What, then, does the story we have just unfolded imply for urban politics? Or more specifically, what are the implications for urban politics New York style? To answer this question, let us go back to the beginning of modern political analysis of New York City, which is that classic of postwar American pluralism, Governing New York City, by Wallace Sayre and Herbert Kaufman. They famously wrote that “[n]o part of the city’s large and varied population is alienated from participation in the system.”1
The pluralist school of political science would come under fire from a wide range of power-elite, structuralist, and, later, regime theorists. The criticisms of pluralism were wide ranging. One was that there were, in fact, important interests that were alienated from the political process. As Bellush concludes in her updating of the Sayre and Kaufman account, “The outcome of the urban political process is not necessarily a little something for all the players.”2 Clearly, the argument goes, the pluralist system wasn’t working for some.
What does the story of New York City’s homeless over the last thirty years imply for this critique of the pluralist picture? Of course, conclusions drawn from the experience of one numerically small group cannot radically change the large picture. However, the homeless, despite being relatively few in number, are an especially significant part of that picture. As Dr. Johnson has noted, and as John Rawls has elaborated upon, the conditions of the worst off members of a community are a particularly important consideration in judging the justice of that community. And, as we have seen, the homeless, many of them, are perhaps the worst of the worst off in New York City. If the city has not addressed their needs, that is a damning reflection on the pluralists’ optimistic evaluation. On the other hand, if the homeless have gotten a fair shake, perhaps there is hope for any disadvantaged group.
New York’s record on the homeless is mixed. An evaluation of it depends on whether we focus on how far the city has come or on what the situation is now in the early years of the de Blasio administration. Very striking progress has been made since homelessness was first pushed onto the city’s political agenda in the late seventies. On the other hand, the city’s shelter population—about fifty-five thousand people in the summer of 2015—is nearly as large as it has ever been since that beginning. A fair evaluation of the city’s record has to take both of these realities into consideration.
It would be hard to find an interest group less likely to be successful in the urban political arena than the homeless. David Snow, Sarah Soule, and Daniel Cress point out the obvious when they write that
[a]t first glance the homeless . . . are unlikely candidates for protest mobilization. The homeless not only suffer severe resource deficits materially and socially, but they are more impoverished in both realms than most other marginalized individuals and groups. . . . They also rarely have ready access to the kinds of material facilities and resources that most social movement activists and organizers take for granted such as a meeting place, a telephone, and a few office supplies.
Other obstacles to the political efficacy of the homeless are also noted, including high rates of disability and the transitory nature of homelessness.3
And yet, despite this unpromising position, New York City’s homeless have been quite successful in certain ways. We have recounted the development of the Department of Homeless Services with an annual budget, recently, of about $1 billion; the demise of such hellholes as the big room of the Men’s Shelter at 8 East Third Street, the Emergency Assistance Unit, the flophouses and welfare hotels; their replacement with state-of-the-art facilities, including the PATH building and Project Renewal’s recovery program; the rise of much-improved temporary accommodations such as those operated by a network of nonprofit providers; and the development of high-quality permanent shelter for the disabled, as was provided for in the various New York/New York agreements. Despite some qualifications that we will get to shortly, the experience of the homeless of New York implies a degree of openness that is consistent with the pluralists’ account of the city’s political system and that tells against the arguments of the pluralists’ critics.
Obviously, one might disagree with this assessment for a number of reasons. Donna Wilson Kirchheimer denies that homelessness policy is a pluralist success story and concludes that the city’s efforts have all fallen “well within the historically acceptable protective functions of the partial U.S. welfare state.”4 But the question here is not whether New York’s response to the homeless has transcended the parameters of the partial American welfare state. It hasn’t, not by a long shot. The question at the moment is whether the homeless “got something” in the way that pluralists conceive of everybody getting something out of the city’s political system. If the homeless got something, that represents, perhaps not a “pluralist’s dream” but a degree of openness that is consistent with the pluralists’ picture of urban America. The pluralists never argued that the degree of openness they found allowed for breaking beyond the boundaries of the American welfare state.
How large, then, are the resources devoted to homelessness in New York City? The Department of Homeless Services is allocated $1.04 billion in the city’s 2014 expense budget of $69.8 billion, only about 1.5 percent of the total.5 By other measures the resources devoted to the homeless in New York City are more impressive. As was noted earlier, in 1978, or the year before the Callahan litigation began, the city government spent $6.8 million on the homeless,6 for an increase of more than 4,000 percent in constant dollars.7 It would be illuminating to compare New York City’s efforts with those of other cities. However, most cities do not have an analogue to New York’s DHS, which makes comparisons difficult. An exception is Los Angeles, with its Los Angeles Homeless Services Authority, which spent $72 million in FY 2014–2015.8 The findings of Daniel Cress and David Snow on “The Outcomes of Homeless Mobilization” for fifteen organizations of the homeless in eight U.S. cities are also telling. The greatest success was that of the Oakland Union of the Homeless, which achieved what the authors describe as the “stunning” victory of getting built a twenty-six-unit, $4.7 million housing project.9 But as we have seen, this outcome is small potatoes compared to New York City’s efforts. The resources New York City spends on the homeless, then, are small relative to the city’s entire budget, but large compared to the city’s past efforts and the efforts of other cities. In short, the city’s homeless, at least by some relevant measures, did indeed end up getting quite a bit. That result, at first blush, seems consistent with the pluralists’ account of city politics as a system open to a wide range of participants.
