Informal social groups and policy programs: A Programmatic Action Framework (PAF) on policymaking

It was summer 2016 when the research group of the later funded project on “Programmatic Action in Times of Austerity” (ProAcTA) first met in the south of France. Six months later, it had worked out a project proposal that was granted by the Agence Nationale de la Recherche (ANR) and the Deutsche Forschungsgemeinschaft (DFG) under the grant numbers ANR-17-FRAL-0008–01 and DFG BA 1912/3-1. This is where the story of the Programmatic Action Framework (PAF) formally begins, although it bases on a variety of previous work on the programmatic approach. Telling this story and presenting the final project results is the core theme of this introductory contribution on “Informal Social Groups and Policy Programs: A Programmatic Action Framework (PAF) on Policymaking.”

Even before the meeting in 2016, the core research group around the programmatic approach, consisting majorly of (in alphabetical order) William Genieys, Patrick Hassenteufel, and Marc Smyrl, postulated the idea that biographical homogeneity may build the basis for cooperation between policy actors and shape public policy over a longer period of time. This idea dates back to the beginning of the 2000s and is rooted in the observation of sectoral elites emerging in France from the grounds of homogenous education and resources (Genieys, 2005; Genieys & Hassenteufel, 2001). Defending the budget and authority in their policy sector, the elite actors have been denoted as custodians of state (Genieys, 2010), struggling with their counterparts, designated as austerians (derived from austerity), over financial and regulative resources. In doing so, they coalesced around a definable policy program, which led the researchers to name the theoretical perspective “programmatic approach” (Genieys & Hassenteufel, 2012).

Building on these bases, the programmatic approach has been extended further and applied to a variety of policy sectors and countries (Darviche et al., 2013). In two more research projects by the French research group (MIRE and OPERA), the programmatic approach sheds light on elite trajectories and their influence on policymaking in France and the United States. Hence, programmatic actors have been found primarily in health policy, from within the state shaping the transformation of health care in France, Germany, Spain, and the UK (Genieys & Smyrl, 2008b; Hassenteufel et al., 2010).

Originally sticking to the idea of programmatic elites and the idea of sectoral competition between elite actors that eventually leads to policy change, Hornung and Bandelow (2020) published the first version of the “Programmatic Action Framework.” As Figure 1 shows, the framework included a variety of elite sociological considerations on the transformation of programmatic actors into programmatic elites, the methodological tools to identify these actors, and a struggle between elites as the major explanation for policy change and sectoral strength of an elite group. Thereby, it keeps the terminology of elites and “austerians” as counterparts to custodians of state, as depicting a conflict between spending ministries and the Ministry of Finance as the ministry seeking to implement austerity measures (Genieys & Smyrl, 2008b, p. 78). It did not yet include testable hypotheses on the success of elites and programs, apart from the implicit postulations that the elements that would identify collective action of a programmatic elite would also be drivers of their success the more present they are.

Although empirical research on the phenomenon of programmatic elites lives on and has been recently applied to Turkish foreign policy (Süleymanoğlu-Kürüm, 2020), the PAF has been further developed by the integration of social psychological insights of the Social Identity Approach (SIA) (Haslam, 2001; Hogg & Reid, 2006; Hornsey, 2008; Light, 2015) and the respective relevance of social group memberships and group dynamics. In this view, the programmatic groups that actors form are informal social groups whose identity influences the preferences and behaviors of actors populating the policy process (Hornung et al., 2019). While the key argument remains the same—programmatic groups form on the basis of biographical ties and use policy programs as means to career-related and ideological ends—the assumptions and hypotheses are slightly modified to further clarify the theoretical argument and distinguish PAF from existing theories of the policy process. The PAF thereby is also suitable to add to the combination of theory and empirics in public administration research, which has been criticized as lacking in contemporary research (Peters & Pierre, 2017).

