Inside Collaborative Networks: Ten Lessons for Public Managers
by Robert Agranoff
Indina University–Bloomingon
This paper offers practical insights for public managers as they work within interorganizational networks. It
is based on the author’s empirical study of 14 networks involving federal, state, and local government
managers working with nongovernmental organizations. The findings suggest that networks are hardly crowding out the role of public agencies; though they are limited in their decision scope, they can add collaborative public value when approaching nettlesome policy and program problems.
Extended discussions were undertaken in the field on two separate occasions with more than 150 public officials, in addition to field observation and examination of network documentation.
Lesson 1: The network is not the only vehicle of collaborative management.
Lesson 2: Managers continue to do the bulk of their work within the hierarchy.
Lesson 3: Network involvement brings several advantages that keep busy administrators
involved.
Lesson 4: Networks are different from organizations but not completely different.
Lesson 5: Not all networks make the types of policy and program adjustments ascribed to them in the literature.
Lesson 6: Collaborative decisions or agreements are the products of a particular type of mutual learning and adjustment.
Lesson 7: The most distinctive collaborative activity of all of the networks proved to be their work
in public sector knowledge management.
Lesson 8: Despite the cooperative spirit and aura of accommodation in collaborative efforts, networks are not without conflicts and power issues.
Lesson 9: Networks have their collaborative costs, as well as their benefits.
Lesson 10: Networks alter the boundaries of the state only in the most marginal ways; they do not appear to be replacing public bureaucracies in any way.
It is time to go beyond heralding the importance of networks as a form of collaborative public management and look inside their operations.
it is well known (1) that “the age of the network” has arrived (Lipnack and Stamps 1994), (2) that hierarchy and markets are being supplemented by networks (Powell 1990), (3) that public managers are enmeshed in a series of collaborative horizontal and vertical networks (Agranoff and McGuire 2003), and (4) that networks need to be treated seriously in public administration
Public manage- ment networks are, in every sense, collaborative connections like social networks, although they not only comprise representatives of disparate organiza- tions but also go beyond analytical modes. They are real-world public entities.
Agranoff and McGuire define collaborative management processes as “the process of facilitating and operating in multi organizational arrangements to solve problems that cannot be solved, or solved easily, by single organizations”
Although it is certainly true that mutual dependency is leading to an increasing number of horizontal relationships crossing many boundaries, lateral connections seem to overlay the hierarchy rather than act as a replacement for them.
Program specialists frequently (and more naturally) work across agency boundaries. Their work is technical or based on specialized knowledge, and it is geared to solving problems, belonging to epistemic communities, and acting on shared beliefs.
For the line administrator, however, it is largely business as usual most of the time, dealing with internal POSDCORB matters, along with increasing collaborative pressures.
Why do networked accountability bodies persist? Because they deliver different forms of public value to their multiple participants.
Networks can perform a great many public service purposes. They not only bring many parties to the table but also have the potential to expand the resource base. The most important element of the resource base is the potential for knowledge expansion, a function that administrators said was indispensable.
From knowledge comes the possibility of new solutions derived by, owned, and implemented by several parties.
The key to sustained network involvement is performance, and the key to performance is adding public value by working together rather than separately.
The first benefit is the value added to the manager or professional, such as learning new ways to collaborate, intergovernmental skills, and how to network, along with enhanced technical and information and communications technology skills. Second are the benefits accruing to the home agency, such as access to other agencies’ information, programs and resources; access to information and communications technology; cross-training of agency staff; and most important, enhanced external input into the internal knowledge base. Third are the collective process skills that accrue from working together over a sustained period of time—for example, developing interagency planning, piloting an adaptation of a new technology, developing a mutual interagency culture that leads to subsequent problem solving, and experimenting with electronic group decision technology. Fourth are the concrete results accrued, such as an action plan, a capability building conference, new interagency strategies, and multiagency policy and program changes. These types of value-adding performance results sustain administrators’ efforts in collaborative undertakings.
It is an accepted fact that bureaucratic structures have become more flexible and permeable over the past century. Today’s organizations are becoming more conductive—that is, they are continuously generating and renewing capabilities, bearing in mind the alignment between internal forces and external demands, including the importance of creating partnerships through internal–external interaction, building alliances and coalitions, forming and reforming teams across functions and organization boundaries, and collaborating to actively manage interdependencies. In this sense, perhaps bureaucracies and standing networks appear a good deal alike because both need to be concerned with managing complex partnerships, with blurring boundaries. The difference is that one structures and creates rules and strategies under the umbrella of one organization, whereas the other must interorganizationally and collectively create structures, rules, and strategies that fit their multiorganizational needs.
