Unelected representatives

8The idea of a world parliament of labor dates from before the First World War. It was originally a demand on the part of various workers’ unions such as the American Federation of Labor in the United States and many European unions at the International Trade Union Conference in Leeds in July 1916. [20] Following the creation of the ILO, this frame of reference – that of popular representation through parliamentary institutions – was used to promote the project of such an international organization. On 11 April 1919, British politician George Barnes, one of the founders of the organization, declared, “We believe that our scheme will give life and strength and vitality to the League of Nations by bringing it in contact with the daily life of the people.” [21] At the opening of the first ILC in October 1919 in Washington, the conference organizers recalled how, in setting out the principles and rules of representation, they had been inspired by “the rules practiced in various parliaments”. [22]

9However, although it may have been a powerful rhetorical tool with which to legitimate the organization to the working classes and the union world, in a political context marked by fear of the propagation of the Bolshevik Revolution, the metaphor was deceptive. Firstly – and contrary to the hopes of the trade union leaders of 1919 (foremost among them Léon Jouhaux) – the ILO has no legislative power as such. Its decisions have no direct effect on national legislations; they must be ratified in order to be enacted. And then there is the non-electoral manner of representation. Not that the latter is peculiar to the ILO alone – indeed, it is a trait common to most international organizations. [23]

10As far as the government representatives who sit on the ILC are concerned, even to this day there is no constitutional or regulatory provision that dictates the way in which they are to be chosen. Despite some national idiosyncrasies, member states usually send diplomats [24] to fill the seats or civil servants from their labor ministry (in some cases, the labor minister himself) in view of the ILO’s functional specialization in questions of employment. In exceptional cases, heads of state themselves make the trip. Insofar as the composition of governments derives – in contemporary democracies at least – from the results of a (legislative or presidential) election, it might be argued that the presence of a labor minister or, a fortiori, a head of state or of government maintains some kind of electoral link, albeit an indirect one. Yet this argument is weakened if one considers the rare and episodic character of this representation “at the highest level” – it often lasts only as long as a speech or an official meeting with the director general.

11If the representativeness of governmental delegates is largely presumed within the ILC, that of employers and workers is subject to greater scrutiny and has to be proved to a greater extent. But here once again, election is not a necessary condition for the validation of nominations. Workers’ and employers’ representatives are subject to a “clause of representativeness” contained in Article 3, Paragraph 5 of the ILO constitution, which reads as follows: “The Members undertake to nominate non-Government delegates and advisers chosen in agreement with the industrial organizations, if such organizations exist, which are most representative of employers or workpeople, as the case may be, in their respective countries.” [25] This article “invents” the notion of the most representative organization, and that of the representativeness of professional organizations in general, [26] and it acts as a double filter designed to select those organizations considered best equipped to ensure the representation of workers and employers (if such organizations exist, that is). For it is not just a question of organizations that are “representative” of workers and employers, but of the “most representative” among them.

12The first selective filter of Article 3 is enacted at the national level by governments, upon which the task falls to nominate the representatives of workers and employers, in agreement with the organizations concerned. The second filter then allows the ILC, if an ILO member protests, to refuse to admit a delegate who fails to satisfy the condition of representativeness. [27] In 1922 a controversy over the nomination of a worker delegate from the Netherlands caused the Credentials Committee (the body in charge of the accreditation of delegates and the examination of complaints relating to the nomination of ILC members) to appeal to the Permanent Court of International Justice (PCIJ) to interpret this clause. [28] In its advisory opinion, the PCIJ confirmed that the notion of representativeness is not defined by the constitution, noting that “the most representative organisations for this purpose are, of course, those organisations which best represent the employers and the workers respectively”. This ruling only serves to accentuate the subjective and mutable dimension of the notion of representativeness in play here (selection of the “most” representative organization can only be the result of an assessment that is relative, and it cannot rely on supposedly “objective” criteria alone). On the basis of jurisprudence, debates, and existing works on the question, one might sum up the criteria that preside over the selection of workers’ and employers’ representatives as follows: a numerical criterion of majority (which organization has the greatest number of members?), a procedural criterion of consultation (have all organizations been consulted by the government before proceeding with the selection of delegates?), and a political criterion of conformity with the spirit of tripartite representation (is the organization independent of the government?). These criteria are not always formulated explicitly. Nor are they cumulative – one of them, such as the criteria of majority, may be invoked in the absence of the others – or mutually exclusive. They are interrelated and combined at the discretion of the international conference, and they are adapted according to national contexts. Depending on the case, certain criteria may be put aside or, at least, relativized. Historically, the number of members that an organization has and, consequently, its status as majority organization, has been the paramount criterion. The criterion of independence from government was the cause of numerous debates concerning German and Italian delegates during the period of Nazi Germany and fascist Italy, and then again during the Cold War in relation to communist delegates. It has gained increased importance again since the beginning of the 1990s, as is attested by the sanctions against Myanmar. [29]

