Public policies are conceived as Programs of action promoted by the public authorities (Meny and Thoenig, 1989) focused on the identification and solution of collective problems, including not only the production of guidelines for action (values, goals, norms) but also the management of the collective action needed to actually implement those guidelines.
However, public authorities (politicians and upper level civil servants) are not the only relevant actors in policy making. The growing diversity and complexity of the social structures (Papadopoulos, 1994) and the resulting changes in the role of the State, highlight other relevant partners (interest groups, businesses, unions, the media) as well as the fragmentation of the major institutional actor (the State), into a myriad of players (the “balkanized” divisions of the public administration tecnostruture, the multiple layers of street level bureaucrats, the newcomers resulting from devolution and privatization processes). This overcrowding of the policy-making process produced models of policy analysis based on new concepts such as “policy network” (Thatcher, 2004) and “thematic network” (Heclo, 1978).
The wide range of actors and instances of policy making implicit in these network approaches make the ground for the concept of “public action” where the role of the State is counterbalanced by the consideration of other actors and instances, both public and private. This provides a more horizontal and circular (and less hierarchical) perspective on the analysis of the policy-making process (Commaille, 2004). Therefore, the public action approach we suggest in this study views the policy-making process as a non-linear, fragmentary and flexible process, with a complex set of actors, instances and interactions, before and after the moment of the formal policy decision (Thoenig, 2004).
The concept of regulation is used to identify and describe the processes through which collective action is promoted, guided and coordinated in order to achieve goals set up as desirable solutions, regarding specific situations conceived as problems. “By definition, regulation involves rules; however, these may be explicitly formulated anddocumented, or they may be tacitly understood by participants in some collective activity” (KNOWandPOL, Orientation 3, regulation: specifications).
Regulation is a polysemic word and this leads to the clarification of its meaning through some conceptual distinctions. Firstly, we distinguish between institutional and situational regulation. Institutional regulation is viewed as a “set of actions decided and implemented by an instance (government, hierarchy of an organization) to guide the actions and interactions of actors over whom it has some amount of authority” (Maroy & Dupriez, 2000). Situational regulation is concerned with the active production of the “rules of the game” (Reynaud, 1997), through processes of mutual adjustment (Lindblom, 1990) among actors, generating a complex system of multiple regulation sources and modes. A second distinction is made between normative and functional regulation. “The goal of normative regulation is to ensure commitment to the political system or to a given set of local or professional norms”. “The goal of functional regulation is to ensure efficiency and effectiveness” (KNOWandPOL, Orientation 3, regulation: specifications).
Finally, an additional distinction is made between bureaucratic regulation, based on the use of legal authority as the main source of power, and post-bureaucratic regulation based on competence as the main source of power.
Following this concept clarification, we may say that, in this study, the regulation of public policies and public action in education is conceived as a process made of a complex set of actions and interactions carried out by multiple actors, producing the coordination of the collective action in the provision of education as a public good.
The goal of the research is therefore to describe and analyze a specific feature of this complex process, that is, the use of PISA within the on-going policy-making processes in each country. PISA is defined as a knowledge regulation tool (KRT). Therefore, a clarification of this heuristic concept must be provided.
Following Salamon (2002) definition, we mean by a tool or instrument of public action “an identifiable method through which collective action is structure to address a public problem”. However, this is a quite general definition that applies to a wide range of instruments from government decrees, to professional best practice principles. Following the project’ review of the literature (Pons & van Zanten, 2007), here “we are concerned not with regulatory instruments in general but with what we called there “knowledge- based regulation tools” (KRTs), assuming that these are “not only knowledge-based but also knowledge-oriented: (they) not only draw on certain kinds of knowledge but produces certain (perhaps other) kinds of knowledge” (KNOWandPOL, Orientation 3, regulation: specifications).
Accordingly, PISA is conceived as a device for shaping the way actors think, act and interact in educational policy-making. The assumption to be tested in the research rests on the idea that, in educational policy and politics, PISA sets a regularized pattern of interactions among individuals and organizations, defining the proper issues to be addressed to, such as quality and effectiveness, and the proper perspectives to address them, such as evaluation and benchmarking. PISA not only draws on certain types of knowledge such as the ones that come from the school effectiveness or the organizational development and learning literatures, but it also produces new questions, new issues, new knowledge as it evolves through its sophisticated research apparatus and the dynamics of the educational policy process, both at the supranational level and at the national level. Defining PSA as a KRT means that it is conceived as an example of the complex and circular relationship between knowledge and policy: PISA as a policy instrument produces knowledge; PISA as a research instrument produces policy.