Navigation – Plan du site

Work Representation and Union Representativeness on the Frontier of Wage Earning:

The Case of Direct Distribution of Printed Matter
Représentation du travail et représentativité syndicale aux marges du salariat : le cas de la distribution directe
Karel Yon
p. 83-98
Cet article est une traduction de :
Représentation du travail et représentativité syndicale aux marges du salariat : le cas de la distribution directe


La réforme de la représentativité syndicale permet-elle de lutter contre l’institutionnalisation des syndicats ? L’article s’interroge sur cette attente, qui se retrouve autant chez les acteurs de la réforme que chez les spécialistes du syndicalisme. En étudiant un secteur professionnel de constitution récente, la distribution directe, il montre que la réforme ne saurait répondre à un problème incorrectement formulé. L’institutionnalisation des syndicats est en effet un processus contradictoire, qui dissocie autant qu’il relie les salariés à leurs représentants. Cette précision conduit à étudier l’impact de la réforme de la représentativité syndicale de manière plus nuancée. La mobilisation des règles de représentativité dans les deux entreprises étudiées montre que les effets de la réforme en termes de syndicalisation s’avèrent ambivalents. Le nouveau régime de représentativité ne garantit pas davantage un contrôle plus étroit des salariés sur leurs représentants. Il accroît en revanche l’interdépendance des arènes de branche et d’entreprise, en tant que celle-là est désormais connectée à celles-ci via la procédure élective.

Haut de page

Texte intégral

Article published in French in Travail et emploi, nº 131, juillet-septembre 2012.

  • 1  “Le dialogue social sera toujours un combat,” interview with François Chérèque, Challenges, n° 117 (...)
  • 2  “Négociation sur la réforme de la représentativité. État des lieux et propositions de la CGT. Fich (...)
  • 3  Legislative record on the law n° 2008-789 of 20 August 2008 on renewal of industrial democracy and (...)

1It is a commonplace idea that the unions are too institutionalized which can lead to their acting in isolation, to negotiating on behalf of everyone without being accountable to anyone. In response to this problem, the law of 20 August 2008 “On the improvement of industrial democracy and the reform of working time” was adopted (See Box 1). Shortly before signing a joint position which would essentially be taken up in the law, Francois Chérèque, the general secretary of the French Democratic Confederation of Labour (CFDT), interpreted this negotiation as a means of bringing the unions and employees closer together: “The trade unionist is present in the company, the employee knows him. The unions are seen as institutions belonging to the decision-making elite. I also believe that unions are too institutionalized. This is why the trade unions’ representativeness in the workplace should be defined.”1 At the same time, a similar argument was made by the General Confederation of Labour (CGT), which criticized the 1966 decree as presenting the unions “as immutable institutions, separate from the expression of the employees they represent,” and called for representativeness which flowed “from the vote of employees in personnel elections [at] all levels (firm, professional sector, labour market area, region, nation).”2 By repealing the irrefutable presumption of representativity of the five “historic” confederations in favour of an electoral-based procedure, the 2008 reform aimed at establishing “strong and legitimate” unions.3 Not only might it encourage the unions to take the employees’ vote into account in their negotiations, in particular by prohibiting minority agreements, it might even play a role in encouraging unionization, forcing unions to have a sufficiently strong and widespread activist base so as to meet electoral tests.

  • 4 Box added to the English translation.

Box 1
The Rules of Union Representativeness and their Reform in 20084

It is crucial for labor unions to hold a representative status for it gives access to numerous rights (namely the right to designate union representatives and thus give legal protection to activists in the workplace, the right to collective bargaining, and the right to receive various state financing).
Before 2008, there were two different means by which trade unions could be considered representative:
1) The most convenient was state recognition. Since 1966, five confederations had benefited from that privilege known as the “irrefutable presumption of representativity” (présomption irréfragable de représentativité), which made them representative at all levels of action: CGT, CFDT, FO, CFTC, CFE-CGC.
2) Other unions had to demand recognition in the courts, and establish proof before a judge of their representative status. Each process of recognition was limited to a specific bargaining area, from a single workplace to a national firm or a whole industry. So it was very difficult and costly for newborn unions to develop.
The procedure decided on in the 2008 act is now based on workplace elections.
The first hurdle to overcome is to be recognized as a trade union, in order to get the right to run for elections. But the law made this quite easy. Hence each of the five above-mentioned unions, as well as additional organizations, are now placed in fair competition with each other and can designate transitory union representatives (RSS, représentants de section syndicale) in all the firms in which they are not yet representative.
The second step is crucial, for it decides recognition as a representative union. All unions that reach a threshold of 10% of the recorded votes are now considered representative. The recognition is effective at the level of the electoral process, which is the firm or a specific unit within it. Indeed, there can be several works councils within a single firm, in relation to several workplace locations.
In addition, a union member now has to be a candidate in workplace elections and reach the threshold of 10% of the votes on his personal name in order to be designated as union representative.
Beyond the firm, there are two additional levels of recognition: the industry level and the national, cross-industry level. Both are calculated by aggregating the results in all firms. The threshold is 8% of the recorded votes at the industry level. At the national level, unions also need to be representative in branches from a variety of sectors including industry, construction, trade and services.
Another important aspect of the reform is that the rules now forbid minority agreements. Any collective agreement has to be signed by unions representing at least 30% of the recorded votes, as a condition for it to be implemented. In addition, unions representing at least 50% of the recorded votes can oppose a collective agreement.
Two additional bills were passed in 2010. The first one concerned the representative status of unions among civil servants which is now based on a similar electoral process. The second one concerned very small firms (less than 10 employees). Workers in those firms do not have elected representatives. To ensure that all workers are called on to take part in defining which unions are representative, they now have to vote every four years, not to elect individuals but to choose the union by which they agree to be represented.

  • 5Structural reorganization, targeted objectives in increasing membership, targeting the underrepres (...)

2The reform of union representation was defended by its proponents as the first step in a solution to the crisis of union representation in relation to union weaknesses within important sectors of the wage earners, whether in the TPE (very small businesses) and PMEs (small and medium enterprises), or in the service sector, among youth and precarious workers. Up to now, organizations’ proactive efforts alone5 have not been able to check this crisis. Testing this assertion would require investigating multiple fields corresponding to the different possible fields of unionization. This article aims at providing an initial contribution to this work of documentation from one industrial sector. Mailbox distribution of unaddressed printed material (flyers, free advertising, and institutional publications), which industry players call “direct distribution of printed matter [distribution directe],” is an interesting case for several reasons. It is a symbolic field in the “new frontier” of unionization. The emergence of this sector is illustrative of the growth of commercial services in our economy where unions are less well established than in industry or the public sector. The very recent establishment of the sector illustrates the reconstruction of the boundaries between public and private sectors. Part-time work, an atypical status which undermines job security, is the dominant form of employment in the sector (see Box 2).

Box 2
Direct Distribution of Printed matter: Institutionalized Job Insecurity?

  • 6 The company names have been changed, although the almost duopolistic structure of the branch makes (...)

The activity of the distribution of unaddressed printed matter developed in the 1970s along with the rise of free advertising media and large scale supermarket distribution, although its origins are even older. With the decline in advertising media, as illustrated by the recent liquidation of Paru/Vendu, and the development of internet-based advertising sites such as Le Bon coin, distribution firms have entered the field of virtual communication while the work of distributors has refocused on the dissemination of advertising brochures, of institutional publications issued by local authorities, and more recently of the new free information press (such as Metro or 20 Minutes).
Although activity is extremely fragmented, it is now mainly dominated by Gratuit6, a subsidiary of La Poste, and Distrib, a subsidiary of a major media group, which together share 95% of the market. The two companies are roughly the same size. The former has a smaller workforce but has higher sales (430 million euros, compared to 312.9 million euros for Distrib). According to the Association for Financing the Performing Arts (AFDAS), which collects the job training taxes of the direct distribution of printed matter branch, in 2010 there were some 43 distribution firms contributing to this approved professional collector organization. Most are micro-companies that are often beyond the rules of the industry. Some firms with more than 50 employees have a regional activity, sometimes subcontracting orders from one or another of the “big” firms.
Low wages and difficult working conditions regularly give rise to articles of “moral denunciation” in the local or national press. Data presented in the social balance sheet of the branch gives an idea of the conditions for distributors. This data covers the two largest firms, Gratuit and Distrib, which are the only members of the Professional Union of Direct Distribution of Printed Matter. In 2011, there were 32,985 distributors, 40% of whom were women, and 3,269 supervisory “staff” personnel. The total size of the branch, according to AFDAS, is 22,500 equivalent full-time employees, which emphasizes the importance of part-time work. The work force is rather old: half of distributors are over 50 years old and only one-eighth of them are under 30. Almost all distributors are on permanent work contracts (CDI [contrat à durée indéterminée]), even though a quarter of those who are newly hired are on limited duration work contracts (CDD [contrat à durée déterminée]. There is a high turnover rate, with almost as many leaving as those hired (32,750 as compared to 33,279). A quarter of all departures are by resignations and almost half are at the end of the trial period, which reflects the difficulty firms have in stabilizing the labour force. Only 5% of distributors are employed full-time, two-thirds are employed on part-time contracts of less than 65 hours per month, often below the minimum level for entitlement to social security benefits (60 hours per month). The average salary reported in 2011 was 628 € for women and 655 € for men. To the extent that distribution is primarily a labour intensive industry, highly dependent on “key account” customers such as the supermarkets, distributors’ remuneration remains the main economic adjustment variable. In practice, distributors often work for periods exceeding the number of hours for which they are paid, which is regularly confirmed by the convictions for “clandestine work”.
The flexible work hours also make direct distribution of printed matter a popular activity with young people, women with family responsibilities, as well as workers who need to combine several salaries. The lack of skills required and the ease of hiring also attract many immigrant workers, retirees seeking extra income (military and retired policemen, impoverished pensioners) or individuals who have had biographical accidents (bankrupt shopkeepers, socially marginalised employees). There is no real collective work: the distributors are spread over several hundred sites which are all over the map and where they are rarely present. They come every week to get their batch of printed matter to distribute and do their rounds alone.