Yes, the homeless got something, but what they got, for the most part, was not what they were most desperate for, permanent housing; what they usually got was emergency shelter and some associated services. From this perspective, Kirchheimer’s criticism of the limits of the city’s homelessness policy has more force, although it needs to be reformulated. Certainly the city’s efforts have not transcended the “protective functions of the partial U.S. welfare state”;10 the real problem is that those efforts have not lived up to the accomplishments of the American welfare state at its best. New York’s shelter system, as impressive as it is in many respects, pales in comparison with LaGuardia’s achievements in slum clearance, providing affordable housing, and establishing the New York City Housing Authority (NYCHA).11 In recent decades, housing reform of such ambition has not been politically possible. New York City has had to develop its homelessness policy under the constraints of conservative dominance, neoliberal globalization, and, often, financial austerity. The result has been a shelter system, not in the grand tradition of the New Deal but one that reflects the flexibility and limitations of what has been called the “post-Fordist welfare state.”12
The homeless have indeed gotten something out of the city’s political system, and yet it cannot be said that they have been active participants in that system. In important ways they have indeed accepted passively the policies that the system has dealt out. The homeless did not make use of the organizing resources or strategies that have been effective for the other urban interest groups described by the pluralists. Kirchheimer correctly reports that
the beneficiary population was a weak political force. The homeless were the poorest people in New York City and had no financial resources for political mobilization. The poor did not have an organizational base from which to press for their own needs. They tended to be inactive in electoral politics, and their support was not heavily courted by elected officials. The homeless also did not exert their force of numbers through organized protests or street demonstrations, which have been important to New York City’s day care and senior citizen movements. About a third of homeless single persons were handicapped by major mental disabilities, and many suffered from substance abuse. Most heads of homeless families were young single women who had less than a twelfth grade education and less than a year of work experience, and they were preoccupied with caring for small children in unstable residences from which they had to move almost every other year.13
New York City’s homeless won the substantial resources they did not through their own political action but through the political action of others on their behalf. It is telling that it was the activist lawyer Robert Hayes who sought out and represented the plaintiffs to the Callahan litigation, and not the homeless who organized themselves and sought out representation. The Coalition for the Homeless has been an effective advocacy organization, but although in its earliest days homeless people were a significant part of the milieu out of which it developed, for the most part the organization has been, as its name says, a coalition for the homeless, rather than of the homeless. At no point did the coalition’s leadership or staff consist mainly of homeless people or formerly homeless people, although there was, as we have seen, some participation of the homeless in the formative days of the coalition.
The exact nature of the experts who were at the center of homelessness politics in New York City has been usefully termed by Kirchheimer as a policy community, which she described as follows:
Within this milieu, a new policy community sprang up to specialize on the issue of homelessness. It was an identifiable community of shared values, which consciously challenged the existing institutional structure. It consisted of people who were committed social advocates, not neutral technicians, and they were not based in the city bureaucracy. The internal structure of this policy community was loose and open, and contacts were ad hoc and informal. Many members knew each other, exchanged information, shared strategies, and at times coordinated particular actions.14
This picture of the homeless of the city as politically inactive needs to be qualified in one respect. The homeless did not organize, but they did occupy public spaces and may be said to have disrupted the routines of passersby who saw them. Research has shown that simply by being visible, the homeless gain sympathy and support.15 Joel Blau, in his book The Visible Poor: Homelessness in America, writes that “perhaps the single most significant attribute of homelessness is its visibility. Visible poverty disrupts the ordinary rhythms of public life. It undermines the rules governing the use of public space.”16 There is no research on how often New Yorkers see homeless people or on how that sight impacts them. However, a 2007 study by the Public Agenda Foundation found that 63 percent of New Yorkers reported that there were some or a lot of homeless people in their neighborhood and that 47 percent reported directly helping a homeless person within the past year.17 My own interest in homelessness was provoked in the early 1980s by the sight of homeless people in the subways that I used to go to work. Mayor Koch would eventually say that simply “seeing people in the street itself” forced the homelessness issue on his attention.18 Homeless people have undoubtedly been visible over the last thirty years in New York City, and that visibility was most likely part of the reason why the city devoted more resources to the homeless. Insofar as visibility disrupted ordinary routines and resulted in more sympathy and resources, it may be thought of as a political phenomenon.
The finding that the homeless did not organize but did disrupt is similar to the thesis of Frances Piven and Richard Cloward’s book, Poor People’s Movements, which claims that “[w]hatever influence lower-class groups occasionally exert in American politics does not result from organization, but from mass protest and the disruptive consequences of protest.”19 On closer examination, however, the case of the homeless in New York City does not confirm the Piven and Cloward thesis. First of all, mass protest did not play much of a role in developing the city’s homelessness policy. The Thompson Square Park riot got attention at the time, but it happened well after the Callahan and McCain litigation had begun and did not have a major impact on policy. And as we saw, the continuing disruption in the park during the early Dinkins administration led to the park’s closure and then reopening after an extensive renovation rendered it less hospitable to the homeless. There is not much evidence to suggest that the city’s homeless people considered themselves to be part of a larger movement or that they shared a set of protest beliefs. The type of disruption they mainly engaged in—inappropriately occupying public spaces—did not require any type of collective consciousness. More coordinated forms of disruption were rare. In January 1989, Parents on the Move (POM), a self-help organization of homeless parents living in the Brooklyn Arms welfare hotel, seized the offices of the city’s Crisis Intervention Services at the hotel and refused to be relocated to anywhere except permanent housing. A survey of twenty members of the group found that thirteen of them considered the organization a success in the sense of “making people feel good about themselves and their homeless situation,” but POM folded soon after the demonstration.20 The Philadelphia Union of the Homeless tried to develop a New York City branch, which “promised civil disobedience,” but according to one sympathetic observer, the group soon “dissipated” when its main organizers moved on to other cities.21 Homeward Bound Community Services (HBCS) organized a sleep-in at City Hall Park that lasted for hundreds of days, but according to a review of the history of the protest, “the encampment fell apart after a while, a victim of disorder, neglect, and the instability of its residents.”22 Robert Hayes praised the efforts of HBCS but acknowledged that “[h]omeless people organizing themselves has been rare.”23
The sheer visibility of the homeless was an important contribution to the politics of homelessness in New York City. Otherwise, however, homelessness politics in the city has mostly been made on behalf of the homeless, not by them. This reality has received very little comment. The one sustained consideration of this fact has been the passionate denunciation by Theresa Funiciello in her book Tyranny of Kindness. Funiciello, a formerly homeless mother, welfare activist, and advisor to the New York State commissioner of social services, Cesar Perales, fulminates against what she calls “the creation and marketing of homeless people,” writing that “[u]nder the rubric of helping homeless people, little empires were built, expanded and strengthened, careers were boosted, and media stars created overnight.” The problem is that “the poverty industry has become a veritable fifth estate. Acting as stand-ins for actual poor people, they mediate the politics of poverty with governmental officials.”24
Whatever one makes of Funiciello’s overwhelmingly negative caricatures of many policy actors, and her blistering invective, mediating the politics of poverty with governmental officials is a pretty fair description of the work of the city homelessness-policy community identified by Kirchheimer. What Funiciello does not succeed in doing is showing, rather than asserting, that this mediated or paternalistic mode of policymaking has been bad for the homeless. Her main argument is that the homeless would benefit immensely if all the resources devoted to their problems were simply delivered directly to them in the form of a check and/or permanent housing, rather than being routed through the public and nonprofit bureaucracies. The problems with such a guaranteed-income-and-housing scheme—which are rooted in perverse incentives, fairness, and behavioral issues—have been developed in detail elsewhere.25 What is interesting here is that implicit in Funiciello’s argument is the recognition that the resources devoted to the homeless in New York are quite considerable. It is also true that since Funiciello’s writing in 1993, most of the poor shelter conditions described in her book have been dealt with. All this is to say that the politics of dependency, with the homeless as a group mostly passively organized by others on their behalf, have been beneficial to the city’s homeless. Or again, the politics of homelessness in New York City are dependency politics, and it’s a good thing too.
