CAIRN-INT.INFO : International Edition

1Among recent changes in public policy, one of the most remarkable shifts has been the increasing importance attributed to scientific knowledge and expertise in decision-making processes. It is true that the trend of relying more and more on a scientific and technical approach to problem solving is not entirely new; in fact, it attests to long-term evolutions in government and public policy. [1] Moreover, although such developments might initially seem limited to highly scientific and/or technical questions, they nevertheless also perfectly illustrate a growing trend whereby public policies are predominantly legitimized through recourse to expertise. [2] However, the evolution of scientific expertise as used by public figures is also the result of more specific changes, highlighted in particular by studies on public health and collective risk. [3]

2It was the introduction of policies to regulate and manage public health risks that first prompted studies on the role played by scientific expertise in public decision-making. These preliminary investigations articulated the queries that still inform today’s research, whether it be Michel Callon and Arie Rip’s article on the hybrid nature of expertise [4] or the research done on occupational and environmental risk regulatory agencies which addresses the boundaries of expertise. [5] In a number of ways, the question of boundaries is at the heart of expertise research, whether through an investigation of the boundaries between science and expertise, [6] between expertise and decision-making, [7] and between certainty and uncertainty, [8] or through the analysis of new institutional measures implemented in France during the last few years. [9] This trend is so strong that expertise is now in fact often defined via this notion of boundaries. [10]

3Nevertheless, by spending too much time investigating the boundaries or the relations between expertise and decision-making, or between experts and other key actors, one risks accepting uncritically the validity of categories such as “experts” or “deciders”, even though these very categories need to be questioned. One also runs the risk of accepting that the “context” in which these expert investigations unfold is itself a permanent feature and that it is hardly influenced, if at all, by these processes. In short, although the tendency of research to autonomize the expertise process in order to better comprehend and explain it has increased our understanding of the internal workings of expert proceedings, [11] in reality such research also tends to underestimate the extent to which the expertise process is itself limited or contextualized by social factors that are external to the logic employed by groups of experts. The framework imposed by the stable definition of any problem limits the sorts of questions posed to the experts and the manner in which they can formulate their opinions, and is thus insufficiently considered by research on expertise. [12] Moreover, the unequal distribution of certain resources among different categories of participants or among social groups, and the subsequent effect of this unequal distribution on the structuring of knowledge and of expert proceedings, are rarely taken into account by contemporary sociological studies of science. Focused on laboratory life [13] or fluid alliances and networks, [14] these studies only marginally address the question of power relations. And yet, re-emphasizing the unequal distribution of certain resources allows one to gain new perspectives on the production (or, in some cases, the absence) of the scientific knowledge which characterizes certain disciplines, and thus on the expert processes connected with these forms of knowledge. [15]

4Building upon prior research that attempted to promote a new political sociology of science, [16] I hope to illustrate the value of undertaking an analysis of expertise and knowledge production in relation to the public policy sector within which such expertise and knowledge is integrated, and the inequalities which characterize this sector. While it is in some senses a continuation of previous studies which sought to articulate the relationship between power relations and the production of systems of representation, [17] this article will also insist on the uniqueness of these processes when they call upon scientifically informed expert opinions. This study thus distinguishes itself from other work which, in the course of analyzing public policy, has primarily focused on its cognitive aspects. [18] These aspects appear on the contrary to be quite strongly correlated with the power relations structuring interactions between groups of key players at different levels of the elaboration of public policy, as has been clearly shown by research on the notion of public policy instruments. [19] I will start my analysis by focusing on occupational health, a sector that often exaggerates certain tendencies observed in other areas of public policy. In particular, this sector is characterized by the existence of contradictory representations of the problems and goals of public policies, as advocated by different social groups. In this context, scientific expertise can never completely rise above these dichotomies. Nevertheless, expertise has been profoundly reformed these past few years by the creation of a Department of Occupational Health (DST – Département santé travail) within the framework of the Institute for Public Health Surveillance (InVS – Institut de veille sanitaire), and by shifts in responsibility for occupational health issues, initially exercised by the French Agency for Environmental and Occupational Health Safety (AFSSET – Agence française de sécurité sanitaire de l’environnement et du travail) then, since 2010, by the French Agency for Food, Environmental and Occupational Health Safety (ANSES – Agence nationale de sécurité sanitaire de l’alimentation, de l’environnement et du travail), which is the result of a merger between AFSSET and the French Food Safety Agency (AFSSA – Agence française de sécurité sanitaire de l’alimentation).

5By granting labor administration supplementary resources, the transformations that affect the production of knowledge and expertise in the sphere of occupational health constitute the main avenue of change in this area of public policy. [20] Nevertheless, despite the emergence of these new resources, what most characterizes this domain is in fact the inertia of power relations between key players, which permanently hinders any possibility for significant change. Thus, I first show that changes affecting public health policy lead to a number of transformations in the realm of occupational health, most markedly in the domain of knowledge and expertise production. I then examine how expert investigations are becoming increasingly specialized, and the resulting transformations in the power relations between the different stakeholders who contribute to the definition of these policies, as well as the limits of such transformations. [21]

Occupational health “belabored” by public health

6The paradoxes associated with occupational health policies stem from their two-way affiliation. Since such policies have an effect on general public health, they are believed to belong to the public health sector. Whereas occupational health policies were originally considered to be policies of social negotiation, pertaining to professional relationships within the workplace, the trend towards considering them as part of public health has become more and more pronounced since the 1990s. [22] Even though this connection with social relationships is no longer exclusive, such policies still rely on close co-operation between unions and employer representatives. Decisions regarding occupational health are first and foremost the result of negotiations between social partners, while the government often limits itself to organizing these discussions and translating their conclusions into legal norms. These characteristics explain why, for a long time, the very notion of expertise was practically unheard of in this sphere of public policy. However, since the 1990s occupational health has gradually become a question of general public health, which has provoked a series of reexaminations of previously established conclusions.

The long absence of science and expert knowledge from occupational health

7Aligning occupational health policies with social relations produced specific modalities of expertise and of association (or, at times, the non-association) of knowledge with public decision-making. This situation remained in effect until the asbestos crisis of the mid-1990s. As noted above, decisions regarding occupational health are first and foremost determined as a function of the power relations established between unions and employer representatives (for example, when consulted within the framework of the Steering Committee on Working Conditions [COCT – Conseil d’orientation sur les conditions de travail, formerly the CSPRP – Conseil supérieur de la prévention des risques professionnels], an advisory body of the Ministry of Labor). It is during these types of social negotiations that expert scientific knowledge may be mobilized. As the public health facets of occupational health are not taken into account as such, no investigation relative to the public health consequences of such decisions is undertaken.

