Article Economic Adjustment, Political Institution- alisation and Social Marginalisation: COS ATU and the First Democratic Government (1994-99) Franco Barchiesi Introduction In the period between its first two democratic elections, the South African political transition has known important peculiarities which differentiate this case from an African scenario characterised, in the same time, by uncertain shifts to multi-party systems and by the consolidation of policies of economic structural adjustment. Three decisive aspects, in particular, apply to the role of labour in the South African transition. First, organised labour provided a decisive contribution, during phases of struggle as well as negotiation, in overcoming apartheid and setting the stage for the establishment of democratic political institutions. Second, labour's role in the transitional and post-transitional phase was rooted in a formal electoral and programmatic alliance between the biggest union federation, the Congress of South African Trade Unio'ns (COSATU), the leading party in government, the African National Congress (ANC), and the South African Communist Party (SACP). Such an alliance immediately defined COSATU's role, beyond the mere representation of worker interests, as a factor of integration of such interests in a project of governance that includes different, and often contradictory, class forces, domestic imperatives and international pressures. Third, South Africa's political transition paralleled a transition towards market-based solutions to challenges of economic competitiveness and social integration. However, the shift to 'neo-liberalism' was in this case, compared to other African countries, less dependent on economic crisis and the constraints imposed by international TRANSFORMATION 38 (1999) ISSN 0258-7696 COSATU and the First Democratic Government financial institutions, and more a matter of contestation over strategic choices between labour, state and capital. This paper will provide an overview of debates and developments that have affected COSATU's role and perspectives during the first five years of post-apartheid representative democracy. Such a discussion will introduce an evaluation of challenges that will likely face organised labour in South Africa after the 1999 general elections. The general argument of this article is that the political transition in South Africa has been decisively affected by COSATU as a governing force and as a factor of social stabilisation in a context marked by macroeconomic free-market orthodoxy. However, the institutionalisation of organised labour, implied in such outcome, selectively and unevenly affected other dimensions of COSATU's life, determining in particular a growing marginalisation of the federation as a critical actor in spheres of industrial and workplace change, and labour market restructuring. The progressive narrowing of policy priorities around free-market orthodoxy will, probably, raise new questions concerning independent working class political representation and alliance politics in the post-election period. Labour and Political Institutionalisation The independent trade unions that emerged from the struggles of the black factory working class during the 1970s and 1980s grew with little explicit organisational or political relationships with the historical liberation movements in exile. National union federations, such as the Federation of South African Trade Unions (FOSATU) in 1979 and COSATU in 1985, derived out of a convergence of various union groupings from different organisational traditions, but generally articulated around strong shopfloor identities and a culture of grassroots control of the organisation (Friedman 1987; Baskin 1991; Barchiesi 1992). COSATU's workplace orientation greatly contributed to its consolidation and expansion. This included the role of the shop steward as an agent of representation and organisation, the prominence of workplace structures over territorial offices and the statutory predominance at any organisational level of elected worker representatives over full-time officials. COSATU's assumption of a precise politicaljole to replace a political opposition heavily hit by the states of emergency in the second half of the 1980s (Webster 1988) led to a close alignment with the ANC-SACP alliance, based on COSATU's adoption of its political programme. 21 Franco Barchiesi Parallel to this, labour gained an unprecedented relevance in the redefinition of the industrial relations system. The transition in this field largely pre-dated political transition itself. COSATU demanded, in fact, a compulsory system of national centralised bargaining by sector complemented by a body for centralised tripartite social and economic policy making. Finally, the articulation and flexibility between centralised bargaining and workplace negotiations became a topic for intense debate (Baskin 1995). Worker mobilisation was crucial in explaining COSATU's role in these institutional transformations - at least while an illegitimate apartheid government remained in place - and worker militancy did not decline even in periods of economic crisis (Innes 1992). This translated into labour's contribution to changes in the Labour Relations Act (LRA), in its influence on government policies, such as the Reconstruction and Development Programme (RDP), and in the unions' participation alongside state and organised business in national tripartite forums on social and economic policy making, which culminated in the 1995 launch of the National Economic Development and Labour Council (NEDLAC). However, the political transition meant for COSATU a growing engagement with new institutional responsibilities. The growth and development of the federation contributed to shape some of the critical choices it has now to face. Steady numerical expansion, increased complexity of the bargaining agenda, the need for a more articulated division of tasks inside the organisation, are part of this scenario. The militant impetus predominant during resistance politics is now being relativised, while a new attitude to compromise and non-adversarialism at a centralised and tripartite level and on the workplace has emerged. Many affiliates approached management-initiated forums for workplace restructuring and productivity enhancement in a way that often combined participation with confrontation in shaping powers, guarantees and procedures of such structures (Mailer 1992; Leger 1992; Buhlungu 1996; Barchiesi 1997). Moreover, unions have been increasingly involved in industry-level negotiations on industrial policy, restructuring, training and competitiveness, which often implied managing the social consequences of labour market mobility, downsizing and retrenchments (Maree and Godfrey 1995; Steinberg and Seidman 1995; Von Holdt 1995b). Finally, COSATU achieved its long-standing goal of being recognised as a player, together with business and government, in centralised tripartite policy-making on macroeconomic reform, social reconstruction, 22 COSATU and the First Democratic Government employment creation and labour market change. As a consequence, South African labour studies were reorientated towards views which largely focused on trade unions as institutionalised worker representatives. This means that unions had to become actors in enforcing 'social pacts' based on the exchange between social stability, industrial peace and economic competitiveness, on one hand, and a 'socially sensitive' view of economic reconstruction, open to redistribution, productivity-related increases and employment security, on the other. Karl Von Holdt's (1995a) concept of 'strategic unionism' defined the combination of levels of union activity beyond traditional working class concerns, in a pragmatic use of radicalism and mobilisation inside both central and localised social institutions. Glenn Adler and Eddie Webster (1995) captured the unions' contribution to the transition under the heading of'radical reform'. They argued that labour's radical programmes and strategies of mobilisation decisively influenced the transition through their interactions with institutionalised pact-making processes that involved reformers in the old regime and moderates in the opposition. Such interaction would arguably avoid South African labour being demobilised or tamed by the transition. Rueschemeyer, Stephens and Stephens (1992:55) argue for the 'inherent ambiguity of organised collective action' to show how union leaderships face contradictory appellations from their own rank-and-file and 'other power holders' respectively, with potentially unpredictable outcomes. The macroeconomic policy choices of the ANC government are a clear example of contradictory pressures facing COSATU. The federation's contribution to the 1994 electoral programme of the Alliance could be mainly found in the Reconstruction and Development Programme (RDP). Its final drafts, however, blurred labour's social-democratic emphasis on demand management, redistribution and domestic-driven growth, creating an uneasy coexistence with outward-looking development and the promotion of foreign investment (Ismail 1994; Munck 1994a; Bond 1999). A neoliberal orientation became predominant in post-RDP policy documents (Adelzadeh and Padayachee 1994; Adelzadeh 1996; International Monetary Fund 1996), culminating in the 1996 Growth Employment and Redistribution (GEAR) strategy (Department of Finance 1996). This shift, in particular, defined growth as premised on orthodox assumptions that combined the stability of macroeconomic indicators with the search for global private competitiveness. Moreover, once the notion of economic growth was recast inside these new boundaries, it became a precondition for social 23 Franco Barchiesi transformation, rather than being reciprocally integrated, as in the RDP's phrasing. GEAR maintained a vague commitment to job creation, which was, conversely, made dependent on fiscal discipline, public spending restraint (a three per cent deficit-to-GDP ratio by year 2000 with further expansion conditional upon the, by now highly unlikely, growth rate of 6.1 per cent), export promotion and privatisation. Moreover, while redistribution was subordinated to economic growth, the growth potential of redistribution was ignored, and a tight monetary policy was privileged over an expansionary budget for the management of the industrial strategy and domestic demand (National Institute for Economic Policy 1996,1998). As Patrick Bond (1999) notices, the rise of a free-market hegemony in South African policy making responds to a logic of 'homegrown structural adjustment' that, unlike experiences with neoliberal restructuring common elsewhere on the continent, depends largely on strategic choices of domestic economic and political actors, rather than on external conditionalities. This feature highlights, therefore, the contradictions for labour implied in the combination of promoting the country's position in global competition for markets and investment, and seeking social compacts through institutionalised negotiations to gain legitimacy, social control, and grassroots cooperation. The rise inside the Alliance of a free-market, neo-liberal response to issues of socio-economic transformation, where primacy of the private sector and labour market flexibility feature prominently, has emphasised, more than COSATU's capacity for mobilisation and far-reaching programmes for change, its nature as representative of specific class forces. As such, it is pressurised to find terrains of mediation with powerful economic actors linked to multinational capital and international financial institutions: The government's conservative economic strategy has generated contradictory demands on labour - greater stability, productivity and flexibility. Greater productivity and competitiveness, for instance, is made to pivot on organised workers' support for training schemes, changes in production and shift systems and pay deals linked to work performance, as well as greater flexibility in the labour market at large. On the other hand, workers' demands for better wages and working conditions are deemed to undermine investor confidence. (Marais 1998:231) 24 COSATU and the First Democratic Government At the same time, macroeconomic orthodoxy can narrow institutional spaces for negotiation and mediation. This contradiction places clear limits to the role of labour advanced by the South African transition. It can moreover question whether a priority on pragmatism and institutional negotiation coincides with rank-and-file views of radicalism and militancy. The social construction of militancy cannot in fact be limited only to the ways this is translated in the formal organisations' discourse (Barchiesi 1999). Episodes such as workers' opposition to the 1995 LRA (Von Holdt 1995c), to the three-year automobile industry agreement of the same year (Bohmkc and Desai 1996; Barchiesi 1997), or to the privatisation of municipal services (Van Driel 1998) confirm the limitations suffered by a purely institution-based understanding of labour's role and outcomes in the transition. Conversely, the diminishing importance of militancy and mobilisation in defining COSATU's space in the transition1 supports Sakhela Buhlungu's (1997) argument that the rise of unions' 'influence' in government has in many ways contrasted with their loss of independent 'power' inside the Alliance. Important factors that facilitate such outcomes are the exodus of experienced union leaders towards political positions in the ANC and government, and the growing involvement of COSATU in economic activities such as Union Investment Funds. The rising relevance of the latter indicates an integration of labour in a changing South African business landscape, for which traditional mechanisms of grassroots control are proving inadequate (Naidoo 1997). Finally, COSATU's shop stewards have been historically aware of dangers of lack of transparency in the definition of the Alliance's political strategy, and of differences in political styles between the federation and the ANC, with particular regard to issues of internal democracy (Von Holdt 1991:17-21). These dilemmas facing COSATU indicate broader ambiguities in approaches to the constitutionalisation of labour in South Africa. Hardt and Negri's (1994) view of constitutionalisation of labour implies legislative and institutional provisions aimed at integrating representative workers' organisations in processes of societal bargaining based on the exchange between productivity and social wage. In this way, strong and disciplined identities based on waged work would ensure the compatibility of worker militancy and capitalist development. Theories of the welfare state (Offe 1984; Esping-Andersen 1990; Kolberg 1992) emphasise that constitution- alisation of labour as defined here is more easily pursued in societies with 25 Franco Barchiesi strong traditions of factory workers' struggle. At the same time they require Keynesian-style employment policies as both a condition for the funding of the system and as a consequence of its pattern of social control. However, high unemployment and policy-imposed constraints to social resources for redistribution prevents this option for South Africa. As a result, COSATU's nature as part of the political ruling alliance defines its role as imposing discipline on its membership for the sake of social stability and economic competitiveness. But the lack of the social bases for the kind of constitutionalisation of labour that such an aim would require makes control of the grassroots more uncertain. Therefore, alternative routes will have to be experimented with. Similar dilemmas have opened, in different contexts, the alternative of a purely defensive role for the union movement. This aspect is in particular emphasised in countries affected by structural adjustment in the presence of historically strong labour movements (Munck 1994b). For example, in many Latin American countries the unions' role in 'social contracts' meant the acceptance by workers, as a consequence of heavy defeats, of a compromise aimed at maintaining some linkage between income and productivity for union members in the formal sector in a context of increasing labour market flexibility. To avoid such a scenario COSATU has balanced until now its position between an allegiance to the ruling coalition and a kind of flexible and piecemeal negotiation on the economic reconstruction policies (Schreiner 1994). However, the degree of flexibility and sacrifices in such an option needs to be evaluated more carefully. In fact, a strong trend is emerging inside the ANC to define labour's contribution to the transition in terms of restraint to uplift the 'more disadvantaged' in a mode of decision-making where the borders between concertation and technocracy are not always clear. The implications that such a situation carries for labour emerge clearly from a discussion document published in circles close to the Vice-President, Thabo Mbeki: Proceeding from the objective reality of the place and role of private capital ... which exists independent of our subjective wishes, the correct strategic decision the democratic movement must take is that the democratic state must establish a dialectical relationship with private capital as a social partner for development and social progress... This is meant to describe a complex, contradictory, co-operative and dynamic relationship, many of whose elements are formed or decided at the international level. What is certain is that there is a need for co-existence and co-operation between the democratic state and private 26 COSATU and the First Democratic Government capital in order to address social development.... (African National Congress 1996:10-11) Here, while