The vertical approach to party leaders

8Building on Michels’ research, [9] the elitist perspective has prevailed in studies on party leaders. Influenced by monism, this perspective has focused on mechanisms of concentration of power and collusion of elites. [10] These studies have mainly relied on the paradigm of the iron law of oligarchy, refining and testing its empirical value. The organizational approaches of Duverger, Katz and Mair are also based on the assumption that power is monopolized to the detriment of a helpless, manipulated base, and that leaders are interchangeable, sharing the same corporatist reflexes. [11]

9There have admittedly been exceptions to this vertical approach: S. J. Eldersveld, whose work is all too often reduced to the elaboration of the “stratarchy” concept, has empirically tested and questioned the postulates of Michels’s iron law by assessing the coherence of party organizations. [12] Likewise, critical sociology applied to parties suggests taking into account the system of internal positions that creates a more or less implicit hierarchy of leaders. [13] Within “dominant coalitions”, Panebianco’s approach, revisiting the concept of leadership, reconsiders the difference between bureaucrats (or believers, entirely committed to the management of the organization, whose ambition is collective only and who have very limited autonomy) and political entrepreneurs (or careerists, who are to a certain extent free from hierarchical constraints, while possessing a charismatic and strategic authority). [14]

10These diverse attempts at exploring internal differentiation at the top levels of organizations remain however grounded in a conception of the party as an accumulated hierarchy of strata or circles: interactions between the leaders are not properly studied. Existing literature on the PS only addresses them in passing, focusing on two main avenues of research. The first one consists in examining the homogeneity and the continuity of the socialist trademark: a result according to Sadoun [15] of the ideological specificity of the party, which has enabled it to resist the temptation of communism and totalitarianism, and according to Grunberg and Bergounioux [16] of “the long remorse of power” that has characterized party leaders throughout the past century, constantly stuck in a cycle where the quest for power was immediately followed by disillusion as their political programme was implemented. The second approach, proposed by Sawicki, [17] consists in exploring the heterogeneity of the party instead, both in spatial and temporal terms, at the level of the activists and of the departmental federations. Rather than focusing on the organization as such, this approach seeks to demonstrate that the party is pervious to multiform social networks, as well as institutions that characterize the French polity such as municipalism or presidentialization. [18] Recently, the study of the primaries and of the presidentialization of the PS [19] has rekindled the debate between these two perspectives, as advocates of the former saw a paradoxical strengthening of the leaders’ power (Lefebvre) and champions of the latter claimed that clarifying the socialists’ relationship to power and opening the party up to sympathizers could only bolster internal democracy. These two approaches aim to draw general conclusions and thus do not grant particular importance to the study of the internal coherence in the group of party leaders; though their methodological and theoretical perspectives are opposed, they paradoxically converge in their judgment of a group that is seen as motivated only by the quest for power, or even an “economy of cynicism” [20] said to counterbalance the centrifugal tendencies of the organization via clear and shaved electoral goals.

11The internal coherence of the group of party leaders has been postulated rather than genuinely investigated. The sociologist François Bourricaud was the one of the first to conclude from the semi-specialization of tasks in parties that leadership has by definition a plural dimension. While major organizations have not become specialized enough for each person to be assigned one function only, these tasks are still too numerous and different to be carried out by one person. This suggests the existence of a new configuration; a system of relationships that remains to be studied. [21]

12

“The problem of authority doesn’t lie with the leader, but with the leadership. If we consider that leadership is carried out by several leaders, the key problem becomes that of their relationships and their harmony, and accordingly we are led to examine how a constellation of individuals can become an efficient coalition.” [22]

The relevance of the “collegiality” concept in the study of political parties

13“Collegiality” has been used in an attempt to conceptualize such coalitions. Introduced in the field of political sociology by Weber, it has been picked up and refined by sociologists of work (organizations and professions); scholars working on political parties could now benefit from reconsidering this notion. A legacy of the Athenian and Roman polities [23] as well as of the Catholic church, [24] the concept appears in Weber’s work without being really elaborated upon. Defined as a government of peers, following the blueprint of Sparta, Venice or Switzerland, it is presented as a form of political regime that is ultimately doomed, due to the inevitability of the bureaucratization of parties and governments. [25] Neglected by political scientists, the concept was picked up by organizational sociologists. It was for instance prominent in English-language studies on the governance of universities, [26] already perceived in the 1960s as organizations defined by consensus-oriented decision-making processes and the ability of the academic community to self-organize without requiring an external hierarchical authority. [27] Sociologists of professions have also used the concept of collegiality as a normative principle, forcing professionals with unequal competencies and hierarchical statuses to work as a “team”. [28]

14These elements apply more or less directly to the political field and political parties. Limitations relating to disciplinary boundaries in studies on parties have often been pointed out; Sawicki and Siméant have for instance have called for more integrative approaches using tools from the sociology of social movements and collective action. [29] Likewise, there are insights to be gained from the sociology of professions and organizations. In the recent literature on political parties, especially research inspired by cartel party theory, there is a consensus that the modernization of parties should be analyzed as a professionalization and a normalization (or homogenization) of these organizations. Parties are also routinely conceived as “enterprises” (in line with Weber and Schumpeter’s works) whose actors are moved by individual and collective interests and pursue strategies. In this perspective, the analyses of competencies and statuses as well as of relationships of power (these concepts in France, when applied to political parties, routinely reference research by Crozier and Friedberg), [30] would benefit from being reviewed and refined using other research on organizations. In the 1980s, neo-Weberian sociologists endeavored to reformulate the distinction between collegial organizations (“experts” or professionals working on non-routine tasks in committees of peers) and bureaucratic organizations (whose members are linked by a hierarchical principle and are expected to carry out partial and short-term tasks). [31] In the formulation proposed by Waters, [32] a collegial leadership is characterized by the specialized expertise of the members; this expertise is the basis of their legitimacy and of the formal equality they enjoy. A collegial body is furthermore defined by self-regulation, collective decision-making (through consensus and persuasion), and the reciprocal monitoring of colleagues’ output. These criteria have been refined, systematized and empirically tested in Lazega’s studies on corporate law firms, and more recently on Catholic dioceses. [33] The Weberian concept has thus been enriched with references to structural sociology (the relationships and practical modalities of exchanges between members ultimately count more than the postulated finality of the consensus) and to the strategic analysis of organizations (peers are entrepreneurs looking for niches and fighting for differentiated forms of status).

