Big questions in Public Network Management research
Journal of Public Administration Research and Theory
Print Media Subtitle:
Authors:
Authors:
Volume:
Issue:
Pagination:
ISSN:
Subject Terms:
J-PART
Lawrence
Jul 2001
Robert Agranoff
Michael McGuire
11
3
295-326
10531858
Public administration
Management
Abstract:
As the use of networks in public management increases, more and larger questions
expand this research arena. In many ways, public network management is in search of a
paradigm equivalent to the hierarchical-organizational authority paradigm of bureaucratic
management. We raise and offer preliminary answers to seven metaquestions that address
the nature of network management tasks, group process in collaboration, flexibility of
networks, self-responsibility and public agency accountability, the cohesive factor of
networks, power and its effect on group problem resolution, and the results of network
management.
Copyright Transaction Publishers Jul 2001
Full Text:
ABSTRACT
As the use of networks in public management increases, more and larger questions
expand this research arena. In many ways, public network management is in search
of a paradigm equivalent to the hierarchical-organizational authority paradigm of
bureaucratic management. We raise and offer preliminary answers to seven
metaquestions that address the nature of network management tasks, group process
in collaboration, flexibility of networks, self-responsibility and public agency
accountability, the cohesive factor of networks, power and its effect on group
problem resolution, and the results of network management. The light we shed on
these issues by examining the black box of networks is designed to contribute to
building an empirically derived knowledge base of network management.
As long as a branch of science offers an abundance of problems, so long is it alive; a
lack of problems foreshadows extinction or the cessation of independent
development .... It is by the solution of problems that the investigator tests the
temper of his steel; he finds new methods and new outlooks, and gains a wider and
1
freer horizon.
We share the optimism and enthusiasm of discovery expressed by mathematician David
Hilbert one hundred years ago in the above quote. At that time, he outlined twenty-three
major problems to be studied over the next century in what is still history's most
influential speech about mathematics. He conveyed to the scientific community that
unsolved problems are a sign of vitality and that research on the most difficult problems
was necessary: "for what is clear and easily comprehended attracts, the complicated
repels us" (Hilbert 1902). In the spirit of Hilbert, we attempt to assess, in at least a
preliminary way, the operational knowledge base of the public network management
field. There are many more questions than answers in network management and, since
just a few can be covered in these pages, we selectively raise some core questions-and
suggest preliminary answers to these questions-that frame any potential theory of
network and collaborative management.
That the study of interorganizational network management is temporally relevant needs
little justification. Knowledge as a factor in social and economic production is making
land, labor, and capital increasingly secondary. Knowledge work increasingly involves
problem solving, problem identifying, and strategic brokering, which constitutes the
symbolic-analytic work identified by Reich (1991, 177). Knowledge creation and
learning occur best when people are reacting to someone else's thinking; thus networks
and other communities of learning are part of today's knowledge program architecture
(Pasternack and Viscio 1998). Networks or teamwork are also the vehicles for arranging
the human capital needed to create new technologies, including those that are managerial.
These assumptions place networks and their management at center stage. Networks, as
the term is used in the literature, typically refers to multiorganizational arrangements for
solving problems that cannot be achiev! ed, or achieved easily, by single organizations.
Public management networks are led or managed by government representatives. Simply
put, networks constitute emergent phenomena that are distinctive managerial vehicles and
that offer challenges for the single organization and its management.
Since the focus here is on interactions that involve public organizations, we draw
conceptual questions and suggest partial answers from a wide range of the policy and
management literature, including our own (Agranoff and McGuire 1998; 1999a; 1999b),
as well as from more contemporary analyses of interorganizational and intergovernmental
relations. In other words, we seek to organize the state of the field on management of
public networks by reviewing extant studies. The results suggest that more is known than
might have been apparent about a range of issues. This accounting or distillation suggests
more detailed but no less important questions. This article presents the results of this
broad examination of the research record.
Increasingly, the capacities required to operate successfully in network settings are
different from the capacities needed to succeed at managing a single organization. The
classical, mostly intraorganizational-inspired management perspective that has guided
public administration for more than a century is simply inapplicable for
multiorganizational, multigovernmental, and multisectoral forms of governing. If
network management is a function distinct from that of hierarchical management, as
2
many eminent scholars suggest (Kettl 1996; Milward 1994; O'Toole 1997a), then focused
research and improved conceptualization on this core public activity must be accelerated.
Indeed, as the use of networks in public management increases, more questions regarding
this research arena need to be answered. Weiner (1990) predicts organizational
management and transorganizational (network) management will become two parallel but
overlapping streams within administrative theory, relying heavil! y on both behavioral
science and management science research and practice. Network management is thus in
need of a knowledge base equivalent to the hierarchical organizational authority
paradigm of bureaucratic management. The primary challenge for empirical research in
public administration is to functionally separate how knowledge issues, problems, and
technologies are identified, brokered, and solved in networks.
We will examine seven basic operating questions related to network management. First,
we will raise the issue of whether there is a functional network equivalent to traditional
management procedure. Is there a POSDCORB in network management? Second, we
will investigate the group decision approaches in network management. Are they the
same as those derived from applied behavioral science or are other learning and decision
processes at work? Third, we will address the issue of flexibility in networks. Do
networks provide flexibility beyond rapid adaptation or procedural accommodation?
Fourth, we will discuss the venerable accountability issue. In what ways do networks
employ mutual self-responsibility, and does this substitute for the loss of public agency
accountability? Fifth, we will explore issues of trust, common purpose, mutual
dependency, and other issues in networks. What is the cohesion factor in networks that is
equivalent to legal-rational authority in organization! s? Sixth, we will consider the issue
of power in networks, which is often neglected, particularly its effect on the synergistic
creativity that reciprocal relationships hope to produce. How does power come into play
and what is its effect on group problem resolution? Seventh, we will pursue the issue of
network results or network productivity. Do public management networks produce results
that would otherwise not have occurred? The light we will shed on these issues by
examining the process-the black box-of networks may contribute to building an
empirically derived knowledge base of network management.
What are the critical functional equivalents to traditional management processes? Is there
a POSDCORB equivalent set of tasks that replaces the standard planning, organizing, and
so forth?
When public or private sector managers are observed operating in their home
organizations, one can clearly describe, in the most general sense, what activities are
performed at any given time and, for the most part, why they are performed. A
vocabulary or nomenclature that has developed and has been refined over many decades
is available to us to help elucidate the manager's daily activities within the single
organization bureaucratic structure. For example, many textbook descriptions of
management include tasks such as planning, organizing, and leading, and for more than
sixty years the field has relied, to varying degrees, on a simple acronym-POSDCORB-to
distill management activities into seven basic tasks. Furthermore, managers have certain
remedies to fix single organizational problems. If an organization appears to lack
direction, planning may be the answer; if communication and coordination are poor
3
across departments within an organization, a reorganization may be in! order. Scholars
have identified some behaviors used in managing networks and when it is best to use
them, but a similar functional and conceptual parallel to traditional management
processes has not yet been developed (Hanf and O'Toole 1992). We open with a purely
suggestive grouping of network management behaviors as a way to summarize the
literature in a substantive way.
One class of behaviors addressed in the literature involves the general tasks of activation.
Activating includes the process of identifying participants for the network (Lipnack and
Stamps 1994) and stakeholders in the network (Gray 1989) as well as tapping the skills,
knowledge, and resources of these persons (Agranoff and McGuire 1999a). All interests
should be included in network processes (Innes and Booher 1999). Network managers
arrange (Klijn and Teisman 1997), stabilize as much as possible (Stone 1999), nurture
(O'Toole 1988), and integrate (Lipnack and Stamps 1994) the network structure.
Activation is a critical component of network management because resources like money,
information, and expertise can be the integrating mechanisms of networks. According to
Scharpf (1978), selective activation of potential participants in the network is "an
essential prerequisite for successful interorganizational policy formation and policy
implementation" and, if it is performed ! correctly, is based on "the correct identification
of 'necessary participants' in policy-congruent networks," as well as the willingness of
these potential participants to devote resources to the network and not be influenced by
actors who may have other interests at stake (p. 364).
Activating the right players with the right resources is the crucial task of governing in
cities through coalitions of public and nongovernmental organizations. We know from
the urban regime literature (Stone 1989 and 1999; Stoker and Mossberger 1994) that
governing coalitions in cities must "bring together resources that are adequate to address
the identifying agenda," otherwise the arrangements will not be effective (Stone 1999, 3).
Such regimes operate according to network principles, but do so informally, usually
through nongovernmental organizations' pursuit of mutually derived agendas. Since the
probability of activating a weak link in the chain-involving the wrong player or the wrong
mix of resources-is greater than zero (see Landau 1991), this task is of crucial importance
to network effectiveness. Deactivation also appears to be an important component of
network management. For this task, network structures are rearranged and shifted if they
are not performing as d! esired. The most common prescription in this regard is to
introduce new actors as a means of changing the network dynamic, shifting the influence
of existing actors, and facilitating fluid leadership roles (Klijn 1996; O'Toole 1988; Klin
and Teisman 1997; Termeer and Koppenjan 1997).
Other management behaviors identified in the literature are consistent with the general
task of framing. Like activation, framing is used both during the formation of the network
and as a management tool when network effectiveness diminishes or is suboptimal.
