Policy Networks Research Paper

1. The Changing Context Of Policy-Making

This definition of policy networks signals a change in the way political decision-making has been conventionally understood. It contradicts the standard notion of a clear separation between state and society, with the state being seen as the highest center for social and political control. As Mayntz (1993) observes, social modernization has brought more importance to formal organizations in all areas of society and with their strengthened position comes an increasing fragmentation of power. Policy networks represent one response to the escalating functional differentiation of social subsystems. With the state no longer being the sole entity capable of organizing society, there is a dispersion of expertise and competence, a multiplication of channels for mediation and agreement, and the involvement of different levels of decision-making from the local to the supranational (Jouve 1995). Consequently, public policy is much less likely to be produced by a unitary state with adequate control over all the resources necessary for implementing decisions in the public interest.

Rather, public policy is more likely to result from strategic interaction among several or many policy actors, each with a particular understanding of the problem and its own individual or institutional self-interest (Scharpf 1997).

1.1 Historical Development

The sharing of political power between public and private actors became the first avenue of enquiry that led to the development of the policy network concept. In the mid-1970s, Schmitter and Lehmbruch (1977) suggested the concept of corporatism be reconsidered as a conceptual tool for understanding the particular relationships between public and private actors that had emerged for the management of macroeconomic policy in several European democracies. The revitalization of the corporatism concept coincided with other suggested new concepts for capturing the changing patterns of power sharing between private and public actors: ‘corporate pluralism’ in Scandinavia, ‘issue networks’ in the US (Heclo 1978), ‘policy communities’ in the UK (Richardson and Jordan 1979), and ‘policy networks’ (Katzenstein 1978).

The early focus in each of these concepts was on broad, macro characteristics of the political system. As these concepts became more familiar, however, a number of scholars noted their potential relevance for understanding private–public relations in more restricted sectoral policy arenas. Drawing on a growing number of empirical studies, these scholars began to develop typologies of different structural patterns of public–private interaction at the sectoral or meso level (Atkinson and Coleman 1989, van Waarden 1992), prompting a certain spate of definitional disputes. By the mid-1990s, however, sectoral and state corporatism, pressure pluralism, clientelism, issue networks, and state-directed networks generally were accepted as core structural types.

In addition to patterns of power sharing between public and private actors, a second line of inquiry developed focusing on the degree of integration among the various actors involved in these horizontal policy-making settings. Rhodes (1990) and Rhodes and Marsh (1995) targeted the degree of integration as their key variable and outlined a typology of different types of ‘policy networks’ that ranged from fairly tightly integrated ‘policy communities’ to very loose and open ‘issue networks.’ The prevalence of large organizations in horizontal arrangements in policy-making also triggered the interest of political sociologists interested in patterns of communication between large organizations. Laumann and Knoke (1987) sought to identify network structures by examining how information was transmitted between policy actors, the nature of the resource transactions between actors, and the relative openness of boundaries in various policy domains. These studies employed quantitative social network analysis developed in sociology. By the mid-1990s, this quantitative approach had incorporated insights from other policy network writings (Pappi and Henning 1998).

1.2 Current Views On The Concept

When speaking of the degree of integration of the policy network, it is usual to employ a concept that refers to the aggregate of actors involved. The term ‘policy community’ is utilized most commonly for these purposes, but terms like ‘policy domain’ or ‘policy subsystem’ are also used. Research has established a number of factors that are related to integration:

(a) Openness of the policy community. The more open deliberations in the policy community tend to be, the less integrated it is likely to become.

(b) Size of the policy community. The larger the number of members of the policy community, the greater the obstacles to high integration.

(c) Frequency of interactions among community members and the degree to which these interactions become institutionalized as opposed to remaining ad hoc and fluid. Institutionalization normally enhances the degree of integration.

(d) Membership stability. The more changeable the membership of a policy community, the less integrated it will be.

(e) Density and structure of information exchange. The denser the ties for information exchange and the more direct relationships are, the more integrated the policy community.