8This situation effortlessly perpetuates itself, as occupational health regulations have for a long time been considered peripheral within the Ministry, whose primary focus is issues of job creation and unemployment. [23] Although working conditions remained a priority on the administration’s agenda and in the minds of politicians until the first half of the 1970s, they were quickly overshadowed by the ensuing economic crisis and mass unemployment. [24] Few political initiatives were taken in the sector of occupational health, which was on the contrary characterized by acceptance of the status quo or, rather, a total lack of decision-making. [25] When scientific facts sporadically made an appearance in negotiations, they were attributed the same importance as other kinds of arguments; no specific procedures were used to differentiate the scientific elements. A certain “blurring” of the lines between scientific and social dimensions was thus observed during negotiations. The expertise provided in such cases was generally informal and anecdotal, since in general, hospital practitioners specializing in occupational pathologies were individually contacted to provide the administration with the necessary data. In this arrangement, the main sources of information regarding occupational hazards were employers – in other words, those who were the most capable of financing scientific expert investigations or commissioning the research capacities of large industrial groups, but who, above all, had exclusive access to industrial sites and procedures as well as primary data. [26] The central role played by industrialists was particularly evident in the extreme case of asbestos, which witnessed the creation of committees combining scientific expertise and the protection of industrial interests; e.g., the French Asbestos Permanent Committee (CPA Comité permanent amiante). [27] However, the importance of this role can also be seen in the continued pressure exerted on researchers at the National Research and Safety Institute (INRS Institut national de recherche sur la sécurité) whenever their work threatens industrial interests, as was the case with glycol ethers and aluminum. [28]

9Even more influentially, the role played by industrialists is above all evident in their ability to prevent the discovery of new information or even to produce certain forms of ignorance. [29] This cost-effective and invisible form of power makes the establishment of scientific knowledge on occupational pathologies quite difficult and contributes to the diseases’ continued neglect. Scientific information is thus patchy and incomplete, and does not allow us to measure the effects of work on the health of workers, and even less to estimate the workplace component of health inequalities across social groups. The only global figures on occupational hazards are those produced by health insurance companies, which only account for fully compensated occupational diseases and therefore greatly underestimate the levels of occupational risk. [30] Even though numerous public reports have confirmed that insurance companies’ figures are not representative of the broader effects of work on health, [31] these numbers continue to act as a smokescreen and to be used as misleading indicators during the development of new occupational health policies.

10This lack of data is one of the main reasons for stagnant policies and thus for the absence of new policies designed to better protect workers’ health. It is therefore at the heart of the power relations between key actors and largely helps to preserve the sector’s current public policy direction. As a result, this situation ends up conferring a very peculiar status on those initiatives that seek to produce new knowledge: as they emerge in a contentious sphere whose precarious equilibrium is maintained only by the absence of new data, such initiatives are immediately viewed as politically motivated attempts to destabilize the current configuration, independently of their authors’ original intentions. Occupational health initiatives are thus especially costly for the actors who spearhead them, as these individuals are opened up to sometimes virulent criticism from those who play a key role in defining policy direction or who occupy an “ownership” position – to use Gusfield’s term – with respect to these policies. [32]

11This explains in part why initiatives that aimed to generate a better understanding of occupational health sometimes originated in areas which were fairly distant from scientific concerns. In many cases, attempts at reform were more or less motivated by activist concerns, especially in the period following 1968, which was characterized by a number of social movements targeting working conditions. [33] During the 1970s, however, some scientists tried to develop epidemiological surveys based on problems flagged by union organizations.


“We wanted to do a proper study of mortality in the shipyards. The idea was to get corporate data, so we needed information from human resources. Theoretically, we could go through the CHSCT (Comité d’hygiène, de sécurité et des conditions de travail – Corporate Committee for Hygiene, Safety and Working Conditions) […] And then we realized that […] if we showed up at the company with a militant CGT member (Confédération générale du travail – French trade union organization), then we’d be sure to find the doors to management closed.
(Inserm epidemiologist, 20 February 2009)

13Even when they were less directly linked to social movements, these initiatives often had the dual aim of improving understanding of occupational health and of increasing the latter’s presence in public debate. An interview with a medical inspector who directed several studies in the field of occupational health and was one of the authors of Les risques du travail, [34] a mid-1980s publication that outlined a plethora of occupational risks, illustrates this clearly:


“I always thought that occupational health surveys were an essential tool for getting people to be aware of occupational health issues. […] My goal is to get out of this zone of social invisibility. To take any measures necessary – and possible – to go beyond this invisibility that we still haven’t overcome.”
(Medical inspector of working conditions, 14 May 2009)

15The exploratory surveys Conditions de travail and Sumer (Surveillance médicale des expositions aux risques professionnels – Medical Monitoring of Exposure to Occupational Hazards) were launched in 1978 and 1982, respectively, and would become the main sources of information regarding working conditions in France. [35] The first investigation relied on the Insee’s Employment survey and followed the latter’s statistical methods quite closely; the second was more innovative, as it used data from workplace physicians. The fact that these surveys were introduced by the Ministry of Labor confirmed the close attention now being paid to occupational health issues. Nevertheless, it was more the case that it was an opportunity for a few motivated individuals to fully devote themselves to these issues than it was illustrative of a genuine desire on the Ministry’s behalf to develop this type of survey, as this account of the beginnings of the Sumer study reveals:


“As always, the DARES was very concerned with employment policies, with salary levels […] And like little ants, we toiled away in our little corner. […] And so Labor Relations Management had left medical inspectors in charge of that. Sure, there were two or three people who were supervising, but really, they were completely swamped by other things… It wasn’t a central initiative coming from the administration, and I think that’s why it worked.”
(Statistician, DARES, 25 May 2009)

17In the same way, the fact that these studies have endured so long can be explained by commitment of the individuals directly involved, as well as by a certain degree of inertia on the part of the institutions in question. As Michel Gollac and Serge Volkoff explain on the subject of the Conditions du travail survey:


“Once set in motion, this type of instrument benefits from the inertia of its supporting structures; this inertia protects it from political, administrative and even scientific trends. At the same time, the publication of results increases the commitment of the individuals involved within the survey’s network and thus consolidates this network over time.” [36]

19The circumstances under which these studies took place show that until the 1990s, questions about working conditions and occupational health were rarely – if ever – addressed from a public health perspective. Rather, they belonged to the social sphere in the broadest sense, or sometimes even to the political domain. In this context, initiatives that sought to produce new instruments of knowledge were most often of activist origins (backed by political convictions or union membership), or were at the very least committed to raising social awareness of these issues. Consequently, until the 1990s, the field of occupational health functioned independently of public health expertise and the initiatives noted most often established statistical indicators based on existing models in the field of social policy.

Transformation in the wake of developments in public health

20The 1990s were characterized by a series of public health crises that revealed the increasing amount of media attention devoted to these questions, [37] as well as the difficulties the health administration (in particular Community Health Protection – Direction générale de la santé) had in dealing with them effectively. [38] In the field of occupational health, the public scandal around asbestos shone a crude spotlight on the absence of independent public experts and on the significant role played by employers in the negotiation of occupational health policies. [39]

21The redefinition of asbestos as a public health issue that threatened the general population led public figures to align their approach to dealing this problem with the predominant approaches used for other health problems and, at the same time, to distance themselves somewhat from the usual ways in which occupational health risks were dealt with. [40] For the first time, the Ministry of Labor based all of its decisions on expertise provided by a group it had officially commissioned from Inserm (Institut national de la santé et de la recherche médicale – National Institute of Health and Medical Research), [41] thereby giving itself the means to independently investigate the scientific aspects of the problem. This investigation and the fact that, soon after, several more expert investigations were requested from Inserm began to overshadow the role of clinicians, who had traditionally been an important source of information for the Ministry, and gave instead a larger role to epidemiologists and hence to the method of risk analysis. [42] This shift was accompanied by a reconsideration of the relationship between scientific expertise and decision-making processes. One example was the implementation, in 1998–99, of a charter that aimed to separate scientific expertise from social consultation during the process of creating or modifying the chart of occupational diseases recognized by the CSPRP. [43]

22These changes occurred during a time when public health administration was itself subject to important upheavals, in particular due to the creation of different agencies for health safety as decreed by the law of 1 July 1998. [44] Among the agencies created in 1998 was the Institute for Public Health Surveillance (InVS – Institut de veille sanitaire), which would become the main instrument for health monitoring used by the Ministry of Health and its Department of Occupational Health (DST – Département santé travail). [45] Even if during this period the asbestos crisis helped to raise awareness of occupational hazards, the creation of the DST did not seem like an obvious next step for those active in public health; in fact, it was not the subject of any initiatives planned by those who had instigated the reform of public health policies. Rather, it was future department managers who sought to convince these reformers to integrate occupational concerns into the creation of an institute which was, at the outset, primarily oriented towards public health (but which, for all practical purposes, excluded occupational issues). These future department managers were listened to attentively, as many were not exactly unknown in the domain of public health. For example, Marcel Goldberg was already well known as a university professor of epidemiology and a research director at Inserm.