  • 7For a summary in French of this literature, see Thomas, 2011.

3Direct distribution of printed matter therefore combines a set of features which are challenges for a “union renewal.”7 Far from the experiments conducted in France or abroad in organizing precarious workers (Béroud, Bouffartigue, 2009), union intervention in this sector has in particular advanced a traditional register of action, that of collective bargaining. The establishment of this branch [branche professionnelle, an industry-wide collective bargaining unit] is the result of a National Collective Agreement (CCN, Convention collective nationale) recently adopted in the mid-2000s, after a decade of negotiations. Union recognition was immediately applied at the institutional level through collective bargaining rather than through collective action, but this does not seem to have put an end to either the precariousness of the sector or to its social conflicts which remain endemic. We thus find a situation similar to that described by Jean-Michel Denis (2008) in the industrial cleaning sector: precarious workers spread across multiple sites, thus preventing unions “from having a territorial density” sufficient to enforce contractual work site standards (p. 48). He also discussed the “classic problem of ‘distance’ or the space between unions and employees, generated by the mechanism of representation, the bureaucratization of organizations, and/or the institutionalization of trade unionism” (p. 45). Would direct distribution of printed matter become a similar example of this disconnection between the workers and their representatives? In this case, would the law of 20 August 2008 “On the improvement of industrial democracy and the reform of working time” help alleviate this problem? Without claiming any possible generalization of its empirical findings, this text nevertheless aims at providing food for thought going beyond the framework of monographs by shedding new light on the question of the institutionalization of unions.

  • 8  This point of view is quite far from the conceptualizations developed around the dynamic and plura (...)
  • 9  A similar analysis can be found in Chris Howell’s theory of “virtual unionism” (1998, 2009) accord (...)

4In most discussion on trade unionism –whether by researchers, journalists, political and administrative actors or by trade unionists themselves– institutionalization is seen in a negative sense: it refers to the process of growing independence, isolation and bureaucratization of organizations which could at least partially explain the crisis of the labour movement.8 For many years Dominique Andolfatto and Dominique Labbé have insisted on the “weak social roots of unionism.” Pointing to an “increasingly questionable representativity,” they portray unionists as “professionals of representation […] who no longer have any links, other than electoral ones, with the employees they are supposed to represent and have only a more or less theoretical knowledge of the real situation and needs of these employees” (Andolfatto, Labbé, 2006, p. 349, p. 351). This analysis perpetuates an old problem: the critique of the institutionalization of unions, born at the dawn of the labour movement with the “iron law of oligarchy” (Michels, 1971), which experienced a significant renewal in interest in France in the 1980s, during an unlikely conjunction of the collapse of union membership and the increase in the institutional power of unions (Adam, 1983). It was specifically formulated by Pierre Rosanvallon, who explained the institutionalization of trade unionism by the establishment of its function as a “social agency”, i. e., “all union functions which are related to an institutional role […]. As a social agency, the union is a kind of ‘public office of social affairs’, a quasi-public agency. This aspect differs from its dimension as a social movement, a force which demands and negotiates” (Rosanvallon, 1998, p. 24).9 Both the justifications as well as the criticisms of the reform of trade union representation are generally based on this diagnosis. For some it is a way of adjusting union representation to the world of real work by reconnecting employees and their representatives through the ballot box (Bévort, Jobert, 2008). Others have predicted that it would have the opposite of this intended effect: the intensification of the institutionalization of unions and thereby the loss of their social roots (Andolfatto, Labbé, 2009). These opposing viewpoints share the view that there is a gap between the world of “real” work and its union representation. It is this assumption that I intend discussing in this article.

5In my view, institutionalization should not only be thought of as a process of differentiation, separating union representation from its social “base.” I intend demonstrating that this dynamic must be described in its contradictory nature, as a movement that also has the effect of binding the unions closer to the workers. This approach is underpinned by a “constructivist” approach to wage labour. As opposed to the hypothesis of the disjunction between the world of “real” work and union representation, it presents both of them as being involved in the institutions of wage labour (Vatin, Bernard, 2007; Friot, 2012). Institutions, as “mechanisms of knowledge and action,” organize economic activities (Salais, 2008), just as much as they shape the “workers” in their social being and subjectivity (Castel, 1995). Unions are doubly involved in this structuring, as established form and establishing force: if “labour law has largely been built by building unions” (Borenfreund, Souriac, 2008, p. 2), this is because unions, in turn, were built by building labour law. In this perspective, the institutionalization of unions should be understood in a wider context, that of the growing legal formalization of labour relations with negotiations by branch being a key vector of this legal network (Didry, 2002). Such an analytical framework, based on a detailed field survey (see Box 3), means breaking with a mechanistic and homogenizing conception of law to consider it in its duality and plasticity, both as an “ideological framework structuring labour relations, and as a symbolic and material resource, whose usage can change the current content and forms of domination” (Pélisse, 2009, p. 82).

6First, I will discuss the relatively recent history of this industry, which is characterized by the joint institutionalization of the branch and of its trade unions. I will demonstrate the limits of an analysis in terms of the disconnection between employees and their representatives. Describing the conditions of institutionalization of the direct distribution of printed matter sector helps, in turn, in putting the implications of the 2008 Act into perspective. In fact, we can only grasp the changes induced by the new regime of representation by comparing the sequence of events following this reform to those which preceded it. In the second part, I demonstrate how the mobilisation of the new rules of union representation have transformed the reality of the unions in the sector, both in terms of employee representation as well as in social dialogue. In this regard, the unique configuration of the sector as a virtual duopoly, facilitates the comparison and analysis of interactions between the branch and the firms.

Box 3
Material and Sources

  • 10 Research done for the Dares (2010-12) conducted with Sophie Béroud, Jean-Michel Denis, Marnix Dress (...)

This article is based on materials collected in the course of on-going research on the implementation of the reform of August 2008 “On the improvement of social democracy”.10 Since March 2010, I have interviewed various actors in the sector: two representatives of the Labour Ministry’s administration; sixteen union negotiators and representatives at the branch level; stewards and union activists at the two main firms of direct distribution of printed matter; the representative of a “free” trade union of the management personnel; and a representative of the corporate management. I also had access to the records of the representative of the General Directorate of Labour (DGT) who had more or less directly followed labour relations in the sector since 2000. These include reports of joint bilateral committees; correspondence of the social partners with the DGT; the internal communications of the DGT; various union documents; and the “liaison files”, minutes of the activities of the Presidents of the Joint Commission (PCM) representing the Ministry of Labour in the collective bargaining sessions. I collected much of the material (union leaflets, testimonies and analyses, newspaper articles) in paper form or on the websites of trade unions and several blogs and forums of sector employees. Finally, I attended a three-day meeting/training session of activists and union representatives from both firms, as well as the meeting of a joint committee and a bilateral working group at the branch level.

Institutionalization and Unionization of Direct Distribution of Printed Matter

7According to the classical pattern of institutionalization, which can be found both in Rosanvallon and his followers (Duclos, Mériaux, 2009) as well as in the neo-corporatist literature (e. g., Offe, 1985), the union which has its power guaranteed by the state is no longer concerned with its real base among the wage earners. The living reality of work gradually escapes it and it risks losing its ability to mediate. This pattern is a priori well suited to explaining the establishment of the union in the direct distribution of printed matter sector essentially “from the top.” Two sequences can be distinguished: the foundation phase corresponding to the establishment of conventional standards and the institutionalization phase, corresponding to the subsequent stage of the routinization of the system of rules.

