The Institutions of Programmatic Action

Abstract

A comparison of French and German health policy between 1990 and 2020 reveals the institutions under which programmatic action occurs. Expert interviews allow for an in-depth analysis of which institutions and processes are necessary for programmatic action to take place. In presenting the results of these expert interviews, this chapter shows how different systems of elite recruitment and policy advice are relevant to programmatic action by at the same time stressing that decentralized and corporatist structures are less directly related to programmatic action.

What can we learn from these results of programmatic action analyses for the institutional conditions under which programmatic action generally occurs? To assess this conclusively in the case of French and German health policy, it is important to keep in mind that the theoretical perspective is born from the observation of French policy-making and is therefore initially strongly informed by the institutions of the French system. However, it is necessary to generalize the influence of institutions on the occurrence of programmatic action in order to enable traveling capacity of the PAF to other political systems. The institutions of policy advice and education turned out necessary in the interviews both with French and German experts. To this end, it is worth abstracting from specific national institutions. This is done in the following.

7.1 Institutionalized Elite Recruitment and Policy Advice—Par Excellence

Programmatic action in France is strongly influenced by the ENA, in line with the history of the PAF. The homogeneity of career paths and the biographical interfaces of programmatic actors are much more common in France. This is because career paths are more hierarchically organized and rigid than in other countries, including Germany. At this point, the elite formation system emerges as a key determinant of the formation of programmatic groups. In the health sector, moreover, the role of commissions, which are often established in advance of reforms, is confirmed in a special way. These expert commissions enable different but central actors in the policy process to participate in the elaboration of reforms and, in this way, to develop policy programs. The firmly institutionalized structures provide a quasi-permanent opportunity for programmatic groups to form. Here, a special role of the institution IGAS becomes apparent. As an institutionalized body of reflection, it is one of the first points of contact for the best ENA graduates. Not only do IGAS actors benefit from the institution’s group-building dynamics, but they can also be appointed to other positions at any time from their IGAS position and then move back. This in itself allows for inter-institutional exchange and facilitates contact between actors who can form programmatic groups. At the same time, political institutions have hardly been subject to change since the Fifth Republic. Path dependency thus makes the formation of programmatic groups permanently possible.

Programmatic action in France is very executive-heavy and centralized because of the majoritarian democratic structures and few veto players. Programmatic groups are small and homogeneous, not only because of the elite formation system, but also because of the few points of contact between actors in the sector and at the subnational level. However, the empirical analysis here does not show that strong federalism and low corporatism must be obstacles to programmatic action. They may also enable programmatic action, because they involve fewer consensus constraints and allow programmatic actors to coordinate on a smaller scale. However, it can be concluded that the substance of the reform program makes less profound change possible as a result because fewer actors are involved.

With regard to the success of programmatic groups, which was the second focus of this study, the French political system offers fundamentally better opportunities for the long-term success of programmatic groups. Here again, the considerations on institutional influences on emerging and existing groups come into play. In France, hardly any distinction can be made between emerging and existing groups during the period under study, as they are recruited through established career paths and take the places of previous actors. The program thus becomes more institutionalized and emerging groups are the successors of the existing group, while the network remains the same. The political system, with its low degree of corporatism and federalism and few veto players, plays into the hands of programmatic groups.

Beginning with the selection of interviewees, 35 interview requests were sent to the actors identified in the chapters above. The selection was again guided by the formal positions these actors held over at least five years. Eleven interviews were conducted between November 2018 and May 2019. To preserve the anonymity granted, the names of the interviewees are not explicitly mentioned, but provided with an ID and the institution where they took place. Figure 7.1 displays the interview IDs and the respective institutions. Access to the transcripts of the interviews can be found in the appendix.

Fig. 7.1

Interview partners—French health policy. Source: Own illustration

On the one hand, the interviews are suitable for cross-checking the previous findings about a policy program and a programmatic group that existed in French health policy between 1990 and 2010, and continued to have influence afterward. In terms of both the content and the actors of programmatic action, several interviewees confirm the interconnectedness. For example, a social adviser from this period in the prime minister’s office notes:

F2: The reforms that have been carried out between 2007 and 2011 are really a continuation of 1995 and 2004. For me, they add, they change, but this is not a questioning of 1995 and 2007. Moreover, I don’t have the feeling […] that the Touraine reform finally really called into question this philosophy of the 1995–2004 years. (Interview F2, 2019)

This is cross-validated by other interview statements.

F2: It seems to me that the two major structural reforms are the Juppé 1995/1996, and it is the 2004 law. So, we stay with that logic. The HPST law is a change, but it does not call into question the policies followed since 1995 and 2004. (Interview F2, 2019)

Both statements reaffirm the great importance of the three reforms, the Plan Juppé in 1995, the Douste-Blazy reform in 2004, and the HPST in 2008. Beyond the reforms themselves, the interviews also reaffirmed the statements made as a result of the biographical analysis regarding the key role played by that several actors in the elaboration of these reforms. Specifically, the names that were dropped by the interviewees and that also appeared in the previous analysis of formal positions were Frédéric Van Roekeghem and Dominique Libault. As one of the project partners once stated: “[Libault] wrote > 99% of the Douste-Blazy law with his collaborator at the DSS” (Hassenteufel, private conversation). The fact that these actors, who biographically held key positions, were also decisive in the reforms that were passed justifies the positional approach, which states that actors in key positions play a dominant role in promoting policy programs. Moreover, the close collaboration between these actors leads them to develop a social identity based on their professional careers. Thus, the starting point for cooperation seems to be anchored in career intersections, which can develop into a programmatic identity if it is linked to a concretely identified reform program that is gradually adopted and implemented. This provides evidence for the biographical identity hypothesis.

