Introduction
‘Traditional’ forms of industrial relations in the global north have been badly eroded by processes of neoliberal globalisation. The International Labour Organization (ILO) – historically closely associated with ‘tripartism’ and conventional industrial relations – has been threatened by these developments (e.g. Cox 1977; Standing 2008). The organisation’s response to these challenges, the 1999 ‘Decent Work Agenda’ (DWA), re-emphasised a handful of protections against the worst abuses of labour rights alongside measures to promote employment. Tripartism has been retained as a pillar of ‘decent work’, albeit articulated more loosely as ‘social dialogue’ making room for other ‘civil society’ actors (Fashoyin 2005).1 The erosion of ‘traditional’ industrial relations, then, would seem to have posed considerable challenges for the ILO.
However, the ILO has also long operated in places where ‘traditional’ labour relations have scarcely ever prevailed – not least sub-Saharan Africa. For students of African political economy, then, the question is whether or not the promotion of ‘tripartism’, or ‘social dialogue’ in its newer form, is likely to be of benefit to African workers. On one level, the question could be answered by critiquing tripartism and the ILO’s approach to labour relations more broadly. Here I take a somewhat different tack: I suggest that, while ‘tripartism’ is a problematic idea, the ILO’s practical efforts to promote tripartism have often had impacts that are difficult to predict from the intents of the ILO’s programming alone. It is less the ILO’s vision per se and more the messy encounters between that vision and the real balance of forces in national and regional political economies that really matter.
In the following section, I outline the concept of ‘tripartite fantasies’ to capture the essential tensions that have developed around the ILO’s interventions in sub-Saharan Africa. Drawing on archival and interview research conducted in Geneva and at an ILO field office in Pretoria, the remainder of the article traces out the ILO’s efforts to provide assistance to African trade unions from 1960 to the present. As Croucher and Wood have recently noted (2015, 353), the role of the ILO in promoting tripartism in the global south remains largely unexamined. While initiatives for trade unions have never been publicised to the same extent as the ILO’s conventions and recommendations, in practical terms such ‘development’ assistance activities have long taken up the bulk of the budget and staff time of the organisation. Meanwhile, the focus here on initiatives dealing with workers’ organisations is justified because these programmes are a point where the tensions surrounding the tripartite fantasy are particularly acute. Workers’ education programming has very often involved efforts to produce particular kinds of trade unions. A similar word is in order on the empirical focal points of the paper. I spend the bulk of the paper on developments from roughly 1960–1990, followed by a brief consideration of contemporary trends. This is a useful exercise because it underlines the more fundamental issues at stake – whatever problems we might find with the tripartite fantasy are not solely down to its obsolescence in the face of neoliberal globalism. I focus on South Africa in the discussion of developments since 1975, for two reasons. First, South Africa has historically been of particular importance for the ILO in the region: assistance to unions under apartheid was a key focus in the 1970s and 1980s and the country has taken on an increasing importance since 1994. The second reason is that South Africa is probably the jurisdiction in sub-Saharan Africa where the ILO’s tripartite fantasy comes closest to being realised – the country has a comparatively large and well-organised trade union movement, and well-established formal tripartite institutions. The limitations of the ILO’s vision in South Africa are often only amplified elsewhere.
Tripartite fantasies
Tripartism is an institutional form more than a specific set of policies. It refers to formal institutions for economic policy-making in which government mediates between workers and employers. Tripartite institutions are nonetheless built on the assumption of cooperation and compromise between workers, employers, and the state – they are often, in short, based on a reformist conception of labour politics. The emergence of the ILO, and of international labour governance more broadly, was very closely linked to efforts to deradicalise labour as a political force. The promotion of international labour standards emerged as a conservative response to revolutionary pressures stemming from the growth of trade unions, the expansion of the franchise, and the rise of socialist parties in the latter part of the nineteenth century in Europe. International agreements on labour standards were first articulated alongside the ‘new nationalism’ linking ‘policies for industrial expansion, protection for workers, and increased state power’ (Cox 1996, 46). The establishment of a formal institution to oversee international labour standards after World War I, moreover, was in no small part a reaction to the Russian Revolution. As Croucher and Wood (2015) note, the history of tripartite institutions is closely linked to the history of class compromise in Europe.