capital's confidence and foreign investment are recognised as primary engines of growth, also thanks to the rhetorical advocacy of allegedly uncontrollable international constraints, on the part of labour the document emphasises the centrality of the continuing and special role of the progressive trade union movement and its leadership to the mobilisation of black workers to understand and adhere to the broader objectives of ' the process of democratic transformation, in their own interest. The instinct towards 'economism' on the part of the ordinary workers has to be confronted through the positioning of the legitimate material demands and expectations of these workers within the wider context of the defence of the democratic gains as represented by the establishment of the democratic state. (African National Congress 1996:12) COSATU's response recognises that in many instances the transition i has been led not by its alliance with the ANC but by 'the old bureaucracy, business advisers, economists from the Reserve Bank, the World Bank, etc' (Congress of South African Trade Unions 1996:6), and that demobilisation and demoralisation of the rank-and-file followed, while the democratic state has been 'confined to limited areas of governance' (African National Congress 1996:8). However, the union federation still does not question , the institutionalisation of its alliance with the ANC. The image of the government as simply captive of uncontrollable forces, instead of being an actor capable of actively and strategically relating with markets and '< institutions, is rather functional to advocate a new 'reconstruction accord' between COSATU and the ANC to revive the RDP. On the other hand, the , nature and contents ofsuch an'accord' are not clearly defined by COSATU. The federation's adoption of a pragmatic approach to 'flexible independence' (September Commission 1997:31-32) supports electoral pacts with the ANC that include COSATU's 'top priorities' as a condition to maintain the Alliance in place even in the absence of a common political programme. Assuming that GEAR, defined as 'non negotiable' by the > ANC, is the biggest stumbling block for such a programme, the definition of the contents of an electoral programme for the 1999 elections raised | fresh contradictions between COSATU and the ANC. This adds to the differences that emerged between the two, for example on the legislation i around pension funds, training, competition policy, municipal privatisation and the definition of labour's role in the framework agreement on 27 Franco Barchiesi privatisation. In these regards COSATU and affiliates' officials have often attacked the ANC and the government on accounts of lack of democracy and transparency (Buhlungu 1997). On the other hand, the 1998 Presidential Jobs Summit, hailed as the most important tripartite job creation initiative during the ANC's first term, fell far short of endorsing a reversal of macroeconomic orientations towards an expansionary, demand-based employment policy. The Summit confirmed in fact GEAR's constraints on state intervention, even if it allowed the 1999 Budget marginally to loosen the deficit targets. Apart from an emphasis on specific projects in labour-intensive sectors, based on ad-hoc financing and with little discussion on the nature of jobs created, the only real long-term gain of COSATU from the Summit has been the recognition in the final Declaration of a focus on 'job retention'. This will probably open up new opportunities for labour to pressurise for meaningful negotiations over retrenchments. However, the prospects for a reduction of COSATU's subordination to the dominant socio-economic policy framework do not look much clearer after the Summit than before it. The Changing Regime of Industrial Relations and Societal Bargaining The tripartist experiment and its uncertainties During the 1994-99 period, COSATU's numerical growth has consolidated and has been, at the same time, paralleled by significant changes in its composition and the relative weight of affiliates (Naledi 1994; Macun and Frost 1994). In fact, for the first time the federation has been able to relativise its once absolutely predominant 'blue collar' image. While its biggest manufacturing and mining affiliates (NUMSA, SACTWU and NUM) had to cope with the impact of work and employment restructuring - which involved retrenchments and decentralisation of production (Andrew Levy and Associates 1998) - public sector unions (NEHAWU, SADTU and SAMWU) have shown the most remarkable trends to the increase in membership.2 These growth patterns are affected, on the other hand, by contradictory developments and contingencies. From one side, larger spaces for union organising have been created by the disappearance of the 'homeland' governments, by a greater freedom of organisation in the public sector, by the affiliation of new unions (especially the teachers' union SADTU) and by new organisational rights for workers previously excluded. On another side, while public sector unions have not been spared 28 COSATU and the First Democratic Government by the 'brain drain' of expert leaders to the ANC, recent commitments by the government to downsize public employment in a context of heightened wage militancy place heavy questions on the future of these unions. The rising subordination of COSATU to the ANC in the Alliance context carries important implications in the sphere of restructuring and change in industrial relations institutions. In particular, it can be questioned whether COSATU has been able, in the first five years of democratic government, to retain its strategic initiative and to advance general transformative programmes through these institutions. On the contrary, there are indications that such structures have developed mainly as vehicles for the legitimation and the acceptance of downward compromises along the lines suggested by the re-orientation of macroeconomic paradigms in the political system. The National Economic Development and Labour Forum (NEDLAC) was started in 1995 from the merger of pre-existing separate bodies with industrial relations and economic development competencies. Its four 'chambers' operate as bodies of tripartite consultation (with the inclusion of a 'community forum') between state, business and labour representatives, and as organs for legislative recommendations to the Parliament (Maree 1998). However, NEDLAC has been marked in the first four years of its existence by the same fundamental ambiguities at the heart of its predecessors' experiences. In fact, it has generally not solved the question concerning its being a proper decision-making body, which would make it an effective vehicle for a real constitutionalisation of labour, or if, on the other hand, it is still to be intended as an advisory forum aimed at limited consensus-building on short-term issues. NEDLAC's own position seems orientated to this latter understanding: NEDLAC is a vehicle for the social partners to mobilise their constituencies into an effective joint strategy for social and economic transformation.... We have reached a moment wherein the leadership of each constituency is willing to consider the possibility of making short-term trade-offs and of mobilising our collective resources in order to create a longer-term outcome which will deliver employment, economic growth and uplift the living standard of all.... There is no other alternative open to South Africa except to seek a meaningful social partnership (NEDLAC 1995:3). The argument continues by suggesting that wage policies will arguably combine, far from Keynesian ambitions, considerations linked to productivity increases with the aims of inflation control - in line with government's macroeconomic approaches and without wage indexation — 29 Franco Barchiesi and a very loosely defined 'floor of improved social benefits' (NEDLAC 1995:6). COSATU's analysis has shown difficulties in dealing with ambiguities that surround NEDLAC, and the debate at the level of affiliates and territorial structures has often proven inadequate. Moreover, the trade union field appears fragmented and heterogeneous, and very few of the existing industrial unions seem capable of an effective engagement in industrial and economic policy-making (Lloyd and Rix 1995). This is partly due to the differential competitive potential of different sectors, which faces some unions with the alternative of sustaining restructuring on the employers' terms or adopting a purely reactive and defensive approach (Hirsch and October 1994). On the other hand, leading businessmen presented NEDLAC's rationale and aims as: '... to reach principled compromises in the face of harsh uncomfortable realities... to facilitate the transformation of South Africa into an economy which will succeed in the global economy' (Bobby Godsell, Business Day, January 3, 1995). Such views seem to limit both