Another, particularly important, factor affecting the level of integration of a policy community is the degree to which its members share basic beliefs, values, and norms. Some actors within a given policy community may come to share a set of deep normative core beliefs and from these develop basic strategies and policy positions (Sabatier and Jenkins-Smith 1993). These ‘advocacy coalitions’ create ‘spaces of meaning’ within the policy community that provide a framework for political action. The degree of integration of the policy community will be affected by how many distinct spaces of meaning it holds. Sabatier and Jenkins-Smith suggest that policy communities will normally contain two to four advocacy coalitions. They add that the number will likely be higher when the policy community is newly formed, and may dwindle to one dominant coalition as policy-making becomes routinized. The concept of an epistemic community developed in international relations has certain properties in common with Sabatier and Jenkins-Smith’s advocacy coalition. All or part of an epistemic community may act as an advocacy coalition. When this development occurs, these epistemic communities often provide important links between policy communities, whether in other nation–states or at the transnational level.

Different types of policy networks may be distinguished depending on whether state actors share political power and on the relative resource balance between private and public actors. State actors may share political power by inviting private actors to participate directly in the formulation or implementation of policy. Here private actors become policy participants. Corporatist and clientelist networks have this feature. Or, state actors may monopolize policy formulation and policy implementation, while listening to, or receiving submissions from, private actors. In this relationship, private actors perform as policy advocates. Pressure pluralism is the most common network with this feature, but it also occurs in issue networks and statedirected networks. The decision by state actors to share political power affects the overall balance of resources between public and private actors in the policy network. Resources may be relatively balanced, they may favor the state actors, or they may advantage civil society actors. Both pressure pluralist and sectoral-corporatist networks feature a balanced relationship. Private actors have the stronger position in clientelist and issue networks while public actors dominate in the state-directed and state-corporatist types.

2. Theoretical Status Of The Concept

Although definitional disputes surrounding the concept have gradually been resolved, the overall theoretical status of policy networks remains subject to debate. Two fairly fundamental criticisms have been directed at the concept. First, Dowding (1995) has charged that the concept is nothing more than a metaphor and will thus fail to produce theories of the policy process. Second, the concept is said to have limited relevance because of an apparent lack of suitability for explaining policy innovations and policy change.

2.1 Policy Networks As Metaphors Only

Dowding has argued that the attempts to provide a middle range theory of policy-making utilizing the policy network concept are bound to fail. He suggests that the driving force of the explanation, the independent variables, are not characteristics of the network as such, but of the components within the networks. Variables such as the relative openness of a policy network are cast as a mere redescription of policy phenomena. Related criticisms argue that policy network studies have failed to explain both the mutual dependence of public and private actors and the conditions under which particular policy networks arise (Thatcher 1998). Critics also lament the lack of specificity of the hypotheses that do exist, a failing bound to render any empirical testing rather unconvincing.

Although useful for spurring discussion and reflection, there are several reasons why these criticisms may be premature. First, the criticisms undervalue the descriptive utility of the policy network concept. One of the most difficult empirical problems faced in the analysis of public policies is determining specifically who the consequential actors might be in any policy domain. Although one might be able to begin to specify these actors by utilizing a good understanding of formal institutions and of the persons occupying key positions, such a specification is almost inevitably bound to be incomplete. Policy-making is replete with informal relationships and the operation of subtle, somewhat opaque norms that empower certain actors in unexpected ways. Accordingly, the methodologies developed for delimiting a policy network have a broader use in policy analysis. By identifying the central actors in policy-making, they assist the analysis of the processes by which policy ideas are translated into particular policy proposals through interchanges between an identified set of actors. This identification process becomes particularly helpful in policy areas featuring multilevel governance as Peterson (1995) has argued in referring to the European Union. This argument also has relevance to such patterns of governance as they emerge in policy areas where globalization has a decided impact.

Second, numerous empirical studies subsequent to the publication of Dowding’s critique have provided convincing evidence of relationships between variations in policy networks and policy outcomes. Josselin’s (1997) study of the negotiation of core directives for a single market in financial services in the EU evaluates the extent to which different network types facilitate or hinder the promotion of member-states’ interests in EU-level negotiation settings. She shows that a relatively closed corporatist network at the national level is a decided asset in promoting a unified position at the EU level. Conversely, however, a more fluid issue network makes such unity more difficult, while providing additional degrees of freedom in the fluid decision-making arenas that exist at the supranational level. Similarly, Daugbjerg (1998) relates the relatively closed character of corporatist networks to patterns of political change. The more closed and cohesive the network, the more incremental the pattern of change and the more stable the utilization of policy instruments. Coleman’s (1996) study of the impact of globalization on the regulation of financial services at the domestic level traces relationships between particular types of policy networks and the varying adaptation of countries to these globalizing pressures.

