The relationship between knowledge and policy is variable throughout the public action process and depends, among other factors, on the academic knowledge available, the predominant modes of regulation, the characteristics of the actors involved and the object of the policies implemented.
In the case of school autonomy and management, the administration exercises a strong influence through its centralization and bureaucratic regulation. This explains the major role played by the political agenda of the Ministers of Education in the parliamentary debate, in the legislative process, in the changes made to the various decrees analyzed, and in the academic production as well as in the very development of the Educational Administration academic field.
It is worth stressing that the influence of the “mode of regulation” in the policy process and in the mobilization of the knowledge  is not exclusive; rather it is intertwined with other equally important influences. According to the mi-nisters we interviewed, the results of academic efforts emanating from the educational sciences (in the area of educational administration) are not “suitable” for resolving what they believe are problems. Therefore, after an initial phase in which the drawing up of the legislation was preceded by expert and academic studies (albeit with the collaboration of administration officials), the Ministry of Education has recently attempted to create a “technostruture” to produce its own “state knowledge”, compatible with its own “needs” (see Barroso et al, 2008). This knowledge is based, above all, on “evaluative research” and on acknowledgement of “good practices” (Lessard, 2008, pp. 562-563) and entails, very often, different areas of the Social Sciences (statistics, business management, social psychology, etc). In this process we are not simply confronted with a “transfer of knowledge”. Here we are witness to an application of trans-disciplinary knowledge, or as Armin Nassehi put it (2008): “We have to become aware of a transformation process rather than a transportation process”.
Besides, in another arena of public action (the schools), we see the emergence of new actors (and “new” knowledge), resulting from the growing number of school managers who have taken master’s courses in this area. Here practical knowledge is “theorized” and the practitioner’s knowledge approaches that of the author as a consequence of the heavy research component required in these courses (at least up until the “Bologna reform”). This is even more evident when the object of public action is school organization as a management unit. To the extent that such knowledge depends heavily on “or-ganizational learning”, the knowledge of actors is of critical importance. This trend is reinforced by continued “bureaucratic-professional regulation” in Portuguese schools despite recent changes which translates into the continued relevance of teacher “expertise” in school organization and management, as well as the power and knowledge of “professionals”.
Indeed, the confrontation between these several kinds of knowledge (“bureaucratic”, “new public management” and “pedagogical-professional”) constitutes one of the distinguishing features of this public action.
• The circulation of knowledge is the result of the circulation of people
This is evidenced in school teachers and school managers who become MPs, in MPs who become members of ministerial offices or Secretaries of State, or in academics who integrate working groups created to support policy decision-making. The same is true of union leaders and civil servants who take post-graduate courses at Universities with professors who integrate working groups to help policy decision or who perform studies for the Ministry of Education. This network includes academics who write thematic articles in the press, university professors who become ministers, union leaders who become professors and vice-versa. Secretaries of State often join university faculties and conduct studies aimed to guide policy decision-making. Some ministers end up chairing the National Education Council while former chairs enter the ministry. The circulation of knowledge also includes experts. These experts participate in working groups or develop studies for the Ministry of Education and are often invited to make speeches at union sponsored events. Crosspollination is also expressed when headmasters take up management positions in the Ministry of Education and senior administration managers acquire positions as university professors, training future headmasters. In this context, knowledge is externalized, above all through the discourse of actors and the constellations of knowledge determined by the structures of the social networks that result from the circulation of actors in different arenas of public action.
• The use of knowledge in the political decision-making process is heavily influenced by the interaction (competition) among the different interest groups
The dynamics of the legislative process are influenced, in general, by “advocacy coalitions”  that share a set of long held ideas and beliefs about education and the governing of schools in particular. These “advocacy coalitions” mobilize knowledge according to their shared interests as well as to its efficacy at manipulating the debate and influencing various actors in Parliament, the Ministry of Education, schools, the Press, unions, universities among others. In this study it was possible to identify various “advocacy coalitions” through the convergence of arguments and actions connected with certain beliefs and ideas about the purposes of autonomy and the methods of bringing it about. A recent “advocacy coalition” formed around the newspaper, Público in particular through the editorials written by the editor, opinion articles, and letters from the readers. The major parties that made up this coalition included an association called “Fórum Liberdade de Educação” set up by the CDS (right-wing parliamentary party), a group of bloggers, and business associations, among other kinds of actors. These groups debated such issues as the defense of a family’s “freedom of choice” regarding schools, “vouchers”, “competition among schools” within education marketplaces, “effective leaderships”, “total quality”, etc. Although these policies are typical of conservative and neo-liberal movements, they are cham-pioned not because of the merits of their ideological standpoint; it is because of apparent reasoning and technical quality of their arguments. Their knowledge is mobilized on the basis of “good practices”, the “foreign example”, the “scientific evidence”, the “business management”, etc. It is destined to persuade teachers, the families of pupils, members of the administration, and public opinion in general about the existence of “a problem” for which they “have a solution”.