the scope of NEDLAC's policy interventions and labour's leverage inside it. To these problems are added further questions which emerged out of NEDLAC's operations. In fact, if this institution can promote organised labour's commitment to growth and competitiveness, including the acceptance of economic sacrifices, however the effective representativity of employers' organisations, and their capacity to enforce agreements among their constituencies are still in question. Moreover, business tends to privilege bilateral contacts with the government and decentralised workplace negotiation, which substantially reduce NEDLAC's roles (Pretorius 1996). In fact, if the Council's ambition is to promote interactions between parties finalised to social cooperation, NEDLAC also recognises that the abandonment by the government of its interventionist approach has created a more confrontational environment between labour and capital. Then NEDLAC's options are limited to promote consent not on broad processes of social reconstruction, but on more detailed aspects such as 'social clauses' or specific sections of the LRA {NEDLAC News Update July 1, 1996 and January 2, 1997). These shortcomings, which sometimes led to bitter clashes, for example during the discussion of the 1998 Budget, are particularly resented by many COSATU affiliates, which are distant from top-level tripartite bargaining and whose members have to face directly the consequences of economic restructuring, downsizing, retrenchments and outsourcing. In the case of 30 COSATU and the First Democratic Government the municipal workers' campaigns against privatisation of local public services, for example, grassroots criticism also touched the increasing subordination of COSATU, reinforced by corporatist tripartite bargaining.3 In particular, a wide perception exists among unions that, if labour cannot decisively shape macroeconomic and industrial policy orientations, there is little point in taking part in NEDLAC structures that are largely confined to details (Marais 1998:230-4). COSATU's dilemmas are reinforced by the absence in South Africa of institutional traditions that in countries with deeply entrenched union organisations have developed along the lines defined by Esping-Andersen (1990) as, respectively, 'social-democratic' and 'corporatist'. While the former provided for a decisive redistributive role for the state in ensuring universal access to basic social rights, the latter replaced such a role with clearly defined and formally regulated negotiating institutions for an effective social policy making based on the allocation of socio-economic resources to different status groups. However, while the GEAR approach precludes a decisive redistributive state role, it is also questionable the definition of a NEDLAC-style body as a 'corporatist' institution (Baskin 1993), due to its noticed ambiguities in policy-making attributions. In understanding the experiment of NEDLAC, it seems more useful to point at its strong ideological component as a vehicle to legitimise mutual restraint between the parties in the name of national unity (Habib 1995). Therefore, rather than having an impact in deciding fundamental socio-economic policy orientations, this experiment is rather orientated towards mitigating the social consequences of macroeconomic frameworks whose decision is outside its scope of action. Von Holdt (1993:22) sees in the corporatist discourse, whose bases are increasingly weakened in many of the areas where it is historically strong (Scandinavia, Germany, Australia), a danger of cooptation and bureaucratisation. In contexts that, similarly to South Africa, are characterised byapparent inequalities and exclusion from formal labour markets, institutions of a corporatist kind can be conducive to conservative alternatives to redistributive policies, which would otherwise encourage radical expectations and a militant rhetoric counterproductive for the aims of 'development' (Heller 1995). From this point of view, NEDLAC is rather similar to Latin American flirtations with corporatism (Gibson 1997; Weyland 1996). These, more successfully than authoritarian forms of decision making, have introduced neoliberal reforms in countries ruled by 31 Franco Barchiesi parties that, like the ANC in South Africa, are supported by multiple constituencies and legitimated by a non-class developmental discourse. In these countries, the populist rhetoric on the 'needs of the very poor' provide an effective tool for the ruling party to mobilise 'peripheral' constituencies (rural, unemployed) to impose restraint on their own 'metropolitan' bases of support (primarily organised labour). In the Australian case, often mentioned by South African advocates of corporatism as an example of union engagement in social compacts, the inclusion of the unions in tripartite institutions that promoted liberal economic policies had, moreover, devastating effects on grassroots' union support (Hampson 1997; Bramble and Kuhn 1999). The conclusion from these cases seems to suggest that neoliberalism and corporatism do not constitute clearly discernible alternatives in processes of societal bargaining hegemonised by a free-market orientation to economic policy. The wide range of restructuring options open for capital and the different meanings of 'corporatism' that can co-exist inside NEDLAC reflect a substantial asymmetry in the state's relations with business and labour respectively. In a context whereby COSATU is organically linked to the majority party in government, this gives a wide autonomy to the government to use tripartism as a device to enforce macroeconomic discipline and wage restraint on a labour movement confined on the defensive (Habib 1997). Steve Gelb (1998) notices that the crucial role of the state in tripartite bodies, combined to its unwillingness to enforce duties and obligations on business, leads to a mixed policy model. This would combine a strong central power holding the prerogative to insulate 'non negotiable' general principles of macroeconomic policy from societal bargaining with an emphasis on market forces in the intermediation of different interests. 'Tripartite bargaining' would then be narrowed down and confined to agreeing on the technical ways to implement those principles. Therefore, NEDLAC, as Leo Panitch (1996) argued, rather than responding to an optimist view of 'democratic corporatism' would function according to the same logic of state-led union subordination that gave corporatism a 'bad name' among trade unions all over the world: There is no doubt, then, that NEDLAC has been designed as a means of institutionalising and harnessing trade union power... It is an illusion to think that this portends the autonomous and effective use of trade union power in the decisions that govern economic life in South Africa... Industrial policy being discussed at NEDLAC ... is in fact a policy which seeks to integrate South Africa into the framework of free 32 COSATU and the First Democratic Government trade, the deregulation of capital flows and exchange rates, the promotion of privatisation, and export-oriented competitiveness. The active union role in such a framework is expected to be one of supporting plant closures and wage restraint (Panitch 1996:7-8). Finally, COSATU's involvement in tripartite policy-making bodies contributes to confine its original socialist discourse to the abstract level of a mainly legitimating device, which is in practice translated into arguments of self-responsibility and restraint in the pursuit of socio-economic reconstruction and development. Changing normative frameworks The 1995 Labour Relations Act is aimed at defining a post-adversarial industrial relations framework where issues of worker cooperation and participation are central. There are two particularly noticeable elements in the law. First, it rejects the view of the parties' 'duty to bargain'. Even if the act presents itself as a departure from a tradition of voluntarism, its outcome is largely a restructuring of voluntarism itself. In fact, differently from the experience with the old industrial councils (to be renamed 'bargaining councils'), the act explicitly supports collective bargaining. However, such support is mainly indirect and provided by an active legislative protection of the right to industrial action. Minister of Labour Mboweni (1995:26) spoke in this regard about'duty to bargain imposed by power relations'. As a corollary, a comprehensive vision of trade union organisational and bargaining rights is enforced. Moreover, the act introduces a new organ, the 'workplace forum', in establishments with more than 100 employees (Benjamin 1995;BaskinandSatgar 1995). These forums are to be constituted on request by a representative union and intended as bodies for consultation and limited joint decision making on issues not related with wages and working