“We feel that our strength is on the factory floor”: Dualism, shop-floor power, and labor law reform in late apartheid South Africa

In the seven years since the Marikana Massacre of 16 August 2012, post-apartheid South African labor relations have entered a deep crisis. The killing of thirty-four striking platinum miners by armed police is now widely acknowledged to have been a symptom as much as a cause of this crisis. As the Financial Mail reported in 2013 with frustration, ‘there is barely any consensus on key issues such as minimum wages, central bargaining and union representation in the workplace’ (Jones, 2013). In the eighteen months following Marikana, South African platinum and gold miners engaged in numerous unprotected wildcat strikes. Many of these workers, dissatisfied with the representation provided them by the National Union of Mineworkers (NUM), or disgusted with NUM’s apparent complicity with management at Marikana, joined a breakaway miners’ union, the Association of Mineworkers and Construction Union (AMCU), which spoke the language of workers’ control and direct shop floor democracy (even if it didn’t always practice it) (Sinwell & Mbatha, 2016; Theron, Godfrey, & Fergis, 2015). The National Union of Metalworkers (NUMSA) and more than a dozen other Congress of South Africa Trade Unions (COSATU) affiliates soon broke away from the main trade union federation, convinced that it had become undemocratic in its structure and beholden to the ruling ANC’s neo-liberal economic program at the expense of workers (Ashman & Pons-Vignon, 2014; Fogel, 2014; McKinley, 2015). In the run-up to the May 2014 national elections, NUMSA explicitly refused to campaign on behalf of the ANC; finally, in April 2017, NUMSA leadership launched an alternative labor federation (24 affiliates, 700,000 members strong) (‘Saftu an inevitable response’, 2017; Whittles, 2017), with an emphasis on independence from political parties and a commitment to democratic practice emanating from the factory floor.

During the lengthy hearings into the Marikana conflict, labor sociologist Sakhela Buhlungu told the Marikana Commission of Inquiry that it is ‘very difficult to service every workplace and to meet with every worker, and this created social distance … .Unions had also become too close to management and the institutionalisation of trade unionism meant that decisions were often taken at a national level, not from the core worker base upwards.’ Observers across the political spectrum have proclaimed that collective bargaining system, union democracy, and basic trade union structures in South Africa are ineffective and often prove fundamentally at odds with the interests of workers. Even one of COSATU’s pre-eminent founders, Jay Naidoo, took to the pages of South Africa’s Sunday Times, urging a renewal of the South African trade union movement’s long tradition of ‘workers’ control’. Like Buhlungu, Naidoo observed that there is ‘a huge disconnect between its leadership and the membership.’ (Mail and Guardian, 2014; Naidoo, 2014; Dlamini, 2017). The Trotskyist left, never reconciled in any case to COSATU’s subordination of its political independence to the tripartite alliance with the SACP and ANC, denounced the federation for its privileged, isolated leadership. (Saville, 2012; “Is Bureaucracy killing,” 2013; Fogel, 2013). Loet Douwes-Dekker, one of the architects of the ‘institutionalization of conflict’ in South African labor relations, told me in an interview that COSATU’s great failing over the years has been to neglect the shop floor, in the misguided assumption that the tripartite political alliance with the ANC and the South African Communist Party (SACP) would allow them to deliver the goods to their membership from above. (Creamer, 2012; Douwes-Dekker, 2014).

The growing gap these commentators observed between the shop-floor and the union office was nothing new. Indeed, its origins can be found in the fundamental transformation of the apartheid labor system wrought by a process initiated in the late 1970s. On May 1st, 1979, South Africa’s Commission of Enquiry on Labour Legislation – better known as the Wiehahn Commission, after its Chairman, Nicolas Wiehahn – released a set of recommendations on the revamping of the apartheid state’s industrial relations. Over the course of eighteen months, from August 1977 through early 1979, the Commission had heard testimony from 184 witnesses and pored over written submissions from 255 further respondents, including trade unions, employers’ associations, labor officials, businessmen, labor activists, and even the secret police in the Bureau of State Security. By a slim majority, Wiehahn and the thirteen commissioners urged the government to grant all South African workers the unequivocal right to join registered trade unions, to dismantle the job reservation that had long barred Africans from certain skilled positions, and to extend to all unions, regardless of racial composition, the opportunity for inclusion in South Africa’s corporatist industrial relations machinery. ‘For the first time,’ liberal industrial relations expert D.W.F. Bendix claimed, the apartheid state ‘has dropped the doctrine of the temporary sojourner and the so-called White economy and has accepted urban Blacks as a permanent, integral part of South African society’ (Bendix, 1979, p. 79).

Despite the enthusiasm of Bendix, other business-oriented liberals, and verligte (‘enlightened’) Afrikaners, there were many objections to the Wiehahn Commission’s report at the time, from both right and left. Those unbendingly committed to apartheid, especially workers in the exclusive white unions of the South African Confederation of Labour (SACOL), understood that the Commission’s recommendations breached a fundamental rampart of apartheid’s ‘separate development’ in labor markets, union organization, and the country’s collective bargaining regime. They reacted with barely suppressed fury. Prior to the hearings, C.P. Grobler, Secretary General of SACOL, had warned that his organization remained ‘firmly opposed to Black trade unions’, which they feared would undercut white workers and become a ‘political football.’ Grobler reached for the excuse that apartheid labor planners had relied on for more than twenty-five years: ‘Blacks, generally speaking, have not yet reached that standard of responsibility which is a prerequisite to trade unionism’, he proclaimed. Upon its release, Arrie Paulus, of the Afrikaner-dominated Mine Workers Union, denounced the Commission’s report as the ‘greatest treachery against the white employes of South Africa since 1922, when hundreds of miners were shot because they opposed the Chamber of Mines who tried to force them to share their work with blacks.’ (Grobler, 1976, p. 171; SALDRU, 1979, p. 25; van Zyl Hermann, 2016). On the left, the ANC, the SACP, the National Union of South African Students, and their allies in the emergent black trade union movement regarded Wiehahn initially as little more than a transparent effort to placate foreign investors. As the African Communist put it, Wiehahn’s report was a ‘State Plan to Shackle African Trade Unions.’ A slightly more sophisticated critique observed that by offering black unions official recognition, but requiring that they register with the government, the Commission sought to increase control over a fractious, unruly, and politicized black working class at the grassroots. Observing the continued political repression faced by labor leaders, the activists centered on the South African Labour Bulletin (SALB) and Durban’s radical Institute of Industrial Education (IIE) concluded that the purpose of the Commission proved to be ‘the extension of control over unregistered unions, in a unitary system which can be sold abroad’ (Nyameko, 1979 1979; SALB, 1979, p. 53).

Despite its signal importance in restructuring South African labor relations, the Wiehahn Commission merits little more than a passing mention in most histories of apartheid and the struggle against it. Histories of the 1970s focus instead on the 1973 Durban strikes that presaged the Wiehahn process, the 1976 Soweto student uprising, the emergence of Black Consciousness and the 1977 murder of Steven Biko, ANC exile politics, and the birth of the mass democratic movement inside the country in the early 1980s. (Brown, 2016; MacMillan, 2013; MacQueen, 2018; Nieftagodien, 2014). The central role of the workers’ movement in toppling apartheid is recognized, but scholars tend to associate this with the birth of COSATU in 1985, rather than with the preceding shop-floor struggles initiated by its predecessor, the Federation of South African Trade Unions (FOSATU). Scholars who discuss the Wiehahn reforms, moreover, tend to replicate views expressed at the time: regarding them as a last-ditch effort to woo the black working-class away from radical politics and preserve apartheid, or at best as a small opening within which workers could organize. Gerald Kraak, in Breaking the Chains, observes that through industrial relations reform ‘the state hoped to incorporate African unions and circumvent the growing militancy of African workers’ (Kraak, 1993, p. 113); David Lewis, in his essay on black workers in From Protest to Challenge, points out that ‘the choice facing the state was either to ban the independent union, or to bring them under the ambit of legislation that was highly restrictive but could generate a greater degree of consensus among employers, the international community, and black South African workers.’ These reforms, Lewis maintains, ‘were intended above all to insulate the economy from political conflict.’ (Lewis, 1997, 211). The comprehensive history of FOSATU, the union federation that actually did the most to take advantage of the Wiehahn reforms, argues that these changes ‘aimed to stifle and divide the expanding black labour movement,’ which since the 1973 strikes ‘was rising up outside of state control.’(Friedman, 2011, p. 28). But intentions and results can differ; as Buhlungu observes, ‘Black unions demonstrated a remarkable ability to use the Wiehahn labour relations reforms … to their advantage’ (Buhlungu, 2010, p. 71; see also Brown, 2016, pp. 104–105). In his comprehensive history of African Workers in South Africa during the 1970s, Steven Friedman observed that the commission’s goal was to ‘grant African union rights under conditions which controlled them’ and deter them from political engagement. Nevertheless, he recognized that such rights offered new space for Black workers to build their unions (Friedman, 1987, p. 153).

