The failure is noteworthy when one considers that the new framework was characterised by progressive intentions and informed by a participatory, Northern European model (Donnelly and Dunn 2006). The new government, supported by a strong labour movement, carried out what was presented as a radical restructuring of the workplace aimed at ensuring racial equity, improving working conditions and democratising firm level decision-making. It built on the reforms which had followed industrial (and political) action by the black trade union movement since the early 1970s. It could be said with hindsight that the Labour Relations Act (LRA) of 1995 walked in the footsteps of the 1981 LRA in the sense that it established institutions in which unions could participate, thus asserting the legitimacy of majority black unions while continuing ‘the ongoing effort to divert union power away from the shop floor’ (Lichtenstein 2013). The crucial decision which was made in the new labour law regime was to leave most of the regulation of labour and pay (including minimum wage) conditions to sectorial bargaining councils, and to leave the establishment of the latter to unions and employers. In sectors where unions were too weak to bargain, sectoral determinations would be adopted by the minister of labour, on the recommendation of a five-member employment conditions commission (ECC), in order to regulate conditions of pay and work.6
This architecture goes a long way towards explaining why so many workers are poor and unprotected in South Africa. It would, however, be naive to think that industrial relations frameworks are the sole, or even the main, reason for labour market outcomes. The restructuring of the labour market has reflected the restructuring of the South African economy. In this section we explore how the new industrial relations system has entrenched the power of capital, first through the inability of collective bargaining (and unions by extension) to protect many workers and then through the very limited effect of direct state intervention in the labour market. Thirdly, the weakening of trade unions is discussed.
Collective bargaining
Until the end of apartheid in 1994 (and since long before its start), South Africa had been characterised by institutionalised racism in the labour market. The large migrant, badly-paid and tightly controlled black workforce was a central feature of the mining economy (Moodie 1994). Whites benefited from institutionalised racism, for instance in the form of job reservations or of the rule that they should always be in a hierarchically superior position. Black trade unions, which developed from the 1970s, first focused on eliminating the most humiliating practices of daily racism in the workplace, such as the blurring between racial and technical lines of hierarchy.7 Their success later provided an essential foundation for the mounting economic and social contestation which would lead to the fall of the apartheid regime. The largest confederation that emerged from this movement, the Congress of South African Trade Unions (Cosatu), has become a member of the Tripartite Alliance which has ruled South Africa since 1994. Since tense (to use a euphemism) labour relations were a key feature of apartheid, a progressive workplace regime was a crucial expectation associated with the democratic transition.
The legislative framework of industrial relations adopted by the new government was based on the Northern European model which entails, firstly, strong social partners including trade unions deeply rooted in workplaces and, secondly, high levels of collaboration between capital and labour. Such a progressive regime has, however, not materialised in South Africa out of the potential contained in the new legislation. The belief that unions would grow stronger in workplaces in the post-apartheid period turned out to be unfounded in most cases. It was expected that social ‘partners’ would collaborate at the national, sectoral and firm levels to agree on policies and pay levels whereas class struggle has proved more visible than class collaboration at all levels, in particular in workplaces.
The cornerstone of the new labour market structure is the Labour Relations Act (LRA) adopted in 1995, which encouraged (or relied on) what Todd (2004) calls ‘process voluntarism’ – effectively leaving most of the regulation of labour relations in the hands of unions and employers. Bargaining councils (BCs, formerly industrial councils) were retained as the primary forum for collective bargaining, and bargaining at sectoral level was preferred. The Act also provided a mechanism whereby unions (or employers’ associations) which were not sufficiently representative to form a BC could seek to establish a statutory council, with limited powers to bargain. There was also provision for a plant-level structure, the workplace forum, influenced by the German work councils.
The number of BCs has declined, after a peak at ninety-nine in 1996.There were forty-seven bargaining councils in 2009. This can be attributed to councils collapsing in some industries, or mergers of regional councils into national councils in others. Councils can have their collective agreement extended to all employers and employees within the scope of the council and not only the members of the party organisations. Companies can (and often do) apply to the BC for an exemption from some or all provisions in the agreement(s); about 80 per cent of applications for exemptions are granted – confirming that the system does not lack ‘flexibility’ (Bhorat and Cheadle 2007).
This seemingly progressive legal regime, however, has been constantly degraded in the past twenty years (Pons-Vignon and Anseeuw 2009). First, the National Economic Development and Labour Council (Nedlac) is hardly the forum for democratic policy consultation which it was intended to be, largely because government has not taken it seriously in many important instances, in particular not when the landmark Growth, Employment and Redistribution (GEAR) programme was adopted in 1996. Beyond Nedlac, and in spite of its continuing participation in the Alliance, Cosatu has been consistently frustrated with the policy options chosen by the ANC and, crucially, the latter’s refusal to have an open debate on decisive policy choices (Cosatu 2002).
In practice, the new architecture has not produced the intended dense web of negotiated conditions. BCs have not extended their scope to any significant extent in the post-1994 period – and in some sectors have declined. The tendency has been away from centralised bargaining (which is typically associated with the best outcomes for workers) to more local, fragmented bargaining. Coleman (forthcoming) thus argues that:
While centralised bargaining is critical for the labour movement … it is not in its current voluntaristic form able to drive the transformation demanded by the South African situation. Indeed the fragility of collective bargaining institutions can be used by employers to attack wage levels.
This is particularly visible in wage agreements, with low wages inherited from apartheid being perpetuated in many cases. The implication is that wages vary greatly across sectors, reflecting the strength (or weakness) of unions: in 2005, the minimum wage for unskilled electrical workers was R185 per week, whereas it was R647 for the same category of workers in metals and engineering (Budlender 2009). Ironically, some of the most favourable wages and conditions have been achieved by the National Union of Metalworkers of South Africa (Numsa) without a bargaining council! (In the auto assembly sector, a national bargaining forum, governed by rules agreed upon between employers and the union, has thus allowed for flexible and effective bargaining.)
Other institutional forms supposed to promote collective bargaining have had even less success. Workplace forums were supposed to be consulted about the restructuring and reorganisation of production, training, pay schemes and disciplinary procedures – but only a handful have been established. Employers were reluctant about co-determination and joint problem solving, while unions ‘feared that (workplace forums) could serve to co-opt workers and reduce their willingness to oppose employer proposals’ (Budlender 2009: