Archon Fung
The
parts of a machine work with a maximum of cooperativeness for a common result,
but they do not form a community. If, however, they were all cognizant of the
common end and all interested in it so that they regulated their specific
activity in view of it, then they would form a community. But this would
involve communication. Each would have to know that the other was about and
would have to have some way of keeping the other informed as to his own purpose
and progress.
-- John Dewey, Democracy and Education (1916)
In this chapter, we lay out the structure of the local units in which the
pragmatic citizens described above join their cognitive and political powers to
develop solutions to public problems. This local unit is the second of SLD's
three institutional levels; it is the theory's atom of social action. All
organizations are composed of basic units governed by rules of action and
decision--a squad or platoon in an army, a firm or production team in
manufacturing, the "nuclear" family, the New England town assembly, or the
local chapter of a national civic association or interest group. In SLD
generally, local units are small and coherent groups of lay citizens and
professionals dedicated to improving some aspect of public life. For the
concrete cases of school reform and community policing discussed below, the
local units are Local School Councils and neighborhood beat groups
respectively. Pragmatic citizens form local units (by joining them from below
and legislatively mandating their construction from above) in order to address
some persistent common need--such as the need to educate their children or to
maintain safe streets--without knowing precisely how to best satisfy that
need.[1] In these local units, they deliberate
experimentally to identify and implement solutions to the common problems that
they face. Using John Dewey's (1935) metaphor for democracy, we call these
units "communities of inquiry" because they govern themselves through a
deliberative problem solving process that is a political form of the ordinary
scientific method of inquiry. Section 7.1 specifies this constitutive practical
deliberation as a five-step formal procedure, while the second and third
sections show how the real world reforms in Chicago community policing and
local school governance have implemented local action units that can be
accurately described as problem solving communities of inquiry.
By way of orientation, briefly contrast the basic powers and structure of SLD's
communities of inquiry with three other kinds of local units: secondary
associations, participatory democratic councils such as idealized town
meetings, and the basic unit of public or private bureaucracies. Unlike
secondary associations in civil society, SLD's local units exercise substantial
power over the use of public resources in say, schools or policing.
Furthermore, SLD's local units must exercise more technical competence than
most secondary associations because they are charged with, and held accountable
for, using their powers responsibly and effectively.
SLD's local units differ from participatory democratic councils in three major
respects. First, participatory democratic bodies (Mansbridge 1980; Fishkin
1991) are often presumed to be omni-competent whereas SLD's local units are
functionally differentiated into, say, some local units for education and
others for public safety. SLD recognizes, indeed springs from, the complex
nature of modern social action and its division according to specialization
renders the burdens of effective participation more tractable. Second,
participatory democratic bodies are most often considered autonomous because
their decisions enjoy the legitimation that comes from unmediated citizen
voice. Local units in SLD, however, are held accountable according to external
performance criteria. Unlike participatory democracy, it is not enough that the
people speak; they have to know what they are talking about. Under SLD, those
units that fail to generate satisfactory outcomes are subject to external
remedial intervention (see Chapter 8). Finally, in perhaps what is less of a
difference than a further specification of participatory democracy, the
decision processes of SLD's local units are deliberative and pragmatic, whereas
participatory democrats may favor a variety of decision making that range from
consensus generating discussion to straight voting; once again, the decisions
in SLD are good because they work, not because they more accurately reflect the
interests and opinions of the people.
We have already discussed many of the differences between SLD's local units
and those of hierarchical bureaucracy implicitly, but restate two major
differences here. First, SLD includes lay citizens such as parents and
neighborhood residents in its decision making, whereas professional
bureaucracies seek insulation from such popular influence. Second, SLD imposes
an alternative logic of decision. Whereas command and control systems attempt
to determine optimal routines from their central offices and their research and
development sections, SLD devolves authority downward to local units
themselves, but demands in turn that local units exercise that authority in a
disciplined, deliberative, measurable, and public fashion.