This connection between the ILO, tripartism, and class compromise in Europe suggests a question: how does the idea of tripartism translate in the rest of the world? Efforts to extend tripartite structures into Africa and elsewhere date to the mid twentieth century. The ILO became increasingly involved in governing colonial labour practices in the 1930s and after World War II (see Daughton 2013; Maul 2012); this extension of the ILO’s mission to the colonised world also set in motion long-standing efforts to promote tripartite institutions outside Western Europe. A number of developing countries, not least in southern Africa, have adopted nominally tripartite institutions of economic governance, often with the support of the ILO. Recent studies suggest some conclusions about the impacts of tripartism in Africa. Fraile and Baccaro (2010, 5) argue that tripartite institutions in developing countries have ‘blunted’ the worst excesses of neoliberal policies, rather than construct alternatives. Studies of southern Africa similarly suggest that the outcomes of tripartite institutions installed during the process of democratisation have often been disappointing for workers (Dibben, Klerck, and Wood 2015; Webster 2013). The political dominance of single-party governments, growing opposition from business, and the global rise of neoliberalism have restricted the capacity of tripartite institutions to advance workers’ interests.
There are also deeper problems with the concept of tripartism. Tripartism is ultimately rooted in a normative preference for reformism, compromise, and consensus building over radical change or agonistic politics. This has meant that the ILO has rarely been attentive to power relations, political conflicts, or structural factors that might stand in the way of the effective functioning of industrial relations institutions. Caraway (2006), for instance, argues that the ILO’s conception of ‘free’ unionism encourages fragmentation. ‘Free’ unions can thus come at the expense of ‘powerful’ unions: indeed, encouraging the formation of rival trade union centres under the guise of promoting ‘free association’ is a means by which governments and businesses have sought to undermine politically troublesome unions in some instances. Selwyn (2013) argues in broader terms that the ILO’s emphasis on institutional reforms, embodied in the DWA, promote ‘top-down’ solutions to labour abuses, leaving it up to governments and corporations to treat workers better. This closes down considerations of the forms of exploitation which create ‘indecent’ work in the first place, or of the place of ‘bottom-up’ mobilisations in securing workers’ rights.
These arguments suggest a more fundamental puzzle. If tripartite institutions rarely work well outside the historical contexts in which they emerged, what does the practice of promoting tripartism actually accomplish? I suggest that it is useful to think about the role of the ILO in this respect as a set of practices linked to a kind of ‘tripartite fantasy’. The ‘tripartite fantasy’ is a particular set of assumptions about what constitutes ideal labour relations. Perhaps most importantly, this involves a troubled conception of the purpose and functions of workers’ organisations. The tripartite fantasy involves a commitment on one hand to ‘free’ trade unionism – meaning freedom of association and autonomy from government and political parties – and on the other, an understanding of the ‘appropriate’ role for trade unions based on depoliticised and non-radical participation through established institutional channels. These two dimensions do not always easily co-exist. The term ‘fantasy’ highlights the fact that this vision has deeply coloured the way in which the ILO’s officials have interpreted and interacted with the structures of political economy with which they have been confronted. The key argument of this paper is that it is in the troubled translation between abstract policy ideas and the contested political economy of labour relations in the real world that the really significant impacts of organisations like the ILO are often felt.
Highlighting this set of tensions offers a pessimistic assessment of the ILO’s ability to improve the situation of African workers, but also underlines the ability of workers’ organisations to twist ILO programming to their own ends. The following pages show that union leaderships have not always advanced the interests of rank-and-file workers in reappropriating the ILO’s resources, but that their actions have often profoundly shaped the practical impacts of the ILO. African workers will not find the tools they need to confront neoliberalism in conforming to the ILO’s vision, but they might still find some use in engagements with the ILO.
