1973. The Politics of Budgeting Federal Aid: Resource Mobilization by Local School Districts . Beverly Hills, CA: Sage. Publius: The Journal of Federalism 37(3): 390 - 412. Provine, Dori Tensions between Civic Membership and Enforcement in the Devolution of Immigration Punishing Immigrants: Policy, Politics and Injustice , ed. Charis E. Kubrin, Marjori e S. Zatz, and Ramiro Martinez, Jr. New York: New York University Press. 42 - 61. 160 Environment and Planning 24(9): 1317 - 1336. R Distribution, 1950 - 1986. Urban Resources 5: 7 - 14. American Political Science R eview 83(1): 193 - 213. Riverstone - Newell, Lori. 2014. Renegade Cities, Public Policy, and the Dilemmas of Federalism. Boulder: Lynne Rienner Publishers. - Huffington Post , 2 April 2012. http://www.huffingtonpost.com/joel - john - roberts/did - americas - ten - year - pla_b_1394905.html (April 25, 2015). Robinson, Scott E., and Warren S. El Policy Studies Journal 38(2): 199 - 216. Govern Journal of Business Ethics 114(1): 107 - 129. Stata Technical Bulletin STB - 13: 19 - 23. On Being Homeless: Historical Perspectives , ed. Rick Beard. New York: Museum of the City of New York. 10 - 19. rban American Review of Public Administration 19(3): 233 - 248. Planning: The Collection Action Problem and the Potential of Social Local Environment 5(2): 153 - 169. - Oriented Learning, and Policy Change in an Science Communications 8(4): 649 - 692. 161 Sabatier, Paul A., and Hank C. Jenkins - The Advocacy Coalition Framework: An Theories of Policy Process , ed. Paul A. Sabatier and Christopher Weible. Boulder, CO: Westview Press. 117 - 166. Hom Georgetown Journal on Poverty Law and Policy 13(3): 545. - Comparative Urba n Studies Project Seminar, Woodrow Wilson International Center for Scholars. Public Administration Review 64 (2), 141 - 151. Schlesinger, Jr., Arthur M. 2003. The Coming of the New Deal: 1933 - 1935. Vol. 2 of The Age of Roosevelt. New York: Mariner Books. Schweik, Susan M. 2009. The Ugly Laws: Disability in Public . New York: New York University Press. Schumaker, Paul & Kelly justice concerns of elected officials: The centrality of the floors principle in addressing urban Policy Studies Journal 41 (1), 70 - 96. Scott, W. Richard. 1995. Institutions and Organizations . Thousand Oaks, CA: Sage. Journal of Human Justice 5(1): 58 - 81. Implementation of Urban Affairs Review 47(3): 433 - 457. Shelly, Bryan. 2011. Money, Mandates, and Local Control in American Public Education . Ann Arbor, MI: University of Michigan Press. Paper prepared for 2005 Midwest Political Science Association Conference, Chicago. - Up Federalism: The Diffusion of Antismoking Policies from U.S. Cities to St American Journal of Political Science 50(4): 825 - 843. 162 American Journal of Political Science 52(4): 840 - 857. Seven Lessons for Scholars and Public Administration Review, 72(6), 788 - 796.. Journal of Public Policy 34(3): 357 - 387. Simon, Tulane Law Review 66(4): 631 - 651. Stress in Regional Science and Urban Economics 41(4): 360 - 371. Economonitor, 5 October 2011. http://www.economonitor.com/blog/2011/10/one - third - of - americans - one - paycheck - away - from - homelessness/ (March 20, 2015). Snow, David A., and Leon Anderson. 1993. Down on Their Luck: A Study of Homeless Stree t People . Berkeley: University of California Press. of Relief: Explaining State Policy Choices in the Devolution Revolution. American Journal of Political Scien ce 45(2), 378 - 395. Antipode 44(4): 1510 - 1531. Evaluation of Alternative Tax and Expenditure Limitation Policies on Kansas Local Public Budgeting and Finance 29(2): 48 - 70. Western Political Quarterly 32(4): 396 - 409. Minnesota Law Review 81: 97 - 132. Stone, Clarence N. 1989. Regime Politics: Governing Atlanta, 1946 - 1988 . Lawrence, KS: Univer sity of Kansas. 