Compared to activation, framing is a more subtle task but is just as necessary. Framing
involves establishing and influencing the operating rules of the network (Mandell, 1990;
Klijn 1996; Gray 1989), influencing its prevailing values and norms (Kickert and
Koppenjan 1997; O'Toole 1997b), and altering the perceptions of the network
participants (Termeer and Koppenjan 1997). The rules of interaction among participants
4
and the perception of that interaction are malleable, and conscious change in the latter
through framing can jump start an unproductive network. A manager can frame the
network context by introducing new ideas to the network (Kickert and Koppenjan 1997),
thereby creating (or celebrating) a shared purpose o! r vision (Mandell 1988; Gray 1989;
Lipnack and Stamps 1994; O'Toole 1997b). In this regard, the manager might offer
suggestions for looking at a problem differently or might recommend an alternative
decision-making mechanism. Although not a simple or straightforward managerial task,
framing "gives shape to purposes, and . . . has great influence in the alignment of various
forms of engagement" (Stone 1999, 7).
In addition to activating the network and framing the interaction of network participants,
the literature indicates that network managers need to induce individuals to make a
commitment to the joint undertaking-and to keep that commitment. A manager must
often sell an idea to potential network participants to secure commitment; support for the
network and its purposes) must be built (Gray 1989; Innes and Booher 1999; Mandell
1990). Mobilizing "requires a view of the strategic whole and an ability to develop and
achieve a set of common objectives based on this whole" (Mandell 1988, 33). Managers
build support for the network by mobilizing organizations (Mossberger and Hale 1999)
and coalitions (Kickert and Koppenjan 1997), and by forging an agreement on the role
and scope of network operations (see Benson 1975). Additionally, the ability to manage
networks is related to the internal support and cooperation of the manager's primary
organization. Managers in city government, f! or example, should have the cooperation of
city councils and their chief executives as a means to more confidently engage in
networking and achieve the strategic purpose at hand (Agranoff and McGuire 1999a). We
distinguish between behaviors associated with activation and mobilization by including in
the latter the important finding that a human relations component-motivating, inspiring,
inducing commitment-is a standard part of the network management repertoire.
Research on network management also addresses synthesizing the network by creating
the environment and enhancing the conditions for favorable, productive interaction
among network participants. Managers must find a way to blend the various participantseach with conflicting goals or different perceptions or dissimilar values-to fulfill the
strategic purpose of the network. The network manager seeks to achieve cooperation
between actors while preventing, minimizing, or removing blockages to cooperation.
This steering of network processes is tantamount to game management in the sense that
the result of the network process "derives from the interaction between the strategies of
all actors involved" (Klijn and Teisman 1997, 99). The strategies of each network
participant and the outcomes of those strategies are influenced by the patterns of relations
and interactions that have developed in the network. Relationships and interactions that
result in achievement of the network pur! pose are the aim of the network manager, and
important management behaviors include facilitating and furthering interaction among
participants (Agranoff and McGuire 1999a; Kickert and Koppenjan 1997); reducing
complexity and uncertainty by promoting information exchange (Innes and Booher 1999;
O'Toole 1988 and 1997b; Mossberger and Hale 1999); changing incentives to
cooperation (Kickert et al. 1997); developing new rules and procedures of interaction
(Klijn 1996; Termeer and Koppenjan 1997); changing positions, relations, and roles of
participants (Klijn 1996; Kickert et al. 1997); helping the network to be self-- organizing
5
(Innes and Booher 1999); and engendering effective communication among participants
(Mandell 1990; Kickert et al. 1997). Synthesizing seeks to lower the cost of interaction,
which can be substantial in network settings.
Although we distinguish among four different management behaviors or tasks, the tasks
identified in the network management literature are nearly seamless in their applicability.
That is, just as a manager in a single organization may lead, plan, and organize
simultaneously within the context of a particular goal, so too are multiple behaviors
utilized in network settings. Synthesizing and activating are management tasks often used
in conjunction. For example, the removal of a network participant can have myriad
effects on the network. Alternatively, if the attitudes and behaviors of some participants
are stifling productive interaction, the manager might try to create an environment within
which greater interaction can take place by improving the incentives of participation or
reassigning (facilitating the reassignment of) various roles within the network. Similarly,
synthesizing the network is often accompanied by refraining, or a conscious attempt to
change the network pa! rticipants' perception of the interaction changes. After or during
deactivation or refraining, managers must mobilize support for the changes, reestablish
the purpose of the network, and make sure all participants are on board.
Beyond identifying these tasks, research must discover the set of managerial decision
rules that network managers currently follow in their quest for performance. What does a
network manager do when trust is lacking among the actors? What should she do? What
does a network manager do when the goals of the network actors are conflicting? What
should he do? What are the best mechanisms to improve decision making in the network,
or to increase productivity, or to address a slacker in the network? Broadly, what are the
critical occurrences or situations that affect network operations, and what are the
preferred responses to these situations? Extensive observation and in-depth interviewing
are required to answer these critical questions.
Are the approaches to groupware-that is, group development that reaches a mutual
understanding and transcends the more immediate and interactive bases of coordination
and communication through hierarchy-substantively different from those derived from
the applied behavioral science approaches that emanate from human relations research?
Groupware describes interagency task group development for reaching jointly arrived at
solutions. Hierarchies have become one among many means to coordinate and control
actions across people, knowledge, time, and space. This type of groupware (which is
different from computer decision support software) produces more immediate and
interactive bases of coordination than does hierarchy. It incorporates principles that are
qualitatively different from interorganizational coordination or mutual adjustment.
Groupware requires that "virtuous circles" have a multiplier effect as multiple cultures,
procedures, and divisions of labor be incorporated into the network (Clegg and Hardy
1996, 11). Decisions are deliberative, creative, and group based, as organization
representatives shed some of their ideology and forge redefined problems, a mutual
course of action, a multi-- organizational implementation technology, and diversified
methods of resource acquisition. Groupware leads to syn! ergistic products. Operating by
groupware clearly requires the application of standard applied behavioral science
techniques, but is more required? Myron Weiner (1990, 456) suggests that in
6
transorganizational (network) management, techniques similar to organization
management are normally employed: group problem solving, force-field analysis, action
planning, team building, process consultation, and others. However, the two domains are
distinctive in at least two ways. First, in network management, empowerment is based on
information rather than on authority. Second, in network management, existing
organizational structures are dependent variables for network systems. The several
organizations working together can be fashioned into new systems, using the flow of
information to link the transorganizational system.
Social capital is one possible ingredient in the interorganizational information flows that
are necessary for developing groupware. Fountain (1998, 104) refers to social capital as
the stock that is created when a group of organizations develops the ability to work
together for mutual productive gain (for related definitions see Coleman 1990; Putnam
1993). Social capital is essential for groups of disparate representatives to work toward
sharing resources held by individual organizations:
Like physical capital and human capital-tools and training that enhance individual
productivity-'social capital' refers to features of social organization, such as networks,
norms, and trust, that facilitate coordination and cooperation for mutual benefit. The
notion of social capital extends our understanding of 'cooperation' or 'collaboration' in
two significant ways. First, linking cooperation to the economic concept 'capital' signals
the investment or growth potential of a group's ability to work jointly. Second, the
concept identifies the structure created from collaborative effort as capital (Fountain
1998, 105).
Fountain argues that the important elements of social capital are trust, norms, and
operations of the network, which are closely related to the values and objectives of the
actors. She suggests that tools that enhance the creation of social capital need to be part
of policies that promote innovation and productivity growth (p. 113). As a result, social
capital has been associated with increased innovation and economic growth within the
biotechnology and information industry, as well as revitalization in manufacturing.
Shared learning is a fundamental component of groupware. Collaborative processes
among organizational representatives can be characterized as joint learning systems.
Inner and Booher (1999, 3) suggest that a new mind-set is needed to overcome traditional
notions of producing specific agreements and actions. The idea is similar to Senge's
(1990, 3) learning organization, "where people continually expand their capacity to create
the results they truly desire, where new and expansive patterns of thinking are nurtured,
where collective aspiration is set free, and where people are continually learning how to
learn together." To Inner and Booher, the most important consequences occur not at the
end, but during the discussion process itself. An environment conducive to learning is
created when network members follow principles of civil discourse; when all are listened
to; and when conditions of sincerity, comprehensibility, accuracy, and legitimacy are
met. Collaborative discussio! ns involve creating shared meaning, pursuing mechanisms
other than arguing and debating, bringing out added knowledge, and formulating ideas
and processes of joint action that can result in "good answers through process" (Inner and
Booher 1999, 5). Senge (1990) suggests that learning organizations require five core
disciplines: personal mastery, mental models, shared vision, team learning, and systems
7
thinking. Like learning organizations, networks require similar collective "cognitive
capabilities" since "the intelligence of a network lies in the patterns of relationships
among its members" (Lipnack and Stamps 1994, 210).
Groupware is also developed through negotiation. Bardach (1998, 232) refers to a
"culture of joint problem solving" that includes an "ethos that values equality,
adaptability, discretion and results." Part of this ethos is to overcome bureaucratic
tendencies (hierarchy, stability, obedience, procedures) through a lively sense of valuecreating possibilities. Bardach (p. 238) also suggests that "collaboration is a matter of
exhortation, explication, persuasion, give and take. To collaborate is to negotiate."
