Archon Fung


Chapter 1:
Introduction

1.1. What's the Matter With Democracy?

Characterized by seemingly contradictory sentiments of self-satisfaction, disappointment, and exhaustion, the end of the twentieth century is a peculiar moment in the history of American democratic government. On one hand, we in United States rightly celebrate the evident worldwide victory of our political model--representative government combined with basic human rights--over various kinds of dictatorships (Fukuyama 1989). While this victory has so far been moral rather than practical--many rulers do not enjoy the electorally expressed consent of their governed--it is nevertheless decisive. There are no broadly credible political forms that fall outside these liberal democratic boundaries.
In the United States and other mature liberal democracies, disappointment rather than elation fills this moment of triumph. Unlike newcomers to constitutional democracy, we do not look upon the machinery of electoral government and the administrative state with the naïve eyes of those at the beginning of a long adventure, but rather with the weary hindsight of an experienced traveler. Our experience with this kind of government has generated at least three disappointments. First, the people lack the enthusiasm necessary to utilize their hard won rights to political representation. All democracies exhibit low levels of electoral participation, especially in lower-level elections, and the trend in recent decades has been downward (Lijphart 1997; Sartori 1987:103). In the United States, the turnout for presidential elections has declined from 60-65% in the 1950s and 1960s to 50-55% in the 1980s and 1990s. In off-year elections, turnout in local elections has ranged around 25% (Teixeira 1992). The second disappointment is that these low rates of participation do not seem to indicate satisfaction or even apathy about the state's performance, but instead have been accompanied by a precipitous decline in popular confidence toward government. Public opinion polls, for example, reveal that the percentage of survey participants responding that they "trust the government in Washington to do what is right "all of the time" or "most of the time" has declined from a peak of 70% in 1966 to less than 25% in 1992 (Putnam 1994).
More practically and urgently, our system of government--elected politicians that control administrative bureaucracies--have disappointed us with their apparent failure to deliver the public goods that are most important to us (Osborne and Gaebler 1992: 1-25). Their mission failures are most evident in America's large urban areas, especially our hard core inner cities. There, the very built environment of the streets, electric lights, and sewers disintegrates like so much melting snow. The schools are some of the worst in the developed world (Chicago Tribune Staff 1988), and urban American homicide rates soar at an order of magnitude above that of the rest of the nation and of other industrialized nations (see chapter 5).
These substantial disappointments of our conventional democratic institutions have drawn few constructive proposals from democratic theorists. While practitioners at the local, state, and federal level have embarked on a rather bewildering variety of experiments,[1] theorists and critics of democracy have failed to follow their lead by offering constructive proposals to amend the basic structure of representative government and bureaucratic administration. With precious few exceptions,[2] even those theorists once committed to radical and participatory transformations seem strangely exhausted and resigned to accept the forms and defects of more conventional institutions (Habermas 1992, 1996; for comment on this trend, see Phillips 1993 and Galston 1993).
It may be that these scribblers have fallen silent because they have discovered a deep truth that we have indeed approached the end of history. Perhaps no amount of creative institutional tinkering can carry us beyond these basic disappointments of representative government and its ironic inability to advance those values most closely associated with the idea of democracy. While this pessimistic thesis may turn out to be true, no one has yet proved it. Given the gravity of democracy's disappointments, then, prudence dictates that we proceed as if it were not true by continuing to search for undiscovered institutions that better vindicate our deep, professed democratic commitments.

1.2. Elements of a Radical Alternative

The challenge is simple: find institutions that better realize our core democratic values--in particular the fairness, effectiveness, and popular determination of public action--than the currently dominant arrangements of electoral representation and technical bureaucracy. Though there must be many rejoinders, this volume offers one sustained response by developing a proposal for radical democratic governance called Street Level Democracy (SLD). Any such response must inevitably take into account numerous contentious speculations concerning, for instance, the abilities of ordinary citizens and the tractability of complex urban problems. Since these tendentious empirical matters are unavoidable, we do not begin theoretically, as many democratic theorists might,[3] but instead develop SLD by examining and extrapolating from very concrete urban initiatives.