A first, central modification has been carried out with regard to the inclusion of the SIA approach, as a psychological model of the individual underlying all further theorizing on actor strategies, preferences, and behavior. With the programmatic approach assuming policy change to be the result of competing programmatic elites that are shaped by their common biographical trajectories, there is a clear combination bureaucratic objective to enhance authority and careers with the biographical social identity of policy actors, or what Thorne (2017) terms professional identity.

Figure 2 visualizes this version of the PAF. It assumes that policy professionals with their respective social identities coexist. At the some point in time, some policy professionals will be bound together by their homogeneous career trajectories and experiences, although they each occupy different positions and hold different resources. Mostly, they are in a somewhat medium position in their respective organization, be it civil service, politics, or sectoral associations. When they meet and collaborate at some occasion, and when this collaboration is repeated and intensified, the actors will become programmatic actors and develop a joint policy program. At this point, the policy program takes over the role as defining element of the social identity of this programmatic group and thus shapes a programmatic identity. In this perspective of PAF, programmatic groups either pursue policy change or policy stability, depending on whether they are already dominant or still emerging. Who wins in the struggle for influential positions and authority is decided by several program-related and actor-related characteristics. These factors of success include actors’ resources, alliances, institutionalization, and the program's coherence, flexibility, and narrative.

However, this picture of PAF implicitly assumes that an emerging programmatic group always faces a dominant programmatic group with whom it competes over authority. It also does not explicitly include the term programmatic group, but only programmatic actors and policy professionals, which are defined in the original programmatic approach as those directly involved in public policymaking (Genieys & Smyrl, 2008a, p. 29) and an established term in policy process research to describe those that are employed rather than elected to positions in which they can directly influence policymaking (Svallfors, 2016). Up to then, however, PAF applications were still confusing in who exactly would or even could be members of programmatic groups. It remained open whether members would only be policy professionals and for the most part bureaucrats, or whether they would also include other policy actors (such as elected partisan professionals). The question here is, also, whether this matters to the key argument of PAF, namely that actors that are directly concerned with the policymaking process—in one way or the other—always seek to increase their authority and advance their careers and use biographical ties and programmatic content to achieve these goals. In other words, studies on ministerial careers in France also indicate that there are many ways to the top, which are less dependent on personal characteristics—apart from ambition—but personal relations that are established through various channels (Dogan, 1979, p. 1 + 16). Furthermore, the question of how institutional opportunities and constraints would further or hinder programmatic action was insufficiently addressed in this PAF version. This critique fostered an update of the PAF, which is presented by Bandelow et al. (2021) in this issue.

The contributions by Bandelow et al. (2021) and Hassenteufel and Genieys (2021) in this issue present the twin pillars of the current state of research on programmatic action, both theoretically and empirically. The first explicitly sets out the theoretical assumptions upon which the model and the hypotheses of the PAF are built. Supposing that policy actors are motivated by and seeking increased authority in a policy sector, their behavior is to follow the strategies that are likely to help them achieve this goal. Further assuming that increased authority is the result of a successfully promoted policy idea and collaboration with other actors, the notion of programmatic groups describes exactly the coming together of policy actors around policy programs to foster their careers. The policy program then becomes the guiding normative principle for programmatic actors, not out of an ideological but a strategic interest. Building on these assumptions, the PAF presents a theoretical model of policy change described in detail in the contribution by Bandelow et al. (2021) and formulates hypotheses on the formation of programmatic groups, and their personal and programmatic success (Figure 3):

Proceeding from the theoretical foundations, Hassenteufel and Genieys (2021) emphasize how to empirically assess programmatic action. They outline the methodological duality in the study of programmatic action that rests on two different but by far not incompatible approaches, that of elite sociology and policy research. Combining these methodologies leads to a unified research protocol that is intentionally standardized to allow for comparability of findings in programmatic action research across contexts. Figure 4 visualizes this dual approach to programmatic action, one elite sociological starting from positional and sociological analyses, and one rather inspired by policy research, with a focus on discourse and relational analyses in the form of networks. Both task 1 and task 2 can therefore be taken as starting points, respectively, proceeding with the other and finally ending with task 3. Even if not applied in full, parts of this research protocol may well be used to gather stand-alone evidence on the existence of programmatic action. This has been done with regard to the biographical analysis (Hassenteufel, 2012) and discourse analysis (Bandelow & Hornung, 2019).