15This does not mean political parties should be considered as “organizations” or “enterprises” like any others, if only because they are usually based on associations of volunteers and because their vocation is to produce standards, ideas, and frame behaviors (which to some extent makes them “cultural enterprises”). [34] Applied to political parties, the criterion of expertise, among others, requires reinterpretation. It refers to a status, for instance being an electoral representative, that can go hand in hand – albeit not necessarily – with a form of thematic specialization: being polyvalent and having highly varied skills are qualities that actually tend to be most valued at the top levels of the organization.

16The concept of collegiality is however useful to the study of collective action between peers, or more broadly of the possibility of an organization without a formal hierarchy. More precisely, it stresses the complexity of the relationship between formal, constitutional apparatuses and practical modes of exchange and decision-making. In this sense, it sheds light on the raison d’être of the party’s governing bodies. Who speaks in the deliberations and negotiations that take place there and why? What gets exchanged (types of arguments, of information) and how? What type of (non) decision-making is produced? Which forms of leadership and internal discipline does collegiality ultimately bring about?

Statutory collegiality

18The PS is not the only political party that asserts its collegiality. The latter is generally a feature of left-wing parties, whereas right-wing parties (at least in France) explicitly emphasize the principle of the leader’s charismatic authority. [36] More precisely, collegiality as an openly stated principle is one of several criteria (how the leaders are chosen, duration and plurality of internal mandates) used to assess internal democracy in political parties. The green party (les Verts), for instance, supports collegiality to the point of questioning the notion of leadership. [37] Collegiality can in that sense be considered as a minimal safeguard against the personalization of power, in order to prevent spokespersons from deviating from the collective line. In the PS (as in les Verts), collegiality is formally based on the rejection of the pre-eminence of a leader (there is no president, but a “Premier(e) secrétaire” [first secretary]) and on a proportional representation system that divides power between several factions. But unlike in les Verts, where collegiality is bolstered by highly binding rules (rotation of leaders according to the revolving-door principle, strict limitation of the number of internal mandates), it is in the PS merely held as a standard, liable to prevent oligarchical tendencies.

19Lastly, in the PS, collegiality relates to the confederal dimension of the organization: leaders are elected according to a double principle – factions and federations must be represented. Collegiality is thus manifest in the way the internal bodies function, but also in how the national bodies recruit: in the Conseil national and Bureau national, members are picked from lists, in proportion to their motion’s score at the Congress. They are thus selected on the basis of their ties to a faction and of the place attributed to them by that faction in the team in charge of representing it. In these two bodies, moreover, positions are undifferentiated.

20The composition of a faction’s list for the Conseil national is based on two constitutional principles: diversity and complementarity. Both the candidate’s individual resources (speaking or writing skills, local and/or national exposure) and collective resources (importance of the candidate’s federation, elective mandates, etc.) are taken into account. Jean Mallot, who was in charge of selecting candidates for the “Nouveau Monde” faction (until 2005 headed by Henri Emmanuelli and Jean-Luc Mélenchon), uses a sports team metaphor to describe the collegiality of the list (“it’s a bit like coming up with a line-up for a football team, you’ve got to have people who can strike and others who can defend”). Lastly, the territorial representativeness of the faction must be guaranteed.

21

“It’s the same old debate. You can argue for the inclusion of a comrade who’s got no mandate, but who’s an up-and-coming 30-year-old. On the other hand, in some federations, it can be worthwhile to put in a comrade who’s the mayor of an important city, because he’s going to bring in people and he might bring people into your faction. Territorial representativeness also counts, because it’s in the best interest of the faction to have the entire territory covered, certain regions, because you have to prepare the future, in view of the appointments, to bring new blood and have experienced people with you. But those things aren’t set in stone; there’s no recipe. Every case is different. The ultimate goal is to further the faction’s interests, by having people present in the governing bodies, and making sure they are efficient…” [38]

22The Conseil national and Bureau national thus appear as heterogeneous agglomerates of competing “teams” formed by the different factions, meeting numerous criteria and whose composition is left to the evaluation of faction leaders (“there’s no recipe”). Collegiality within these governing bodies therefore has to contend with all these different sub-groups. [39]

23Formalized in the party’s constitution, collegiality plays out in practice as a form of party culture, materialized in distinctive forms of language (although not specific to the PS) – use of the “tu” form, of the term “comrades”, the predominance of “we” over “I” – and in spatial layouts (square tables at the Secrétariat and Bureau national, no raised platform). It is also embodied in the interactions between peers.

Collegiality in practice

24In order to grasp the actual impact of collegiality on the leaders’ and the party’s activities, we have to move away from an exclusive conception of the notion. The relative dimension of collegiality is clear in the way the governing bodies work in practice – members are formally equal and their votes are equally important. As we will see, the obligation to publicize decisions taken by the collegial bodies has implications on the production of political decisions, but it also affects the boundaries of the topics addressed, and more broadly the role of these bodies in the party’s hierarchical system. The statutory constraint can result in a minimal application of the statutes, in the sense that only issues that do not raise a challenge to the deliberative function of the governing bodies are put on the agenda (in order to avoid those that are not likely to be settled by consensus).