conditions, referred to centralised bargaining. They are aimed at facilitating the establishment of a 'conflict-free' plant level environment (Du Toit 1997). In fact, the government recognised (Ministerial Legal Task Team 1995) the importance of the new LRA in creating a climate of social stability conducive to growth and foreign investment. To this aim is also functional the new emphasis on 'speeding up' the resolution of industrial disputes, substantially relying on mediation and arbitration, to the detriment of traditional labour and ordinary judicial channels. Therefore, the new LRA acquires a much broader scope than what is commonly envisaged in a piece of labour legislation. It also calls the unions 33 Franco Barchiesi to responsibilities at different levels, from self-discipline in the use of their power in order to make the economy viable, to a new role as brokers of stabilisation and industrial peace for the sake of growth and development (Kester 1995). However, the separation contained in the act between distributive issues to be negotiated through collective bargaining at the central level, and production-related aspects, to be settled in the workplace, could inhibit a holistic and strategic approach by labour to bargain over productivity and work organisation. This is all the more likely if one considers that the right to strike is substantially confined to collective bargaining issues (Etkind 1995). Resentment in this regard has been expressed by important COSATU structures, such as the shop-steward council in the Wits Region, the biggest and most militant (Von Holdt 1995c). This problem is compounded by the uncertain demarcation in the act between consultative and joint decision-making issues at the workplace level. It also seems that other aspects will continue to depend on balance of strengths in the various industries, namely the complementarity between workplace and central negotiations and the extension of contracts to non-parties. In this way, the capacity of the LRA to provide coherence to the industrial relations system can be questioned (Klerck 1998). At the same time, the continued relevance of issues of power and conflict in the workplace, regardless to the LRA's stated aims to emphasise areas of 'common interest', questions the capacity of new plant-level structures to facilitate the desired shift to post-adversarialism in industrial relations at the point of production. This latter possibility has been, conversely, raised by those who advocate 'German-style' co-determination as a possible outcome of the establishment of workplace forums (Webster et al 1997; Godfrey et al 1997). In this view, such forums would systematise and provide new legal grounds to isolated experiments with workplace participation (such as in the case of Volkswagen South Africa) in which unions have been proactively engaged in the past, albeit in rather unstable forms. Therefore, the LRA would, at the same time, officially recognise and support a fundamental shift in management's thinking and institutionalise workplace industrial relations in a system of joint responsibility. Here, labour would be enhanced and upgraded (which would include better training, remuneration and job security) through its contribution to quality and productivity. However, as I have shown elsewhere (Barchiesi 1998), no convincing evidence exists of a fundamental divide among South African employers on long-term strategies to adopt forms of 34 COSATU and the First Democratic Government workplace co-determination. Rather, in a context where co-determination is facing a deep crisis due to industrial restructuring in its own 'home' countries, such organisational framework is adopted in South Africa with much more contingent aims. They relate, primarily, to controlling and co-opting shop stewards from strong union 'shopfloor structures and involving them in managing the social and employment consequences of usually unilateral forms of restructuring largely based on cost-cutting technologies and outsourcing (Kraak 1996; Buhlungu 1996). While in many companies (particularly in industries such as motor and automotive), the workers' response has led to the replacement of shop-stewards involved in such schemes, their intended institutionalisation of labour at the grassroots has collapsed. Other companies could, on the other hand, maintain their productive and market performance, in the presence of weaker unions, by resorting to traditional and more authoritarian management styles. Therefore, rather than heralding a new era in managerial thinking, South African experiments with co-determination are to be understood as part of a continuum in strategies of control, whose adoption largely depends on balances of power on the shopfloor. Recent research has underlined workplace arrangements and voluntary and flexible collective bargaining as successful strategies in dealing with global competitive pressures (Ferner and Hyman 1992; Belanger et al 1994). Following such trends, workplace forums have been theorised as providing union organisations with 'an opportunity to become centrally involved in restructuring' (Webster 1996:36). However, the nature of such 'involvement' could be heavily affected by the limitations to the exercise of workers' collective power contained in the LRA. This latter, in fact, reinforces at the workplace level the identification of unions with responsibilities defined at a mainly institutional level, a trend already entrenched in the South African experiment with tripartism. Critics of the Act have convincingly argued (Etkind 1995; Kester 1995; Klerck 1998) that this dispensation can substantially reduce rank-and-file workers' bargaining power, especially in weaker and less competitive sectors. COSATU itself has been well aware of such shortcomings, whose negative effects in terms of protection of employment levels constituted a highly contentious point at the Presidential Jobs Summit. In fact, the labour constituency at the Summit identified a crucial challenge as: Statutory support is potentially inadequate: The inability to negotiate and strike over retrenchments; the voluntarist nature of centralised 35 Franco Barchiesi bargaining.... The combination of escalating retrenchments, historically high unemployment levels, large numbers of unorganised workers in certain sectors, and the difficulties of regulation in small businesses place further constraints on collective bargaining. Collective bargaining has an important but limited role to play, and must be seen within the context of the broader macroeconomic and industrial policy-making environment. (Labour Constituency to the Jobs Summit 1998) The LRA's diversification of bargaining levels seems, moreover, already contradicted by the expansion of plant and company level bargaining in many sectors (Joffe and Lloyd 1996), and by the rise of non-standard forms of employment in industries such as retail, construction and clothing and in establishments not covered by collective agreements (Baskin 1998). In these areas, dynamics of employment creation in precarious and unstable occupations are indeed paralleled by powerful pressures by employers for the decentralisation of bargaining, which have led during the 1990s to the collapse of several bargaining councils. It has been observed in many industrialised countries that the crisis of centralised income and industrial policies often leads to declining union density in the sectors more vulnerable to competition (such as automobile and metal). Many national trade union centres have had, in fact, to face fragmentation of collective bargaining in sectoral, tendentially individualised adjustments to accommodate new requirements for flexibility (Visser 1994:310-1). Trends towards decentralisation, which imply growing vulnerability for the workers at low skill levels and those exposed to international competition, are actually encouraged in South Africa by emphases on 'regulated flexibility' and 'modest minimum wages', contained both in GEAR and in policy formulations of the Department of Labour (Labour Market Commission 1996). A recent ILO study (Standing etal 1996) recommended from this point of view to combine centralised bargaining with forms of 'local level bargaining', arguably more responsive to the productive and competitive potential of different areas of the country. However, 'decentralisation' of bargaining is not the only crucial characteristic of such trends. In fact, they also express an on-going change in power relations, where the firm's profitability and market performance becomes the main criteria of adjustment for worker wages and working conditions. This is not only in opposition with the stated aims of the LRA of making worker power and independence a driving force of socio-economic change, but it even contradicts the emphasis placed by South African tripartism on centrally negotiated social and productivity pacts. It