Introduction

The idea of policy networks assumed growing importance in the public policy literature of the 1990s. Defined as “(more or less) stable patterns of social relations between interdependent actors, which take shape around policy problems and/or policy programmes” (Kikert et al, 1997, p 6), the concept emerged originally in the US in the early 1950s as a critique of pluralistic explanations of political decision making. Pluralist theory posits a (more or less unlimited) number of groups competing (with more or less equal degrees of influence) for the attention of a largely disinterested state. Network analysis in contrast argues that a small number of groups enjoy a privileged relationship with the state at the expense of other interests. The approach was given particular form in the concept of ‘iron triangles’: a metaphor for the symbiotic relationship seen to exist between policy makers, government agencies and selected interest group(s) within a particular area of policy making (Peters, 1986).

As well as being distinct from pluralist analysis, network theory also differed from the other main model of interest group representation: corporatism. Unlike pluralism, in which all pressure groups are seen to have a roughly equal ability to influence the policy process, corporatist theory highlights the privileged role of certain, selected groups. Because of their key role in society, these groups have a ‘representational monopoly’ (Schmitter, 1979) that is recognised, licensed or created by the state. In the UK, for example, corporatism described the relationship between the state and the organised representatives of ‘capital’ and ‘labour’ that characterised the development of economic and industrial policy in the 1960s and early 1970s (Cawson, 1986). Despite obvious similarities to their own approach, policy network theorists such as Marsh and Rhodes (Marsh and Rhodes, 1992; Rhodes, 1997; Marsh, 1998) argue that the corporatist model may only be applicable in certain contexts. Policy making is more complex than the corporatists suggest. Rather than taking a ‘monolithic’ view of policy making, which sees all areas of government policy dominated by the same powerful groups, Marsh and Rhodes argue that it is important to “disaggregate policy analysis” (Marsh and Rhodes, 1992, p 4) and examine the particular forms of interest group representation that characterise specific policy areas.