Foundation of the Branch and Union Recognition

8For a long time the distribution of unaddressed printed matter was dispersed among very different professional worlds: on the one hand, the postal world, and on the other, that of the press and advertising. The first attempt to introduce advertising into the postal delivery rounds was in 1953-56 and it was cancelled due to union resistance. It was repeated at the beginning of the 1970s on a voluntary basis. The postal agents were divided between those who accepted and those who refused to do this work which was accompanied by additional compensation (Cartier, 2003). In the private sector of advertising agencies and media companies, the activity was, as in the delivery of the press, mainly paid by piecework wages. Piecework wages are based on a commercial law service contract: “the firm buys a global service counted in number of pieces to be produced, negotiating a unit price which in the industry’s jargon is called a ‘chantier’ [job] or a ‘lot’”(Célérier, 2012, p. 91). Here the job is defined as the number of mailboxes to which one has to deliver. For firms that use this type of service, this is a way of externalizing costs as well as occupational hazards. Workers, called distributors, perform some work at home (preparing bundles of prospectus to be put into mailboxes) and provide their own work tools (walking shoes and clothing to cope with the weather, vehicles to get to the work site, trolleys to carry bundles, etc.). They are solely responsible for the conditions for the realization of their service, which can particularly lead to mobilizing the whole family, the spouse and children who make a free contribution to the job. Denigrated by the postal worker unions, which were attached to the spirit of public service and saw this as commercial “dirty work,” the activity was in the meantime barely visible in the private sector. Giving rise to multiple informal arrangements negotiated privately between distributors and contractors, the press and advertising unions considered it just as peripheral as it was difficult to understand.

  • 11Interview, 24 January 2012.
  • 12  More specifically, there was a union presence in some companies, but it was tolerated to the exten (...)

9Paid by piecework, distributors were issued contracts that offered no guarantee of compliance with the hourly SMIC (minimum wage), which was illegal. That is why in the early 1990s the General Directorate of Labour [DGT, Direction générale du travail, a division of the Labour Ministry] invited the firms involved in this field of activity to reach an agreement with the union confederations presumed to be representative so as to give contractual protection to workers in line with the generalization of contracts by branch launched in the late 1970s (Tallard, 2004). A leader of the FO (Force ouvrière, Workers Power) union of publishing, media and advertising, who was one of the architects of the sectorial contract, underlined the reactive nature of this union “concern”: “We plunged into the subject, what is direct distribution of printed matter? Of course there was no unionization.”11 Unions were invited to submit proposals: expanding the scope of an existing agreement or negotiating an ad hoc text. The position of this unionist was to negotiate an annex to the collective agreement for the logistics of direct publications. But the creation of a specific employer organization, the Union of direct distribution of printed matter (SDD, Syndicat de la distribution directe) in 1994, reoriented the negotiations towards the production of an autonomous text. Initially, there was an evident convergence of interests between employers and unions. For the unions, negotiation at the national level was a way of compensating for their absence from the workplace. Such negotiating by branch is a well-known practice. It has particularly helped FO find a place in the union movement (Bergounioux, 1975). For its part, forced by the public authorities “to legalize its activities,” the SDD was prepared to tolerate the unions at the level of the branch as long as they kept out of the firms.12 Should this really be described, however, as unionization from above with, as a result, the original sin of “rootless” organizations confined to the arena of bi-lateral negotiations and supported more by institutional recognition rather than by that of the distributors themselves? In reality, the dynamics of the negotiations progressively changed the situation.

  • 13  Appeals to the labour tribunals regularly led to the reclassification of the distributors’ contrac (...)
  • 14  Interview, 24 January 2012.

10In the initial phase of development of contractual rules, collective bargaining at the branch level helped establish a union presence. While the first union negotiators were external to the industry (they were full-time union representatives from the world of media and advertising), they shared a common desire of linking up with the distributors. The unions established a presence in firms by advancing their goal, announced during bi-lateral negotiations, “to make distribution a real profession” and challenging piecework wages in the courts, in industrial tribunals13 and in workplace strikes. The circumstances and the rhythm of bargaining at the branch level were thus determined by the exchange of blows that were played out in other areas. Similarly, the participation of workers from the branch in trade union delegations was one of the issues in a confrontation with the SDD, which the unions at times accused of not keeping its commitments to cover travel costs, a condition for the non-Parisians who were not full-time staff to participate in negotiations. The presence of distribution workers provided experience-based knowledge which encouraged discussion, but as evidenced by the minutes of meetings, it was the full time union federation representatives who conducted the negotiations. In addition to calling on a variety of skills, including rhetorical and legal, and coordinating union action on these multiple fronts, the full time staff ensured continuity while workers from the ranks changed regularly: “For us, our delegates were always being fired,” a CGT official told me in an interview.14 He understood this as a testimony to the integrity of his delegation, blaming some unions for having recruited workers who were not clearly independent vis-à-vis the employer. But it also demonstrates the fragility of the establishment of the unions, confirming the declaration of an FO full-time staff worker who was interviewed and said that before the establishment of the CCN, “union representation was ridiculous and ephemeral.”

  • 15  Gratuit was created in 1987 as a subsidiary of La Poste [the public Postal service] offering firms (...)
  • 16  This is particularly true of the CGT federations of the postal service and of book publishing and (...)

11The negotiations, which had been blocked, were reopened with the arrival on the employers’ side of an actor who was more open to “social dialogue.” The 2003 merger-acquisition by the La Poste group of the company Gratuit,15 a traditional distribution market operator, marked the beginning of this public operator’s role in the private distribution market. Following the acquisition, the unions of La Poste began to penetrate the sector, organizing the distributors, sometimes in competition with the unions of the same union confederation,16 and taking part in the negotiations. Gratuit proclaimed its intention of spreading the group’s “social model” among the distributors and encouraged a number of concessions. The company generalized part-time permanent work contracts (CDI [contrat à durée indéterminée]), made commitments in terms of improving working conditions, vocational training, etc. These company agreements preceded and pushed forward the conclusion of the branch agreement.

  • 17  Modulation refers to the potential created by a collective agreement to vary the workweek without (...)

12A compromise was finally signed in the National Collective Agreement (CCN) of direct distribution of printed matter. The text was signed on 9 February 2004 by the SDD and the unions and federations affiliated to the five confederations. It was “étendu” [made compulsory for the entire industry] by order of the Ministry of Labour on 16 July 2004 and came into force the following year, on 1st July 2005. The main change brought about by the establishment of the CCN was in officially abandoning piecework wages. Since then, wages in the distribution sector have been based on a time measurement, recorded in permanent work contracts. But the status of distributor, structured by the CCN, is nonetheless atypical: it combines a modulated part-time work contract17 to a particular system of compensation, called “préquantification” that makes an a priori evaluation of the time required to accomplish the distribution jobs. This system addresses the employers’ concern for “flexibility” in meeting the demands of the market through a special exemption in the common labour law which imposes an ex post count of hours worked. The compromise of the CCN, endorsed by the General Directorate of Labour, exchanges this special exemption status in return for strong union recognition granting important union rights concerning both the branch as well as the firms (see Box 4).

Box 4
Union Rights in the National Collective Agreement (CCN) for Direct Distribution of Printed Matter

The CCN lays down common provisions across the entire sector in terms of trade union rights and representation of employees. Three of them deserve to be highlighted. The first concerns firms: given the dispersion of distribution centres and the frequency of “short part-time contracts”, the CCN provides that a coefficient multiplier of 1.7 should be applied to the calculation of actual full-time equivalent jobs. This calculation is used in defining the workplaces, determining the number of positions in the employee representative bodies (IRP), and the number of hours of union delegation. The second provision relates to hours of union delegation: they are paid as time worked when taken during the working hours of the employee representative (RP [représentant du personnel]). When taken outside of working hours, they are considered as overtime and calculated on the basis of the minimum wage. The third provision relates to the branch: CCN provides that each signatory trade union organization has a “technical advisor” in charge of monitoring the affairs of the branch in the person of a full-time employee of one of the firms in the sector, temporarily assigned to this function. This is a rather unique provision: it is rare that the employers’ organisation directly pays for union activity.

13The process of union organizing was thus promoted “from above,” but it was not limited to “summit unionism”: union recognition went beyond the scope of the arena of bi-lateral negotiations to become rooted inside the firms themselves, even during the negotiations, and this was then included in the text of the CCN.

Anxiety Over Representativity: the Difficulties in Legitimizing the Branch Compromise

14While the “unionization from above” hypothesis is wrong, would it be more appropriate to counterpose the initial foundation phase of institution, the foundation phase conducive to unionization since it necessitated the active mobilization of interests, from the subsequent phase of institutionalization, marked by the demobilization of distributors and the establishment of their representatives’ own interests? Thus we once again find the pattern in which once legal representation has been established, it is no longer necessary to activate social representativeness. The initial phase of unity is followed by a phase of demobilization or differentiation of interests, resulting in a “risk of the loss of recognition for organizations whose future is now guaranteed and who no longer have the interest or capacity to demonstrate their social representativeness” (Offerlé, 1998, p. 72). In fact, there were various expressions of distrust of the organizations associated with the professional branch compromise, leading to the emergence of new players claiming to represent the distributors. However, these phenomena cannot be summarized as a general pattern of the disconnection between institutionalized unions and discontented distributors.