Finally, Dowding’s critique is addressed by some scholars who are using more formal models. Pappi and Henning (1998) elaborate a complex model that suggests direct relationships between network types (pluralism, corporatism, clientelism, state-direction) and policy outcomes. Scharpf’s ‘actor-centered institutionalism’ introduces network-based negotiation settings as one of several core types of forums for policy-making. Drawing on game theory, he sets out a series of propositions that provide a basis for modeling negotiations between actors in such networks and deriving likely differences in policy outputs based on general welfare and distributive justice criteria. Both of these theoretical approaches would appear to hold considerable promise for the development of middle level theory where policy networks will be an important independent variable.

2.2 Policy Networks And Policy Change

Policy networks have been conceptualized primarily using structural variables. Such structural characteristics are helpful for building descriptions of the policy process at a particular point in time and, more recently, for positing testable explanations of variations in policy outcomes. Dowding (1995) questions, however, whether these same variables are sufficient for explaining how and why policies change and why networks themselves change. Two avenues of investigation have been pursued for addressing these criticisms. The first focuses on the relationship between policy ideas and policy networks and the second draws on an analysis of boundary rules.

Sabatier and Jenkins-Smith (1993) argue that policy networks include one or more advocacy coalitions that are composed of actors who share deep core beliefs, near core policy beliefs, and secondary beliefs about the use of policy instruments. Changes in the relative dominance of advocacy coalitions in a policy network may prompt policy change. Or members of coalitions may negotiate adjustment of secondary beliefs that, in turn, open the door to policy change. Paralleling Sabatier and Jenkins-Smith’s work, Jobert (1995) argues that a ‘political forum’ and a ‘scientific or expert’ forum exist alongside a policy network. When politicians begin to experience problems that cannot be addressed well within a policy network, they begin to reach to these other fora for advice. To the extent that this advice is acted upon, the policy network may be dissolved, with new arrangements taking its place.

What both these theoretical arguments stress is the importance of looking at the interaction between policy ideas and who shares which ideas within a given policy network. Coleman et al. (1996) follow up on these insights by arguing that the degree of sharing of policy ideas in a given policy network will determine the approach to policy negotiations among actors in the network. They hypothesize that the more ideas are shared, the more policy discussions will take on aspects of what Scharpf (1997) has called ‘positive coordination’ or a ‘problem solving’ decision-making style. They suggest that positive coordination is more likely in closed networks like corporatism than in more open networks such as pressure pluralism. By coupling policy network structures to policy ideas, they then trace out different patterns for how policy change occurs. Scharpf’s work stresses the importance of looking at actor preferences and cognitive orientations in studying policy processes. The richness of his theoretical work offers further opportunities for improving theories of policy change.

The second avenue for improving explanations of policy change focuses on different types of boundaries within policy communities. In their retrospective on policy network analysis in the UK, Rhodes and Marsh (1995) note that a policy community can have several policy networks. These, in turn, may operate at two levels; a central level whose members dispose of considerable resources, and a peripheral one composed of those with few or no resources. Changes in the actors at the center of a policy community and thus in the dominant policy network, or movements from the peripheral networks to the dominant one, will be crucial intervening variables in explanations of policy change. Similarly, new actors may build peripheral networks over time and then come to contest the exclusivity of the existing dominant policy network. To the extent that new actors enter an existing policy network, they may bring about a change in its structure. For example, entry of new actors into a previously closed corporatist network may foster sufficient dissent that the state refuses to involve private actors any longer in policy making. Once such actors are excluded, the network takes on pressure pluralist characteristics, altering, in turn, the policy process and perhaps policy outcomes (Coleman et al. 1996).