As this book has demonstrated, New York City homelessness policy, as it has developed since the beginning of the Callahan litigation, has passed through three moments: a moment of entitlement, a paternalistic moment, and then a post-paternalistic moment. These moments can be summarized as follows.
As it was developed by advocates working through the courts in the early 1980s, and by the city bureaucracy reacting to that pressure, city homelessness policy began by focusing on rights and offering shelter with no quid pro quo asked of the clients. Delivering on that entitlement was the central challenge to policy during the Koch and most of the Dinkins years. The courts and various advocacy groups such as Coalition for the Homeless are primarily concerned with this aspect of homelessness policy. These interests pushed policy in the direction of developing a shelter system that was large, court supervised, and primarily concerned with delivery of emergency accommodations.
The entitlement or right to shelter completely transformed the city’s homelessness system. The system grew tremendously in the early eighties. While 7,584 individuals were sheltered in 1982, 21,154 were sheltered in 1985. Spending grew from $6.8 million in 1978, the year before the litigation to establish a right to shelter began, to $100 million in 1985. To cope with the rapidly expanding demand, the city rushed to open large, barracks-style shelters where hundreds of clients would sleep in cots laid out in open spaces. During these years the city also relied on commercial welfare hotels to shelter homeless families at the cost of $72 million in 1986. The shelter system during these years was satisfactory neither from a conservative nor from a liberal point of view. The right to shelter was absolute, and unbalanced by any requirements to work, participate in rehabilitation, or seek permanent housing.
Shelter quality was often very poor during the Koch and Dinkins years, when the focus was on establishing the right to shelter. That a right or entitlement to shelter should result, at least at first, in poor-quality shelter is not a paradox but a necessary outcome of the advocates’ decision to apply the rights premise to the problem of homelessness. If homelessness had been conceived of as an “ordinary” social problem, like poverty or mental illness, it would have been addressed legislatively, just as other social problems are. A legislative solution—such as developing more affordable housing—would likely have been an incremental solution as various interests fought to have an influence on policy. An expanded affordable housing program would thus probably have been implemented over time, with new housing coming on line over a period of some years. This approach would have left more people homeless for longer but would have avoided the implementation problems associated with suddenly providing access to shelter on a day certain.
If homelessness was to be addressed through the right to shelter, the main forum in which policy would play out would be the courts. Courts, being relatively independent of multiple political interests to satisfy and less sensitive to implementation concerns, were free to impose a nonincremental solution to homelessness, one in which the basic right to shelter was established with the signing of the Callahan consent decree on August 26, 1981. The result was a sudden crush of demand for shelter, one that the city had to meet immediately in a catch-as-catch-can fashion. Large barracks-style shelters based in armories—with all the problems of privacy, security, and hygiene they implied—were a foreseeable part of that response; there was no time to go through the politically demanding process of opening up many smaller, neighborhood-based shelters as the advocates had wanted. Use of the infamous welfare hotels expanded, as they were the supply of emergency shelter for families most available to hand. And few services were offered to clients, as the main concern was simply providing the mandated emergency shelter. Approaching the problem of homelessness as a matter of right or entitlement meant that, in the short run, the city would create a system that guaranteed the right to free, low-quality shelter.
Dinkins tried to extend this entitlement by providing shelter clients with rapid transfers into permanent housing, still without making any demands of clients. Entries into the shelter system began to go up. As we saw, by 1991, key figures in the Dinkins administration came to believe that this entirely rights-based policy had created a perverse incentive for more potential clients to enter the system. Later research by Cragg and O’Flaherty indicated that this belief was incorrect and that more placements would have decreased, not increased, the shelter population. This observation raises the tantalizing possibility that had the Dinkins administration persisted in its original course it might have come as close to solving the homelessness problem as was possible at that time.
But even if the Dinkins administration had known that it had not created a perverse incentive, political considerations suggest that it was likely that the generous distribution of permanent housing units would not have continued for long. During the early 1990s, strong opposition was building to purely rights-based, services-oriented welfare programs in general, which, at the national level, would result, by 1996, in the passage of the quite demanding Personal Responsibility and Work Opportunity Reconciliation Act. In New York City, a similar process and a similar disappointment with the city’s service-oriented welfare program, BEGIN, would result in the very work-oriented welfare reforms under Giuliani.26 In this political environment, it is inevitable that the city’s purely rights-based homelessness policy would, sooner or later, have become the object of a similar, paternalistic reform.
That the change in direction came sooner, under Dinkins, rather than waiting for the arrival of Giuliani, is probably to be attributed to the arrival of an effective local policy entrepreneur, Andrew Cuomo. As chairman of the New York City Commission on the Homeless, usually known as the Cuomo Commission, Andrew Cuomo acknowledged the concerns of Nancy Wackstein, Cesar Perales, and other Dinkins administration officials who believed that the homelessness system had become over-generous. He borrowed the policy idea of “mutual responsibility,” which had been developed in the national welfare reform debate, and ingeniously applied it in detail to city homelessness policy. He outmaneuvered traditionally liberal welfare policy actors, such as Barbra Sabol, the commissioner of the Human Resources Administration under Dinkins, and eventually sold his plan to the possibly somewhat reluctant Dinkins. The paradigm that the Cuomo Commission produced—one that insisted that homeless clients do something, such as work, search for housing, or participate in rehabilitative programs in return for shelter and services—was essentially paternalistic and would become the basis for policy under Giuliani.
Lawrence M. Mead defines paternalistic policies as “social policies aimed at the poor that attempt to reduce poverty and other social problems by directive and supervisory means.”27 Many of the homelessness policy developments since the Cuomo Commission have been consistent with the paternalistic paradigm. That paradigm rests on two assumptions. First, paternalism subscribes to the “underlying cause” theory of homelessness. It assumes that there is “something wrong” with the homeless that importantly contributes to their homelessness. This might be mental illness, substance abuse, or simply a “dysfunctional” or underclass lifestyle that, it is thought, makes housing hard to hold onto. Paternalism assumes that this underlying cause has to be addressed, otherwise the subject will remain homeless. Therefore, the second assumption of paternalism is that if clients are to receive housing, whether shelter or some form of permanent housing, they have to participate in a rehabilitative regime that will address the underlying cause of their homelessness. Such participation could involve enrollment in a substance abuse rehabilitation program, or mandatory compliance with psychiatric treatment, or, for homeless families, participation in various sorts of housing-search, work-preparation, or family-living programs.
