FRIEDMAN AND REITZES ARTICLE DEMOCRATISATION OR BUREAUCRATISATION?: CIVIL SOCIETY, THE PUBLIC SPHERE AND THE STATE IN POST-APARTHEID SOUTH AFRICA1 Steven Friedman and Marine Reitzes The Global Revival of Civil Society Civil society, once of interest only to political theorists, is enjoying a global revival. The collapse of Stalinist states; the demise of the one-party state in Africa; and the overburdened welfare states of Western Europe have prompted disillusionment with the state and the rise of theoretical anti-statism. In the East, West and South, varieties of the state have failed to meet expectations. The welfare state is increasingly unable to deliver and is seen as a source of bureaucracy, not emancipation (Keane, 1998:1-30). The growth of anti-democratic statist structures, and the Weberian nightmare of an ever-expanding bureaucracy, characterises modern societies. In the West, liberal democracy is seen by many as an illusion: citizens have little, if any, say in decisions. In the East, identification of 'the people's party' and monolithic state with the assumed homogeneous 'will of the people' has been recognised as spurious, and the notion of the morally regulative state discredited. On the right, these developments prompted a renewed theoretical assault on the post-war consensus in liberal democracies in support of the welfare state. On the left, grassroots social movements in civil society came to be seen as a more plausible route to popular empowerment than the state. It is in the light of these dynamics that citizens and theorists have come to demonise the state and deify civil society. Civil Society in South Africa In South Africa too, the idea of civil society fired the imagination of activists and commentators and has come to mean all things to all people, different things to different people. TRANSFORMATION 29 (1996) 55 ARTICLE FRIEDMAN AND REITZES Here, ideas were influenced by circumstances peculiar to a transition from authoritarian rule.2 Since the majority was denied representation in the state, resistance organisations were forced to mobilise against the state. As opportunities for legal organisation opened in the late-1970s and early-1980s, this mobilisation was led by a network of civic, youth and other movements which coalesced into the United Democratic Front. The trade union movement, whose renaissance posed the first organised challenge to the white monopoly of power, combined 'routine' workplace activity with anti-apartheid mobilisation. Because these movements mobilised independently of the state and resembled the social movements which kindled left enthusiasm for 'civil society* elsewhere, they inspired a local variant of left 'civil society' theory. When activists within these movements, or intellectuals sympathetic to them, demanded a greater say or more resources, they did so in the name of 'civil society'. The trend was strengthened by the fact that apartheid established subordinate institutions which claimed to meet black demands for representation, on white terms and within parameters set by an automatic white majority in the central legislature. The (successful) attempt to deprive these institutions of legitimacy became a prime focus of resistance activity; this implied an emphasis on showing that 'elected' authorities were less representative than those formed in 'civil society' by 'the people'.3 For some, the defective representativeness of apartheid local government came to illustrate the deficiencies of all local government: it became common to stress that it was 'not enough' to extend the franchise to all - it was necessary also to 'empower the people' and their vehicle, 'civil society'. Two consequences flowed from this. The first is that the demand for the democratisation of the state was not restricted to a call for universal franchise; democracy was held to be incomplete unless 'civil society' was assured a share in decisions. This implied that it needed to be incorporated within the state. The second was that civil society, conceived of as those associations which participated in 'the struggle', was held to possess a capacity to change society which states, even representative ones, lack. Civil society is thus burdened with the expectation that it can provide a panacea for many ills. It was these concerns which prompted the drafting of the RDP, which began as an attempt by the Congress of South African Trade Unions to ensure that a post-election government did not ignore the concerns of organised labour. This concern was soon adopted by social movements within the alliance who saw the RDP as a means of binding a universal franchise government to redistributive social programmes and a role for 'civil society' in the new order. While the RDP 'base document' which emerged was the result of elaborate negotiation within 56 TRANSFORMATION 29 (1996) FRIEDMAN AND REITZES ARTICLE the ANC alliance, it does contain plentiful references to the 'empowerment' of civil society: Democracy for ordinary citizens must not end with formal rights and periodic... elections... Without undermining the authority and responsibilities of elected representative bodies ... the democratic order we envisage must foster a wide range of institutions of participatory democracy in partnership with civil society ... and facilitate direct democracy ... (ANC, 1994:120-1). The document thus proposes 'sectoral forums... people's forums, referenda... and other consultation processes' (1994:120-1)). This not only demonstrates the concerns noted here, but appears to commit the majority party to supplement democratic institutions with civil society forums. Post-Election Civil Society The election of a majority government has disturbed this apparent consensus between 'civil society' and the majority