To a much greater extent than in any other American city, two Chicago programs set into place structures for direct democratic governance of educational and public safety institutions in its neighborhoods. In 1988, a state reform law devolved control over basic issues to elected governance councils--composed of parents, teachers, and principals--located each school. In 1994, city-wide policing reform established a parallel structure in community policing. In each of the city's 280 beats, residents meet frequently with police to jointly establish public safety priorities for the neighborhood as a whole and to direct police power toward the resolution of those priorities. Throughout the chapters that follow, we use the empirical developments in school and police governance to develop abstract radical democratic intuitions into workable institutions and to explore speculations about what citizens and their institutions can do. We use the theoretical lens of SLD as an abstract structure, in turn, to advance the interpretation of these agency reorganizations beyond mere description into a template for democratic reform. By using each as a step for the other in this way, we hope to move the project of building a workable democratic proposal beyond what the theorist or the empiricist working alone could accomplish.
The nub of the proposal is that radical democracy has a lot to offer modern complex society. Contrary to the view that the sheer scale and complexity of tasks set to the modern nation state prohibits direct forms of democracy, I argue (counter-intuitively) that precisely these modern features of complexity and scale have frustrated the received institutions of elite-mediated mass democracy and hierarchical administration, and that increasingly complex and diverse contexts of public action open spaces for bureaucratic reconstruction along more decentralized, participatory lines.
From the realm of past ideas, SLD borrows freely from two intellectual streams--participatory democracy and pragmatism--to construct a decentralized, participatory, experimental, and deliberative institutional alternative to our received form of democratic state action (electoral representation-cum-bureaucracy). Following the central claim of participatory democrats, SLD contends that those most concerned with a particular site of public action should be given the power, resources, and responsibility to carry out that action. In the case of education, for instance, it recommends decentralization of authority to the school level (site-based management) and governance structures that include parents, school administration, faculty, and (where appropriate) students. Whereas participatory democracy principally seeks the devolution of democratic power and authority from the center to the periphery, pragmatism sees decentralization as just the first piece--necessary but far from sufficient--of a much larger puzzle. Always attentive to results, pragmatism forces participatory democrats to focus more closely on what their institutions will look like and what they will do, "after the devolution."
To answer this question, SLD builds upon the decentralizing instinct of participatory democrats by giving contemporary institutional form to John Dewey's (1927, 1935) notion that the core of democratic activity is social exploration that engages both ordinary people and experts.[4] The proposal treats decentralized governance groups, such as residents and police in a neighborhood beat or a school's governance council, as democratic communities of inquiry that find and implement fair and effective means to carry out functions such as the education of children or maintenance of public safety. The image of inquiry is properly evocative on several dimensions. First, participants in SLD realize that the solutions to their problems are not obvious--that answers must be sought out because the routines prescribed by experts have for the most part failed them. Second, deliberation--the practice of decision-making through reasoned discussion rather than interest--settles disagreements about proper courses of group action. Third, as in inquiry, all such decisions are tentative and subject to revision based on further evidence, assessment of previous experiences, or in light of others' comparable experiences.
But decentralized democracy and deliberative pragmatism are not completely compatible. Most justifications for decentralization--for instance that it maximizes opportunities for political participation or that allows public action to be tailored to local preferences--do indeed unambiguously favor tiny polities. But does localism optimize the capacity for effective problem-solving inquiry? Pragmatists should favor small units for at least two reasons. First, front-line operators (patrol officers, teachers, production workers) know more about their specific local conditions, and so are likely to know what will work and what will not in light of that knowledge. Second, those who are close to the point of action are better positioned to assess the results of past efforts and bring that new knowledge to bear on future public choices. If individuals in a particular community of inquiry treat each other reasonably, are capable of formulating good guesses about the best course of action, can deliberately decide on a collective course, command the resources necessary to follow it, and improve upon past decisions through assessment, then self-regulating communities of inquiry may indeed generate fair and increasingly effective solutions to their respective public problems.