Only one other scholar I am aware of, Danelle van zyl Hermann, has looked at the archives of the Wiehahn Commission in depth, but her research emphasizes how industrial relations reforms eroded the privileges previously enjoyed by white workers and their unions (van Zyl Hermann, 2016). Industrial peace, international legitimacy, and the preservation of the free enterprise system were the Commission’s stated goals; for the more liberal members, if achieving these entailed dismantling fundamental aspects of apartheid (such as the enforced job reservation that created skilled labor shortages) this was certainly a price worth paying, even at the cost of losing support from the most intransigent sectors of the white working class. Discussions of Wiehahn that examine black workers’ responses to the reshaping of South Africa’s industrial relations terrain, on the other hand, focus on the ensuing ‘recognition debate’ (Baskin, 1991, 27; Kraak, 118–19; Lewis, 213–14). The reforms initiated by the Wiehahn process confronted the independent trade unions of FOSATU, organized during the 1970s without benefit of state legitimation or the corporatist bargaining frameworks long established for white unions, with a dilemma. Should they take advantage of the official recognition now accorded them, which would enhance their position vis-à-vis employers? Or would this constitute a sell-out to the apartheid state? (Bonner, 1983; FWN, November 1979, pp. 1–2; Friedman, 1987, pp. 156–158). As important as this debate was, it obscures another related tension created by the Wiehahn reforms. As the door cracked open for black workers and their unions, they faced an equally pressing question about where bargaining should take place. Prior to Wiehahn, the new black unions had secured an unofficial foothold by organizing single factories and then forcing the employer to the bargaining table directly, outside of any state-sanctioned bargaining mechanisms. This meant that immediate concerns of the workers on the factory floor could be addressed by the bargaining team, backed up by the direct threat of strike action; and that shop stewards could report back to the membership on a regular basis, and then transmit their concerns forward in a consultative fashion. The result was a high degree of shop-floor participation, a direct expression of workplace grievances, and a keen sense of empowerment when this process wrung concessions from an employer that workers might encounter directly. After Wiehahn, admission to corporatist collective bargaining structures created pressure for the FOSATU unions to bargain at the much further removed industrial council level, where multi-plant unions and professional negotiators squared off against organized employers. While this might effectively regularize collective bargaining procedures, it also threatened to dissipate hard-won shop-floor power and union democracy, shunt aside local workers’ plant-based concerns, and entangle unions in lengthy legal processes that failed to redress workers’ immediate needs. As the Natal branch of the sweet workers union put it when faced with their employer’s demand that they join the industrial council in their industry, ‘we feel that our strength is on the factory floor. Our union is democratic, controlled by our members.’1 Such worker resistance to industry-wide bargaining, a 1982 report by the Investor Responsibility Research Center noted, ‘attracted far less notice than the debate over registration, even though it represents a far greater challenge to industrial relations in South Africa.’ (Hauck, 1982, p. 25).

This article traces the consequences of the growing fissure identified by the IRRC, that between the shop-floor power sought by the increasingly militant independent black trade unions the government sought to tame, and the corporatist industrial council system that had historically excluded African workers and their unrecognized unions. As the IRRC predicted, this tension over ‘levels of bargaining’ had a profound effect on the trajectory of labor organizing in the final years of apartheid, the shape of the trade union movement that emerged from that struggle, and consequently, the labor conflicts raging today in post-apartheid South Africa between those seeking to restore workers’ control to the shop floor, and those hoping to dampen it.

Reforming industrial relations

One of the most pressing tasks identified by the Commission was the integration of the two, distinct systems that had been imposed by the segregated South African industrial relations regime. Such ‘dualism’ was, of course, first and foremost a division by race, since African workers legally did not constitute ‘employees’ with the rights, protections, benefits, and status conferred by the state on that legal category. But dualism went beyond that, cutting to the core of the employer-employee relationship in South African manufacturing establishments. White workers – or, to be more precise, non-African workers, since Indian and Coloured workers did have limited employee rights – benefited from direct access to the industrial council system, which for half a century had facilitated national-level, industry-wide pattern bargaining between registered trade unions and regional, supra-regional, or national-level employers’ associations. The parallel and subordinate system for African workers, first instituted in 1953 by the Native Labour (Settlement of Disputes) Act, permitted only single plant committee-based arrangements, providing no statutorily protected collective bargaining mechanisms (Lichtenstein, 2005). As late as 1977 this legislation continued to deny African workers direct access to the corporatist bargaining arrangements at the industrial council level (although they were often covered by agreements reached by white unions in their industry, to their detriment). As the Wiehahn Commission’s report put it, South African labor relations rested on ‘two systems of inverse composition existing side by side within the same economy – industrial councils representing an industry-level superstructure with little or no statutory infrastructure at the plant or factory level, and the committee system representing an enterprise-level infrastructure with little or no statutory infrastructure at the industry level’(Commission of Inquiry into Labour Legislation, 1982, p. 24). This dualism excluded black workers from participation in making industry-wide agreements, even while it inadvertently encouraged them to shop-floor activism that negated the corporatist bargains struck over their heads by white unions at a more centralized level. For the first twenty-five years of apartheid, this system, while not impervious to organized activity by black workers, managed to confine their industrial actions to a relatively narrow terrain. And then, in 1973, the lid blew off.

The 1973 Durban strikes engaged in by up to 100,000 unorganized African workers unleashed a sharp struggle over the nature of the anemic in-plant committee structures to which African workers had access.2 Surprised by the apparently spontaneous mass strikes, employers and the state sought to enhance ‘communication’ in the workplace by expanding the number and scope of what appeared to be limited and ineffective in-plant consultative committees. Few employers, however, intended to encourage worker-elected structures empowered to negotiate. Instead, they envisioned the far tamer instrument of newly enabled ‘liaison committees,’ which allowed the employer to appoint up to half of its members. Black workers remained torn between dismissing these nascent shop-floor structures as stooge committees or infiltrating them with newly militant shop-floor activists, using them as a Trojan horse for shop steward committees and, eventually, independent black unions affiliated with FOSATU (Lichtenstein, 2015).

Liberal industrial relations reformists, like Bendix and his colleagues at the Institute of Labour Relations, acknowledged that the new legislation passed in the wake of the 1973 strikes was designed ‘not only to prevent the formation of black trade unions, but also to block the development of the all of a sudden dangerous works committees.’ By 1979, it was clear that employers remained resistant to empowering shop-floor organizations in any meaningful way (Bendix, 1979, pp. 86–87). Nevertheless, works committees served as embryonic shop-steward committees, and thus contained the seeds of a worker-control orientation for the emergent independent unions that crystallized after 1979 in FOSATU. Bendix himself, whose views appeared broadly representative of the liberal, market-oriented industrial relations experts who backed the Wiehahn process, noted in 1978 that ‘From an initial rejection of works, but particularly liaison committees, these institutions have become accepted by [black unions] as a relatively effective generator of black workers’ consciousness. There is also no doubt,’ he concluded, ‘that works committees are in the process of assuming plant union character, a fact not appreciated by management in many instances.’ Many union organizers felt the same way (Bendix, Piron, & Swart, 1978, p. 23; Lichtenstein, 2015).

