The actors of economic policy decision-making

9In the wake of the Socialist Party’s election victory in May 1981, the winds of change seemed to be blowing in the ministerial cabinets, with the rise to power of the “intellectual middle classes and the business world” [5] (Dagnaud and Mehl 1982). However, this did little to stem the tide of “technocratism”: 65% of the socialist government—Dagnaud and Mehl’s “rose elite”—still came from the ranks of France’s administrative elite. Pierre Birnbaum (1985) also notes that, despite the increased presence of representatives from the liberal professions and the trade union and non-profit worlds, the proportion of senior civil servants and members of the grands corps at the highest levels of government remained stable. Florence Descamps (2018, 38) confirms this, noting that between 1981 and 1984, senior civil servants from the Ministry of Finance remained well represented in the most important cabinets: those of the president, the prime minister, the Ministry of Finance, and the Budget Ministry. This was an underlying trend: Frédéric Sawicki and Pierre Mathiot (1999) reach the same conclusions for the 1981–1993 period. They note that, far from a sea change, there was instead a partial replacement of the traditionally dominant section of the senior administration—the grands corps—by a previously dominated section—civil administrators (ibid., 4). [6]

10These studies on social recruitment for ministerial cabinets help explain how some of the state’s broad political orientations persisted no matter who was in power. However, in the particular case of the “turn,” they do not explain exactly how the regulatory process that resulted in a gradual redefinition of the rules of economic policy actually unfolded. We thus found it necessary, as Sawicki and Mathiot suggest, to reconstruct “the space of the deciders” (ibid., 7) associated with this specific decision. We therefore carried out a targeted investigation, taking inspiration from Pierre Bourdieu’s (2005 [2000]) study of housing policy reform in the 1970s. Bourdieu begins with the principle that “to understand ‘state policy’ [..] it would also be necessary to know the state of opinion of the mobilized, organized fraction of the ‘opinion-makers’” (2005 [2000], 92). By proceeding by “trial and error” and using criteria “of reputation” as well as “institutional” criteria, he reconstructs a group of a hundred or so individuals, including senior civil servants, politicians, members of ministerial directorates or the parapublic sector, and private-sector representatives (ibid., 99). We did the same thing for the case of the “turn.”

11To reconstitute this decisional space, we made a list of every person cited in our corpus of sources (Insert 1) for whom there was at least some indication that they had participated in the regulatory process (i.e., at least one position expressed in the controversy). This produced a list of about thirty people. While drawing up the list, we encountered the phenomenon of saturation, where new sources no longer offered new names to add to the list (Eloire et al. 2011), so we concluded that we had made a complete inventory of the main players in the controversy. Besides President Mitterrand, this list includes the main members of the government who dealt with economic issues: Pierre Mauroy, the prime minister; Jacques Delors, minister of the economy and finance; Laurent Fabius, minister for the budget; Michel Rocard, minister of territorial development; Jean-Pierre Chevènement, minister of industry; Gaston Defferre, minister of the interior and decentralization; and Pierre Bérégovoy, secretary general of the Élysée and later minister of social affairs.

12In addition to these leading political players, the list also includes certain members of their entourage—cabinet directors and economic advisors. For the Élysée, this includes Jacques Attali, François-Xavier Stasse, Alain Boublil, Jean-Louis Bianco, Élisabeth Guigou, Christian Sautter, Hervé Hannoun, André Rousselet, and Charles Salzmann; for the prime minister’s office, Jean Peyrelevade, Henri Guillaume, Robert Lion, and Bernard Brunhes; and for the Finance Ministry, Philippe Lagayette, Jérôme Vignon, and Pascal Lamy. Note that no names appear from the entourages of Fabius, [7] Rocard, Chevènement, Bérégovoy, or Defferre, whereas those of Delors, Mauroy, and Mitterrand are very active.

13We also include important senior civil servants: for the Treasury, its director Jean-Yves Haberer, his successor Michel Camdessus, and his deputy Daniel Lebègue; for the Bank of France, its governor Renaud de la Genière. Finally, we listed various figures who were involved informally (with no official connection to the state). For Mitterrand, this includes editorialist and press baron Jean-Jacques Servan-Schreiber, Schlumberger CEO Jean Riboud, and Christian Goux, an economist and Socialist Party elected official. For Mauroy, this includes Crédit Lyonnais president Jean Deflassieux and the economists Pierre Uri and Jean Denizet.