13And yet nowhere in either Article 3 or the related jurisprudence is the election of workers’ and employers’ representatives mentioned as a prerequisite for participating in the ILO. This does not mean, though, that members attach no importance to it. Among the 60 or so workers’ and employers’ delegates questioned between 2011 and 2014, many stated that they attributed their legitimacy to an electoral mandate won earlier in their career as activists or professional representatives, even when no specific election had taken place to nominate them for a seat at the ILO. The importance accorded to election transpires, for example, in the statement of the Ghanaian workers’ representative: “When it comes to the ILO, representation is a central slogan, and representation means having a mandate […] so people who come to the ILO are those who have been elected by those they represent.” [30]

14This kind of statement (which the institution itself is in no position to confirm) indicates the recognition of the electoral principle as the foundation of the legitimacy of representation. Even if, as we have seen, election occupies only a marginal place in the ILO’s procedural definition of representativeness – at least at the stage of the accreditation of members – what about the substantial side of representation and of the interests present within the organization? Here again, the tripartite and international character of representation entails a series of compromises in the parliamentary logic, though there is a much greater assumption of these on the ILO’s part.

Representation by group: an inversion of the logic of general interest and particular interest

15The ILO’s tripartism is not limited to a representation of the state via three types of actors. Since 1919, it has further shuffled the cards of national representation by establishing a system of representation centered on the notion of the group – the Governmental Group, the Workers’ Group, and the Employers’ Group – thus foregrounding a particular type of solidarity that one might describe as “transnational”. [31] Concretely, this means that, apart from in plenary sessions, delegates do not sit as representatives of their nation, but are assembled according to their membership of one of these three groups. In 1919, Edward Phelan suggested in the British memorandum that there should be at the ILC “three representative bodies which meet separately”. [32] This system, established at the Washington conference, has remained in force until the present day. In the following extract, the Australian employers’ representative praises it on the grounds that it allows actors who have no other space of common representation to discuss and deliberate upon questions of the foremost importance to them: “One of the good things about the ILO is its tripartite structure […] the three main actors in the world of work have the opportunity to discuss matters within their own group, their ‘constituency’, at the International Labour Conference.” [33]

16Likened to a “constituency” – a term that, in French (circonscription), as in English, carries the idea of political representation of a district and, more generally, of a group to be represented in the tradition of democratic representation [34] – the group, over and above the state, thus imposes itself as the representative entity of reference for ILO members. There is a notable difference here, however, between workers and employers on one hand, and governments on the other: whereas the former are absorbed by the group dynamic from the outset, the latter maintain, and even cultivate, their national specificity. The workers’ and employers’ groups express themselves almost exclusively through the intermediary of their respective spokesperson (the president of the group), but since 1919 the governmental group has been divided along national and regional lines. Today, the greater part of the work of negotiation takes place at the level of the governmental group, within regional subgroups (the Nordic group, the West African group, the Latin American group, and so forth). Thus, as the government representative of Canada remarks, “Hardly any attempt is made to coordinate the governmental group […] [I]t tends to be very fragmented in regard to the questions covered, and it is too difficult to arrive at a common opinion on these questions.” [35]

17In the 1960s, Robert Cox attributed this difference in attaining unity to the greater diversity of national interests represented by the governments, in contrast to workers and employers, who are “naturally” inscribed within a logic of interest groups, allowing them to develop transnational solidarities. [36] This reading, however, tends to essentialize the interests in question, and to take as given the existence of such transnational solidarities. It disregards the potential overlapping of affiliations (to the group, to a nation, to a region), and above all the important work done by international workers’ and employers’ organizations, foremost among them the International Trade Union Confederation (ITUC) [37] and the International Organization of Employers (IOE), [38] in structuring representation. Adopting the language of the sociology of interest groups, one might say that the latter act as veritable “entrepreneurs of representation”. They supervise the secretariat of each group and therefore take an extremely vigilant position in regard to the composition of delegations, doing so precisely with the purpose of constructing a common interest and imposing true group discipline upon the workers’ and employers’ delegates during negotiations and votes. ILC workers’ and employers’ delegates thus express themselves only rarely in their own name during plenary sessions. The work of negotiation is the responsibility of the secretaries and spokespeople of the groups, who also fulfill strategic functions within the ITUC and the IOE – bodies whose voice is relayed to the ILO by civil servants working in the services dedicated exclusively to the monitoring of relations with workers and employers.