Workers’ education and decolonisation
The ILO launched a programme for workers’ education in 1956 – which was expanded into sub-Saharan Africa after 1960. The Director General’s report to the first African Regional Conference included a full chapter on workers’ education (ILO 1960). The ILO understood the purpose of workers’ education in terms of the tripartite fantasy – seeking to encourage the formation of trade unionism suited for consensual bargaining in workplaces and national-level institutions. Workers’ education, in this respect, was often concerned with bringing about ‘appropriate’ subjectivities on the part of workers. Training materials from a mission to Zambia in 1967, for instance, included the following on the conduct of union stewards:
A man who readily concedes plainly improper grievances but fights hard for those with some substance gains the respect not only of his members but also often of management … A man who shouts and threatens on every occasion and who generally puts on a dramatic performance for the benefit of his members – and then ends up being made a fool of and losing the case – is not a man who commands respect. (ILO 1968, 32)
The practical activities of the ILO’s experts were often impacted by struggles stemming from this political-economic context. The Workers’ Education department (WED) launched a pilot project on trade union education in West and Central Africa in 1965. The activities of this project illustrate the conflicts in which the ILO found itself entangled. For instance, shortly before one expert was scheduled to make a trip to Upper Volta to lead a seminar in January 1966, workers launched a general strike to protest against austerity measures adopted by the government. In response to the strike, the government banned all trade union meetings – including the ILO seminar (ILOA 1966a). After the strike and associated protests brought down the government, the new regime reversed course and allowed the seminar to go ahead. The participating confederations even established a joint committee for workers’ education, in order to receive further ILO assistance (ILOA 1966b). Ultimately, however, the joint committee lasted only a few months before splintering into rival committees, which prevented the education of any more than a few dozen workers (ILOA 1965).
These national struggles were overlaid with conflicts at the regional and global levels between conservative nationalist unions (often closely affiliated to single-party governments), ‘independent’ unions with affiliations to the International Confederation of Free Trade Unions (ICFTU), and radical unions with links to the communist-led World Federation of Trade Unions (WFTU). This meant that in any given country multiple government and trade union donors were often operating separate workers’ education programmes. It concerned ILO officials that ‘donor organisations seem sometimes to be more concerned with outbidding each other to establish clients for themselves’ (ILOA 1972c). Several pan-African union confederations were also in place. The most notable were the All-African Trade Union Federation (AATUF) – based initially in Accra, then in Dar es Salaam – and loosely linked to the WFTU, although typically more defined by anti-colonialism than communism; and the African Trade Union Confederation (ATUC), based in Dakar, whose members were mostly affiliated to the ICFTU. More conservative unions favoured the formation of a single regional federation backed by the Organisation of African Unity (OAU) – along the lines of the Organization of African Trade Union Unity (OATUU) eventually established in 1973 (see Agyeman 2003).
Trade union unity?
By the early 1970s, the ILO clearly saw the promotion of ‘unity’ as a potential antidote to the factional divisions that frequently derailed projects on the ground. ILO officials responded positively to a resolution on trade union unity by the OAU’s Conference of African Labour Ministers (CALM) in 1969 (ILOA 1969). Yet, it was impossible to reconcile any compulsory policy of ‘unity’ with the principle of freedom of association. An internal comment on a draft speech to be delivered at another CALM meeting in 1972 noted that ‘Trade union unity whether at the national or continental level is a delicate topic … It would be advisable … to inform the conference that the ILO attaches great importance to the development of responsible, free, and democratic trade unionism’ (ILOA 1972a). ‘Unity’, then, was appealing but difficult to balance with the ILO’s commitment to ‘free’ unionism.
In spite of the pitfalls, though, the ILO supported the formation of OATUU. Almost immediately after its formation, the Secretary General of OATUU wrote to the ILO asking for an ‘extensive Worker’ [sic] Education Programme at national, regional, and continental levels’ (ILOA 1974). WED started lobbying for assistance to OATUU under the ILO’s technical assistance budget (ILOA n.d.a). It was difficult to avoid the fact that OATUU was not an ‘independent’ representative of workers’ interests. OATUU (and many of its member unions) made little effort at establishing any independent fundraising capacity – e.g. a check-off system or regularised collection of member dues – relying instead on the OAU, international donors, and governments (Ananaba 1979, 219–220). OATUU was committed from the start to the solidification of singular ‘non-political’ (read: closely controlled by the ruling party) trade union centres. In short, OATUU was dependent on its ability to leverage assistance from both the OAU and the ILO – a potentially difficult balance made possible largely by the ILO’s interest in ‘unity’.
Given the importance of its external linkages, OATUU sought to cement its role in the ILO’s WED activities in the region (ILOA 1977). In 1978, the OAU Assembly called for a ‘joint action programme’ of the ILO, OAU and OATUU on ‘labour and related social questions’ (ILOA 1978a). The ILO’s regional director responded positively, noting (ironically) the difficulties raised by participating in uncoordinated projects with international confederations or bilateral donors, which ‘have their own objectives which may or may not be consistent with ILO objectives’.2 In 1979, OATUU’s General Council passed a resolution that ‘any assistance of a multilateral nature … by any extra-African organisation must necessarily pass through OATUU as the only expression of willingness for progress and for the promotion of African workers and Trade Unions’ (ILOA 1979). No formal machinery was ever put in place, but OATUU was increasingly closely involved in WED activities. One of the most notable areas of activity was aid to national liberation movements (NLMs) in southern Africa.
