Permeability and Openness of Boundaries and Borders

In applying systems concepts to the study of social organization, Katz and Kahn conclude that “living systems, whether biological organisms or social organizations, are acutely dependent of external environments,” asserting that they “must be conceived as open systems” (1978, p. 208). “Unless ‘energy’ of some sort is imported (see Katz and Kahn, 1978) into the social system that system will tend to break down” (Peters, 2008, p. 65). According to systems theorists, “that a system is open means, not simply that it engages in interchanges with its environment, but that this exchange is an essential factor underlying the system’s viability, its reproductive ability or continuity, and its ability to change” (Buckley, 1998, p. 44). The importance of the external environment to the regulation of a social system has been long recognized as an important systems feature. As Mintzberg observes, “... external controls forces the organization to be especially careful about its actions. Because it must justify its behaviors to outsiders, it tends to formalize [these external controls]” (1979, p. 290).

We have discussed the multiplexity of social ties formed between two or more nodes. We may also view social ties in terms of their permeability or openness. Strong, formal ties give rise to tight bonds between actors, making it hard to bring new actors into the network. Granovetter’s classic discussion of the “strength of weak ties” is built on the premise that social systems, as living systems, will require some kind of exchange of energy or resources with their environment. Weak ties are more permeable, and in their capacity to break we find their value. Weaker ties are more amenable to the development of new ties than are stronger ties. Open systems with permeable boundaries possess a greater capacity to build bridges or links to nodes or entirely other networks. The more that a system limits its exchanges with its wider environment, the more closed and essentially bonded a network is.

Linze Schaap discusses this permeability as a matter of network closure. “Closure,” he observes, “occurs when certain actors are excluded from the interaction, for example because other actors fail to appreciate their contribution or dislike their presence and attendance” (Schaap, 2008, p. 118). He distinguishes between two kinds of closure: social and cognitive (Schaap, 2008, p. 119). “Social closure or exclusion means that actors are excluded from the interaction, excluded from membership of the governance network.” Social closure is related to the capacity of the governance network to include or exclude members. Schaap suggests that cognitive closure can occur “when knowledge, information, ideas, or proposals are ignored and denied access to the agenda” (Schaap, 2008, p. 120). Cognitive closure may exist even when social openness between parts of the system exists.

Schaap suggests that when closure is used to create social exclusion it can be applied as a conscious and unconscious strategy (Schaap, 2008, p. 119). The capacity of a governance network to regulate its borders matters a great deal. “Closedness implies that steering signals [generated from the outside] do not penetrate into the system” (Kickert and Koppenjan, 1997, p. 55). The role of “steering signals” from their external environments is particularly relevant when accountability and governance structures are considered. Systems theory views these signals as feedback, a topic we will turn to later in this section.

The open or closed nature of governance networks may be found in discussions of differences between iron triangles, issue networks, and policy communities. Criticizing the iron triangle as being too closed and narrowly construed, Hugh Heclo posited the issue network to account for the relatively scale-free and open nature of policy networks. He defined issues networks as being comprised of “a large number of participants with quite variable degrees of mutual commitment

or of dependence on others in their environment____Participants move in and out

of networks constantly” (Heclo, 1978, pp. 102—3). “Issue networks tend to be the broadest, most extensive, and evanescent of the various kinds of networks” found in the public administration and policy literature (Gage, 1990, pp. 130—31). The clash or complementarily of interests between potential policy actors serves as the primary coupling force in scale-free issue networks. According to Koppenjan and Klijn (2004), scale-free issue networks form:

• 1. When the need for interaction arises for the first time between actors who were not previously aware of their mutual dependencies
• 2. When new problems or actors manage to penetrate existing networks, thus creating chaos so that new forms of consensus must be developed in order to tackle previously unknown, politicized problems, or to enable interaction between old and new participants
• 3. When problems cut across networks so that actors from different networks must learn to interact with one another (Koppenjan, 2008, p. 145)

Issues networks arise when two or more policy actors recognize the mutual dependencies that develop when conditions change, ill-structured, wicked problems materialize, or as the result of the fragmentation of an existing network. Issue networks may evolve into more formalized and closed policy communities (Rhodes,

1997). Policy communities tend to include “limited numbers of participants and a conscious exclusion of some groups” (Schaap, 2008, p. 112). Because of their closed nature, policy communities may tend to favor the status quo. In essence, policy communities tend to exhibit the closed qualities of iron triangles, but involve a potentially wider array of policy actors.

We suggest that governance networks exhibit qualities of openness and closedness that will likely be structured through a variety of boundary-forming and boundary-brokering activities, which in turn will likely have differential effects on the level of policy change from the status quo (Adams and Kriesi, 2007). This view of the permeability of governance network boundaries also allows for the possibility that some parts of the network may exhibit greater degrees of openness or closedness than others.