Paternalism thus conditions the right to shelter on various “good behaviors” of the clients. Paternalism began to manifest itself as the limits of the early, entitlement-based, emergency-oriented system became apparent. An obvious problem was the poor quality of the shelter that was provided by the jerry-built system. The unconditional right to shelter also proved to be problematic. Behavioral problems, such as substance abuse, nonwork, and criminal activity, enacted by some of the homeless, seemed to require that the entitlement to shelter be conditioned on participation in work, treatment, and job-search programs. It was thought that strong conceptions of the rights of the mentally ill sometimes had to be limited in order to provide necessary protection and therapy. This line of thinking resulted in the efforts of Project HELP to involuntarily transport mentally ill street people to psychiatric services, which represented an early manifestation of the paternalistic impulse. Improving shelter conditions made more acute the possibility that a perverse incentive to become homeless in order to qualify for housing had been created that had to be counterbalanced in various ways. This set of challenges is of particular concern to mayors and administrators who, unlike the courts or advocates, are responsible for the actual operation of the shelter system, and for spending. These actors therefore pushed policy in a paternalistic direction, one in which rights are conditioned on good behavior and participation in programs such as drug-treatment, job-search, and housing-search activities.
The development, under Giuliani, of an eligibility process for homeless families was a first, necessary step in moving from a purely rights-based policy in a paternalistic direction. Bloomberg’s efforts to develop an eligibility process in the shelters for singles continued this direction. Not-for-profitization also expanded greatly during the paternalistic moment, partly because of the demonstrably better management of nonprofit shelters as compared with public shelters but also because creating a two-tier system of public intake shelters and nonprofit program shelters made it possible to require client participation in rehabilitative programs without violating the right to shelter as it was developed through the courts. And of course the program shelters themselves, focusing as they did on rehabilitation and quid pro quo, were also a paternalistic policy development. Even the improvement in shelter conditions, while obviously pursued on humanitarian and legal grounds, has reinforced the paternalistic regime, since if one is to demand a quid pro quo, it helps to offer a benefit that is worth having.
The limits of paternalism started to manifest themselves by the final years of the Giuliani administration. It was then that DHS executives began to notice that substantial numbers of shelter clients refused to take up the paternalistic deal offered by the then newly developed program shelters: admission to a better-quality shelter in return for participation in a rehabilitative regime. That deal was also being offered to street dwellers, to whom outreach teams would offer access to permanent, supported housing, but only if they demonstrated themselves to be housing ready by abstaining from substance abuse, participating in psychiatric treatment, and in general behaving themselves. A frustratingly high number of street dwellers declined this deal, making inroads against public homelessness hard to achieve. The option of involuntarily transporting street dwellers who refused treatment and shelter to hospitals, as was attempted for a time by Project HELP under Koch, turned out to be impractical. So paternalism ended up not being able to reduce homelessness, which left city policymakers in the difficult position of having to manage the homelessness problem apparently indefinitely without being able to end it.
One response to these problems was Housing First. At the least, Housing First substantially modifies the paternalistic quid pro quo of housing in return for compliance with various sorts of rehabilitative programs by offering the housing before expecting compliance, or rather, by offering housing and not expecting, but perhaps hoping for, compliance. Now, if we interpret Housing First as simply a variation on paternalism—that is, if we see it as a tactic for in fact increasing compliance with treatment by offering the treatment after, rather than requiring treatment before housing is provided—then Housing First seems not to deliver on the paternalistic quid pro quo. Housing First clients use, for example, substance abuse services less than treatment-first clients do, even though there is no difference in actual substance abuse among these two groups.28 Housing First clients also use psychiatric treatment services less than treatment-first clients do, though the differences are for the most part not statistically significant.29 Housing First, then, does not work if it is conceived of as a perhaps counterintuitive tactic for increasing compliance and thus making real the paternalist quid pro quo.
Housing First is more radical than that. Housing First doesn’t increase, and in some cases even lessens, compliance, but it improves outcomes, and thus obviates the need for compliance. Housing First clients, despite the fact that they use substance abuse services less, do not engage in substance abuse any more than treatment-first clients do.30 And, whatever their use of psychiatric services, Housing First clients show fewer psychiatric symptoms than do treatment-first clients.31 If mental illness and substance abuse are supposed to be underlying causes of homelessness, Housing First does a better job of addressing them than do treatment-first programs that paternalistically insist on the use of services that are supposed to address those underlying causes. In other words, Housing First radically undermines the paternalistic paradigm.
Housing First strikes at both the underlying cause and quid pro quo assumptions of paternalism. First, consider the underlying-cause assumption. It is important to understand that, unlike some of the formulations of the antipsychiatry movement of the sixties and seventies, such as those of Thomas Szasz and R. D. Laing, Housing First philosophy does not deny that its potential clients are mentally disabled. What is denied is not the reality of disability but that this real disability precludes subjects from being housed. In other words, Housing First admits that there is “something wrong” with its clients. Housing First’s point is that providing permanent housing need not be delayed until this disability is dealt with. Housing First is not housing only; clients’ disabilities are dealt with by the extensive support services provided with the permanent housing. But access to the permanent housing is not contingent on use of the services. Disability remains a real and serious problem, but it is no longer a “cause” of homelessness in the sense that provision of housing is futile if the disability is not dealt with first.
But if there is no underlying cause of homelessness, demanding a quid pro quo of clients—housing in return for compliance, with a rehabilitative regime to address the cause—makes no sense; thus Housing First undermines the mutual responsibility or quid pro quo assumption of paternalism. Another way of putting this is to say that Housing First rejects the “directive and supervisory” nature of paternalistic policy. Under paternalism, clinicians or front-line workers or operators have to provide clients with what Mead calls the “help and hassle” they need to overcome their disabilities, which are presumed to be the cause of their poverty or homelessness. When disabilities are no longer thought of as causes of homelessness, the role of the providers is reduced as their intervention is no longer considered essential to overcoming privation. This is not to say that clients’ disabilities do not need to be treated, but that housing no longer need be contingent on treatment and that therefore clients can choose when and if to take up treatment. Tellingly, Housing First rhetoric designates program beneficiaries not as “clients” but as “customers,” and speaks of its “shift from ‘expert’ service provider to the ‘expert’ consumer of the provided service.”32 Tsemberis has said that “the whole point of consumer-driven services is putting the people that the services are intended to serve in charge of the program. Turning the asylum over to the inmates.”33 For all of its clear-headedness about the reality of mental illness, Housing First clearly represents a fundamental break with paternalism. It was under the Bloomberg administration that Housing First policies began to be widely implemented, especially in new outreach techniques to the street homeless, including the use of Safe Havens, and in changing intake procedures for homeless families to deemphasize diagnosis of underlying causes and instead stress quick planning for rapid rehousing.