In attempting to weld the existing industrial councils and emergent plant-specific committee structures into a unitary system, Wiehahn precipitated a renewed struggle over the sources of shop-floor organization and power. In particular, employers, personnel experts, white unions, and state labor managers began to recognize the dangers inherent in the shop-floor control embedded in the committee system if its power were to be transferred to the independent black unions. The threat, as ILR pundits saw it was that ‘The eventual possible recognition of black trade union rights in addition’ to the power granted at shop floor level to committees would ‘place black labour in a power position unequalled by labour in any other industrial relations system’ (Bendix et al., 1978, p. 23).

Indeed, much of the initial testimony before the Commission by representatives of the South African business community, across a wide range of employers and multiple economic sectors, expressed anxiety about further empowering black workers on the shop floor. The Transvaal Chamber of Industries, for example, claimed in their submission that ‘Works Committees … are in effect entrenched groups even more powerful than Trade Unions and that they could be very dangerous indeed from the employers’ point of view.’ They hoped the more easily dominated liaison committees could be opened to all races, diluting the power of African workers while extending managerial shop-floor control over non-African workers as well in the name of a more ‘equal’ and unitary system. Similarly, the Durban Chamber of Commerce urged the Commission to recommend a tripartite system, resting at the plant level on local multiracial ‘Works Councils’ modeled after the liaison committees, with 50 per cent employer representation. The director of the important iron and steel industry’s new ‘Black Labour Unit’ testified that ‘once you set up an Industrial Council you’ve got to recognize the level of agreements that that Industrial Council can relate to and can negotiate,’ adding that ‘what would hurt is if legal rights were given to the lower level concept [the committees] to upset the horizontal situation [of ICs] that we’ve got.’3

That said, E.P. Drummond, Director of SEIFSA, the steel and iron industry’s trade association, told the Commission that he envisioned participation of Black unions at both the IC level and the committee system at plant level. ‘I don’t mind a little bit of both because I can play the one off against the other. The system should be very flexible,’ he said. The Motor Industry Manufacturers Association, representing 111 plants with nearly 7000 black employees, testified that the existing legislation in fact ‘functions well mainly because liaison committees are gradually being accepted by Bantu employees.’ True, as workers ‘become more sophisticated’ perhaps they should be admitted to the negotiating machinery of the Industrial Conciliation Act, but ‘to allow Black unions now may very well create confusion and there is the possibility that they could be used by activists to upset the political equilibrium of the State.’ Several employers expressed disquiet with the fact that potential committee members were receiving training from the wrong people, namely the activist workers’ advisory associations such as the IIE and the Urban Training Project (UTP). These organizations, they suspected, had a ‘leftist tilt’ emphasizing ‘conflict of interest between employer/employee.’ These employers hoped instead that the ILR or else the moderate, segregated (as opposed to exclusionary) white-led unions grouped in Trade Union Council of South Africa (TUCSA), which had long cultivated paternalistic relations with parallel black unions, might take the lead in providing trade union training to African committee members. For their part, TUCSA unions like the Garment Workers’ Union also proved anxious about the potential for a ‘revolutionary Black trade union movement’ emerging from penetration of the committee system. As Anna Scheepers told the Commission, under the system set up after 1973 ‘the workers in conjunction with advocates of [racial] polarisation, were manipulated to plan for works committees.’4

Resistance to the empowerment of black organizations should hardly come as a surprise. To be sure, in the late apartheid era, management sought new ‘human relations’ mechanisms and personnel policies to quell endemic shop-floor conflict, stabilize industrial relations, improve productivity, and rationalize workplace negotiation made unwieldy by the dualistic system. Above all, as J.D. Farrell, one of Bendix’s close colleagues at the Institute of Labour Relations pointed out, employers confronted ‘A real need of more communication between management and workers’ (Farrell, 1978, p. 11). But they hardly wanted to encourage black workers to join powerful, independent, militant shop-floor unions, with the ability to bargain directly over daily conditions and grievances in the workplace. As one metalworking firm told the ILR in a 1978 survey, ‘Black trade unions will do more harm than good.’5 Even deliberately cautious and explicitly ‘non-political’ workers’ aid societies, like the UTP, found themselves stymied by employer intransigence. The UTP had reported with frustration in 1975 that ‘There has been no improvement in the attitude of employers … the opposite is more often the case. Management generally appears to have forgotten the … 1973 strikes and is assuming it can work through the [liaison] committee system.’ As Donovan Lowry of the UTP recalls in his memoir, during the 1970s, rather than accept unions, ‘many employers tried to retain the coercive practice of the past, while others attempted to introduce disempowering improved human relations, with the state encouraging them to do this’ (Lowry, 1999, p. 97).

Indeed, as Farrell observed, most employers continued to imagine that in their firm, black workers had decent conditions, were paid enough, and remained satisfied with the liaison committee system, even while dissatisfaction might reign in the larger social order, or even on the shop floor of their competitors. Based on a 1977 survey he conducted of nearly 1600 firms, employing over 200,000 black workers, Farrell found that 2/3 of respondents remained satisfied with the current industrial relations system and 80 per cent of the participants believed that ‘black workers in general are satisfied with their wages.’ Most significantly, Farrell concluded that ‘There is no doubt that employers are strongly against the official recognition of black trade unions’ – even though many acknowledged that such resistance ‘may have a negative influence on labour relations’ (Farrell, 1978, p. 12).

The struggle over industrial dualism

Most of the participants in the Wiehahn process envisioned a reformed industrial relations system that, while eliminating the most egregious forms of racial dualism, such as the restricted definition of ‘employee’ or the imposition of a job colour bar, would continue to throw up a firewall between weak organization at the plant level and the existing corporatist structures of collective bargaining in the industrial councils. Faced by 1979 with the inevitability of official state recognition of African trade unions, reformers hoped that once the new unions acceded to government oversight by registering, they would enter the industrial councils, where they would be outmatched by the long-established white-led unions and well-organized employers’ associations. At the same time, while existing in-plant committees would become legally integrated into a single, non-segregated industrial conciliation act, the revamped system would reproduce the ability of management to use the liaison committee structure to outflank black unions and to thwart shop-floor power at the local level, now under the guise of the ‘works councils’ open to all races. As the editors of the IIE’s South African Labour Bulletin recognized in their stinging critique of the Commission’s report, under this arrangement ‘the policing and interpretation of the Agreements and all management-worker issues arising in the company would be the ambit of the Workers’ Council structure not of trade union shop floor organization … .There can be little doubt that managements faced by independent trade unions will make every effort to seal the factory against trade union organisation by promoting in-company Workers’ Council structures,’ the editors of the journal concluded (SALB, 1978, pp. 7–9). As Eddie Webster of the IIE put it, ‘If this were the case, in a whole range of day-to-day issues, which are the bread and butter of active trade unionists, trade union officials and shop stewards would be completely shut out with the unavoidable consequences of losing touch with the rank and file.’ As he concluded, ‘the implications of this proposal may well lead the registered unions to become mere benefit societies’ (Webster, 1979, p. 554).

Understandably, the new black unions grouped in FOSATU regarded this as an effort to muscle aside the vibrant shop-floor structures they had struggled to build inside South African factories since the 1973 strike wave. These unions, concentrated in metals, textiles, chemicals, and transport, faced a difficult choice: accept the devil’s bargain of registering with the government, and thus gaining access to industrial councils where they might easily be overwhelmed by white unions (each union got one seat at the IC table, regardless of the size of its membership), or remain in what was now disingenuously called ‘the unorganised sector.’ If they chose this latter path, the new unions would find themselves competing with workplace committees empowered to ‘negotiate and enter into statutory agreements,’ an option many employers naturally preferred to in-plant union recognition (Commission of Inquiry into Labour Legislation, 1982, p. 74). As the SALB editors pointed out, the new dispensation would interdict the ‘informal’, non-state sanctioned agreements that the unregistered unions had been making with employers for nearly a decade. Only registered unions, the Wiehahn Commission proposed, would have the right to negotiate and conclude legally binding agreements with employers or sit on industrial councils. Thus, by allowing black unions to join the country’s industrial relations system, the Commission sought to undercut the very independence black workers had gained by virtue of being excluded from that system in the first place (South African Labour Bulletin, 1979, p. 68).