Negotiations have to take into account all interests-personal, organizational, partner
organizations, and the collective (Galaskiewicz and Zaheer 1999). Bardach (pp. 245-46)
suggests that there may be stages in network negotiations that are analogous to the
contentiousness-stalemate-constructive dynamic in private-- sector negotiations: the
mutual understanding of a latent opportunity to create public value; the recognition that
collective efforts and risks are manageabl! e; and the process of collaboration proceeds.
Negotiations in networks seek consensus only after members have fully explored the
issues and interests, and only after significant effort has been made to find creative
responses to differences (Innes and Booher 1999). Such processes are supported by trustbased accommodations and by the dualism manifest in the agency delegate to the
network and the network delegate to the agency being the same person. Weiner (1990)
suggests that whether collaborating or negotiating, such dual loyalty necessitates a totally
new mind-set and value stance for managers who work within transorganizational
systems.
An example of groupware at work occurred as a collaborative task force of governmental
and nongovernmental leaders from Iowa, under the sponsorship of the National
Governors' Association, created a strategic plan for rural development. The Rural Policy
Academy collaborative-comprising state program administrators, Agricultural Extension
Service executives, agriculture and agribusiness representatives, business agency heads
of peak interest associations, credit and capital executives, and civic leaders-approached
its task as a temporary organization, going through process facilitation, brainstorming,
problem identification, team building, survey feedback, and so on. Most essential were
five task teams or work groups, involving expanded memberships and built around
mutually derived goals such as business development and retention and public
infrastructure. Over time, each work group found that individual and organizational
positions receded in favor of new group understanding! . Members also learned about
different policy arenas and developed new mind-sets regarding rural problems. For
example, few members of either the business or the agriculture work team knew about
the difficulties of access to venture capital or the diminution of an entrepreneurial ethos
among young Iowans. Obviously, the work groups and the entire Rural Policy Academy
could not incorporate all individual and organizational interests brought to the table.
Healthy doses of negotiation ensued over dearly held interests and concerns, including
the future of small hospitals in nonmetropolitan areas and the consolidation of local
government facilities, offices, and programs. Complex issues like these required
considerable negotiation to reach any policy recommendations (Agranoff 1991). Despite
8
such obstacles, groupware was developed through difficult negotiation, and social capital
was built as members saw different ways of looking at problems.
A number of behavioral process questions remain that could be answered with systematic
observation of network processes. The foregoing discussion suggests that information is
an important ingredient in interorganizational transactions. Is the information that
participants bring to the network an empowering or an overpowering force in
collaborative discussion? If the techniques of applied behavioral science are virtually the
same but the clustering is different, what is the distinctive clustering in networking?
Social capital's contribution to groupware also raises questions. Is social capital necessary
to reach collaboration or can other features (resources, legal requirements, perception of
mutual gain) make up for capital deficiencies? Is social capital sufficient to reach mutual
outcomes? Can collaboration occur when social capital is low or absent from an
interagency situation? With respect to joint learning, are there new and different
dimensions of such learning? If j! oint learning through process is the key, what is the
process? If negotiations are part of the collaborative process, to what extent are network
negotiations different from those between single organizations? In the process, when does
joint, learning-based, mutual decision making end and negotiation begin? Does a
negotiation imbalance (too much negotiation) reduce the chances of creative synergy?
Most important, how does negotiation within the network contribute to the development
of group abilities to recognize and solve problems?
How do networks provide flexibility beyond rapid adaptation or procedural
accommodation?
Flexibility is one of the most venerable attributes of networking. The search for flexibility
has been bound-up to a considerable degree with the need for different institutional
frameworks that facilitate mobility of resources, rapid access to skills and technology,
efforts to stabilize market or systemic shocks on small organizations, and policing of
unfair competition (Piore and Sabel 1984, 29-35). Networks are said to facilitate the
arrangement of entities in a way that falls somewhere between the openness of the market
and the rigidity of the hierarchy. As Powell (1990, 301) suggests, "Networks are `lighter
on their feet' than hierarchies. In network modes of resource allocation, transactions occur
neither through discrete exchanges or by administrative fiat, but through networks of
individuals engaged in reciprocal, preferential, mutually supportive actions." Adaptive
efficiency is the theme Alter and Hage (1993, 39) use to depict interorganizational
adaptation associ! ated with "quality, flexibility and innovativeness" (p. 14) because
networks are superior to individual organizations in achieving such efficiencies.
Sengenberger and Pyke's (1992) assessment of networks in industrial districts, involving
regional and municipal governments along with private-- sector enterprises, highlights
the advantages of flexibility:
At their best, they represent a type of industrial organization that meets competitive
challenges through differentiated high-quality products, flexibility of adjustment, and the
ability for innovation. The ability to offer quality, design flair, choice, flexibility, speed,
and innovation is itself a product of a particular kind of organization, based, as we have
indicated, on a peculiar mixture of competition and co-operation. Often people refer to
the combination of flexibility and specialized production units typical of industrial
9
districts as "flexible specialization" and contrast it to Fordism, the inflexible organization
of production on mass market lines, employing dedicated machines and specialized, often
unskilled or semi-skilled, workers (P. 5).
Fulop (1996) found that the need to accommodate to rapidly changing technology,
markets, and management approaches is an important reason participants join networks in
Australia. Agranoff's (1998) study of public-private rural enterprise alliances found that
needs to leverage public money, to harness emerging technology, to adapt the workforce
to technical changes, and to provide services to small firms are additional advantages of
networks. Agranoff and McGuire's (1998) study of city government networking with
other local governments and the private sector, for economic development purposes,
revealed that most cities become involved in multiple networks. These networks are
forged for different purposes: strategy making, resource exchange, or promotion of
particular projects. The presumption is that local governments engage in such multiple
arrangements in pursuit of flexibility.
If one accepts these advantages at face value, the question remains, How do networks
provide such flexibility? The basic question assumes rapid adaptation and procedural
accommodation as a means to overcome the natural tendency of bureaucracies to slowly
respond to complex environmental changes. In 1997, the South Dakota Federal-State
Rural Development Council convened a group of officials from the federal government,
state government, state association, private sector, and tribes to respond to the devastating
impact of the 1996-1997 winter blizzards and spring floods. The network was able to
move federal assistance through the state and local governments in ways that had not
been accomplished before, particularly the programs of the Federal Emergency
Management Agency (FEMA), the Department of Housing and Urban Development
(HUD), and, to a lesser extent, the Small Busines Administration (SBA). State officials
brokered remediation programs through nonprofit and statewide as! sociations, and
mayors were pleasantly surprised at how quickly food and temporary shelter were
forthcoming. Not only did most network participants respond rapidly, but a number of
standard procedures that normally slow down processes were overlooked or put on the
back burner. HUD, for example, decided to streamline its application form, particularly
the survey of conditions portion, and agreed to accept a shorter Housing Authority of
South Dakota form until the two could be merged. FEMA actually moved its response
and remediation teams into some towns before an assessment of conditions could be
made. Some extra SBA funds were moved to the state's Small Business Development
Center (SBDC) before the expected multiple applications were filed. How did all this
happen? It was a result of the joint effort to gauge the extent of the problem, identify
barriers to problem resolution, and ask the appropriate agencies to respond in an aberrant
fashion. When the emergency responses subsi! ded, the council worked with the U.S.
Department of Agriculture/Rural Development (USDA/RD) and FEMA to develop a
strategic plan for weather related emergencies (Radin et al. 1997). In this way, flexibility
and adaptation were enhanced by the network, not only by convening many decision
makers, but by making the adjustments and committing the resources to solve the
immediate problem.
10
Do networks reach beyond these standard advantages of flexibility and facilitate new
means of dealing with political problems, bureaucratic obstacles, and resource
deficiencies? If so, how do network efforts create flexibility that would not otherwise
occur through hierarchies? Certainly, managers representing city hall seek out
adjustments across organizational jurisdiction boundaries without operating in a network.
But what are the political issues that managers must address in networks that they cannot
address from their positions in the hierarchy? Do they garner support from different
levels of government and nongovernmental organizations to solve problems that the
actors have mutually identified? One example of flexibility is found in urban regimes that
can make major urban improvement efforts work in ways that city government cannot.
Urban revitalization networks, such as those in Baltimore, are shadow governments that
have circumvented cumbersome city hall bureaucrat! ic procedures, overcome public
procurement rules, leveraged private investments with public dollars, and ignored
political criticism from neighborhood and other citizen groups. Hula (1990, 201)
estimates that thirty to forty of such networks existed during William Schaefer's
administration. While these networks made the Inner Harbor development possible, some
have questioned the propriety and accountability of the administration's failure to operate
in the public, their circumvention of standard public procurement property disposal and
purchasing processes, and their lack of public responsiveness.
There are also bureaucratic issues that managers can facilitate by achieving flexibility in
networks that hierarchies cannot do, or do only with extreme difficulty. Can network
identification of rules and obstacles publicly reveal outmoded or silly regulations? Do
networks create new operating rules that allow organizations to circumvent some of their
own procedural obstacles? Do the new rules help organizational officials save face? Does
the moral and numerical power of the coalition of organizations help recalcitrant
bureaucrats back-off from their rules and procedures? Do network solutions that run up
against established procedures present a logic of revision that otherwise recalcitrant or
rule-bound officials cannot avoid? Does the sheer number of organizations within
networks create pressures to change some procedures?