  • 18  The FAPT is the Federation of Postal and Tele­communications Activities; the FILPAC is the Federat (...)

15The dispute initially arose from within the unions themselves at the time the collective agreement was signed. Some unionists did not accept the compromise over the special exemption. Their action in the field had been on behalf of the standardization of wages in the sector. One interviewee mentioned the case of a distributor representing the CFTC (French Confederation of Christian Workers) “who came to commission meetings with the labour code under his arm” and was dismissed by his confederation who accused him of blocking the conclusion of the negotiations. Some of the unions from La Poste were amongst the most critical of the outcome of the negotiations, seeing the formalization of the status of distributor as ultimately a risk for the status of the postal worker. Thus, the FAPT-CGT refused to sign the CCN, but was deterred from demonstrating its opposition following internal arbitration inside the confederation, the CGT’s signature being provided by FILPAC.18 But tensions occurred mainly between the players involved in the bilateral arena and those activists present in the firm, as evidenced by the testimony of a CFDT union representative:

  • 19  Interview, 12 April 2012.

“[the CCN] was implemented at Gratuit on 5 July [2005], we went on strike on July 5. […] I would not deny that as a central delegate, in 2005, I sent a registered letter to my federation, saying it should denounce this contract.”19

16In their concrete application in the workplace, the new rules of the branch compromise decreased direct compensation. A distributor, a former CGT militant who started at Gratuit in 2000, spoke up with others against the CCN: “We compared pay slips.” Another employee, a former activist of the National Union of Autonomous Trade Unions (UNSA) who had worked at Gratuit well before its takeover by the group La Poste, noted that the application of the contract had divided his salary by three. This is what was summed up by the FO negotiator interviewed:

  • 20  Interview, 24 January 2012.

“Why is it that we, the trade unions, were attacked when this agreement was published? It is that in the context of this reform there has been a fairly substantial decrease in the amount of reimbursed expenses, mileage, etc. But what we wanted was a real professionalization, with wages which would lead to having rights in social questions, welfare and others. So there has been a transfer. But the transfer was not equal.”20

17However, criticism of the agreement was not the reaction only of the “old” distributors who remembered the working conditions prior to the establishment of the branch. If that had been the only question, then the turnover and repression of protesters would eventually have exhausted the source of criticism. But social conflict was continuous in the sector. Paradoxically, the distributor status itself, while strengthening the flexible organization of employment, helped disseminate the dispute. By objectifying the condition of atypical employees, it led distributors to think of themselves as employees while in practice it kept them at the frontier of wage earners. Profane criticism thus developed among distributors alongside expert criticism conducted in the name of the law by labour inspectors dissatisfied with this infringement of the Labour Code.

18Some distributors did not hesitate at interpreting the industry’s compromise agreement as a cynical exchange between employers favouring their own economic interests and union representatives preoccupied with their bureaucratic interests. In this context, the mobilization of distributors was a manipulation by union officials to influence the negotiation, before abandoning them to their fate once the agreement had been signed. Also advanced was the argument that the unions were “bought” with a post as technical advisor, in which the SDD financed a full time official for the branch for each union signatory to the agreement.

19This analysis was given even more credit by the fact that all the confederations recognized as representative in the industry supported the compromise. The divorce between the distributors and the unions is explicitly proclaimed on the “forum of the Union of employees of direct distribution of printed matter and of postal activities (USDDAP).” This internet site, created in 2005, illustrates the emergence of forms of action alternative to the trade unions. This site denounces “the active complicity of the trade union federations of the branch in their assistance provided to employers in the exploitation of the distributors and the massive and illegal practice of undeclared work.” On this site, the distributors centralize legal information and share advice on negotiating their employment contract or defending their rights against the employer. A section of the forum even announced in 2007 the establishment of an USDDAP guide,

“which will include on the one hand an assessment of the depots, on the other hand an evaluation of shop stewards and union delegates. This information will allow applicants for distributor or postal worker jobs to know in which depot of his area it is best to submit an application. It will also assist active employees to know for whom to vote in the various elections, particularly in employee elections, and who to avoid.”21

20Some expressions were simply totally exterior to unionism, as in this letter from a distributor to the Director of the General Directorate of Labour on January 25, 2011, which asked that “the legal rights of the Labour Code no longer be diverted by agreements between partners.” It is true that the role of technical advisor gave some reality to the representation of a threatening world of “social partners,” quite distant from the distributors. The office created the position of an expert in distribution activities, socially distinct from the activity itself. This was not compensated for by the profile of committed trade unionists: among the currently employed advisors, only one of them is a former distributor, and he had no longer been one for quite some time before he came to office (he had been promoted to warehouseman and then became a full-time union staff worker), and another is totally outside of the sector as he is a retired postal worker.

21It is too schematic, however, to link these diverse phenomena to the rejection of unions by the distributors. First, because the establishment of SUD-PTT, a union which had been outside the sector, was also fed by the criticisms of the unions which had signed the CCN. In fact, many representatives of this organization had come from competing unions (UNSA, CGT, CFDT, etc.) which were accused of being too conciliatory vis-à-vis the employers. But the union signatories of the CCN also contributed to shaping the criticisms of those employees dissatisfied with their condition. Unions, whatever their confederation, are in fact the most common instrument utilised in the many labour court disputes over the recognition of the work of the distributors. They are present in the strikes on compensation issues or working conditions that regularly affect the warehouses. Thus, we should not forget that the formalization of labour relations has also led to publicizing and legitimizing the unions, both directly and indirectly. Through the signing of the CCN, the employers in the sector had recognized the legitimacy of union representation, which went beyond the level of the industry, and was expected to be closely linked to greater union presence in the firms. Thus, in the wake of the CCN agreement, new agreements or amendments in application of the new rules on trade union rights were negotiated in various companies in 2005. This particularly concerned the application of the coefficient factor used to determine the number of representative offices. At Distrib, the company was obliged to redefine its relationship to the unions, as evidenced by a supervisory manager:

  • 22  Interview, 7 December 2010.

“There was nevertheless a negative image of unions. Even I also had this view… […] The agreement has led to many advances. I do not know if this is because of writing up the rights and obligations of employees… there was a strong rise in social policies. […] We had to form a social relations network.”22

  • 23  In the context of the liberalization of the postal market, Distrib decided at the time to invest i (...)
  • 24  The calculation of such remuneration is further defined by the introduction of computer software.

22The CCN has also accompanied –and probably accelerated– an economic reorganization of the sector which has also contributed to the increasing union presence. Along with the entry of new operator, Gratuit, on the market, its main competitor, Distrib, embarked on a strategy of acquisitions. Some of the firms which it acquired already had union representation, especially in the southwest of France. The market for the direct distribution of printed matter was actually the front line in a war which the two main operators had been preparing with the liberalization of the postal market in the background.23 The increased competition led to the creation of a virtual duopoly through successive purchases and the liquidation of less competitive firms, while economic profitability imperatives within each firm were becoming stronger. It is just as much due to the establishment of the CCN as it is to the integration of local structures into national entities that labour relations have progressively been transformed from a local mode to a bureaucratic one, ruling out –and sometimes making technically impossible24– informal local arrangements. Several of the activists encountered justified their union involvement during the 2000s by a deteriorating climate in the workplace due to these changes.

23In addition, while it created the conditions for greater union involvement of employees, the CCN did this based on a utilitarian relationship with the unions, maintaining an interest in unionization which was less oriented towards defending the “gains” of the agreement than it was to obtaining benefits for oneself. There is abundant evidence from various union activists about union representatives or elected delegates without any “trade union consciousness.” In a manner similar to that which exists in agency employment, holding a union post is a complement to one’s initial wage (Grollier, 2010). Posts can be used by the distributors as a “bonus” or even as a cost imposed on the company in order to raise the stakes in negotiating a dismissal (and, after 2008, a contractual termination) or in pressuring the employers to offer a promotion. They could also be offered as recompense by the employer himself. These practices, which are deviant with respect to the activist norm of legitimate disinterestedness, can be understood in the specific context, both social and legal, which forms them. While part-time work along with job insecurity and low wages continue to dominate the sector, the CCN established a relative unlinking of the opening of worker rights and of trade union rights, while the absence of work groups and the personalization of relations with lower supervisory personnel facilitated individual strategies. In addition, interviews with unionized distributors show how utilitarian interests and an “interest in disinterestedness” (Bourdieu, 1984) intertwine and are continuously redefined in the course of a union career.