As Metal and Allied Workers Union (MAWU) organizer Bernie Fanaroff put it in a 1982 speech he made to the union’s shop stewards, employers ‘wish to remove collective bargaining … from the `emotional’ atmosphere of the shop floor into the “rational” (read “bureaucratic”) calm of the Industrial Councils … .[T]hey wish to restrict or remove as far as possible the participation of the mass of membership in the collective bargaining process.’ This ran directly counter to the new unions’ organizing tactics and philosophical commitment to workplace democracy (Fanaroff, 1982). Minutes from the internal deliberations of employers suggest that Fanaroff was on to something. In a 1980 meeting of Federated Chamber of Industries’ executive committee, employers acknowledged that ‘the ultimate objective is to work towards a unified industrial relations system in which collective bargaining about remuneration and other conditions of work take place primarily at industry level within the industrial council system, where it tends to be more depersonalized and to acquire a more professional and rational character’ (South African Federated Chambers of Industry, 1980). In much the same way, the expansion of the liaison committees, now rechristened works councils, was in fact designed to ‘remove from unions many of their traditional shop floor functions.’ ‘Once the Works Council is formally recognized,’ the SALB concluded, ‘management will have a powerful argument and institution with which to keep trade unions at bay’ (South African Labour Bulletin, 1979, p. 68). Even non-radicals like Bendix had to concede that these committees remained ‘house trade unions – the trade union by the grace of the boss and for the boss’ (Christian Concern for Southern Africa, 1982, p. 62).

In a subsequent report issued in 1980, the Wiehahn Commission went into much greater depth on the emergent dynamics of a revamped industrial relations regime. Released with more than a year of experience with the revised system, this final part of the report expressed concern that some independent black unions continued to operate outside of the new rules of the game, having refused registration. Some, the commission complained, had gone so far as to ‘embark on direct industrial action wholly in conflict with well-tested procedures for mediation, arbitration, and conciliation’ (Commission of Inquiry into Labour Legislation, 1982, p. 454). Indeed, the commissioners cautioned against ‘a new form of dualism’ (Commission of Inquiry into Labour Legislation, 1982, p. 455) that seemed to be emerging. As the newly established National Manpower Commission (NMC) warned a year later, employers faced a new generation of urbanized, educated black workers with a ‘heightened awareness of [their] rights as workers.’ ‘Many black trade unions are also aware that their bargaining power is stronger at plant level than at some remote institution, where they are overshadowed by the entrenched power of white and coloured trade unions and the collective strength of an employers association,’ the NMC admitted (Department of Manpower, 1981, p. 46). Moreover, organizers found the highly visible, democratic in-plant negotiation preferable to the remote industrial councils for unions trying to secure loyalty from a workforce. Few black workers had any notion of how the IC worked, and indeed had suffered years of low wages based on agreements negotiated by white unions. A shop steward from a sugar mill complained that in his industry white unions had dominated the IC and negotiated low wages for Africans; disputes were settled by IC employees rather than shop stewards. ‘I cannot see how we can have anything to do with the present Industrial Council system,’ he concluded. A metal worker complained that MAWU could not overcome the power of the skilled (white) workers in that industry’ s IC, which was ‘controlled by the craft unions’ (FOSATU Worker News, 1982b, p. 3). By way of contrast, as Fanaroff put it, the plant-based unions built during the 1970s had ‘thrown up leadership which is itself very conscious of and directly answerable to the shop floor membership’ (Fanaroff, 1982, p. 2). Shop-floor textile workers organizer Obed Zuma observed in some of his handwritten notes that ‘house agreements’ were far more democratic than those struck by the IC, since shop stewards met regularly with workers in the plant, prepared demands, and then reported back. ‘All workers are greatly involved in the process of bargaining. Also they bargain directly with their employers,’ Zuma concluded (Zuma, n.d.). The contrast with IC agreements couldn’t be greater.

FOSATU, the IIE, and allies of the independent black trade union movement regarded the new industrial relations system with deep suspicion, a hostility only confirmed by the efforts of employers to bend the new arrangements to their own advantage. SEIFSA, for example, wasted little time before sending out guidelines on ‘the development and participation of black workers in trade unions’ to its 5000 member firms, who together employed half a million workers, 80 per cent of them Africans. Not only was the metal, steel, and engineering sector crucial to setting bargaining patterns in South African manufacturing, it also was the sector with one of the best organized and most militant independent African unions, MAWU, which by 1980 represented 10,000 black workers. Thus, SEIFSA’s response to the Wiehahn recommendations set the tone for how manufacturers would seek to work with the new order. The metal employers’ vision of the new industrial relations regime appeared disappointingly narrow, at best. For starters, SEIFSA did not even bother to consult with the metal trades unions, black or white, representing workers in the industry, before issuing guidelines. Moreover, SEIFSA’s circular made clear that recognition should be granted only to those black unions that chose to register with the government and gained entry into an existing industrial council. The first requirement was by no means a path chosen by all black unions, and the second often came up against the objections of white workers, who retained the veto power to block admission to industrial councils. As Eddie Webster subsequently remarked, it was no surprise that ‘the attempt by Wiehahn to incorporate black workers into the Industrial Council by excluding them from the shop floor was seen by MAWU as a direct threat to the principles of democratic trade unionism.’ SEIFSA advised ‘its members that they should have no dealings whatsoever with unions other than through the employers’ association at Industrial Council level … .Employers were warned not to negotiate with shop stewards at plant level and not to grant unregistered unions stop orders’ (Webster, 1983, pp. 15–17).

Tellingly, the new SEIFSA guidelines denied black trade union organizers access to factories, noticeboards, or meeting facilities, limiting their approach to workers, hamstringing their organizing efforts, and making them vulnerable to police outside plant gates. While the guidelines promoted the formation of workplace committees, there was no effort to distinguish the old, employer-dominated committee structure from genuine trade unions. As Douwes-Dekker concluded, the SEIFSA guidelines implied ‘that trade unions should not be given or acknowledged a role in the place of work’ (Douwes-Dekker, 1980, p. 20), an approach sure to reinforce black workers’ hostility to the committee system, and obviously designed to thwart the metalworkers’ growing shop floor organizational strength.

African workers proved well-aware of their employers’ continued preference for the committee system. ‘If they had a positive attitude to the union, the first thing they’d have done would have been to allow them to come into the company and talk to the people. And that isn’t allowed,’ one black metalworker interviewed in 1980 observed. Another chimed in, ‘That’s the reason they’ve set up the employees [works] council, to force the union out of the way … .The fact that union workers have to stand outside the doors in order to talk to people, is sufficient to make the workers wary and wonder – why don’t they come inside? There must be something wrong with them. … . The workers are quite frankly afraid of being spoken to at the door. They want people to come into the company to talk. And that’s just what management won’t allow,’ he concluded (Christian Concern, 1982, p. 48).

In a similar fashion, the largest employer in South Africa’s textile industry, the Frame Group, sought to keep FOSATU’s National Union of Textile Workers (NUTW) out of its plants by striking a deal with its rival, the far tamer TUCSA union, the Textile Workers Industrial Union (TWIU) and pressuring African workers to join the latter. ‘Now when it is totally unavoidable to recognize a union,’ the Fosatu Worker News pointed out, ‘it is hardly a surprise to find [Frame] trying to force workers into a TUCSA union that relies on management for support’ (FWN, 1983).