party. But it has changed less than it seems to have done. At first glance, the tensions are acute. In the months after the election, cracks in the alliance between the ANC and COSATU were the subject of much reportage. Tensions between SANCO and the ANC have also been marked. It has become a cliche to note that alliances which held against a common enemy fragment once that foe is no more. This could mean that the ANC now finds enthusiasm for civil society dispensable. But these analyses are simplistic. Firstly, the tensions predate the election: this is shown by COSATU's difference with the ANC over the strike right (Atkinson, 1994:140-1) and in tensions between the two over unions' role in post-election decision-making.4 This latter tension illustrated a wider ANC concern to limit the role of civil society forums in policy-making.5 Despite these differences, the ANC and its 'civil society' partners remained allies during the 1994 and 1995 election campaigns. Secondly, the divide between ANC politicians insisting on the primacy of elected government and activists seeking to subordinate it to 'civil society' forums has narrowed as the transition has progressed. In 1990, some activists still insisted there was no need for local elections because civics were a democratic form of local government6 - or that local councils ought to submit all decisions to 'people's assemblies'7 As these arguments were submitted to scrutiny and debate, and as the prospect of power in formal institutions neared, they were modified. The final RDP draft, with its caveat that representative institutions not be undermined, shows this. TRANSFORMATION 29 (1996) 57 ARTICLE FRIEDMAN AND REITZES Thirdly, enthusiasm for incorporating 'civil society' in formal decision-making institutions has not disappeared as the ANC has entered government. The clearest expression is Nedlac, three of whose chambers comprise representatives of government and the two most organised private interests, business and labour: the fourth includes 'community organisations'. While Nedlac is formally an advisory body, it does have statutory status and the presence of the government opens the prospect that binding agreements will be concluded which will be enacted into law. Nedlac seems likely, therefore, to become one channel through which the RDP document's promise of civil society's 'empowerment' will become realised. Nor are plans for 'civil society' incorporation restricted to Nedlac: the post-election government was also concerned to promote local development forums (LDFs). A document prepared by the Gauteng RDP Commission suggested that LDFs become representative structures: member organisations would have to prove 'a certain membership', 'provide a list of activities' and submit proof of 'continuous representation' every six months. The Commission also hoped to encourage sub-regional forums which would aim to provide a 'link between provincial and local government'. Commission documents see these forums as co-ordinators of development; it says they will link with metropolitan government where possible (unpublished document, Gauteng RDP Commission). Local RDP committees and similar institutions to deal with water supply have also been established. The impression that a deep chasm has opened between the ANC in government and its erstwhile partners in civil society is, therefore, oversimplified. Differences have emerged which were partly hidden by the anti-apartheid alliance. These may grow and might ensure for 'civil society' a less significant role in government than its advocates hope. For the present, however, Nedlac's composition and the search for forums to drive or influence development suggests that the tensions are about the extent and form of 'civil society' participation in government decisions, not the principle. This begs a need to examine this trend and to inquire into its effects on democracy. Are we moving towards an enrichment of representative institutions or their subversion? And is an interest in incorporating 'civil society' in development decisions likely to increase the prospect of democratic and effective development or abridge it? Civil Society: why and who? These questions cannot be answered without some reflection on the nature and purpose of civil society. 58 TRANSFORMATION 29 (1996) FRIEDMAN AND REITZES ARTICLE We can regard civil society as 'a public realm of private individual association' (Reitzes, 1994:100) or of 'organisations that are autonomous from the state but interrelate with it... that interact with the state but don't want to take it over' (Chazan, 1993:14). Its existence is necessary to democracy since it provides a vehicle for citizens' participation in public life and a check on the exercise of state power: one of its prime purposes is to 'civilise' the democratic state. But civil society also buttresses that state by binding citizens to the rules of democratic politics, so 'civilising' private associational life. Precisely because they are concerned to hold the state to account but not take it over, civil society institutions integrate citizens into the norms of democratic life: an antagonistic relation of civil society, or its actors, to the economy or state arises only when ... the institutions of economic and political society serve to insulate decision making... from the influence of social organisations, initiatives, and forms of public discussion (Cohen and Arato, 1992:x-xi). This calls into question the identification of civics and other sections of the 'liberation' movement with 'civil society' since these movements were indeed part of a bloc which sought to take over the state. Their formal incorporation in state decision-making, therefore, raises a host of problems. The first is that, if they are best viewed