Module 4 through 7 has an associated group discussion that should focus on discussing the course content for that Module. Each discussion will span the two-weeks of the Module. Each stude

Jill M. Purdy is associate professor in the

Milgard School of Business at the University

of Washington Tacoma, where she also

serves as academic director of the Center

for Leadership and Social Responsibility.

Her research focuses on the intersection

of private enterprise, the public sector,

and society at large. Current research

projects examine social entrepreneurship

and the social practices of multinational

corporations internationally. She serves on

the editorial board of Negotiation and

Confl ict Management Research.

E-mail: [email protected]

A Framework for Assessing Power in Collaborative Governance Processes 409

Public Administration Review,

Vol. 72, Iss. 3, pp. 409–417. © 2012 by

The American Society for Public Administration.

DOI: 10.111/j.1540-6210.2012.02525.x.

Jill M. Purdy University of Washington Tacoma

Th e growing use of collaborative methods of governance raises concerns about the relative power of participants in such processes and the potential for exclusion or domination of some parties. Th is research off ers a framework for assessing power that considers authority, resources, and discursive legitimacy as sources of power and considers the participants, the process design, and the content of collaborative governance processes as arenas for power use. A case study of a collaborative governance process is presented and analyzed using the power framework. Implications for the design of collaborative governance processes are discussed, including the benefi ts of a multidimensional defi nition of power, tools for managing power imbalances among participants, and strategies that participants can use to participate more fully in collaborative governance processes.

State and federal agencies are using collaborative processes so often that their use has become institutionalized (Cheng 2006). Collaboration

describes “a process through which parties who see dif- ferent aspects of a problem can constructively explore their diff erences and search for solutions that go beyond their own limited vision of what is possible” (Gray 1989, 5). Collaborative governance processes such as multistakeholder roundtables, dispute resolu- tion processes, community advisory councils, and regulatory negotiations bring together organizations from the government, business, and nonprofi t sectors to collaborate on problems of mutual concern. Th e benefi ts of collaborative processes include greater responsiveness to complex situations and more deliberation than traditional governance processes (Leach 2006). Collaborative governance may produce more eff ective, effi cient, and fl exible policies (Sousa and Klyza 2007) with greater public acceptability.

However, using collaboration to govern also has fl aws and weaknesses. Gerlak and Heikkala describe “many institutional and political obstacles to collaboration, including confl icting agency goals and missions, infl exible administrative and legal procedures, and constrained fi nancial resources” (2005, 658). Th e

incomplete legal foundation for collaborative processes raises questions about authority, transparency, and accountability (Bingham 2009). Th e context of col- laborative governance may not fairly balance private interests and public authority (Sousa and Klyza 2007). Critical interests may not be represented (Leach 2006), and collaborative processes may bias decisions toward the participants with greater resources. Finally, collabo- ration can be a way of advancing self-interested goals such as increasing power (Huxham and Vangen 2000).

Many of these concerns are linked to power dis- parities among participating organizations and how power aff ects such issues as representation, participa- tion, and voice. Th is article off ers a framework for assessing power and how power is used in collabora- tive governance processes. Power is described along multiple dimensions, including authority, resources, and discursive legitimacy, and three arenas for power use are considered here: the participants, the process design, and the content of collaborative governance processes. A case study off ers examples of how the framework can be used to assess power. Th e article concludes by discussing the implications of the power framework for understanding and designing collabo- rative governance processes, including how power imbalances may be addressed.

Sources of Power and Arenas for Power Use in Collaborative Governance Collaboration occurs in the context of public manage- ment when stakeholders work together with gov- ernment to create new policies or to address public problems. Authors refer to this variously as cross-sec- tor collaboration (Bryson, Crosby, and Stone 2006), new governance (Bingham, Nabatchi, and O’Leary 2005), collaborative public management (O’Leary and Bingham 2007), or collaborative governance (Carlson 2008). Th e term “collaborative governance” is used in this article to refer to processes that seek to share power in decision making with stakeholders in order to develop shared recommendations for eff ective, lasting solutions to public problems. Collaborative

A Framework for Assessing Power in Collaborative Governance Processes

410 Public Administration Review • May | June 2012

interact in complex relational webs. Hardy and Phillips (1998) propose three sources of power that are particularly useful for understanding interorganizational dynamics: authority, resources, and discursive legitimacy. Each of these sources of power is dis- cussed here.