23From the point of view of public health administrators and professionals – who are in general not particularly sensitive to occupational health issues – addressing occupational hazards was a way to expand their field of expertise, at the expense of groups outside the field who it was therefore easy to challenge (especially since occupational health specialists are generally not well known within the broader medical world). The only external players that could potentially be a source of conflict were university specialists in occupational medicine and the institutional bodies with existing links to the Ministry of Labor (in other words, actors outside the public health sector). Once past the lack of awareness of occupational issues, the inclination was largely to approach this area in the same manner as all the other risks addressed by the health system – that is to say, by relegating non-medical authorities to a merely peripheral role. This tendency was not appreciated by the traditional actors within occupational health, in particular by workplace physicians and hospital staff in charge of occupational pathology departments.


“People involved in occupational health were a little afraid of being swallowed up and no longer having an identity […] [They] were afraid they would be forced to integrate into public health. And anyways, the desire for this integration is still sort of present; you can see it quite clearly with some public health figures.”
(Professor of occupational medicine, today in charge of a public health agency, 16 January 2008)

25The National Mesothelioma Surveillance Program (PNSM – Programme national de surveillance du mésothéliome) was the first project launched by the Department of Occupational Health and confirmed the influence of the asbestos crisis in relation to contemporary changes in occupational health. [46] Under the aegis of one project, this program brought together both public health and occupational health teams, working with the relevant ministerial departments (DGT and DGS), thus bolstering the new department’s legitimacy. The transformation of expertise in the domain of occupational health thus first occurred as a consequence of the asbestos problem. With this grave issue, officials knew that they had no “margin of error”, and that their decisions would be closely monitored by stakeholders capable of alerting the media if necessary. The resurgence of interest in the asbestos issue also had an influence on more long-term changes regarding the role played by expertise in other occupational health discussions.

The production of new knowledge on occupational health

26Following the PNSM, the Department of Occupational Health implemented a number of different programs to address specific pathologies (musculoskeletal or mental disorders), specific economic sectors (the meat industry, the medical and paramedical sectors), or a more generalized cluster of issues, such as the programs attempting to classify all occupational diseases (MCP maladies à caractère professionnel). [47] Even if the launching of these programs fell under the purview of the InVS, they often relied heavily on teams who already had experience in the specific problem area.

27In the Pays de la Loire region, for example, several initiatives of this type were developed, including an epidemiological study on MSDs (musculoskeletal disorders) developed by a professor of occupational medicine who is today internationally recognized by the scientific community. Based on this existing expertise, the InVS proposed the establishment of an epidemiological surveillance program for MSDs, which attested to the complementary relationship between scientific research and epidemiological surveillance. In 2002 a “sentinel network for the epidemiological surveillance of MSDs’ was instituted in this region, originally just for a two-year period. [48] The success of this second large-scale program of epidemiological surveillance (the PNSM being the first) continued to confer legitimacy on the DST and also provided a model for future projects – for example, the occupational disease reporting program.

28As decreed since 1946, all medical doctors are required to report any physical harm linked to occupational activity. And yet, since 1976 the law has not specified to whom such reports should be made. [49] During the 1990s, several initiatives were taken to identify occupational diseases at the national level using data from regional occupational health inspectors. In the Pays de la Loire, this data collection relied on the commitment of workplace physicians to the development of occupational health initiatives. [50] Before becoming independent, this project was initially integrated within the MSD surveillance program. Thus, Department of Occupational Health officials did not start from scratch when committing to this project, as projects addressing occupational diseases already existed in the region. Nevertheless, by proposing a new, more systematic methodology with the more explicit goal of epidemiological and public health monitoring, the DST gave a new dimension to this data, which also benefited from new means of dissemination. While in 2003 the investigation of occupational diseases was tested in the Pays de la Loire, today it is undertaken in seven regions of France and seems on track to continue expanding.

29The InVS hence acts as a catalyst for teams ready to become involved with investigations in the Institute’s area of expertise or lead epidemiological research projects. But even if it initiates these surveillance programs, the InVS does so thanks to dedicated teams that already possess the relevant knowledge and know-how. When these programs were first launched, the department’s means were extremely limited and it was forced to function in tandem with external teams. With the gradual development of a larger team, the InVS has been able to pilot some programs itself: for example, Samotrace, a study monitoring mental health risks linked to work, that was created in 2006 and led by a doctor who was a member of the InVS, rather than an external researcher.

30Once designed and endorsed by the DST, an epidemiological study benefits from recognition that goes beyond the sole purview of workplace health specialists and extends to the domain of public health, administration and research. By granting greater visibility to the conclusions of these studies and reaching a larger public, the InVS helps to alter the power relations that structure occupational health policies, since one of the conditions enabling the persistence of current policies is in fact a lack of awareness regarding the effects of work on health.

31Changes in public health and its increasingly large area of overlap with occupational health have thus brought about important transformations in the field of occupational health itself. The creation as well as the sustainable nature of the DST can thus primarily be explained by the fact that it is becoming more and more difficult for occupational health specialists and activists to maintain their previous modes of operation. Obtaining the means to measure the effects of work on health and having them evaluated by experts operating somewhat independently within the framework of an agency are two of the concessions that these professionals had to make in order to ensure a minimum degree of consistency between public health and occupational health policies. The fact that the impetus behind the creation of the DST came from outside the occupational health sector and that it was more the result of individual attempts at reform rather than a veritable effort spearheaded by public health officials is quite revealing of the types of changes affecting occupational health. The occupational health sector cannot refuse to integrate these changes; nevertheless, in the process it is confronted with other modes of thinking that, on the contrary, seek to maintain the status quo by granting a key role, if not an exclusive one, to negotiations between social partners.

Is expertise increasingly essential – or increasingly “tamed”?

32The first few years after the turn of the millennium saw a more systematic organizational approach to expertise and knowledge production. Although it had mostly operated in the background up until that point, the DGT gradually began to take the lead in the acquisition of expert resources. Its goal was to develop an assessment tool that could be used more readily by both government and professionals (workplace physicians, risk prevention engineers) involved in the implementation of relevant policies. [51] In so doing, the systematization of previously observed changes has become clear, with a shift from experts solicited on a case-by-case basis to agencies specially designed to produce knowledge and expertise. The gradual formation of state expertise has not been without its own ambiguities, however. Although it altered the power relations between key figures, it did so not by allowing for an escape from prior relations of domination, but rather by merely changing the ways in which such relations may be updated.

From AFSSE to AFSSET: the integration of occupational health into AFSSET

33Although the creation and growing influence of the InVS’s Occupational Health department has allowed for an increase in the production of epidemiological data on occupational hazards and thus for potentially useful data to inform relevant policies, the fact that this department is firmly based in the public health and epidemiological sectors is at odds with the predominant logic behind occupational health. In addition, even if the DGT shares the job of heading the department with the DGS, the InVS is itself under the sole authority of the Ministry of Health, with the result that the DST does not appear to be an organ that can be directly deployed by the administration. This is especially problematic as, since the 1990s and the decade’s public health crises, tension has frequently existed between the heads of the DGS and the DGT, with some DGT leaders fearing that occupational health would fall under the DGS’s purview and thus be neglected. Such tensions are common and can particularly be seen with regard to health safety agencies. Structurally, the DGS tends to favor a reduction in the number of agencies: the larger the agency, the greater the control the DGS can wield over it. The DGT, on the other hand, favors the existence of a specific occupational health agency, or at least an occupational health and environment agency with greater autonomy to make its own choices. The development of expert agencies within the occupational health sector thus illustrates the power struggles between these different ministerial branches and the resulting instability. The creation of a Department of Occupational Health in 2005 within the French Agency for Environmental Health Safety (AFSSE – Agence française de sécurité sanitaire de l’environnement), which became the French Agency for Environmental and Occupational Health Safety (AFSSET – Agence française de sécurité sanitaire de l’environnement et du travail), revealed that Labor and Environment were proritized over Health. Conversely, the July 2010 fusion between AFFSET and the French agency for food safety (AFFSA) which resulted in the French Agency for Food, Environmental and Occupational Health Safety (ANSES Agence nationale de sécurité sanitaire de l’alimentation, de l’environnement et du travail) seemed to suggest that the health sector was regaining influence and importance. It is still too early to predict the consequences of this merger; as such, the discussion below will only include references to AFFSET during the period when occupational health started to become a main concern, i.e., from 2005 to July 2010.