24Therefore the institutionalization of the sector through the branch agreement did not mean cutting the unions off from the distributors. Instead, it had two partly contradictory effects: on the one hand, it promoted the dissemination of the union form as a legal resource and its adaptation inside firms through various channels, and on the other hand, it stabilized an arena of additional representation having its own logic, that of the branch. This reconfiguration has created a structural tension, a tension that not only opposes distributors and union representatives, but also runs through each of the unions present in both areas. The landscape of direct distribution of printed matter, as shaped by the CCN, thus resembles an unstable arrangement that the 2008 reform may have finally shaken.

The Contradictory Dynamics of “Industrial Democracy”: What the 2008 Reform Changed

25The 2008 reform gives greater importance to personnel elections than had been the case in the previous configuration. By subordinating the legal representation of the union to the electoral expression of the employees, it expressly aimed at bringing the two above-mentioned dimensions of representation (legal and social) closer to each other in the name of “industrial democracy.” We have previously seen that the law lends itself to multiple uses. It should not, therefore, have a mechanical effect. The way in which the new rules of representation were applied in the two main firms in the branch demonstrated this. Beyond its application, however, the 2008 reform involved a reconfiguration of the legal system that put an end to the structural disconnection between representation in the firm and representation in the branch.

Ambivalent Unionization in the Firm

  • 25  This little known organization boasts an existence since 1953. It has a website: http://www.c-a-t. (...)
  • 26  Interview, 7 December 2010.
  • 27  SUD-PTT, which is developing a group strategy, gained a foothold in the direct distribution of pri (...)
  • 28  Indeed, if the union represented by an RSS has not reached the threshold of representation after t (...)

26The reform of trade union representation reinforced the trend to the unionization of labour relations that had been advanced by the institutionalization of the branch. This is particularly clear in the case of Distrib. While granting new means to the trade unions, the firm maintained, according to the unionists who were interviewed, a policy of “containment” of unions. This was reflected in particular by its support for non-union candidates in the elections of shop stewards (DP [délégués du personnel]) and to the works council (CE [comité d’entreprise]) which led to the “neutralization” of the employees’ representative bodies (IRP [instances représentatives du personnel]). The supposed close relations between the majority of the CE and management meant that the latter did not have to face any obstacles in those proceedings which required the council’s opinion, whether in the dismissal of protected employees, the monitoring of the typologies of préquantification or in the firm’s economic and social policy. When the IRP were re-elected in 2010, the application of the 2008 Act changed the rules of the game. Those candidates who previously had not been members of a union shifted to the lists of an autonomous union, the Autonomous Confederation of Labour (CAT).25 Its founder in the company acknowledged having had the idea following a training session organized by management and led by consultants who were external to the company.26 The suppression of the irrefutable presumption of representativeness has also enabled SUD-PTT to participate in the negotiation of the pre-election protocol agreement and to establish its organisation.27 Some transitory union representatives (RSS [représentants de section syndicale]) were appointed prior to the election, but only in those warehouses where there were chances of their getting elected.28 All the unions increased in number of votes and expanded their organisations, reflecting the efforts made in confronting the new question of representativeness. The increase in voter turnout in the first round (which reached 38.33% as compared to 17.05% in the previous election in 2007) and the decline in blank and spoiled ballots (from 18.38% to 10.31%) thus has a two-fold explanation: it was not only labour, but also the firm’s management who were interested in the outcome of the election. In a context in which more or less restrictive injunctions to bargaining at the firm level were increasing, the fact that the new rules subordinated unions’ authority to sign agreements to their electoral representativeness led the employers, according to some interviewees, to bet on an independent union in the first round rather than on a second round open to non-union candidates. With 13.4% of the votes cast in the first round, the CAT won representativity. SUD did not reach the 10% threshold, but by approaching 8%, the union received enough votes to allow it to claim representativity at the branch level. For the first time in the firm’s history, a union majority had won a majority of the works council, thus activating a number of legal powers (especially in terms of economic expertise) which until then had remained dormant.

  • 29  Interview, 18 January 2012.

27The situation at Gratuit, where union presence had become an established fact, was more complicated due to the presence of many organizations: CGT, CFDT, FO, CFTC, CFE-CGC (French Confederation of Management Personnel-General Confederation of Executives), UNSA, SUD-PTT. The public operator had brought with it the language of “social dialogue” and an expertise in the management of collective labour relations which sharply differed from the anti-union policy that had previously existed in the sector: “Voting is an act of a ‘corporate citizen’ which should be encouraged so as to fully express democracy in the firm,” read the pre-electoral Memorandum of Agreement in 2007. The effort made in electoral organization at the time had led to greatly increasing the participation rate which went from less than 30% to 49.9% between the 2005 and 2007 elections. It reached 62.2% in the next election in 2011. Management was accustomed to negotiating –“in five years I had to sign 36 agreements,” said the current Director of Human Resources29– and had its preferred partners, which changed with the issues but often included two “small” unions: UNSA and the CFTC. In a firm which has an established practice of social dialogue, hostility vis-à-vis the employer does not serve as social cement: it is competition rather than cooperation between unions that seems to prevail. For several years, a local CGT union existed concurrently in the firm alongside a federal CGT representation, the organizational translation of the structural tension between the levels of the branch and the firm. During the personnel elections in October 2011, lists of SUD-Gratuit competed with lists of SUD-PTT. This dissent was facilitated by the structure of the IRP which, unlike its counterpart at Distrib, is decentralized. Not all unions presented candidates in the six works councils and not all the unions attained the threshold of representativity. The 2011 elections thus resulted in a union representation with variable geometry. Only the CFDT and CGT were representative organizations in all works councils. FO, CFTC and SUD-PTT were respectively representative in five, four and three of the six works councils. The UNSA was the big loser. It was only recognized as representative at one work unit and lost its representation at the national level of the firm thus depriving the management of a privileged partner in the firm’s negotiations. SUD-Gratuit, which had presented lists in two regions, did not attain representation at any level.

28The new representativity rules authorized various legal “tactics” on the part of trade unions or employers. In the Memorandum of Understanding concerning the shop steward elections, Distrib management introduced a clause excluding heads of warehouses from the electorate. By depriving them of an electoral sanction, it cut the grass under the feet of unions such as the CFTC which had started their unionization, designating some of them as union delegates. Similarly, SUD-PTT managed to deal a fatal blow to the CGC by challenging the categorical identity of the organization all the way to the Court of Cassation. In the first round of the 8 October 2010 elections, the CGC union had only presented candidates in the second and third electoral colleges and had its representativeness recognized on this categorical basis. Since the quorum was not reached, a second round of the elections was held on 19 November 2010. This time the CGC introduced lists in all the three electoral colleges. Crossing this fact with the union constitution which says that it represents “subject to certain conditions, the employees,” the Court concluded, in a decision of 31 January 2012, that the representativeness of the CGC should be measured by combining all three colleges, which had the effect of depriving the union of its representativity and cancelling the appointment of its central union delegate. The site of the FO-Distrib section wrote that “while management systematically attacks the designation of union employees whenever it can, it took the trade union SUD to attack the union of management personnel in order that justice was done.”

29For Gratuit, the subdivision of the professional representation in the regional works council was determined by the 2007 elections. The General Directorate of Labour had responded to a request of the CGT for arbitration by imposing this decentralization. One month later, the Gratuit management denounced the section of the firm’s 2005 collective agreement concerning trade union rights: it associated its proposals for the regionalization of trade union rights to a reduction of some of its resources, which led to an “Agreement on Social Dialogue in Gratuit” in January 2009 with only three unions signing (CFTC, FO and UNSA). A clause in the agreement, of innocuous appearance at the time, was to have important consequences:

“The provisions concerning union representatives will apply until the next elections in all works councils. They shall cease to apply as of right after the announcement of the election results so as to be in conformity with the law of 20 August 2008 for the section concerning the renewal of industrial democracy. A new negotiation will begin within six months preceding the date of the first round.”

  • 30  Interview, 5 October 2012.
  • 31  In several judgments of 10 November 2010, confirmed by a judgment of 18 May 2011, the social chamb (...)
  • 32  While the works council is national at Distrib, the company agreement which organizes the designat (...)
  • 33  Only four judgments were in favour of unions and will have to go to the Court of Cassation.
  • 34  “Préalable à toute négociation,” a statement of the union coordinating committee CFDT-CFTC-CGT-FO- (...)
  • 35  This data should continue to evolve in 2013 following a management plan to reorganize the perimete (...)