18The change of scale in the passage to the international stage – and, moreover, within a tripartite framework – thus transforms the classical conception of general interest and particular interest. Whereas at the national level the government, particularly in the French tradition of representation, takes on the role of legitimate representative of the general interest, looming over the particular interests embodied by interest groups, this logic is inverted at the international level. In the ILO, on the contrary, it is the workers and employers who are in a position to manifest a common interest (which they are unable to put to work in their struggles on a national scale), whereas governments seem to be at once divided and disorganized.

19This discipline that reigns within the workers’ and employers’ groups has contributed greatly to the persistence of the parliamentary metaphor, with workers’ and employers’ groups often being compared to political parties. [39] If the internal functioning of the ILO groups is similar in many regards to that of national political parties, though, here the analogy is again deceptive. For, unlike a national representative assembly, where the respective weight of political parties varies according to electoral results, in the ILO the numerical equilibrium between the three groups will never change, so long as appointments are made upstream of each of the group representatives and independently of any electoral consultation whatsoever. There is no need for delegates to rely on the group in order to take up positions that have been attributed to them from the outset. By guaranteeing each group its position, the non-electoral nature of representation thus plays a role in engendering and reinforcing transnational solidarities over time, at least for the employers’ and workers’ groups.

20This rather freely adapted and self-serving use of the parliamentary metaphor in a non-electoral context marked by a reconfiguration of the interests present can be explained by a twofold concern for the legitimation of the institution itself. The first concern, touched upon above, is due to the historical and social context within which the ILO emerged. In 1919, plenipotentiaries were obsessively focused on the containment of the Bolshevik Revolution. The representation of workers and employers seemed to supply governments with a two-pronged response to workers’ unions in their demand for social democracy on the one hand, and to the revolutionary movements on the other, since they were able to brandish from the outset the “reformist” ambition of an institution that aimed to bring representatives of opposed interests into dialogue. The longevity of the metaphor, however, is better explained by the emergence of a changed international context following the Second World War. As a unique organization in terms of structure and mandate alike, the ILO at first found itself in competition with the United Nations and its Economic and Social Council (Ecosoc), and then, from the 1970s on, with other organizations that, like the Organization for Economic Cooperation and Development (OECD) and the International Standards Organization (ISO), regularly challenged the functional legitimacy of the ILO by adopting norms that infringed on its field of competence. [40] In a context marked by increased inter-institutional competition, the ILO placed greater emphasis on its tripartite representation as a “comparative advantage” [41] distinguishing it from other international organizations, and in doing so drew on its social democratic frame of reference. [42] Although not entirely without foundation, this claim on the part of the ILO to be the organization that is most “open” to society as a whole does tend to sidestep the powerful logics of selection of representatives that have structured its tripartite representation since 1919.

Secular anticommunism as a structuring feature

22Although the creation of the ILO cannot be seen purely as a response to communism, insofar as it was based upon a loose network of ideas and individuals committed to social reform from the end of the nineteenth century, [44] its anticommunism exerted considerable influence upon the systems that would regulate access to representation. Although from the 1930s the ILO adhered to the principle of universal representation at the level of states, by the same token authorizing the participation of communist worker and employer delegates at the level of the ILC, international union and employers’ organizations (foremost among them the IFTU, then the WCL and the ITUC for workers and the IOE for employers), the majority of which were in the ILO, deployed an arsenal of institutional machinery in order to block access for communist delegates – considered to be mere mouthpieces for their governments – to positions of power. And there are many such positions: presidency of the workers’ or employers’ group, of certain key commissions (the Credentials Committee, the Committee on Freedom of Association), and, finally and above all, of the Governing Body, the limited executive organ that is appointed partially by election every three years by the ILC (see below) and that sets the strategic orientation of the organization.