As this cursory yet demanding list of requisites already suggests, pragmatists may rightly voice reservations about prematurely eviscerating organizational centers. Centralized power, if not authority, might support efforts of inquiring communities in several critical ways. First, some units will be less capable than others, and that this difference in capability will correspond somewhat with background inequalities of wealth and power. The center can provide capacity building support--training in the individual skills of analysis and social skill of deliberation--to disadvantaged communities. Second, a central organization can enforce fairness in case the self regulating mechanism of deliberation fails--when there are momentary violent disputes or when one faction of a community entrenches itself and is not subject to reasonable appeals. Third, inquiring efforts of communities will benefit from access to the experiences of other communities. If we think of each community as conducting independent experiments in public policy, then random variation dictates that some will invent effective solutions more quickly than others. Connections between communities can diffuse these best-practices. These revelations about the character of centralized power and that connections between inquiring communities can be more or less conducive to communities' problem-solving capacities lead quickly to the conclusion that the overarching institutional structure encompassing these communities is itself an object of experimentation. A fourth function of the center, then, is to continually modify itself in ways that enhance the capacities of its component communities of inquiry. A fifth function of the center is not distinctive to pragmatist goals; decentralized units engaged toward similar ends will often encounter situations in which they each will benefit from common actions such as the creation of a common pool resource or joint action on a problem that overlaps jurisdictions.
The constitutional architecture of Street Level Democracy, then, defies the simple-minded dichotomy between power that is either centralized or dispersed. For a given public purpose--education and public safety are the ones that I will discuss--SLD recommends that the bulk of decision making authority and control over public resources be devolved to small, geographically based, operational units composed of both line level public servants and consumers of public services. In the case of education, the natural unit is the school and participants include teachers, parents, and the community. The appropriate unit in public safety is the neighborhood, and participants are police officers and residents. These units are charged with overcoming the obstacles that stand between the condition in which they find themselves and more desirable states--effective schools or safe neighborhoods. The central office or network that connects these operational units is muscular, but its goals are far different from the task definition and supervision roles of bureaucratic head-offices. Instead, it seeks to support component communities of inquiry by ensuring the integrity of their deliberative experimental processes, spreading the lessons generated by them, holding them accountable to their own plans and promises, and progressively redistributing capacity to those less advantaged.
These elements of Street Level Democracy--decentralization, direct citizen participation, deliberative problem solving, and a muscular center that simultaneously supports local units and holds them accountable--constitute an institutional set that can supplement and to a limited extent replace our received institutions with a directly democratic form that just might be able to overcome our disappointments with electoral/bureaucratic government.

1.3. Outline of the Argument

In the next chapter, we expand on this general discussion about the relationship between the institutions of governance and democratic values. In particular, we lay out five core democratic values--effectiveness, fairness, autonomy, deliberation, and solidarity. We say that one form of governance is more democratic than another just in case it more ably advances each of these values. Taking democracy seriously, then, entails searching for forms of governance that are more and more able to realize these core democratic values. With this theoretical democratic ruler in hand, the argument proceeds in three parts that make up the bulk of this volume. The parts explain the "how," "what," and "whether it works" of Street Level Democracy. The first part shows how SLD institutions can be built at all, given the resilience of hierarchical agencies, by describing how reformers constructed them in Chicago during the 1980s and early 1990s. Part II explains exactly what SLD is by providing a theoretical blueprint that lays out the roles of citizens, the local units in which they participate, and the "supportive center" that links these units together. The third part explores whether or not SLD works by empirically examining its operational characteristics in the streets of Chicago.