25

“Hollande has known for a few months [since the internal referendum on the TCE, the Treaty of Rome] that he’s in the minority in the Bureau national (BN): look at the BN’s agenda, it’s very sparse, there’s a number of points that he doesn’t put on the BN’s agenda. That has to mean something for the life of the party, that means there’s a point to the BN… can you imagine if there were only the SN [essentially made up of members of the majority] left? The BN has a proportional system, at least […] The BN is not a municipal council or a regional council where things work in terms of opposition/majority, right/left, the point is to provide some policy input, right, it’s not because you can’t communicate in the BN that there’s in fact no point to the BN; it’s precisely because whoever wrote the statement knows it will be submitted to the BN that he wrote it that way […] He knows otherwise it’s not going to make the cut.” [40]

26Actors’ use of the party’s governing bodies is thus very clearly informed by the collegial constraint, which in the PS goes hand in hand with the internal pluralism created by the proportional system. These adjustments to the constraint can be made negatively, through avoidance strategies, but also positively, in the drafting of texts prior to deliberation. Generally speaking, the proposals submitted to the Bureau national (statements, resolutions…) are drafted with a view to achieving consensus (perceived in opposition to the arithmetic rationale that prevails in municipal councils) and avoid open confrontation, which is detrimental to the party’s internal cohesion and public image.

27The implementation of collegiality can therefore result in watered-down agreements; it must also contend with other types of regulation. As Waters explains, personal leadership and collegiality co-exist in most bureaucratic organizations; it is mostly the proportion of each of these forms of regulation that varies. [41] Bureaucracy, for instance, is more suited to the efficient monitoring of the organization’s material resources or the recruitment of staff. The party’s national headquarters, in the rue de Solférino in Paris, can be considered as a predominantly collegial organization, where decisions are taken collectively by the national bodies but work is essentially performed by the permanent members, who are salaried by the organization and bound by their subordinate hierarchical position within the organization. In addition to co-existing with bureaucracy, collegiality must also reckon with various forms of leadership. Indeed, within a collegial system, peers or associate-rivals compete to obtain a privileged status within the group.

28

“Members of a group compete to obtain a form of status, even if the latter entails responsibilities, for instance that of the long-term future of the organization. These members exert a form of leadership.” [42]

29This leadership is visible within the factions, which function in an explicitly hierarchized manner. Within the collegial bodies themselves, the superiority of the faction leaders manifests itself in the existence of exclusive meetings between them before and sometimes during the debates (for instance in the case of electoral committees) and by the implicit rules on speaking at meetings: it rarely happens, for instance, that a participant with lesser credentials directly addresses a faction leader. Leadership is thus present even within the collegial structures. According to Lazega, it results from specific expectations: other members look for the leaders to bring solutions to the problems of collective action, particularly regarding the medium- and long-term future of the organization. It also results from a concentration of social resources (exogenous or endogenous to the organization), which is demonstrated by the fact that faction leaders often have experience of ministerial and/or major representative positions, and as leaders of youth movements. In order to genuinely account for the collegial phenomenon, we therefore have to observe that there is no such a thing as a “democracy between peers” here.

30The practice (or the proportion) of collegiality within the organization also varies according to the context. The ethnographic study that provides the material for this paper began after the socialist defeat in the 2002 presidential elections and was completed in 2008 at the time of the Reims Congress, after a third consecutive defeat in the same elections. Conversely, between those two dates, the PS won all mid-term elections (municipal, regional, cantonal, European…), which paradoxically contributed to a weakening at national level by boosting the independence and the increasing importance of the local representatives. Though factions were constantly undone and reformed, Hollande managed to retain his leadership thanks to makeshift majority coalitions, without the support of a structured faction. This weakening of the national leadership worsened after the failure of the referendum on the Treaty establishing a Constitution for Europe [Treaty of Rome]: the active campaigning of a number of socialists against the treaty (Mélenchon), at odds with the line voted by the majority of the party, publicly revealed the limitations of the leadership’s coercive capacity. Is the importance granted to the concept of collegiality in the study of the leaders a bias introduced by the study (as the ethnographer can only perceive what is temporally and geographically accessible to her), itself a result of the context? My observation took place during a period when the national leadership was weak, in an organization whose ties were loosened and weakened, and where collegiality might have been all the more obvious as the “associate-rivals” were “condemned to make do with each other”, forced to cooperate since no leader was strong enough to impose his or her point of view. [43] Yet, I argue that the concept remains relevant even in periods of strong national leadership, for instance in the “party of Epinay” led by Mitterrand, or with Jospin at the helm between 1995 and 1997.

31Available research on Mitterrand’s and then Jospin’s leadership stints shows in both cases the importance of the efficient use of the collegial dimension in the party leadership. Mitterrand stepped back from the collegial structures (and bypassed them through informal and bilateral structures) but still formally abided by them, leaving them enough autonomy so that all party sub-groups (factions, federations) could have room for expression. [44] Jospin also revived collegiality, albeit under other forms: he promoted the devolution of the party as a response to decentralization (for instance at intermediate levels by normalizing and professionalizing the Premiers secrétaires fédéraux, now in charge of implementing decisions taken at national level) and remotivated the national bodies (which became teams with diversified competencies and expertise, and not just purely strategic combinations of sub-groups). [45]

32In other words, there is no single relation between collegiality and the power of personal leadership. In the process of comparing these different phases, I have evidenced different uses of collegiality. In its (very rare) purest configuration, observable in weakly bureaucratized organizations, it generally results in a form of “weak leadership”. [46] In the parties, it tends to consist of a pattern of “pockets” that resist bureaucratization and personal power to varying degrees; it can also end up being just one way of managing the organization among others, a form of bottom-up strategy used by the top level to legitimate its own power.

33Collegiality thus sheds light on how top-level groups function, specifically in this case the national governing bodies of the PS. Insofar as the latter are first and foremost places for debate and deliberation, collegiality has practical effects on the conditions of public speaking and on the standards of debate.

Differentiating sources of legitimacy and status

36In collegial bodies, everyone has the right to speak (to vote and to veto), but not all members speak from the same position. As we have seen, a government of “peers” does not mean that all members are strictly equal or that they identify among themselves. The behaviors of the governing body leaders are thus the product of constant adjustments between differentiation and equality, between division and construction of a unity that constitutes the institution’s façade.