as adjuncts of the liberation alliance, incorporating them into government decision-making is not to include civil society - it is to give the election winner two bites at the representational cherry: through public election and official selection. At best, this would simply formalise a change of power in which one set of interests with the ear of the governing party is replaced by another: at worst, it would insulate government from the full range of interests in society by placing between them and it an artificially selected 'civil society'. The dividing line between some elements of 'civil society' and the state with which they wish to interact has become exceeding thin. Many leaders of the social movements have been absorbed into the post-apartheid state: others may merely be waiting their turn - SANCO's ex-president found himself demanding a say on behalf of civil society one day, joining the state as an MP the next (Business Day, Jan 25,1995). SANCO sees itself as a watchdog on local government while arguing with the ANC about how many of its activists are to become members of the government over which it wants to watch (Sowetan, Jan 26, 1995). Argument over the roles of elected representatives and civil society may really be about how the spoils of the acquisition of state power are to be divided. Another is that, if civil society is indeed to be civil, some preconditions must be met: an inclusive constitution; inclusive legal citizenship; a culture of rights TRANSFORMATION 29 (1996) 59 ARTICLE FRIEDMAN AND REITZES and duties; representative democracy; political tolerance; legal equality of all individuals; and a legitimate government and state. Under apartheid, none of these existed, and so many institutions which were identified as 'organs of civil society' were misidentified. Their purpose was not to participate in a democratic polity, but to fight against an undemocratic one. In the view of some critics, the consequences were neither civil nor democratic: ... ANC-SACP approval was given to those who flew its flag, 'enemies of the people' were targeted, and 'unity' was turned into a demand for political conformity ... the central problem was that the unity of the 'people' tended to be conceived in terms of an abstract and monolithic 'general will', discounting the actual and divergent empirical wills of its constituent members. ... (Fine, 1992:71-83). In other words, the dictates of 'struggle' in a polarised society prompted the resistance movement, of which 'civil society' social movements were part, to claim and demand a uniformity which did not exist and which contradicts the very notion of civil society, one of whose premises is diversity. Civics, youth congresses and the like became not the vehicles of some people (of even very many people) but of the people. This explains the tendency among resistance organisations to equate the social movements allied to them with the whole of civil society, not a part of it. In reality, they are only a part. While the apartheid state may not have provided civil society with the preconditions to become 'civil', it did not destroy all independent associational life and so there existed organisations which were independent of it and which might have reason to interact with it. Among the disenfranchised, Narsoo makes a helpful distinction between organisations of 'survival' and of 'resistance'. The former 'were the burial clubs, stokvels ... hawkers associations, and even football clubs. Their basic project was to survive the rigours of apartheid and to provide some sustenance collectively' (Narsoo, 1991:27). To this may be added institutions such as churches, whose membership comprised both the enfranchised and disenfranchised, and associations within 'white' society, from powerful business associations through to special interest groups which sought to influence state policy on specific issues. If plans to incorporate 'civil society' in state decision-making are to exclude these organisations, we are again likely to see not an 'empowerment' of civil society, but the exclusion of large parts of it from political institutions. Finally, the genesis of 'civil society' in the context of a struggle for hegemony between two contending blocs forces a re-examination of the vehicles whereby 'civil society' was incorporated into state decision-making during the transition 60 TRANSFORMATION 29 (1996) FRIEDMAN AND REITZES ARTICLE period and which have, according to the RDP document, become models for their post-apartheid incorporation: forums. These multi-interest negotiation vehicles varied in their inclusiveness and representativeness. But, with some exceptions, it is questionable whether they were primarily intended as vehicles for interaction between private interests and the state. They arose at a time when the liberation movement was concerned to prevent an undemocratic state, which was a party to the negotiation process, from unilateral decision-making designed to give it an unfair advantage, either by introducing changes for which it could claim credit or by creating realities which an elected government would be unable to undo; they were, therefore, not only creatures of the transition, but products of the contest for state power (Shubane and Shaw, 1993). This explains why they existed as much, if not more, to prevent state decisions as to influence them. It also begs a need to consider whether institutions which aimed to serve one purpose under particular conditions can serve another under different ones. This raises another question which has been implicit in this discussion but which needs now to be made explicit: the extent to which the election of a representative government has altered the parameters of the civil society debate. Ignoring Democracy?: the political sphere The establishment of a constitutional democracy has created some of the necessary conditions for the emergence of a strong and 'civilising' civil society. To be sure, only the necessary conditions, not the sufficient ones, exist: much work must still be done to realise the Utopia of civil society. But these conditions cannot be legislated. At best, the formal requirements for the possibility of their existence can be legislated and, to a certain extent, materially provided. The question is how? The RDP document is correct to insist that the holding of elections does not ensure democracy. Party support does not necessarily indicate preference for particular policies. Any party wishing to govern in a democracy is obliged to present to voters a policy package: those who support it will do so because they endorse most, but not all, those policies - witness evidence that most ANC voters disagree with its policy on capital punishment. And a mandate conferred every five years does not give representatives unlimited latitude since conditions, and public attitudes, may change in this period. These points argue for a continued role by citizens in politics between elections. But representative democracy offers advantages which civil society does not. The most obvious and important is that representativeness and accountability to the citizenry is a precondition for participation in government, but not in civil TRANSFORMATION 29 (1996) 61 ARTICLE FRIEDMAN AND REITZES society. This is not to contradict the preceding paragraph. It is to point out that, however conditional their mandate, public representatives can only acquire that status if they win voter support and that they will lose it if, at the next election, they lose that support. Precisely because civil society is a realm of voluntary association and diversity, its constituents do not have to be representative of, or accountable to, the citizenry. A crucial element now exists which was absent under apartheid and which has not been fully recognised by some advocates of a formalised role for civil society in government: state decisions are now formally taken by elected representatives of political parties which compete for support among the entire electorate. The absence of such a reality under apartheid explains why the relationship between civil society and the state is often conceived of as that between the citizenry and the bureaucracy. This was the case until last April; formal 'political society', the realm in which competitors for power compete for public support, was closed to 80 percent of society. Now it is formally open to all and it is questionable whether democracy can survive, or even be said to exist, if this sphere is not the primary vehicle by which citizens indicate preferences to public representatives. Given that '[t]he political role of civil society ... is the generation of influence through the life of democratic associations and unconstrained discussion in the cultural public sphere' (Cohen and Arato, 1992:ix-x), the target of such influence should not be the state itself, but political society, partially constituted of political parties and parliament: Between civil society and the state there has to be some general form of mediation, for if each ... interest of civil society lobbies the state on behalf of its own private concerns - no matter how justified - then judgement of their claims and determination of priorities between them are left in the hands of one body alone, the state executive. [It] is in principle the representation of the state interest in civil society; the party system is in principle the representation of the private interests of civil society in the state. If the state executive is not to be the sole mediation between state and civil society, then the party system of representation is essential (Fine, 1992). In other words, the incorporation of civil society in the state through forums runs two risks. The first is that it allows those who have not submitted themselves to the test of election to exert as much, if not more, influence as those who have. The second is that it may bypass the political system, if public demands are placed 62 TRANSFORMATION 29 (1996) FRIEDMAN AND REITZES ARTICLE not at the door of parties and their representatives in parliament but at that of state officials who are meant to be subordinate to them. None of this assumes that the party system or parliament is automatically responsive to public opinion. Our current system of election is highly unresponsive since representatives are accountable to party leaders, not the electorate. And, while the current parliament has attempted to turn the formerly clandestine standing committee system into a means of eliciting public responses and of holding bureaucrats and ministers to account, complaints from some committee chairs question whether this system is operating as well in practice as it seems to do in theory.8 But this suggests that the route to more effective participation by all civil society associations in the affairs of government lies not in the establishment of networks of forums but in parliamentary and electoral reform, a topic hardly mentioned by local civil society theorists, but one which so concerns one of their British counterparts, John Keane, that he devoted a chapter of his Democracy and Civil Society to an analysis of 'Dictatorship and the Decline of Parliament' (Keane, 1998:1-30). Civil society's strength depends on that of the democratic state for it is the latter which provides it with the liberties, public order, material conditions and access to public decision-making without which civil society collapses or is subordinated. In the face of a coercive state, institutions of civil society can lose their