34Starting in 2005, the occupational health sector began to rapidly consolidate within AFFSET, including the 2007 reworking of the organizational flowchart, which now grouped together environmental and occupational aspects of the same risk in each sector. However, the creation of AFFSET was not without its own problems. Pubic health agencies were already quite numerous, and some experienced overlap with AFFSET’s field of influence. Moreover, other organizations, such as the National Institute for Research and Safety (INRS – Institut national de recherche et de sécurité), also claimed specific expertise in the field of occupational health. The positional complications that accompanied AFFSET’s birth [52] are among the factors that explain the agency’s particularly rigid stance on a number of issues, including its support for the Afnor NF-X 50-110 standard used to structure its expert opinion production, and its avoidance of any ties that could appear to hinder the agency’s independence. [53]

35For the DGT, the creation of AFSSET was an important step towards the constitution of an independent source of expertise that could be utilized in the course of its routine operations. Although AFSSET is under the joint tutelage of three different ministries (Health, Environment and Labor), the DGT provided more than 50% of its financing and played a crucial role in its conception and launch. In addition, the agency’s operating methods facilitate its relationship with the DGT. For example, the fact that both workers’ and employers’ unions are present on AFFSET’S Board of Directors allows these organizations to be linked with expertise programs and informed of the latter’s progress, which consequently lessens the degree of suspicion sometimes observed between these organizations and the InVS’s Department of Occupational Health. On a disciplinary level as well, AFFSET’s expertise is broader than that of the InVS. Very quickly, AFFSET had access to expertise resulting from occupational health and engineering research, which could be applied with greater ease in a corporate setting and/or by professionals directly intervening in the workplace, such as local public health insurance agents.

36AFFSET’s creation led to a more precise definition of the connection between scientific expertise and decision-making, based on the model that existed in public health administration: namely, a relationship that exhibited a firm boundary between risk assessment and risk management, as well as between the production of expert knowledge and decisionmaking processes. Even if this is merely a reference standard, and in practice operations deviate significantly from the theoretical model, the assertion of such a strict separation is one of the elements justifying the growing independence of health safety agencies. [54] And yet this separation is far from possessing the same legitimacy in the occupational health sector that it has in the public health sector. It is sometimes seen as the DGS’s “trademark” and believed to be an attempt at applying public health expertise to occupational health issues. On the contrary, some actors argue that maintaining a certain level of permeability between the fields of risk assessment and risk management, made possible thanks to a lack of publicity about occupational health policies, is one of this sector’s fundamental elements.

Restructuring expertise to transform power relations: an acceptable risk?

37The current changes that expertise and knowledge production are undergoing have important effects on the power relations between the different groups of actors contributing to policy design. The government’s ability to support some of its decisions with proven scientific data is one way it boosts its influence. By relying on scientific knowledge, the state makes any further investigation rather more difficult, as any new inquiries would need to be undertaken on scientific grounds and thus utilize specific resources. Consequently, a greater degree of legitimacy is conferred on the government when in discussions with its partners. In this light, the question of implementing Occupational Exposure Limit values (OEL values) offers an interesting subject of analysis, as during the period studied one can observe a growing trend towards the “hardening” of the scientific apparatus which frames decision-making processes. Although the preparation of these values has been a focus for the Ministry of Labor’s attention since the 1980s, its concomitantly created groups of experts have always maintained a rather informal status, constituting just the first step in the internal investigation of these cases by Ministry services. The preparation of limit values was only formally recognized in 2005, with a preliminary scientific evaluation entrusted to AFFSET. Even if this process is still in its infancy (so far, AFFSET has only assessed a few toxic substances, and no decision has yet been made within this new framework), one can surmise that future decisions will emphasize the “scientific” nature of these values. [55]

38However, one must nuance this idea of the government’s growing influence relative to other players in two important ways. First of all, in the domain of labor relations, the pursuit of consensus among social partners informs many of the decisions taken. Even if one believes that the government might have much to gain from granting itself more latitude, in reality conserving the relative influence of other negotiating parties remains important, in order to not exacerbate social tensions and/or be forced to play the role of an arbitrator, something which labor administration has never wanted to do. An overly radical strengthening of expertise resources could paradoxically weaken labor administration, which derives a good part of its legitimacy from its ability to maintain good relations with all the social partners involved. Secondly, even if administrative figures are granted more power in the broadest sense, one cannot speak of an empowerment of the DGT as such, since this movement is accompanied by the redistribution of roles and of margins for manœuvre among the different actors involved. The legitimization of administrative action through expertise takes place at the cost of the break up of arrangements of state actors, some of whom consequently demand greater autonomy. Moreover, this fragmentation of groups of actors is occurring at a time when the rules that dictate their relations are becoming more and more public, thus making informal negotiations less frequent. It is nevertheless possible to imagine that having frequent recourse to public expertise will subsequently reduce the repercussions caused by the unequal resources of employer organizations and workers’ unions during negotiation and decision-making. The scope of this movement will depend, however, on the degree of autonomy that administrative agencies and actors will be able to preserve (or not) with regard to economic leaders.

39In addition to changes in power relations between key players, the actual modalities of their confrontations will also be redefined, especially insofar as media relations and public debate are also gradually transformed. As mentioned above, issues regarding occupational risks have usually been discussed with relative discretion, by specialists employing arguments that were not a priori intended to escape their hushed enclosures. The appearance of new players and new modes of acting has modified these traditional rules of operation in several different ways.

40The establishment of agencies was at first accompanied by new ways of relating to public space. In fact, the expert reports produced by the InVS and AFFSET are always made public on their website as soon as they have received approval from their governing bodies. This has led to a considerable expansion in the number of people who can obtain information about occupational health. Within the context of specialized and confined decision-making and negotiation processes, it is possible to revisit prior agreements and challenge scientific elements which had been agreed upon during previous meetings. Conversely, once information has been confirmed by an expert committee and made public it is a lot more complicated to contest it or pretend it has not been endorsed. Although it occurred prior to the creation of agencies, the asbestos case offers a revealing example of this change in the rules of internal debate. Although many of asbestos’s hazardous aspects continued to be disputed before its publication, Inserm’s 1996 collective expert report reinforced the status of knowledge on the dangers of asbestos and put an end to most of the controversy. By definitively confirming an estimated number of asbestos-related deaths, the report forced officials to consider these facts when making any new decisions. The publication of AFFSET’s reports on OEL values will no doubt have the same effect, significantly reducing the spectrum of possible solutions to a problem, even if one can already observe the emergence of new spaces for negotiation, no longer around the level of values decreed, but rather on the question of when they are to be implemented. [56] In this manner, the public investigation of an issue by experts contributes to the stabilization of its terms by preventing the re-examination or questioning of past decisions, something which is always possible within more discrete environments.