14The list presents certain similarities to Bourdieu’s. That said, it is shorter (34 names vs. Bourdieu’s 97), which might seem surprising: Could a decision as important as the “turn toward austerity” really have been made by so few people? This is at least what our study indicates, and it can be explained by several reasons. First, the controversy in question was largely “underground” according to the journalist Serge July (1986, 82), who only broke the story in Libération on March 14, 1983. This discrete character fits with the limited number of participants. Second, the regulatory process of the “turn” was of a specific nature: it was much more informal and even collegial (Lazega 2001) than the bureaucratic, formalized nature of the reform studied by Bourdieu, which was the subject of an official commission in 1974–1975 (2005 [2000], 104), or the 1981–1982 bank nationalization reform, which was the subject of weekly interministerial meetings (Morin 2020, 63). We can also compare our list with Descamps’s (2018), which features 68 individuals “involved in the development of economic and financial policy decisions” from 1981 to 1984. His list includes more names, but she uses a different selection method: she begins by drawing up a list of state entities—the Élysée, the prime minister’s office, the ministries of the Economy, the Budget, Research, Industry, Social Affairs—and then includes every advisor connected with these entities. This reflects a nominalist strategy (Laumann et al. 1983), whereas we combined the nominalist and realist strategies, [8] which may limit the number of actors retained according to the nature of the controversy under examination.

15The absence of representatives of certain institutional groups in our population—such as the political parties within the executive (the Socialist Party [PS] and the Communist Party [PC]), the unions close to the left (in particular the French Democratic Confederation of Labor [CFDT] and the General Confederation of Labor [CGT]), and the National Council of French Employers (CNPF)—could merit further attention. We therefore put our list to the test by expanding the scope of our inquiry and looking for potential new actors. For the PS, we consulted the book of Lionel Jospin, who was the party’s first secretary at the time. For the PC, we consulted the books of three government ministers from the party: Anicet Le Pors, Marcel Rigout, and Charles Fiterman. [9] We consulted the books of Henri Krasucki for the CGT, Edmond Maire for the CFDT, and Yvon Gattaz for the CNPF (Appendix 2). This second phase of our inquiry did not lead us to add any names to our initial list, but rather it reaffirmed our demarcation of the decisional space of the “turn.”

16The borders of our study population reveal the existence of a “hard core” of decision-makers. We did not seek to reconstruct the entire “nebula” of other interactions these actors might have had with other individuals on other subjects. The protagonists of the “turn” were completely immersed in the political and ideological context of their time, and they were indeed major actors in it.

Establishing the choice between stimulus and austerity

17The socialist government’s economic decisions [10] and their macroeconomic consequences have already been studied at length (Ross et al. 1987; Halimi 2000; Bonnaz and Paquier 1993; Asselain 2001). The government’s shift away from its initial program happened gradually (Insert 2). Although in May 1981 the policy of stimulus was the government’s unquestionable horizon, the question of austerity later emerged as a necessary alternative (Asselain 2001; Margairaz 2001). The actors were thus confronted with a contradiction between their desire to pursue their original goals—increasing purchasing power, full employment, and developing public services—and the need to respond to certain difficulties encountered in high office, such as the increase in public spending due to the stimulus policy and high inflation (13.4% in 1981 and 11.8% in 1982, according to INSEE). As early as 1982, the executive was also concerned by the mounting unemployment rate, which the stimulus policy was failing to bring down (the rate climbed from 6% in 1981 to 6.6% in 1982). In addition, certain advisors were particularly worried about the decline of two economic indicators: the balance of trade (which fell from -0.8% of GDP in 1981 to -2.1% in 1982) and foreign exchange reserves, which were constantly the subject of “confidential” memos. [11] The reason these advisors found these indicators so concerning is that they were worried that their decline would lead to speculation against the franc on the financial markets, and that the Bank of France would not be able to contain it. [12]

18The questions posed by the monetary and budgetary context thus grew more pressing for the government and its advisors as the controversy developed. Two contrasting positions on budgetary issues were already being staked out in the first days of Mitterrand’s seven-year term. On one side, “those closest to the president [and] the prime minister were convinced that change must be achieved through intense economic stimulus” (Bauchard 1986, 59). On the other side, according to presidential advisor François-Xavier Stasse, “the Rocardians [such as himself] […] warned prior to 1981 that the economic program […] risked engendering major economic and financial imbalances” (Grunberg 2006, 2). On the latter side was Jacques Delors, whose “obsession was with reducing budgetary expenditure” (Attali 1993, 99), and who, as early as November 1981, “raised the specter of ‘the pause’ [in state spending]” (Favier and Martin-Roland 1990, 409).

19A division emerged on monetary issues too. On one side were advocates of restrictive management of the franc, such as the governor of the Bank of France, Renaud de la Genière, who was prepared to resign because he was “preoccupied by budgetary slippage and the threat to the stability of the French economy posed by the stimulus policy” (Duchaussoy 2011, 142), but who noted with astonishment that the new executive did not intend to devalue the franc in May 1981 (Pfister 1985, 246; Mauroy 2003, 172). On the other side were those who supported the “principal of strong devaluation accompanied by withdrawal from the EMS,” such as Jean-Pierre Chevènement, who believed that “staying in the EMS while instituting even moderate stimulus” would be an error (Chevènement 1985, 47 and 270.)