At the national level, as Douwes-Dekker recognized, the push to centralized council bargaining had the unfortunate by-product of lulling employers into the conclusion that ‘the trade union as organised expression of collectivity has no role or responsibility for the interaction between management and the union members in the place of work,’ thus deferring recognition. For those employers who sought to minimize the in-plant power of the new unions, Douwes-Dekker feared that ‘the committee system is seen as a viable alternative to exclude the trade union from the workplace,’ one that rested on a paternalistic ‘human relations policy’ that ignored the basic conflict of interest between workers and employers. He bemoaned the fact that the Wiehahn recommendations ‘reinforced rather than questioned the beliefs expressed by the Institute of Labour Relations, which employers now endorse, regarding the exclusion of the trade union in the place of work,’ since the management-dominated committee structure favored by employers lacked legitimacy among black workers. As Douwes-Dekker bluntly pointed out, the Wiehahn Report ‘is silent on legitimate union activities in the plant and the linkage between unions and committees.’ The persistent apartheid goal of using the committee structure, embedded in the 1953 Act, to blunt shop-floor organization remained in place (Douwes-Dekker, 1980, pp. 20–23).

Douwes-Dekker complained that employers and labor officials imagined that trade union negotiations at the centralized IC level would suffice to industrially enfranchise the black working class. But without a constant presence in the workplace, how would new members be recruited without company interference? How could unions function democratically? How could local grievances be heard? In essence, the vision of trade-unionism that had developed at the grassroots among black workers in the wake of the 1973 strikes differed substantially from the centralized, corporatist system into which the Wiehahn process promised to induct them. As he observed, ‘the black trade union movement in its organising campaigns, also found that workers perceived the disciplinary nature and strength of the trade union expression in the place of work’ (Douwes-Dekker, 1980, p.27, my emphasis). As the UTP had noted in its submission to the Commission, ‘The industrial council system has entrenched injustices to Black workers.’6

Douwes-Dekker was no radical, and his view remained distinct from activists in FOSATU and the IIE who continued to see shop-floor power as a crucial component of revolutionary trade unionism and even a vehicle for eventual socialist political transformation. His perspective might be characterized as a de-politicized syndicalism, one that regarded shop-floor power as a path to economistic trade unions that would remain sheltered from political action and co-optation by the radicals, even while serving as an effective training ground for a democratic proceduralism that might eventually spread to South African society as a whole (Douwes-Dekker, 2014). Similarly, but with a good deal less sympathy for black workers’ aspirations, many in the newly empowered liberal industrial relations establishment saw de-radicalized African trade unions as the best guarantee of a peaceful transition to a more liberal and capitalist South Africa, one that would abandon apartheid in the name of a stable economy and free markets. As a management personnel textbook published at the time observed, ‘if one accepts that the white man’s present powers and privileges must and will inevitably diminish …, then it could be argued that the danger lies not in recognizing African trade unions but in failing to do so’ (Orpen, 1976, p. 45). Such a view rejected out of hand the verkrampte apartheid hard-liners’ insistence that African workers were still not ‘ripe’ for unionism, and thus could too easily fall prey to manipulation by political agitators. At the same time, the industrial relations establishment hoped to divert the emerging independent black unions away from participation in the radical, socialist-based politics of social transformation pursued by many activists advocating workers control at the plant level.

Defending the shop floor

In the 1981 labor legislation that emerged from the reform process initiated four years earlier, the apartheid state finally granted African workers (including, grudgingly, migrant workers) the unequivocal right to join registered unions and full recognition as ‘employees.’ While the government touted overseas its new labor code as a radical restructuring, at home the consequences proved far less revolutionary than they claimed. As Brand and Brassey (1981) pointed out in the pages of the Industrial Law Journal at the time, strong doubts persisted about the fairness of the industrial council system, which had always worked against black interests because the councils were dominated by white unions. Even under the new act, they concluded, ‘Industrial councils are likely … to remain unrepresentative and employees who are not actually represented are likely to be suspicious of their conciliatory and negotiating role.’ As a result, FOSATU unions did not abandon the struggle for shop-floor negotiating rights as preferable to absorption into the industrial councils, or at least as complementary to the corporatist process. The Chemical Workers Industrial Union (CWIU), for instance, said that while ‘not opposed to representative bargaining on an industry wide level,’ it refused to ‘give up its right to negotiate wages and working conditions at the factories where it is representative’ (FWN, 1981, p. 4). For its part, MAWU feared that if it joined the powerful metal industry council it would cede control over national bargaining to SEIFSA. ‘SEIFSA remains the most powerful employers’ body in South Africa outside the Chamber of Mines, while the Steel and Engineering Council represents the Industrial Council problem in its most intractable form,’ the 1982 FOSATU annual report acknowledged. ‘At this stage of its development,’ noted a position paper, ‘MAWU only has strength at an individual factory level … .It is thus only at this level that MAWU can negotiate from a position of strength and representavity and simultaneously ensure worker control and democracy’ (FOSATU, 1982b).7

Employers ‘want workers to negotiate at a place where they are weak, and not in individual factories, where they are strong,’ FOSATU Worker News cautioned. Indeed, FOSATU’s 1982 Congress debated whether its unions would refuse to enter IC’s ‘on terms which are to their disadvantage,’ and would seek instead to ‘negotiate a system of plant-based industry-wide bargaining,’ despite what they had begun to recognize as ‘the growing need for large-scale bargaining structure.’ ‘Only when workers can match the employers’ strength in powerful, national unions,’ FWN concluded, ‘will industry-wide bargaining make sense’ (FOSATU Worker News, 1981, p. 4; FOSATU Worker News, 1982a, p. 3). A position paper prepared for the Congress complained that ‘bosses all over South Africa are trying to force Black workers into Industrial Councils … .Workers are asking, why is this happening, and what should be the workers’ reply?’ Industrial councils, FOSATU organizers noted, traditionally had been used to gain benefits for white, Indian and Coloured workers ‘at the expense of African workers’; to prevent workers from striking with mandated cooling-off periods; to replace shop steward power with ‘neutral’ arbitration; to ‘set up professional bureaucracies separated from and not controlled by the ordinary union members’; and, in general, to weaken workers’ organization on ‘the factory floor’ (FOSATU, 1982a). As former FOSATU organizer Mike Morris suggested in a 1990 article, ‘Black unskilled workers experienced the industrial council system in an extremely concrete way every week when their pay packets reflected the industrial council deductions,’ even while the agreements struck there by the white craft unions to which they did not belong disadvantaged them. These aspects of the IC system ran directly counter to FOSATU’s ‘principles for democratic collective bargaining,’ including plant-based negotiations ‘subject to mass mandate,’ a strong shop steward system designed to police agreements on a daily basis, and direct control of subscription funds. As the position paper noted, ‘bosses at present want to use industrial councils to stop plant-based bargaining’ (FOSATU, 1982a; Morris, 1990, p. 151). Indeed, an independent survey done in 1981 showed that employers (and white unions) much preferred the IC system as a collective bargaining mechanism, while black unions ‘showed a preference for collective bargaining at the plant level’ (Nattrass & Ardington, 1981, p. 14).

Much of the ensuing legal and bureaucratic struggle over the levels of negotiation at which the emergent black unions would be permitted to exert their newfound power and legitimacy took place within a pair of new institutions created by the government in response to Wiehahn, the National Manpower Commission (NMC) and the Industrial Court. While bemoaning the conservative makeup of the NMC (‘the [Labour] Minister it seems is not to have a single advisor with any known sympathy to the non-racial trade union movement’), the leadership of FOSATU reassured its members that ‘labour relations aren’t determined by manpower Commissioners whispering in the Minister’s ear. They grow out of factory struggles and FOSATU is better represented on the factory floor’ (FWN, Aug. 1980, p. 5).