autonomy, serving as conveyor belts for state policy (Narsoo, 1992:27). Furthermore, the state can use the concept to legitimate its actions and claim support which may not exist. More specifically, our history of polarisation raises the prospect that the post-apartheid state will become a vehicle for former constituents of hegemonic blocs, informed by a totalising and adversarial legacy, unable or unwilling to nurture a diverse, plural, society. In that event, civil society will collapse. If democracy is a necessary condition for a vigorous civil society, it is not only society which must be civilised and democratised, but the state. The route to a stronger civil society may lie in opening the state to the widest possible public influence through reforms which maximise its accessibility to all, rather than to those able to gain admittance to state-sanctioned vehicles for civil society participation. While the possibility of access to state decisions has been opened, the capacity to use it has not. Pluralist theory, which sees formal democracy as a system in which all interests have an equal opportunity to influence public opinion and, therefore, decision-making9 tends to ignore the reality that some citizens have far greater capacity to do this than others. Here, ability to participate is shaped TRANSFORMATION 29 (1996) 63 ARTICLE FRIEDMAN AND REITZES not only by differential access to resources needed to organise, but by a factor as basic as that public debate is not conducted in the languages spoken by most citizens. Leaving democracy to the market-place of ideas and organisations alone may well ensure that the voices which most need to be heard remain stilled. But it is difficult to see how the remedies on offer from the RDP document or government officials appointed to implement it address that problem. However well-meaning the process of selecting participants in public forums and similar vehicles, these are bound to become channels for those who would in any case be heard under conditions of unrestrained pluralism. If the problem is differential capacity to organise, it cannot be solved by finding new representational channels for the organised. Bureaucratising Civil Society: corporatism and its limits To point out that democratic states give all extra-parliamentary groups the framework for incorporation into civil society, enabling them to make claims on the state, does not necessarily mean that these states rely only on the parliamentary system and pluralist rules for dealings with private interests. On the contrary, a look at industrial democracies shows that most establish arrangements which allow strong, organised, private interests guaranteed access to the state. These arrangements, labelled 'democratic corporatism', not only co- exist with democracy but are held by some to be essential to it since they commit organised interests whose consent is vital to democracy's functioning to its outcomes in exchange for the right to negotiate policy. Are not these arrangements the inspiration for the forums proposed by the RDP offices? An answer lies in understanding why corporatism has emerged and who participates in it. Its purpose is not to 'empower civil society' but to formalise the participation of interests who already have power and whose demands the state needs to incorporate if it is to govern effectively. While democratic corporatism is, therefore, invariably initiated by the state (Cawson, 1986) - which alone has the formal power to establish a statutory institution - it does this as a response to existing power relations. This is why it is common for the state to formalise a corporatist arrangement only after strong private interests have begun to do this themselves. Because the chief aim of the exercise is effective government rather than popular empowerment it is essential to successful corporatism that the parties are able to bind their constituents to negotiated agreements. If they are not, the government - and their negotiation partners - might as well leave policy-making to the pluralist market-place. And, for a variety of reasons, international experience shows that producer interests command the power to do this, 64 TRANSFORMATION 29 (1996) FRIEDMAN AND REITZES ARTICLE consumer interests do not (Cawson, 1986). Business organisations, trade unions and professional associations tend to be candidates for corporatism, rather than tenant coalitions, consumer unions or parent associations. It is this dynamic which gave birth to Nedlac. While there is an important and lively debate on the extent to which business or labour in this country is equipped for corporatism - which has, of course, been heightened by events in 1996 indicated that both sides' commitment to negotiated compromise is limited, there is no doubt that both are strong producer interests, with definable constituencies who pay dues, that both are organised and that they have at least the potential to bind crucial constituencies to negotiated agreements. Whatever Nedlac's fate, three of its chambers represent an attempt to introduce a mode of interaction between the state and strong private interests which has a proven ability to co-exist with democratic institutions and to enhance industrial efficiency and equity. The same cannot be said of the decision to establish a development chamber which will include organisations who can show they 'represent a community interest at national level'; have a direct interest in the RDP; and are democratically constituted and able to seek mandates. The members of this chamber are unlikely to be able to bind citizens to decisions - many do not even recruit members.10 It is, therefore, unclear why their inclusion is held to increase the prospect of achieving social consensus on policy. On the