20Studying the positions taken by the actors at different stages of the controversy gives some idea of the comparative weight of budgetary and monetary themes, and of the power relations between partisans of each camp (Table 1). Despite its imperfections, [14] we believe that this data serves as a good indication of how the controversy unfolded. While budgetary issues appear to have been evoked relatively constantly throughout the controversy, monetary questions were raised more often toward the end of the controversy than at the beginning. There are also marked differences regarding the power relations between camps within each theme. With respect to budgetary issues, the number of stimulus advocates decreases over time, while the number of austerity advocates increases. Similarly, with respect to monetary issues, the number of advocates for successive devaluations (within the EMS) increases, as does the number of advocates and opponents of withdrawal from the EMS. It was around this last theme that the conflict finally crystallized.

21By combining all the information on individual position-taking by theme, it is possible to determine each actor’s position, at each phase, regarding the choice between stimulus and austerity. Stimulus advocates are those who supported both an expansionist budgetary policy and significant devaluation that could lead to withdrawal from the EMS, while austerity advocates are those who favored both decreases in budgetary spending and, if necessary, EMS-supervised devaluations. While the former decreased in number, the latter increased (Table 2), and two camps gradually emerged over time. This evolution in positions and in power relations within the decisional space likely had an impact on Mitterrand’s final decision in favor of austerity. [16] Furthermore, there was the impression that the advocates for stimulus were not as unified as the austerity camp. With regard to the former, presidential advisor Charles Salzmann writes: “they were not all pursuing the same objective” (1996, 88). While some, such as Jean Riboud and Jean-Jacques Servan-Schreiber, were seeking above all to liberate French businesses from European constraints, others, such as Pierre Bérégovoy, Laurent Fabius, and Jean-Pierre Chevènement, were more concerned with loosening the monetary constraints imposed on the franc within the EMS. But how should this evolution in opinions be understood? We begin to address this question by studying the field of the controversy.

Two efficient forms of capital: Statist and political

23As we have discussed, the actors who made up the decisional space belonged to various fields: political, bureaucratic, academic, or managerial. However, through a multiple correspondence analysis (MCA), we can reconstitute the structure of this space. To do so, we rely on individual sociodemographic data, in the belief that position-taking on economic matters is linked to sociological variables, including the generation actors belong to, their academic and professional career, and their political engagement (Table 3). [17]

24With respect to age distribution, the population can be divided into three nearly equal parts, with the under-40 group slightly overrepresented compared to the 40–54 and 55-and-older groups. As for higher education, 68% attended a grande école (École nationale d’administration [ENA], École Polytechnique, Sciences Po), and 40% held a university degree in economics, while 29% held a law or humanities degree. With respect to professional careers, 59% were senior civil servants in a grand corps, whether administrative [18] or technical [19]; others belonged (or had belonged) to industry, banking, teaching, or the press. As for political engagement, 44% were registered members of the Socialist Party, [20] but this proportion rises to 75% if we include more informal connections to the party, such as belonging to expertise networks (Fulla 2016). Seven of them favored Mitterrand’s motion at the Metz Congress in 1979. [21] Nearly half had authored one or more works on economic issues, and a third on political issues. [22]

25The MCA includes 11 active variables corresponding to 26 modalities (Table 3), projected onto a factorial design (Figure 1). The first (horizontal) axis delivers 23% of the total information. It defines actors according to academic and professional career (triangles) and age (circles). On the right of the diagram [23] are younger individuals who attended grandes écoles (ENA, Sciences Po), belonged to the grands corps (particularly administrative), and worked in the upper reaches of the civil service. On the left are older actors who tend to be defined negatively—that is, as not belonging to the grands corps, not having attended a grande école, and not having pursued a career as a senior civil servant. We argue that this axis can be interpreted as the expression of a first form of capital, which we call statist capital, because it is produced by and within the state through participation in its most central institutions: grandes écoles, grands corps, and the senior civil service. [24]

26The second (vertical) axis delivers 16% of the total information. It identifies individuals according to political engagement (squares) and, in particular, proximity to the Socialist Party. In the upper portion of the diagram [26] are those actors who had an electoral mandate and/or function in the Socialist Party, who participated in the 1979 Metz Congress and the 1981 presidential campaign, and who had authored works on politics—some also belonged to the administrative grands corps. In the lower portion are senior civil servants in the technical grands corps, who are defined negatively by the weakness of their connections to the political field (unaffiliated with the Socialist Party, no electoral mandate). We argue that this axis can be read as a second form of capital within the decisional space of the “turn,” which we call political capital, since it defines individuals according to their degree of belonging to a political party (the Socialist Party in this case).

