Theoretical Perspectives on Policy Implementation

The term “implementation” as a popular concept in contemporary discourse among scholars of public policy dates back to the work of Jeffrey Pressman and Aaron Wildavsky (1973) in the early 1970s. Research on policy implementation provides the essential link between political and economic analyses of policy implementation and the organizational/institutional analysis of public administration (Hjern and Hull 1987). This research has been through some major phases of development. Three phases – commonly referred to as the first, second and third generations – can be identified in the literature (Goggin 1990; Howlett and Ramesh 1995; Pal 2006). Elaborating on these approaches is beyond the scope of the present work, but a brief overview, however, will serve as a context for advancing our understanding of the multi-actor implementation framework proposed here.

When it was originally developed as a field of inquiry, research on policy implementation was marked by the emergence of a top-down approach in the scholarly literature (Bardach 1977; Pressman and Wildavsky 1973; Sabatier and Mazmanian 1981). The theoretical and empirical assumptions of this approach were immediately criticized as excessively mechanistic and unable to do justice to the realities of policy delivery in democratic societies. The critics who espoused a bottom-up approach were unified by their effort to examine the politics and processes of policy implementation, starting from the frontlines of public administration, where street-level public officials often interact with organized societal interests (Barrett and Fudge 1981; Elmore 1981; Kickert 1997; Klijn 1996). The debates on the relative merits of the top-down and bottom-up approaches were grouped under the label of “first-generation implementation research” (Hill and Hupe 2002).

A consequence of the normative schism between the two traditions was the theoretical impoverishment of first-generation research on policy implementation. A new generation of scholars emerged in the late 1970s and early 1980s – a second generation of research – who synthesized the insights of the top-down and bottom-up approaches into a conceptual framework that consisted of a set of theories of implementation (O'Toole 1986; Palumbo and Calista 1990; Sabatier 1986). This synthesis approach, however, has its own problems – especially its tendency to be little more than a combination of variables from the two perspectives, which leaves the reader with a long list of variables and complex diagrams of causal chains (Exworthy and Powell 2004; Linder and Peters 1987; Sinclair 2001).

A third generation of researchers, who distilled the large number of variables into a manageable framework, eventually emerged in the late 1980s and early 1990s (Winter 1990). They hoped to develop more elegant theories that could lend themselves to broader generalizations and more longitudinal inquiries (Goggin 1990). As Laurence O'Toole Jr. (2000) notes, however, this effort proved too ambitious, because very few scholars have so far been willing to undertake such inquiries. In the 1980s, moreover, the process of policy implementation was influenced by structural changes in public administration towards decentralization, devolution of responsibilities, partnerships, and the restructuring of accountability relationships in service delivery (Kettl 2000; O'Toole 2000; Pal 2006). As a result of such transformations, public policies are increasingly being implemented in concert with non-state actors in cooperative or collaborative partnership arrangements. These new inter-organizational partnerships are not merely a passing fancy but are likely to be permanent features on the landscape of policy implementation (Kernaghan, Borins, and Marson 2000).

The central concern shared by theoretical perspectives on policy implementation, organization and governance is to understand how government organizations interact with their external environment in the delivery of policies.

As a result of transitions towards complex and multi-actor policy processes, the focus of research on implementation shifted from trying to build meta-theory towards explaining concerted action across institutional boundaries (Lindquist 2006; O'Toole 2000). Thus, one notices the broadening of the approach to research on policy implementation into a multi-focus perspective that looks at a multiplicity of actors, loci and levels (Hill and Hupe 2003). In federal systems, for instance, the different levels of policy action consist of federal, provincial or state and municipal jurisdictions and their agencies. The loci of policy action often consist of constellations of ideational and interest coalitions within and outside the state within a policy subsystem (Sabatier and Jenkins-Smith 1993).

Source: Canadian Public Administration (Journal) Volume 54, Issue 1, pages 121–142, March 2011, Policy implementation in multilevel environments: Economic development in Northern Ontario, Charles Conteh.