Authority is the socially acknowledged right to exercise judgment, make a decision, or take action (Greenwald 2008). Authority is determined by relative status within the institutional context in which the participating parties are embedded. While authority may

be achieved coercively, most often it results from social agreement to delegate power over defi ned areas to a particular organization or role. As Mary Parker Follett noted, the “true state is utterly dependent upon us for its appearance . . . we do not have a sovereign state until we make one” (1918, 315). Th e authority of a government agency to set and enforce rules is accepted because citizens share

a belief in the “rationalization and the regulation of human activity by legal and bureaucratic hierarchies” (Friedland and Alford 1991, 248). Th is bureaucratic logic permeates most collaborative govern- ance processes because they are convened by governmental bodies, resulting in a rational-legal interpretation of authority. Th e author- ity of government is tied to its rights to establish and enforce rules, while the authority of nongovernmental organizations (NGOs) and citizens stems from their right to participate in governance or to pursue legal action.

While authority may appear to be an objective form of power, it can shift over time as the social order is renegotiated. Th e negotiation of authority can occur at the macro level through legislative and judicial processes or at the micro level through group-level discus- sion (Follett 1918). For collaborative governance processes, viewing authority as malleable recognizes that participants with authority can share it with others, for example, by committing to enact the group’s recommendation rather than accepting it as advisory.

Th e “power over” perspective frames authority as a trump card that dictates which participant ultimately gets to decide an issue. Th e “power to” perspective suggests that authority is vital to the success of collaboration—it “gives teeth to a collaborative eff ort” (Straus 2002, 40). Without authority, the shared problem or common goal that brought participants together in a collaborative governance process may not be solved or achieved even if the group reaches consensus. From the “power for” perspective, authority also can be used to sanction the participation of stakeholders who might other- wise be marginalized. Empowering others through participation in decisions “increases the total capacity for eff ective action rather than increases domination” (Kanter 1977, 166).

Resource-based power recognizes the dependencies among organiza- tions involved in collaboration and their ability to deploy resources. Resources include tangibles such as fi nancial resources, people, technology, and supplies; and intangibles such as knowledge, culture, and capabilities. Resource power is distinct from authority; for example, the information held by government agencies gives them infl uence that extends beyond their authority to make rules (Freeman and Langbein 2000). Organizations depend on human

governance is a cross-sector concept spanning the public, private, nonprofi t, and citizen domains (Policy Consensus Initiative 2005).

Prior research identifi es two primary concerns about power in cross- sector collaborations: convening stakeholders and managing power imbalances (Bryson, Crosby, and Stone 2006). First, adequate power is needed to convene stakeholders, but government agencies often act as both conveners and participants in collaborative processes, raising questions about their ability to dominate such processes (Broome 2002). Second, actors who are less powerful in terms of resources, voice, or legitimacy may be ex- cluded from collaborative processes or may be co-opted by more dominant parties (O’Toole and Meier 2004). Despite the importance of such concerns, little theory exists to guide conveners, participants, and researchers in understanding how power shapes collaborative processes and outcomes. Huxham and Vangen note that “there is no coherent body of litera- ture on power in collaborative settings” (2005, 174).

Analyzing power in collaborative processes is challenging because they are ambiguous, complex contexts in which participants, social structures, and processes can change rapidly (Huxham and Vangen 2000). Th eories addressing personal forms of power are inadequate for understanding collaborative processes, while models emphasiz- ing structural aspects of power are limited by their focus on a single organizational context. To understand power in collaborative gov- ernance processes, one must consider power’s political, economic, and social aspects (Bierstedt 1950), as well as its structural, relation- al, and cognitive aspects (Hardy and Phillips 1998). A framework is needed that addresses the sources of power and the dynamics of power so that we can understand both power attributes and how relational power changes over time.

Conceptualizing power in terms of its sources implicitly assumes that power is a resource that can be expanded, diminished, or trans- ferred. However, a purely zero-sum, resource-based view of power is inadequate for collaborative contexts. Power can be used to advance the joint eff orts of the collaborators, resulting in mutual gain, or to empower others to participate more eff ectively in the collaboration, resulting in altruistic gain. Huxham and Vangen place the three power orientations of own gain, mutual gain, and altruistic gain on a continuum and label them “power over,” “power to,” and “power for” (2005, 175). Concern for others’ interests is most likely to emerge under conditions of good communication, trust, and shared goals (Norris-Tirrell and Clay 2010), emphasizing that context and process as well as resources play a role in determining power orientation.

Th e framework off ered here considers both the sources of power held by collaborative governance participants and the arenas that collaborative processes provide for the use of power. Th rough these twin foci, the framework describes the kinds of power held by participants in collaborative processes and reveals how power can be exercised structurally and relationally in collaborative processes.

Sources of Power: Authority, Resources, and Legitimacy In collaborative governance processes, organizations and coalitions with varying degrees of cohesiveness, resources, and political clout

[L]ittle theory exists to guide conveners, participants, and researchers in understanding

how power shapes collaborative processes and outcomes.

A Framework for Assessing Power in Collaborative Governance Processes 411

governance practice. To be authentic requires the use of appropriate organization, methods, and tools; facilitative leadership; and delib- erative space free of coercion” (Booher 2004, 44). Th e three essential elements embedded in Booher’s defi nition of authentic collaborative governance processes are the participants, the process design, and the content. Each of these components provides opportunities for the exercise of power.

Participants describe who is involved in a collaborative process and who leads it. Vital participants in collaborative governance processes are sponsors who identify an issue and initiate action, conveners who gather diverse people to work on common problems, and a

neutral facilitator to negotiate interests, inte- grate resources, and establish accountability (Policy Consensus Initiative 2005). Partici- pants should include those with formal power to make a decision, those who can block a de- cision, those aff ected by a decision, and those with relevant information or expertise (Straus 2002). Participation is often determined by leaders whose interpretation of the situation

determines which stakeholders are invited to collaborate and which are excluded. Fung (2006) notes that the degree of democracy in a collaborative process can be determined in part by who is invited to participate. However, issues of participation extend beyond who is invited. Collaborative governance processes are voluntary, so some invited parties may choose not to participate because of mistrust of collaborative processes or a preference to address the situation through alternative means. Th ose who do participate vary in experi- ence and eff ectiveness with such processes.

While many studies of collaborative governance exist, public administration research has not looked closely at process design for collaboration (Bingham 2009). Process design describes the where, when, and how of collaborative governance, infl uencing the nature of interaction and the modes that are used for communication and decision making. Processes for collaborative governance must be designed with fl exibility to allow trial and error without creating ambiguity and confusion (Straus 2002), but the decision making of process design occurs before the actual content of the collabora- tion occurs. Process design determines whether participants feel fairly treated, and these perceptions of procedural justice infl uence satisfaction with the outcomes of the process (Brockner and Siegel 1996). Process design also helps convey status within a group, signaling who holds a leadership role and whether participants are equal.