27Within the decisional space, thus structured by two forms of capital—statist and political—we can identify four subgroups of actors (Figure 2). In the bottom-right corner are actors with both statist capital (on axis 1) and technical skills (on axis 2). In the top right are actors with administrative statist capital (on axis 1) and political capital (on axis 2). The top-left corner contains actors with more extensive professional and political experience (on axis 1) and political capital (on axis 2) accumulated over the long term, which distinguishes them from those in the top right. Lastly, the bottom-left corner comprises a heterogenous group of economic experts from universities or the private sector, who are less closely connected to the political and statist fields. Beyond these two forms of capital, the MCA suggests the existence of three competing forms of status in economic policy decision-making: political experience status, connected to a dominant and long-term position in the political field; political-statist status, tied to the simultaneous but more recent possession of both forms of capital [27]; and finally statist-technical status, which is linked to participation in the primary institutions of the statist field (grandes écoles, grands corps, senior civil service).

28It is possible to connect the structure of the decisional space to the positions taken by the actors in May 1981 (Figure 3). Stimulus advocates (underlined) tend to be on the side of political capital, whereas austerity advocates (circled) are more often on the side of statist capital. This opposition reflects three differences: 1) a difference in academic and professional socialization: on the left of the diagram are those with political and activist training and careers; on the right are those from the grandes écoles and grands corps; 2) a generational difference: those on the left tend to be older, while those on the right tend to be younger; and 3) a difference in status: on the left are political experience status and political-statist status, characterized by temporally dominant positions within the executive (president, minister, secretary general); on the right is statist-technical status, characterized by temporally dominated positions (advisory and expertise).

29Hence, at this stage in the regulatory process, the socialist leaders were faithful to the policy agenda promoted throughout the presidential campaign. They justified their position either on the level of conviction—“no one contested the over-20% increase in public spending contained in the budgetary documents drafted by Fabius; only Delors expressed reservations and made clear his misgivings” (Pfister 1985, 245)—or in terms of duty to the electorate: “those closest to the president [and] the prime minister were convinced that change must be achieved through intense economic stimulus” (Bauchard 1986, 59), and “none of the foremost ministers of the era, including Rocard, regret these initial decisions” (Favier and Martin-Roland 1990, 113).

Figure 1

The decisional space of the “turn”: The cloud of modalities

The decisional space of the “turn”: The cloud of modalities

30However, certain ministerial cabinet members and senior civil servants did not engage in this kind of discourse. For example, Jean Peyrelevade writes that “right before the first devaluation in October 1981, which failed, certain members of the prime minister’s cabinet sounded the alarm. One day, in his office, Mauroy asked us in his typical fashion: ‘So kids, where are we headed?’ We replied: ‘Right off the cliff’” (Bouchet-Petersen 2013). In fact, in late 1981, “[Peyrelevade] was, along with Lagayette and Stasse, one of the first to sound the alarm and preach austerity” (Baume 1993, 51). The tone was the same at the Bank of France, which “was hostile to this radical paradigm shift. It instead fought for increased respect for equilibrium [and] the struggle against inflation […] The Bank advocated a far-reaching and restrictive policy” (Duchaussoy 2011, 74). Its governor, “in his letters, exhorted the finance minister Jacques Delors to take a softer line on current policy. The Bank feared that it might prove dangerous and eventually threaten the goal of reducing inflation that had been institutionalized for the past five years” (Duchaussoy 2010, 4).

Figure 2

The decisional space of the “turn”: The cloud of individuals

The decisional space of the “turn”: The cloud of individuals

31Although austerity was initially a temporally dominated, minority position that was considered illegitimate, its supporters gradually grew in number. It then became legitimate and dominant, and was eventually instituted by Mitterrand in March 1983. We will now discuss an underlying social mechanism that enabled its spread.

Figure 3

State of positions taken in the controversy: May 1981

State of positions taken in the controversy: May 1981

Status competition

32A comparison of the state of positions taken in 1981 and in 1982 (Figure 4a) highlights a social mechanism that is important for the regulatory process: “social homophily” (Eloire 2014), which holds that those who are similar to one another will come to share the same views. The austerity advocates tend to be grouped at the bottom of the diagram, which is where this position originated. The analysis shows that more and more of the actors we identified as undecided in 1981 became pro-austerity, and that these new recruits were sociologically close (in terms of their position within the decisional space) to those identified as being the first to advocate a policy of austerity—hence the idea of a social homophily effect.

33This mechanism led to other notable changes of opinion from June 1982, specifically those of the prime minister Pierre Mauroy and the minister of the economy Jacques Delors (upper-left quadrant), who at that point became ardent supporters of austerity. Even though, with Mitterrand’s election victory, stimulus had won at the polls, the reality of certain economic difficulties ended up legitimizing a different opinion in these men’s minds. The fact that Mauroy and Delors had already voiced their doubts about Mitterrand’s economic program even before the campaign, for example at the 1979 Metz Congress, facilitated the about-face.