37In a closed system, where recruitment is partly based on cooptation [48] and the functioning of the group relies on a bounded solidarity, the types of reciprocal influence that leaders can and do seek to exert on each other essentially depend on the type of status they can aspire to. In Weberian sociology, status is the composite product of social achievements and resources: it chiefly means having political and social resources, a degree of prestige liable to confer recognition of individual responsibility and relative autonomy in the collegial bodies. Yet as Lazega points out, “for Weber, collegiality is only possible because statuses derive from heterogeneous sources […]. For instance, the most competent, the most popular and the most committed official representative or associate each has a form of status”. [49] The stability of this polycracy therefore rests as much on differences of status as on the interdependence and/or reciprocal neutralization of the leaders, who each control different resources.

38However, the diffusion of power within the collegial bodies and the façade of the formal equality between members both make it difficult to identify internal hierarchies and different statuses. This is particularly the case in the PS, which is characterized by complex recruitment rules and highly diversified electoral and organizational resources. The internal evaluation of members’ statuses depends partly on the “formal” and “informal” rules in play at the time of recruitment (i.e., during the process of cooptation on to the motion’s lists proposed to member vote). [50] Among the formal rules listed in the party’s constitution are the representation of factions and federations, but also gender equality and even turnover (since 2003, the constitution has stated that the governing bodies must include at least 30% of new members each time it is re-elected). The informal criteria tend to relate to the internal and external social resources that prospective members can use to support their application: the ties to networks that can serve the party, the position within a given faction or the recognition of electoral legitimacy, while often not explicitly taken into account, are key parameters.

39Based on a questionnaire submitted to Conseil national members in 2003, [51] I have evidenced five typical trajectories to the Conseil national using principal component factor analysis (PCA): those who specialize in organizational tasks (federal leaders, parliamentary assistants…), those with an electoral mandate, those who were members of youth movements, experts (grandes écoles, academics), and members of a ministerial cabinet. The variables that match these trajectories are organized on two opposing lines: one between internal recruitment (acquisition of activist and electoral resources) and external recruitment (recognition of external expertise accumulated in a grande école or in a ministry), and the other between “delegated” resources (youth movements, premiers secrétaires fédéraux, parliamentary assistants) and “individual” resources (a stint in a grande école, an electoral mandate, a party mandate). One of the main findings of this survey is that trajectories involving internal, electoral and/or activist experience are, contrary to generally accepted ideas (especially regarding the importance of having attended a grande école), much more frequent than external trajectories. [52] Electoral resources appear to be increasingly indispensable as one climbs the hierarchical ladder, and a mandate as MP is particularly effective at opening doors to the Conseil national. Between 1993 and 2003, 35% of Bureau national members and 44% of Secrétariat national members had an MP mandate, as well as 13.5% of Conseil national members; 31% of BN and SN members had been mayors. Experience of several mandates remains the most effective strategy, as 76% of BN members and 79% of SN members had served at least two different mandates. [53]

40Reflecting this disparity of resources, public speaking is very unequally distributed within the governing bodies. According to the 1999 minutes of the Bureaux nationaux (weekly meeting of around 70 members proportionally distributed among the factions, in charge of establishing the party’s policy line), nearly 50% of participants almost never speak (between 0 and 5 times over a year) and almost two thirds speak fewer than ten times. Yet over this period as most faction leaders were members of the government they did not attend Bureau national meetings and were therefore unlikely to monopolize the discussion. The number of contributions does not so much reflect internal influence as it reveals different uses of the governing body, which may in some cases relate to diverse statuses. Tellingly, the most active contributor was the Premier secrétaire (François Hollande, more than 130 contributions), followed by the secrétaires nationaux in charge of fields affecting the core of the party’s activity (finances) or requiring rapid reaction to current developments (international relations). The three speakers who contributed more than twenty times were important figures in the party: Pierre Mauroy (viewed as a moral authority in the party, 21 contributions), Jean-Luc Mélenchon (25 contributions) and Marie-Noëlle Lienemann (29 contributions).

41As we can see, this inequality in speaking publicly simultaneously matches differences in title, function, prestige and strategic uses that do not necessarily coincide: the Premier secrétaire does have a higher status within the party, but also plays the role of an arbiter, which explains the fact that he contributes often. Mélenchon and Lienemann, as leaders of the Gauche socialiste faction, used the Bureau national as a platform to systematically express their minority’s point of view; Mauroy also used it because it was one of the few political arenas where he was not just one grand élu (important elected representative) among others, but it was also where the legitimacy of his career (as a former SFIO member and co-founder of the PS, Mitterrand’s close friend, former Prime Minister and former Premier secrétaire) would be recognized. The unequal number of contributions also reflects a conception of representation that is specific to the PS, where a number of party leaders play a role of mirror-representative (the “silent” are meant to reflect the diversity of the party by the sole fact of their presence) rather than being the political mouthpiece for the sub-groups they are supposed to represent.

42Furthermore, my interviews show that members have a high degree of awareness of their status and of the expectations that come with it, and even of the extent to which they have the legitimacy to participate. Barbara Romagnan, who became a BN member at the age of 30, representing the NPS faction, [54] explains that she found it difficult to adjust her behavior to what was expected of her considering her relatively low status.

43

“I remember my first Bureau national meeting because I was the first to speak; I was really worried about the prospect of playing a merely ornamental role. […] Not for a second did I think I was up to what I thought were their standards. You know, I wouldn’t have moved mountains to be there in someone else’s place, but right then I thought one of the ways I could justify my presence apart from gender equality was by speaking out. I was prepared. […] The thing is, pretty much no one gives a damn what I might say. There are people that need to talk, because you understand they’re on this or that side, so even if no one cares what they say, it makes sense for them to speak, whereas no one gives a rat’s ass about me. So if I don’t have anything to say, not only are they not going to listen to me, but there’s no point to me being there, I’ve got no legitimacy.”