23This objective complicity between the ITUC and the IOE, which minimized the influence of communist delegates, particularly during the Cold War era, rests upon three principal, mutually entwined institutional dynamics. Firstly, there is the principle of the autonomy of non-governmental groups: according to this principle, each group is the sole master of its own internal procedures. Secondly, there is the adoption of a majority vote to appoint representatives to the Governing Body, which has been unfavorable to communist delegates, who tend to attract less votes. And finally, there is the principle of co-optation that presides over the constitution of the lists of candidates to be elected every three years to the Governing Body, the composition of which is negotiated upstream, together with the ITUC and the IOE, meaning that, in the absence of an agreement made with one of these two organizations, no unaffiliated and unsupported delegate has any chance of gaining a position of responsibility within the organization. Thus, not only are the employers’ and workers’ delegates subject to relatively strict conditions of entry to the ILO stemming from Article 3 (mentioned above), but they must also, on pain of being marginalized or confined to an opposition role, conform to the party line of the dominant group within the institution.

24It is these same mechanisms that made it possible, particularly in the workers’ group, to ostracize the Christian Unionists, who gathered first under the International Federation of Christian Trade Unions (IFCTU) and subsequently under the World Confederation of Labor (WCL). Even if the end of the Cold War and the shifts in wage-earner and employer unionism have rendered this kind of control obsolete in theory, [45] the question has occasionally cropped up again in regard to delegates from authoritarian regimes where the principle of freedom of association is not respected. The case of Chinese workers and employers, repeatedly discussed since the end of the 1980s, and whose representatives, after numerous negotiations with the ITUC and the IOE, were finally allowed a seat on the Governing Body, are frequently the subject of debate within the groups. These ideological logics, in contributing to the marginalization of certain groups within the Governing Body, only add to an almost century-old structural dynamic that tends to make the Governing Body a space reserved for the most powerful members.

The Governing Body: representation as return on investment

25Since 1919, the ILO’s Governing Body, today composed of 56 incumbent members and 66 deputy members, has been an arena of representation highly coveted by members of the ILO. One of the delegates encountered at the ILC in 2013 compared it to the “kitchen” of the organization, where everything is prepared in advance to be “served” to the members of the ILC:

26

“When you go to a restaurant, you are served a prepared dish, but you don’t know how it has been made in the kitchen […] Here, Conference is the final product. But I prefer to see how it’s cooked. The Governing Body: that’s where everything is prepared.” [46]

27And indeed, it is certainly at the ILC that international labor conventions are adopted; but so-called strategic questions (budget, agenda, election of the director general) are discussed within the Governing Body. The latter is appointed partly by the ILC, by a vote of ILO members every three years. Appointments are only partly made on this basis because, as in the great majority of international organizations, the founding members secured a certain number of permanent seats that are not open for election. [47] At the ILO, these permanent seats (which today number ten out of the 28 governmental regular seats) boast a certain distinction in relation to the permanent seats of the Security Council, insofar as they are allocated according to the economic power of the member states qualified as states “of chief industrial importance”. In principle, then, the list is subject to change; but in practice it has seen only marginal changes, the most recent revision dating from 1983. Today it includes Germany, Brazil, China, France, the United Kingdom, the United States, India, Italy, Japan, and Russia. Although this list offers only a glimpse into the unequal distribution of representation on the Governing Body, it is an invitation to investigate the underlying criteria, both implicit and explicit, that govern representation in the most important organs of the ILO.

28These permanent members include the principal contributors to the ILO budget: the United States, Japan, the United Kingdom, Germany, France, and Italy. This logic is not systematic, however, since Brazil, Russia, India, and China (the “BRIC” countries) contribute relatively little to the budget. Conversely, “major” contributors such as Norway and Canada do not belong (or, in the case of Canada, no longer belong) to the club of permanent members. Since the 1970s, the level of budget contributions to the ILO has been abandoned as a criterion to determine the list of permanent states, in favor of the sole criterion of gross domestic product (although the two are linked, insofar as a country, in principle, contributes to the budget of an organization in proportion to its wealth). Nevertheless, to the present day most members continue to consider this contribution as a factor in justifying their status as permanent member – an impression that contributes to the promotion of a “shareholder” conception of representation. For if representation is considered as a return on investment, the secure position of the wealthiest states within the executive organ of the ILO must rest upon a shared belief that the effectiveness of the ILO depends to a great extent upon the material (financial and human) investment of the major powers.