34Delors was the first to express his misgivings. He tried in vain “to convince Mauroy and Mitterrand of the urgency of the situation”: “I suggested a pause in the announcement of the reforms. Despite devaluation, the situation remained uncertain on the foreign exchange market […]. I deliberately employed the shock word ‘pause’ to move public opinion” (Duchaussoy 2011, 80–81). At first, Delors was on his own: Mauroy initially condemned his opinions, before converting to them in early 1982 (Bauchard 1986, 59). In June 1982, a first austerity plan was put in place, but Mitterrand continued to hesitate between stimulus and austerity. But the social homophily mechanism continued to work its effects, and by March 1983 the number of austerity advocates had further increased (Figure 4b): the stimulus advocates now occupied a very limited place in the decisional space. From 1981–1983, we count 19 changes of opinion, versus 15 unchanged opinions. 5 of the changes were from austerity to stimulus, while 14 were from stimulus to austerity.

Figure 4

State of positions taken in the controversy

State of positions taken in the controversy

35The analysis demonstrates the importance of competition between different forms of status in the regulatory process. These statuses were hierarchized: decision-making positions within the government were occupied by those who possessed political capital, while those with statist capital tended to occupy advisory and expertise roles. Paradoxically, though, the opinion of those with statist-technical status ended up winning out in the controversy. This opinion was first adopted by two actors with political experience status (Mauroy and Delors), but who, holding the same type of status as the president, were unable to influence him. It was in fact the change of opinion on the part of certain actors who possessed the non-congruent status, political-statist status (Attali, Fabius, and Rocard), that appears to have been decisive in the final phase of the controversy. In order to better understand this other aspect of the regulatory process, we will now discuss the relational infrastructure of the decisional space.

The structure and evolution of the collaboration network

37Jacques Attali describes the first days at the Élysée as follows: “New men without the slightest idea of the inner workings of the state moved into the Élysée. Four teams organized around the president […] including mine, which was responsible for international summits and ‘having ideas.’ […] With me, four collaborators: Jean-Louis Bianco, Pierre Morel, Ségolène Royal, and François Hollande” (Attali 1993, 25). Among the names cited, Bianco was “his childhood friend, as well as a graduate of the ENA and a conseiller d’état like himself,” and Hollande and Royal were “two young senior civil servants who had graduated from the ENA a year earlier” whom he “took under his wing” (Auffret 2009, 73 and 92). At the prime minister’s office, “initially, Mauroy had been hoping to have his banker friend Deflassieux by his side. The latter had declined the invitation and suggested Peyrelevade to head the cabinet. But the young École Polytechnique graduate wasn’t Mitterrandian enough to be entrusted with that much responsibility. Hence the choice of Lion” (Ravaï 1997, 25). As for the Finance Ministry, Delors explains: “Lagayette immediately set about on drawing up his cabinet and consulted me. [We] were connected by a deep friendship. He was a member of Échanges et Projets. [28] […] The assistant directors also played an important role […]. Pascal Lamy, foremost among them, thought it best to later join Mauroy’s team. I want to underscore how solid and united the group was, which counted for a lot during the difficult decisions of March 1983” (Delors 2004, 129–30).

38As these examples illustrate, the respective entourages of the members of the executive (Sawicki and Mathiot 1999; Eymeri-Douzans and Bioy 2015) were founded on the basis of recommendations and past collaboration links. In 1981, in fact, 53% of members of the Socialist Party’s economic committee were members of ministerial cabinets (Fulla 2016, 784). The notions of “team” and personal or professional “friendship” (Lazega 2001; Eloire 2010; Penalva-Icher 2010) were at the forefront in the makeups of the cabinets, which tended to be of limited size. For example, according to the 1981 Bottin administratif, the president’s and the prime minister’s cabinets each counted 25 members, the minister of the economy’s comprised ten, and the budget minister’s had just eight. Working in the cabinets was an intense, collective experience—sharing ideas, coming up with opinions on a wide range of subjects, and rapidly producing memos, analyses, assessments, speeches, and so on. The archives we consulted bear the traces of this work: advisors communicated with each other regularly, sending each other documents that they annotated and sent back. Since each cabinet was attached to “its” ministry, it was potentially in competition with other cabinets in the case of disagreement. [29] But each cabinet could also cooperate with the others depending on the circumstances. The cabinets thus formed a relational infrastructure, the systematic study of which is necessary to understanding the regulatory process.