163 Journal of Urban Affairs 15(1): 1 - 28. Urban Affairs Review 40(3): 309 - 341. Large Council - Public Administration Review 59(1): 44 - 53. Swift, Tracey A., Christopher Humphrey, and Vishal Gor. British Journal of Mangement 11(1): 31 - 45. American Political Science Review 108(3) : 605 - 641. Journal of Public Administration Research and Theory. Teske, Paul, and Mark Schneider. Public Administration Review 54(4): 331 - 340. Tiebout, Charles M. 1956. "A pure theory of local expenditures." The Journal of Political Economy 64(5), 416 - 424. Tsemberis, Sam. 2010. "Housi ng First: ending homelessness, promoting recovery and reducing costs." In How to house the homeless York: Russel Sage Foundation: 37 - 56. t, Consumer Choice, and American Journal of Public Health 94(4): 651 - 656. Unger, Irwin. The Best of Intentions: The Triumphs and Failures of the Great Society under Kennedy, Johnson, and Nix on. New York: Doubleday Books. United States General Accountability Office. 2010. Recovery Act: Contract Approaches and Oversight Used by Selected Federal Agencies and States . Washington, DC: Government Printing Office. United States Interagency Counci l on Homelessness. 2010. Community Approaches to Homelessness: Supplemental Document to the Federal Strategic Plan to Prevent and End Homelessness . USICH Briefing Paper. http://usich.gov/resources/uploads/asset_library/BkgrdPap_Community.pdf (October 15, 2014). 164 United States Interagency Council on Homelessness. 2011. 10 Year Plan Update . http://w ww.homelesshub.ca/sites/default/files/rx1twghk.pdf (December 5, 2012). United States Interagency Council on Homelessness. 2014. About USICH. http://usich.gov/about_us/ (September 16, 2014). United States I nteragency Council on Homelessness. 2015. Permanent Supportive Housing. http://usich.gov/usich_resources/solutions/explore/permanent_supportive_housing (July 1, 2015). US Census Bureau. 1993. Population: 1790 - 1990. https://www.census.gov/population/ censusdata/table - 4.pdf (August 11, 2014). US Census Bureau. 2010. National Data Book . Area and Population, Table C - 1: Cities. http://www.census.gov/compendia/databooks/2010/tables/cc_C - 01.pdf (April 3, 2015). U.S. Code Title 42, Chapter 119, Subchapter I, § § 11301. US Department of Housing and Urban Development. 1994. Priority Home! The Federal Plan to Break the Cycle of Homelessness . Washington: Department of Housing and Urban Development. US Department of Housing and Urban Development. 2005a. 2005 CoC H omeless Population and Subpopulations Report CA - 600: Los Angeles City and County CoC. https://www.hudexchange.info/resource/reportmanag ement/published/CoC_PopSub_CoC_CA - 600 - 2005_CA_2005.pdf (June 25, 2015). US Department of Housing and Urban Development. 2005b. 2005 CoC Homeless Population and Subpopulations Report CO - 503: Metropolitan Denver Homeless Initiative COC. https://www.hudexchange.info/resource/reportmanagement/published/CoC_PopSub_CoC_CO - 503 - 2005_CO_2005.pdf (June 25, 2015). US Department of Housing and Urban Development. 2008. 2008 CoC Homeless Populations and Subpopulations Report All States, Territories, Puerto Rico, and DC . https://www.hudexc hange.info/resources/reportmanagement/published/CoC_PopSub_NatlDC_2 008.phf (July 20, 2015). US Department of Housing and Urban Development. 2009. 2009 CoC Homeless Populations and Subpopulations Report All States, Territories, Puerto Rico, and DC . https://www.hudexchange.info/resources/reportmanagement/published/CoC_PopSub_NatlDC_2 009.phf (July 20, 2015). 165 US Department of Housing and Urban Dev elopment. 2010. 2010 CoC Homeless Populations and Subpopulations Report All States, Territories, Puerto Rico, and DC . https://www.hudexchange.info /resources/reportmanagement/published/CoC_PopSub_NatlDC_2 010.pdf (July 20, 2015). US Department of Housing and Urban Development. 2012. PHA Homelessness Preferences: Web Census Survey Data . http://www.huduser.org/portal/datasets/pha_study.html (February 13, 2015). US Department of Housing and Urban Development. 2013. Ending Homelessness Through the Housing Choice Voucher and Public Housing Programs, Notice PIH 2013 - 2015 Webinar. Offi ce of Public and Indian Housing. Accessed May 15, 2014 at http://portal.hud.gov/hudportal/HUD?src=/program_offices/public_indian_housing/homelessnes s . US Department of Housing and Urban Development. 