The first part, chapters 3 through 5, reviews the transition from centralized bureaucracy to decentralized pragmatic organization that has taken place in the Chicago Police Department (CPD) and the Chicago Public Schools (CPS). Until less than ten years ago, the CPS and CPD resembled the large urban bureaucracies that can be found in many American cities and were exceptional only in the tragedy of their failures. Changes in these two bureaucracies, then, may point out a trajectory of reform applicable to other cities and bureaucracies and shed understanding on how such revolutionary reform occurs.
I argue in that such a shift has indeed taken place and that it was a conscious response to "performance gaps" between public expectations of these urban agencies and what they actually were able to accomplish (Downs 1967). Whether or not large bureaucracies were ever the most efficient organizational form to keep the domestic peace or educate the nation's children, they certainly perform these functions less well than they once did and much less well than we would like them to do. In the poorest of inner-city neighborhoods, agency failures approach complete breakdown. Diverse reformers with quite different ideologies and motives banded together and used this undeniable breakdown to set in place two central planks of Street Level Democracy: decentralization and popular participation.
Whereas the Progressive watchwords of centralization and hierarchy gave birth to the great urban bureaucracies, the contemporary ideas about effectiveness recommend decentralization. Professional reformers, then, advocated administrative decentralization--site based school control in education and neighborhood patrolling in the police--as a core component of reform. In both education and policing in Chicago, remnants of the American New Left had formed minor social movement advocacy groups that emphasized the importance of community participation. Professional reformers in the late 1980s, perhaps because they saw it as one manifestation of "customer-orientation," were not hostile to these social movements and incorporated local voice and control into their reform packages. Thus, quite happily for advocates of neighborhood government and participatory democracy, decentralization and participation became two central reform principals in both the CPS and CPD.
In the years immediately following these dramatic reforms, those charged with educating children and preserving safety have discovered major deficiencies in these new organizations. While some used the new latitude of local control to make their schools more effective and neighborhoods safer, others were less successful. In some schools and neighborhoods, few participated. Others simply failed to take advantage of the new freedom and continued to execute the tried and ineffective routines in which they were practiced. Some seemed to lack the skills of analysis, deliberation, planning, and assessment necessary for self-directed action. The more and less successful alike wondered whether other similarly situated groups in the city had developed effective solutions to the same problems. In response to these implementation difficulties, the central offices of the police department and the school system, together with a host of interested organizations, developed additional institutional responses that resembled the recommendations of Street Level Democracy that come from its pragmatist moment: supportive mechanisms such as training, dispute facilitation, assessment devices, channels for information sharing and best-practices diffusion, and common pool resources.
The project's second part builds a constitutional architecture of Street Level Democracy that draw upon these CPS and CPS restructurings and upon the method of constructive political theory. How exactly do these pieces of Chicago reform fit together into a coherent alternative and why is the alternative promising? The ideal institutional account of SLD describes the political conception of a community of inquiry at three levels. Chapter 6 lays out the role of a citizen of a community of inquiry. That citizen must embrace the public ends for which the community is convened (a safer community or more effective schools) as her own ends, possess the capacities necessary to participate on an equal footing with others, and abide by the requirements of reasonable deliberation. Chapter 7 describes the local unit of SLD--a neighborhood beat group or school governance council. At this level, SLD is a deliberative problem-solving procedure that requires participants to identify and prioritize problems, strategize about how best to solve those problems, implement solutions, and constantly evaluate their efforts in order to improve their effectiveness. Finally, chapter 8 describes the requirements of an administrative "Supportive Center" that links and assists these communities of inquiry by pooling information, assuring accountability, and providing technical assistance.