That the methods established to research programmatic action are indeed suitable to investigate programmatic action also in other parts of the world, beyond Europe and “Western” democracies, is demonstrated by two more contributions in this issue. These follow partly the elite sociological approach to programmatic action and partly that of policy research. Duque (2021) applies the former perspective to higher education policies in Chile and Colombia. Comparing quality assurance policies, he poses the question why despite similar policy instruments used in the two countries, the approaches to assure quality in higher education differed substantially after the adoption of the policy program. An integration of institutional perspectives of layering and path dependence corroborates the finding extracted also from interviews that while one programmatic group was able to institutionalize its policies, the other suffered from instrumental layering that weakened the original ideas of its policy program.

Studying health policy in Brazil, Davidian (2021) equally applies the PAF in an elite sociological way to the policies of social protection. Brazilian health policy reforms have been found to be essentially shaped by the sanitaristas, a group of actors from medical professions and healthcare professionals. Tracing the influence of this group as a programmatic group, she shows how the programmatic actors under the authoritarian regime first needed to stay at the universities, and they managed to occupy central positions in the context of institutional transformation and decentralization. Thereby, the contribution outlines how the PAF may enhance the explanations of stability in times of institutional change.

Enhancing the original field of application of French defense policy, Faure (2020) analyzes French warfare policy against the backdrop of the PAF. Two competing groups whom Faure denotes as custodians and austerians, respectively, advocated the import of external or further use of national fighter jets. The victory of the sovereigntist public action program can be explained by closer interpersonal ties at the level of decision making and a larger termed programmatic clique that also involves industrial actors. Thereby, the contribution is strongly inspired by the original view on programmatic groups as competing elite groups that are struggling over authority in a sector.

Adding more explicitly to the institutions of programmatic action and policy research, Hornung (2021) investigates the institutional characteristics in France and Germany that have led to the success or failure of policy programs beyond the financial crisis. She finds that a crucial explanation for long-term policy program stability in the stability of institutions themselves. However, the durability of policy programs, at the top of their cyclical pattern (Bandelow & Hornung, 2020), seems to also depend on the extent to which the policy program fits the existing institutional structures of a policy sector. If it contradicts the institutions, chances are high that the policy program will only insufficiently be institutionalized, which will eventually seal its end.

The published contributions on the PAF in this issue show numerous tying knots for further studies. One concerns the extension of PAF analyses to other policy sectors, which Duque (2021) exemplifies by his application to higher education policy. Another concerns the application to other political systems. Here, also, Duque (2021) and Davidian (2021) have made the case for a fruitful transferability of the PAF to authoritarian regimes like Chile, Colombia, and Brazil. These contributions turn attention to the institutions relevant to programmatic action, something which Hornung (2021) also looks further into. Future research should therefore place more emphasis on how political and sectoral institutions impact programmatic action. This also concerns the type of programmatic groups that one is likely to find in different systems, as Faure (2020) shows for warfare policy.

Apart from these research agendas concretely addressed by the contributions in this issue, the core concepts of the PAF are open to further refinement. Especially the notion of programmatic groups as social groups in the understanding of social psychology would profit from a further cross-fertilization of policy process research and psychological research. This encompasses among others the following research questions: In what way do group dynamics impact programmatic action and the change of programmatic groups and policy programs? How do social group memberships and programmatic group membership in particular affect policy actors and individual preferences and behavior? To what extent do programmatic group characteristics play out in inter-group relations? How can the concept of programmatic action be combined with other theoretical perspectives, such as policy instruments (Brunn, 2020)?

What should also be addressed and reflected on in further studies is the relation between programmatic action and the public. Partly, this has been done with regard to discourse analyses through which programmatic actors can be identified via shared discourse elements in the media. But programmatic actors may also instrumentalize and use the public to push for their policy program and to provide a favorable environment to foster programmatic action.