is, indeed, the largely 36 COSATU and the First Democratic Government market-driven restructuring of employment and the labour market, which is taking place, that seems significantly to reduce power and influence of an increasingly institutionalised union movement. The final section of this paper will examine this aspect in more detail. Changing forms of employment: the challenge of social margin- alisation GEAR'S recommendations for increased labour market flexibility are now echoed by a rising intellectual consensus (Nattrass and Seekings 1996; South Africa Foundation 1996; Baskin 1998) supportive of the idea of a 'two-tier labour market'. This concept aims to recognise and entrench, from a normative point of view, an actual reality which sees growing discrepancies in wage rates, working conditions, union rights and collective bargaining coverage between different geographical areas, companies of different size and productions with different skills and technology contents. The already noticed trend towards a decentralisation of bargaining would therefore be addressed, in such a perspective, through a process of 'regulated deregulation' (FernerandHyman 1992). This means that targeted categories of semi- and unskilled jobs, small businesses, and workers in rural and recently industrialised areas would know a relaxation of negotiated minima, and a loosening of regulations on hiring and firing. In this argument, global economic constraints limit the options for such workers to, at most, 'an American-style welfare state regime with a very inegalitarian labour market, where the state provides minimal and stringent means-tested public welfare' (Nattrass and Seekings 1996). Therefore, a system of individualised rewards dependent on capital's need for flexibility would not only create a lower category of labour rights, but also substantially limit labour's institutional role in redistributive policies and employment creation. On the other hand, studies by the International Labour Organisation (1999; Standing et al 1996) recognise that the South African labour market is already extremely flexible, characterised by significant wage gaps between sectors, by the expansion of precarious and non-standard forms of labour, and by employers' exclusion of basic employment protections (Rees 1995; Standing 1997). Trends towards the casualisation of labour are affecting the structure of the working class inside and outside the workplace, following the neoliberal advocacy for the deregulation of employment relations (Horwitz 1995; Standing etal 1996; Crankshaw and Macun 1997;Brosnanetal 1998). The 37 Franco Barchiesi use of non-standard forms of employment is heavily conditioned by peculiar social, economic and institutional factors rooted in strong local specificities (Peck 1996), which threaten the regulatory potential of centralised bargaining and collective organisations. Moreover, the implementation of such strategies is facilitated by existing and long-standing labour market segmentations by race, education and gender, and by patterns of household subordination that facilitate the entrance of the weakest sectors of the labour market in these kinds of jobs (Klerck 1994; Kraak 1996). The decentralisation and subcontracting of production in South Africa is facilitating the rise of forms of 'atypical' employment, especially in labour-intensive sectors (mining, retail, construction, clothing) where most new jobs are now created in such forms (Bezuidenhout 1997; Kenny and Bezuidenhout 1999). A series of diverse employment arrangements (temporary, casual, contract, part-time, homework), generally precarious, limited over time and with reduced benefits has emerged. This facilitates a proliferation of different employment contracts — and often different employers in the case of contracting out of segments of production to third party firms - inside the same production runs, with a resulting fragmentation of workplace bargaining units. In such sectors, these patterns of employment are providing employers with increasingly common solutions to problems of control and reduction of wage costs. In the most extreme cases, the subcontracting of jobs through labour brokers amounts to an absolute commodification of labour, which is deprived of any bargaining power, since the contractual relationship between employees and the company is replaced by a purely commercial relationship between core and client companies (Kenny and Bezuidenhout 1999). However, the expansion of 'atypical' labour has been documented in recent years also in some of the industries affected by the most intense dynamics of technological innovation, such as petrochemicals and the automotive sector (Klerck 1994; Crankshaw and Macun 1997; Barchiesi 1997). Here, outsourcing and the breakdown of unions' constituencies are encouraged by government-sponsored tariff liberalisation and market opening to foreign competition. Such employment practices provide, in fact, capital with cost-cutting advantages to 'produce for the formal sector according to formal sector standards, under "informal" sector conditions' (Theron 1996:9; Bezuidenhout 1997). These dynamics are, moreover, challenging the labour organisations' representative capacity inside 38 COSATU and the First Democratic Government collective bargaining, and they finally confirm observations made in similar contexts that traditional union constituencies are the most likely to be expelled from the labour market by dynamics of precarisation of employment (Humphrey 1996). At the same time, unprotected, unstable and exploitative jobs proliferate for an increasingly vulnerable, and largely female, workforce. Combining aspects of self-employment with the subordination to the factory hierarchy these processes envisage the extension of capitalist command over the territory through loose networks of relations, that partially utilise the 'arypicals' and homeworkers' desire for self-management of time as a vehicle to increase working time and labour flexibility. At the same time, the separation of union organisations from potential bases of support and their difficulty to influence societal bargaining outcomes are increased (Rees 1997). Following a more general trend (Hyman 1997), the rising heterogeneity in forms of employment, combined with pressures to fragment industrial relations in conditions of rising labour weakness in the workplace, adversely impacts on the organisational cohesiveness and the collective solidarity of South African unions. Therefore, the institutionalisation of labour organisations provided by the democratic political system and by the new industrial relations regime does not widen COSATU's options in consolidating democracy in the 'homegrown structural adjustment' context. Conclusions This paper has explored the challenges and contradictions faced by COSATU in its double role as participant in a political democratising coalition and as a representative of worker rights and powers threatened by a free-market macroeconomic restructuring endorsed by that very coalition. Gills, Rocamora and Wilson (1993:23) argue, in their four-countries analysis of recent transitions from authoritarianism to democracy, that: 'The onset of formal representative government changes the conditions under which labour and other popular movements must operate. The pattern emerging ... can be summed up under the headings "mobilisation and realignment'". In the South African case, labour mobilisation was necessary to create a political space for the transition. However, the social forces that were mobilised to that end have been soon after marginalised from the centres of power by the realignment of the ruling party around procedural democracy and conservative economic and social policies (Hippler 1995). COSATU's 39 Franco Barchiesi institutionalisation inside the Alliance worked mainly as a strategy of worker responsibilisation and discipline inside the development and competitiveness discourse than underpinned South Africa's 'homegrown structural adjustment'. At the same time, dynamics of labour institutionalisation did not address COSATU's difficulties in representing material changes in forms of work, employment and power relations in production. The South African job market is narrowing and new entrants obtain extremely low levels of security, rewards and protection. As a result of labour market and employment flexibility, competition between different segments of the working class is likely to increase, undermining union organisations and decentralising interactions between social actors in production. On the other hand, such threats might challenge organised labour to seek a constituency and alliances beyond waged employment, to define broad policy alternatives in terms of redistribution, social wage, and expansion of public intervention in socio-economic policy. This could allow possibilities to overcome 'from below' the problems