Such
a community is formed, then, by citizens brought together by their common need,
say for safer neighborhoods or more effective schools. The purpose of the
community is initially limited to the satisfaction of that need. Since these
citizens are unsure about what they ought to do but know that they must do
something, much of the community's activities can be accurately characterized
as inquiry about appropriate actions: specification and prioritization
of problems facing the group, analysis about the causes of those problems, and
formulation strategies to solve the problem that take account of the groups
limited capabilities. These communities analyze, however, only in order to
guide the collective actions that will satisfy their common need. Since both
their analyses and actions will inevitably be imperfect, collective action is
also experimentalist--implementation reveals flaws in analysis that then
feed back into the mill of inquiry.[2]
Finally, the group makes decisions through deliberation. Each
participant aims to identify the most promising strategy to satisfy the need
which he shares with the rest. If there were an authoritative expert who could
optimal strategies, command rather than deliberation would be the appropriate
method of decision. If the participants had opposing interests, and the goal of
the group was to advance the greatest happiness of the greatest number, then
the aggregation of interests through voting might be an appropriate decision
method. But since the goal is to find solutions to common problems rather than
to aggregate opposing and since we stipulate that there is no dispositive
expert, then deliberation--full and open discussion of available options--seems
appropriate not only to decide the best course, but also to gain the allegiance
necessary to implement it. If, after ample deliberation, participants still
hold conflicting opinions regarding the optimal course of action, voting could
be an appropriate mechanism to tentatively settle the divergence and arrive at
the single opinion that collective action requires.[3]
SLD institutionalizes these organizational principles of deliberation,
experimentation, and inquiry through an ideal problem-solving procedure. At
this most abstract level, a community of inquiry in SLD is just a group of
pragmatic citizens who are trying to satisfy a collective need through the
following five-step (D1-D5),[4] iterated
procedure.
D1. Identify and Prioritize. Parties are taken to share a common but
vague concern, for instance the perception that their neighborhood is unsafe or
that their school could do better. Participants begin by dividing this general,
daunting concern into component problems such as a crack house on the corner or
the dilapidated school building. Prioritizing these component problems builds a
consensus on what exactly the problem is and yields a schedule that will assist
in the allocation of collective resources.
D2. In the second step of our ideal procedure, parties propose, justify, and
select a provisional strategy to address the concrete concern developed at
D1. The rational capacities--the instrumental reason--of parties is called into
play here. At this stage, the deliberative process should forge a number of
robust proposals, or strategies, to address the common concern. Each of these
proposals constitutes a hypothesis about how best to address the concrete
problem. This stage also requires parties to be reasonable. Some may attempt to
disguise their private interest as the general interest by making proposals
geared to advance their private interests, at the expense of other parties, at
the same time that it solves the general problem. Since parties are called on
to justify their proposals, and proposals which cannot be justified in terms of
the common good are excluded, we hope that deliberation between rational and
reasonable parties will generate a menu of strategies, each of which seems
(prospectively) effective and fair by the lights of everyone.
A complete proposal will have at least these elements: a set of tasks to be
done, a division of labor which assigns tasks to parties, a set of expectations
about what each of the tasks will accomplish vis-a-vis the concrete issue
before the group, and provisional methods with which to assess whether or not
parties completed assigned tasks and whether successfully accomplished tasks
yielded expected effects.
With a set of seemingly effective proposals before them, the parties
deliberate again to select one that seems more promising than the rest.
Deliberation again is geared toward choosing an effective and fair proposal,
and geared toward achieving consensus. Since a menu of proposals, all of which
seems both fair and effective, is before the group, there is no reason to
suppose that consensus can be achieved. Each proposal is, after all, a guess
about what the world is like and how it will respond to human action; these are
complex matters about which reasonable and bright people often differ. One
unobjectionable way to proceed, therefore, is to vote on the proposals and
adopt the majority or plurality winner as the provisional hypothesis for the
group. In giving his vote, each party gives her guess about which of these
solutions is most effective, and all realize that the social choice which
results is their best guess and nothing more.
D3. Implementation is the third step of the procedure. Parties attempt
to carry out the tasks assigned to them by the proposal selected in D2. Each
may fail to carry out her task for a variety of reasons, for instance she may
shirk or the task itself may be more demanding than anticipated.
D4. Monitoring and Evaluation. Following implementation, the parties
deliberate about how things went. It is hoped that the resources for
intersubjective agreement on assessment will have been progressively
constructed at D1 and D2.
Working backwards, they first assess the degree to which component tasks of
the solution were successfully implemented. So, the group assesses whether or
not particular parties failed to deliver on their commitments and whether the
tasks assigned were too demanding. This level of assessment yields information
about the reliability and capacities of various parties.
The group then evaluates whether or not the accomplished tasks yielded
expected benefits. Each task in the solution adopted at D2 itself represents a
hypothesis about the intended effects of various course of action. The action
at D3 can be viewed as executing experiments formulated at D2, and one of the
points of evaluation is to attempt to ascertain the validity of those rough
hypotheses. Finally, the agenda (D1) is only a provisional guess about the
components which make up their common concern and may itself require revision
in light of evidence. With a full evaluation in hand, parties can assess their
entire solution in parts: what worked, what did not, whether new strategies
need to be formulated, and whether the agenda needs to be revised.