30The negotiation took place, however, after the elections. According to a central union representative interviewed “having the negotiations prior to the election would have allowed unions, if they were not representative, to see what would happen to them. The management did not want this, as it would involve greater expenses.”30 This postponement of the negotiations had additional consequences following the legal decision of 10 November 2010. In that decision, the Court of Cassation decided that the works council and not the shop stewards determined the default perimeter for the designation of the union representative.31 In practice, this meant the disappearance of the workplace union representative at Gratuit, since the level of representation of the works council was the region.32 Following the elections, the management had most designations of union representatives annulled by the District Court,33 while proposing negotiations on a draft agreement which recognized “a separate institution understood in the sense of the works council,” as the perimeter for the designation of union representatives. In a joint statement in November 2011, the unions accused management of wanting to “suppress union representatives on all production sites so as to further isolate employees and remove unions from the workplace.”34 Despite the opposition of a majority of organizations, an agreement was signed on 7 June 2012 with FO, CGC and CFTC, which barely represented the required 30% relative majority (32.8%). Paradoxically, the application of the reform at Gratuit thus led to a restriction in union representation. In 2010, according to the firm’s social balance sheet, 393 union delegates had actually been appointed according to the rules in force at the time. According to the document signed in 2012, depending on the configuration of representative organizations and taking all organizations into account, the maximum number of delegates that could possibly be named had been reduced to 179.35

31The comparison of the two companies allows us to qualify the assumption that the reform would be a fulcrum for trade union development. On this point, the question of representative posts is crucial: the first step towards a commitment is usually obtaining a designated or elective representative position, because it offers legal protection. This is why the active units of the union are frequently limited to employees with union posts. In sectors with little job security, fear of employer retaliation is much stronger as it can be both more dramatic and less visible. However, the necessary electoral benediction for union representative [DS, délégué syndical] restricts the pool of nominations. It prevents using the post of union representative as a position for the recognized establishment of a union section, which the creation of the post of transitory union representative [RSS, représentant de section syndicale] only partially offsets: besides the fact that the latter office has fewer means available, it is allowed to establish unions only within those perimeters where an organization has not demonstrated its representativeness. Specifically, it is impossible for the CGT, FO, CFDT, CFTC and the CAT, all representative organizations at Distrib, to appoint union representatives in the workplaces in which they have not previously presented candidates in the elections. Paradoxically, only SUD-PTT, CFE-CGC or new outsiders can be legally established through the designation of transitory union representatives. At Gratuit, the contradiction was resolved in a manner unfavourable to union representation, the employer’s position being based on the legal decisions reducing the perimeter for the designation of a union representative to that of the works council. The centrality given to the criteria of the electoral base in determining representativity initially resulted in legitimizing a definition of trade union activity centred on social dialogue rather than on the workplace. In the direct distribution of printed matter sector, this change had the effect of moving the reality of the sector a little further away from the ideal scheme which dominated the branch compromise: the acceptance of a special exemption status for distributors by union negotiators was conditional on establishing monitoring bodies at the branch level (technical advisors) as well as at the unit level (IRP), the connection between these two levels to be ensured by the unions. In addition, the new legal framework leads to a closer link between the conditions of union representation and the collective agreements (pre-election protocols and agreements on union rights) which in each firm organize the perimeters of employee representation. It follows that the “effects” of the law are even less understandable outside the specific configurations for each firm, as illustrated by the paradox of a restriction of trade union presence in the firm assumed to be the most open to social dialogue while the unionization of the firm seen as the most anti-union was advanced.

Towards the Definitive Destabilization of the Branch Compromise?

32Do the new rules, while not necessarily strengthening trade union presence, nevertheless create closer representative links? The most direct and intended effect of the reform was that industrial democracy should become a substitute for the failure of union democracy as a means of increasing workers’ control over their representatives. In fact, control is less direct than indirect through the power of sanctions which gives voters the right of the non-renewal of outgoing candidates. It is therefore impossible to measure its effects in a single electoral cycle. But the extent of turnover among employees in direct distribution of printed matter, as opposed to the tendency to lengthen the term of office (it is now four years, not counting the extensions following the disagreements during the negotiations of the pre-election protocols), makes the introduction of such a constraint on the action of representatives highly unlikely. Moreover, the exercise of that control would imply that there is a “public sphere” (Habermas, 1993) enabling employees to become aware of the issues in the arena of social dialogue and to submit them to a reasoned critique. Has, for example, the new emphasis on elections given a higher profile to the controversial question of the status of distributors? Has this enabled distributors to choose among the unions with full knowledge of their positions on the issues? The possibilities of voters taking up such protest issues depend on how union players themselves build their campaign. In consulting union members and their election material, the status of distributors is at best one question among others, and probably the least concrete. It appears as the background against which more immediate issues are played out: the transformation of the profession linked to mechanization, working conditions and low pay. In addition, the direct impact of the works council elections primarily concerned social and cultural activities for the employees, and the campaign centred in part on this point. These limits are written into the separation between the procedures that organize social relations of representation and the questions to which they are supposed to respond. On the one hand, there is the tangible function of the works council, the framework for the organization of elections (the firm or the workplace) and, on the other hand, the questions of social dialogue, the most decisive of which are decided on the branch level. Insofar as elections, which now determine the relevant trade union landscape, take place in firms around multiple issues, the bilateral arena continues to work at “limiting the space for the circulation of controversy and debate” and to differentiate “responsible” actors, who are integrated into this debate, from those who are “irresponsible” and outside it (Henry, 2005).

  • 36  A first text, published in 2007, was denounced by SUD-PTT and suppressed by the State Council in a (...)
  • 37  The Council of State ruled in SUD’s favour recalling that only a law could allow a system of labou (...)
  • 38  “Temps de travail : le gouvernement passe en force !” FILPAC-CGT leaflet, July 2010.

33It is thus the legal disputes rather than the election campaign that has contributed to publicizing the question of the status of distributors, especially in denouncing the decrees of the Ministry of Labour, which had twice sought to juridically secure the system of préquantification.36 But these actions in the legal field, while they expose the precarious condition of the distributors, have not led to a clarification of the union positions. At the same time that they go along with the criticism of the status of the distributors at the level of the firm, the union representatives at the branch level have never completely dissociated themselves from it on the bilateral level. Thus, the only union signatory of the CCN which denounced the decree of 2010 (FO) did not do so in the name of the illegality of the measures (unlike SUD), but rather in the name of the “misapplication” of “good rules.”37 As for the CGT, while it now declared in its leaflets to favour a return to the effective control of working time based on a system which is “self-declarative or which controls work and is activated by [employees] at the beginning and end of their shift”,38 it has not attempted up to now to denounce the CCN nor to attack the decree before the Council of State. While the decision of the high court symbolically reinforced the position of SUD and the CGT, leading FO to advance the idea of an effective control of working time by geolocation, it has not meant an end to negotiations to amend the CCN in the same framework of a juridical exemption. The SDD went as far as discussing the need to change the law, a possibility that is not rejected by some of the unions. The decision of the Council of State is still too recent and there may still be shifts in the position of the unions. The change in the parliamentary majority and with it, a shift of labour policies, could also affect the balance of power in the short term.

34Beyond these economic uncertainties, the reform of trade union representation, however, could have the effect of undoing the structural tension by permanently destabilizing the branch compromise. At the branch level, the former system of representation particularly practiced “social partnership”: it combined screening by the state (the irrefutable presumption of representativeness of the five “historical” confederations) and mutual recognition (a branch is established through the signature of a CCN, and the actors, particularly the employers, are established for signing a CCN). However, by basing union representativity at the branch level on the vote of the employees rather than the recognition by the employers’ representatives, the new system partially disconnects representation from the logic of social partnership, while linking the bilateral arena to the firms. The likely evolution of the trade union representativity landscape at this level could jeopardize the compromise which is constitutive of the branch and which is written into the CCN. As we have seen, the acceptance of an exceptional status for distributors had in return involved a strong institutional recognition of the unions. However, the new rules of representation tend to dissociate the terms of this exchange: on the basis of representativeness conquered at the polls, union actors such as SUD-PTT who reject the exceptional status of distributors will claim the benefits of trade union rights in the CCN. At the same time, those union actors whose representativeness is threatened may be encouraged to reconsider their attitude towards the system. Up to now, the social partners have turned away from such fateful ordeals. In the aftermath of employee elections at Distrib (see table), SUD-PTT questioned the social partners, demanding to be invited to bilateral meetings and simultaneously asked the General Directorate of Labour to initiate a study of representativeness. Its request went unanswered and was renewed in the fall of 2011 following new elections at the IRP at Gratuit. Based on the 2008 text which provides that the list of representative organizations is to be established at the end of an election cycle, the SDD and those unions which were considered representative under the earlier system postponed their decision until 2013. Meanwhile, the CGC’s loss of representativity at Distrib could provoke a chain reaction resulting in it losing its categorical representation at the level of the entire branch.