41In addition to the media’s influence on groups participating in the elaboration of occupational health policies, the procedures that make expert reports public also have far-reaching consequences in terms of the appropriation of these issues by a broader spectrum of society. In effect, one can assume that the public dissemination of scientific knowledge about occupational health contributes to an increase in the number of individuals potentially interested in these issues. When such information reaches more diverse circles, it can help give more visibility to occupational issues that had hitherto been rather ignored. [57] Connections made with environmental themes likewise contribute to the decompartmentalization of occupational health, making it more and more difficult to justify maintaining this sector’s previous operating mode. More generally, increased public interest in these issues is progressively imposing new forms of justification and legitimization for stakeholder decisions. In effect, although compromise can be reached and maintained within a closed environment, without it being necessary to explicitly delineate its nature, the burden of potentially having to justify one’s actions or explain the terms of compromise considerably limits the number of possibilities. [58]

Science and expertise still embedded in social issues

42The consideration of these gradual changes in the domain of occupational health should not lead us to an idealized depiction of science and expertise as the sole catalysts behind these transformations, thanks to their position outside of the sector’s many conflicts and divisions. On the contrary, expert knowledge production programs are the subject of many conflicts and divisions, illustrating that the reality which they inform is likewise prone to controversy and contradictory definitions. Case in point: the Sumer study, which is now recognized as one of the leading studies monitoring labor conditions (it was used as an assessment tool for many of the objectives outlined in 2004 and 2011 public health legislation), is nevertheless subject to frequent investigations which re-examine its claims. While the first Sumer study dates back to 1983, each of its new revisions (1987, 1994, 2003 and 2009) was commissioned on a case-by-case basis, involved important discussions between employers’ organizations and labor unions, which sometimes directly threatened the study’s continued existence. As this study paints a rather unflattering picture of working conditions, it has always been problematic for employers and their representatives. The main critique that these individuals make of the Sumer study concerns the representative nature of the workplace physicians who take part in the study. As these doctors are volunteer participants, employer representatives judge them to be biased activists. This criticism was the impetus for the Sumer validation study led by Nicolas Dodier at the end of the 1980s. [59] As the participation rate for workplace physicians increases and the total number of medical professionals having contributed to at least one edition of the Sumer study also grows, such criticism has tended to become increasingly limited. In order to overcome these difficulties, those in charge of knowledge and expertise production instruments must labor intensively to strengthen their scientific legitimacy. [60] Convincing the greater public of the scientific rigor of a study or institution is a crucial element in limiting the questioning of its results, which may otherwise be subject to endless discussion within the context of occupational health policy negotiations. The fact that it was quite difficult to put an end to the Sumer study in 2008, despite the government’s close ties with employer interests and the latter’s outspoken criticism of the study, was largely due to the fact that the study had been labeled as scientific and worthy of public interest by the National Council for Statistical Information (CNIS – Commission nationale de l’information statistique).

43This embedding of science within social relations is likewise particularly apparent in the ways in which scientific knowledge is used by decision-making figures. In fact, despite sometimes very categorical rhetoric about the separation between science, expertise, social negotiations, and political decision-making, the boundaries between these different domains are extremely debatable, to the point that the establishment and reinforcement of these boundaries is itself at stake in many of these confrontations. [61] In the case of negotiations initiated in the context of agencies like the Steering Committee on Working Conditions (COCT – Conseil d’orientation sur les conditions de travail), many participants from both employers’ organizations and labor unions profess professional competence with regard to the issues being discussed. [62] Doctors, workplace physicians and hospital practitioners specializing in occupational pathologies are indeed often members of the negotiating delegations. Their involvement in discussions attempting to identify occupational illnesses, or to elaborate an appropriate regulatory response, is thus tied as much to their membership of a delegation as to their scientific knowledge of the issue at hand. Such cases of multiple affiliations arise across the board in the field of occupational health, as legal arbitration always takes place as a function of power struggles among the different negotiating partners. Consequently, scientific arguments always appear as weapons in a more or less direct confrontation, and the ability to combine resources plays a major part in the course of negotiations. Although scientific resources are becoming more and more important, they are nevertheless only one resource among many competing others; they are far from being widely accepted by all participants as the most critical source of information. The separation between expertise and negotiation is even denounced as harmful by some labor union representatives, who find that the division introduces biases which strengthen the pre-existing inequalities between employer and labor representatives.


“It’s completely absurd and futile to try to divorce an intangible form of expertise, taken out of context, allegedly based on knowledge – and whose relevance can be debated, I think, because scientific knowledge is just scientific knowledge. In fact, it’s very very far from reality, because we only know what we are looking for, what we are able to look for, what we are paid to look for, what we are granted access to, etc. etc. We can clearly see this pseudo-distinction between people who make decisions (who are not us) and people who possess knowledge (but who don’t know anything about occupational health, because they’re not on the ground, they’re not inside companies, they don’t have the right data, they don’t have the necessary means), so it’s absurd. The right solution is to allow this rather complicated sort of alchemy to just happen, as in the end it works pretty well.”
(Professor of occupational pathologies, labor union representative for a COCT commission, 19 February 2009)

45More fundamentally, the biased nature of science reflects the difficulties experienced by scholarly research when it attempts to measure and publicize the effects of work on workers’ health. The inability of scientific researchers (mainly in the fields of public health and epidemiology) and experts to conduct such studies stems from several characteristics of scientific work, as can be revealed by a sociological approach to science. [63] First of all, presenting the problem in exclusively scientific or epidemiological terms means that data is produced much too slowly to meet employees’ more immediate needs (prevention, identification, etc.). [64] Depending on whether one adopts a positive or negative perspective, this type of argument leads people to view workers as either “guinea pigs”, subject to human experimentation, or “sentries”, on the lookout for environmental hazards that are found with a much higher frequency in professional settings. [65] Moreover, purely academic expertise might underestimate the importance of occupational hazards, which possess certain characteristics (e.g., the fact that they essentially concern labor collectives and are strongly linked to the subordinate position in which employees find themselves) that make them especially difficult to identify for researchers, who often ignore the social environment of individuals and focus mainly on individual risk or aggregated group risk, regardless of their status. [66] In addition, epidemiological studies favor the most stable groups of employees, hence underestimating the risks faced by the growing number of employees who work for subcontracted companies or as temporary workers, as well as by employees with diverse career paths, thus adding to the continued lack of social visibility for these labor subgroups. [67] On a larger scale, as they depend on data provided by industrial and commercial stakeholders, activities of knowledge and expertise production are at the mercy of the latter’s strategies of dissimulation or manufactured ignorance. [68] Taken together, these arguments highlight the distance that exists between scholarly science (or the various different forms of scholarly science, the issue of occupational risk naturally giving rise to many long-standing debates) and social movements seeking to better identify and address health problems caused by working conditions. [69] Consequently, the transformations analyzed in this article are ultimately limited by the almost structural inability of scholarly science to extricate itself from the social framework within which it operates, and thus to support the challenging of existing power relations.


47The case of occupational health clearly illustrates the importance of studying the capacity for knowledge and expertise production in relation to the power relations that exist between the different actors on the public policy stage. First of all, going beyond general notions of expertise, this type of analysis allows us to insist on the specificity of each expert report process and its significance in relation to the issues that inform a certain activity sector and the relationships among its concerned parties. Therefore, both the absence of expertise which can permanently change the face of occupational health, as well as the gradual emergence of tools and ways of thinking imported from the public health sector cannot be understood without taking into account the power relations affecting the actors who help to steer these policies (labor union representatives, corporate representatives, state officials) and who define occupational health policies first and foremost as elements of social negotiation. Likewise, such an analysis demonstrates how issues of knowledge and expertise are inextricably linked to issues of power, and how these are two sides of the same coin. The tension surrounding the implementation of new research and expertise programs and the difficult application of research and expertise instruments to these problems clearly demonstrate this.