contrary, it may weaken the forum's ability to function by adding to the organisations from whom agreement must be sought, but not to those able to bind important parts of the society to agreements. The fourth chamber may be an attempt to strengthen the non-business side of the Nedlac table. But the assumption by some on the left that 'empowering civil society' is a short cut to outcomes which cannot be achieved through the representative system because powerful conservative forces exist to obstruct them seems to ignore the reality that civil society: ... comprises all organisations which are not part of the coercive apparatuses of the state. It is a terrain which includes not only social movements, but also ... capital. Civil society is not, therefore, in itself 'progressive' - it is a terrain of contest between conflicting 'progressive' and conservative forces... Those... who ignore the presence of conservative interests in civil society allow their powerful role to be disguised (Narsoo, 1992:5). More likely, however, given that a persistent criticism of Nedlac is that it excludes marginalised groups, is that the fourth chamber is designed to incorporate these in the corporatist fold. This is flawed for two reasons. The first TRANSFORMATION 29 (1996) 65 ARTICLE FRIEDMAN AND REITZES is that it is by definition implausible that any organisation represents those unable to organise. Second, it seems based on the fallacy that those who are not included in state-engineered institutions are necessarily denied a say. This is to misconceive the nature of civil society and the democratic state, and the relation between the two: civil society exists precisely to give a voice to those who wish the state to take cognisance of their needs. Critics of corporatist institutions tend to forget that they have three parties and that one of these is democratic government which remains subject to the same electoral pressures it faces when there is no corporatism. Those who do not believe that corporatist agreements reflect their constituency's interests remain able to seek to thwart these through pluralist tools, from quiet lobbying of ministers to mass legal demonstrations. There is a strong danger that the incorporation of 'community groups' into Nedlac or other forums will serve not to empower civil society but to bureaucratise it. Community organisations were required to apply to Nedlac and their applications are subject to approval by a committee chaired by a representative of the RDP office (The Star, Dec 28,1995). Since the criteria are subjective (what does 'a community interest at national level' mean?), 'civil society' representatives are, in effect, chosen by a government official. Similarly, the Gauteng RDP Commission's LDF document - which seems to envisage a web of forums leading from the grassroots to the Commission - suggests that the 'government must be in control of [their] launch', and 'the (ANC) alliance ... must be consulted in the establishment of LDFs'. Any proposed LDF project 'should be reported to the RDP Commission and the [National Working Committee]' which 'makes the assessment and ratifies the project' (Gauteng RDP Commission document). Given that the criteria for membership are vague, and it is not clear who would decide whether they are met, this raises the spectre of government officials choosing the forum's membership. These proposals seem to intend to construct civil society from the top down: to co-opt it and determine the nature and extent of its participation. Either 'civil society' participation is being evoked to legitimate state policy, and/or this is an attempt to co-opt selected organs of civil society and to use them as conveyor belts for state decisions. The idea of organs of civil society being managed by the state by definition negates the idea of civil society. Alternatively, the state is attempting to defer its ultimate policy-making responsibility to institutions which it defines and controls within strictly circumscribed limits. These proposals raise the spectre of a co-opted and bureaucratised civil society: far from positing civil society as an alternative to the Weberian nightmare, they TRANSFORMATION 29 (1996) FRIEDMAN AND REITZES ARTICLE raise the possibility of it becoming an arm of the state. Far from democratising society, this may reinforce a tendency already, in the view of some, prevalent in industrial societies: The... exercise and equilibrium of power now takes place directly between the private bureaucracies, special-interest associations, parties, and public administration. The public as such is included only sporadically ... and even then it is brought in only to contribute its acclamation (Calhoun, 1992:22). The point of these criticisms is not simply to highlight defects in particular documents. It is to draw attention to an exercise whose assumptions probably lead it inevitably in the direction described here. Democratic corporatism is usually a response to a need to incorporate into state decision-making groups which have demonstrated their representativeness and their organisation: no committee is required to decide whether COS ATU or Business South Africa are sufficiently representative to merit inclusion in Nedlac. Once the state sets about selecting representatives of civil society, one of two outcomes seems likely. The first is an attempt to avoid the danger of a state-selected civil society by engaging in an elaborate exercise to ensure representativeness and inclusiveness. But why should officials be better than the electorate at determining representativeness? Alternatively, the candidates may be selected not because they are held to be representative, but because they are seen as politically compatible - in which case, the arrangement is likely to