In the 14 public management networks studied, four types of public value were queried, and managers found substantial benefits in each dimension.
- The first benefit is the value added to the manager or professional, such as learning new ways to collaborate, intergovernmental skills, and how to network, along with enhanced technical and information and communications technology skills.
- Second are the benefits accruing to the home agency, such as access to other agencies’ information, programs and resources; access to information and communications technology; cross-training of agency staff; and most important, enhanced external input into the internal knowledge base.
- Third are the collective process skills that accrue from working together over a sustained period of time—for example, developing interagency planning, piloting an adaptation of a new technology, developing a mutual interagency culture that leads to subsequent problem solving, and experimenting with electronic group decision technology.
- Fourth are the concrete results accrued, such as an action plan, a capability building conference, new interagency strategies, and multiagency policy and program changes. These types of value-adding performance results sustain administrators’ efforts in collaborative undertakings.
Virtually all of the 14 networks studied operated with some form of council or board, elected by the entire body of agency representatives, very much like the board of directors of a nonprofit organization.
The real work in all of the networks studied was done in either standing committees (e.g., finance, technology transfer, tele-medicine, educational applications, transportation technical review) or focused and usually shorter-term workgroups (e.g., ortho-infrared mapping, bicycle and pedestrian, broadband usage, community visitation, water and wastewater treatment)
Because all networks do not really make decisions, it is prefer- able to refer to many of their deliberative processes as “reaching agreements” rather than “decisions,”
In collaborative bodies, decisions and agreements are necessarily based on consensus, inasmuch as participating administrators and professionals are partners,not superior–subordinates.
they are co-conveners, co-strategists, co–action formulators, co-programmers, and so on. It is also true that public agency administrators possess neither ultimate legal authority (except, of course, within one’s home agency domain) nor control over all technical information. Authority in the network is shared with the many stakeholders at the table: other administrators, program specialists, research scientists, policy researchers, and interest group and advocacy association officials. Among the partners, it is unlikely that any single agency or representative at the table will have the legal authority or financial resources to completely approach a problem.
informational: wherein partners came together almost exclusively to exchange agency policies and programs, technologies, and potential solutions. Any changes or actions were voluntarily taken up by the agencies themselves.
developmental: wherein partner information and technical exchange were combined with education and member services that increased the members’ capacities to implement solutions within their home agencies and organizations.
outreach: wherein the activities of the developmental network were engaged; in addition, however, they also blue-printed strategies for program and policy change that led to an exchange or coordination of resources, although decision making and implementation were ultimately left to the agencies and programs them-selves.
action networks, wherein partners came together to make interagency adjustments, formally adopt collaborative courses of action, and deliver services, along with information exchanges and enhanced technology capability.
Creating a collective power of new possibilities.
The issue is to bring about enough cooperation among disparate community elements to get things done.
In order to open up new possibilities, the networks studied used six distinct predecision or agreement learning strategies.
They prepared for brokered consensus through (1) group discussion or exchange of ideas; (2) political negotiation of sensitive concerns and intensely felt needs; (3) direct applica- tion of technology or preestablished decision rules or formats; (4) application of preestablished, formulaic procedures (e.g., those related to regulations, grants, or loans); (5) data-driven decisions or agreements (e.g., market studies, usage patterns, traffic or accident counts); and (6) predecision simulation or electronic base groupware or other decision techniques.
“Knowledge is a fluid mix of framed experience, values, contextual information, and expert insight that provides a framework for evaluating and incorporating new experiences and information”
Whereas data refer to discrete, objective facts, and information is a message in the form of a document or an audible or visual communication, knowledgeis more action oriented, both in process and in outcome. Knowledge management has two dimensions: explicit knowledge, which can be codified and communicated easily in words, numbers, charts, or drawings, and tacit knowledge, which is embedded in the senses, individual perceptions, physical experiences, intuition, and rules of thumb
In the networks studied, the process of knowledge management in many ways defined the major focus of their standing committees and working groups.
First, essentially all of them began by surveying the universe of data and information that their partners had developed or could access, plus external databases of use to them.