48Finally, this tension illustrates a paradox that is specific to the sectors of public policy that rely heavily on scientific expertise. By modifying the information on which public policy and action depends, changes to knowledge and expertise production have now become one of the main vectors through which one can hope to lastingly transform this sphere of public policy and thus bear upon the interests favored by public policy. Nevertheless – and here is the paradox – as this vector represented by scientific research and expertise becomes increasingly central and necessary, it is also so strongly influenced by the power relations among the social groups of any one domain that the degree of innovation furnished by this research and its ability to potentially disrupt the status quo ultimately seems rather weak. In reality, far from providing a fulcrum which would allow one to influence power relations from the outside, the transformation of expertise has only had a limited effect on the relationships among actors involved in policy direction. Therefore, in the case of occupational health one is forced to state the following paradox: endowed with new instruments for the production of knowledge and expertise, labor administration’s power is gradually being strengthened with respect to the other stakeholders involved in the elaboration of policy. Nevertheless, despite the important upheavals studied above, in the context of these developments a certain number of signs continue to demonstrate the relative stability of power relations. For the moment, therefore, nothing seems to show that these changes will steer occupational health policy towards a better awareness and consideration of workers’ health.


  • [1]
    Olivier Ihl, Martine Kaluszynski, Gilles Pollet (eds), Les sciences du gouvernement (Paris: Economica, 2003); Olivier Ihl (ed.), Les “sciences” de l’action publique (Grenoble: Presses Universitaires de Grenoble, 2006).
  • [2]
    Cécile Robert, “Expertise et action publique” in Olivier Borraz, Virginie Guiraudon (ed.), Politiques publiques I, La France dans la gouvernance européenne (Paris: Presses de Sciences Po, 2008), 309-55; Laurence Dumoulin, Stéphane La Branche, Cécile Robert, Philippe Warin (eds), Le recours aux experts. Raisons et usages politiques (Grenoble: Presses Universitaires de Grenoble, 2005).
  • [3]
    For two summaries, see Céline Granjou, “L’expertise scientifique à destination politique”, Cahiers internationaux de sociologie, 114, 2003, 175-83; or Pierre-Benoît Joly, “La sociologie de l’expertise: la recherche française au milieu du gué”, in Olivier Borraz, Claude Gilbert, Pierre-Benoît Joly (eds) Risques, crises et incertitudes: pour une analyse critique (Grenoble: CNRS/MSH Alpes, 2005), 117-74.
  • [4]
    Michel Callon, Arie Rip, “Forums hybrides et négociations des normes socio-techniques dans le domaine de l’environnement. La fin des experts et l’irrésistible ascension de l’expertise”, in Jacques Theys, Véronique Liber, Marie-Pierre Palacios (eds), Environnement, science et politique. Les experts sont formels (Paris: Germes, 1991), 227-38.
  • [5]
    Ronald Brickman, Sheila Jasanoff, Thomas Ilgen, Controlling Chemicals. The Politics of Regulation in Europe and the United States (Ithaca: Cornell University Press, 1985); Sheila Jasanoff, “Contested boundaries in policyrelevant science”, Social Studies of Science, 17(2), 1987, 195-230.
  • [6]
    For two very different perspectives, see Philippe Roqueplo, Entre savoir et décision, l’expertise scientifique (Paris: Inra, 1997); and Paul-André Rosental, L’intelligence démographique: sciences et politiques des populations en France, 1930-1960 (Paris: Odile Jacob, 2003).
  • [7]
    CRESAL, Les raisons de l’action publique. Entre expertise et débat. Actes du colloque CRESAL-CNRS, Saint Étienne, 13-14 May 1992 (Paris: L’Harmattan, 1993); Olivier Borraz, Les politiques du risque (Paris: Presses de Sciences Po, 2008).
  • [8]
    Jacqueline Estades, Elisabeth Rémy, L’expertise en pratique. Le cas de la vache folle et des rayonnements ionisants (Paris: L’Harmattan, 2003); Céline Granjou, Marc Barbier, Métamorphoses de l’expertise. Précaution et maladies à prion (Paris: Maison des sciences de l’Homme/Quae, 2010).
  • [9]
    Daniel Benamouzig, Julien Besançon, “Administrer un monde incertain: les nouvelles bureaucraties techniques. Le cas des agences sanitaires en France”, Sociologie du travail, 47(3), 2005, 301-22.
  • [10]
    The organizers of a recent conference on this theme had proposed precisely such an approach: Yann Bérard, Renaud Crespin (eds), Aux frontières de l’expertise: dialogues entre savoirs et pouvoirs (Rennes: Presses Universitaires de Rennes, 2010).
  • [11]
    Stephen Hilgartner, Science on Stage. Expert Advice as Public Drama (Stanford: Stanford University Press, 2000); C. Granjou, M. Barbier, Métamorphoses de l’expertise.
  • [12]
    These framing effects are clearly highlighted in Joseph R. Gusfield’s The Culture of Public Problems: Drinking-Driving and the Symbolic Order (Chicago, London: University of Chicago Press, 1981); for recent examples in the domain of public health, see Claude Gilbert, Emmanuel Henry (eds), Comment se construisent les problèmes de santé publique (Paris: La Découverte, 2009).
  • [13]
    Bruno Latour, La science en action. Introduction à la sociologie des sciences (Paris: Gallimard, 1995).
  • [14]
    Michel Callon, “Some elements for a sociology of translation: domestication of the scallops and the fishermen of St-Brieuc bay”, in J. Law (ed.), Power, Action and Belief: a New Sociology of Knowledge? Sociological Review Monograph (London: Routledge and Kegan Paul, 1986), 196-223. Madeleine Akrich, Michel Callon, Bruno Latour, Sociologie de la traduction. Textes fondateurs (Paris: Presses de l’École des Mines, 2006).
  • [15]
    Dominique Pestre, “Thirty years of science studies: knowledge, society and the political”, History and Technology, 20(4), 2004, 351-69; Scott Frickel, Kelly Moore (eds), The New Political Sociology of Science. Institutions, Networks, and Power (Madison: University of Wisconsin Press, 2006), in particular, David J. Hess’s chapter, “Antiangiogenesis research and the dynamics of scientific fields: historical and institutional perspectives in the sociology of science”, 122-47.
  • [16]
    Scott Frickel, Kelly Moore, “Prospects and challenges for a new political sociology of science”, in S. Frickel, K. Moore (eds), The New Political Sociology of Science, 3-31.
  • [17]
    For example, see Bruno Jobert, Pierre Muller, L’État en action. Politiques publiques et corporatismes (Paris: PUF, 1987).
  • [18]
    See Peter A. Hall, “Policy paradigms, social learning, and the state: the case of economic policymaking in Britain”, Comparative Politics, 25(3), 1993, 275-96; for a discussion of the role of cognitive aspects in the analysis of public policy, see also Pierre Muller, “L’analyse cognitive des politiques publiques: vers une sociologie politique de l’action publique”, Revue française de science politique, 50(2), 2000, 189-208.
  • [19]
    Pierre Lascoumes, Patrick Le Galès (eds), Gouverner par les instruments (Paris: Presses de Sciences Po, 2004).
  • [20]
    Emmanuel Henry, “La reconfiguration de l’expertise en santé au travail: vecteur de transformation de l’action publique?”, in Nathalie Dedessus-Le-Moustier, Florence Douguet (eds), La santé au travail à l’épreuve des nouveaux risques (Paris: Tec et Doc/Lavoisier, 2010), 259-70.
  • [21]
    This article is the product of a collective research project called “Pratiques de l’expertise scientifique”, sponsored by the CNRS and the Institut des sciences de la communication (Claude Gilbert, Emmanuel Henry, JeanNoël Jouzel, Pratex. Analyse de l’expertise en pratique (Paris: CNRS, Rapport pour l’Institut des sciences de la Communication (ISCC) of the CNRS, 2009)). The study traced the history of several past investigations or consulting agencies in the field of occupational health based on documentary analyses and interviews with those in charge of them. This article has benefited from the painstaking attention it received from Isabelle Bourdeaux and Renaud Crespin, whom I would like to thank here.
  • [22]
    Laurent Duclos, Guy Groux, Olivier Mériaux (eds), Les nouvelles dimensions du politique. Relations professionnelles et régulations sociales (Paris: LGDJ, 2009).
  • [23]
    For example, in the context of DARES (Directorate of Research, Studies and Statistics – Direction de l’animation de la recherche, des études et des statistiques), see Étienne Pénissat, L’État des chiffres. Sociologie du service de statistique et des statisticiens du ministère du Travail et de l’Emploi (1945-2008) (Paris: École des hautes études en sciences sociales, 2009).
  • [24]
    Pierre Mathiot, Acteurs et politiques de l’emploi en France (1981-1993) (Paris: L’Harmattan, 2000).
  • [25]
    Peter Bachrach, Morton S. Baratz, “Two faces of power”, The American Political Science Review, 56(4), 1962, 947-52.
  • [26]
    Gerald E. Markowitz, David Rosner, Deceit and Denial. The Deadly Politics of Industrial Pollution (Berkeley: University of California Press, 2002).
  • [27]
    See Emmanuel Henry, “Militer pour le statu quo. Le Comité permanent amiante ou l’imposition réussie d’un consensus”, Politix, 70, 2005, 29-50; and on an international scale, see Jock McCulloch, Geoffrey Tweedale, Defending the Indefensible. The Global Asbestos Industry and its Fight for Survival (New York: Oxford University Press, 2008). Unlike in the US or the UK, where works by historians and sociologists have clearly illustrated the role played by industrialists in the structuring of scientific knowledge, thanks to their records being made available during legal trials, it is very difficult to have access to similar data in France.
  • [28]
    Mireille Jarry, Françoise Lalande, Jean Roigt, Inspection générale des affaires sociales (Igas), Contrôle du fonctionnement de l’institut national de recherche et de sécurité pour la prévention des accidents du travail et des maladies professionnelles (Paris: Igas, 1999).
  • [29]
    On the logic of ignorance production, see Scott Frickel, M. Bess Vincent, “Hurricane Katrina, contamination, and the unintended organization of ignorance”, Technology in Society, 29(2), 2007, 181-8.