The ILO, OATUU, and national liberation movements
The ILO played a significant role in international opposition to apartheid – South Africa was forced to withdraw from the organisation in 1964. The ILO hosted a ‘Workers’ Conference on Action Against Apartheid’ in Geneva in 1973, which called for aid to ‘the oppressed people of South Africa’ (ILO 1973, 2). There were also internal proposals as early as 1972 for technical assistance to exiled unionists (ILOA 1972b). Assistance to workers on the ground, however, was difficult for the ILO to provide – ironically, precisely because South Africa had been pushed out of the ILO. Officials noted that if unions in host countries were to organise educational activities for exiles, the ILO could provide assistance without raising issues of jurisdiction (ILOA 1972d). The ICFTU was initially more successful in establishing links with workers’ organisations in the white-ruled countries. The formation of OATUU in 1973 and the formal support of the OAU for NLMs thus gave the ILO an important entry point. Nonetheless, OATUU’s connection with a particular vision of ‘trade union unity’ that sat awkwardly with the ILO’s tripartite fantasy created some notable tensions. These were compounded as OATUU and the ICFTU became increasingly aligned with exiled and independent unionists, respectively.
The first major ILO event for NLMs was a workshop held in Lusaka in 1978 with the Zambia Congress of Trade Unions (ZCTU) and OATUU in support of trade unions affiliated to national liberation movements in southern Africa. South Africa, notably, was represented by the exiled South African Congress of Trade Unions (SACTU). The ILO officials involved delivered rather banal talks about administrative structures, ‘normal’ trade union functions, and different ways to provide trade union education. One official included an appendix with a table giving a detailed breakdown of 18 different techniques for facilitating discussion (ILO 1978, 59–63). The ILO’s archive files relating to the workshop even contain printed lyrics to British and American trade union songs – e.g. ‘Solidarity forever’, ‘Roll the Union on’, ‘We shall not be moved' (ILOA n.d.b). Rather like the mission to Zambia 10 years earlier, then, the ILO’s contributions centred on producing the particular kinds of subjectivities required by the tripartite fantasy.
Meanwhile, the OATUU and ZCTU officials present addressed the conference to talk mostly about the need for labour movements to work with governments for national ‘development’. An OATUU speaker critiqued the idea of trade union ‘independence’:
Some people think that collaboration of a trade union organization with a government is a sell-out … In Africa, governments and workers are allies in the struggle for development. Who is subordinate to who is not even the question. (ILO 1978, 38, emphasis added)
Meanwhile, SACTU was increasingly coming under pressure from the formation of independent trade unions in South Africa. The Federation of South African Trade Unions (FOSATU) was founded in 1979 and the Council of Unions of South Africa (CUSA) in 1980. The independent federations organising in South Africa scarcely had contacts with OATUU. FOSATU depended on the ICFTU for the bulk of its budget – in 1982, 84% of its operating costs were paid out of ICFTU assistance (WHP n.d.a). These developments posed significant challenges for SACTU, which suddenly found itself competing for international resources (Plaut 1984, 118–119). The ICFTU, meanwhile, was concerned with limiting SACTU’s access to international assistance. A 1984 discussion paper of the ICFTU’s Coordinating Committee on South Africa suggests that ‘It is disquieting to note that SACTU is increasingly trying to take credit for developments inside South Africa as this could seriously prejudice the progress of the independent trade union movement’ (WHP 1984).