F8: There was a trio on the implementation of the 2004 health insurance reform: Fréderic Van Roekeghem, Thomas and Jean-Marc Aubert. But on the HPST law, it was more with Olivier de Cadeville. It’s on the institutional side, Dominique Libault. (Interview F8, 2019)

F5: Dominique Libault is the man in the organization of the ministry. It’s someone who makes his whole career- (I1: Yes, in the DS.) in the DSS. (Interview F5, 2019)

F6: [Libault] tried very hard to launch reforms around both the universality of social protection and its sustainability over time, in particular with the CSG, the financing law, the CMU, and finally tried to rethink social protection in a universal dimension, but by maintaining, as [he] always sa[id], in the French model, an identity of Social Security in relation to the State. (Interview F6, 2019)

And yet, many interviewees point to Raymond Soubie as an important individual in the genesis of the reforms, especially the 2008 reform. Frédéric Van Roekeghem, “who knew music by heart, who was in Alain Juppé’s cabinet in ninety-five, who was director of Mattei’s cabinet in two thousand and four and who became general manager of the CNAM” (Interview F2, 2019), is associated with one of the major health care reforms, especially the 2004 reform, but also with the other two reforms. In particular, the influence on health policies that he exerted together with Raymond Soubie, whom he has known since that time, is also visible when other health policy actors are asked about this issue (see also Chastand, 2012):

F10: The Van Roekeghem-Soubie axis, in power relations, was nevertheless two goldsmiths. They are people who greatly influenced the policies that were made, there is no doubt about it. (Interview F10, 2018)

F6: In fact in the immediate entourage then of Nicolas Sarkozy, President of the Republic, whether on health insurance or pensions, the one who really counted was Raymond Soubie. (Interview F6, 2019)

The interview with a former staff member of Frédéric van Roekeghem also mentions other actors who were involved in the reform process. They state that collaboration with the finance ministry (Bercy), the health ministry (Matignon), and the presidency (Elysée) in preparing the reform was intense. They also explicitly name Éric Aubry as one of the key players in the health ministry.

F8: But very frequent with Bercy who was one of our allies anyway. [...] we were also in frequent interaction with Matignon and the Elysée. Matignon it must have been Eric AUBRY, at the Elysée, [...] Marguerite BERARD [...] Julien SAMSON. (Interview F8, 2019)

Biographical connections become even more apparent here. While studying at the ENA, Éric Aubry was part of the 1980–1982 cohort that also included Anne-Marie Brocas, who participated in the Soubie Commission and later took a position in the DSS when the Plan Juppé was being prepared. This ensured that the policy program, once elaborated, lived on in the collaboration between the actors who formed a programmatic group around it and also placed several of them in key positions that were first established through this program.

With regard to the power resources that programmatic groups use for their success, the interviews reveal a particular relevance of certain institutions and networks. The results thus also confirm the relevance of the institutions identified as relevant at the outset, but focus even more narrowly on the CNAM and the DSS. One of the long-time members on the CNAM leadership assigns considerable power to the DSS in drafting legislation and assesses the CNAM as dependent on or at least subordinate to the control of the health ministry. It is often instrumentalized as a buffer against the interests of unions and health professions representative bodies, they said.

F8: When you are in the DSS, if there is a change in the regulations, you design the measure, write the text and monitor its financial impacts. […] So, I would say that the CNAM’s position was part of a somewhat complicated institutional and inter-ministerial game, but not at all beyond the control of the state. I would say that the role of the CNAM was rather to be in the front line, in a way also to protect the state on a number of subjects, particularly in negotiations with health professionals, which is a subject that can be quite costly. (Interview F8, 2019)

The expectations that can also be confirmed is that a programmatic group uses the institutionalization of its policy program through institutions as a guarantee for the stability and survival of its ideas. Strengthening certain institutions through the policy program also allows them to place their members in the positions created by those policies. One interviewee in the DSS clearly states that the reforms have helped to strengthen the very department in which they were drafted:

F6: This whole ’95–2007, 2008 phase is a very strong affirmation phase of the DSS. Well, I have to say—beyond the fact that I am a director in [*anonymized*]—I also play in this role. Now, quite honestly, I have a certain strength of conviction and this is a time when the DSS is very strong. (Interview F6, 2019)

This interpretation is also cross-validated by another interviewee:

F7: In fact it allowed the Social Security Directorate, in a way, to become very autonomous from the decision in the health sector. (Interview F7, 2019)

The interviews confirm, first, the results of the previous document analyses, discourse network analyses, and biographical analyses that the three major health policy reforms of 1995, 2004, and 2008 were interrelated both regarding the content and the actors surrounding the reforms. Moreover, the reforms can be consistently traced to multiple reports and working commissions in which the same actors prepared and subsequently pursued the jointly developed proposals that compose the policy program. Because the interviews highlighted the specific role of IGAS and DSS in the recruitment of career actors, they provide evidence for the biographical identity hypothesis and the science policy hypothesis. Consequently, these two institutions prove to be central to the careers of programmatic actors:

F8: I would say the world of social and the world of social budgets, it’s a small world. So, most of the profiles went through IGAS or the DSS. So, the positions are held by people who have the same culture, the same profiles. (Interview F8, 2019)

The career paths of policy actors, the institutions through which they pass, appear as a predisposition for the emergence of programmatic actors. In particular, some identified positions develop as turning points in the careers of individuals and as a central point in the education of individuals, which is the basis for the formation of programmatic groups.