39In order to recreate the collaboration links within the decisional space, we made use, first and foremost, of data from the Bottin administratif [30] to construct a network of co-membership links (Finez and Comet 2011; Pette and Eloire 2016), on the basis that two actors are linked if they belonged to the same entity at the same time (Élysée, prime minister’s office, Finance Ministry, etc.). We set the following rules: there is a link between two actors if they both belong 1) to the same entity, for cabinet members, 2) to the government, for links between ministers and the prime minister, 3) to the circle of those close to Mitterrand, for links between the president and ministers. From a methodological perspective, we thus obtained a bimodal network (individuals–entities or roles), which we then transformed into an undirected unimodal network (individuals–individuals). We carried out this operation three times, for the May 1981, June 1982, and March 1983 phases, taking into account the relational changes from one period to another. Certain actors entered the controversy gradually, such as Jean Riboud, whom the president began to consult on economic issues in 1982, even though the two men had known each other since 1971 (Boublil 1990, 22 and 36). Other actors left the controversy, such as Daniel Lebègue and Robert Lion, who left the prime minister’s cabinet for the Treasury and the Deposits and Consignments Fund, respectively. Still others circulated between entities, such as Élisabeth Guigou, who was at the Treasury in 1981, then the Finance Ministry in 1982, and finally the Élysée in 1983.

40We also made use of information that is less systematic than that of the Bottin administratif, but that is just as valuable and important for understanding this regulatory process. This data came from qualitative sources, such as friendships, close relationships, and temporary alliances (Appendix 4). This reconstructive work was facilitated by the relatively small population size, by the abundance and precision of available sources and testimonies, and by preexisting research on the “turn.” A description of the structure of the three graphs shows that the number of links, as well as their density [31] and average degree, [32] increased from one phase to the next (Table 4). This clearly reflects the increase in relational activity within the decisional space, and therefore the reinforcement of the phenomenon of cooperation between competitors that is under examination here.

Cooperation between competitors

41The increase in the number of links between the first and second phases of the regulatory process can be explained by the phenomenon of cooperation between competitors. In late 1981, certain actors belonging to initially separate entities began to establish connections, as Jean Peyrelevade explains: “The first serious warning was the devaluation in 1981. It was a failure. The finance minister submitted a plan to us that was basically nonexistent. […] We had to […] make allies with Delors, with Stasse at the Élysée, and with Lagayette in Finance. That was done to counterbalance the influence of Attali, who was serving the president illusions” (Roussel 2015, 404). This strategy led, in June 1982, to an alliance between Mauroy and Delors: “‘Let’s team up,’ the prime minister said to his finance minister. Delors agreed. That evening, a decision was made to ‘force austerity through, and fast’” (Baume 2013, 58).

42The role of intermediary was thus played by Stasse for the Élysée, Peyrelevade for the prime minister’s office, and Lagayette for the Finance Ministry. Their interactions had an effect on the evolution of the structure of the collaboration network. “We were completely isolated,” explains Peyrelevade, “but Mauroy was brave: he imposed this policy with the help of Delors, Lagayette, and Stasse, our great supporter at the Élysée. We were opposed to consumption-driven stimulus” (Baume 2013, 60). This isolation, clearly visible in the 1981 network, in which the austerity advocates (black nodes) are few in number, dispersed, and on the periphery (Graph 1), diminished over time. In the 1982 network, the austerity advocates are more numerous, interconnected, and central thanks to the inter-entity links that had been established (Graph 2). This trend is confirmed in the 1983 network, where the austerity advocates form a dense relational block (Graph 3). Conversely, the stimulus advocates (gray nodes) are initially centralized, but little by little are overtaken by the other camp, and in the final phase of the regulatory process appear much less united.

Graph 1
Network of collaboration and position-taking in 1981

Graph 1
Network of collaboration and position-taking in 1981

Graph 2
Network of collaboration and position-taking in 1982

Graph 2
Network of collaboration and position-taking in 1982

Graph 3
Network of collaboration and position-taking in 1983

Graph 3
Network of collaboration and position-taking in 1983

Legend: The shapes represent functions: square = minister/president; circle = cabinet member; diamond = senior civil servant; triangle = outside the executive.
The colors represent opinions: gray = pro-stimulus; black = pro-austerity; white = unknown/no explicit opinion.

43Thus, the intermediacy mechanism initiated by the joint relational work of Stasse, Peyrelevade, and Lagayette not only increased the network’s overall activity by increasing its density (Table 4), but it also modified the nature of the cooperation links. The network, which was initially shaped by links internal to each entity, gradually opened up to links between the different entities. [33] This makes it possible to obtain the E-I index, [34] which shows that the network of actors involved in the “turn” was less and less structured by intra-entity links and increasingly by inter-entity links (Table 5). In 1981, the ratio is negative (there are more intra-entity than inter-entity links); in 1982, the same is true, but the ratio is less negative; and finally, in 1983, the ratio is positive (there are more inter-entity than intra-entity links).

44The phenomenon of cooperation between competitors thus depends on a first relational mechanism: intermediacy. However, the analysis of the evolution of the relational infrastructure of the regulatory process reveals a second mechanism: the construction of social niches. [35] The dense block of austerity advocates in the 1983 network can be interpreted as such a niche. [36] Indeed, the actors that made up the block occupied structurally equivalent positions (Lazega 1998) in the network and therefore played similar roles in the collective division of decision-making work. By defending austerity as a group, they gave it collective force within the decisional space. The relational work, which started out informal, became increasingly organized.