29Far from being confined to the governmental members alone, this logic has also been widespread since the origins of the organization within the groups of workers and employers. Although there are no permanent seats for the latter, in a sort of mirror effect, we find on their benches, once again, the representatives of workers and employers from the most important industrial countries.

30Alongside these overt logics in which power conditions representation via financial investment, there are other, more discreet logics in play, at least in the recent period, which my fieldwork has been able to bring to light. These relate more to the symbolic investment of members within the ILO. The investment is also connected with the exemplary behavior of members in social matters (this is particularly the case with the Scandinavian countries, which are often billed as “star pupils” of the ILO) as well as in meetings (respect for procedure, constructive contributions to meetings, a sense of compromise), and above all outside of official sessions, where certain delegates do not hesitate to display, or even to conspicuously stage, their “devotion” to the institution: late informal meetings, coordination of negotiations, preparation and writing of reports, communication of information to non-members of the Governing Body, and so forth. One example in this regard is the case of the UK representative, who attributes its status as a permanent member to its greater involvement in the work of the ILO, seeing it more as a responsibility (or even a constraint) than as a privilege:

31

“I think we have to contribute more. This gives us a responsibility, not only to be representative, but to really act for the good of the ILO […] So I see this more as a responsibility than as a privilege […] We have the responsibility to be present when discussions are taking place. What’s more, you’ll notice that, in the meeting rooms, when everyone has left, I’m still there, working.” [48]

32Ultimately, then, the logics of power seem to be undone as investment is individualized in emblematic personalities, such as the charismatic former president of the Workers’ Group, Roy Trotman, who was born in Barbados and was president and spokesman for the Workers’ Group from 2003 to 2011. Being elected or retained as a representative within the Governing Body thus compensates the most “deserving” [49] members, those who have truly paid their dues.

33Where the logic of material investment takes political representation out of the classical democratic framework, with representation becoming almost “unconditional”, as is the case for the permanent members, the logic of symbolic investment brings it back in touch with a more traditional contractual relation between representative and represented. The representative is in the service of the represented and makes the “work of representation” the very essence of his legitimacy – a conception that goes back, classically enough, to what Hanna Pitkin characterizes as “substantial” representation (“acting for”). [50] But such legitimation through action implies the valorization of certain skills at the individual level, once again reinforcing the selective dimension of representation.

Valorization of technical skills: the specialization and individualization of representation

34Since the creation of the first international administrative unions at the end of the nineteenth century, there has been a recurrent technical feature of the work of representation at the international level. [51] At the ILO, the technical nature of debates relates as much to their legal character (for the most part what is being debated are normative instruments) as it does to the complexity of the procedures that frame them within the institution. Many of the delegates questioned during my study conceded that they had been “completely lost” during their first attendance at an ILC meeting, and even more so at those of the Governing Body. When the learning of codes of both a procedural and a linguistic kind [52] is not the object of any specific training provided either by the national institutions or by the services of the ILO, it must take place on the ground, and it demands participation over time, particularly for those who hope to take up a position of responsibility. In this way, a dynamic is set in motion that, by placing the emphasis on technical skills, encourages members to specialize in international matters.

35The graphs below (Figures 1-3), constructed on the basis of a sample of a just over a thousand representatives who had sat on the Governing Body since 1919, illustrate this tendency toward the specialization of representatives over time. Although workers’ and employers’ organizations have always for the most part sent their leaders (presidents and general secretaries), the number of people specifically in charge of international questions has grown (particularly on the workers’ side) since 1945, a tendency that has continued over the last twenty years. The category “other” is mostly made up of vice-presidents, secretaries, treasurers, or retired ex-secretary-generals.

Figure 1

Position of workers’ and employers’ representatives within their own organization, 1919-2014

Position of workers’ and employers’ representatives within their own organization, 1919-2014

Figure 2

Ministerial origin of government representatives, 1919-2011

Ministerial origin of government representatives, 1919-2011

Figure 3

Position of government representatives within the labor ministry, 1919-2011

Position of government representatives within the labor ministry, 1919-2011

36On the government side, whereas the graphs illustrate the relatively weak presence of ministers (although this has changed slightly over the last ten years), it can nevertheless be observed that the work of representation remains essentially in the hands of labor ministry officials, whom Article 11 of the constitution, although relatively silent on the process of selection of governmental representatives, requires to be “qualified”.