OATUU, meanwhile, responded to these developments by seeking to lock in the ILO’s commitment to the exiled unionists. The Secretary General wrote to the ILO in 1982 requesting help in setting up a seminar for South African trade unions. OATUU fell back on familiar arguments about ‘unity’:
There are many organisations … who are very active in South Africa, under the pretext that they are helping the workers in South Africa and those who are in the neighbouring states … . These organisations have caused confusion inside South Africa, where they are now having five national centres, but they are also beginning to cause confusion even in the independent states in Southern Africa. (ILOA 1982a)
Yet, WED was simultaneously pursuing projects with the independent federations. WED was not especially successful overall, but did manage to secure funding from the Canadian Labour Congress (CLC) and the Canadian International Development Agency (CIDA) for a workers’ education project for unions in South Africa in 1983. FOSATU was generally receptive, although the ICFTU saw the project as an infringement on their ‘territory’.3 The ILO also hosted meetings of the ICFTU’s Coordinating Committee on South Africa (COCOSA) in Geneva, in which FOSATU and CUSA representatives participated. In general, though, the ILO was clearly hamstrung by internal debates, the relationship with OATUU, and their inability to work in South Africa itself. A FOSATU official noted in 1982 that ‘it is very clear that within the ILO this whole program of support is very controversial and raises a number of problems about the ILO’s relation to South Africa as a non-member and to the liberation movements’ (WHP 1982). Assistance to independent federations in South Africa was difficult for the ILO to provide, ironically despite the organisation’s broader preference for ‘free’ trade unions.
The situation became somewhat less complicated as the decade wore on. The relationship between the ICFTU and the independent federations became difficult. FOSATU and CUSA members were marginalised in COCOSA. A FOSATU official reported of the 1981 meeting that ‘The real business of the meeting took place without us being referred to, and a FOSATU contract was agreed to without me being called in to discuss it’ (WHP 1981). FOSATU developed a cautious approach to internationalism (WHP 1983b). To a considerable extent, this approach was carried over after the merger of independent unions into the Congress of South African Trade Unions (COSATU) in 1985. COSATU refused to affiliate to the ICFTU, preferring bilateral relationships with other national confederations.
Further, OATUU’s 4th Congress in early 1985 collapsed. A group of unions raised a number of allegations about the secretariat misusing funds, manipulating votes, and a failure to ‘take seriously the preoccupations of the African working class’ (ILOA 1985a). The Secretary General refused to resign, but a rival Provisional Coordinating Committee was established (despite his objections). One of the immediate consequences was that WED cancelled a Danish International Development Agency (DANIDA)-funded multi-country project for OATUU (ILOA 1985c). The OATUU Secretary General responded by accusing WED of having ‘people or organizations who are inimical to OATUU’ influence the cancellation of the project and that ‘our confidence in your impartiality and fair judgement has been greatly shaken’ (ILOA 1985b). The specifics are less significant for present purposes than the fact that the entire basis for OATUU’s existence was under threat. OATUU’s model of ‘pan-African’ unity was rooted in a particular model of single-party government and extensive incorporation of ‘civil society’, all supported by the rents collected from control over primary exports – that was profoundly in crisis. With its patrons in the OAU under strain, OATUU persisted mostly because of its ability to leverage resources from northern donors through the ILO – despite the gap between the ILO’s tripartite fantasy and OATUU’s political commitments. This greatly amplified the threat from the cancellation of the DANIDA project.
The ILO thus started pursuing avenues of cooperation with COSATU. WED officials met with COSATU leadership in Lusaka in March of 1986 to discuss establishing a workers’ education programme for COSATU, but the inability of the ILO to work in South Africa remained a stumbling block.5 Indeed, ironically, after the leadership situation in OATUU was resolved in favour of the provisional committee group, COSATU officials started to explore the possibility of establishing links to OATUU. In the end, OATUU being increasingly limited in terms of resources, this amounted to very little. ILO events were nonetheless important to these developments insofar as they provided opportunities for COSATU and OATUU officials to meet (WHP n.d.b). Nonetheless, COSATU’s generally cautious approach to international unionism helped reinforce the situation in which the ILO continued to deal much more easily in ‘global’ policies to mobilise pressure against apartheid than in direct assistance to workers on the ground. COSATU did seek to work through the ILO in some instances – most notably, COSATU protested to the governing body of the ILO in 1988 over reforms to the Labour Relations Act that the union felt would infringe on the right to strike and favour racially constituted unions over non-racial ones. A ‘Fact-Finding and Conciliation Commission’ would eventually be sent to South Africa, although this was delayed several years because of South Africa’s status in the ILO. Notably, rather unlike SACTU – which sought to draw on the ILO and other international links to compensate for their organisational weakness – the ILO here provided an alternative outlet for COSATU in responding to proposed policy changes from a hostile government rather than a source of material and political resources. The broader point here is that, whatever the shortcomings of the tripartite fantasy in its own right, what often mattered most in the ILO’s interactions with the movement against apartheid were the struggles between the actors involved to reappropriate ILO spaces and programmes towards very different ends.