  • [30]
    Annie Thébaud-Mony, De la connaissance à la reconnaissance des maladies professionnelles en France. Acteurs et logiques sociales (Paris: La Documentation française, 1991); Ellen Imbernon, Estimation du nombre de cas de certains cancers attribuables à des facteurs professionnels en France (Paris: Institut de veille sanitaire, 2002); in terms of Spain, see Alfredo Menéndez-Navarro, “A camel through the eye of a needle: expertise and the late recognition of asbestos-related diseases”, International Journal of Health Services, 2011, 121-35.
  • [31]
    For the most recent, see Noël Diricq, “Rapport de la commission instituée par l’article L. 176-2 du code de la sécurité sociale” (Paris: Ministère de la Santé, de la Jeunesse, des Sports et de la Vie associative, 2008).
  • [32]
    J. R. Gusfield, The Culture of Public Problems.
  • [33]
    Françoise Piotet, “L’amélioration des conditions de travail entre échec et institutionnalisation”, Revue française de sociologie, 29(1), 1988, 19-33.
  • [34]
    Bernard Cassou, Dominique Huez, Marie-Laurence Mousel, Catherine Spitzer, Annie Touranchet-Hebrard (eds), Les risques du travail. Pour ne pas perdre sa vie à la gagner (Paris: La Découverte, 1985).
  • [35]
    For a description of how these were implemented, see Michel Gollac, Serge Volkoff, “Mesurer le travail. Une contribution à l’histoire des enquêtes françaises dans ce domaine”, Document de travail du Centre d’études de l’emploi, 127, July 2010, 1-18.
  • [36]
    M. Gollac, S. Volkoff, “Mesurer le travail”, 9.
  • [37]
    Jean-François Girard, Jean-Michel Eymeri, Quand la santé devient publique (Paris: Hachette Littératures, 1998).
  • [38]
    See Aquilino Morelle, La défaite de la santé publique (Paris: Flammarion, 1996); or “La sécurité sanitaire: enjeux et questions”, special issue of the Revue française des affaires sociales, 3-4, 1997.
  • [39]
    Emmanuel Henry, Amiante: un scandale improbable. Sociologie d’un problème public (Rennes: Presses Universitaires de Rennes, 2007).
  • [40]
    Emmanuel Henry, “Quand l’action publique devient nécessaire. Qu’a signifié ‘résoudre’ la crise de l’amiante?”, Revue française de science politique, 54(2), 2004, 289-314.
  • [41]
    This process, which has existed since the beginning of the 1990s, consists of a group of experts gathered together for several months to address one specific question, with the aim of offering a scientific overview of a problem identified in existing literature (Jeanne Etiemble, “L’expertise collective à l’Inserm”, in Geneviève Decrop, Jean-Pierre Galland (eds), Prévenir les risques. De quoi les experts sont-ils responsables? (La Tour d’Aigues: L’Aube, 1998), 63-77).
  • [42]
    Theodore M. Porter, Trust in Numbers. The Pursuit of Objectivity in Science and Public Life (Princeton: Princeton University Press, 1995).
  • [43]
    French legislation demands that compensation for occupational risks be first measured against the definitions of a set number of pathologies with occupational causes. The definition of occupational illnesses is accomplished in the form of charts that link pathologies and the types of occupations that are likely to produce them. On the negotiations surrounding the creation or the modification of these charts, see Marc-Olivier Déplaude, “Codifier les maladies professionnelles: les usages conflictuels de l’expertise médicale”, Revue française de science politique, 53(5), 2003, 707-35.
  • [44]
    D. Benamouzig, J. Besançon, “Administrer un monde incertain”.
  • [45]
    François Buton, “De l’expertise scientifique à l’intelligence épidémiologique: l’activité de veille sanitaire”, Genèses, 65, 2006, 71-91.
  • [46]
    For more information about this project, see the Bulletin épidémiologique hebdomadaire, 40, October 2003.
  • [47]
    To get an idea of the DST’s activity, see the InVS’s website (<>).
  • [48]
    Catherine Ha, Yves Roquelaure, Annie Touranchet, Annette Leclerc, Ellen Imbernon, Marcel Goldberg, “Le réseau pilote de surveillance épidémiologique des TMS dans les pays de la Loire: objectifs et méthodologie générale”, Bulletin épidémiologique hebdomadaire, 44-5, 2005, 219-21; and “La surveillance épidémiologique des troubles musculo-squelettiques”, special issue of the Bulletin épidémiologique hebdomadaire, 44-5, 15 November 2005.
  • [49]
    Véronique Daubas-Letourneux, “Produire des connaissances en santé au travail à l’échelle régionale. Le signalement des maladies à caractère professionnel dans les pays de la Loire”, Revue française des affaires sociales, 62(2-3), 2008, 213-35.
  • [50]
    As shown by the existence, since 1986, of the Association for Ergonomical and Epidemiological Research in the West of France (AREEO – Association pour la recherche en épidémiologie et en ergonomie de l’Ouest): Véronique Daubas-Letourneux, Annie Thébaud-Mony, Produire des connaissances en santé au travail à l’échelle régionale. Étude sociologique du dispositif de signalement des maladies à caractère professionnel dans les pays de la Loire (Nantes: Maison des sciences de l’Homme Ange Guépin, 2006).
  • [51]
    These professionals have themselves experienced great changes over the past few years, particularly with the decreasing importance of workplace physicians and the rise of other professional bodies: see Stéphane Buzzi, Jean-Claude Devinck, Paul-André Rosental, La santé au travail. 1880-2006 (Paris: La Découverte, 2006), and Pascal Marichalar, “La médecine du travail sans les médecins?”, Politix, 91, 2010, 27-52.
  • [52]
    We won’t dwell here on the controversies surrounding AFSSE’s first expert reports, in 2003, on mobile telephony: on this subject, see Olivier Borraz, Michel Devigne, Danielle Salomon, Controverses et mobilisations autour des antennes-relais de téléphonie mobile (Paris: Centre de sociologie des organisations/CNRS/Sciences Po, 2004).
  • [53]
    On similar mechanisms in the French Food Safety Agency, see C. Granjou, M. Barbier, Métamorphoses de l’expertise.
  • [54]
    Olivier Borraz, Les politiques du risque (Paris: Presses de Sciences Po, 2008).
  • [55]
    Emmanuel Henry, “‘License to expose’? OELs, scientific expertise and state in contemporary France”, forthcoming in Soraya Boudia, Nathalie Jas (eds), Toxic World. Toxicants, Health and Regulation in the Twentieth Century, Pickering and Chatto.
  • [56]
    E. Henry, “‘License to expose’?”.
  • [57]
    Renaud Crespin, Emmanuel Henry, Jean-Noël Jouzel, “Peut-on sortir de la méconnaissance des effets du travail sur la santé? Quelques pistes de réflexion pour les sciences humaines et sociales”, in Santé au travail. Quels nouveaux chantiers pour les sciences humaines et sociales? (Grenoble: CNRS/MSH Alpes, 2008), 253-63.
  • [58]
    Claude Gilbert, Emmanuel Henry, “La définition des problèmes publics: entre publicité et discrétion”, Revue française de sociologie, 53(1), 2012, 35-59.
  • [59]
    Nicolas Dodier, Jugements médicaux, entreprises et protocoles de codage. La contribution des médecins du travail à l’enquête SUMER sur les risques professionnels (Paris: La Documentation française, 1990); and L’expertise médicale. Essai de sociologie sur l’exercice du jugement (Paris: Métailié, 1993).
  • [60]
    Hilgartner, Science on Stage.
  • [61]
    Sheila S. Jasanoff, “Contested boundaries in policy-relevant science”.
  • [62]
    M. Callon, A. Rip, “Forums hybrides et négociations des normes socio-techniques dans le domaine de l’environnement”.
  • [63]
    D. Pestre, “Thirty years of science studies”; S. Frickel, K. Moore (eds), The New Political Sociology of Science.
  • [64]
    Annie Thébaud-Mony, Travailler peut nuire gravement à votre santé: sous-traitance des risques, mise en danger d’autrui, atteintes à la dignité, violences physiques et morales, cancers professionnels (Paris: La Découverte, 2007), 190 and passim.
  • [65]
    On the original link between scientific research on environmental risks with that on occupational risks, see Christopher C. Sellers, Hazards of the Job. From Industrial Disease to Environmental Health Science (Chapel Hill, University of North Carolina Press, 1997).
  • [66]
    Marcel Goldberg, “Cet obscur objet de l’épidémiologie”, Sciences sociales et santé, 1(1), 1982, 55-110; Marcel Goldberg, Maria Melchior, Annette Leclerc, France Lert, “Les déterminants sociaux de la santé. Apports récents de l’épidémiologie sociale et des sciences sociales de la santé”, Sciences sociales et santé, 20(4), 2002, 75-128.
  • [67]
    Annie Thébaud-Mony, “Construire la visibilité des cancers professionnels. Une enquête permanente en SeineSaint-Denis”, Revue française des affaires sociales, 2-3, 2008, 237-54; Annie Thébaud-Mony, Béatrice Leconte, “Mémoire du travail et des expositions professionnelles aux cancérogènes. Enquête en Seine-Saint-Denis (France)”, Pistes, 12(3), 2010.
  • [68]
    Robert Proctor, Cancer Wars. How Politics Shapes What We Know and Don’t Know about Cancer (New York: Basic Books, 1995); G. E. Markowitz, D. Rosner, Deceit and Denial; David Michaels, Celeste Monforton, “Manufacturing uncertainty: contested science and the protection of the public’s health and environment”, American Journal of Public Health, 95(1), 2005, 39-48; Annie Thébaud-Mony, “Les fibres courtes d’amiante sont-elles toxiques? Production de connaissances scientifiques et maladies professionnelles”, Sciences sociales et santé, 28(2), 2010, 95-114.
  • [69]
    This has already been well established with regard to environmental issues by a certain number of sociological studies, as well as studies on the history of science. For example: Phil Brown, “Popular epidemiology and toxic waste contamination: lay and professional ways of knowing”, Journal of Health and Social Behavior, 33(3), 1992, 267-81; Marcel Calvez, “Les signalements profanes de clusters de cancers: épidémiologie populaire et expertise en santé environnementale”, Sciences sociales et santé, 27(2) 2009, 79-106; Barbara L. Allen, “Shifting boundary work: issues and tensions in environmental health science in the case of Grand Bois, Louisiana”, Science as Culture, 13(4), 2004, 429-48. See also Alan Irwin, Citizen Science. A Study of People, Expertise, and Sustainable Development (London: Routledge, 1995).