that the restructuring of work and employment creates 'from above'. However, it is highly questionable that COSATU will find a space to look for such alternatives without straining existing relationships with a ruling party that has repeatedly confirmed its allegiance to 'fiscal discipline'. Such allegiance has been, indeed, confirmed even in dramatic ways, as shown by President Mandela' s and Deputy-President Mbeki's interventions at the 1998 Congress of the SACP, which harshly rebuffed criticisms of GEAR from other Alliance partners. On the other hand, the history of the first five years of tripartism and democratic policy-making in South Africa shows that COSATU has been substantially unable to reverse dominant patterns of economic adjustment by exercising its influence inside the Alliance. There are reasonable grounds to suppose that such ability will further decrease if the representative potential of COSATU unions will continue to be threatened by employment reduction and labour market fragmentation. On the other hand, if such trends will not be reversed through policy adjustments inside the Alliance, the question of possibilities and prospects for an independent working class politics could raise to new prominence in public debates. Notes 1. The findings of research by the Sociology of Work Unit, University of the Wirwatersrand, on COSATU members' attitudes to representative democracy, presented at a "SWOP Breakfast Seminar" on March 26,1999, clearly show a 40 COSATU and the First Democratic Government substantial drop in worker attendance to unions' affairs and activities, and a decreasingly positive attitude to industrial action. On the other hand, most workers interviewed reported that they ignored the existence of policies such as the Growth, Employment and Redistribution strategy, or of institutions such as NEDLAC. 2. The National Union of Mineworkers (NUM) and the National Union of Metalworkers (NUMSA) of South Africa were, at the beginning of 1998, still the largest COSATU unions, with respectively 320,000 and 270,000 of the federation's approximate 1.9 million members. They are followed by the two public sector unions NEHAWU and SADTU (teachers), which together with another public workers' union, SAMWU (municipal workers), recruit about 450,000 members. Union densities are 46 per cent for SADTU, 35 per cent for SAMWU and 24 per cent for NEHAWU (NALEDI Policy Bulletin 1(4) 1998). 3. A very useful collection of documents and publications can be found, in this regard, on the SAMWU's website (http://www.samwu.org.za). References Adelzadeh, A. 1996. 'Growth and Development: Labour and Business Perspectives on Economic Development'. In: J Baskin (ed), Against the Current: Labor and Economic Policy in South Africa. Johannesburg: Ravan. Adelzadeh, A and V Padayachee. 1994. 'The RDP White Paper: Reconstruction of a Development Vision?', Transformation 25. Adler, G and E Webster. 1995. 'Challenging Transition Theory: The Labor Movement, Radical Reform, and Transition to Democracy in South Africa', Politics and Society 23(1). African National Congress. 1996. The State and Social Transformation. Discussion Document, Johannesburg: African National Congress. Andrew Levy and Associates. 1998. Annual Report on Labour Relations. Johannesburg: A. Levy and Associates. Barchiesi, F. 1992. '"Unity in action": Pratiche della democrazia e lotte operaie nel sindacalismo indipendente sudafricano, 1973- 1990', Vols. 1 and 2. Unpublished dissertation, Faculty of Political Sciences, Bologna: University of Bologna. 1996. 'The Social Construction of Labor in the Struggle for Democracy: The Case of Post-Independence Nigeria', Review of African Political Economy 69. 1997. 'Flexibility and Changes in Forms of Workplace Subjectivity: A Case Study of the South African Automobile Assembly Industry'. Unpublished MA dissertation, Johannesburg: University of the Witwatersrand. 41 Franco Barchiesi 1998. 'Trade Unions and Organisational Restructuring in the South African Automobile Industry: A Critique of the Co- determination Thesis', African Sociological Review 2(2). 1999. 'Beyond the State and Civil Society. Labor Movements and Economic Adjustment in African Transitions: South Africa and Nigeria Compared'. Forthcoming in G Bond and N Gibson (eds), Contested Terrains and Constructed Categories: Contemporary Africa in Focus. Boulder, CO: Westview Press. Baskin, J. 1991. Striking Back: A History ofCOSA TU. Johannesburg: Ravan. 1993. 'Corporatism: Some Obstacles Facing the South African Labour Movement', Research Reports 30, Johannesburg: Centre for Policy Studies. 1995. "The Need for Centralized (but Flexible) Bargaining', South African Labour Bulletin 19(1). 1998. 'South Africa's Quest for Jobs, Growth and Equity in a Global Context'. Paper presented at the 11th Annual Labour Law Conference, Durban. Baskin, J and V Satgar. 1995. 'South Africa's New LRA: A Critical Assessment and Challenges for Labour', South African Labour Bulletin 19(5). Belanger, J, P Edwards and L Haiven (eds). 1994. Workplace Industrial Relations and the Global Challenge. Ithaca, NY: ILR Press. Benjamin, P. 1995. "The New Labour Relations Bill', SouthAfrican Labour Bulletin 19(1). Bezuidenhout, A. 1997. 'The Subcontracting of Labour in South Africa: Breaking Mirrors and Extracting Smoke'. Paper presented at the Congress of the South African Sociological Association, Umtata. Bohmke, H and A Desai. 1996. 'Everything's Going Wrong: Toyota,' South African Labour Bulletin 20(6). Bond, P. 1999. Elite Transition. Uneven Development, Neoliberalism and Globalization in South Africa. London: Pluto Press. Bramble, T and R Kuhn. 1999. 'Can Corporatism Survive? A Comparative Study of Australia and South Africa'. Forthcoming, South African Journal of Labor Relations. Brosnan, P, F Horwitz and P Walsh. 1998. 'Workplace Flexibility and Labour Cost Reduction in Three Countries: A Study of Small and Medium 42 COSATU and the First Democratic Government Sized Businesses'. Paper presented at the World Congress of the International Industrial Relations Congress, Bologna. Buhlungu, S. 1996.'Trade Union Responses to Participatory Management: A Case Study'. Unpublished MA thesis, Johannesburg: University of the Witwatersrand. Buhlungu, S. 1997. 'Flogging a Dying Horse? COSATU and the Alliance', South African Labour Bulletin 21(3). Congress of South African Trade Unions. 1996. 'A Draft Program for the Alliance'. Discussion paper presented to the Executive Committee, Johannesburg: COSATU. Crankshaw, O and I Macun. 1997. 'External Labour Market Flexibility: Is There Employment Flexibility in South African Industry?' Paper presented at the Conference on 'Labour Markets and Enterprise Performance in South Africa', January, Pretoria. Department of Finance. 1996. Growth, Employment and Redistribution: A Macroeconomic Strategy. Pretoria: Government Printer. Du Toit, D. 1997. 'Industrial Democracy in South Africa's Transition', Law, Democracy and Development 1(1). Esping-Andersen, G. 1990. The Three Worlds of Welfare Capitalism. Cambridge: Polity Press. Etkind, R. 1995. 'Rights and Power: Failings ofthe New Labor Relations Act', South African Labour Bulletin 19(2). Ferner, A and R Hyman. 1992. 'Introduction. Industrial Relations in the New Europe: Seventeen Types of Ambiguity'. In: A Ferner and R Hyman (eds), Industrial Relations in the New Europe. Oxford: Blackwell. Friedman, S. 1987. Building Tomorrow Today: African Workers in Trade Unions, 1970-1984. Johannesburg: Ravan. Gelb, S. 1998. 'The Politics of Macroeconomic Policy Reform in South Africa'. Paper presented at the Conference on 'Democracy and the Political Economy of Reform', January, Cape Town. Gibson, E L. 1997. 'The Populist Road to Market Reform. Policy and Electoral Coalitions in Mexico and Argentina', World Politics 49(3). Gills, B, J Rocamora and R Wilson. 1993. Low Intensity Democracy: Political Power in the New World Order. London: Pluto Press. Godfrey, S, P Hirschsohn and J Maree. 1997. 'Where is Management Going? Employer Strategies with Regard to Worker Participation 43 Franco Barchiesi and Workplace Forums'. Paper presented at the Annual Congress of the South African Sociological Association, Umtata. Habib, A. 1995. 'How Much of Civil Society Could Be Included in a Corporatist-Type Agreement or Structure?'. In: R Humphries and M Reitzes (eds), Civil Society after Apartheid. Johannesburg: Centre for Policy Studies. 1997. 'FromPluralismtoCorporatism: South Africa's Labour Relations in Transition', Politikon 24(1). Hampson, I. 1997. 'The End of the Experiment: Corporatism Collapses in Australia', Economic and Industrial Democracy 18(4). Hardt, M and A Negri. 1996. Labor ofDyonisus. A Critique of the State Form. Minneapolis, MN: University of Minnesota Press. Heller, P. 1995. 'From Class Struggle to Class Compromise: Redistribution and Growth in a South Indian State', Journal of Development Studies 31(5). Hippler, J. 1995. 