37Since this specialization implies both a financial investment [53] (visits to Geneva, training) and a human one (personnel dedicated to international matters), it contributes to the individualization of representation and dissuades both states and professional organizations from frequently appointing new members. The tendency is yet more marked in the Governing Body, where it is not unusual to see the same individuals sitting for ten years at a time – just over three terms – with certain people attaining records of longevity of 20, 30, or even 50 years. Here one might suggest a parallel with Antonin Cohen’s analysis of representation in the European Parliament: “Two characteristics stand out, delineating the contours of a small supranational parliamentary power elite: longevity and legal capital.” [54] These characteristics also give us the profile of an “‘expert’ MEP”. [55] As early as 1929, Ernest Mahaim, then the Belgian representative, used the figure of the “expert” to characterize representatives (in particular workers’ and employers’ representatives) sitting on the ILC: “The conference is not made up of plenipotentiaries. It has a unique composition, which mixes together and places on the same footing representatives of states […] and bosses and workers whom we might compare to experts.” [56] Once again, then, the valorization of technical skills contributes to muddying the contractual nature of representation by blurring the boundary between representatives and experts.

38The tripartite dynamic and the dynamic of internationalization thus combine to give representation within the ILO a highly specialized character. This process once again challenges the traditional framework of democratic representation by conferring a technocratic dimension to representation, something that becomes particularly evident when, going beyond the factual data alone, we examine the complaints of actors who feel “underrepresented” within the ILO.

Representation contested

40In 1919, the systems of representation in force at the ILO were subject to lively challenges on the part of those who felt themselves to be marginalized, or even excluded, from the arenas of representation. [57] Schematically, it is possible to distinguish two major types of claims made during this period: those that took aim at tripartism, and in particular at the insufficiency of the categories “governments”, “workers”, and “employers” to represent the whole of society; and those that targeted the internal practices of representation within the ILO.

41There have been recurrent criticisms of the ILO’s tripartism. A distinction can be drawn between those that point out the incapacity of this type of representation to take into account actors situated in the interstices of the foundational categories of tripartism (for instance, cooperatives, intellectual workers, and independent workers, none of whom fit into a standard employment relationship), [58] and those that criticize the primacy granted to professional organizations and, among them, to specific types of professional organization (namely interprofessional organizations, to the detriment of sector-specific organizations) in the representation of workers and employers. Other criticisms have focused on the historical benefits of representation accruing to certain organizations whose representativeness has been regarded as questionable, at least as judged on the basis of their numbers. This was the basis of the dispute between Force ouvrière and the Confédération générale du travail following the Second World War, the government having nominated Léon Jouhaux, who became secretary of the new confederation, principally because of his central role at the ILO since 1919.

42As early as 1919, and then above all from the 1970s and up to the present day, the question of non-organized workers has been posed ever more acutely, owing to a twofold tendency: the almost universal decline in rates of unionization in industrialized nations, and the major, sometimes predominant, role played by informal work in so-called developing countries as well as in some industrialized ones. For these reasons, the question of non-unionized workers was, and remains, for both national and international institutions, a problem for the legitimacy of delegates: In whose name do the representatives speak? Who do they represent? Their members, or everyone who, whether or not affiliated to a union or employers’ organization, nevertheless necessarily belongs to the wider collective from which the organization emanates?

43At the ILO, this question was acutely brought to light by the case of the Indian NGO Self-Employed Women’s Association (SEWA), created in 1972 and often cited as the typical case of a body promoting an alternative vision of the organization better suited to speak in the name of so-called “informal” workers. [59] Situated at the confluence of the union, cooperative, and feminist movements, SEWA was at first viewed with suspicion by established authorities and unions, at the national and international level alike. In 1981, SEWA split from the Textile Labor Association (TLA) and turned first to the secretariat of the ILO, emphasizing its technical strengths and its expertise in initiatives to encourage the employment of women, so as to obtain international recognition. In a 1982 letter addressed to the director general, the secretary general of SEWA, Ela Bhatt, argued that the problems specific to workers in the informal sector and, in this case, independent female workers, had been constantly neglected by the structures in place, and that only a “separate movement” could make their claims “visible” and “audible”. [60] SEWA openly challenged the traditional model of professional representation, indirectly accusing the union organizations of having, in spite of their universalist discourse, privileged the interests of employed workers in the industrial sector. Thus, as well as the obstacles placed in the path of the recognition of SEWA as a legitimate representative of the interests of independent female workers by the national Indian unions (themselves already in competition every year during the composition of the Indian ILC delegation), SEWA also had to contend with suspicion from the ICFTU. In spite of the support of the secretariat, and then the International Textile, Garment, and Leather Workers’ Federation, which included SEWA in its delegation as an observer, it would not be until 2006 that SEWA was admitted as an affiliate organization of the ITUC.