This article analyses contemporary changes in the production of knowledge and expertise pertaining to occupational health, by relating them to the characteristics of this public policy sector and recent developments within it. The article first shows how changes in public health policy bring about certain changes in occupational health, for instance by necessitating the creation of new expert agencies. It then examines the institutionalization of expertise on occupational risks, along with the impact this has on the modalities of decision-making in a specific field. It explains that in spite of huge changes, the power relations characterizing this public policy domain remain relatively stable. Finally, apart from the case of occupational health, this article again raises the question again of the production of knowledge and expertise, by relating them to the power plays structuring a public policy sector.

Emmanuel Henry
Emmanuel Henry is an associate professor of political science at Sciences Po Strasbourg and a researcher within the Centre for European Political Sociology (Groupe de sociologie politique européenne – GSPE, UMR Prisme, CNRS, Université de Strasbourg). He is the author of Amiante: un scandale improbable. Sociologie d’un problème public (Rennes: Presses Universitaires de Rennes, 2007) (Asbestos: An Improbable Scandal. Sociology of a Public Issue); and has recently co-edited (with Claude Gilbert) Comment se construisent les problèmes de santé publique (Paris: La Découverte, 2009) (How Public Health Problems Are Shaped). His work focuses on the sociology of public policy while paying attention to the different kinds of inequalities that characterize our contemporary society, especially through a reflection on how public issues are shaped and treated. He is currently studying the importance of expertise in the domain of public health and is the director of a collective research program subsidized by the French National Research Agency (ANR – Agence nationale de recherche); he likewise continues to research the role of expertise in the domain of occupational health (GSPE, IEP de Strasbourg, MISHA, 5 allée du Général Rouvillois, CS 50008 F-67083 Strasbourg Cedex).
Translated from French by 
Sarah-Louise Raillard
Latest publication on cairn or another partner portal
Uploaded on on 03/03/2014
Distribution électronique pour Presses de Sciences Po © Presses de Sciences Po. Tous droits réservés pour tous pays. Il est interdit, sauf accord préalable et écrit de l’éditeur, de reproduire (notamment par photocopie) partiellement ou totalement le présent article, de le stocker dans une banque de données ou de le communiquer au public sous quelque forme et de quelque manière que ce soit.
Loading... Please wait