Finally, what resource deficiencies do network flexibilities plug? Are networks involving
governments better than hierarchies at securing multiple resource contributions such as
money, human resources, information, technology? Better at sharing such resources? Do
networks leverage resources easier than hierarchies do? These are the kinds of empirical
questions that need to be answered if we are to understand the real contributions that
flexibility can make in public management networks.
In what ways do networks employ mutual self-responsibility, and does this substitute for
the loss of public agency accountability? Is there residual accountability that falls
between the cracks of organizations, and does it fall in any direction?
The issue raised here involves the question of control and the difficulty of establishing
accountability in public management networks. Indeed, this is one of the most frequently
raised questions in the public administration field (Kettl 1996; Milward 1996;
Frederickson 1997). In single organization settings (networks), do bureaucrats
(government network managers and network members) operate in accordance with
executive direction and the policy preferences of those in authority to whom they report,
11
melding personal responsibility with constitutional duty (Rohr 1989)? Or do they
contradict the norms of representative democracy by assuming too much discretion and
disregarding the important direction of legislators and elected officials of the executive
branch (Burke 1986)? Such venerable issues are magnified for more-complex network
settings, where multiple parties operate with limited authority and even less direct contact
with executive principals or legislative oversight m! echanisms. Is this strong concern
with accountability issues in networking warranted? Is there a demonstrable "leakage of
authority" in collaborative settings (Bardach and Lesser 1996, 198) and, if so, can its
impact be determined through empirical study?
The concern for loss of accountability in multiorganizational settings first emerged as
contracting relationships were becoming plentiful. In such arrangements, there is some
fear that the employees of contracting organizations will pursue purposes inconsistent
with those of elected officials (Milward 1994). Because contractors assume a critical
intermediary position in service delivery systems, they have autonomy from each of the
other governmental players (Kettl 1993). Kettl describes the discrepancy between intent
and reality in the contracting relationship:
The philosophy of contracting out presumes that the basic relationship between
government and contractor will be that of principal and agent. The contractor's job is to
act as agent of the government's policy. The relationship is fractured, however, if
contractors create independent political ties with policymakers and thus outflank their
administrative overseers. In such cases contractors are less agents than partners, helping
to shape the very design of the program, free of any significant oversight, and beneficiary
of state and local governments' dependence on their performance (p. 176).
Two important factors render discussions of accountability in network unique. First,
accountability relationships in multi-- organizational networks are quite different from
those found in the dyadic linkages of contracting or the bi-level interaction of levels of
government portrayed in some definitions of federalism. Myriad applications of
principal-agent theory to contractual and extrajurisdictional activity exist in the literature
(Waterman and Meier 1998). Even in intergovernmental policy implementation, Chubb
(1985) describes the federal system as a two-tiered hierarchy and argues that a principalagent model provides an accurate description of the impact of federal grants on policy
making and managerial behavior. This results in an oversimplification of the multiple
two-way interactions that constitute policy implementation. For settings where multiple
interests-jurisdictional, organizational, sectoral-are involved in policy design or
implementation, establishing jus! t who is the principal and who are the agents is a near
impossible, maybe even meaningless, exercise. As O'Toole points out in his description
of networks, "IT]hey do not have the formal wherewithal to compel compliance with
such cooperative undertaking" (1997c, 445).
Second, because in networks there is no obvious principal or agent, and no exigent
authority to steer the activities of the network in harmony with elected officials, the issue
of accountability is miscast. With no single authority, everyone is somewhat in charge,
thus everyone is somewhat responsible; all network participants appear to be accountable,
but none is absolutely accountable. Additionally, the expectations of accountability to a
legally constituted democratic entity are not consistent with networks that may possess
12
insufficient democratic legitimacy; interaction between civil servants and representatives
of private interest groups, other governmental layers, and implementing organizations
makes it difficult for representative bodies to influence policy. In general, one can argue
that the public sector's responsibility for affecting the public interest is compromised and
limited by the use of networks.
However, it may be that networks force us to shift our concern away from hierarchical
accountability to notions of responsibility, responsiveness, and the fostering of
democratic ideals. From this perspective, O'Toole asks, "Does networked public
administration pose a threat to democratic governance or offer the prospect of its more
complete attainment?" (1997c, 458). He suggests that consideration of the
aforementioned dimensions offers a more complete assessment of the effect of networks
on democratic governance, and he concludes that network management "provides both
complications and opportunities to facilitate parts of the democratic ideal" (p. 458). In a
similar call for redirecting traditional public administration concerns with accountability,
some argue that accountability means organizational systems based in oversight and
reporting arrangements. This accountability to concern obfuscates issues of effectiveness
and performance (Bardach and Lesser 1996). Accountabi! lity for means accountability
for results, accountability for setting wise priorities, accountability for targeting, and
accountability for system modification and design. Accountability for these outcomes is
the objective of the collaborative or network and accountability to is the means to that
objective. From this perspective, bureaucratic notions of accountability are simply
inapplicable in networks, if they ever were in bureaucracies.
Since a control orientation is problematic for collaborative structures, some steps can be
taken to maintain the flexibility of collaboration while ensuring accountability.
Wondolleck and Yaffee (2000) offer suggestions derived from their study of natural
resource management. First, collaborative processes should be developed as
supplemental to, not exclusive of, normal decision-making processes. Decisions made
collaboratively "still need to pass muster under legally derived administrative
procedures" (p. 237). Second, they argue that accountability can be ensured by using
mechanisms that allow decisions to be reviewed by independent, ostensibly objective,
sources. Third, where possible, collaborative decision making should strive toward
achieving agreed-upon performance measures that capture the intent of policy objectives.
Standard setting can create incentives for collaborative players to seek solutions that
achieve the standard. Finally, elected officials and other stak! eholders are too often
uninformed about the results of collaborative decision making. Wondolleck and Yaffee
suggest that home agencies of collaborative decision makers should monitor and evaluate
the effectiveness of solutions created collaboratively.
Some descriptive and normative issues must be sorted out. To whom or what do (should)
network players feel responsible? To whom or what do (should) network players feel
responsive? If we analyze the network as a single entity rather than as many pieces, then
how does personal responsibility get translated into network accountability? Are there
certain collaborative structures, such as hub-spoke structures (Harrison and Weiss 1998),
that are (or are perceived to be) more accountable because there is a single organization
13
through which most information passes and from which the network manager emerges? If
so, does collaboration still require some single player that can be held accountable?
We must be able to measure the outcomes and performance of networks in order to assess
how accountable a particular network is to its stakeholders and for achievement of its
stated goals. The groundbreaking work of Provan and Milward (1995; Milward and
Provan 1998) in this regard is very important. How do we know if a network is shirking
its responsibility if we don't know what the network is doing or how well it is doing it? If
there is variation within and across networks in the level of personal responsibility
assumed by the players, is performance affected? The most basic issues of network
management are bound up in our ability to assess effectiveness and then compare that
effectiveness against some baseline to which the network can be held accountable. As
with all empirical questions, conceptualization and measurement development is
imperative.
What is the cohesion factor in networks that is the equivalent to legal-rational authority in
organizations?
This is one of the most difficult questions in network management, not so much because
nothing has been said about this area, but because of so many apparent alternate
interpretations. Since Weber, scholars and public managers have explicitly or implicitly
accepted that legal authority vested in a hierarchy keeps people operating in a
bureaucratic structure and is the reason people allow themselves to be led, so long as
expectations are within a zone of acceptance. Networks are not based on the legal
authority paradigm, however; they are "structures of interdependence, involving multiple
organizations or parts thereof, where one unit is not merely the formal subordinate of the
others in some hierarchical arrangement"(O'Toole 1997a, 45). When people from
different organizations are not in legally bounded authority relationships, why would they
decide to come to the table, work together on problem clarification and solutions, reach
agreement, and follow through on implementat! ion? The literature suggests the
importance of trust, common purpose, mutual dependency, resource availability, catalytic
actors, and managerial ability.
In the absence of the legal charter, it is commonly accepted that people join, remain, and
work together because of some element of trust. Trust in collective behavior is linked to
the obligation. Such fiduciary obligations are said to be essential in holding networks
together because they involve the obligation to attend broadly to the concerns of specific
others beyond the boundaries of specific, measurable transactions (Barber 1983). Trust
does not require common belief, but obligation and expectation. Ferguson and Stoutland
(1999, 44) relate to such expectation through four trust dimensions: participant motives,
not exploiting or betraying purposes; competency, possessing the knowledge and skills to
do what is expected; dependability, holding the necessary resources; and collegiality,
showing respect and fairness.