D5. Reiteration. The experience of D1-D3 made public through
deliberation at D4 equips the group to attempt another round at the solution of
its common problem. Since we expect the solution to be neither a complete
failure nor a complete success, the parties will be motivated to continue some
process of cooperative action to the extent that they still have a problem in
common. So the ideal procedure is iterative.
We can expect the quality of later iterations of the proposal formulation and
selection (D2) and evaluation (D4) stages to increase for three general
reasons. First, previous rounds generate more public information about
each of the parties and about the common problem. So, initial expectations
about the skills, reliability, and trustworthiness of each is subject to
revision through the addition of information. Also, experience and public
reflection upon attempts to address the problem yields information about its
contours. Second, we can expect the limited practical rationality
(described in Chapter 6 as instrumental reason about social problems) and
reasonable (described in Chapter 6 as the power of moral restraint to
subject oneself to the bonds of a better argument) capacities of all parties to
improve because of the principle of learning by doing. Since these three
features--the amount and quality of information, the rational capacities of
parties, and their reasonableness--largely determine the character of
deliberation, we can expect future rounds of proposal generation and evaluation
to improve.
Implementation (D3) will improve for two rough reasons. First, the parties
themselves, again through the principle of learning by doing, will gain the
knowledge and skills required for various implementation tasks. Second, public
knowledge of the skill level and reliability of each increases with future
iterations, and so the tasks assigned to parties will become more suited to
individual interests and skills.
With this abstract procedure in hand, we now consider its appearance in
decentralizing reforms to policing and public schools in Chicago. The following
two sections show how those organizations have implemented deliberative
problem-solving in their local units of public action. We can think of
community policing reform as creating 279 communities of inquiry--one in each
beat, and educational reform as creating some 560 school-based communities of
inquiry. These examples show not only how the abstract procedure can be
operationalized to solve complex public problems, but also demonstrates that
agencies which until quite recently appeared to be among the most retrograde
and hierarchical have indeed embarked upon this reform trajectory.
Policing reformers have operationalized the ideal deliberative problem solving
twice in the course of developing community policing institutions. The first
was the Joint Community Police Training (J.C.P.T.) and organizing program for
neighborhood residents that took place from 1995 to 1996, discussed briefly in
Chapter 5 above. The deliberative fora of monthly neighborhood beat meetings
between residents and patrol officers and the actions that those discussions
produce constitutes the second implementation of deliberative inquiry in
community policing.
Though a training initiative in name, J.C.P.T. actually involved substantial
community organizing and problem-solving activity. To repeat briefly, the city
funded approximately 100 field staff--some police but mostly "civilian"
personnel--to mobilize residents around issues of community policing throughout
the city. In each beat, designated organizers were charged with generating
resident participation by working with existing neighborhood based
organizations (NBOs) or by direct door-to-door canvassing of neighborhood
residents. Trainers would then lead residents through a five meeting
problem-solving curriculum over four months' time. At the end of this period
(determined by funding constraints), program-designers hoped that residents
would be able to sustain problem-solving involvement without professional
staffing or support.
J.C.P.T. embraced learning-by-doing as its pedagogical method, and the "doing"
followed exactly the steps of our deliberative problem solving procedure. In
the first session, trainers facilitated discussion among residents to select a
the most important crime and disorder problem in their neighborhood and to
analyze the causes of that problem (D1). Program designers stipulated that
situation had to possess three features in order to qualify as a problem: (i)
it had to occur in a definable location, (ii) there had to be identifiable
offenders, and (iii) victims also had to be identifiable by at least categories
(e.g. motorists). These three aspects of problematic situations then form a
"crime-triangle."
In the second session, residents and patrol officers developed strategies to
attack each of the three sides of the triangle (D2). Strategies often involved
police and resident capabilities such as increasing patrol visibility,
deployment of un-marked units, petitions, negotiations, and demonstrations.
Often, however, they often called upon participants to leverage
resources not readily available to the group--various city services, an
alderman's office, civic organizations. Strategies also included dividing the
labor of implementation among group participants.
Between the second and third sessions, participants attempted to implement
these strategies (D3), and in the third session participants discussed the
successes of their efforts (D4), and devised new strategies if those chosen in
the second session seemed not to be working (D5). The fourth meeting consisted
of a wrap-up session to celebrate any victories, solidify resident commitment
to this problem-solving process by reviewing often surprising accomplishments,
and set in place resident leadership who would take responsibility for
continuing the process absent staff support.