A third arena for the exercise of power is the content of the col- laboration, or what issues are addressed and what outcomes are pursued. Deciding the scope of the issues is an important oppor- tunity for the use of power (Altheide 1988). Th e content of initial agreements in a collaboration aff ects the outcomes of the process (Bryson, Crosby, and Stone 2006), and the interpretations that people use to identify issues and understand alternatives are closely linked to the success of the process (Gray 2004). Finally, decisions about the content of a collaborative process determine who has a legitimate claim to participate in the process and how the proc- ess will unfold (Gray 1989), linking content-related power use to participants and process design.

resources to represent them in collaborative governance proc- esses, while information and knowledge resources are needed to comprehend and analyze the issues. Financial resources can allow organizations to gain expert advice or representation in collaborative processes, increasing their infl uence. In addition to using resources for their own benefi t, organizations can use resources to infl uence other participants in collaborative processes by rewarding them for support or compliance or by punishing them for dissension or noncompliance.

In collaborative contexts, the relational and perceptual aspects of power may be as important as the objective ability to command resources. For example, those with fi nancial resources often behave as if they hold power, while those lacking them typically feel disem- powered, even if they have alternative sources of power available (Huxham and Vangen 2005). Similarly, “people who are thought to have power already . . . who look like they can command more of the organization’s resources . . . may also be more infl uential and more eff ective in getting the people around them to do things and feel satisfi ed about it” (Kanter 1977, 168–69).

Discursive legitimacy refers to the ability of an organization to repre- sent a discourse or speak on behalf of an issue in the public sphere (Hardy and Phillips 1998). Organizations exercise discursive legiti- macy when they act on behalf of the values or norms of a society, such as the rule of law, the logic of economic rationality, or principles such as democracy or respect for diverse cultures. An organization with discursive legitimacy draws its power from the status of the val- ues or logic it represents. Th is form of power acknowledges that “in- terorganizational power relations cannot be fully understood without attention to the larger pattern of societal dominance” (Benson 1975, 233). Participants who lack authority or resources can exert power if they are perceived to speak on behalf of a societally important ideal, such as ecological preservation or racial equality.

Some parties have discursive power based upon the discourse that is rooted in society. Put simply, this means that power is attributed to them because of the way we, collectively, talk about them in relation to others. (Huxham and Vangen 2005, 176)

A participant in collaborative governance may represent relatively low status people within a society, such as immigrants, but may exert power by linking to a societal value, such as acceptance of diversity. Th is type of discursive power is strongest when the value represented is widely shared and the organization’s claim of repre- sentation for it is relatively uncontested.

Discursive power also stems from the ability to manage meaning by infl uencing how information is presented. Th is is particularly important in collaborative contexts because parties in collaboration are engaged in complex negotiations around developing common meanings and mutual identities (Huxham and Vangen 2005).

Arenas for Power: Participants, Process, and Content While collaborative processes are widely used in government, “not every process calling itself collaborative is an authentic collaborative

In collaborative contexts, the relational and perceptual aspects of power may be as important

as the objective ability to command resources.

412 Public Administration Review • May | June 2012

including meeting frequency and formality. Th e right to design col- laborative processes is linked to authority. Th e structure of public sector collaborations may be imposed by a host organization (Huxham and Vangen 2000); however, process design can also be a collaborative proc- ess (O’Toole and Meier 2004). An organization with authority may challenge deep structure notions about power by sharing its power to design the process. Once a process has been designed, the type and du- ration of interactions determine opportunities for participation and for the use of power (Sharfman, Gray, and Yan 1991). Processes can limit participants to listening as spectators, or they can enable participants to deliberate and negotiate (Fung 2006). Resources may infl uence such process factors as meeting frequency, meeting location(s), and options for participating in person, by telephone, or by videoconference. Th e costs of hosting and participating in collaborative processes, such as travel, lost work time, meeting space, note taking, and communication, are also linked to the availability of resources. Discursive legitimacy aff ects the modes and frequency of expression during meetings. Greater discursive legitimacy can lead to domineering behavior and one-way fl ows of information, or it can be used to perform a gatekeeping func- tion that ensures equitable participation from all parties. In the deep structure, discursive legitimacy aff ects how participants communicate about the process to each other and to their constituents.