Another response to the limits of paternalism was Culhane’s refocus of policymakers’ attention on chronic homelessness. By demonstrating that a relatively small percentage of shelter clients accounted for a greatly disproportionate percentage of shelter use, Culhane identified a strategy by which a dramatic impact on homelessness might be made: concentrate on housing the chronically homeless and thus reduce shelter use strikingly. Thus the possibility of achieving something like an end to homelessness began to seem within striking distance. Making use of Culhane’s analysis, Philip Mangano of the United States Interagency Council on Homelessness, proselytized to any state and locality that would listen the importance of developing ten-year plans aimed at ending homelessness. As we saw, one of the jurisdictions that headed Mangano’s call was New York City, where Bloomberg ambitiously adopted not a ten- but a five-year plan to “make the condition of chronic homelessness effectively extinct in New York”34 and reduce the city shelter population by two-thirds. Thus the never-ending stalemate of managing homelessness that the paternalistic paradigm seemed to hold forth would be “overcome,” as the mission statement of DHS under Bloomberg came to read.
Also integral to the goal of ending or at least dramatically reducing homelessness would be the strategy of prevention. Prevention represents a step away from the paternalistic paradigm because it assumes that in most cases there is nothing so very “wrong with” potentially homeless families and singles that they can’t be maintained in housing with a little bit of help. Here again it was under Bloomberg that prevention was prioritized and raised to coequal status beside the task of providing shelter in the DHS mission statement. The main programmatic manifestation of that reemphasis was the Home Base program, which Bloomberg initiated, very publicly subjected to scientific evaluation, and made a centerpiece of his homelessness policy.
Thus it was during the early years of the Bloomberg administration that a post-paternalistic approach to homelessness began to be embraced in New York. The main elements of this paradigm were articulated by that administration in the 2004 overall plan to reduce the shelter census by two-thirds within five years. Not only was this goal not met, but by the end of the Bloomberg years the shelter population was at an all-time high of about sixty thousand people. What had gone wrong? Had the post-paternalistic paradigm spectacularly failed?
Rather than fail, post-paternalism was broached and then abandoned under Bloomberg. As was discussed earlier, the post-paternalistic concern with immediately planning to rehouse homeless clients implies that sufficient resources must be deployed to make those plans work, which implies that rental subsidies ought to be available. But under Bloomberg, first the priority assignment of NYCHA units and Section 8 vouchers to shelter families was stopped, then the limited rental subsidy program, Advantage, which was implemented partly to compensate for the end of these preferences, came to an end. The end of the subsidies amounted to a break with or a failure to follow through on the logic of the post-paternalistic paradigm and also precipitated the dramatic rise in the shelter census.
How is it that subsidies came to an end under Bloomberg? The administration seems to have ended the preferential access to NYCHA units and Section 8 vouchers for shelter families because it believed that policy created a perverse incentive for families to enter the system. On the basis of a temporary decline in admissions to the family system, Linda Gibbs at the time believed that “we had confirmed what everybody had expected, which is that there were a number of [families that] came toward the shelter because they were trying to do their best to access Section 8 and the Housing Authority Units. And that—when that wasn’t available anymore—did cause decrease in demand.”35 Why the Advantage program came to an end is less clear. The city and the state got into a disagreement over funding, and after weeks of intense lobbying by the city to maintain the program, the state withdrew its funding, the city declined to support the program on its own, and Advantage came to an end. Having that happen on their watch had very unfortunate consequences for the Bloomberg administration. Rapidly rehousing shelter clients became much more difficult, the shelter census rose to unprecedented levels, and the very promising beginning Bloomberg made in introducing a new direction in homelessness policy came to a disappointing end.
De Blasio revived and greatly expanded the post-paternalistic paradigm that came to an end in the later Bloomberg years. The new administration did not take up some of the rhetorical flourishes of its predecessor. De Blasio declined to promise to overcome homelessness or set a specific goal for shelter census reduction, but that is the case because it was acknowledged that it would take some years before new policies could be expected to make a dent in those problems. But the Bloomberg emphasis on prevention was reinforced with the expansion of Home Base and the consolidation of various prevention programs at HRA. Subsidies, in the form of the LINC programs, consistent with the interest in rapidly rehousing rather than rehabilitating clients, were introduced. Some iterations of LINC were aimed at the chronically homeless. Homelessness is highly relevant to de Blasio’s political commitment to fighting inequality, so it seems likely the new paradigm will be implemented as consistently in his administration as political reality will allow.
Another remarkable feature of homelessness politics in New York City is that the rise of a well-funded, high-quality system of care took place, mostly, during the tenure of what John Hull Mollenkopf describes as a “conservative dominant coalition.”36 In his book A Phoenix in the Ashes: The Rise and Fall of the Koch Coalition in New York City Politics, Mollenkopf describes how Koch was able to forge a “conservative, white, pro-growth coalition” at a time when biracial and neighborhood-oriented coalitions were in power in other cities.37 Mollenkopf shows in detail how the political and economic environment of New York City caused this conservative coalition to be in power more often than alternative progressive coalitions. The course of politics during the years we have discussed bear out Mollenkopf’s prediction, with Koch, Giuliani, and Bloomberg all putting together variations on the conservative coalition and with Dinkins being the sole representative, before the election of de Blasio, of an alternative progressive coalition. Although Mollenkopf notes the increase in spending for the homeless under Koch, his few remarks on homelessness present the issue as one of concern primarily to white liberals and other constituents of the Dinkins progressive coalition.38 But if that is the case, why was expansion and improvement of the homelessness system pursued not only by Dinkins but also by Koch, Giuliani, and Bloomberg, all, within the New York City context, conservatives of various stripes?