F5: Well, the IGAS is a caricature of what the French technocratic system is like. They are people who do the ENA, they come out well ranked at the end of the ENA, and they will have the opportunity to have a golden parachute because they do a few years as inspectors. […] They do this job for a few years, and then they take positions of responsibility in the Administration or in the operators who are around the Administration. And they come back—so it’s a great thing because it means you can come in and out whenever you want. (Interview F5, 2019)

In addition to the homogeneity of careers, which emerges here as a central factor for collaboration and ongoing careers, there is a need to bring these actors together for intellectual reflection and development of strategies. For such purposes, it is necessary to create bodies that institutionalize such efforts. Newly created institutions with many types of actors involved are one way to achieve this goal. Or, as one long-time DSS member states:

F6: The idea was to find bodies in which there is dialogue, consultation, information, cold reflection, and so these high councils, we started with the HCAAM in two thousand and four, or rather the COR first of all, excuse me, the COR, the HCAAM, the family council, the HCFiPS and I think it plays a useful role. I was also talking about the fact that administrations are often taken by urgency, that there is nevertheless a lack of strategic thinking that I find a little strong today in the system and the high councils can make it possible to bring a little strategic thinking with resources that are quite limited in the end, to propose medium-term visions. For example, the HCAAM, the recent HCAAM report on the health system, nevertheless inspired the health law quite a bit. (Interview F6, 2019)

Besides, the question of where content of policy programs comes from is a key element in the analysis of programmatic action. In the previous chapter, the occurrence of programmatic action revealed the importance of working commissions in bringing together actors who later use the policy program to place themselves in the key power positions in which they implement that program. Key institutions where policy proposals are developed include IGAS and the directorates in the health ministry, DSS, (Direction de la Recherche, des Études, de l'Évaluation et des Statistiques (Research, Studies, Evaluation and Statistics Branch); DREES), and DGOS.

F5: There are many prospective reports because in fact it is much more used as a suggestion box […] the Administration has been much strengthened in the study department since we created a studies department, the DREES, in the years two thousand, ninety-eight or ninety-nine, so there is a studies department at the Ministry of Social Affairs that also does evaluation, […] and then there is a national health insurance fund that has study services. […] in France the production of expertise and evaluation of public policy is done—(I1: At IGAS.) by the administration, that’s it.—(I2: Internally.)—That’s it.

F5: What must also be said is that IGAS gives resources, expertise to those who leave it, which means that we are then very well equipped for political positions or responsibilities—

F5: It gives you a general culture—that’s it. […] You know everyone because you spend your life meeting people in the field. (Interview F5, 2019)

In addition to the well-known and established institutions, the creation of new institutions, including the HCAAM, during the implementation of the policy program also served the purpose of creating a new body for the generation of policy proposals.

F9: But in 2004, HCAAM was indeed created with the idea of creating a consultation process that allows proposals to emerge. (Interview F9, 2018)

Besides the formally installed institutions, temporary working commissions, explicitly established for the preparation of health care reforms and composed of several actors who are repeatedly part of these commissions, play a central role in the continuous promotion and implementation of a policy program. Such working commissions, of which the Soubie Commission can be considered the first, guaranteed the continuous participation of programmatic actors in the decision-making process and allowed them to translate their ideas into concrete policies.

F8: There was an actor at the time too, it was Gérard Larcher. Because Sarkozy had entrusted him with a mission for which I was the rapporteur. So I wrote Gérard Larcher’s report after I left for the CNAM and Larcher even wanted to reunite the commission after when Roselyne Bachelot had drafted her bill. (Interview F8, 2019)

To conclude the French chapter at this point: The different methodological approaches, from the analysis of legal documents, media and public discourses, actor biographies, and finally qualitative interviews, cross-validate the findings on a programmatic group between 1990 and 2020. The reforms adopted during this period can clearly be traced back to a policy program and programmatic actors linked by biographical career trajectories and shared policy ideas. The institutions of programmatic action are located, at least in France, in central institutions of the education system, whose graduates have increasingly oriented themselves toward the top positions in health policy. Certainly, the French education system and the recruitment process of administrative and political staff through the ENA facilitate homogeneity of career trajectories and increase the likelihood that actors will eventually meet and collaborate. Such an education system is unique in comparative politics, but it explains how institutions can form the elites that later occupy key positions in the policy process (Hassenteufel & Le Galès, 2018, pp. 296-297). In health policy, the social security system has become a more prominent career path for these actors since the 1980s, as evidenced by the increasing number of actors occupying positions created in the health care system (Genieys, 2005).