By way of justification, I contend that SLD as an institutional alternative is more fair and effective than the bureaucratic form that it replaces. In Chapter 9, I argue that SLD is more effective than centralized bureaucracy in the complex and unstable problem environments posed by modern citizes because it yields several mechanisms that are unavailable to hierarchical administrations. Whereas bureaucracy discourages civic engagement, SLD depends upon it and can thus draw upon the information and energy of ordinary citizens. While the separation that professionalism imposes between public servants and citizens sows distrust, partnership partially constitutes SLD and so is likely to generate a capillary-level of trust between civil society and the state that in turn makes cooperation possible. One genetic defect of bureaucracies is that they persistently strive toward the impossible goal of eliminating the discretion of their low-level operatives. SLD, on the other hand, increases discretion at the operational level but attempts to harness it and make it accountable by inviting public direction. In the impulse to rationalize, bureaucracies seek clear divisions of labor between different functions--e.g. policing, sanitation, housing inspection, education, parks and recreation--that make difficult the coordinated, strategic deployment of these energies. Since SLD focuses on particular problems, it can more easily orchestrate and recombine such specialties to resolve problems such as drug houses, truancy, and school-to-work. Another constituent characteristic of professional bureaucracy is that they seek autonomy--isolation from politics, civil society, and other professions--which cuts them off from potentially constructive and transformative feedback about the effectiveness of their strategies. SLD, on the other hand, incorporates continuous assessment as a central feedback mechanism of its deliberative procedure.
Chapter 10 argues that SLD will also generate public action outcomes that are more fair than insulated hierarchical agencies. One objection is that SLD depends too much on the frail method of deliberation. Outcomes are fair to individuals within a community of inquiry when deliberation regulates that community's decision processes. One function of the center is to detect and correct persistent failures of deliberation within these units. Another objection to decentralizing schemes like SLD is that better off communities will leave disadvantaged ones behind, and the response is twofold. First, unfairness between localities is inevitable whenever background social inequalities are geographically correlated, and SLD is no more vulnerable to this criticism than bureaucracy. Another function of the center is to redistribute capacity, if not resources outright, to more needy units.
One persistent criticism of bureaucracy that has been made in many ways is that the consequences of a good idea at the top frequently get messy and generates unintended consequences at the bottom. With only the historical discussion of SLD as forward-looking citywide reform (in a city of some three million residents) and a conceptual constitutional defense that is necessarily abstract, SLD is similarly left wide open to the objection that it sounds good to practitioners (Part I) and theorists (Part II) but, hard reality being what it is, cannot work. Some common variations on this theme are: people won't participate (Olsen 1965; Riker and Ordeshook 1968); only rich people will participate (Nagel 1987; Verba and Nie 1987); only educated people are capable of participating effectively (Verba et. al. 1995); people will not subordinate their self-interest to the constraints of reasonable deliberation (Austen-Smith 1992); ordinary citizens are not knowledgeable enough to interact with experts; and a community of inquiry requires a homogenous community.
Part III of the project uses both city-wide data and a series of six neighborhood-level case studies to explore the extent to which these and other objections weigh against SLD as an institutional alternative to hierarchical bureaucracy. The Chicago reforms in public education and the police department provide a ideal large-N laboratory in which to explore of how and whether flesh-and-blood citizens in all of their situated complexity can deliberate and solve problems as SLD asks them to do.
Chapter 11 uses a variety of quantitative city-wide data on participation in school reform and community policing institutions to explore whether SLD "works" in the sense that various kinds of citizens participate in the deliberative opportunities that it constructs. Some of this chapter's findings may surprise skeptical critics. Over the period for which data was available, participation levels are not stunningly high, but are sufficient to sustain SLD's deliberative problem solving in both community policing and local school governance. Unlike nearly every other channel of political participation, residents from poor neighborhoods participate at rates equal to or greater than those from wealthy ones. In community policing, participation rates of poor neighborhoods are greater than those of wealthy ones. Weighing against the concerns of feminist theorists that deliberation favors male over female participants, it turns out that many more females than males participate in both community policing and school governance. Contrary to theorists (Putnam 1993) who argue that high levels of social capital--civic associations and civic norms--are critical components in "making democracy work," empirical researchers have found that neighborhoods poor in social capital are as effective as those who possess it richly in both school and police governance in Chicago.