Trust is required as public and nonpublic organizations attempt to redefine their usual
legal-based (hierarchical, contractual) relationships (Nohria 1992). As there is more trust,
there is less need to constantly monitor compliance (Alter and Hage 1993). Sabel (1992)
suggests that mutual obligation and expectation is key: "Trust-based governance
14
structures have rich, consultative institutional structures whose very existence belies the
assumption that the agents expect their actions automatically to be harmonized by the
confluence of belief." Indeed, Fountain (1994) suggests that trust as a social relation may
be on a par with exchange as a lubricant in network behavior. Operationally, Perrow
(1992) suggests several forces that build trust within small-firm networks: sharing and
discussing information; similarity in processing and techniques; experience in working
with another firm; long-term relationships, even if contact is intermittent; similar size,
power, or strategic! position among firms; rotation of leadership; similar financial
rewards; and economic advantages to support shared meaning.
Others maintain that public or collective good, as manifest in a shared belief or common
purpose, contributes to holding the network together (Alter and Hage 1993). Mandell
(1999, 46) refers to the sum of these qualities as a program rationale, which describes the
mind-set or commitment to the whole that holds a network together when traditional
methods of coordination and control are not operative. From this perspective, program
structures are not merely aggregates of individual organizations; the network itself is
more critical than the component organizations, and focus on the individual organization
is relevant only for the understanding of how and why each organization contributes to
the overall effort (Provan and Milward 1991). These program structures display common
features: reciprocal relationships, many suppliers of resources, overlapping and dynamic
divisions of labor, diffused responsibility for actions, high potential for poor
coordination, massive information ! exchanges, and need for information input from all
actors (Hanf, Hjern, and Porter 1978). Program rationale keeps actors working in spite of
the difficulty of maintaining these network features.
Mutual dependency, particularly oriented to the availability of resources, is yet another
explanation of cohesion. The interorganizational literature dating back a few decades
suggests that actors in a network are in some form of interactive dependency, usually
based on resource exchanges (Pennings 1981). Most organizations in a network are thus
strategically interdependent, but some are more resource dependent than others
(Yuchtman and Seashore 1967). All interorganizational interactions-communication as
well as joint activity-are ultimately and fully dependent and seeking an adequate supply
of resources. This orientation is the defining activity and is equated with forces of control
within networks. Alter and Hage (1993) believe that both private and nonprofit
organizations (including subnational governments) are resource dependent and must
avoid single vertical markets (funders) while they engender multiple horizontal resource
markets. While the extent of commitment and ! resources shared may vary greatly,
parties possess some measure of resources (funding, expertise, support, human resources)
to move the network forward.
Leadership and guidance ability within self-managing systems is another prominent
contributor to network cohesion identified in the public network management literature. It
is commonly understood that network leadership and management require the "principles
of 'soft' guidance" as a replacement for command and control (Windhoff-Hentier 1992).
The principles of steering ability are critical in holding the network together. All the
managerial tasks we have discussed take advantage of network self-management
propensities, using minimal coercion and resources, "balancing social forces and interests
15
and enabling social actors and systems to organize themselves" (Kooiman 1993, 256).
Bardach's (1998) thorough examination of success and failure in collaborative settings
relies heavily on various aspects of capacity building (i.e., human resources, use of best
practices, availability of critical skills, ability to ward off hostile forces), but points most
heavily to leadership as th! e central factor holding the network together. The leader as
craftsman elicits common goals, creates an atmosphere of trust, brokers organizational
and individual contributions, and deploys energies in accord with some strategic plan. It
is important to note that these are not personal attributes, talents, or conditions of
individuals; they are roles in a system of strategic interactions. Cohesion in the network
from this perspective is thus dependent on the integrative, creative, and purposive
opportunities seized by those who apply the craft of network management. Building on
numerous cases, Bardach illustrates how capacities are accumulated through various
stages of investigation, decision, operation, avoidance of obstacles, building of
momentum, and rewarding of performance-all the time holding the structure together.
This type of managing against the hierarchical grain, he concludes, means that leaders
have to create interagency collaborative capacity, thus moving toward ! tangible resource
exchanges plus "cooperative dispositions and mutual understanding of the individuals
who are trying to work together on a common task" (p. 307).
Is it possible that more than one of these forces can be at work within the same network
configuration? One network study of six rural enterprise alliances-nonmetropolitan
partnerships among producers, distributors, labor unions, employer associations, credit
institutions, and government agencies-revealed several cohesive dimensions at work.
Mutual dependence clearly held the alliances together; scarcity of technical information,
advanced equipment, design expertise, market information, venture capital, and operating
funds were among the resources exchanged. Generally speaking, a program rationale also
existed in the underlying belief in creating rural-based industry by promoting small
production units. The obligations and expectations regarding trust within the alliances
developed over time, even among competing entrepreneurs. Also, the private partners had
to trust the various governments' reciprocal actions, whereas the governments had to
place faith in program success w! hen loans or grants were awarded. Finally, all of the
alliances were held together by strategic, operational implementation and assessment
abilities, a form of steering that used a catalytic leader or project manager to build a type
of interorganizational collaborative capacity (Agranoff 1998).
Clearly, a public network management theory must come to grips with an equivalent to
hierarchical authority, but the answer may not lie along a single channel. If this is so, then
questions other than, What is it? may be relevant. For example, what measures of shared
purpose, trust, resource dependence/interdependence, and steering or collaborative
capacity are needed to hold networks together? How do leaders use each cohesive force
to reach the aims of the network? Is an operating deficit in one cohesive area
complemented by a surplus in another? At what point do deficits in one or more of these
cohesive factors contribute to a functional collapse of cohesion, and thus to a collapse of
the network?
How does the often neglected or misunderstood role of organizational power in network
management come into play?
16
Network power, or the ability to get action by partners or organizations under
circumstances where actors are under dual responsibility roles to both organizations and
networks, is clouded by the rhetoric of networking. A focus on joint decision forces
within networking such as those we have mentioned, as well as perceptions of equity,
mask the potential for coercive behavior in network management. Klijn (1996), for
example, characterizes policy networks as coequal, interdependent, patterned
relationships. It is possible, however, that different actors occupy different role positions
and carry different weights within networks. Some sit in positions with extensive
opportunity contexts, filling "structural holes" (Burt 1992, 67), creating unequal
opportunity, while others may be less willing or able players. Organizational
representatives also differ with regard to the resource dependencies they may bring to the
network (Rhodes 1981), leading to power differences. Clegg and ! Hardy (1996, 679)
suggest that " [w]e cannot ignore the facade of 'trust' and the rhetoric of 'collaboration'
used to promote vested interest through the manipulation and capitulation by weaker
partners." Power concerns should be at the core of any general theory of network
management because we must know whether power moves hinder the kind of synergistic
creativity that reciprocal relationships are purported to produce.
Power in networks can be portrayed neutrally, or at least dualistically, as a property that
either prevents or facilitates action. Collaboration can involve changes in structures of
domination or asymmetries of resources employed in the sustaining of power relations
(Gray 1999). Schapp and van Twist (1997, 66-- 67) refer to individual veto power in
networks because "actors in the network are able to cut themselves off from the steering
interventions of other actors." Such veto power can be used in different ways: to exclude
certain actors, to ban certain points of view, or to close potential actors outside of the
network. On the other hand, there is an enabling component to network power. The use of
power does not always include getting the dominant to act, or altering their social control.
In the social production model identified by Stone and his associates (1999, 354), it is
assumed that social forces (and society's most vexing problems) are characterized by a
lack of co! herence and that many activities are autonomous with many middle range
accommodations instead of a cohesive system of control. In this type of situation, the
main concern is how to bring about enough cooperation among disparate community
elements to get things done. This is a power to that, under many conditions of ultra
complexity, characterizes situations better than power over. Moreover, the social
production model holds out hope of constituting new possibilities under conditions of
flexible preferences, or "bringing about a fresh configuration of preferences through
opening up new possibilities" (p. 355). The key task in such power configurations is that
of building critical skills within networks of relationships while enlarging the view of
what is possible. It is different than mobilization against the powerful, but it entails
enlarging ways of thinking about preferences. The social production model juxtaposed
against veto power illustrates just two dimensions of power i! n networks. As any
political scientist can relate, power has many faces. In the minimum, it can inhibit or
facilitate collective action, including action within networks.
Are there important operating dimensions of power? Gray (1989) characterizes
collaboration as power sharing. The exercise of collaboration involves how power is
acquired and challenged in the formation of collaborative efforts. Resource mobilization
17
theorists (Alinsky 1971; Gamson 1975; Bachrach and Baratz 1963) suggest that two
sources of power are critical to the problem-setting phase: the power to mobilize and the
power to organize. Gray (1999) offers others: the power to strategize; control of
information, particularly in the direction-setting phase of collaboration; and the ability to
exercise influence or authorize action, which is paramount to implementation. Gray sees
this collaboration sequence as a power sharing that implies at least a temporary
redistribution of power, enabling stakeholders with differential power levels to engage in
domain transformation. She cautions, however, that alternative strategies of engagement
can also characterize stakeholder involveme! nt: compliance, when one organization
bows to the wishes of the other (or others) because of resource leverage or undermining
of legitimacy; contention, or challenging practices held or decisions made by more
powerful stakeholders, leading to discursive legitimacy; and contesting, or a means of
engagement within the domain of other stakeholders, particularly the more powerful
ones. Also, equality of power can be eroded by invoking formal or legal authority,
resource control, or cooptation of low-power stakeholders. These sequences illustrate that
power is a property in network management that can facilitate or inhibit collaborative
actions.