A second, older, “advocacy coalition” formed around one of the main teachers’ unions (FENPROF), “progressive” pedagogical militants, some academics (from various institutions of higher education), former “members of management boards”, and left-wing parliamentary parties. They argue for “democratic man-agement”, “pedagogical priority regarding the administrative questions”, “teacher participation”, and “collegial departments”, among other positions. Their knowledge is mobilized on the basis of the “school experience”, “critical thinking” (by some academics) and of the results of the research that are “congruent” with the policies advocated.
In addition to these two “advocacy coalitions” other “less ideological” as well as less well defined and volatile ones have formed throughout the process, organized to support or criticize certain legislative decrees (as was the case with “Decree-Law 172/91” and “Decree-Law 75/2008”). These groups included university professors, other experts with post-graduate training and school managers. The knowledge mobilized here issues, in the main, from academic works or the discourses and practices of school managers. At times expedient “alliances” are formed (or other less formal connections) among members of these groups and the two “advocacy coalitions” mentioned above. These connections explain the migration of a lot of “concepts” from one scene of public action to another and from one group of actors to another. This is the case, for example, of the concepts of “autonomy”, “leadership”, “evaluation”, “quality”, “participation”, “project” that are constantly fine-tuned in order to be compatible with the ideas promoted in each situation.
• The relationship between knowledge and policy can happen either to construct problems or to define solutions
In the case of the public action under analysis, “the reinforcement of school autonomy” was presented as a solution by the Ministry of Education, even before “its absence” was perceived as a problem by the schools. This shows two distinct but complementary logics that coexist in this process to establish the transfer between the scientific field and the politic field. On the one hand, there was a logic centered on the search for solutions, which translated into “knowledge-based policy” (or “knowledge-informed policy”). This logic was accepted by interviewed ministers and in the creation of the working groups, and consisted of first gathering information and then deciding afterwards. On the other hand, there is a logic that started as a set of ‘a priori’ solutions incorporated into the political “agenda” where one was expected to “find” appropriate problems. In this case, as suggested by Kingdon (1995), “it is the policies that are in search of the problems”, meaning that the intended use of this applied knowledge was, above all, to justify the need to take certain deci-sions.
These two logics go a long way to explaining the “supply and demand” game in the processes of transfer between knowledge and policy (and vice versa), that is apparent throughout this public action, both in the constitution of the working groups, and in the use of the available academic knowledge. On the one hand, the demands made by ministers of education of the working groups are based on a prior “policy goals” (“reinforcing autonomy”, “creation of strong leaderships”, “quality evaluation”, etc) and the expectation is that the working groups legitimize and bring about the achievement of these goals. On the other hand, commissioned university studies (or leveraging research already available in the academic world) are prior conditions for the congruence between the research results and the policy orientation. In another words, the use of knowledge in policy determines the political conditions of knowledge production.
In this context, as mentioned earlier, the influence of knowledge (especially the available academic knowledge) in the public action under study was conditioned by three major factors: feasibility (operational knowledge that could be easily translated into action was preferred); convergence (knowledge that is related with pre-existing beliefs is preferred); agenda (knowledge that fitted into “agenda setting”). Or, as Delvaux, building on Kingdon, puts it (Delvaux et Mangez, 2008, pp. 79-80): certain kinds of knowledge (the kinds incorporated into beliefs or put into language) tend to be mobilized to either gauge or get through the test of feasibility; the test of acceptability; and the test of relevance.
 For elaboration on this point see Delvaux (Delvaux et Mangez, 2008, pp. 46-52).
 Accepting the idea proposed by Sabatier and Jenkins-Smith (1993, p. 5): “An advocacy coalition consists of actors from a variety of public and private institutions at all levels of government who share a set of basic beliefs (policy goals plus causal and other perceptions) and who seek to manipulate the rules, budget, and personnel of governmental institutions in order to achieve these goals over time”.