44The other major form of challenge has taken aim not so much at the fundamental question of tripartite representation but rather at the dynamics of internal representation at the ILO. This critique is the direct result of factors discussed in the second section of this article: the predominance of an anticommunist social-democracy-type unionism on the side of workers, the promotion of the market economy and free enterprise on the side of employers, and the general domination of industrialized nations (“Europeans”, “Western”, or “Northern” ones) in all arenas of representation within the organization. Since 1919, countries such as India and Brazil, but also Canada, have challenged the Eurocentric and undemocratic character of representation within the Governing Body, as much from the point of view of governments as from that of workers and employers. The argument was soon being made for a democratic and equitable representation guaranteed by the election of all members, against the existing mechanisms of representation, denounced as “privileges”. [61] Here lies the whole ambiguity of the Governing Body, which members expect, despite its executive role, to also be representative of the organization as a whole. This demand for representativeness is also found at the level of ILO officials, whose recruitment (in particular in terms of their nationality) is scrupulously monitored by members.

45Such challenges became more radical in the 1960s, to the point of paralyzing the ILO’s work, as the question of the “democratization of structures” was systematically forced onto the organization’s agenda, something that continued up until the end of the 1980s. [62] Among the propositions put forward to palliate what the marginalized members considered as “deficits of representativeness” were measures already often in force at a national level, such as quota systems (particularly by region), proportional representation, rotation, the broadening of the Governing Body, or even the pure and simple abolition of unelected seats. From this point of view, the ILO can hardly be regarded as an innovator relative to national institutions. Less forceful but present nonetheless, the claim for greater participation of women would ultimately also be expressed in the ILO by feminist organizations. [63]

Representation reformed: between adaptation and controlled improvisation

46In spite of the apparently strict and selective nature of representation at the ILO, the regulatory arrangements that govern the relation of representation, and above all the actors from which it originates, bear witness to a certain capacity for adaptation.

47In terms of the challenges directed at tripartite representation in general, in particular those from countries where professional organizations are absent or poorly structured, it was agreed in 1919 that it was a matter for the governments concerned to choose “on their sole responsibility” the appropriate organizations. Although the workers’ and employers’ groups would prove extremely vigilant in ascertaining these organizations’ independence from government, there was tacit consensus as to the necessity of “filling” the pre-established categories. Anything – even the appointing of representatives from organizations whose representativeness was weak, or indeed highly questionable – was preferable to sending an incomplete delegation. Tripartism having been the founding stone of the organization’s legitimacy, the members’ commitment to preserving it led them to make certain compromises. This spirit is summed up well by Canada’s representative for employers, who are represented at the ILO by the Canadian Employers Council, an organization that specializes in international questions, but which is not recognized as being the most representative at the national level: “Maybe we’re not the most representative, but we’re all they’ve got.” [64]

48Once an arrangement is reached, a “ratchet effect” (effet de cliquet) tends to assert itself, so that the nominated organizations rarely change (this also occurs for reasons related to specialization, as mentioned above), reinforcing the impression of a “premium on seniority” as far as representation is concerned. A change of representative organization is not impossible, but it requires a prior agreement with the structures already in place.

49As far as the complaints of NGOs are concerned, the ILO has partially resolved the problem in two different ways since 1919 and the controversy ignited then by the absence of cooperative movements, which were very influential at the time. [65] In the 1920s, it was decided that representatives of cooperatives, which strictly speaking belong to none of the three categories of tripartite representation, could be included in delegations as “technical advisors”. Unlike a delegate, a technical advisor takes part in negotiations but has no vote. In fact, technical advisor status was originally devised not so much to make up for delegates’ lack of competence in certain “technical” matters as it was to allow for the representation of certain categories of workers (foremost among them agricultural workers, but also women workers) when questions concerning them were on the agenda. Although this type of representation did not satisfy the cooperatives, the status of technical advisor has since then been widely used to allow the representation of minority professional organizations, in particular those from nations with several unions and employers’ organizations, and also other NGOs. Moreover, the repeated demands of cooperatives for representation distinct from that of governments, employers, and workers led to the creation, in 1948, of the status of observer (or consultative status, since it carries no voting right) – a status that today allows certain NGOs that do not fit into the classic framework of professional representation to be represented at the ILO.