Dynamics within the Federal-State Rural Development Councils again illustrate these
complexities of network power. In their formative years these multijurisdictional councils
did a reasonably adequate job of bringing together the disparate actors and following the
five-step sequence of collaborative power that we have identified. In regard to certain
problems related to rural development, particularly those that dealt with
intergovernmental regulatory or programmatic barriers to economic and community
development, these networks opened up numerous new possibilities. Thus, in regard to
the intergovernmental program issues explored by the councils, power was leveled
among partners. Federal, state, association, private-sector, and tribal stakeholders were
able to explore identified problems in depth, develop workable solutions, and get
implementing agencies to yield power and procedure in the interest of solving the
problem. In this sense, true social production was possible. !
Lurking behind the overt pretenses of network power sharing, however, was the
contesting and sometimes contention or compliance enforcing of the most powerful
stakeholders. In virtually every state the interest or noninterest of the governor's office
proved to be a key force in determining the issues that the councils addressed. In one or
two states, councils' agendas proved to be their governor's agenda. In other states, the
work of councils could come to a halt should their governors be uninterested or unwilling
to support it. Under very limited circumstances could most councils tread on the turf of
an administration's rural agenda, if one existed. Rural policy agendas belonged to the
governor, so councils were relegated to marginal cross-- jurisdiction discussions, issue
papers, and demonstrations. Moreover, at a network operating level most councils have to
defer to the power of their two most important stakeholders, the state departments of
economic development (agenci! es also very close to their governors) and USDA/RD,
housed in each state and headed by a presidential appointee. These two agencies
constitute the sources of discretionary funding for rural development projects. In one
state the power conflict within the council between these two major stakeholders was so
intense that the rest of the council tried to mute it by precluding either party from
18
membership in its steering committee. But this was not a deterrent for either, as they
wielded their levers behind the scenes. Most important, in this state the powerful
stakeholders limited the council's agenda to issues that did not interfere with their
interests (Radin et al. 1996; Radin et al. 1997). In virtually every state that has a council
(thirty-five in total), both the power to and the power over is exhibited.
The power issue must take center stage in network research. What techniques can be used
to channel negative stakeholder engagement into positive engagement? Do networks need
to manage power in collaboration in different ways than do hierarchies? How does an
operating network recognize when collaboration is breaking down because the power
balance is breaking down? How is power channeled from stakeholders protecting their
interests into social production? How do power brokers in networks build effective social
production? What are the negative effects on creative problem solving of unequal
application of power?
Power is obviously a more complicated concern in networks than we can possibly depict
here. As an attribute of management, the subject needs to be raised because so much of
the rhetoric of networking emphasizes processes that imply mutuality. One popular book
on network management, for example, extolls five key features of mutuality: unifying
purpose, independent members, voluntary links, multiple leaders, and work at integrated
levels (Lipnack and Stamps 1994). It says much less about the ability of key stakeholders
to dominate or how such domination may erode unifying purpose, independent members,
and so on. Some empirical research has revealed the reality of power. The work of
Milward and Provan (1998) found that network effectiveness was associated with
stakeholder exercise of controlling power among community mental health centers, a
monopoly provider that dominated both service delivery and funding provisions for
others in the network. Power concerns must be moved to ! the core of network
management research.
Do public management networks produce results that otherwise would not have
occurred? Do networks discover processes and solutions that would not have emerged
from work through a single organization?
These questions address the origin of a public management network. That networks are a
unique institutional form, consisting of processes different than the spontaneous
coordination of markets or the visible management of hierarchy (Powell 1990), is the
premise upon which the study of networks is based. The critical issue in question is
whether public management networks produce solutions and results that otherwise would
not have occurred through single, hierarchical organizations. Are networks required for
achieving results in particular problem areas? When the public demands action on certain
public issues, are multiple players drawn together to fulfill that demand because, if the
demand is to be fulfilled, it can be only through a network? Or, when the public demands
action on certain public issues, do multiple players work jointly to fulfill those demands
because governments will not do so?
One dominant perspective argues that the pace and quality of social change at this point
in history are the primary determinants of the emergence of network forms. This social
change thesis is prominent in the writings of futurists (Toffler 1980), business consultants
19
(Lipnack and Stamps 1994; Peters 1992), organization theorists (Clegg 1990), and in
much of the literature on public management networks. Just as the bureaucratic
organization was the signature organizational form during the industrial age, the
emerging information or knowledge age gives rise to the network, where persons link
across internal functions, organization boundaries, and even geographic boundaries. The
world is characterized by extreme complexity and diversity (Dunsire 1993; Kooiman
1993). Power is dispersed, not centralized, and tasks are becoming unified, rather than
subdivided and specialized. Society worldwide demands greater freedom and
individuation, rather than integration. In such a world, a! n organizational form based on
individuation, dispersed power, and unification is necessary; the network is that form.
A related perspective is the problem change thesis, which asserts that the types of
problems or issues that society addresses collectively are increasingly wicked-"problems
with no solutions, only temporary and imperfect resolutions" (Harmon and Mayer 1986,
9). Tame problems are readily defined and easily decomposable into neat, technical
solutions (O'Toole 1997a), but these have given way in large part to wicked problems.
For most of the problems that emerged in the first part of the twentieth century, a
bureaucratic organization was ideal-problems were easily defined, goals were clear, and
objectives were measurable. The metaphor of the wicked problem stands in contrast to
traditional bureaucratic policy making and implementation. When there is little or no
agreement about the appropriate division of moral, institutional, or legal responsibility for
any particular public problem, who can say how that problem might be solved (Harmon
and Mayer 1986)? For wicked problems, ag! reement is forged by jointly steering courses
of action and delivering policy outputs that are consistent with the multiplicity of societal
interests. Other, more nonconventional modes of organizing, like networks, have
emerged to do just that.
Is the emergence of networks a completely natural phenomenon as the social change and
problem change arguments indicate? The empirical literature offers mixed answers, most
of which focus on the types of policy instruments adopted as the key determinants of
network emergence. For example, the emergence of intergovernmental networks within
the context of municipal wastewater treatment assistance does not appear to be "natural"
(O'Toole 1996) in that their creation was necessary in order to continue delivering such
assistance. In fact, the network emerged from a decision by the federal government
(EPA) to reduce its role as grantor to municipalities and also from the ensuing initiative
of states to create separate state revolving loan funds as a partial replacement for the lost
federal support. This is evidence for the claim that "efforts from political levels to trim
the scale of bureaucracy and extent of direct administrative responsibility for
accomplishing public purposes! . . . accentuate the networking impulse" (O'Toole 1996,
239). The decision to utilize state revolving loan funds resulted in more complex
governing configurations involving new actors with needed resources and technologies.
O'Toole concludes that "decisions to shift programs to the states, deregulate, privatize,
and employ market-based mechanisms have consequences for interorganizational
arrangements and programs in practice" (p. 239). It is unclear whether a networked policy
structure is the best choice for providing assistance to cities for wastewater treatment, but
evidence indicates that it certainly is the best-perhaps the only choice-for the type of
policy instrument being used, which, in this case, is a state-level revolving loan fund.
20
Similarly, as the acceptable technology for treating people with severe mental illness
shifted in the 1960s from directly provided institutionalization to a reliance on services
provided by community health and social service organizations, collaborative service
implementation networks of these provider agencies emerged (Provan and Milward 1991
and 1995; Weiss 1990). Did these public management networks produce results in the
mental health field that otherwise would not have occurred? Once direct provision of
government services in hospitals was viewed as an unacceptable policy instrument and,
consequently, vertical integration of services through one provider was no longer
appropriate, service implementation networks became the only, therefore the best,
organizational form through which mental health services could be delivered. Given the
type of policy instrument adopted by governments, networks did indeed produce results
that otherwise would not have occurred. Harrison a! nd Weiss (1998) use numerous cases
to demonstrate how a more inclusive approach to workforce development results in
networks of inter-- organizational relationships that connect the residents of low-- income
neighborhoods to employers with training positions through mediating institutions such
as community colleges, government agencies, and various community-based
organizations. Since the most politically acceptable means for employing persons in
America is for these persons to find jobs on their own in a competitive market, rather
than have government give jobs to them, extensive networks of employers, trainers, and
the unemployed clearly become the best organizational fit for delivering such services.
Like the empirical research cited in these last two paragraphs, any examination of public
management networks clearly must not disregard the political context within which
networks emerge and operate. Even in network settings, structure follows strategy.
Finally, over and above the societal level changes occurring and the shift in the types of
acceptable policy instruments used in governing, decisions made in networks may simply
be better decisions. Not better in the sense of more efficient-there is nothing particularly
efficient about making decisions jointly-but better in the sense of being more effective,
since, ideally, those involved in any given network are not merely steerers, but
stakeholders, suppliers, clients, even customers. If the basic problem and challenge of
policy and strategy making in many policy settings is for multiple governmental and
nongovernmental organizations to jointly steer courses of action and to deliver policy
outputs that are consistent with the multiplicity of societal interests, then a policy
decision that laboriously, even painfully, meets that test is bound to be viewed as the best
decision. When all relevant interests are considered in decision making, then, at the
minimum, the decision! will have wider agreement, which still may be the test of the best
policy (Lindblom 1959). To some degree, network decision making can be viewed as
more rational than individual decision making is. Multiple parties means multiple
alternatives to suggest and consider, more information available for all to use, and a
decision system that is less bounded by the frailties of individual thinking. Additionally,
decisions in networks may not be the product only of a more rational process, but they
may also occur as a result of a synergy that can develop when multiple players pursue a
common solution. Synergy means that the commitment and interaction of the participants
stimulates new alternatives that otherwise would not have been considered.