In addition to the education system, the strong incorporation of advice and scientific thinking in the processes of decision-making can be observed in France. The bringing together of actors in bodies of intellectual reflection to develop reform proposals are ideal conditions for the formation of programmatic groups. This occurs primarily through the ministerial cabinets, which involve numerous advisors who can act as a group, but also through the IGAS, which symbolizes an institutionalization of the scientific advice fed into the policy process. Moreover, the institutions of ENA, IGAS, and the advisory and decision-making positions in the ministry are closely intertwined, and actors frequently move between these positions. In the study of French public administration, scholars argue that the resulting professional bureaucracies are characterized by great expertise and networking and a consequential influence on policy-making (Bezes, 2016, p. 260). It remains to be evaluated, against the backdrop of the German case, to what extent these institutions can be generalized independently of the institutions setting and the political system.

With regard to the long-term success of the programmatic group in France, one can conclude that the programmatic actors succeeded in networking in such a way that it was possible to consistently fill positions, even against the backdrop of political changes. The bureaucratic network is thus stable and independent of macropolitical changes, maintaining itself through the very institutions that led to its formation. Indeed, this is due in large part to the educational system that predetermines these career paths. However, the programmatic group has also managed to establish institutions and new agencies (HCAAM, HAS) from which its members continue to benefit because they give them influence in policy processes and create positions that strengthen the network as a whole. Here, too, it remains to be evaluated for the German case whether such mechanisms exist analogously in other systems.

7.2 Federalism, Corporatism, and Institutional Change in Germany

In contrast to France, programmatic action in German health policy did not persist. This was due to institutional changes, but also due to particular institutional settings and lacking institutions of elite building and policy advice as the following paragraphs will show.

Federalism reforms have weakened the requirement for consent and thus the need to involve subnational actors in health policy decisions (Zohlnhöfer, 2009, p. 58). In German health policy, only reforms that affect the organization of hospital policy require approval by the Bundesrat (Bandelow et al., 2020). Even in the early stages of programmatic action, Rudolf Dreßler announced in 1996 that he would appeal to the mediation committee for the reforms planned by the black-yellow coalition, threatening to instrumentalize federalist structures. Nevertheless, the subnational level is not to be neglected, as it most recently represented a countervailing power again in the discussion about the supervision of health insurance funds. The decentralized structures in Germany therefore enable and hinder programmatic action in equal measure, as they promote group formation but tend to make success more difficult.

The corporatist bodies were also changed in the course of programmatic action. Institutions originally conceived as “bargaining corporatism” became “competitive corporatism” (Rhodes, 2001, p. 177), changing the traditional role of corporatism. Instead of negotiating social pacts, corporatist actors now compete more with each other and make decisions in the shadow of hierarchy, with the state playing an increasing role in negotiations. Parapublic institutions, described as another “node” besides parties and federalism (Katzenstein, 1987, pp. 4, 35), have been partially deprived of their role. Because of these nodes of parties, federalism, and parapublic institutions, the German political system offers several points of contact for programmatic actors. However, the formation of programmatic groups does not take place within these nodes, but rather when the nodes spill over and come together in new institutional forums.

Moreover, the increasing importance of coalition negotiations for German politics also has an impact on policy processes. For example, most reforms adopted in Germany in the past were strongly influenced by the coalition agreement. In contrast to the expert commissions that were very common in the past, the access of non-political actors to the elaboration of policy ideas is currently often realized through coalition negotiations—if they manage to enter them.

Findings from the previous analyses suggest that programmatic action occurred during the first period under study (1990–2010), but not in the second period under study (2011–2020), unlike in France. A qualitative analysis through interviews is intended to confirm these findings for the German case and shed more light on the resources, strategies, and power of the programmatic group. To this end, a total of 37 interview inquiries were sent out and 20 Interviews were conducted between May 2018 and January 2020. Not all of these were conducted in the formally relevant institutions as shown in Fig. 7.2; some served as key informant interviews and some served to understand the underlying structure of the health care system. Figure 7.2 visualizes the interview IDs according to the institution to which the interviewee belongs. The remaining six interviews cannot clearly be assigned to one of the institutions. All are available from the appendix.

Fig. 7.2

Interview partners—German health policy. Source: Own illustration. The figure only shows the 14 interviews that have been conducted in key institutions of the health care system. The other six interview partners not included in this figure were experts from within the system that are, however, not clearly assigned to a particular institution

Federalism and corporatism were crucial to the success of the programmatic group. The role of the subnational states in this phase of programmatic action is relevant in that the arena of the Bundesrat and state coordination was actively used to introduce the ideas of the Enquete Commission into the political process. The fact that the programmatic group occupied central positions in the states enabled it to coordinate as a group at that level as well, and to use the resources and decision-making processes at state level directly:

G6: And then we built up, as it were, a counter position, a reform perspective for the SHI system. And we fed that into the A-country process. And of course we made massive use of the findings of the Enquete. (Interview G6, 2019)

Corporatism has an ambivalent role. On the one hand, the major reforms in the 1990s were deliberately adopted without the participation of stakeholders and the powerful actors from sickness funds and physicians’ associations in order to directly overcome the expected resistance. On the other hand, the positive list, which was adopted and then not implemented, showed that the success of reforms adopted by political actors in retrospect depends on their implementation by self-governance. The interviews suggest that implementation is one of the biggest problems facing self-governance when it comes to health care reform. As a consequence, the analysis shows that in corporatist settings it is necessary for programmatic groups to involve corporatist actors who ensure broad support for the policy program. In doing so, they fulfill the original function of corporatist actors to communicate decisions to their members.