2014a. 2014 CoC Homeless Population and Subpopulations Report CA - 600: Los Angeles City and County CoC. https://www.hudexchange.info/resource/reportmanagement/published/CoC_PopSub_CoC_CA - 600 - 2005_CA_2014.pdf (June 25, 2015). US Department of Housing and Urban Development. 2014b. 2014 CoC Homeless Population and Subpopulations Report CO - 503: Metropolitan Denver Homeless Initiative COC. https://www.hudexchange.info/resource/reportmanagement/published/CoC_PopSub_CoC_CO - 503 - 2005_CO_2014.pdf (June 25, 2015). US Department of Housing and Urban Development. 2014c. Mayors Challenge: Frequently Asked Questions . https://www.hudexchange.info/reso urce/3795/mayors - challenge - faqs (September 20, 2014). US Department of Housing and Urban Development. 2014d. Mayors Challenge: Mayors and Staff . http://portal.hud.gov/hudportal/HUD?src=/program_offices/comm_planning/ veteran_information/mayors_challenge/mayors_and_staff (September 20, 2014). US Department of Housing and Urban Development. 2015. Mayors Challenge . http://portal.hud.gov/hudportal/HUD?src=program_offices/comm_planning/veteran_information /mayors_challenge (May 16, 2015). US Dep artment of Housing and Urban Development. HUD Exchange. 2014. COC Homeless Populations and Subpopulations Reports: 2005 - 2014 . https://www.hude xchange.info/manage - a - program/coc - homeless - populations - and - subpopulations - reports/ (May 5, 2015). 166 US Department of Housing and Urban Development. HUD Exchange. 2015. Resources for Chronic Homelessness . https://www.hudexchange.info/homelessness - assistance/resources - for - chronic - homelessness (February 12, 2015). US Department of Housing and Urban Development. Office of Community Planning and Development and Office of Special Needs Assistance Programs. 2007. Defining Chronic Homelessness: A Technical Guide for HUD Programs . US Department of Housing and Urban Development. Office of Policy Development and Research. 2002. Evaluation of Continuums of Care fo r Homeless People . http://www.huduser.org/publications/pdf/continuums_of_care.pdf (January 10, 2015). US Department of Housing and Urban Development. Office of Public and Indian Housing. 2013. PHA 2010 Notice PIH 2013 - 15: Guidance on Housing Individuals and Families Experiencing Homelessness through the Public Housing and Housing Choice Voucher Programs . http://portal.hud.gov/hudportal/documents/huddoc?id=pih2013 - 15.pdf (June 25, 2015). Impli Publius: The Journal of Federalism 29(3), 51 - 73. Wagner, David. 2005. . New York: Rowman & Littlefield Publishers. Wagner, David, and Jennifer Barton Gilman. 2012. Confronting Homelessness: Poverty, Politics and the Failure of Social Policy . Boulder, CO: Lynne Rienner Publishers. American Political Science Review 63(3): 880 - 899. Wallack, Lawrenc e. 1993. Media Advocacy and Public Health: Power for Prevention . Thousand Oaks, CA: Sage. American Journal of Political Science 24(4): 715 - 729. Westphal Administrative Science Quarterly 40(1): 60 - 83. H YouTube video, 35:28. https://www.youtube.com/watch?v=8oJnYi6XPVU (October 20, 2014). 167 El Paso Times , 15 December 2003. Urban Geography 10(1): 62 - 85. g Local Paper. Wong, Kenneth K. and Paul E. Peterson. 1986. Urban Response to Federal Program Flexibility Politics of Community Development Block Grant. Urban Affair s Review 21(3), 293 - 309. Public Administration Review 67(2): 249 - 264. Yates, Douglas. 1977. The Ungovernable City: The Politics of Urban Problems and Policy Making . Cambridge, MA: MIT Press. - Nonprofit Sector Relations: Non profit and Voluntary Sector Quarterly 29(1): 149 - 172. Political Research Quarterly 58(3):497 - 509. El Paso Times , 15 December 2003. Urban Geography 10(1): 62 - 85. Wolman, Hal, Robert McManmon, Michael Bell, and David Paper. Wong, Kenneth K. and Paul E. Peterson. 1986. Urban Response to Federal Program Flexibility Politics of Community Development B lock Grant. Urban Affairs Review 21(3), 293 - 309. Public Administration R eview 67(2): 249 - 264.