More empirical research is necessary to sort out the distinctive contribution of networks
as a decision-making entity. Is the emergence of networks in public policy and
21
administration due to a shift in the types of policy instruments used by governments
(McGuire 2000)? Do we see a greater number of networks because of a shift in the level
of political acceptability of some forms of governance? Conversely, do networks select
policy instruments that are fundamentally different than those selected by bureaucracies?
Does the choice to utilize networks emerge from the political calculus of government
institutions, or is that choice predetermined by the natural forces of change? Research
that examines how and why networks emerge within the context of particular policy and
program areas is necessary to better understand what can be expected from network
forms of governance and how to evaluate the solutions that emerge from such forms.
CONCLUSION
In order to produce useful information on the structure and process of management in
networks involved in public action, one must address relevant questions. These questions
are not answered here. However, we are able to make some tentative conclusions
regarding network management. There are apparent common network management
sequences like activation, framing, mobilizing, and synthesizing. We need to know more
about how these processes unfold. Some of these are behavioral, as collaborative learning
that draws on social capital plus negotiation supplement more standard behavioral
science techniques. We need to know more about how negotiation plays off against
collaborative learning. Apparently, many political, bureaucratic, and resource deficiency
issues can be facilitated through flexible processes by managers. It is apparent that
accountability in networks is defined and perceived differently than in single
organizations. We also need to know more about the contribution o! f personal
responsibility and responsiveness to network performance. Additionally, the cohesion
developed through trust and a program rationale should be measured against the force of
authority in hierarchies.
Not all is peace and harmony in networks. We know that power needs to be confronted as
a core element in network management, both in terms of control and its social production
forms. The impact of control on synergistic collaboration needs further elaboration.
While we know that networks create value by acting as important means of bridging
multiple organizations and dealing with difficult problems, there is more to be explored.
Is the emergence of networks attributable to changes in policy instruments? Is the
emergence of new policy instruments attributable to the availability of network forms of
organizing? There is little research on the productive outcomes of networks regarding
whether solutions would be different if carried out through single organizations. Simply
put, we need to know more about how networks produce.
In the opening chapter of the volume that emerged from the first national Public
Management Research Conference, Hal Rainey addressed the topics of what public
management is and what public management should be, by asking "whether public
administration scholars might do better in advancing both the identity of the field and its
research and theory if fewer of us ruminated on these topics and more of us simply
identified important theoretical and research questions and worked on providing answers
to them" (1993, 9). So it is with public network management. It is not so much that
scholars who study networks ruminate needlessly about normative issues-indeed, the
subject is rarely introduced (O'Toole 1997a)-but rather that so few do empirical research
22
with the expectation that it will lead to additional research, and too few provide empirical
researchers with questions to consider and hypotheses to test. To be relevant,
management research must inform action: data for data's sa! ke is not useful. As it is now
a core task of governance, network management must be placed up front as an essential
arena of examination in the fields of public management and administration.
REFERENCES
Agranoff, Robert.
1991 "Iowa's Rural Policy Academy: Process and Strategic Plan." Report submitted
to Aspen Institute/Ford Foundation evaluation study of National Governor's
Association Rural Policy Academy, Washington, D.C.: Aspen Institute.
1998 "Rural Enterprise Alliances: Challenges to Public Management."
International Journal of Public Administration 21:153375.
Agranoff, Robert, and McGuire, Michael.
1998 "Multinetwork Management: Collaboration and the Hollow State in Local
Economic Policy." Journal of Public Administration Research and Theory 8:67-91.
1999a "Managing in Network Settings." Policy Studies Review 16:18-41.
1999b "Expanding Intergovernmental Management's Hidden Dimensions. "
American Review of Public Administration 29:352-69.
Alinsky, Saul D.
1971 Rules for Radicals. New York: Vintage Books.
Alter, Catherine, and Hage, Jerald. 1993 Organizations Working Together.
Newbury Park, Calif.: SAGE.
Bachrach, Peter, and Baratz, Morton S. 1963 "Decisions and Non Decisions: An
Analytical Framework." American Political Science Review 57:641-51.
Barber, Benjamin.
1983 The Logic and Limits of Trust. New Brunswick, N.J.: Rutgers University
Press.
Bardach, Eugene.
1998 Managerial Craftsmanship: Getting Agencies to Work Together. Washington,
D.C.: Brookings.
23
Bardach, Eugene, and Lesser, Cara. 1996 "Accountability in Human Services
Collaboratives: For What? and For Whom?" Journal of Public Administration
Research and Theory 6:197-224.
Benson, J. Kenneth.
1975 "The Interorganizational Network as Political Economy." Administrative
Science Quarterly 20:229-49.
Burke, John P.
1986 Bureaucratic Responsibility. Baltimore: Johns Hopkins University Press.
Burt, Ronald.
1992 Structural Holes: The Social Structure of Competition. Cambridge, Mass.:
Harvard University Press.
Chubb,John.
1985 "The Political Economy of Federalism." American Political Science Review
79:994-1015.
Clegg, Stewart R.
1990 Modern Organizations: Organization Studies in the Postmodern World.
London: SAGE.
Clegg, Stewart R. and Cynthia Hardy. 1996 "Conclusion: Representations." In
Stewart R. Clegg, Cynthia Hardy, and Walter R. Nord,
eds. Handbook of Organization Studies. London: SAGE.
Coleman, James S.
1990 Foundations of Social Theory. Cambridge, Mass.: Harvard University Press.
Dunsire, Andrew.
1993 "Modes of Governance." In Kooiman,ed.
Ferguson, Ronald F., and Stoutland, Sara E.
1999 "Reconceiving the Community Development Field." In Ronald F. Ferguson
and William T. Dickens, eds. Urban Problems and Community Development.
24
Washington, D.C.: Brookings.
Fountain, Jane E.
1994 "Trust as a Basis for Organizational Forms." Paper presented at conference,
Network Analysis and Innovations in Public Programs, University of WisconsinMadison. September 30-October 1.
1998 "Social Capital: Its Relationship to Innovation in Science and Technology."
Science and Public Policy 25:103-15.
Frederickson, H. George.
1997 The Spirit of Public Administration. San Francisco: JosseyBass.
Fulop, Liz.
1996 "Small Firm Networks in Western Sidney." Working paper, Department of
Management, University of Wollongong, Wollongong, NSW.
Galaskiewicz, Joseph, and Zaheer, Akbar.
1999 "Networks of Competitive Advantage." Research in the Sociology of
Organizations 16:237-61.
Garrison, William O.A.
1975 The Strategy of Social Protest. Homewood, Ill.: Dorsey.
Gray, Barbara.
1989 Collaborating: Finding Common Ground for Multiparty Problems. San
Francisco: Jossey-Bass.
1999 "Theoretical Perspectives on Collaboration." Paper presented at
Collaboration Research Group, Technology. University of Sidney, Sidney, Australia.
Hanf, Kenneth; Hjern, Benny; and Porter, David 0.
1978 "Local Networks of Manpower Training in the Federal Republic of Germany
and Sweden." In Kenneth Hanf and Fritz W. Scharpf, eds. Interorganizational
Policy Making. London: SAGE.
Hanf, Kenneth, and O'Toole, Laurence J.
1992 "Revisiting Old Friends: Networks, Implementation Structures and the
25
Management of Inter-Organizational Relations." European Journal of Political
Research 21:163-80.
Harmon, Michael M., and Mayer, Richard T.
1986 Organization Theory for Public Administration. Glenview, III: Scott,
Foresman.
Harrison, Bennett, and Marcus, Weiss. 1998 Workforce Development Networks.
Thousand Oaks, Calif: SAGE.
Hilbert, David.
1902 "Mathematical Problems: Lecture Delivered Before the International
Congress of Mathematicians at Paris in 1900." Bulletin of the American
Mathematical Society 8:437-79.
Hula, Richard C.
1990 "The Two Baltimores." In Dennis Judd and Michael Parkinson, eds.
Leadership and Urban Regeneration. Newbury Park, Calif.: SAGE.
Innes, Judith E., and Booher, David E. 1999 "Consensus Building and Complex
Adaptive Systems: A Framework for Evaluating
Collaborative Planning." Journal of the American Planning Association 65:412-23.
Kettl, Donald F.
1993 Sharing Power: Public Governance and Private Markets. Washington, D.C.:
Brookings.
1996 "Governing at the Millennium." In James L. Perry, ed. Handbook of Public
Administration, 2d ed. San Francisco: JosseyBass.
Kickert, Walter J.M.; Klijn, Erik-Hans; and Koppenjan, Joop F.M.
1997 "Introduction: A Management Perspective on Policy Networks." In Kickert,
Klijn, and Koppenjan, eds. Managing Complex Networks. London: SAGE.
Kickers, Walter J.M., and Koppenjan, Joop F.M.