44Vincent Peillon, who became a BN member in 1994, representing the only minority motion (“motion n^o 2”) against that of Premier secrétaire Henri Emmanuelli, recounts a similar experience: as a young party member, he felt obviously inferior in status to the party’s leading figures; the situation was made even more uneasy because of the internal conflicts between faction leaders, who used the BN as a sounding board for their policy stances, thereby reducing other members to the status of spectators.

45

“When we came to the Bureau national, we were very young and overwhelmed. Some of us were 22-23, like Christophe Clergeau, Christine Priotto; some never sat on the Bureau after that, actually. It was really intimidating, first because of Emmanuelli, of DSK, there was a very heavy, unpleasant mood, they treated each other like dogs. We were there, and well, we spoke, but we had no mandate, no legitimacy as elected representatives… And then you had Mermaz, Mauroy, the leading figures…” [55]

46The balance of the debates that take place within governing bodies thus rests on a clear differentiation of the participants’ statuses. The latter have a direct effect on the contributors’ legitimacy, but also on the content of their contributions, and therefore on their reception as well. Indeed, status, such as it is perceived by peers, encompasses, in addition to a title and the individual and collective resources that come with it, the degree of “representativeness” of the contributor – in other words, whether he or she is entitled to speak in the name of persons other than himself or herself. Whereas Romagnan’s relatively low status reduces her degree of legitimate participation to silent presence, a higher status conversely results in an entitlement to express oneself in a way that is immediately interpreted in the light of the internal competition.

Sustaining cooperation: the lateral control of the actors

47The conditions of public speaking depend not only on the actors’ statuses, but also on collective devices. In order to sustain the collegial structure (and the institutional façade of the party), the diversity of statuses and resources must go hand in hand with operating systems liable to ensure a minimal degree of cooperation. These operating systems are mainly the result of informal practices.

48The conditions of cooperation are as a rule guaranteed by compliance with a number of rules that normalize exchanges. Compliance is ensured by a ritualized organization of seating in the room, of public speaking and, to a certain extent, of the content of the contributions. [56]

49Since the mid-1990s and under Hollande’s leadership, Conseil national and Bureau national meetings have followed the same protocol. In the CN, which meets four times a year for three or four hours, the president of the CN (an essentially honorific position) and the Premier secrétaire are flanked by a number of personalities at the center of the platform. After a short introduction by the president, and the presentation of activity reports by the secretaries in charge of federations and elections, the contributors speak for a short time: up to fifteen minutes for faction leaders, between 3 and 10 minutes for the others. The Premier secrétaire traditionally brings the CN to a conclusion. In the audience, the members are grouped on the basis of their faction or of their motion. In some rooms where the layout makes it possible to have members sitting on opposite sides, such as the “third basement” room of the Assemblée nationale, [57] factions are always positioned in the same spots: on the left, the left of the party (former Gauche socialiste, Nouveau Monde, NPS…), on the right, the Strauss-Kahn and Rocard supporters, and in the middle, the “marais” [members who are not particularly identified with a given faction] and the close allies of the Premier secrétaire.

50Bureau national meetings follow a similarly well-oiled ritual. Every Tuesday at 5pm, the MPs come from the Parliament and the Premier secrétaire walks down the stairs from the large second floor room (salle Marie-Thérèse Eyquem, rue de Solférino) where he has had a final preparatory meeting with his head of cabinet; the participants chat in small groups, waiting for the meeting to begin. These few minutes allow the elected representatives and the federal executives, back from their constituencies, to tweak their statements and to have informal contacts, which is more difficult for them during the rest of the week. Inside the room, the topography of the participants is a rather accurate manifestation of the PS’s representation of the distribution of power within the party. The tables are arranged so that all participants can face each other. The Premier secrétaire presides, flanked by his head of cabinet and the secretary in charge of the federations and organization. He has a central position, but there are no other signs of his pre-eminence: all the tables have been pushed together, apart from that of the permanent members, which is behind the right side of the quadrangle. The participants are seated on the basis of their faction allegiances: those from the majority are closer to the Premier secrétaire, whereas minority members sit on opposite sides of the quadrangle. Seating changes reflect transfers of allegiances to a faction or motion.

51Regardless of current events, the BN meeting follows a three-part blueprint. The Premier secrétaire opens proceedings by giving his political analysis of the past week (in the opposition, critiquing government action; when in power, pointing out government measures that should be promoted) and introduces the first general debate. The second part, devoted to the “expression of the party” allows the secrétaires nationaux to present statements defining official positions; the third part is dedicated to the “life of the party” (internal affairs, organization and mobilization of the members).

52To ensure that everyone complies with the standards and “plays the game”, mechanisms of reciprocal control are necessary. Among other things, this means, as Lazega shows in his study of corporate law partnerships, “handling opportunistic behavior”, since members depend on each other and ultimately on the unity and the smooth running of the collective body. In order to be accepted, control must be reciprocal, but first and foremost it must be informal, to avoid direct power struggles liable to endanger the group’s internal solidarity.

53Thus, a “lateral control regime” is at work: firstly, because once all the informal solutions have been exhausted, there is recourse to third party “sanctioners” who were not involved in the conflict (chosen for their experience, their loyalty to the organization, etc.); and secondly, because these “sanctioners” are not hierarchical superiors, but peers. [58]

54Devices encouraging cooperation can be produced within party sub-groups before the meetings: the balance reached within the governing bodies is at some level the product of partial sub-balances, which may come into conflict but also neutralize each other. Usually before each BN session, factions have a preparatory meeting meant to determine common positions on the issues to be discussed and frame the contributions of their affiliates. More broadly, factions act as spaces for socialization and exchange that result in a harmonization of behaviors and practices. Jean Mallot (then regional councilor and Premier secrétaire fédéral of the Allier department, Nouveau Monde faction) particularly emphasizes the latter dimension:

55

“If I want to speak, nobody’s going to stop me. If someone sets me right, I’ll check during the faction meeting that I didn’t say something stupid. But anyway, if it’s about a really substantial position, we’ll discuss it beforehand. I don’t see any of us speaking out in the BN on immigration quotas without having discussed it beforehand in the faction meeting. If we’re just going to restate the position adopted by the faction last year on job outsourcing, then we don’t need to consult, we know. We’ve got our act together: if we’re going to play, we’re all going to be playing the same tune.” [59]

56Attempts at framing contributions can admittedly only be effective up to a point, partly because those who consider their own status to be higher disregard them (including by not attending these meetings and/or not following collective instructions). Yet, the preparatory meetings at least provide an opportunity for exchanging information, and allow faction members to anticipate possible conflicts and refrain from making things difficult for their allies.