While these city-wide statistics usefully sweep away substantial peremptory challenges against SLD and participatory democracy generally (e.g. "no one will participate" or "only the rich will participate"), they nevertheless treat neighborhood processes as black boxes of participatory democracy and thus cannot be used to examine more fine grained claims concerning SLD's deliberative mechanisms of problem solving. To answer these questions, chapters 12 through 16 explore the operations of SLD in six neighborhood-level cases: three school governance and three community policing groups. I observed the deliberations of each of these groups over a period of approximately ten months between 1996 and 1997. The cases are distributed across two dimensions of "initial conditions:" wealth and interest dispersion. Roughly, we expect that richer, more unified communities will exhibit more successful deliberation and problem solving.
In each case, we examine whether or not parties conducted their discussions according to the deliberative process described in chapter 7, and whether their joint action yielded fair and effective solutions. Chapter 13 examines community policing in a neighborhood separated by lines of race and class; chapter 14 explores SLD in two quite poor but unified communities; chapter 15 examines two groups that are both poor and internally conflicted; and the case of Chapter 16 is a wealthy school where the participants have a long history of solidarity and cooperation.
Not surprisingly, these case studies bear out expectations about the effects of initial conditions: wealthy groups find it easier to deliberate and solve problems than poor ones, and fair deliberation is more frequent in the absence of entrenched conflict. Three more surprising finding of these six cases relates to the relative performance of SLD compared to prior institutional arrangements--bureaucratic command-and-control in shorthand--in each of these neighborhoods. First, SLD arguably generated outcomes that were superior to those of the command-and-control mode in each case, across all examined variation in initial conditions. Like Mae West, even when the quality of deliberative problem solving was bad, it was still better than hierarchical bureaucracy. Second, the gains in problem-solving capacity in switching from command-and-control to SLD institutions seems greater for poor communities than for wealthy ones. In the wealthy areas that we examined, residents had for decades employed rich networks of political influence, voice, and self-help to circumvent the problem-solving deficiencies of school and police bureaucracies. For them, SLD reforms simply added additional channels and methods to an already satisfactory system. In the poor areas that we examined, by contrast, SLD reforms introduced channels of political participation, voice over city services, and problem solving methods for residents who lacked these capacities in any meaningful degree. Finally, SLD's success under inhospitable conditions depended much more on the actions of the "Supportive Center" described in Part II. When teams and individuals dispatched from the CPS and CPD central offices performed their facilitative and technical assistance functions well, groups in poor and divided communities deliberated more effectively.
Now the purpose of a study like this is not to establish definitive conclusions, but rather to open new avenues for debate and further exploration. The successive aims are threefold. Rhetorically, the project implicitly argues that more democratic theorizing should concern itself with the design of real institutions that advance core democratic values. Second, I hope to show by example that such efforts are fruitful by offering a feasible and intriguing proposal called Street Level Democracy for radical, deliberative democratic governance that can solve pressing urban problems and advance other democratic values such as fairness and deliberation. Third, the empirical exploration of real world reforms that approximate the design of Street Level Democracy show these institutions operate roughly according to the specified design. Therefore, they offer some basis for optimism about democracy's unrealized possibilities and hold open the way for further exploration. Hopefully, fulfilling these three tasks will advance practical and theoretical discussions about overcoming democracy's disappointments one small step.


[1] We shall explore two in detail in this volume.
[2] See Unger (1987a); Dorf and Sabel (1997); Cohen and Rogers (1992).
[3] See, for example, Barber (1984). For a purely theoretical argument about why no institutional form can overcome the disappointments of electoral-bureaucratic government, see Zolo (1992).
[4] Dewey himself gave institutional expression to his notions of social pragmatism only in the case of school organization, and only weakly there. Some of his contemporaries, however, did worry about the problems of institutionalization (Follett 1930) and others even built nascent forms of these experiments (Phillips 1919; Devine 1919; Monney-Melvin 1981).