G1: And we then had a quasi-supervisory discussion with the GKV-SV and said: “So if you don’t do it yourself, then we will have to do it ourselves in the near future or change the law”. Then the other party said: “You know, you can change the law as often as you want. If we don’t want to do it, we won’t do it”. (Interview G1, 2020)

At the same time, however, the German health care system is so centrally and closely organized that the respective actors, in constant exchange with each other, are quite capable of jointly implementing reforms with a common goal, as long as the corresponding interests are largely safeguarded. However, these exchanges are becoming increasingly informal and rarely allow for far-reaching reforms. Instead, these encounters can be described as everyday business.

G3: Let me say that we are dealing with a centrally managed health care system. So it’s like the GDR (German Democratic Republic): small, healthy, yes? And that simply means that the staff of the ministry and the staff of the self-government know each other. [...] So in the informal sector, it is incredibly tight. (Interview G3, 2020)

Bringing together the findings from the empirical analyses of French and German programmatic action in health policy thus makes it possible to show which institutions are similar in the two otherwise highly different countries and could thus be generalizable as institutional conditions that favor or hinder programmatic action. However, in order to conclusively assess which institutions enable or block programmatic action, it is also necessary to provide a case in which the PAF is refuted and programmatic action did not occur. To confirm the institutional argument, one should then find that the institutions relevant to the occurrence of programmatic action are not present in the case under study. Therefore, the following subchapter uses expert interviews to ask about the reasons for the absence of programmatic action in German health policy since 2011.

The absence of programmatic action is also perceived by the key actors in the health care system and attributed to various factors. As much as the country level was used in the phase of programmatic action between 1990 and 2010 to feed policy proposals and the ideas of the Enquete Commission into the policy process, the states can act as a barrier. In the current phase, for example, there is a central need for reform with regard to cross-sectoral care and the question of how many hospitals are needed in which functions and at which locations. At this point, then, it becomes clear that the states can serve as both drivers and barriers to programmatic action, depending on whether programmatic actors are also directly at home in the respective structures:

G3: Unfortunately, I have to say that this is very much due to the steering of the Bund-Länder working group. If we don’t have the solutions that we have, I don’t think we can get them in the first few months. There’s a real gap because you can only do this with the states, because we have to get to the hospitals and that’s why you can’t do it alone, so there has to be an agreement with the states. (Interview G3, 2020)

G4: Many things really did work better at state level in the past. We were better organized, […]. The Conference of Health Ministers simply signed off, in my view. [...] And this is, in other words, at their expense, although on the other hand, the real problems we have cannot be solved without the participation of the states. (Interview G4, 2019)

Stakeholders also report a deterioration in the relationship between the ministry and the self-governance. This is mainly due to the fact that, in the eyes of the current health minister, self-governance no longer fulfills essential tasks or does so inadequately. Obvious examples are the lack of progress in the area of digitalization, which is criticized, but also the mutual blockades and hostility on essential questions of reimbursement and service provision.

I2: “Has the relationship between the BMG and self-governance changed in recent years?”—G1: “Yes, yes, deteriorated. That is the question of how to look at it. You can also say that it has improved. (Laughs.) No, the relationship is. So Spahn’s statement in the committee: “I am in favor of self-government if it works.”—I1: “So in this respect it [the state] has gained more control over self-government now the leadership than before?” —G1: “Yes, of course, it allows less and intervenes more often. The house finds this a bit ambivalent, of course, because it means more work. But in principle, we all think it’s good not to be fooled. I think that is already the prevailing opinion.” (Interview G1, 2020)

A particular role is played here above all by the distinction between the scientific, factual, and subject-related level on the one hand and the party-political level, which is also partly state-controlled and oriented to the interests of self-governance, on the other. Several interviewees point out that the scientific level is responsible for the impulses, but these must then be renegotiated at the interest level. This is perhaps the biggest difference from programmatic action in France, which has to contend with fewer federal and sectoral interests due to its leaner institutional structure. These institutions were overcome by the old programmatic group.

G3: So there is always the phase where a high level of expertise is involved. Where one really tries to name and solve the problems and discuss possible solutions. And then there is always a second phase where the interests of the countries play a major role, which are very different. In other words, in terms of size, city-state countries, and so on. Quite a few things. Hospital structures that are completely different in Bavaria than in Saxony or something like that. And even if there is the CSU in Bavaria and the CDU in Saxony, there must be some kind of agreement between the two. And then the A- and B-states sit among themselves and try to agree on a line based on the professionalism of the work and then there is a question of colors and political exchange. And then the professionalism is moved back to the second row. Because of the compromises that have already been found. (Interview G3, 2020)

As important as scientific input and the involvement of scientific actors in reform processes is for the formation of programmatic groups and programs, it is dangerous if these scientific actors subsequently outlive their positions. It is true that it is of great importance for a programmatic group to create positions for its members. However, if these positions are not exchanged and become institutionalized over time, this is both an indicator of a programmatic group’s success and a barrier to new programmatic actors seeking access to the system. In this respect, programmatic action thrives on ever new scientific impulses that must be constantly renewed and exchanged, just as science thrives on doubt.