3. Methodological Issues

The empirical analysis of policy networks has centered on two variables: the degree of integration of policy communities and the division of power between state and civil society actors. Scholars have approached these empirical issues from a variety of methodological standpoints. Among these, the use of quantitative network analysis, elite interviews, and close examination of documents and archival materials are the most commonly used. The use of these methods varies depending on the particular empirical task in question.

3.1 What Are The Boundary Rules Of The Policy Community? Who Belongs?

Determining the boundaries of a policy network is one of the more difficult, yet more important empirical tasks in this approach. Understanding who belongs and who does not provides key information about likely political demands and patterns of political influence, factors, in turn, important to explaining policy outcomes. In addition, as noted above, observations of changes in policy network membership may be crucial to approaching the explanation of political change. Accordingly, boundary questions must be approached diachronically, with Sabatier and Jenkins-Smith (1993) suggesting a time perspective ‘of a decade or more.’

Those working in the public policy field have tended to approach boundary specification by looking at relationships and then at activities and events. In the relational approach to boundary specification, they seek to identify the actors involved in the formulation and implementation of policies in selected policy domains. With these relationships in focus, they then examine variations in the organizational attributes, political ideas, and degree of participation of these actors in selected activities and events. A mix of quantitative network analysis and qualitative techniques are used in these studies. The relationships of concern are usually determined using ‘snowball-sampling’ techniques to set up the right set of elite interviews. This methodology does have its weaknesses in that reliance on elite interviews is costly both in time and in money and it is difficult to replicate these interviews over time to get a dynamic perspective on boundary change.

3.2 Does The State Share Power Directly With Policy Community Members? What Is The Balance In The Distribution Of Resources In The Policy Community?

The answers to these questions provide the empirical base for determining the type of policy network in place in a given policy community. In beginning, to assess the nature of the policy network, it is essential to understand well the organizational attributes of the actors in the policy community. One of the crucial empirical legacies of the corporatist literature were some empirical guidelines for assessing such policy capacity.

The policy capacity of interest associations and of associational systems may be evaluated by looking successively at three types of phenomena. First, it is useful to assess whether associations in a policy network compete with one another for members and are part of a more comprehensive associative order. The development of an associative order is evaluated by examining the degree of horizontal and vertical integration among associations. Second, some evaluation of the internal resources of the associations is useful for assessing likely policy capacity. Frequently, however, these kinds of analysis are still insufficient, particularly for determining the degree to which associations participate in policy-making. Normally, before final determination of the structure of a policy network, it is useful again to turn to elite interviews for additional information on the extent to which civil society actors participate directly in policy formulation or implementation.

4. Future Challenges

The concept of a policy network was designed in the first instance for the analysis of policy making at the national and more local levels. As globalization has increased the internationalization of many policy areas, however, the likelihood rises that several policy networks will be interacting with one another across different levels of decision-making. Future research, therefore, will need to focus on improving understanding of the relationships between policy networks. Some of these networks will have the additional property of being transnational in character, that is, they will include actors from both the nation-state and the international levels.

The concept of policy community mediators offers one avenue for research on the relationships between policy networks (Coleman and Perl 1999). In many internationalized policy domains, boundary rules will create overlapping memberships between policy networks, with mediators emerging from among those actors with joint memberships. Mediation can take place at two levels. First, a mediator may operate as a pragmatic policy broker who looks for opportunities to lessen conflict within and between policy networks in order to move the policy process forward. Second, perhaps even more important, mediation takes place on an intellectual plane, involving the ‘translating’ of policy ideas and policy paradigms between policy networks. Drawing on Kuhn’s work on scientific paradigms, Schon and Rein (1994) suggest that translation is crucial to resolving conflicts between different policy paradigms or frames. Mediators construct a new view of the world that has a place for the policy concerns of both networks and that provides a language and set of norms that enable them to speak to one another.

Success in this second role requires that the mediator be accepted as legitimate in both policy communities. Legitimacy will derive from having both adequate expertise and a socialization experience that permits straddling the boundaries of the communities concerned (Jouve 1995). Research on linkages between policy networks and on the role of policy mediators will be crucial to further theorizing on policy networks in this globalizing era.