Tripartite fantasies in contemporary South Africa
There have obviously been dramatic transformations in both the relations of production and in political institutions across the region since the late 1980s – the end of apartheid, the crisis of the postcolonial state, and the widespread and rapid growth of informal work are all notable. Significantly, though, while tripartite institutions might succeed at times in ‘blunting’ neoliberalism (Fraile and Baccaro 2010), the limits of the tripartite fantasy persist. The ILO continues to have to fall back on trying to encourage trade unions to behave ‘properly’ according to the terms of the tripartite fantasy. This emphasis, indeed, has been deepened by the incorporation of promotion of national institutions for ‘social dialogue’ in the DWA (Fashoyin 2005). South Africa is increasingly a key terrain on which these problems are grappled with. COSATU’s role in post-apartheid South Africa – especially its involvement in policy-making through the National Economic Development and Labour Council (NEDLAC) – came closer to the ILO’s tripartite fantasy than any other workers’ organisation in the region. Indeed, its significance for the ILO in Africa extends well beyond South Africa itself – the ILO has even arranged study visits to NEDLAC for trade unions in Zimbabwe and Swaziland.
In practice, it is hard to escape the conclusion that tripartism in South Africa has not done much for workers. South Africa’s corporatist institutions have been challenged by employers seeking greater flexibility, and democratisation more broadly has not always delivered on ‘expectations of more and better jobs’, especially as employers are often seen ‘bypassing the new labour laws’ (Webster 2013, 210). Economic policy after South Africa’s transition away from apartheid has been profoundly shaped by a history of compromise between the ANC and big business, especially leading firms in finance and mining (Bond 2000; Bassett 2008; Fine 2012). Despite the existence of institutions like NEDLAC, trade unions have often played a relatively marginal role in making policy. Neoliberal economic policies, then, have tended to prevail.
The weaknesses of tripartism are compounded by growing challenges facing the labour movement. As a number of authors have argued, the close alliance of COSATU’s leadership and the leading factions of the ANC seems less and less likely to serve the interests of rank-and-file workers (Bassett and Clarke 2008; Beresford 2012).4 Of course, these political links with the ANC have long been contested within COSATU (Buhlungu and Tshoaedi 2012, 3–4; Southall and Webster 2010). COSATU and formal tripartism have also been undercut to some extent by structural changes in the labour market. It is increasingly difficult for conventional unions to represent the frustrations of poor communities and individuals faced with persistently high unemployment, increasingly precarious work, and growing inequality in South Africa (Scully 2016); an inability reflected in part in a growing wave of community protests (Alexander 2010; Paret 2015). The point is that, while COSATU doubtless remains a crucial instrument for the representation of workers’ interests, both in political terms and in the workplace, it faces growing structural and political challenges, which deepen the limitations of institutionalised tripartism.
The moralism underpinning the tripartite fantasy, however, often obscures the underlying political and economic shifts to the ILO’s officials, meaning the ILO is unlikely to be of much direct help. The close links between COSATU and the ANC are viewed with increasing unease at the ILO, but are attributed largely to leadership failures:
Some people are saying for instance that COSATU, which has very close links to the ANC … some of the members are starting to fight that kind of bureaucracy, because they’re finding that the leaders are not looking after the main members of the organization. And you can see in the situation in Marikana that the workers are lost and they don’t know who is there to help them and they try to sort out their problems by themselves. (ILO official, Pretoria, 16 October 2014)
In [South Africa], they’ve almost moved away from [tripartism], and that’s been their downfall I believe in the Marikana, and all sorts of the challenges they’re having around social dialogue … . And countries that do take it up – and you just need to look at the Swedish and Scandinavian models, even though they’re developed economies – you’ll see how entrenched social dialogue is in their societies, which helps them get through difficult times. (ILO official, 27 October 2014)
The presence of multiple trade union centres continues to pose operational problems for the ILO, which have been reinforced by the fact that COSATU has shown little interest in workers’ education activities. While the ILO is keen to promote the unity of South African trade unions, this has opened up opportunities for the two much smaller ‘national’ confederations (the National Council of Trade Unions and Federation of Unions of South Africa) to draw on ILO support, and for COSATU affiliates to participate in programming independent of the central federation. The National Union of Metalworkers of South Africa (NUMSA), for instance, was providing a financial education course for retrenched members, with ILO assistance, at the time of its expulsion from COSATU. For the ILO, then, the desire to promote ‘unity’ continues to run up against the need to find willing participants for its programming. Here, though, is at least a small hint of the kind of positive role the ILO could play (in spite of itself): sectoral federations have been able to draw on the ILO for certain limited kinds of material and organisational support. Even for trade unions, like NUMSA, committed to far more radical positions than those of the ILO, it might be possible to draw on ILO support on particular issue areas.