50At the level of the Governing Body, the ILO has innovated by creating a status that was not foreseen in the founding texts: that of deputy member, which is modeled on technical advisor status and gives (at least partial) satisfaction to countries that feel marginalized, by ensuring a de facto broadening of the Governing Body. Moreover, insofar as the majority of decisions are taken by consensus – with the notable exception of the election of the director general – the difference between regular member and deputy member is in fact largely symbolic. The creation of another status for members who are not elected to the Governing Body, that of state observer, which allows states not elected to the Governing Body nevertheless to take part in its working process, is a device whose success is clearly attested to, judging by the ever-growing number of states that take advantage of this status to sit in on the meetings of the Governing Body. [66]

51As far as the underrepresentation of developing countries is concerned, many types of arrangements have been put in place. Firstly, constitutional and regulatory arrangements that allocate a precise number of seats for each broad geographical region (Africa, America, Europe, Asia-Pacific) have been established. Secondly, there are arrangements between members, whether it is a matter of members of the same region making agreements between themselves for the rotation of members, or of directors of employers’ or union organizations organizing themselves upstream of elections in order to make space for an “outsider” candidate. This was the case for workers’ delegates from the USSR during the Cold War, and it is the case today for China, the ITUC having committed in the twenty-first century to presenting on the list one candidate less than the number of available seats, thus opening the way for a Chinese delegate to be elected.

52If the rules governing representation, when subjected to criticism, can be extended in this way and can be subjected to various reappropriations by the actors concerned, these more or less formal transactions – which testify to a certain institutional improvisation or even imagination – are all the more possible when they concern arenas and positions which are unfamiliar, or are considered relatively unimportant, by national institutions. Conversely, certain positions always seem non negotiable: permanent seats on the Governing Body, which always remain in force, and certain positions (the presidency of a group, for example) or mandates within committees seen as strategically important.

53* * *

54Although the ILO uses the reference to parliament for the ends of organizational legitimation, I have shown that it applies at best only partially and at certain points, given the marginalization of the electoral relation, controlled access to representative arenas, and the specificity of the required profile of the representative – all of which accentuate and reinforce over the long term the technocratic dimension of representation. Moreover, to a great extent the ILO reproduces the relations of power that structure the international scene: relations between state and non-state actors, between industrialized and developing states, between European and non-European states, and between states of the North and South. Yet in allowing various critiques to be aired, in more or less formal spaces, it remains able, through partial reforms, to defend its claim to be representative. Such critiques appear to represent limits on the claim of ILO members to represent the world of labor at the global level, universally and authentically. Analysis of their reception, however, allows light to be shed on some limited yet tangible extensions of both the conceptions and mechanisms of representation at work within the ILO. For, in spite of the durable and apparently fixed nature of the criteria for the selection of representatives, mechanisms of representation can be invented, challenged, and reappropriated by the actors concerned, for ends other than those for which they were originally conceived. This does not take place, however, without generating a certain confusion over the status and role of the representative and what exactly distinguishes him or her from the official, the expert, or the charismatic leader.

55From the point of view of a more general reflection on contemporary forms of representation, the ILO constitutes an interesting entry point from which to interrogate the principles and categories of democratic representation (election, general interest and particular interest, group, party, delegate, representative, and expert) that the organization has espoused since its creation. If on first sight this configuration of tripartite, international representation seems to scramble the categories of the classic analysis, it also calls for a broadening of their scope beyond the national framework. The ILO, like any institution, is affected by the classic dilemmas of representation within an organization (the delimitation of members from nonmembers, the choice of criteria for the selection of representatives, and the arbitrage between functional and democratic logics of representation).

56Although analysis of these dilemmas through the prism of the challenges of actors in the ILO allows light to be shed on a relative, controlled improvisation of the institution as far as representation is concerned, it nonetheless leads to a blind spot, opening up new perspectives for research on the contemporary mutations of representation. The question of why and how relatively marginalized actors (at the national or international level) manage to find room for maneuver so as to make themselves heard or better represented could be answered by looking into the existence (or otherwise) of untapped resources within international organizations, and also by exploring the hypothesis that these organizations (and particularly those which specialize in the social domain such as the ILO) have not particularly captured the interest of national elites.