45Throughout the first phase (1981–1982), the work of developing pro-austerity arguments was thus carried out separately within each entity. Our sources discuss the creation within the presidential cabinet of a first subgroup, sometimes referred to as “the Élysée five” (Favier and Martin-Roland 1990, 484), which included Attali, Guigou, Sautter, Stasse, and Bianco. According to Attali, the group met “every day” starting in 1982: “The memos multiplied: How can imports be reduced without leaving the EMS? How can we devalue without worsening unemployment? How can we show that a floating exchange rate will send us into a vicious cycle?” [37] (Attali 1993, 325). An important element of the “turn” was “the solidity and solidarity of the Élysée five” (Duchaussoy 2011, 120).

46A second subgroup is also mentioned: the “group of ‘four’ of Matignon [the prime minister’s office] (Peyrelevade, Lebègue, Brunhes, Guillaume)” (Daniel 1988, 144). Guillaume writes: “From the start, our little group took a realistic view. We wanted austerity […]. With Peyrelevade and Lebègue, we drew up a secret plan. […] We were completely isolated” (Baume 2013, 60). But in the end, their work influenced the prime minister’s position. According to Peyrelevade, “Mauroy said to us: ‘Mitterrand had a good idea—study it for me.’ We told him every time that it didn’t hold water, and he’d yell at us to make us follow our line of argumentation to the end” (Favier and Martin-Roland 1990, 443). Guillaume writes that “little by little, Mauroy started listening to us. It was a very difficult period for us: we had to apply an economic policy that went against our own ideas, all while advocating for a change of direction” (Baume 2013, 52). [38]

47Contact between these subgroups was not immediately established, because “the Finance Ministry and the prime minister’s office believed that Attali and Bérégovoy were anti-austerity” (Favier and Martin-Roland 1990, 415). Mauroy sent a memo directly to Mitterrand in which he argued in favor of austerity, being careful “not to mention it to Attali or Bérégovoy out of fear that they might strike back” (Asselain 2001, 423). Attali confirms this: “In principle, neither I, nor Bérégovoy, nor Stasse were supposed to know about the memo, even though it was in keeping with our own analysis” (Attali 1993, 234). Rather than bringing together individuals, the intermediacy mechanism brought together previously established subgroups of advisors who were already convinced about austerity and who were set on preventing withdrawal from the EMS. [39]

48Faced with this pro-austerity front, the regulatory process was marked by the emergence of another subgroup. Attali explains that, in August 1982, “the president asked me to convene […] an ‘informal group.’ They would soon be referred to as the ‘evening visitors’” (Attali 1993, 299; Rimbert 2005; July 1986). This group, which included ministers Fabius, Chevènement, Defferre, and Bérégovoy, advisors Boublil, Rousselet, and Salzmann, and businessmen Riboud and Servan-Schreiber, met several times up until March 1983: “We met regularly […] to discuss these topics with Mitterrand in the evening, in the Élysée library” (Fabius 1995, 74). This select group sought to develop an “other policy,” an alternative program to austerity, which Mitterrand was “uncomfortable” with (Pfister 1985, 238). The group defended the idea of “floating the franc [editor’s note: withdrawal from the EMS] to enable a reduction in charges and to stimulate investment” (Attali 1993, 312–13).

49However, in contrast to the inter-cabinet social niche discussed above, the sources do not mention the existence of any collaboration links between these “evening visitors” outside of their meetings with the president. Rather than a social niche, this subgroup appears to have been more like an assembly of individuals, some of whom were influential and charismatic, such as Riboud, who is often presented as the leader. But the Riboud line, which focused on restoring the efficiency of French businesses, did not exactly correspond to the Socialist Party’s 1981 agenda: “The advocates of the ‘other policy’ were not all pursuing the same objective” (Lacouture 1998, 88). Mitterrand listened so attentively to the advocates of the “other policy” only because he himself was hostile to the constraints of the EMS: “We are not in control of our own policy. By remaining in this system, we are putting ourselves at the whims of the waves—and all to Germany’s benefit” (July 1986, 96). [40] Moreover, according to Guigou, the president harbored “a great distrust of financiers and technocrats. Attali, Bianco, Sautter, Stasse, and I represented that world. He knew how much we were meeting every day to refine our arguments. He said to himself: ‘They all studied at the same school, they’re all from the same background, so it’s normal that they think the same thing’” (Guigou 2000, 56).