Indeed, it wasn’t long before the NMC, established to oversee all aspects of South Africa’s new industrial relations dispensation, began to express frustration at the black unions’ persistent determination to pursue shop-floor power rather than be absorbed quietly into the new corporatist blueprint. In its first annual report, the NMC complained that ‘Some employee organisations are also as yet too often inclined to use their bargaining power outside the formal system and also to depend more on negotiation at shop floor level.’ (NMC, 1982, p. xi). A year later, things had only gotten worse. At the end of 1982, the NMC reported that ‘the past year … more than previous years, was characterised by a greater measure of activity outside the statutory system in labour relations, so much so that such activities are at present considered to be a fundamental part of the country’s labour relations practice’ (NMC, 1983, p. 108). As attorney John Brand observed, ‘the bargaining unit is … the battleground between employer and trade union.’ The former hoped for larger units, which are ‘more difficult for a union to organize and keep organized.’ On the other hand, he noted, ‘the smaller unit … gives the union a better chance of … getting recognized’ and ‘allows the individual [worker] more say in the decision-making process’ (Brand, 1985). Thus, not surprisingly, the NMC complained that some black unions ‘indicated that the industrial council/conciliation board system is not considered suitable to deal with the problems of the workers at the lowest levels (undertaking/workshop levels) and that they are therefore not prepared to become part of the system’ (South Africa. National Manpower Commission, 1983, p. 112). FOSATU’s annual report in 1982 confirmed that ‘Despite a hardening of attitudes by management, the unions have continued to gain recognition in the plants.’ They counted 173 factories with functioning steward committees recognized by management, ‘with rights to grievance handling, regular meetings with management, time off for shop stewards meetings, negotiating rights and union rights such as stop orders and access’ (FOSATU, 1982b, p. 4). There had only been ninety-nine the year before. ‘I think FOSATU should continue to push for more house agreements because we believe that in this way workers are having a direct channel to their problems,’ noted organizer Obed Zuma (n.d.).

To illustrate this worrisome development, the NMC pointed to the existence of more than 200 local ‘recognition agreements’ outside of the new statutory mandate, negotiated between the black unions and employers without state oversight or the imprimatur of an industrial council agreement. By 1983, the NMC concluded that these in ‘many cases, were actually the only de facto means of formalising relations between employers and unions.’ Much to the Commission’s dismay, ‘bargaining at the local level is developing into an established part of the country’s labour relations system.’ This development flew in the face of the strenuous efforts of employers and the state to prevent exactly this from happening. Indeed, the NMC complained that the bargaining conducted at the industrial council level (where white unions continued to predominate) was supposed to govern local agreements, but ‘certain people would prefer to reverse the order and claim that the former in fact supplement the latter!’ (South Africa. National Manpower Commission, 1984a, p. xx). Even while MAWU reluctantly joined the metal council in 1983, the union’s leadership issued a statement to its members declaring that ‘MAWU remains committed to shop floor organisation as being the most important – industry-wide organisation is secondary’ (MAWU, 1983). Meanwhile, fearing a complete upending of the labor system they were at such pains to construct, the NMC insisted that ‘It is of great importance for the maintenance and promotion of industrial peace’ that industrial relations be ‘regulated by a unitary system and for the Government to do everything possible to make the formal or statutory machinery … so attractive that most of the parties concerned, including the newer [black] trade unions, will join it’ (South Africa. National Manpower Commission, 1984b, p. 187).

So concerned was the NMC about these developments that during the mid-1980s they issued a pair of special reports on ‘levels of collective bargaining.’ Together, these reports reveal the strategy pursued by the new black unions, as they sought to build their power factory by factory rather than submit to a centralized industrial relations regime designed to discipline and subsume them. In its initial special report, released in 1984, the NMC continued to raise concerns about the recalcitrance of black workers and their unions to avail themselves of the new bargaining mechanisms made available to them. The report acknowledged that between 1924 and 1979 most black workers, excluded from what the NMC regarded as an otherwise sound system of labor relations, could not ‘effectively give expression to their needs – at any rate not through the statutory institutions provided for them.’ With the post-Wiehahn changes in labor legislation, however, ‘it was expected that they would possibly enter the system in order to be able to express their needs more clearly.’ With some head-scratching, the special report observed that black workers and their new unions ‘apparently have objections to various aspects of the statutory system, and as a result some of them do not as yet want to register or still shy away from participation in industrial councils’ (NMC, 1984b, p. vi).

Of course, black workers had every reason to be wary of the industrial councils. For many, this was alien territory, long a bargaining table to which they had been barred access and where, at best, white employers and white workers had struck deals that included blacks without any representation or consultation. Not only did the new unions fear the suffocating embrace of the state, but they correctly recognized that incorporation into the IC system would leave them as junior partners to established white unions, which hardly had their interests at heart. Moreover, as the FOSATU (1982a) position paper described above argued, ‘in most industrial councils non-racial [i.e. FOSATU] trade unions represent less than 1/10 of the workers, and are therefore weak in the industry as a whole.’ Finally, as even the NMC admitted, many black workers were primarily ‘concerned with job security which could be better negotiated at factory or plant level.’ As Fanaroff pointed out, for black workers job security and limits on managerial prerogative were the key issues, because to be retrenched also meant to be sent back to the African reserve. ‘A worker losing his job also loses his accommodation and his right to be in the urban area’ (Fanaroff, 1982, p. 8). This was the sort of security best negotiated at plant, not IC, level. Above all, the new trade unions saw the long-established IC system as too cumbersome, ‘undemocratic and bureaucratic,’ nothing more than ‘an ideal system for reactive, lazy, good-for-nothing [white] trade union leadership’ This was both a racial critique and a tactical evaluation of what unions might be able to do at the shop-floor level. Black unions were organized from the ground up, and had begun ‘exercising their power base at this level’ (NMC, 1985, pp. 118–119). As one black trade union leader put it, ‘The Black trade union movement does not have power, in the majority of cases, at Industrial Council level. It is evident, therefore, that it will continue to bargain at plant level,’ even if some unions agreed out of apparent necessity to join the Council in their industry as well (FOSATU, 1982b, pp. 25–27).

By 1986, the NMC estimated that between 400,000 and 500,000 black workers were covered by over 500 factory-based agreements struck between unions and employers outside of the recently established statutory bargaining process. In some instances, these agreements supplemented collective bargaining arrangements governing wages and basic conditions developed at the IC level; in others, they had displaced them altogether. Most of these local pacts covered black workers in unskilled and semi-skilled categories, and included grievance procedures, election and recognition of stewards, access to the factory, dues check-off, the right to in-plant meeting and notice-boards, and a no-strike clause. They often governed seniority, health and safety, disciplinary and lay-off procedures as well–in other words, all those procedural matters falling beyond the scope of wages and basic conditions of employment, and touching on matters of shop-floor control so essential to the black working class under apartheid (NMC, 1986, pp. 13–14), which had no other source of representation or power. And at times, they offered a floor on in-plant wages that exceeded that mandated at a national level by IC agreements.

The NMC conducted a close examination of fifty-five of these agreements, finding that well over half of them applied to workers already covered by industrial council agreements and/or Wage Board recommendations. They concluded that black workers and their unions pushed for this additional local level bargaining because they ‘feel more at home at this level …, possibly because they feel that they have a strong following at this level.’ There was ‘a feeling among Black workers that they are not treated by their employers with respect for their human dignity,’ and local bargaining was understood as the best way to rectify this. By contrast, blacks felt that ‘the industrial council system was designed and is managed and run by Whites for Whites and is of no advantage to the Black worker.’ Despite its occasional sympathy for organized black workers, the NMC showed its cards when it dismissed this sentiment because ‘this objection has a political basis and cannot be regarded as a weakness in the system as such’ (South Africa. National Manpower Commission, 1984b, pp. 109, 117).

Towards a new corporatism

In addition to re-introducing the dualism complained of by the NMC, another unintended consequence of Wiehahn was increased industrial unrest, the very thing the reforms had been designed to rein in. On the one hand, black trade union membership increased rapidly after the new legislation of 1981, so that it had become ‘larger than the total figure for the three other population groups.’ At the same time, since many black unions resisted incorporation into the industrial council system, and continued ‘to deal directly with employers at plant level,’ shop floor conflict became endemic and led to increased police repression of workers (Vose, 1985, pp. 452–53).