57Both at faction level and at party level, cooperation seems in fact better sustained when the devices used are either negative or entirely implicit. The normalization of the debates therefore first entails sanctioning and/or stigmatizing deviants (i.e., those who fail to meet the expectations created by their status, or who do not conform to collectively accepted language standards). Again, this chiefly points to the differentiation of statuses within the governing body, and in such a case to the weakness of the position of whoever is stigmatized. Romagnan, for instance, told me how the leaders of her faction scolded her when she spontaneously decided to criticize a position put forward by a party dignitary:

58

“It’s like that time when Juppé was indicted. Emmanuelli was a guest on Karl Zéro’s show [a satirical infotainment TV show] and said ‘I wish him good luck’. That was shocking, it wasn’t the same thing, before the law on the funding of political parties and afterwards… [60] I get that he wasn’t going to be too harsh, but saying this… And then at the next BN meeting, Pierre Mauroy asks to speak to point out that Emmanuelli’s comment was very dignified, and I was just stunned. Okay, if someone had attacked him, it’d make sense to defend him, but between that and asking to speak in a BN meeting to say that… I didn’t feel like stepping out of that room without having said something. So I was really embarrassed, but I raised my hand and I said, ‘well, I found Henri Emmanuelli’s comment very shocking, it wasn’t the same situation’ and all that, right… And then I see everybody staring at me, just staring… even Benoît [Hamon]. So I sent him a note to ask if something was wrong; he answered ‘Was your contribution settled upon in the BN preparatory meeting?’ and I told him ‘no’ and he answered that it made things a bit tricky. So then we don’t have the right to speak about tricky stuff anymore? […] I was very embarrassed, but I would have been even more embarrassed if I hadn’t said anything.” [61]

59In this case, Romagnan was scolded by her faction leaders and faced disapproval from the members because she broke the rules that generally prevail within the governing body, and which are coherent with the principle of collegiality (avoiding ad hominem criticism and direct confrontation with other participants). Her transgression was judged all the more harshly as she did not possess sufficient status to break the rules: by slamming a faction leader (Emmanuelli) and a leading party figure (Mauroy), Romagnan upset informal but very real hierarchies. The weakness of her status made it impossible for her to express the position of her faction, even though she derived what little legitimacy she had from her faction.

60The perpetuation of rituals is therefore the primary vehicle for socialization. The transmission and reproduction of standards in terms of behavior and language occur through a process of imitation; the occasional transgressions and the scolding that immediately follows them only make it clearer that these implicit rules have a binding value.

61This socialization happens in several stages. Initially, newcomers tend to find the coded functioning of the body uncanny. This coded nature contributes to the effect of a closed-off group, appearing as a complex universe based on tacit routines that make little sense to the non-initiated. Subsequently the new members start to learn the rules, in different ways depending on how far ahead they are in their political career. This happens chiefly through imitation. Beginners in politics are merely expected to be present, as Jean-Claude Pérez explains – he entered the BN in 1993 at the age of 29, coopted by the Fabius faction when he had been Premier secrétaire fédéral of the Aude department for only two years.

62

“Sure, they’re intimidating, the first BN meetings… It’s hard to make things out, to analyze the political situation, to unpack what’s being said, what’s going to become an official statement or not, what is said between the lines, the rifts. Actually in 1993, I was there as an observer: I told myself, ‘Here I am, lucky enough to see how they play in the premier league’. It wasn’t my league, I didn’t have the skills required to play in that league but I was lucky enough to sit next to those who could… Also, nobody tells you how things go. Nobody welcomes the newcomers, there’s none of that. There’s a ritual and regardless of who the Premier secrétaire is, the BN meetings always start and end in the same way. I’ve seen Mauroy, Fabius, Rocard, Emmanuelli and Hollande. It’s like politics in general: you get thrown into the pool, and if you’re good, you’ll manage, but if you can’t manage you’ll drown. It’s natural selection, things are tough and there’s no love lost.” [62]

63There are no rites of passage for joining the “premier league”. Behaviors conform only as a result of the ritualization of the official body, which works in an unchanging, de-personalized manner (neither historical circumstances nor the personal styles of the successive Premiers secrétaires have an impact). Only the long-term regulars can understand what is really at stake behind the debates (power struggles, positions resulting from former quarrels or strategic imperatives created by the situation itself).

64By contrast, the pre-existing ties between members who enjoy a higher status and join the governing body at a more advanced stage in their career undoubtedly facilitate the learning process for them. The socialization that occurs in these meetings forms a continuum with other collective experiences (ministers’ offices, parliamentary groups and committees, faction meetings…); as a result, spelling out the rules to them appears unnecessary. This is what Claude Bartolone explains – he was already the number two in the Fabius faction, had been a mayor for five years and an MP for nine years when he became a BN member in 1990:

65

“My first BN meeting wasn’t really overwhelming, because all those people around the table, you’ve already had the opportunity to work with them on very varied occasions”. [63]

66The process of learning the codes is therefore not so much a “natural selection”, as Pérez puts it, as a long acculturation to specific standards. This acculturation is made more or less easy by experience of other such collective settings; those with no experience tend to be initially passive, worried about making a misstep. This ritualization, which affects all members, ultimately glosses over differences in status (those who have the longest experience of the governing bodies are not necessarily the most important figures): what counts instead is how far you understand the rules of the game. Conversely, those who break the rules, by upsetting the agenda, threatening pre-established solidarities, failing to acknowledge differences in status (for instance, answering back to the Premier secrétaire), or making explicit traditionally implicit power struggles, run the risk of being stigmatized, and ultimately of losing their status. This is another manifestation of the closed nature of this polycracy of peers.