The figure below is a worksheet from J.C.P.T. training materials on which
participants could record each of steps of public action. To illustrate just
how closely the ideal deliberative problem solving process was implemented in
J.C.P.T., I have the marked spaces in the form as they correspond to the steps
(D1-D5) laid out above (7.1):
Figure 7.1. J.C.P.T. Problem-Solving Worksheets.
In addition to this short-lived organizing and training program, deliberative
problem-solving has also being formally implemented at the core of Chicago
police operations. A General Order to the patrol division--the rank-and-file of
the Police Department--issued in April 1996 institutionalizes the procedure
through three complimentary devices: beat meetings, beat teams, and a set of
supervised instructions on problem-solving (Chicago Police Department 1996;
Fung 1997c). Once again, beat meetings are public sessions typically held
monthly in each of the city's 279 police beats. Similar to J.C.P.T., police and
residents are to use these sessions to identify crime and disorder problems in
the neighborhood, develop and implement strategies, evaluate results, and
re-iterate these problem solving steps. At this stage, however, beat-meeting
problem-solving is typically less effective than J.C.P.T. due to lack of
trained facilitation. Whereas J.C.P.T. staff had themselves received
substantial training in the goals and procedures of problem solving, the police
and residents who attend beat meetings usually have not benefited from such
orientation.
The 1996 General Order directs police to form "beat teams" that consist of
officers directly responsible for serving each beat--typically five patrol
officers and their sergeant. These officers meet regularly in "beat team
meetings" to choose priority problems, develop strategies, and discuss
effectiveness of various strategies. Though orders instruct them to "give...
special attention to the problems identified during beat community meetings" in
the selection of priorities, police may over-ride these resident
recommendations because "beat community meetings may not be representative of
the entire beat, and the problems they identify may not be representative of
the problems on the beat." Community side participants can respond
(deliberatively) to objectionable police decisions, however, at successive beat
meetings.
The General Order requires line-level police to document their problem solving
activities to enable monitoring and improvement of future effort through
post-facto analysis. By capturing action on the written page, these forms show
how decision authority has been extensively devolved to operational units and
that those units follow deliberative problem-solving at laid out in 7.1 above.
Consider the "beat plan form" which might more appropriately be labeled a
"problem-plan form" since a single beat typically has three or four such forms
open at any given time--one for each open problem. As with J.C.P.T., the form
leaves complete operational discretion to patrol officers, yet imposes the
generative structure of cognition and action which I have described as
deliberative problem solving:
Figure 7.2. Beat Plan Form, Side 1
Figure 7.3. Beat Plan Form, Side 2
In the space that I have marked "D1," officers record the specific origins of
this problem as a priority issue. In most cases, problems become priorities
when they are raised at community beat meetings. In the spaces marked "D2,"
police in the beat team develop a series of strategies to address these
problems through analysis of the problem, a guess about the time required to
address it, and particular action items (strategies) together with the
assignment and definition of tasks necessary to implement those strategies.
Moving to the second side of this form, officers continuously monitor each
other's implementation efforts and the effectiveness of those efforts in the
space marked "D3/D4." Finally, in the space marked "D4," officers record the
results of summary self-assessment after the problem has been "solved."
Recall (from chapter 4) that the a 1988 Illinois Public Law radically
decentralized the governance structure of the Chicago Public Schools. For each
of the 560 elementary schools, the legislation created an elected Local School
Council composed of the principal, two teachers, six parents, and two community
members. Each LSC was empowered to hire and fire the principal, allocate the
school's discretionary monies, and help determine the allocation of staff
resources. The law also requires each LSC to develop a School Improvement Plan
(SIP) that guides the exercise of these powers.[5]
According to the general language of the enacting legislation, each school's
SIP is three year plan "to improve educational quality."[6] In practice, it is a working document, updated yearly,
that states a school's vision of itself as an excellent educational
institution, lists the most urgent steps necessary to move the institution to
that point, and assigns those tasks to particular individuals in the LSC or
staff. The principal of a school typically develops the plan in consultation
with school staff, the LSC, and other members of the community, and the LSC
must approve the document each year. SIPs are modified annually according to
changing circumstances and results of implementation efforts, and so compose
part of a "continuous planning" process.[7] The
changing activities of staff, LSC members, and others who work with the school
can be broadly viewed as the actualization of this ever-changing plan.
In order to ease the task of composing SIPs, an office of the CPS recommended
a format that nearly every school has chosen to follow; though some school have
much better SIPs than others, they all look similar. This paperwork reflects,
and thus allows us to infer, the character of Local School Council deliberative
problem solving. The form has four sections. In the first section, a school
states it vision for itself and the final section records budgeting decisions.