Part of the answer is the creative use made of the court system by Robert Hayes, Steven Banks, and others within the city’s homelessness-policy community. To a considerable extent, the courts forced the conservative mayors to attend to a problem they otherwise might have ignored. But the power of the courts is not the whole answer. Consider, for example, Koch’s decision to sign the Callahan consent decree. Koch himself claims that he signed the decree on the advice of his corporation council, Allen Schwartz, whom he claims said, about the Callahan decree and other decrees, “it’s better that we sign these decrees, because otherwise they’ll impose even heavier sanctions.”39 But most of the other participants on the city’s side reported that one key consideration was that they felt that signing the decree was simply the right thing to do. Robert Trobe and Bonnie Stone of HRA both felt that the city could not take the position that it would not mind if some homeless people froze to death on the streets as a result of there being no decree. And according to Len Koerner of the city’s Law Department, “at some point during the litigation in Callahan, everybody agreed that it wouldn’t be so terrible if men had a right to shelter. . . . [E]verybody agreed that it was a good thing to do and therefore we signed it.”40 No doubt the city’s decision to sign the Callahan decree was driven as well by very practical considerations, including the possibility that had the city decided to litigate the case, it might have lost and been faced with more demanding requirements. But a sense that signing the decree was the right thing to do played a hard-to-quantify role. Thus an administration representing the city’s conservative-dominant coalition with few strictly self-interested reasons to care about the homeless found itself moved by what Kelman calls “[p]ublic spirit . . . [the] inclination to make an honest effort to achieve good public policy.”41
Another key decision of the Koch administration that was motivated in part by public spirit was his decision to invest $4 billion in his ten-year housing plan. About 150,000 units would eventually be developed. Only 10 percent of those units would go to homeless families, partly because a program that featured benefits for moderate-income families was politically more tenable. Nonetheless, building support for the plan was not easy. In his recent biography, Ed Koch and the Rebuilding of New York City, Jonathan Soffer writes, “In practice the mayor got flack for his program from both the Left and the Right. The Right wanted an unregulated private housing market. . . . [T]he Left. . . . deplored Koch’s decision to build significant amounts of middle-income housing, demanding that the needs of the poor should come first.”42 Mollenkopf argues that Koch “felt compelled” to adopt his housing plan, despite his political base, which included “real estate developers, investors . . . their lawyers . . . the city’s business elites and the leading voices for the establishment,” because of “intense conflicts” that his otherwise pro–private developer and pro-gentrification policies precipitated.43 But a sense that rebuilding low- to moderate-income New York was simply good public policy was also important.
Public spirit was perhaps also a factor in the Bloomberg administration’s early efforts to emphasize and forge a new direction in homelessness policy. Exactly why that administration did not follow up on its insights is hard to determine. Too much concern with perverse incentives seems to have played a role. And perhaps Bloomberg’s public-spirited desire to address homelessness was in the end overcome by the fact that the problem was simply not central to his main political concern of economic development.
Public spirit was perhaps less obvious during the Giuliani administration than under Koch or Bloomberg. Certainly Giuliani devoted considerable attention to homelessness policy—more, perhaps, than one might have expected of a conservative. Giuliani’s main policy direction was implementing the recommendations of the Cuomo Commission. No doubt Giuliani believed that this represented good public policy. But it also represented good political instincts, for during his second, successful campaign for mayor, Giuliani achieved considerable political mileage from promising to be a more energetic supporter of the Cuomo Commission than Dinkins seemed to be. Other signature Giuliani policies, such as the campaigns against squeegee men and in favor of public order, were essentially what would be expected from a mayor with a white, middle- and upper-middle-class base.
Public spirit played a role in the Dinkins administration. The most striking feature of Dinkins’s homelessness policy was its U-turn from a conventional liberal approach emphasizing rights and rapid access to permanent housing to a paternalistic style that focused on rehabilitation, eligibility, and quid pro quo. It is true that this paternalistic turn was more a matter of intention than implementation under Dinkins. We have discussed how Dinkins was slow to embrace the change of direction supported by the Cuomo Commission, perhaps for ideological reasons, so that when he eventually did, little time was left for implementation. But Dinkins did finally accept the commission’s recommendations even though they flew in the face of the progressive policy direction outlined while he was Manhattan borough president and endorsed by the city’s liberal coalition. A number of factors explain this shift. The apparently incorrect perception that a perverse incentive had been created by the rapid rehousing of homeless families had an impact. So did the emergence of Andrew Cuomo as a skilled policy entrepreneur. But the simple fact that the Cuomo Commission put together a convincing case that persuaded Dinkins that a change in direction was the right policy decision was also important.
The development of an improved shelter system during a time of domination, mostly, by a conservative ruling coalition becomes less incongruous if it is placed in the context of other policy developments during that time. We have already mentioned the rebuilding of the city’s marginal housing stock. The decline in crime and the development of safer public spaces is well known. A less noticed accomplishment is the vast improvement in public transportation since the early 1980s.44 We usually think of a conservative-dominant coalition as being narrowly concerned with balancing budgets and economic development. During the time period covered here, New York City’s conservative coalition did more than that.
In that it has achieved things one might not expect of a city dominated for a considerable period by a conservative coalition, New York City resembles Atlanta as it is described by Clarence N. Stone in Regime Politics: Governing Atlanta, 1946–1988. Stone describes a governing coalition in which the downtown business elite plays a preponderant role. Yet the Atlanta governing coalition has a record of accomplishments that go beyond a narrow business focus. Stone writes,
Governance requires the power to combine necessary elements for a publicly significant result—whether it is building a downtown expressway system, developing new housing for blacks in the outer area of the Westside, hiring black police officers in a Jim Crow city, redeveloping substandard areas next to the business district, peacefully desegregating the school system in an era of massive resistance, launching a mass-transit system, putting on a National Black Arts Festival, or rebuilding Underground Atlanta as a major entertainment district. Atlanta’s postwar governing coalition has accomplished all of this and more.45
In Atlanta, the downtown business elite governs, not by domination, as is described in Floyd Hunter’s classic account of the power structure of that city during the 1950s,46 but because the resources it commands and the internal coordination it has achieved makes it an attractive partner to any interest group hoping to achieve significant change. The result is that Atlanta’s business elite usually works in partnership with the city’s black middle class, and anyone else who wants to get something done. Atlanta’s governing coalition is “activist but not progressive.”47 It focuses, as the phrase goes, not on exercising power over the populace, but on deploying power to “carry on great enterprises.”48
New York City’s dominant governing coalition looks nothing like Atlanta’s. In particular, the black middle class in the Big Apple plays nothing like the role that it does in Atlanta. The regimes resemble each other, however, in both being activist but not progressive, which is the phenomenon we are trying to understand. Stone’s analysis of Atlanta is also relevant because in New York, with its decades-old tradition of civic enterprise, power is most usefully understood as power to, rather than power over. But the way in which power to is exercised in New York City is very different than in Atlanta, or perhaps than in any other American city.
Exercising the power to in New York City involves getting things done in a highly fragmented political environment. One sense in which that environment is fragmented is that there are simply very many formal governments. It is still true that, as Robert Wood famously wrote in his 1961 book 1400 Governments, the New York City metropolitan area is “a governmental arrangement perhaps more complicated than any other that mankind has yet contrived or allowed to happen.”49 If we focus on just the formal city government, it too is fragmented in various ways. It is a strong-mayor form of government, meaning that the executive and legislature are independent of each other. Unlike most cities, New York encompasses five separate counties, which are also boroughs and which were each represented on the city’s Board of Estimate before that entity was abolished in 1989. Moreover, as Theodore Lowi wrote in his classic account, At the Pleasure of the Mayor, New York City government is “highly articulated,” in the sense that it consists of many different agencies trying to do many different things.50 Each agency is lobbied by and responds to its own set of interest or “satellite” groups, with the result that “there are as many ‘governments’ as there are areas of public responsibility.”51 Those who would exercise power to carry on great enterprises in New York City face a blooming, buzzing confusion of a political environment.