Bibliography:

1. Atkinson M M, Coleman W D 1989 Strong states and weak states: sectoral policy networks in advanced capitalist economies. British Journal of Political Science 19: 47–67
2. Coleman W D 1996 Financial Services, Globalization and Domestic Policy Change. St. Martin’s Press, New York
3. Coleman W D, Perl A 1999 Internationalized policy environments and policy network analysis. Political Studies–London 47: 691–709
4. Coleman W D, Skogstad G, Atkinson M 1996 Paradigm shifts and policy networks: cumulative change in agriculture. Journal of Public Policy 16: 273–301
5. Daugbjerg C 1998 Linking policy networks and environmental policies: nitrate policy making in Denmark and Sweden 1970–1995. Public Administration 76: 275–94
6. Dowding K 1995 Model or metaphor? A critical review of the policy network approach. Political Studies–London 43: 136–58
7. Heclo H 1978 Issue networks and the executive establishment. In: Beer S H, King A (eds.) The New American Political System. American Enterprise Institute for Public Policy Research, Washington, DC
8. Jobert B 1995 Rhetorique politique, controverses scientifiques et construction des normes institutionnelles: esquisse d’un parcours de recherche. In: Faure A, Pollet G, Warin P (eds.) La Construction du Sens dans Les Politiques Publiques: Debats Autour de la Notion de Referentiel [The Construction of Meaning in Public Policies: Debates surrounding the concept of the Referentiel ]. L’Harmattan, Paris
9. Josselin D 1997 Money Politics in the New Europe: Britain, France and the Single Financial Market. Macmillan, Basingstoke, UK
10. Jouve B 1995 Reseaux et communautes de politique publique en action. In: Le Gales P, Thatcher M (eds.) Les Reseaux de Politique Publique. Debat autour des Policy Networks [Public Policy Networks: Debates on Policy Networks]. L’Harmattan, Paris
11. Katzenstein P 1978 Conclusion: domestic structures and strategies of foreign economic policy. In: Katzenstein P J (ed.) Between Power and Plenty: Foreign Economic Policies of Advanced Industrial States. University of Wisconsin Press, Madison, WI
12. Laumann E O, Knoke D 1987 The Organizational State. University of Wisconsin Press, Madison, WI
13. Mayntz, R 1993 Policy-netzwerke und die Logik von Verhandlungssystemen [Policy networks and the logic of negotiation systems]. In: Heritier A (ed.) Policy-Analyse: Politische Vierteljahresschrift Sonderheft 24. Westdeutscher Verlag, Opladen, Germany
14. Pappi F U, Henning C H 1998 Policy networks: more than a metaphor? Journal of Theoretical Politics 10: 553–75
15. Peterson J 1995 Decision-making in the European Union: towards a framework for analysis. Journal of European Public Policy 2: 69–93
16. Rhodes, R 1990 Policy networks: a British perspective. Journal of Theoretical Politics 2: 293–317
17. Rhodes R, Marsh D 1995 Les reseaux d’action publique en Grande Bretagne. In: Le Gales P, Thatcher M (eds.) Les Reseaux de Politique Publique. Debat autour des Policy Networks [Public Policy Networks: Debates on Policy Networks]. L’Harmattan, Paris
18. Richardson J J, Jordan A G 1979 Governing Under Pressure: The Policy Process in a Post-Parliamentary Democracy. Robertson, Oxford, UK
19. Sabatier P A, Jenkins-Smith H C 1993 Policy change over a decade or more. In: Sabatier P A, Jenkins-Smith H C (eds.) Policy Change and Learning: An Advocacy Coalition Approach. Westview Press, Boulder, CO
20. Scharpf F W 1997 Games Real Actors Play: Actor-centered Institutionalism in Policy Research. Westview Press, Boulder, CO
21. Schmitter P C, Lehmbruch G 1977 Trends Toward Corporatist Intermediation. Sage Publications, London
22. Schon D A, Rein M 1994 Frame Reflection: Toward the Resolution of Intractable Policy Controversies. Basic Books, New York
23. Thatcher M 1998 The development of policy network analyses: from modest origins to overarching frameworks. Journal of Theoretical Politics 10: 389–416
24. van Waarden F 1992 Dimensions and types of policy networks. European Journal of Political Research 21: 29–52