The elusive locus of decision-making

68The search for consensus has several consequences: the diffuse character of power makes it difficult to pinpoint the origin of a decision, not only within the governing body but also within the complex, stratified organizational chart of the PS. The collegial authority is characterized by what Noble and Pym call the “receding locus of power”. [65] Members sometimes acutely experience this feeling of not being able to situate themselves within the decision-making chain; they have the impression that the real decision is always taken elsewhere, as Bartolone explains:

“That’s where I realized there really wasn’t a definite locus of power, too. Within the group, you felt like that was the Comité directeur [former name of the Conseil national], when you were in the Comité directeur you thought it was in the government, in the government they felt it was at the Élysée, and at the Élysée they assumed it was in Mitterrand’s office. So that was a blow to my Cartesian conception of decision-making. But it also opened new horizons: since there wasn’t a single locus, it was clear you had to act to manage to get your ideas across in a pedagogic manner, and that there was more leeway that you might imagine.” [66]

Negotiation and/or deliberation

69Theoretical studies of collegiality do not systematically link collegiality with a specific type of argumentation. Indeed, one would have a hard time concluding that collegial bodies are more negotiation-oriented or deliberation-oriented. Initially, we might assume that collegiality and deliberation would tend to go hand in hand. According to Elster, [68] the latter is defined as a situation in which all actors attempt to persuade one other (since the goal is that all participants end up sharing the same opinion); furthermore, the participants are supposed to share common interests (and we can assume that the members of the governing bodies’ interests mostly converge with those collectively defended by the party). However, Elster goes on to point out that deliberative situations are fragile and rare: it is frequently the case that a deliberation consisting of a pure exchange of arguments gets polluted as internal issues (and interests) that do not relate to the discussion at hand are brought up. Deliberation is therefore always liable to turn into negotiation, where diverging interests square up – particularly when limited resources are at stake – and often results in a zero sum game (if one wins, the other loses).

70As in other collegial assemblies, [69] the practice of persuasion between peers in the PS constantly swings back and forth between negotiation and deliberation. While we can find virtually pure examples of negotiation (such as commissions d’investiture, where candidates are appointed) and deliberation (in the absence of short-term political stakes, for instance when assessing the efficiency of a public policy), making out shifts from one to the other tends to be difficult. The object of the decision and the nature of the different bodies in which it is discussed, depending on whether they are in charge of elaborating the decision or enforcing it, only influence the balance between negotiation and deliberation, especially insofar as collegial organizations, particularly in political parties, cannot be considered as Taylorist operations endowed with standardized procedures. They are based on flexibility, which is supposed to allow them to constantly adjust their resources to the objectives set.

71The commission des résolutions that meets on the final night of each PS Congress – a key moment in party life, since it determines for a period of three years both the party’s ideological line and the sharing of resources between the factions (depending on whether their leaders have agreed to “synthesize” or not) – is an extreme case, which magnifies the stakes of the discussions between socialist leaders. Being a part of this crucial closed session moreover means that one has gained one’s credentials in the party (even without the status that befits such a prestigious honor, as the committee is increasingly open to low-level members, experts, permanent members of staff or assistants likely to best negotiate the interests of their faction).

72As its meetings take place after the members have voted, this committee has no statutory existence. Although it is crucial to the functioning of the party, it is the result of a jurisprudential practice whose objectives have varied over time: the attribution of positions in the Secrétariat national, for instance, which took up nearly the entire running time of the Rennes Congress committee, is today no longer discussed in that body, currently devoted to debating the texts. [70] In the end, the duration and the form of the debates depend on the political context and on the balance of power established by the members’ vote. If the vote was close, or competing motions had a roughly similar weight, or a presidential nomination is imminent, then reaching an agreement on a composite motion becomes a more difficult prospect. Yet such an agreement is the official objective of all the socialist congresses; it manifests the unity of the party and allows minority members to join the leadership, both at national and federal level.

73Consider the example of the Le Mans Congress committee, which I attended in November 2005, a few months after the referendum on the Treaty of Rome, and a year before Ségolène Royal’s nomination. The debates were essentially negotiation-oriented: the discussion of the texts was guided by strategic considerations, because the balance of power was already established and opinions were already formed. Interactions were thus based both on solidarity (the agreement was going to have to satisfy the different factions and maximize chances of electoral success in the medium- and long term) and on competition, since limited resources needed to be shared: “Associate-rivals recognize one another in that one’s victory can only be achieved at the expense of the others.” [71]

74At the same time, only deliberation on the text of the composite motion was going to bring the negotiation to a satisfactory end. The final text went beyond the mere addition of competing texts: it introduced “amendments” to the majority motions as well as new formulations intended to reassure all the participants.

75The weaving together of negotiation and deliberation is particularly significant when amendments perceived as markers of the factions’ ideological and programmatic identity are discussed. In Le Mans, this was for instance the case of the debate on introducing the phrase “Sixth Republic”, defended by the NPS faction (represented in the committee by Arnaud Montebourg and Vincent Peillon), into the composite motion, against the majority’s wishes.