In the content arena, authority allows an organization to set the agenda and establish other participants’ expectations regarding the outcome of the process. Other participants may draw on indirect sources of authority such as legal rights to shape the topics or scope of discussion. Resources provide organizations with the ability to collect, share, and interpret information about the issues and topics under discussion. For example, an organization with the capability of producing meeting records can control the scope and depth of documentation, which might infl uence future meetings. Discursive legitimacy can infl uence the prioritization of issues as participants assert the dominance of one issue or perspective over another. In the deep structure, discursive legitimacy operates by infl uencing the framing of issues. Frames are lenses that people use to interpret information and make sense of a situation (Gray 2004). Participants come to the process with cognitive frames that defi ne what the issues mean to them, then engage in interactive issue framing to construct a shared meaning through dialogue (Dewulf et al. 2009). Power can be used to promote certain outcomes by imposing particular frames, or it can be used to prevent the discovery of common frames and shared meanings. Collaborative initiatives can be derailed if participants are unable to understand each others’ frames (Gray 2004).

Th e framework presented here allows assessment of the power sources of participants to a collaborative process, how power can be

Framework for Assessing Power in Collaborative Governance Processes A framework for assessing power in collaborative governance processes can be created by juxtaposing the three sources of power with the three arenas for power described earlier. Th e framework, shown in table 1, describes how authority, resources, and discursive legitimacy can be used to infl uence the participants, process design, and content of a collaborative process. Th is framework goes beyond rights-based approaches to power to allow a more nuanced investi- gation of power and its use in collaborative governance processes.

Th e framework acknowledges that collaborative processes are not objective, predetermined structures but are themselves imbued with interests and power (DiMaggio and Powell 1983). Power can be visible on the surface of interactions as overt infl uence attempts, but it also can be deployed more subtly to frame conversations and to promote or marginalize participants’ voices. Frost (1989) describes such eff orts to shape perceptions of social reality as “using power in the deep structure.” Th e deep structure is a shared system of mean- ing that operates in the collective unconscious of actors (Frost 1989; Hardy 1994). Deep structure uses of power attempt to reinforce or change taken-for-granted roles, structures, and modes of interaction (Everett and Jamal 2004; Frost 1989). Th e framework includes both surface uses of power, such as the inclusion of a particular stake- holder in a collaborative process, and deep structure uses of power, such as the language used to present issues for discussion.

In the participation arena, authority can be used to determine who is invited to participate and how broadly participation extends. More inclusive processes invite open public participation or self-selection, while less inclusive processes rely on the participation of elected offi cials or expert administrators (Fung 2006). However, merely constructing an open, inclusive process may not ensure that invitees participate. Resources may infl uence the number able to partici- pate and how well-informed or expert participants are. In the deep structure, power can be used to infl uence the degree to which people engage in the collaborative process. Discursive legitimacy can aff ect the status of representatives and their ability to participate. Partici- pants need to believe that they deserve a seat at the table (Hardy and Phillips 1998) and must be trusted to make commitments on behalf of their organizations to participate fully (Leach 2006). On the other hand, participants representing coalitions or collaborative networks may participate more vigorously because they represent the collective power of a group (Butterfi eld, Reed, and Lemak 2004).

In the process design arena, authority shapes beliefs about who owns the process and expectations about how the collaboration will proceed,

Table 1 Framework for Assessing Power in Collaborative Processes

ARENAS FOR POWER

Formal Authority Resources Discursive Legitimacy

PROCESS ELEMENTS

Participants Selection of participants Limits on participants

Number of representatives Expertise of representatives

Status of representatives Use of coalitions

Process Design

Ownership of the process Interaction expectations for the process Number, length, and location of meetings

How the process is paid for Frequency of voice Methods of voice Communication about the process

Content Setting the agenda Outcome expectations for the process Use of indirect authority such as legal rights

Distribution of information Understanding and analyzing the issues Production of meeting records

Prioritization of issues Framing of the issues to be addressed

A Framework for Assessing Power in Collaborative Governance Processes 413

state, and tribal governments and demonstrate that the project operates in the public interest. Th e process for licensing a