It has become conventional wisdom on both the left and right that South Africa’s "political miracle" was purchased at the price of leaving the economic pillars of apartheid capitalism intact. On the right it has been argued that South Africa has accepted the inevitability of the market and that it has pragmatically adapted to the new international economic order: it has removed exchange controls, embraced an export-oriented development strategy, practiced strict fiscal discipline, and committed itself to privatisation. (Johnson, 1995) A number of left analysts fundamentally agree with this position, arguing that the global economy is hegemonic and that there is no alternative to its prescriptions of neoliberal economic and social policies outside of an unspecified intensification of conflict. (Lehulere, 1996; Barchiesi, 1997; Amaral, 1997) The Australian political commentator John Pilger, captures the sense of disillusionment on the left when he argued recently that a "historic compromise" between the ANC and the apartheid government left economic power in the hands of the corporate white elite; all that has changed is "the inclusion of a small group of blacks into this masonry." (Pilger, 1998:602)

Both sets of arguments curiously fit comfortably within a conventional political science elite analysis of transitions. In much of this literature it is argued that a transition from authoritarianism to democracy must involve an economic pact between elites on both sides that leaves the fundamental structures of capitalist society in place, thereby ensuring the loyalty of the propertied classes while producing a conservative outcome. Globalisation apparently renders such a conservative pact all the more essential not only because it constrains options for growth, but it also provides domestic capital with an exit option. However, there are persuasive reasons to believe that a conservative pact cannot deliver growth, and therefore cannot establish a fundamental condition for the consolidation of democracy. Yet we are sceptical of the contrary argument that economic crisis will in the long run be favourable for democratic forces.

In recognising the existence of a "double transition"- states that are simultaneously consolidating democracy and liberalising their economies - we accept that these two processes are potentially contradictory and may simultaneously undermine both democracy and economic growth. However we argue that they are potentially reconcilable, though on terms neither the advocates of a conservative pact nor their left critics would recognise. We argue that a class compromise oriented towards growth and redistribution can better provide the economic foundations for sustainable democracy. Such a compromise is not only desirable, but may be possible even in societies with limited economic resources.

Arguments in favour of the inevitability of a conservative pact exaggerate the power of globalisation, by presenting it as an omnipotent that deprives states of manoeuvring room in their choice of development policy. The arguments for a conservative pact further ignore constraints on capital and assume a frictionless mobility. In this such arguments make the mistake of equating all capital with financial capital, overlooking the fact that many firms are highly immobile and that employers may not only be profit-maximisers, but also risk-avoiders interested in steady, long-term, and sustainable profit. A number of analysts, most notably Albert (1992) identify these features, among others, as distinguishing Germany - the "other capitalism" - from the dominant American model. (See also Streeck, 1992; Roemer, 1996, 1995)

Above all, arguments for a conservative pact ignore the power of movements and organisations, the importance of institutions, and the strategic choices that collective actors can make. (Smith, 1998) Movements resisting economic liberalisation may be able to create new institutions through which they can process their demands. In turn these institutions may be able to create new rules of the game allowing what we describe as "bargained liberalisation" to occur, in which prominent groups renegotiate the terms on which a country engages with the global economy. These institutions may expand the universe of choices open to policy makers, while constraining the state and capital from options they would otherwise have chosen.

In essence, the impact of markets on domestic economies is not uniform, but is mediated through specific national configurations of organisations and institutions. (Cox, 1996) While markets constrain the choices that governments make and limit the options unions pursue, they do not predetermine outcomes. (Smith, 1998) In this paper we examine the institutions that have been constructed by workers, employers, and the state during the transition to democracy in South Africa, the choices they have made in developing economic policy, and the possibilities that may be open for a class compromise. The South African experience thus has implications that go beyond the case study and has relevance to other countries attempting to consolidate democracy under conditions of economic liberalisation.