'Democratization of the Third World after the End of the Cold War'. In: J Hippler (ed), The Democratization of Disempowerment: The Problem of Democracy in the Third World. London: Pluto Press. Hirsch, A and L October. 1994. 'The Clothing and Textile See-Saw: What Future for Tripartite Industrial Policy?' South African Labour Bulletin 18(6). Horwitz, F. 1995. 'Flexible Work Practices in South Africa: Economic, Labour Relations and Regulatory Considerations', South African Industrial Relations Journal 26(4). Humphrey, J. 1996. 'Responses to Recession and Restructuring: Employment Trends in the Sao Paulo Metropolitan Region, 1979-87', Journal of Development Studies 33(1). Hyman, R. 1997. 'Trade Unions and Interest Representation in a Changing Europe'. Mimeo, Industrial Relations Research Unit, Coventry: University of Warwick. Innes D. 1992. 'Labour Relations in the De Klerk Era', South African Review 6(3). International Labour Organization. 1999. Country Studies on the Social Impact of Globalisation: South Africa. Geneva: ILO. International Monetary Fund. 1996. IMF International Economic Position. Submitted to the Seminar for Trade Union Leaders in Southern Africa on 'Policies for Economic Growth and Development 44 COSATU and the First Democratic Government in Southern Africa,' hosted by IMF/IBRD/ILO, March 31 - April 3, Harare. Ismail, F. 1994. 'Industrial Strategy and Economic Development in South Africa: 1990-93', Review of African Political Economy 59. Joffe, A and C Lloyd. 1996. Collective-Bargaining Agreements 1995. Volume One: Innovations and Trends. Johannesburg: NEDLAC. Kenny, B and A Bezuidenhout. 1999. 'Contracting, Complexity and Control: An Overview of the Changing Nature of Subcontracting in the South African Mining Industry'. Paper presented at the Symposium on 'Leaner and Smarter Outsourcing in the Mining Industry', Johannesburg. Kester, G. 1995. 'Towards Effective Worker Participation in South Africa'. In: National Labour and Economic Development Institute (ed), Opinions on the New Labour Relations Bill. Johannesburg: Progress Press. Klerck, G. 1994. 'Industrial Restructuring and the Casualization of Labour: A Case Study of Subcontracted Labour in the Process Industries', South African Sociological Review 7(1). 1998. 'Between Corporatism and Neo-Liberalism? Collective Bargaining and South Africa's New Labor Relations Act', African Sociological Review 2(1). Kolberg, J E (ed). 1992. Between Work and Social Citizenship. Armonk, NY: M.E. Sharpe. Kraak, A. 1996. 'Transforming South Africa's Economy: From Racial Fordism to Neo-Fordism?', Economic and Industrial Democracy 17(1). Labour Constituency to the Jobs Summit. 1998. CreatingJobs in South Africa: Key Issues and Strategies. Labour Input to the Presidential Job Summit, April. Labour Market Commission. 1996. Restructuring the South African Labour Market. Report of the Presidential Commission to Investigate Labour Market Policy, RP 83/1996, Pretoria: Government Printer. Leger, J. 1992. '"Talking Rocks": An Investigation of the Pit Sense of Rockfall Accidents amongst Underground Gold Miners'. Unpublished PhD dissertation, Johannesburg: University of the Witwatersrand. Lloyd, C and S Rix. 1995. 'Unions and Democratic Institutions'. Discussion Paper, Johannesburg: NALEDI. 45 Franco Barchiesi Macun, I and A Frost. 1994. '"Living Like There's No Tomorrow": Trade Union Growth in South Africa, 1979-1991', Social Dynamics 20(2). Mailer, J. 1992. Conflict and Cooperation. Johannesburg: Ravan. Marais, H. 1998. South Africa: Limits to Change. The Political Economy of Transformation. London: Zed Books. Maree, J. 1998. "The COSATU Participatory Democratic Tradition and South Africa's New Parliament: Are They Reconcilable?', African Affairs 386. Maree, J and S Godfrey. 1995. 'Trends Towards Corporatism in South Africa: Industrial Relations and Industrial Strategy in the Textile Industry'. In: A van der Merwe (ed), Industrial Sociology: A South African Perspective. Johannesburg: Lexicon. Mboweni, T. 1995. 'Interview: The World of Work and the Economy', South African Labour Bulletin 19(1). Ministerial Legal Task Team. 1995. 'Explanatory Memorandum to the Draft Negotiating Document in the Form of a Labor Relations Bill', Government Gazette, 16259, Pretoria: Government Printer. Munck, R. 1994a. 'South Africa: The Great Economic Debate', Third World Quarterly 15(2). 1994b. 'Workers, Structural Adjustment and "Concertacion SociaF' in Latin America', Latin American Perspectives 21(3). Naidoo, R. 1997. 'Unions and Investments. Preliminary Assessment and Framework Development', Discussion Papers, RR 01/ 97, Johannesburg: NALEDI. NALEDI. 1994. Unions in Transition: COSATU at the Dawn of Democracy. The National Labour and Economic Develop- ment Institute (ed). Johannesburg: NALEDI. National Economic Development and Labor Council. 1995. Discussion Document on a Framework for Social Partnership and Agreement Making in NEDLAC. Johannesburg: NEDLAC. National Institute for Economic Policy. 1996. From the RDP to GEAR: The Gradual Embracing of Neo-Liberalism in Economic Policy. Johannesburg: NIEP. National Institute for Economic Policy. 1998. Poverty Elimination, Employment Creation and Sustainable Livelihood in South Africa. Johannesburg: NIEP. 46 COSATU and the First Democratic Government Nattrass, N and J Seekings. 1996. 'Citizenship and Welfare in a Labour-Surplus Economy: the Case of Post-apartheid South Africa'. Paper presented at the Australian Political Science Association Annual Conference, Perth. Offe, C. 1984. Contradictions of the Welfare State. London: Hutchinson. Panitch, L. 1996. 'Cosatu and Corporatism: a Response to Eddie Webster', Southern Africa Report 11(3). Peck, J. 1996. Work Place. The Social Regulation of Labor Markets. New York, NY: Guilford. Pretorius, L. 1996. 'Relations between State, Capital and Labour in South Africa: Towards Corporatism?' Journal of Theoretical Politics 8(2). Rees, R. 1995. 'Assessingthe94/95 BargainingRound',SouthAfrican Labour Bulletin 19(6). 1997. Organizing Downwards. Preliminary findings of a research as personal communication. Johannesburg: NALEDI. Rueschemeyer, D, E H Stephens and J D Stephens. 1992. Capitalist Development and Democracy. Chicago, IL: University of Chicago Press. Schreiner, G. 1994. 'Restructuring the Labour Movement after Apartheid', South African Labour Bulletin 18(3). September Commission - COSATU. 1997. Report of the September Commission on the Future of the Trade Unions. Johannesburg: COSATU. South Africa Foundation. 1996. Growth for All. Johannesburg: South Africa Foundation. Standing, G. 1997. 'Globalization, Labour Flexibility and Insecurity: The Era of Market Regulation', European Journal of Industrial Relations 3(1). Standing, G, J Sender and J Weeks. 1996. Restructuring the Labour Market: The South African Challenge. Geneva: International Labour Office. Steinberg, J and G Seidman. 1995. 'Goldmining's Labour Markets: Legacies of the Past, Challenges of the Present'. Labour Studies Research Reports 6, Sociology of Work Unit, Johannesburg: University of the Witwatersrand. Theron, J. 1997. 'On Homeworkers'. Institute of Development and Labour Law Occasional Papers, 1/96. Cape Town: University of Cape Town. 47 Franco Barchiesi Van Driel, M. 1998. 'Local Government Privatization through the Back Door', South African Labour Bulletin 22(3). Visser, J. 1994. 'Mutamenti sociali ed organizzativi del sindacato nelle democrazie avanzate', Giornale di diritto del lavoro e di relazioni industriali 16(2). Von Holdt, K. 1991. "The COSATU/ANC Alliance: What does COSATU Think?' In: South African Labour Bulletin 15(8). 1993. 'Impressive Gains, Organisational Crisis', South African Labour Bulletin 17(6). 1995a. 'From Politics of Resistance to Politics of Reconstruction? The Union and 'Ungovernability' in the Workplace'. Paper presented to the Second International Conference on Emerging Union Structures Stockholm. 1995b. 'The Auto Sector. NUMSA Signs First "Post- Apartheid" Agreement', South African Labour Bulletin 19(4). 1995c. 'The LRA Agreement: "Worker Victory" or "Miserable Compromise"?' South African Labour Bulletin 19(4). Webster, E. 1988. "The Rise of Social-Movement Unionism: The Two Faces of the Black Trade Union Movement in South Africa'. In: P Frankel, N Pines and M Swilling (eds), State, Resistance and Change in South Africa. London: Croom Helm. 1996. 'Defusing the Molotov Cocktail in South African Industrial Relations: The Burden of the Past and the Challenge of the Future'. Mimeo, Sociology of Work Unit, Johannesburg: University of the Witwatersrand. Webster, E, I Macun and T Rosenthal. 1997. 'Actually Existing Co-determination ': Case Studies of South African Enterprises. Mimeo, Sociology of Work Unit, Johannesburg: University of the Witwatersrand. Weyland, K. 1996. 'Neopopulism and Neoliberalism in Latin America: Unexpected Affinities', Studies in Comparative International Development 31(3). 48