1997 "Public Management and Network Management: An Overview." In Kickert,
Klijn, and Koppenjan, eds.
26
Klijn, Erik-Hans.
1996 "Analyzing and Managing Policy Processes in Complex Networks."
Administration and Society 28:90-119.
Klijn, Erik-Hans, and Teisman, G.R. 1997 "Strategies and Games in Net
works." In Kickert, Klijn, and Koppenjan, eds.
Kooiman, Jan, ed.
1993 Modern Governance: New Government-Society Interactions. London: SAGE.
Landau, Martin.
1991 "Multiorganizational Systems in Public Administration." Journal of Public
Administration Research and Theory 1:5-18.
Lindblom, Charles E.
1959 "The Science of Muddling Through." Public Administration Review 19:79-88.
Lipnack, Jessica, and Stamps, Jeffrey. 1994 The Age of the Network. New York:
Wiley.
Mandell, Myrna P.
1988 "Intergovernmental Management in Interorganizational Networks: A Revised
Perspective." International Journal of Public Administration 11: 393-416.
1990 "Network Management: Strategic Behavior in the Public Sector." In Robert
W. Gage and Myrna P. Mandell, eds. Strategies for Managing Intergovernmental
Policies and Networks. New York: Praeger.
1999 "Community Collaborations: Working Through Network Structures." Policy
Studies Review 16:42-64.
McGuire, Michael.
2000 "Collaborative Policy Making and Administration: The Operational Demands
of Local Economic Development." Economic Development Quarterly 14:276291.
Milward, H. Brinton.
1994 "Mapping the Linkage Structure of Networks." Paper presented at
conference, Network Analysis on Innovations in Public Programs, LaFollette
27
Institute of Public Affairs, University of Wisconsin.
1996 "Symposium on the Hollow State: Capacity, Control, and Performance in
Interorganizational Settings." Journal of Public Administration Research and
Theory 6:193-95.
Milward, H. Brinton, and Provan, Keith G.
1998 "Principles for Controlling Agents: The Political Economy of Network
Structure." Journal of Public Administration Research and Theory 8:203-21.
Mossberger, Karen, and Hale, Kathleen.
1999 "Information Diffusion in an Intergovernmental Network: The
Implementation of School-toWork Programs." Paper presented at the annual
meeting of the American Political Science Association, Atlanta, September 2-5.
Nohria, Nitin.
1992 "Is a Network Perspective a Useful Way of Studying Organizations?" In Nitin
Nohria and Robert G. Eccles, eds. Networks and Organizations: Structure, Form
and Action. Boston: Harvard Business School Press.
O'Toole, Laurence J.
1988 "Strategies for Intergovernmental Management: Implementing Programs in
Interorganizational Networks." International Journal of Public Administration
11:417-41.
1996 "Hollowing the Infrastructure: Revolving Loan Programs and Network
Dynamics in the American States." Journal of Public Administration Research and
Theory 6:225-42.
1997a "Treating Networks Seriously: Practical and Research-Based Agendas in
Public Administration." Public Administration Review 57:45-52.
1997b "Implementing Public Innovations in Network Settings." Administration and
Society 29:115-34.
1997c "The Implications for Democracy in a Networked Bureaucratic World."
Journal of Public Administration Research and Theory 7:443-59.
Pasternack, Bruce A., and Viscio, Albert J.
1998 The Centerless Corporation: A New Model for Transforming Your
28
Organization in Growth and Prosperity. New York: Simon and Schuster.
Pennings, Johannes M.
1981 "Strategically Interdependent Organizations." In Paul Nystrom and William
0. Starbuck, eds. Handbook of Organization Design. New York: Oxford. Perrow,
Charles.
1992 "Small Firm Networks." In Nitin Nohria and Robert G. Eccles, eds. Networks
and Organizations: Structure, Form, and Action. Boston: Harvard Business School
Press.
Peters, Tom.
1992 Liberation Management. New York: Knopf.
Piore, Michael J., and Sabel, Charles F.
1984 The Second Industrial Divide: Possibilities for Prosperity. New York: Basic
Books.
Porter, David 0.
1981 "Accounting for Discretion in Social Experimentation and Program
Administration." In Katherine L. Bradbury and Anthony Downs, eds. Do Housing
Allowances Work? Washington, D.C.: Brookings.
Powell, Walter W.
1990 "Neither Market nor Hierarchy: Network Forms of Organization." Research
in Organizational Behavior 12:295-336.
Provan, Keith G., and H. Brinton Milward.
1991 "Institutional-Level Norms and Organizational Involvement in a ServiceImplementation Network." Journal of Public Administration Research and Theory
1:391-417.
1995 "A Preliminary Theory of Interorganizational Effectiveness:
A Comparative Study of Four Community Mental Health Systems." Administrative
Science Quarterly 40:1-33.
Putnam, Robert.
29
1993 Making Democracy Work: Civic Traditions in Modern Italy. Princeton, N.J.:
Princeton University Press.
Radin, Beryl A.; Agranoff, Robert; Bowman, Ann O;M.; Buntz, Gregory C.;
Romzek, Barbara S.; and Wilson, Robert H.
1997 The National Rural Development Partnership After Seven Years: A Glimpse
at Eight State Rural Development Councils in 1997. Washington, D.C.: National
Program Office, NRDC.
Radin, Beryl A.; Agranoff, Robert; Bowman, Ann O'M.; Buntz, Gregory C.; Ott,
Steven J.; Romzek, Barbara S.; and Wilson, Robert H.
1996 New Governance for Rural America: Creating Intergovernmental
Partnerships. Lawrence: University Press of Kansas.
Rainey, Hal G.
1993 "Important Research Questions." In Barry Bozeman, ed. Public Management:
The State of the Art. San Francisco: Jossey-Bass. Reich, Robert B.
1991 The Work of Nations. New York: Knopf
Rhodes, Roderick A.W.
1981 Control and Power in Central-- Local Relations. Aldershot, U.K.: Gower.
Rohr, John A.
1989 Ethics for Bureaucrats: An Essay on Law and Virtue, 2d ed. New York:
Dekker.
Sabel, Charles F.
1992 "Study Trust: Building New Forms of Cooperation in a Volatile Economy." In
Sengenberger and Pyke, eds.
Schapp, L, and van Twist, M.J.W. 1997 "The Dynamics of Closedness in
Networks." In Kickert, Klijn, and Koppenjan, eds.
Scharpf, Fritz.
1978 "Interorganizational Policy Studies: Issues, Concepts, and Perspectives." In
Kenneth Hanf and Fritz W. Scharpf, eds. Interorganizational Policy Making.
London: SAGE.
30
Senge, Peter M.
1990 The Fifth Discipline: The Art and Practice of the Learning Organization. New
York: Doubleday.
Sengenberger, Werner, and Pyke, Frank.
1992 "Industrial districts and Local Economic Regeneration: Research and Policy
Issues." In Pyke and Sengenberger, eds. Industrial Districts and Local Economic
Regeneration. Geneva: International Institute for Labor Studies.
Stoker, Gerry, and Mossberger, Karen. 1994 "Urban Regime Theory in Com
parative Perspective." Environment and Planning C: Government and Policy
12:195-212.
Stone, Clarence N.
1989 Regime Politics. Lawrence: University Press of Kansas. 1999 "The Atlanta
Experience Re
examined: Toward Refining the Analysis of Urban Regimes." Paper presented at
the annual meeting of the American Political Science Association, Atlanta,
September 2-5.
Stone, Clarence; Doherty, Kathryn; Jones, Cheryl; and Ross, Timothy. 1999
"Schools and Disadvantaged
Neighborhoods: The Community Development Challenge." In Ronald F. Ferguson
and William T. Dickens, eds. Urban Problems and Community Development.
Washington, D.C.: Brookings.
Termeer, C.J.A.M., and Koppenjan, loop F.M.
1997 "Managing Perceptions in Networks." In Kickert, Klijn, and Koppenjan, eds.
Toffler, Alvin.
1980 The Third Wave. New York: Morrow.
Waterman, Richard W., and Meier, Kenneth J.
1998 "Principal-Agent Models: An Expansion?" Journal of Public Administration
Research and Theory 8:173-202.
31
Weiner, Myron E.
1990 Human Services Management, 2d ed. Belmont, Calif.: Wadsworth.
Weiss, Janet A.
1990 "Ideas and Inducements in Mental Health Policy." Journal of Policy Analysis
and Management 9:178-200.
Windhoff-Hentier, Andriene.
1992 "The Internationalization of Domestic Policy: A Motor of Decentralization."
Paper presented at for ECPR Joint Sessions, Limerick, Ireland.
Wondolleck, Julia M., and Yaffee, Steven L.
2000 Making Collaboration Work: Lessons from Innovation in Naural Resource
Management. Washington, D.C.: Island Press.
Yuchtman, E., and Seashore, Stanley E.
1967 "A System Resource Approach to Organizational Effectiveness." American
Sociological Review 32:891-903.
Robert Agranoff
Indiana University, Bloomington
Michael McGuire
University of North Texas
An earlier version of this article was delivered at the fifth national Public
Management Research Conference, Texas A&M University, College Station,
December 3-4, 1999.
Reproduced with permission of the copyright owner. Further reproduction or distribution is prohibited
without permission.
32