G3: And there in the first place Wasem and Rothgang, who really did it to perfection. What I cannot blame them for. I begrudge them every cent they have. The problem is that they are becoming more systemic, have become systemic. Because the large number of expert opinions means that they have a pool of information that is no longer available to anyone else who is to work in the same field. And I think that is terrible from a scientific theory point of view. So that is... That is not possible. (Interview G3, 2020)

The current institutional conditions in Germany also make programmatic action difficult, as the institutional circumstances have changed considerably. The increased importance of coalition negotiations plays a central role in this context. The coalition agreements of 2013 and 2017/2018 are each characterized by a level of detail that offered the future health ministers only little room for maneuver. Hermann Gröhe, for example, worked out the measures set out in the coalition agreement in great detail. Jens Spahn does the same, but manages more than Gröhe to be known and visible to the public and to give the policies his own touch. The challenge posed by the Corona pandemic offers Jens Spahn an additional opportunity to present himself in public and recommend himself for higher office.

G5: But the last two coalition agreements were exceptional in the health care sector. They were so determined and so specific and so to the point, that is what we want. In the expression. In the form. Up to that point. There was not before in any coalition agreement of the last decades. So. And due to these specifications—and then there is the fact that there were two ministers of health. Gröhe, who really worked through the coalition agreement to perfection. And also Spahn, who, in addition to his own accents, Chapeau, has finally worked through the coalition agreement one to one. (Interview G5, 2020)

In the coalition negotiations themselves, there are also opportunities for administrative actors to exert their own influence on policy formulation. While this is limited at the subordinate working group level, this influence can be very substantial and its success often depends on the personal relationships of trust between the individual actors at this level. If actors at this level have a shared history or bond based on analogies, this increases their mutual trust and they are able to leverage their past cooperation—and possibly the ideas that emerged from that collaboration—to achieve major policy changes. Thus, the institutional shift in Germany toward coalition negotiations has also led to changes at the working level. Despite their similar design to the grand coalitionary compromises of the 1990s, coalition negotiations do not function as a group-forming and group-identity-creating institution. Rather, it is the case that administrative actors who have previously worked together can use coalition negotiations as a venue to push their proposals through—assuming that the politicians give them enough space to do so.

G5: How do we do that now with the health coalition negotiations? And the fact is that in such rounds, it is only the politicians who say that we have first, second, third - we have to have all that. And then they name headings and then they somehow say at a certain point, well, go ahead. And then [we] sat together in the evening and wrote down what they wanted. […] Then we formulated what we wanted. And then we presented it to them. Then they said, yes, that’s good. Let’s go like this. So a lot of things are really on a level of trust. (Interview G5, 2020)

Despite the prominent role of coalition negotiations for German politics, there are also repeatedly commissions in the health sector that are set up as a result of coalition agreements or that can also work out a reform informally, which is then implemented politically. Here, too, it is evident that cooperation in the commissions is essentially characterized by trust. Current examples of such commissions in the German health care system are the Scientific Commission for a Modern Remuneration System (Wissenschaftliche Kommission für ein modernes Vergütungswesen; KOMV) (BMG, 2020a; KOMV, 2019) and the Federal Government/Länder Working Group on “Cross-Sectoral Care” (Bund-Länder-Arbeitsgruppe “Sektorenübergreifende Versorgung”) (BMG, 2018; Fricke, 2019), but also the Concerted Action on Care (Konzertierte Aktion Pflege; KAP) (BMG, 2020b).

G8: Then there are more often commissions, even provided for by law, such as the Federal and State Commissions, which may not have a constitutional framework or anything like that and can legislate, but if they agree on a position, on a draft law, on key points, whatever, then that is how it will be. (Interview G8, 2019)

G5: […] and these are coalitions of people who trust each other, even if they have completely different political, let’s say party books. (Interview G5, 2020)

Knowing that these commissions still exist, one might question the statement that programmatic action is no longer present in German health policy today. De facto, however, the overview of health policy reforms, the discourse network, and the analysis of professional actor biographies suggest that there is no programmatic group that builds on common biographical trajectories and currently translates its shared ideas into policy. Instead, the powerful role of the current health minister and coalition agreements has been much more influential in determining health policy in recent years than social groups and policy programs. However, administrative actors can still influence policy.

G1: “Well it is clearly different, yes. In the sense that more impulses come from him. Which does not mean that you have less influence. The debate is simply broader, I would say. So, what I did not know until now was a minister who reads the central statements of the associations on the legislative projects all by himself and then tells the specialist level ‘please do, please check’ on each point. But that is what he does”. (Interview G1, 2020)

These findings lead to the argument that programmatic action only started in the phase between 1989 and 1992 and that the PAF provides a working explanation for policy change until 2011. Thus, German health policy can be explained using different approaches depending on the phase. Before the 1990s, there were no major reforms and health policy did not yet exist as a policy sector with a ministerial portfolio (Döhler & Manow, 1997). This was only achieved in the 1990s by the programmatic group, which then lost its influence in the 2010s. Since then, German health policy has been in a phase of pluralization and fragmentation of interests with little problem pressure. However, it remains to be seen whether the now pressing problems of digitalization and financing, accelerated by the Corona pandemic, will drive the formation of a new programmatic group. Currently, however, the scientific impulses that once enabled programmatic action are now stuck. Advisors work according to orders instead of their ideas emerging from free thinking. If such open-ended scientific exchange is not institutionalized, even the system’s own experts will not be able to effect substantial policy change.