Table: Election Results for Union Organizations in the Branch

Table: Election Results for Union Organizations in the Branch

* AM [agents de maîtrise]: supervisory personnel.
Reading note: Results in bold type are relevant for legal representation.
The representation by category of the CFE-CGC at Distrib was annulled by the Court of Cassation. The results have been recalculated following the annulation of the elections in the electoral college of employees in Burgundy Rhone-Alpes (BRA). New elections were held only in this electoral college in May 2012. The presence of the decimal number of votes counted for FO, UNSA and SUD-PTT at Gratuit is explained by the fact that these unions presented a single list at BRA. The votes were split according to a specified key distribution decided by the organizations. On the branch, the results are only indicative because they do not cover all employees covered by the CCN.
Source: Calculated by the author based on the results compiled from each firm.

* *

35Beyond the original question –does the 2008 law help tighten the links between employees and their unions?– this text questions a central motif in the arguments made in justification of the reform, but also in most analyses of trade unionism: is the problem of unions really one of their excessive institutionalization? Contrary to common readings that combine “institutionalization” and “de-unionization”, a return to the origins of the direct distribution of printed matter sector demonstrates that institutionalization also produces unionization. Far from cutting off the trade unions from their base, the launching of a process of institutionalized negotiation “from above” at the branch level resulted instead in the rapid establishment of a union presence “on the ground” in firms and workplaces. It forced the unions to mobilize workers in the sector and in return weighed on the sectorial negotiations by branch and firm. But for unions to thus be established as representatives of the distributors, it was necessary that the forms of work of the distributors themselves had to be changed. The establishment of the CCN has reconfigured the relationship of the distributors to their own work as well as their relation to trade unionism. Observing union conditions under the microscope of the formalization of labour relations allows us to consider the institutionalization of trade unionism in its contradictory dynamics. That labour relations are “concretized by the law” does not mean they are petrified by it. The law does not “cause” practice, it directs it as “one of the references on which to build interactions” (Lascoumes, Serverin, 1995, p. 165). Legal representativity cannot therefore be considered as an external variable which acts on the social representativity of trade unions. Rather, it is a resource that actors –trade unionists themselves, but also employers, government officials, etc.– mobilize through multiple channels.

36Such a perspective allows us to study the impact of the reform of trade union representation in greater nuance, without yielding to a priori praises or denunciations. The mobilisation of the rules of representation in the two firms shows that the effects of the reform in terms of unionization proved to be ambivalent. For the effects of the new system of representation to de demonstrated, it had first to be internalized in the specific conditions of each firm’s configuration. It thus seems that it does not necessarily furnish the unions with new resources. It may even involve having to do better with the same, if not fewer, resources. Far from being neutral, the procedure which derives union representativeness from employee elections establishes the union actor on the basis of the arena of social dialogue. In the case of direct distribution of printed matter, this generates new constraints which are different in each firm, in terms of the establishment of unions in the workplace. Thus the law acts as an indicator of the fragility of unions rather than as a lever to strengthen them. While failing to ensure this strengthening of trade unionism, does the reform of representativity lead to greater control by the employees over their representatives? We have seen that the reform has already begun to undermine the compromise which was at the basis of the constitution of the branch. But the destabilization of the branch organization is not subject to the employees’ arbitrage in support of organizations that are hostile to it. Nor is it subject to heightened control of officials by their constituents, or of more transparent collective discussions of issues in the industry: the new representation rules do not imply any opening up of issues for deliberation. They provide, however, an unexpected development related to new legal mechanics, as they connect the previously disconnected designative procedures.

  • 39  Another contribution to the critique of institutionalization, the ethnographic method, is proposed (...)
  • 40  … or not: thus Rosanvallon (1998, p 247). He concluded his discussion by pointing out this contrad (...)

37The new rules of representativeness therefore respond to the “problem” of the institutionalization of unions even less, to the degree that this diagnosis is poorly formulated. It seems more appropriate to consider the reform as a means of reconfiguring union “institutionality.” This reformulation puts an end to the strictly negative vision underlying the theory of institutionalization39 suggesting that there might be a living unionism,40 free from any petrified form and outside of the institutions and the law. This is also the limit of a critical analysis based on the Marxism of Althusser, which relies on a base / superstructure dichotomy (Higelé, 2012). The issue is not so much trying to “oppose” or “limit” the institutionalization of trade unionism as to distinguish the different forms of “institutionality” running through it. This could, with Michels (1971), set the task of identifying those positions likely “to strengthen and stimulate the individual’s intellectual capacity for criticism and control” (p. 301), in order to strengthen what he called “democratic tendencies” against the “oligarchic tendencies.” Such a perspective would also operate the junction with the sociologically more robust definitions of institutionalization, which do not reduce the concept to its legal dimension but rather consider all the norms guiding social practices.

Haut de page


Adam G. (1983), Le pouvoir syndical, Paris, Dunod.

Andolfatto D., Labbé D. (2006), Histoire des syndicats (1906-2006), Paris, Seuil.

Andolfatto D., Labbé D. (2009), Toujours moins ! Déclin du syndicalisme à la française, Paris, Gallimard.

Bergounioux A. (1975), Force Ouvrière, Paris, Seuil.

Béroud S., Bouffartigue P. (dir.) (2009), Quand le travail se précarise, quelles résistances collectives ?, Paris, La Dispute.

Bévort A., Jobert A. (2008), Sociologie du travail : les relations professionnelles, Paris, Armand Colin.

Borenfreund G., Souriac M.-A. (dir.) (2008), Syndicats et droit du travail, Paris, Dalloz.

Bourdieu P. (1984), « La délégation et le fétichisme politique », Actes de la recherche en sciences sociales, n° 52-53, pp. 49-55.

Cartier M. (2003), Les facteurs et leurs tournées : un service public au quotidien, Paris, La Découverte.

Castel R. (1995), Les Métamorphoses de la question sociale : une chronique du salariat, Paris, Fayard.

Célérier S. (2012), « Le salariat dans la chair. Ambivalences du tâcheronnat dans les industries de viande de volaille », in Cingolani P. (dir.), Un travail sans limites ? Subordination, tensions, résistances, Toulouse, Érès, pp. 81-100.

Denis J.-M. (2008), « Conventions collectives : quelle protection pour les salariés précaires ? Le cas de la branche du nettoyage industriel », Travail et emploi, n° 116, pp. 45-56.

Didry C. (2002), Naissance de la convention collective. Débats juridiques et luttes sociales en France au début du 20e siècle, Paris, EHESS.

Duclos L., Mériaux O., (2009), « Métamorphoses de l’institution corporative et crise des politiques de l’intérêt », in Duclos L., Groux G., Mériaux O. (dir.), Les Nouvelles dimensions du politique. Relations professionnelles et régulations sociales, Paris, LGDJ, pp. 33-53.

Friot B. (2012), L’enjeu du salaire, Paris, La Dispute.

Grollier S. (2010), « Quelle syndicalisation des travailleurs de l’intérim ? », Savoir/Agir, n° 12, pp. 27-33.

Guillaume C., Pochic S. (2009), « La professionnalisation de l’activité syndicale : talon d’Achille de la politique de syndicalisation à la CFDT ? », Politix, n° 85, pp. 31-56.

Habermas J. (1993), L’espace public : archéologie de la publicité comme dimension constitutive de la société bourgeoise, Paris, Payot.

Henry E. (2005), « Militer pour le statu quo. Le Comité permanent amiante ou l’imposition réussie d’un consensus », Politix, n° 70, pp. 29-50.

Higelé J.-P. (2012), « Les formes de la délibération interprofessionnelle : le sens du dialogue », Sociétés contemporaines, n° 86, pp. 85-111.

Howell C. (1998), “Virtual Trade Unionism in France : A Commentary on the Question of Unions, Public Opinion, and the State”, in Chapman H., Kesselman M., Schain M. (eds), A Century of Organized Labor in France : a Union Movement for the Twenty-First Century?, St. Martin’s Press, pp. 205-212.

Howell C. (2009), “The Transformation of French Industrial Relations : Labor Representation and the State in a Post-Dirigiste Era”, Politics & Society, vol. 37, n° 2, pp. 229-256.

Lagroye J., Offerlé M. (dir.) (2010), Sociologie de l’institution, Paris, Belin.

Lascoumes P., Serverin É. (1995), « Le droit comme activité sociale : pour une approche wébérienne des activités juridiques », in Lascoumes P. (dir.), Actualité de Max Weber pour la sociologie du droit, Paris, LGDJ, pp. 155-177.

Michels R. (1971), Les partis politiques : essai sur les tendances oligarchiques des démocraties, Paris, Flammarion (1re éd. 1914).

Mischi J. (2011), « Gérer la distance à la “base”. Les permanents CGT d’un atelier SNCF », Sociétés contemporaines, n° 84, pp. 53-77.

Offe C. (1985), Disorganized capitalism. Contemporary transformations of work and politics, Cambridge, The MIT Press.

Offerlé M. (1998), Sociologie des groupes d’intérêt, Paris, Montchrestien.