Such fragmentation is supposed to frustrate change. As one textbook on urban politics put the matter, “In addition, political fragmentation makes it difficult to assemble sufficient political resources to address problems. It may be impossible to arrive at consensus among so many different jurisdictions, and this may lead to ‘immobilism and non-decisionmaking.’ . . . Urban political structures change slowly in an incremental, evolutionary fashion.”52
But as we have seen, homelessness policy in New York City has over the last thirty-six years developed in a series of striking leaps with dramatic, short-term changes in funding, administration, and policy “philosophy.” How was this possible?
Part of the answer is that New York City’s fragmented governance system creates many opportunities for policy entrepreneurship, and policy entrepreneurship is capable of precipitating nonincremental change.
Policy entrepreneurship refers to political activity by middle men (and women) who simplify and distill expert knowledge and disseminate it in a broader political arena. Political actors—private individuals, elected officials, organizations—resort to entrepreneurship when they find it difficult to organize a potential interest group. Policy entrepreneurs stand a better chance of getting their ideas adopted in fragmented governance structures. By hypothesis, policy entrepreneurs represent an unorganized interest. Such interests will not necessarily prevail in a general political contest (for example, in a national election for a parliament), where voters must weigh many different policy ideas and make one vote that is only slightly influenced by any one of them. Policy entrepreneurs have a better chance when there are many political contests. For one thing, the more contests, the more chances of an unorganized interest prevailing. Further, where there are many political contests, some of them are likely to be nonelectoral. Court cases are a good example. In court, a policy entrepreneur need only convince one citizen—the judge—to adopt a policy idea.
But fragmented governance structures are favorable to policy entrepreneurs for another reason. Fragmented governance almost necessarily implies separation of powers. The separate election and independent standing of the executive and the legislature is the most fundamental of all government fragmentations. In the centralized system of parliamentary governments, policy entrepreneurs, if they cannot capture a majority of the all-powerful legislature, are completely out of luck. In a separation-of-powers system, policy entrepreneurs have a chance of getting their way if they can achieve the less daunting task of capturing the executive. Of course, capturing the executive is of little use if the executive has little power. In a strong-mayor system, in which a sole executive has direct authority over the bureaucracy (as opposed to weak-mayor systems, in which the mayor is but one member of a council to whom the bureaucracy is collectively responsible), policy entrepreneurs, when they capture the executive, have a better chance of seeing their ideas implemented.
Finally, policy entrepreneurs have better chances when the political environment is not only fragmented but competitive. Where there are lively contests between two parties, there are two happy outcomes political entrepreneurs may achieve. Both parties may struggle to “own” a particular policy idea. This is what happened in Washington in the 1980s with tax reform and in the 1990s with welfare reform. Republicans and Democrats competed with each other to see who could come up with the flattest tax, the most demanding welfare system. Nonincremental policy change is the likely result of such a “bidding up” competition. Another possibility is that one party may win possession of a policy idea and go on to capture some branch of government with it. In that case the winning party has a clear mandate to implement its idea. Again, nonincremental change is a likely result.
In most respects, fragmented New York City fits the bill of an entrepreneur-friendly environment. To begin with, in New York City, courts have jurisdiction over administrative decisions by the bureaucracy. Hayes’s strategy of translating a political demand—“help the homeless”—into a litigable right—“the right to shelter”—is a textbook case of judicial political entrepreneurship. It was also a classic example of the “rights strategy”—converting a potentially unpopular political demand into a right—which is particularly effective in the context of American political culture and a strong judiciary.53 This strategy helped Hayes to sell his cause to the public and the media, for an unprepossessing interest had been transformed into a hard-to-deny cause.
It is precisely because in New York policymaking power is fragmented between the courts and the other branches that advocates could achieve real power by winning in the courts. With the court effectively on their side, advocates had great say in remaking the shelter system, as the mayor had few choices besides defying the court or going along. Thus, for practical purposes, the nonincremental change of a right to shelter for singles, with all its fiscal and programmatic implications, had arrived.
As we have discussed, McCain and related cases and regulations are good examples of how a fragmented political environment, usually thought to generate many veto points, can sometimes create many opportunity points.54 The fact that the state, as well as the city, had jurisdiction over the shelters was crucial here. As we saw, the question of what homeless families were entitled to bounced back and forth between various cases and various courts and the state Department of Social Services, receiving new additions or modifications each time. The result of many incremental steps turned out to be the nonincremental step of creating what was effectively a right to shelter for families.
When nonincremental change is achieved, future changes are more likely to be nonincremental as well. For past policy is an influence on future policy. Where past policy has been characterized by dramatic breaks, future dramatic breaks are legitimated. Moreover, when one interest wins a striking victory, other interests mobilize to fight back. By 1989 advocates for the homeless were seen, correctly or not, as having gotten almost everything they wanted. When the right to shelter appeared to have its own problems, the advocates for the homeless, correctly or not, were seen as having been discredited. Yet the idea that radical change was not per se bad had been established. It was inevitable that a counter idea would start to crystallize in such an environment.
That New York City homelessness policy was based on the idea of a right to shelter was a potential barrier to the crystallization of a counter idea. Significantly, the counter idea that did crystallize—mutual obligation—specifically did not reject the right to shelter but only stressed that with a right should come a responsibility. In this way the popular rights premise was not challenged directly. Two features of New York City politics account for the sudden success of this counter idea. First was the newly competitive political environment. As a Republican competed strongly for the mayoralty for the first time in decades he, Giuliani, was naturally eager to find some idea to use against the Democrat. Of course, Dinkins too had an incentive to embrace new ideas, but to do so in the case of homelessness policy would have required of him a 180-degree turn. That was by no means impossible, but Dinkins was ideologically and psychologically not disposed to it, and Giuliani had the easier task of picking up a new idea and running with it.
Further, New York City’s strong-mayor government made the counter idea of mutual obligation relatively easy to implement. In a weak-mayor system, Giuliani would have had to resell his idea to the other members of the council, and then turn it over to a general administrative officer or city manager to implement. As an institutionally strong mayor, Giuliani could choose and hold accountable his own policy team. Thus a second bout of nonincremental change, the paternalistic moment, came about.