Sections two and three document a school's problem solving activities and thus
are most salient here.
In the second section of an SIP, titled "Analysis of Current Conditions," each
school lists its priority activity areas, and then reflects upon the strengths
and weaknesses of that area. This section corresponds to the prioritization
(D1) and evaluation of previous strategies (D4) of the ideal deliberative
experimentalist procedure. The following figure reflects one elementary
school's analysis of it own language arts program:
Figure 7.4. School Improvement Plan Excerpt A
The third section of the SIP form, labeled "Establishing Goals, Plans, and
Monitoring Progress," lists strategies, tasks necessary to implement them,
assignment of those tasks, and monitoring provisions; it documents each LSC
problem solving plans at stages D2-D4. Here is part of the same school's SIP
that addresses the aspects of the reading program identified as weak earlier in
its SIP:
Figure 7.5. School Improvement Plan Excerpt B
So, lack of reading comprehension is one priority problem (D1) and the school
has selected Accelerated Reading and SQUIRT (Super Quiet Uninterrupted Reading
Time) programs as its strategies to address that problem (D2). To implement
this strategy (D2), classroom teachers will emphasize students' reading aloud,
more time will be devoted to silent reading, and use of the existing
computerized "Accelerated Reader Program" will be increased. If this
computerized instruction seems fruitful, the school will expand its facilities.
The second column lists multiple monitoring activities, again devised by school
personnel, that involves student testing, teacher self-assessment, and
principal supervision (D4). Finally, the third column lists target dates for
monitoring and implementation (D4).
Updated annually, SIPs serve as a base-line plan to guide staff and LSC
activities throughout the school year. LSCs monitor progress on the document at
their monthly meetings, use SIP's goals to allocate monies and a tool for
principal evaluation, and implement many of its objectives in on-going
committees. It is thus a product, record, and motor of deliberative problem
solving activities in each Chicago public school. As with the two illustrations
from policing, the SIP's wide-open structure decisively illustrates that its
purpose is not to assure compliance with particular instructions, but to
inspire ground level actors to articulate their views about what most needs to
be done and how best to do those things.
Having described the characteristics of the citizen in the previous chapter
and the structure of the local unit in which she participates above, we turn
now to the third and final element in the architecture of Street Level
Democracy. In the next chapter, we describe the nature of central authority
which both supports problem solving deliberations of SLD's communities of
inquiry and holds them accountable to high standards of moral and practical
performance. Unlike most radical democratic proposals, SLD does not advocate
simple decentralization, but rather proposes a kind of mutually supporting
federation. The next chapter describes the functions and structure of the hub
of SLD's federations--its "supportive center."
[1] This is of course a common form of
argument for the analytic origin of the state out of various states of nature.
The most prominent examples come from Locke (1960) in his Second Treatise on
Government and Rousseau (1987) from the Social Contract. More
contemporary versions come from Robert Nozick (1974: 10-53) and John Dewey
(1927: 29-36).
[2] So, John Dewey suggests that "policies and
proposals for social action be treated as working hypotheses, not as programs
to be rigidly adhered to and executed. They will be experimental in the sense
that they will be entertained subject to the constant and well-equipped
observation of the consequences they entail when acted upon, and subject to
ready and flexible revision in light of observed consequences"(1927: 203).
[3] On this epistemic conception of
voting, see J.J. Rousseau in The Social Contract:
When a law is proposed in the people's assembly, what is asked of them is not
precisely whether they approve or reject, but whether or not it conforms to the
general will that is theirs. Each man, in giving his vote, states his opinion
on the matter, and the declaration of the general will is drawn from the
counting of votes. (SC, IV.2.viii)
See also Cohen (1986); Fung (1995), Joshua Cohen, "An Epistemic Conception of
Democracy," Ethics 97, no. 1 (Oct. 1986): 26-38. For an experimentalist
conception of epistemic democracy, see Fung (1995).
[4] This procedures is a pragmatic version of
the more general deliberative procedure offered by Joshua Cohen (1989).
[5] Illinois Compiled Code of Statutes,
Chapter 105, Article 34 (1996).
[6] Illinois Compiled Code of Statutes,
Chapter 105, Article 34, para. 2.4, "School Improvement Plan" (1996).
[7] John Dewey, "The Economic Basis of the New
Society," in The Political Writings, Debra Morris and Ian Shapiro eds.
(Cambridge: Hackett Publishing Co., 1993). p. 171.