Black workers at a Colgate-Palmolive plant near Johannesburg, for instance, spent a year trying to wrest a recognition agreement from the company. Despite Colgate’s alleged commitment to reform – the multinational was a signatory to the Sullivan principles, binding it to non-racial conduct in its South African operations – management refused to meet with the CWIU for over a year. Finally, in February 1981, the company agreed to discuss unionization of the plant’s 300 workers, but only on the condition that final wages and conditions be negotiated at the industrial council level, not in a plant-based agreement. ‘The Union’s role,’ organizers complained, ‘was therefore to be involved in shop floor grievances only.’ The CWIU pointed out that few of the workers at the plant were even covered by the existing industrial council, and in any case that the minimum wages established at the council level fell far below those already paid at the plant. Without plant recognition, the union threatened a strike and asked FOSATU to call a national boycott of Colgate products.8 In the face of the boycott, and on the eve of the strike, after a few months Colgate agreed to bargain wages directly with CWIU at the local level. (Friedman, 1987, pp. 248–50) In the wake of this dispute, FOSATU announced that ‘the question of where negotiations take place is clearly the next battle … .we are likely to see more struggles of this nature in the future’ (Forrest, 1981). A similar dispute occurred at a bakery plant in Natal, where the sweet workers union had organized 450 of the 500 black workers. Here too, there was strong and long-standing resistance to removing bargaining from the shop floor. Faced with management’s insistence that they join the Biscuit Industrial Council, the union resisted, claiming that ‘the Unions on the Council are weak and the management in the industry are able to dictate very much the pace of change.’ Organizers pointed out that in contrast to their local situation, ‘no workers are involved in the wage negotiations at the Industrial council [level].’ As a consequence, ‘any agreements reached on the Council that is not accepted by our members, we will refuse to be party to.’9

As the NMC itself documented, such efforts to defend shop-floor gains led to increased conflict. The number of strikes rose rapidly during these years, from 200 in 1980, to over 400 in 1984, to nearly 800 by 1986. As the Commission remarked in its 1986 report, ‘Black workers are more and more aware of their powerbase in the economy through the withholding of their labour’ (NMC, 1987, p. iv). At the same time, under the velvet glove of new opportunities lay the mailed fist of the security services. Bannings, arrests, and extra-judicial killings took their toll on the leadership of the new unions during these same years, most notoriously in the 1982 death in detention of FOSATU organizer Neil Aggett (Naidoo, 2012). In the two years between April 1981 and April 1983 at least 400 trade unionists and workers were detained, including thirty union organizers and officials (Boyer & Davis, 1984). As a 1985 ILO report on South Africa put it, while ‘racial measures may be less overt in certain areas of the labour field, control over the Black labour force and its trade unions is now applied through security legislation, influx control, and the `homelands’ system.’ Labor actions unilaterally deemed by the state to be ‘political’ continued to lie outside the new industrial relations regime (ILO, 1985, pp. 4, 13).

If employers resisted the formation of truly independent unions inside their workplaces, sought to dilute the power of the new unions in the industrial councils, and called on the power of the state to break militant unions, the new Industrial Court offered another avenue for black workers to redress their grievances. Workers brought 399 cases to the Court in 1984, 801 in 1985, and 2042 in 1986, ranging from objections to unfair dismissals, to demands for union access to factory premises, to the enforcement of local recognition agreements (South Africa. National Manpower Commission, 1987, p. 46). Naturally, one key area of contention was ‘levels of bargaining.’ Industrial council policy had been a major policy issue during FOSATU’s second annual conference in 1982. Larger and faster-growing affiliates like the metalworkers’ union discovered that they increasingly ‘faced … the need to engage in wider forms of collective bargaining than plant-based bargaining in order to exert their collective strength.’ MAWU, for example, had organized 135 factories in the industrial areas around Johannesburg, and found it nearly impossible to negotiate separate agreements at them all (FOSATU, 1982b, pp. 14, 24). ‘Whether MAWU likes it or not the minimum conditions of employment are negotiated annually in the Industrial Council,’ the union noted in a press release announcing its new approach to corporatist bargaining (MAWU, 1983). As Eddie Webster observed in 1983 in the pages of the SALB, ‘A vital tension has emerged inside MAWU between the growing need for stable organisation and the desire for mass participation in the organisation’ (Webster, 1983, p. 18; Forrest, 2011,pp. 123–125)

Yet even while cautiously admitting the necessity of national-level collective bargaining, by the mid-1980s unions like MAWU still looked to the Industrial Court to preserve plant-level bargaining and democratic membership participation in crafting workplace agreements. After all, as Mike Morris observed, ‘the industrial council structures ran counter to the basic organizational principles underlying the independent trade union movement’ (Morris, 1990, pp. 151, 154). FOSATU organizers envisioned ‘Industry-wide collective bargaining in which plant-based bargaining is guaranteed,’ while ‘Bosses … want to use industrial councils to stop plant-based bargaining and the wage “leap-frogging” which come from these,’ a FOSATU discussion brief stated (FOSATU, n.d., p. 4). MAWU negotiator Bernie Fanaroff envisioned a centralized bargaining process that involved ‘consultation with [union] membership at all stages of negotiation,’ rather than just the kind of centralized negotiation by union officials at IC level which is then announced to members with no sense of the bargaining process (Fanaroff, 1982, p. 10). For example, Fanaroff recalled (2016), MAWU demanded that employers allow shop stewards to ‘get a direct, real-time feedback on the [industrial council] negotiations … We were very insistent that the locus of power was with the shop stewards … .All that’s been lost, I’m afraid,’ he ruefully concluded.

An important Industrial Court ruling in 1985 made this balancing act more difficult to maintain, when the Court ruled against MAWU’s contention that shop floor agreements should supersede industrial council negotiations. In South Africa’s metal industry, the NMC observed, ‘[SEIFSA] adopted a policy of encouraging the new trade unions to participate in the industrial council on the one hand and strongly opposing shop-floor bargaining on issues included in the industrial council agreement on the other.’ Meanwhile, however, unions in this sector ‘pressed for shop-floor bargaining over and above the industrial council agreements’ (South Africa. National Manpower Commission, 1987, p. 8). At Hart, a Durban-area manufacturer of plastic and aluminum containers, MAWU had a decade-long history of militancy. Workers struck in August 1984 in a dispute over wages, long service allowance, and funeral benefits. ‘We wanted to negotiate all these grievances outside of the Industrial Council,’ the workers insisted, ‘but management insisted that they would only negotiate … at the Industrial Council.’10 Seeking to give legal weight to the many shop-floor agreements metalworkers had forced on recalcitrant employers, the union argued unsuccessfully before the Industrial Court that ‘bargaining at industry level was directed towards establishing minimum conditions of employment while bargaining at plant level was concerned with actual conditions determined by the circumstances of the particular employer.’ The Court endorsed SEIFSA’s contention that ‘agreements on substantive matters such as wages, overtime rates, and hours of work should be negotiated at industrial council level. Attempts by trade unions to bargain on these matters at company level should be resisted by employers’ (MAWU v. Hart, 1985, pp. 479, 481).

This effectively stymied MAWU’s efforts to join the IC ‘without abandoning the fundamental commitment to rank-and-file democracy’ (Morris, 1990, p. 156) that the unions had emphasized since the 1973 strikes. Black workers sought to preserve their hard-won shop floor power. But organizational imperatives as the unions expanded, employer intransigence, and the new legal regime all inevitably pressed the new unions towards a more corporatist orientation. In one sense, this represented a major triumph for workers who only a decade before had not even been considered ‘employees’ under apartheid labor law. By the end of the 1980s, the SALB reported, ‘where militant unions have entered industrial councils … they have been able to tilt the balance of power in their favour’ (Toerien, 1989, p. 81). In response, some employers, especially those in more competitive labor-intensive sectors even began to ‘revert to plant-level bargaining in order to destroy the national organization of workers in the industry’ (Morris, 1990, p. 157). By the end of the 1980s, unions appeared to embrace centralized bargaining, and employers, now playing a weaker hand, resisted it. As a consequence, when discussions of a post-apartheid Labour Relations Act began in the early 1990s, COSATU (by then the leading labor federation) wanted to see the industrial council system that had once excluded black workers shored up, not dismantled (Wood, 1998). As the trade union movement foresaw wielding real power in a post-apartheid dispensation even the most militant backers of trade union democracy and the socialist project acknowledged that some kind of corporatist road should be pursued (Bird & Schreiner, 1992). This meant restructuring the industrial council system to incorporate the black working class.