resemble not democratic corporatism, but its authoritarian counterpart, in which 'civil society' is co-opted into the state or created by it and then used to endow unrepresentative decisions with a spurious legitimacy (Narsoo, 1991:5). An unintended consequence of the inclusion of 'community organisations' in Nedlac may be to prevent the representation of the really marginalised by assuming that their interests are already represented by those selected to join the council. One further consideration suggests scepticism about the extent to which state initiatives of this sort can 'empower' civil society. Even if civil society groups selected to participate are fully representative, this does not automatically ensure that they possess the capacity to participate in state institutions. Capacity problems have hampered the role of popular organisations in many of the forums which emerged during the transition. This is not an argument for the monopolisation of public policy by technically trained elites: on the contrary, it emphasises the need for elected representatives and interest group leaders to ensure that technical proposals are communicated to their constituencies in ways which allow them to exercise an informed choice as to whether the plans serve their interests. But it does illustrate that TRANSFORMATION 29 (1996) 67 ARTICLE FRIEDMAN AND REITZES incorporation into forums is not an unmixed blessing to those in civil society 'fortunate' enough to be selected: they could find themselves forced to take joint responsibility for decisions they were not fully equipped to take. Similarly, as unions found after their corporatist experiment began, it takes great care to avoid estrangement from their constituency once they are caught up in the pressures and technicalities of forum business. The danger may be particularly acute in development forums, where insistence on consulting a constituency can be portrayed as lack of seriousness about delivery. Organisations in civil society which did represent a constituency when they entered forums may soon cease to do so because they entered them. The reality that incorporation into official structures implies constraints as well as opportunities applies even more forcefully if organisations outside state control are offered a role in implementing official programmes. The concern among some non-governmental organisations that plans to offer them a share of state development funds may subject them to unacceptable controls (Business Day, Aug 23, 1994) reflects this. For highly organised interests with secure power bases, the trade-off may hold more benefits than costs - this is one of the key rationales for democratic corporatism. For the rest, the cost-benefit calculation may point strongly in the opposite direction. However well-intentioned they may be, 'inclusive' and 'representative' forums which stem not from the state's need to formalise dealings with already influential constituencies but from the theories of government planners or the demands of elites claiming a representativeness which is unproven could become a vehicle not for empowering civil society but for shackling it. Some Development Implications A democratic state needs to provide not only a framework for civil and political society, but also basic services without which civil society cannot function. The stated concern of state policy to encourage development among citizens hitherto deprived of it is a crucial potential contribution to the development of civil society. Indeed, since many citizens lack the capacity to participate fully in the associational life likely to impact on government policy, development is arguably a more substantial contribution to a strong civil society than attempts to synthesise it through the creation of official institutions. This is not to say that the elected government should simply set about developing society. This paper has already dealt with the normative objections to such an approach: the 62 percent of the electorate who support the majority party are united in their rejection of minority rule, but they have many competing and conflicting development interests. Even among public representatives 68 TRANSFORMATION 29 (1996) FRIEDMAN AND REITZES ARTICLE elected on the same party ticket, there is no unanimity on development strategy - witness vigorous debate between the ministry of housing and ANC provincial housing MECs on housing policy. There are also severe difficulties in the way of any attempt to develop 'from the top down'. Recent experience has shown that the assumption that 'communities' are united in their preferences may ensure the rejection of development projects, the refusal of recipients to pay for them and violent mobilisation against them (Friedman, 1993). The notion of 'community' is itself questionable, since it describes people who share a residential space but not common interests and preferences. The complex 'community' dynamics which may ensure the success or failure of a development project cannot be discerned by representatives simply because their party won a majority in an election. Local elections may help, not only because they may allow the representation of local interests which cannot be adequately expressed in national government but because they include a degree of direct election by constituents which may ensure that representatives are more accountable to specific sectors of the population. But election is no guarantee of the ability to discern development preferences, both because geography does not necessarily coincide with interest and because some groups most threatened by development initiatives, such as illegal immigrants or criminal gangs, are unlikely to make their presence, let alone their preferences, known to public