Pélisse J. (2003), « Consciences du temps et consciences du droit chez des salariés à 35 heures », Droit et société, n° 53, pp. 163-186.

Pélisse J. (2009), « Judiciarisation ou juridicisation ? Usages et réappropriations du droit dans les conflits du travail », Politix, n° 86, pp. 73-96.

Piotet F. (dir.) (2009), La CGT et la recomposition syndicale, Paris, Presses universitaires de France.

Rosanvallon P. (1998), La question syndicale, Paris, Hachette (1re éd. 1988).

Salais R. (2008), « Conventions de travail, mondes de production et institutions : un parcours de recherche », L’Homme et la société, n° 170-171, pp. 151-174.

Tallard M. (2004), Action publique et régulation de branche de la relation salariale, Paris, L’Harmattan.

Thomas A. (2011), « Universitaires engagés et nouveaux cadres syndicaux aux États-Unis : une alliance pour faire face au déclin des syndicats ? », Genèses, n° 84, pp. 127-142.

Tournay V. (2011), Sociologie des institutions, Paris, Presses universitaires de France, coll. « Que sais-je ? ».

Vatin F., Bernard S. (dir.) (2007), Le salariat. Théorie, histoire et formes, Paris, La Dispute.

Haut de page


1  “Le dialogue social sera toujours un combat,” interview with François Chérèque, Challenges, n° 117, 27 March 2008, pp. 66-68.

2  “Négociation sur la réforme de la représentativité. État des lieux et propositions de la CGT. Fiche 3. Une représentativité des organisations syndicales administrée,” accessed 28 April 2008 on the CGT web site.

3  Legislative record on the law n° 2008-789 of 20 August 2008 on renewal of industrial democracy and reform of working time –Explanatory Memorandum. Légifrance website, accessed 5 June 2009:;jsessionid=D99E2D863E86CC701892AF7F0435223E.tpdjo

4 Box added to the English translation.

5Structural reorganization, targeted objectives in increasing membership, targeting the underrepresented sections of the wage earners, renewal and diversity of trade union leaders, etc. These questions were formulated by the CFDT from the mid-1980s (Guillaume, Pochic, 2009) and by the CGT in the early 2000s (Piotet, 2009).

6 The company names have been changed, although the almost duopolistic structure of the branch makes this operation artificial.

7For a summary in French of this literature, see Thomas, 2011.

8  This point of view is quite far from the conceptualizations developed around the dynamic and pluralistic field of the sociologies of institutionalization (Lagroye, Offerlé, 2010; Tournay, 2011).

9  A similar analysis can be found in Chris Howell’s theory of “virtual unionism” (1998, 2009) according to which “trade unions in France can be seen, in almost a feudal sense, as an estate in which the unions represent an interest –that of labor– by virtue of state sponsorship and public confidence rather than deep roots within the working class” (Howell, 1998, p. 209). [NB: footnote added to the English translation.]

10 Research done for the Dares (2010-12) conducted with Sophie Béroud, Jean-Michel Denis, Marnix Dressen, Maïlys Gantois, Cécile Guillaume and Donna Kesselman. Direct distribution of printed matter (DD) is only one sector among others covered in this survey.

11Interview, 24 January 2012.

12  More specifically, there was a union presence in some companies, but it was tolerated to the extent that it accepted the logic of piece work, as evidenced by some company agreements negotiated in the early 1990s. Otherwise repression was the norm.

13  Appeals to the labour tribunals regularly led to the reclassification of the distributors’ contracts to permanent full-time contracts (CDI). The problem of the distributors’ employment contract has repeatedly been pointed out by labour inspectors or in the courts, such as the district courts, starting from the related issue of employee representation (this in fact is determined by calculations using full-time equivalents).

14  Interview, 24 January 2012.

15  Gratuit was created in 1987 as a subsidiary of La Poste [the public Postal service] offering firms direct marketing services. It extended its activity to the distribution of unaddressed advertising by merging with a distribution firm bought from a media group which was a rival to the group which owned Distrib.

16  This is particularly true of the CGT federations of the postal service and of book publishing and distribution.

17  Modulation refers to the potential created by a collective agreement to vary the workweek without resorting to overtime if the annual average weekly hours do not exceed the time specified in the work contract. This device has been adopted in more than half of the RTT (reduction of working time) agreements between 1998 and 2000 (Pélisse, 2003).

18  The FAPT is the Federation of Postal and Tele­communications Activities; the FILPAC is the Federation of Workers in the Book, Paper and Communication industries.

19  Interview, 12 April 2012.

20  Interview, 24 January 2012.

21Source: the forum of the USDDAP:, accessed 4 April 2012; access reserved to identified users of the forum.

22  Interview, 7 December 2010.

23  In the context of the liberalization of the postal market, Distrib decided at the time to invest in the sector of addressed mail. In 2006, the company became the first private operator to acquire a “postal license.” Although Distrib finally decided not to compete with the Post Office in a market which has been fully liberalized since 1 February 2011, the company participates, through the Union of Postal Operators, in the negotiations leading to a collective agreement for postal activities.

24  The calculation of such remuneration is further defined by the introduction of computer software.

25  This little known organization boasts an existence since 1953. It has a website:

26  Interview, 7 December 2010.

27  SUD-PTT, which is developing a group strategy, gained a foothold in the direct distribution of printed matter when Gratuit, owned by La Poste, entered the distribution market. The union was able to impose its representativeness in Gratuit despite the opposition of the company and could present lists in the first round of professional elections in 2007.

28  Indeed, if the union represented by an RSS has not reached the threshold of representation after the elections, another employee must be designated RSS and the status of protected employee disappears after six months.

29  Interview, 18 January 2012.

30  Interview, 5 October 2012.

31  In several judgments of 10 November 2010, confirmed by a judgment of 18 May 2011, the social chamber of the Court of Cassation reversed legal precedent: it aligned its definition of the concept of a distinct entity, as the framework for the designation of a union representative, on that formulated by the State Council, that is, a separate institution understood as corresponding to the framework of the election of the works council and not to a specific workplace.

32  While the works council is national at Distrib, the company agreement which organizes the designation of union representatives by work place of at least 50 full-time equivalent employees is still in force.

33  Only four judgments were in favour of unions and will have to go to the Court of Cassation.

34  “Préalable à toute négociation,” a statement of the union coordinating committee CFDT-CFTC-CGT-FO-SUD at Gratuit, November 2011.

35  This data should continue to evolve in 2013 following a management plan to reorganize the perimeters of the company.

36  A first text, published in 2007, was denounced by SUD-PTT and suppressed by the State Council in a judgment of 11 March 2009. A second text, adopted in 2010, was denounced by SUD and FO and again suppressed in a decision of the State Council on 28 March 2012.

37  The Council of State ruled in SUD’s favour recalling that only a law could allow a system of labour remuneration to be exonerated from monitoring real work time.

38  “Temps de travail : le gouvernement passe en force !” FILPAC-CGT leaflet, July 2010.

39  Another contribution to the critique of institutionalization, the ethnographic method, is proposed by Julian Mischi (2011) who shows that at the “base”, the relationship between employees and union representatives is less clear cut than it is a permanent trade-off between distance and proximity.

40  … or not: thus Rosanvallon (1998, p 247). He concluded his discussion by pointing out this contradiction: “Trade unionism is torn between sociological legitimizing, now insufficient, and an impossible legitimization of a political nature.”

Haut de page

Table des illustrations

Titre Table: Election Results for Union Organizations in the Branch
Légende * AM [agents de maîtrise]: supervisory personnel.Reading note: Results in bold type are relevant for legal representation.The representation by category of the CFE-CGC at Distrib was annulled by the Court of Cassation. The results have been recalculated following the annulation of the elections in the electoral college of employees in Burgundy Rhone-Alpes (BRA). New elections were held only in this electoral college in May 2012. The presence of the decimal number of votes counted for FO, UNSA and SUD-PTT at Gratuit is explained by the fact that these unions presented a single list at BRA. The votes were split according to a specified key distribution decided by the organizations. On the branch, the results are only indicative because they do not cover all employees covered by the CCN.Source: Calculated by the author based on the results compiled from each firm.
Fichier image/png, 14k
Haut de page

Pour citer cet article

Référence papier

Karel Yon, « Work Representation and Union Representativeness on the Frontier of Wage Earning: », Travail et Emploi, Hors-série | 2013, 83-98.

Référence électronique

Karel Yon, « Work Representation and Union Representativeness on the Frontier of Wage Earning: », Travail et Emploi [En ligne], Hors-série | 2013, mis en ligne le 01 janvier 2014, consulté le 17 janvier 2018. URL : ; DOI : 10.4000/travailemploi.6301

Haut de page


Karel Yon

Ceraps (UMR 8026) –Université de Lille 2 ;

Articles du même auteur

Haut de page

Droits d’auteur

© La documentation française

Haut de page
  • OpenEdition Journals