Overall, it appears that the PAF can be a working explanation for policy change, but it can also be refuted. Counter-evidence for programmatic action is the finding that the actors in a given policy sector act on the basis of preferences derived from core beliefs and that there is a normative (rather than programmatic) opposition. Empirically, this is sometimes clearly evident in the interviews: One actor said in the interview that he was closer to an actor from another party with whom he had worked trustfully for a long time, than to any of his party friends. If he had said that he trusts his party friends more, the PAF would be refuted and the statement would be more consistent with the ACF or partisan theory. Moreover, the absence of programmatic action can be shown by the decline of an existing programmatic group. If there is no programmatic group, the sector disintegrates into economic interests where there are hardly any biographical links between the acting actors and thus hardly any trust. The result is that the individual actors fight for their own profit within the rules of the system without developing a vision for sustainable reforms of the sector. If it had turned out in the current situation that Sonja Optendrenk, for example, is closely networked within the SPD and jointly develops programs in commissions with Lauterbach’s research assistant, this would have been more indicative of programmatic action. However, these biographical links between central actors from different parties and interests are not currently to be found.

What has also changed in Germany is the institutional setting of coalition negotiations, which now play a greater role in shaping policy than was the case before 2013. In contrast to the French case, the programmatic group was not able to change the sectoral structures in the health care system in its favor, or if it was able to do so, it did not grant further recruits to fill the seats it had left. This is partly explained by the lack of institutionalization of the education system. In France, the ENA is designed to bring in actors who are already connected to those in power and educates them to fill those seats. In Germany, there are some opportunities for these actors to meet and establish similarly stable career networks. Nevertheless, this needs to happen more actively in Germany, for example in the context of unique opportunities such as Enquete commissions or working groups that also include expert advice, comparable to the professionalization of the French bureaucracy. Only then is it possible for programmatic action to take hold in Germany. If the programmatic group fails to create these opportunities, its success will end. And if these opportunities do not exist again, there is no new programmatic group.

While both the presence and absence of programmatic action can be observed in German health policy, the interviews conducted with programmatic actors and experts suggest the ways in which political institutions influence these phenomena. First, the extent to which a policy sector has not yet been touched by scientific knowledge may facilitate the generation and dissemination of new ideas to shape the system. Such an endeavor requires that the actors who interact have an equal starting position. The fact that the health care system in Germany was just emerging and evolving at the beginning of the 1990s, and the experience with failed reforms in the years before, were extremely favorable for the formation of programmatic groups. Nevertheless, programmatic action did not arise automatically here, but rather as a result of systematic involvement of scientific insights:

G4: that in the 1990s and 2000s we were such an unrepeatable network of health professionals. What we had in common was that, as economists, social scientists, or lawyers, we were involuntary pioneers because the health care system, and especially the SHI system, was a “terra incognita” in the academic field until the 1980s. [...]. (Interview G4, 2019)

Second, and related to this, the increasing reliance on scientific research and findings in preparing reforms had a substantial impact on programmatic action in Germany. Again, several expert commissions were created in the early 1990s to provide space and opportunity for policy actors to discuss ideas and develop reform proposals. The involvement of scientists in these efforts ensured an outside perspective on problems and solutions that could be taken up by the actors in these commissions. Scientific advice was therefore strongly integrated into the creation of policy programs.

I: “[…] do you think that policy change is above all also triggered by just such scientific impulses? Do you think that this is what is actually needed again?”

G6: “Yes, I am convinced of that, I can see that. So, in the background, it is certainly also the case with many—I think that politicians who are not completely pigheaded will always seek the advice of scientists”. (Interview G6, 2019)

G5: So since I was there, we have always discussed the topics of the Council of Experts with each other and worked them out and then partly decided ourselves, but it was always a relationship like that. (Interview G3, 2020)

By interviewing representatives of these institutions, the empirical analysis also allows an assessment of the relevance of these institutions by looking at the relative influence they had in the reforms between 1990 and 2010. The first thing that can be confirmed is the importance of the Enquete Commission, which was already shown in the previous analysis of the reforms. Some of the above interviews indicated above were conducted with former members of the Enquete Commission. It is crucial to note how often the Enquete Commission and the Lahnstein compromise, in which many of the former members of the Enquete Commission participated, are repeatedly mentioned by health policy actors as a point of reference when explaining the adopted reforms:

G6: And we worked more or less on the side, more on longer-term or more structural things. There were so many hearings and so on. And that’s why there wasn’t really much where we had direct influence. But it paid off in the long run, especially in topics like what played a role in Lahnstein and so on. That’s where we did a lot. (Interview G6, 2019)

In order to introduce the ideas of the policy program, the programmatic group used the connections it had built up through the commissions in which it had prepared its policy program. Some of its members, such as Franz Knieps, who was head of the key department in the health ministry, took a central position in the decision-making process. They were also able to work with high-ranking politicians, such as the then health minister Ulla Schmidt, to push through the proposals. With regard to Ulla Schmidt, the importance of biographical connections is explicitly mentioned by one interviewee, who had even attended the same high school as her.