76

– François Hollande “There’s no need to debate this any further. We’re not in favor of a prime-ministerial regime. I respect Arnaud [Montebourg]’s position, he’s been defending it in the party for many years, but if the amendment is drafted that way, ‘Sixth Republic’, ‘prime-ministerial’, we cannot accept that.”
– Bernard Roman [MP, majority motion, president of the Parliament’s legal committee 1997-2002]: “That’s a contentious point in our discussion, even though we’ve all made a lot of progress on the subject. What we’re pushing for is a strengthened parliamentary regime. NPS suggests the Sixth Republic terminology, we suggest the phrase ‘A new Republic’. In our text, we can include the rebalancing of the executive/legislative poles.”
– Laurent Fabius: “That’s the problem with motion 1, it never says at any point whether the regime should be a parliamentary one or a presidential one.”
– Hollande: “I agree, the hybrid nature of the Fifth Republic requires a clarification; we want to go towards more parliamentarism, and make the President of the Republic more accountable.”
– Vincent Peillon: “This is a fine debate, and we all care very much about it. But once again, we don’t want to choose, so we’re sticking to a half-hearted compromise, and everyone knows what that’s going to cost us. So I suggest to you that Bernard [Roman] and Arnaud [Montebourg] come up with common wording.”
– Hollande: “Fine, but changing the Republic in the commission des résolutions is out of the question.”
– Montebourg [very disappointed]: “Okay, but then Thierry Mandon [mayor of Ris-Orangis, close ally of Montebourg] goes in the group, I don’t.” [72]

77Here, the failure of the deliberation (leading to the adoption of the intentionally vague “new Republic” formulation) results from its over-determination by strategic issues. The fact remains that the committee ritual forces the protagonists to exchange viewpoints and come up with wording that is admittedly consensual, without conveying the socialist message to the public very clearly, but that ultimately allow all the factions to defend the same political position. In other cases, deliberation focuses exclusively on form: the protagonists agree on content, but using an expression formulated by a given faction amounts to acknowledging its symbolic victory. This is illustrated by the debate on the phrase “frontal opposition” (to the government), which is identified by all the actors with the Fabius faction.

78

– Hollande: “We’ve been in opposition for three and a half years, and I don’t think we should get drawn into describing this opposition. The opposition’s role isn’t to act, anyway.”
– Fabius: “All the same, we’d like the phrase ‘frontal opposition’ to be introduced.”
– Hollande: “We’re okay with ‘total’ or ‘global’ opposition, but there’s no reason to acknowledge we’ve failed as an opposition.”
– Fabius: “Lots of people have been using this phrase.”
– Hollande: “Yes, but since we’re attempting a composite motion, we have to come up with solutions, words that everyone can accept.”
– Jean Glavany: “This is an important semantic issue. I agree on the content, but using such words seems judgmental and vexing, and that’s how it’s perceived in the sections. As if we could be in favor of an alliance with the right. To reach a synthesis, we need to trust each other.”

79The phrase “strong and credible opposition” was ultimately chosen. As Jean-Marc Ayrault pointed out during a recess of a meeting of the majority motion, “if we accept Fabius’s amendments, it boils down to saying that he won the Congress’. This example shows the extent to which deliberation and negotiation go hand in hand, and also demonstrates that collective action between peers is characterized by debates on rules and values. [73] During the entire Congress and regardless of the context of the meetings, there were constant discussions on the topics legitimately up for debate, the ways to conduct these debates (what type of conciliation is acceptable? which attitude should be adopted?), and which objectives to pursue (coherence of the majority, public image of the party). The group of party leaders can be likened to a Russian doll structure: it is made up of embedded individuals and groups that struggle to make their own value system prevail.

80* * *

81Promoting a new, updated approach to political parties, which have already been the focus of landmark studies in political science, requires us to draw from other types of literature. Here, I have attempted to show the benefits of using the concept of collegiality, refined and systematically formalized by sociologists of professions, to study party leaders, in order to go beyond the tautological definition of party elites based on their superior hierarchical status or resources. I am not arguing that such a shift doesn’t come with any losses, or that the concept is sufficient to account for the entire reality of these party leaders. For instance, it does not allow us to study the exchanges between the party leadership and external social groups (voters, grassroots party members, the media, interest groups, etc.). It does not encompass the entire range of the leaders’ activities or of the decision-making process. Yet, it is useful to the understanding of all the operations through which the leaders share party resources and produce collective goods such as public stances and political programs. It is precisely on that level that collegiality has a greater impact, given that its constraining effects are strongest here (because exemplarity is required towards the lower levels of the party and the public), and it is harder to achieve (internal conflicts of interest tend to flare up during the process of coming up with the programs). This approach in terms of collegiality, more than those using the more general concepts of “role” and “institution”, allows us to precisely characterize the tension at work between equality and differentiation between peers, and to relate cultural constraints to the exercise of power, by calling attention to the devices produced at meso-level by the organization itself.

82This concept can be extended to all political organizations; although collegiality exists in relatively binding forms in parties that openly defend internal democracy and pluralism, we can assume that it also exists in any organization where bodies are characterized by a lack of formal hierarchy. The co-existence of egalitarian standards and of the inequality of resources, the self-regulation that entails constant debate within the organization concerning the rules and the reciprocal monitoring between peers, the collective, consensus-based decision-making process, should all be observable, on a greater or smaller scale, and with more or less notable consequences on party outputs, both on the left and on the right of the political spectrum.

83Lastly, collegiality implies the presence of associate-rivals. In this sense, it sheds light on the PS’s resistance to centrifugal tendencies, beyond the sharing of common interests. The PS is distinctive in that it formally regulates the interdependence of the actors and turns it into a normative constraint that is essential to the culture of the party. Admittedly, using this notion leads us to emphasize modes of conflict resolution rather than the conflicts themselves, addressed only insofar as they underlie the inseparable negotiation and deliberation which characterizes internal exchanges in the PS. Finally, collegiality opens up interesting avenues regarding the study of party leader as an occupation. One of its key dimensions is the ability to plead and negotiate against and with one’s associate-rivals, not so much for one’s own benefit, but rather for one’s relevant sub-group (federation, faction).

84Moving beyond the iron law of oligarchy requires us to take into account the differentiations and similarities fostered by the cultural constraints that apply to reciprocal relationships between “peers”. Taking this horizontal perspective into account can only be beneficial to our understanding of the density of partisan phenomena.