Les lois Auroux de 1982 ont impulsé, en France, un mouvement de décentralisation des relations professionnelles. On peut effectivement constater, au cours de la dernière décennie, un développement notable des accords d'entreprise. Aux yeux de l'observateur, ce mouvement ne doit pas occulter, malgré tout, ni la pérennité des accords de branche, ni l'importante variable des accords d'entreprise selon les secteurs et les tailles d'entreprise ni, enfin, le fait que la dynamique n'est pas nécessairement la même d'un thème de négociation à l'autre. Une enquête menée dans deux secteurs (électronique et santé) permet de spécifier les limites de la décentralisation en matière de négociation sur l'emploi.
The 1982 Auroux laws in France served to bring unions and enterprises closer together. Whereas historically labour relations had been based on confrontation between a labour movement in which the "revolutionary" spirit had prevalled for a long time, and on a management which had carefully avoided the introduction of collective bargaining in firms, the 1980s were characterized by a significant move towards decentralization of negotiations to enterprises and establishments. In the first part of this article, it is shown that, in fact, the Auroux laws contributed to this change: in 1981, barely 1500 enterprise — level agreements were registered in France; in 1992 there were 6370 such agreements. Nevertheless, it is worth pointing out three complementary phenomena.
First, the law related to employees' freedom of expression has had limited success: according to a recent report, in reality groups for self— expression seldom function and do so poorly. Weak union involvement and management participation policies (quality circles) have undeniably contributed to weakening the interest in and importance of these groups. Secondly, the growth of enterprise — level collective bargaining has not resulted in a decline in industry — wide agreements (their number has remained stable throughout the decade); nor is their growth homogeneous (such agreements are increasingly the reality in large industrial sector firms). Finally, the dynamic varies according to the objects of negotiation. Hence, the following observations: a certain importance given to salaries in negotiations at both the industry and enterprise levels, industry — level negotiation giving way to enterprise — level negotiation regarding work time; a renewal of energy though the volume of agreements remains limited; and interprofessional industry — and enterprise — level negotiations on the issue of employment.
The aim of the second part of the article is, precisely, to evaluate the impact of negotiations on the management of employment relations in two sectors that are at first glance opposites: this interest is based on a study conducted by the two authors in which they examine the management of redundant personnel in a sector in crisis (electronics) and the management of labour shortages in a protected public sector (health care).
Firstly, during recent years programs have multiplied in the electronics sector to deal with layoffs. In order to institute preventive action, in 1990 and 1991, the social actors established joint committees on employment whose mission was to make forecasts of employment in the industry in order to formulate ways to deal with the effects of fluctuating employment (forecasts of movement of the workforce, establishment of special training programs, and so on). In reality, for reasons related as much to the desire of employers to bring back the management of employment relations to the level of the enterprise as to the difficulty in adjusting trade union structures to the reality of a specific industry (electronics and not the metallurgical industry, which remains the usual framework for negotiations), these committees have scarcely been effective.
Furthermore, it was observed that there is a real asymmetry between the enterprise (strategic level where important decisions about employment are made) and the establishments. Within the latter, in spite of rights given to union organizations, they, as well as the local administrations which are sometimes just as powerless as themselves, can only negotiate the consequences of strategies elaborated at the group and enterprise levels. Thus one of the major limitations of the policy of decentralization of industrial relations can be clearly seen here. As regards the health care sector, its System of industrial relations is structured around a statutory logic regulating recruitment, remuneration and employment of permanent personnel. Since the hospital reforms of July 31, 1991, with a concern for democratization similar to the one which was behind the Auroux laws, dialogue between all actors can take place within a series of councils who are responsible for both investment programs and internal regulations, etc. The fact remains that in the case of electronics, decisions about the volume of employment are made at the central level (the Ministry): the volume of employment is thus a fact that the actors cannot negotiate because it is the result of political arbitration and certain financial constraints. This weak union influence on determining factors of employment policy was one of the reasons, among many others, for the emergence of new forms of collective action during the second half of the 1980s. During the conflicts that characterized this period, nurses and auxiliary nurses went beyond union and institutional channels of communication in aid of sectorial organizations. This was significant each time, the conflicts having a direct impact on topics that normally cannot be debated within the framework of joint committees: such is the case of the level of employment, re-establishment of indexation and recruitment terms. This is amply demonstrated in the two sectorial studies that were conducted. These two case studies tend to show that in France the movement towards decentralization not only increases the tendency to negotiate salary changes, management of work time, and so on in a compartmentalized way, but also makes it more difficult for union organizations to negotiate employment policy determinants at the enterprise level. Furthermore, beyond differences between the private and public sectors that many researchers have rightly emphasized, it appears to us that the two sectors examined continue to have certain common characteristics which reflect quite well the difficulty experienced by the current French trade union movement in adapting itself structurally to the new social and economic context.