But at what cost to the factory-based power associated with the organizations built by FOSATU over the previous decade? To what degree did ‘the involvement of independent unions in state promoted forms of collective bargaining carry the risk of distancing the union leadership from the shop floor?’ (Matiko, 1987). In the long run, both structural forces and tactical political choices at the moment of transition from apartheid to democracy did much to strip the union movement of its shop-floor orientation, and subordinate it to a national liberation project spearheaded by the ANC and its communist party allies. By 1994, as the ANC introduced a new political and economic order, COSATU took a firm stance in favour of ‘centralized bargaining forums in all sectors.’ Indeed, they hoped to ‘compel centralized bargaining,’ and their position was ‘ambiguous … on the crucial issue of bargaining at more than one level’ (Godfrey, Maree, Du Toit, & Theron, 2010, p. 82). under the new dispensation, the ‘institutionalisation of conflict means fewer struggles and strikes in the workplace,’ Karl van Holdt observed, warning against the ‘dangers of corporatism,’ as labor-management-state partnerships conducted centralised negotiations. He worried the result would be ‘demobilisation of the mass base of the unions, and an alienation of members from the leadership.’ In such a pact, ‘labour would have gained influence, but at the expense of power’ (van Holdt, 1993, p. 48). Indeed, a 1995 study of ninety-six unionized firms done by the ILO and the Sociology of Work Program in Johannesburg discovered how rapidly shop floor power had dissipated during the transition. Only four companies had full time shop stewards in the plant, and 29 per cent of shop stewards had not even met with a union official in the past year. Nearly a third of the stewards had received no training whatsoever, and of those who had, many had merely attended a weekend workshop (Standing, Sender, & Weeks, 1996, p. 165). In short, the workplace networks that once energized the black trade union movement had begun to vanish. A 2007 study concluded that after a decade and a half of a post-apartheid Labour Relations Act, ‘Firm and plant-level bargaining is in decline, and seems to have largely disappeared within the jurisdictions of bargaining councils’ (Godfrey, Theron, & Visser, 2007, p. 99). This paved the road to Marikana.

Conclusion

Despite the profound limitations of the Wiehahn reforms and the ongoing effort to divert union power away from the shop floor, the changes initiated in the late 1970s helped set the stage for the important role an unshackled black trade union movement would play in the final push towards liberation in South Africa. By simultaneously legitimating black trade union organization while trying to limit its radical shop-floor contestation, the reforms deeply imprinted the character of the South African labor movement. The new dualism inadvertently introduced insured that when black unions emerged into the open during the 1980s, many retained their deep suspicion of the state, their intense commitment to democratic proceduralism, and their desire to defend workers’ power in the factories rooted in a dense network of militant shop stewards’ organizations. Yet, at the same time, the rapid growth of the new unions in key sectors during the 1980s soon made entry into the industrial councils imperative, as mass-based unions sought stability, legitimacy, and national pattern bargaining. Striking separate bargains at dozens or even hundreds of plants was hardly a practicable alternative Together, the subordination of trade union struggles to the larger nationalist agenda of the ANC, the necessity of defending union gains against an employer counterattack, and an understandable desire to wield influence at the level of the national political economy, made corporatism a foregone conclusion by the time the tripartite alliance of COSATU, the ANC, and the SACP dislodged the apartheid state in 1994.

Nearly two decades later, the massacre of workers engaged in a wildcat strike, and at odds with a union leadership closely tied to management and the state, suggests the cost of this bargain proved high in the long run, as the once vibrant shop floor character of the union movement atrophied in post-apartheid South Africa. NUMSA’s current search for an alternative labor politics builds on a long tradition of workers’ control, shop floor democracy, and struggle unionism that independent unions like its predecessor – MAWU – forged during the 1970s. Denigrated then by the SACP and its allies in the ANC as ‘workerism’ (meaning trade union economism), nationalist liberation politics eroded this tradition but never liquidated it entirely. Then, as now, the ANC and the SACP demanded that the labor movement subordinate its sectoral interests to the larger needs of the Struggle, the Transition, or the National Democratic Revolution. Then, as now, workers were expected to modulate shop floor militancy in the interest of larger strategic political goals, such as centralized collective bargaining. Then, as now, shop floor democracy, the power of shop stewards closely knit with comrades in the workplace, and the tradition of report-back and workers’ control, were expected to take a back seat to national-level negotiations, a growing class of union office-holders, and a labor federation that grew closer to management than to workers, replacing democracy with labor bureaucracy. But because it had been so central to the birth of the new unions in the 1970s, ‘workerism’ remains a powerful, if buried, tendency within the South African labor movement; the recent conflicts on the Platinum Belt represented its rushing to the surface, like a dormant volcano coming to life. The effort on the part of the ANC and the SACP to suppress and overcome this tendency has been a long, drawn out struggle, and one hardly unique to South Africa. Wherever working-class movements have joined with a national bourgeoisie in a revolutionary process, they have found that for them the struggle continues after liberation, and their former allies become their antagonists, if not their masters. The question is usually this: how long will it take the working class to see the writing on the wall? Marikana, it seems, was the revelation. Will the new union federation recently launched under NUMSA’s umbrella successfully take advantage of this moment to reassert an alternative tradition?

Acknowledgments
Versions of this article were presented at the following seminars and conferences: African Studies Association, Association of Indian Labour Historians, Emory University’s workshop “Twenty Years Later: South Africa and the Post-Apartheid Condition”, Southern African Reading Group at New York Law School, and the University of Johannesburg Centre for the Study of Social Change. I thank all the participants in these venues for their valuable feedback.

Disclosure statement
No potential conflict of interest was reported by the author.

Notes
1. Maggie Magubane to FOSATU General Secretary, 19 May 1983, SACTWU Papers, D7.2.1.5, Historical Papers, Cullen Library, University of the Witwatersrand, Johannesburg, 4.

2. The single best account of the Durban strikes remains that produced by the IIE in their immediate aftermath (Institute for Industrial Education, 1974).

3. Submissions and Testimony, box 32, 24 November 1977, p. 15; box 31, 12 October 1977; box 9, 16 February 1978, 2825–27, 2860, Commission of Enquiry on Labour Relations, 1977–79 records, South African National Archives, Pretoria.

4. Minutes, Industrial Relations Machinery Subcommittee, 22 September 1977, box 24, p. 31; Memorandum of Evidence, 30 January 1978, box 32; Submission of Natal Chamber of Industries, 25 October 1977, box 31, pp. 4–6; Submission of the Garment Workers Union, September 1977, box 33, Commission of Inquiry, 1977–79.

5. Submissions and Testimony, box 37, Institute of Labor Relations survey, 28 March 1978, p. 16, Commission of Enquiry on Labour Relations.

6. Submissions and Testimony, box 9, testimony of Douwes-Dekker and UTP, p. 2753, Commission of Inquiry, 1977–79.

7. ‘The Metal Industry Industrial Council – A Forum for bargaining?’,22 June 1982, Phil Bonner papers. See also ‘Memo RE SFAWU vis a vis the National Industrial Council for the Biscuit Manufacturing Industry,’ 19 May 1983, SACTWU Papers, D7.2.1.5., Historical Papers, Cullen Library, University of the Witwatersrand, Johannesburg.

8. CWIU memo to FOSATU affiliates, 31 March 1981, FOSATU papers, C1.13.1.13.7, Historical Papers.

9. ‘Memo RE SFAWU vis a vis the National Industrial Council for the Biscuit Manufacturing Industry,’ 19 May 1983, SACTWU Papers, D7.2.1.5.

10. Hand-written notes, Hart dispute, FOSATU papers, C1.13.3.8.20, Historical Papers.