50Despite all this, Mitterrand still ended up flipping to the side of austerity, but “it unfolded, according to Attali, in an atmosphere of violence and harassment on the part of those advocating withdrawal from the EMS” (Duchaussoy 2011, 121). On the one hand, “Mitterrand was very impressed by our resistance” (ibid.); on the other, he saw that “the advocates of withdrawal from the EMS were inconsistent, since Fabius and Bérégovoy had backpedaled” (Favier and Martin-Roland 1990, 473). The pro-austerity social niche had had its effect: the weight of the position shared by these actors was reinforced by the links they had forged and the collective work they had undertaken. From the perspective of the relational infrastructure, the construction of a social niche within the decisional space occurred through a steady increase in the density of links between austerity advocates, which went from 11% in May 1981 to 36% in March 1983, and through a simultaneous decrease in the density of links between advocates of the “other policy,” which went from 24% to 20% over the same period.

51The analysis therefore clearly demonstrates the effect of cooperation between competitors on the evolution and resolution of the controversy between stimulus and austerity. Although the pro-austerity position came from actors with statist-technical status, who were dominated within the social structure, over time this position gained more and more supporters in various corners of the decisional space. The relational work paid off. For example, we found in the archives two handwritten letters signed by Stasse: one is dated March 4 and is addressed to Bérégovoy. It begins “My dear Pierre” and aims to “address [to him] certain reflections, in a personal and very friendly capacity” on withdrawal from the EMS. The other, dated March 22, 1983, is addressed to Fabius. It begins “Dear Laurent” and constitutes, according to the presidential advisor, “one last chance to convey a few messages before the president ends the economic arrangement.” [41] The non-congruence of status, tied to Attali’s and Fabius’s roles—as men who simultaneously possessed the two opposing forms of capital (statist and political)—afforded them the opportunity both to attend the “evening visitors” meetings and to be sympathetic to pro-austerity arguments. In this way, the social and relational mechanisms worked together to make the regulatory process work, and to influence the unfolding of the controversy.

52***

53The “turn toward austerity,” a pivotal moment in French political and economic life, has been the subject of numerous studies since March 1983, and it continues to generate new research and interpretations today. This article has taken part in this movement, while turning a new lens on the controversy. Rather than describing an episode whose historical, political, and economic aspects are already well known, this article has sought to shed light on its sociological dimension by analyzing the structure and dynamics of the regulatory process. In theoretical and methodological terms, it combines the sociology of fields and neo-structural theory. Following the historical sociology of Ivan Ermakoff with respect to two decisive moments (the respective abdications of the German and French parliaments in 1933 and 1940 when confronted with an authoritarian power), the present article underscores the indeterminate nature of collective decision-making and the importance of social and relational mechanisms in resolving uncertainty.

54The neo-structural approach adopted here considers both the structural and singular aspects of the regulatory process. The actors are conceived of as being both the products and producers of the social and relational structures they are embedded in. In this respect, the multiple correspondence analysis and the social network analysis serve as complementary tools for analyzing their multiple forms of interdependence. Just as forms of capital explain economic representations, collaboration links explain the spread and adoption of a particular representation of economic policy: austerity. The explanations proposed in this article are thus situated beneath macro- or mono-causal explanations (Fabiani 2009) that take the “turn” to be an inevitable consequence of the “external constraints” weighing on the French economy at the time or of the “conversion” of the elites to the dominant neoliberal ideology. This article does not deny the fact that, at the end of the regulatory process, the socialist government’s choices may indeed have been adjusted to the international economic and ideological context. But this result was in no way inevitable: counteracting forces were certainly at work, but they were unable to triumph.

55The decisions of March 1983 continue to be discussed because, in retrospect, they appear to be a kind of mold for the regulation of the economy that was adopted in the following decades. Not only do they fit perfectly within the “reform trajectory” (Bezès and Palier 2018) that the French state began taking in the 1960s—and in which they constitute a pivotal moment—but they also fostered its continuation. In fact, from a budgetary policy perspective, the turn toward austerity heralded the policy of “disinflation” that was led with great success by the minister of the economy, Pierre Bérégovoy (Masset-Denèvre 1999), and of which the 1984–1986 financial liberalization reforms were a major component (Ménard 1999). From a monetary policy perspective, the decision to remain in the EMS and to peg the franc to the German mark for good presaged the decision, in the Maastricht Treaty of 1992, to opt for a single currency, with the introduction of the euro in 1999 (Denord and Schwartz 2009). Lastly, the trace of the budgetary and monetary orientations that resulted from the “turn” is visible in the socialists’ support for the austerity policies put in place in the European Union following the global financial crisis of 2008 (Lemoine and Eloire 2019).

56In conclusion, the decision to adopt a policy of austerity does not appear to have been simply the result of external constraints or the formulaic application of the dominant neoliberal ideology; it was also the incidental result of normative and relational work carried out by actors within the decisional space. This work was normative because the controversy helped legitimize a particular representation of economic policy (that of austerity), and relational because this legitimization depended on the collective force of interacting individuals.