G6: Clearly in the background, of course, we also tried to exert political influence, and not least through Franz Knieps and others, that then- and Ulla Schmidt, whom I of course also know, she graduated from the same, three or four years before me at the same high school in Aachen, Reinhardt-Gymnasium. (Interview G6, 2019)

Examining a little further which actors were members of the group and for what reason, it becomes clear again that biography is the determining factor for programmatic action. Depending on their professional background, each programmatic actor was characterized by a driving force that led them to engage in and shape programmatic action. For one type of programmatic actor, it was their health economics educational background that drove them in their efforts to introduce competition into the health care system. For others, it was rather the experience they had in the health care sector and/or their party-political streak that led them to embrace the program of competition. Either way, their involvement in commissions and their cooperation made them a group, and the policy program was what their combined preferences yielded.

G2: While some, so to speak, by training economists, socio-economists […] saw it from a competitive point of view, that is, ideological aspects, rather to create such a competitive order between health insurance companies and to further develop and optimize it intertemporarily, others, now also from my social democratic side, were driven by the injustice of the existing system. (Interview G2, 2019)

Despite these different drivers of action, there was a common strategy in promoting the content of the policy program. This strategy was to use science and scientific methods and results to inspire and legitimize the reform program, specifically that of evidence-based medicine, at a time when the affordability of health care was a real problem. So, the programmatic group took advantage of the situation that there was an urgent need for reform because of the problem pressure, and answered that problem with an evidence-based solution that at the same time helped them in communicating reform.

G5: There was a time when the competition order was established for this reason. Because, basically, the competitive order should be the counterpart of this, let’s say, eminence-oriented medicine. Competition means data, means clarity, as a stringent DRG […] In the next legislative procedures, the points were still picked out and worked through on the basis of the Enquete on structural reforms. […] We were able to see that politicians only acted when the financial situation required it, that is where we are with the topic, and if any external influences, the pressure was too great that they had to act, but no longer with a proper guiding vision and the big competition topic, that is long gone. (Interview G5, 2020)

7.3 Intermediary Conclusion: Institutions of Programmatic Action

By recognizing the different institutions that exist in the two countries and the fact that programmatic action takes different forms depending on institutional conditions, the empirical analysis has uncovered institutions that are similar in both cases and can be seen as the central driving forces of programmatic action. However, these institutions are different from those normally studied by comparative politics and policy process research. In short, the institutions relevant to programmatic action are those that bring together and enable long-term cooperation among hybrid actors active in different institutional environments.

The test of the hypotheses formulated in the theoretical part shows that the institutions often considered in comparative politics can have an impact on programmatic action, but are not crucial for the existence of programmatic action. The programmatic group in France has succeeded in permanently strengthening and perpetuating its policy program through institutions. The recruitment of actors to fill these central positions (e.g., in the DSS or UNCAM) also occurs reliably through these networks. Germany has also managed to create new institutions. However, these do not fit into the original logic of the system, so they are not congruent with the prevailing institutional realities and have become alienated. Consequently, programmatic actors are no longer to be found here. As expected, decentralization and self-governance, as well as the strong role of parties in Germany compared to France, have an ambivalent effect here. They can promote programmatic action when they complement each other—as happened in Germany between 1990 and 2010—and serve as a power resource to push through policy programs. But they can also act as veto players and hinder programmatic action. In France, the traditional non-participation of corporatist actors has repeatedly led to strikes throughout the programmatic action phase, generating resistance to planned reforms. Programmatic groups in the French majoritarian democracy therefore also face hurdles in overcoming sectoral or subnational resistance.

Nevertheless, a stable elite formation system and the institutionalized integration of scientific expertise into the political decision-making process prove to be relevant institutions of programmatic action. When these institutions are stable, programmatic action is repeated. However, if these institutions change over time, this spells doom for programmatic groups if they cannot use the new institutional structure to their advantage. Figure 7.3 visualizes the institutions of programmatic action that were present in both French and German health policy and to which programmatic action can be traced.

Fig. 7.3

Empirical evidence in support of hypotheses. Source: Own illustration. Favorable influence of political institutions on programmatic action

In the case of the relevant institutions of the elite formation system and the constant integration of scientific expertise, it also becomes clear that it is primarily the individuals who become programmatic actors who are active in different arenas. They succeed in linking different arenas and, as a programmatic group, in exploiting the different resources inherent in this linkage. In this respect, they resemble boundary spanners who span multiple arenas rather than multiple issues (Brandenberger et al., 2020).

Regardless of institutions, the role of programmatic groups also shows that they are particularly successful when there is high problem pressure and a need for reform in a sector. In both France and Germany, the financing of the health care system faced growing challenges in the early 1990s due to rising expenditure. These were addressed with different instruments. While in France the effects of the financial crisis and, more recently, the Corona crisis put the health care system under constant cost pressure, the policy program implemented in Germany largely averted the financial crisis. In Germany, the measures introduced in the pharmaceutical sector and low unemployment prevented the health care system from running into financial problems. In this case, therefore, the policy program virtually abolished itself, since it successfully solved the problems and there was subsequently no longer any problem pressure for further programmatic action.

It also striking, however, that in Germany and France the discourse and apportionment of blame differed. In Germany, it was primarily the ruling black-yellow coalition at the beginning of the 1990s that was blamed for the failure of cost containment, with the SPD leading the way. In France, there was more of a conflict between the central government in Paris and the sectoral or regional players. This is sometimes also due to the political system, so that different actors have to be brought together in programmatic groups depending on the country. In contrast to France, the role of party-political actors is stronger in Germany, and many sectoral and regional actors must also be involved. Thus, the success of a policy program depends in part on the support of other groups of actors (Sciarini et al., 2021).