In a somewhat similar vein, despite COSATU’s reluctance to draw on ILO technical assistance, the ILO is identified as a potential ‘strategic lever of power’ with which to ‘further a working class agenda and to facilitate the transformation of global multilateralism’ (COSATU 2012, 23). Calls for policies to promote ‘decent work’ have also been increasingly incorporated into COSATU’s official platforms and congress resolutions. Insofar as COSATU has drawn on the language of creating ‘decent work’ to articulate calls for redistributive policies and for greater state intervention in the economy, this represents a potentially progressive way of engaging with the resources produced by the ILO. But this strategy is also potentially fraught. The Zuma government’s ‘New Growth Path’ (NGP) announced in 2010 is also explicitly framed around the DWA (EDD 2011). The NGP does propose expanded social grants for poor communities and investments in infrastructure, but also threatens to roll back previous gains made by organised labour – most notably by moderating wage increases for workers earning between R3000 and R20,000 per month in order to limit inflation (EDD 2011, 58). These aspects of the NGP (rightly) drew criticism from COSATU – whose response also noted the document’s ‘over-reliance on corporatist social consensus’ (COSATU 2011) – and from a number of academics (Fine 2012).
The point is that whatever positive impacts the broader DWA might have in South Africa or elsewhere are likely to come about through these kinds of struggles to define and appropriate the language of ‘decent work’ rather than through the ILO’s own efforts (see also Bernards 2015; 2016). The ILO is still highly limited in its ability to act in the interests of workers in its own right by the constraints posed by the tripartite fantasy. The fantasy contains a set of assumptions that remain deeply moralistic. Marikana, where police killed 34 striking miners at a Lonmin platinum mine in August 2012, appears in the above quotes as a kind of spectre of the disorder that might result from failures in leadership or the abandonment of tripartite values by workers, government and employers. Of course, this assessment fails to do justice to the massacre as a complex phenomenon with multiple causes and ambiguous consequences (Bond 2013; Alexander 2013; Chinguno 2013). The tripartite fantasy is, ultimately, a fantasy precisely because it ignores the realities of political contestation and power.
Conclusion
The fundamental question here is whether or not the ILO can be expected to make a positive contribution to workers’ livelihoods. The obvious conclusion from the evidence above is that the tripartite fantasy is ill-suited to the task, and this is as true for contemporary workers faced with struggles over neoliberal development strategies and precarious labour markets as much as those involved in struggles over political independence or against apartheid in the postcolonial era. The moralistic and ‘top-down’ character of the ILO’s ideas often inhibits their usefulness. Yet, there is perhaps reason for hope, in an ironic sort of way, in the contradictory and contested character of interventions in practice. This is the real value in tracing out the myriad struggles through which the tripartite fantasy has been enacted. The results of the ILO interventions discussed above have more often than not been more reflective of the balance of forces ‘on the ground’ in any given context than of what the ILO wanted. Paradoxically, this means that ILO interventions have at times been both better and worse for workers than we might expect from the objectives of the organisation on paper.
So, the big point is that much depends on the actions of workers themselves in their relations with the ILO. Materially and organisationally weak trade unions have been able to draw on the ILO’s support to maintain themselves. This has at times been a substitute for undertaking the more difficult work of building stronger organisations, giving voice to rank-and-file workers, or forging links with other social movements. OATUU was able to take advantage of the ILO’s WED programming as part of a broader strategy drawing on international resources to make up for its weak organisational base. However, the ILO has also at times made it easier for unions to operate in harsh environments. Here, the limited assistance the ILO did provide to the independent South African federations under apartheid is notable. Even things as simple as providing safe spaces to meet for South African unionists with potential allies (in Geneva or at regional events) are noteworthy. So too are current efforts by unions and others to define ‘decent work’ in South Africa and elsewhere. Ultimately, there is little reason for African workers to put much faith in the tripartite fantasy, but there is scope for the ILO to benefit workers in spite of itself.