We are still left with the problem of distinguishing incremental and nonincremental change. The changes wrought by Callahan v. Carey in the New York City shelter system, even in budgetary terms, surely count as a nonincremental change. But the changes in the shelter system stemming from The Way Home, while extensive and important, did not generate such a striking quantifiable change. Was this second bout of change, then, really nonincremental? We might respond by noting that magnitude is not necessarily the main measure of nonincremental change. The feature of nonincremental change that especially interests us is that it breaks with the incrementalist model by producing changes that that model would predict are not possible during times of ordinary politics. Put another way, nonincremental change is the product of the “politics of values and ideas,” with its key features of expertise, policy entrepreneurs, institutional fragmentation, competitive political environments, and, especially, public ideas. Defined in these terms, the paternalistic reform of the shelter system under Giuliani was nonincremental, for it did involve public ideas, policy entrepreneurs, and so on, as discussed here.
Thus the first two moments in New York City homelessness policy—the entitlement moment and then the paternalistic moment—can be considered nonincremental changes that were precipitated by policy entrepreneurs marketing attractive public ideas in a competitive, highly fragmented political environment. What about the third policy moment we have identified, the post-paternalistic moment characterized significantly by the rise of the Housing First “philosophy” and movement?
Housing First might be considered an instance of “policy feedback,” in which a new policy is forged in reaction to perceived failures or successes of past policy.55 As we saw, one of the earliest manifestations of the paternalistic reaction to the entitlement moment that characterized most of the Koch and Dinkins administrations was Koch’s attempt to involuntarily transport mentally ill street dwellers to Bellevue Hospital under the Project HELP outreach program. Whatever one thinks about the mental competency of Project HELP’s most famous client, Joyce Brown, and of the ethics of mandatory psychiatric treatment, the policy of involuntary transportation proved to be a failure, as mental health law, a strong advocacy community, and limited hospital facilities made it impossible to permanently remove mentally ill homeless people from the streets against their will. Tsemberis was hired as director of Project HELP shortly after the Joyce Brown affair cost the earlier director her job, and he saw the futility of the revolving door between the streets and psychiatric wards that the project created. And later, after the development of the paternalistically oriented program shelters during the Giuliani administration, more policy actors would become aware of the limitations of the paternalistic model as many clients of the shelters for singles declined to accept the quid pro quo of better services for enrollment in the therapeutic services of a program shelter and remained in the city-operated general intake shelter. Housing First was conceived and implemented as a feedback response to the failures of paternalism to reduce the number of street people to acceptable levels and to enroll enough clients in therapeutic services.
There appeared at this moment of policy failure a charismatic policy entrepreneur, Sam Tsemberis. He faced a formidable challenge in marketing his proposed policy change. With the report of the Cuomo Commission and its adoption and implementation by the Giuliani administration, the paternalistic reaction to the initial entitlement moment had crystallized. Housing First represented a radical rejection of the “underlying cause” and “quid pro quo” assumptions of paternalism. Moreover, Housing First had roots in a number of ideologies that were unpopular among many mental health service providers. Tsemberis has described himself as a follower of R. D. Laing, the radical psychiatrist whose work seemed to question the whole idea of mental illness. Although Tsemberis never denied the reality of mental illness, his insistence that the homeless were “experts in their own lives,” and his emphasis on making services customer-centered rather than clinician-centered meant that Housing First was likely to meet with at least initial opposition from the mental health service providers who were ex hypothesi to be put out of charge of their programs. Tsemberis, along with his coauthors, has also noted that “[m]ainstream housing and treatment providers were also extremely skeptical about the efficacy and consequences of the PHF (Pathways Housing First) model of service delivery. Some voiced concern for consumers, saying that placing homeless persons with mental illness in permanent housing was reckless and potentially damaging.”56 How then did this radical departure from settled policy make any headway?
Again, part of the answer is that New York City’s unusual political environment is conducive to policy entrepreneurship. Deborah Padgett, former president of the board of Pathways to Housing and a Housing First researcher, has argued that “[p]athways . . . could have only happened in New York. . . . because we have such a chaotic, fragmented system that this little upstart could take root.”57 At the time Pathways to Housing was founded, New York City’s homelessness service system had been privatized, or more accurately not-for-profitized. Housing and services were provided not by a single entity like the city, but by a wide range of private organizations. In this environment it was possible for a single, small provider with a distinctive philosophy to spring up. Convincing a single city- or state-wide provider to abandon its settled practices and embrace a radical policy departure would have been much more difficult.
Another aspect of New York’s fragmented policy environment was also crucial: Authority over homelessness policy was shared among the city, state, and federal governments. At crucial moments Tsemberis received help from the state and federal governments at a time when the city, having bought into the paternalistic paradigm, was less sympathetic. The first support for Pathways to Housing came not from the city but from New York State, whose progressive commissioner of the Office of Mental Health, Richard Surles, was more open to the radical message of Housing First than the Giuliani administration would have been. Later, the federal government also played a hand in the institutionalization of Housing First when Phil Mangano, executive director of the U.S. Interagency Council on Homelessness, became a convert to the cause. Mangano, who had given up on the strategy of trying to coordinate the behemoths of the federal bureaucracy, instead conceived of himself as an advocate and was in search of an idea he could market to cities and states. During Mangano’s tenure on the council (2002–2009), the scholarly literature supporting Housing First had come to a critical mass, providing him with material to disseminate. As we saw, it was Mangano who was instrumental in selling Housing First, and the idea of a plan to end homelessness, to the Bloomberg administration.
This brings us to another feature of the new politics of public policy that played a key role in New York City’s, and the world’s, adoption of Housing First: professional consensus. Housing First went from being a radical initiative to international standard operating procedure in large part because the academic literature overwhelmingly supported Tsemberis’s claim that his approach provided better outcomes in housing retention, cost savings, and other key goals than did traditional approaches. Indeed, it was with an idea of developing such a consensus that Tsemberis from the beginning subjected his approach to rigorous evaluation; he knew that such convincing evidence would be necessary to overcome initial skepticism. Tsemberis also showed good judgment in taking the results of arcane scientific analysis and boiling them down into one easily communicated public idea: “Housing First.”
Overall, Housing First is a striking example of how an idea, based in professional consensus, effectively communicated by policy entrepreneurs, and developing a toe-hold in a fragmented policy environment, can effectuate nonincremental change in what has traditionally been thought to be a situation unfriendly to dramatic innovation.
We are not in a position to make sweeping claims about the nature of urban political change. But perhaps urban political science can look at the history of homelessness policy in New York City and consider the possibility that fragmentation leads, not to stasis or gridlock but to change.