Second, this information then used to develop their “own source” explicit knowledge using resources such as libraries, map inventories, strategic plans, fact sheets and policy guides, focused studies, surveys, conferences and workshops, electronic bulletin boards, process reviews, long-range plans, models and simulations, and market studies.
Third, tacit knowledge was rarely formally codified, but it was regularly approached through stakeholder consultations, best practices booklets, workgroups as “communities of practice,” study project report panels, expert presentations, specialized workshops, SWOT workshops, hands-on technical assistance, community leadership development sessions, forums on “what works,” direct agency outreach, help desks, and public hearings.
Fourth, the networks tried to organize the explicit/ tacit interface not through codification but through informal feedback on the myriad of knowledge management activities in which they engaged, usually through some informal post-project assessment or at its board or steering committee meetings.
Fifth, most of the networks directly served some of the knowledge management needs of their partner agencies by producing formal reports, responding to data requests, supplying modeling and planning data, circulating policy reports, sponsoring in-agency forums and report sessions, providing technical expert linkages between the network and specific agencies, and in some cases, providing agency-requested studies.
In the same way that organizations seek structured predictability, networks try to use their open-ended processes of coordinating purposeful individuals who can apply their unique skills and experiences to the local problem confronting the collaborative undertaking
Beyond the formal structure of the governing body and working committees and groups were four elements of power.
First, virtually every network had a champion (and in two cases, two champions)—a visible, powerful, and prestigious public agency head or nonprofit chief executive officer who organizes or sustains the network. The presence of the champion in the network signaled to others in the field to “stay in” and “cooperate.”
Second, there was a political core, normally comprising the primary participating department heads or federal government state directors and chief executive officers of the non- governmental organizations. These managers tended to be part of the governance structure, they sent a message to other participants that the network was important to be involved with, and they were the people who were most likely to be involved in high- level interagency negotiations and resource accommodations.
Third, there was a technical core, primarily workgroup or committee activists who knew the most about a particular topic (e.g., watershed management, planning, geographic information systems, finance, regulation, information and communications technology, and so on). Because a great deal of the work was bound up investigating problems, creating knowledge, and looking for feasible solutions, their work was at the core of network activity, and the most knowledgeable of these individuals held considerable operating power.
Finally, there were paid staff who held the network together through their support efforts, which in the 14 networks ranged from one or two persons who devoted to the network full time to 18 full- or part-time participants in one action network.
This power structure is deep, and the four dimensions overlap in practice—it is every bit as real as those in the organizations from which representatives are drawn.
If managers give up or add to the job of internal operations to engage in cooperation, they obviously do this at some cost.
Many line managers are said to be protective of agency autonomy for one of four reasons: (1) the agency manager knows best, and therefore should carry out its mission and programs; (2) loss of autonomy is associated with the loss of control and guidance of the agency; (3) people place a greater value on losses than on gains; and (4) autonomy reduces uncertainty
Real costs associated with network participation that the managers and professionals articulated. Six general cost categories were indicated:
(1) time and opportunity costs lost to the home agency as a result of network involvement
(2) time and energy costs resulting from the protracted decision-making process, based on nonhierarchical, multiorganizational, multicultural human relations processes
(3) agreements not reached because of the exertion of organizational power or the withholding of power
(4) network gravitation toward consensus- based, risk-aversive decision agendas
(5) resource “hoarding,” or agencies’ failure or unwillingness to contribute needed resources
(6) public policy barriers embedded in legislation, coupled with legislators’ or other policy makers’ unwillingness to make needed changes, which, in turn, frustrated collaborative decisions.
All of these appear to thwart progress within networks.
To a degree, the deliberations of the network and the in- volvement of nongovernmental organizations clearly influenced the courses of action taken by government, and in some cases, new programs and strategies emanated from network deliberations.
Three Large Caveats
First, when it comes to policy decisions, it is almost always the public institutions that make the ultimate call, and in the case of implementation, it is the agency.
Second, in virtually every public management net- work, it is government administrators at federal, state, and local levels who are the core or among the core actors in the network. They are able to inject legisla- tive, regulatory, and financial considerations right into the network mix, which hardly marginalizes them.
Third, many collaborative efforts outside the network form are more tightly controlled by the government, in the form of grant expectations, contract provisions, or loan conditions, tying the nongovernmental orga- nization to the public agency in a tighter way.
Today’s wicked policy problems, dispersed knowledge and resources, first- and second-order effects, and intergovernmental overlays guarantee that managers must engage other governments and nongovernmental organizations