representatives. In any event, the severe limitations on their autonomy which newly-elected local governments face, expressed chiefly in the demand that they restrict themselves to service delivery or to implementing infrastructural programmes on behalf of national government or the provinces, constrains a strong representational role at this level of government. All this would seem to argue forcefully for the formal incorporation of civil society in development decision-making. But such a strategy will, by definition, include only visible, organised, interests. Not only does this beg the question of how, say, illegal immigrants are to be prevented from mobilising against development by an 'inclusive' forum which will inevitably exclude them; it also does not explain how the preferences of those who are not organised are to be discerned by consulting those who are. And, since the interests included in forums are likely to be those who already have means of voicing preferences, the exercise may be not only redundant but may serve to further insulate decision-makers and development from those at the base of society whose interests most need to be heard. It could be argued that there are visible organisations within 'communities' who, while they may not represent all interests, can derail any initiative to which TRANSFORMATION 29 (1996) ARTICLE FRIEDMAN AND REITZES they are not party and that something akin to a development forum is needed to prevent them doing so: this point was generally, and probably accurately, made about civics during the 1980s and early-1990s. But it is questionable whether there still are such organisations: certainly, the civics' capacity to derail development has probably been sharply impaired not only by the departure of key personnel into government but also by the reality that the authority against which resistance would have to be mobilised was elected by the constituency which activists would have to galvanise into resistance. Even if there are such groups, it is unclear why a forum is needed to consult them. Finally, the ability to derail development is a necessary but not sufficient criterion for inclusion in a quasi-corporatist development forum. As this paper has argued, a sufficient condition is also the ability to bind constituencies to compromises and civics' ability to do this is open to question. The apparently insoluble dilemma is that development may be derailed by unorganised or not visibly organised interests but that these cannot, by definition, be included in decision-making forums. The answer, to the extent that there is one, lies not in replacing forums with some other technique for representing the unrepresented. It lies rather in acknowledging that democracy not only confers on elected representatives a mandate, but also a responsibility. To insist that development strategies may fail if they do not grasp the full range of interests among beneficiaries and seek to gain their consent is not to insist that this problem can be addressed by the establishment of more state institutions. On the contrary, it could be argued that the need to establish elaborate structures for the inclusion of all interests is articulated only when there are no accountable public representatives who risk voter rejection if development plans fail. Since we now have - or will have, if necessary political reforms are made - a polity which allows all with a stake in development to hold to account public representatives, we should now expect those representatives or those they delegate to take responsibility for implementing programmes which do win the voluntary consent of beneficiaries. It is not at all clear why a minister or MEC or mayor should need a forum to understand and to respond to his or her electorate - or to be aware of the existence of groups who may derail that which his or her constituents want. This does not mean that our representative institutions are necessarily equipped to take up this task: a focus group research project found a significant gulf between the development expectations of voters and the expectations their public representatives believe them to have (Charney, 1995). There are also parts of our society in which civil society in the sense that the term has been used here does not exist or does so in very attenuated form. Rural 70 TRANSFORMATION 29 (1996) FRIEDMAN AND REITZES ARTICLE residents who participated in the focus groups were asked who in their area was qualified to express their needs: none mentioned civics or interest associations - the ANC or the chief were usually cited (often after some thought) (Charney 1995). Two points can be made about this. The first is that it illustrates the danger of manufactured consultation with civil society: the responses gleaned by the study suggest that any 'civil society' organisations selected for consultation would not represent the respondents and, most likely, many others like them. The second is that those who identified the ANC as their interlocutor are expressing a confidence in elected representation and in the party system which provides an opportunity and a responsibility to public representatives in their areas. The weakest interests do not need more forums: they need representatives willing and able to understand their interests, to mediate between them where they conflict, and to translate development plans into understandable options. They need also an institution with the capacity to translate their preferences into uniformly applied policy and the state is the only one capable of this. They need also the power to remove representatives who do not accept responsibility for gauging their preferences and implementing policy which enjoys wide consent. The beginning of an answer to our development dilemmas, like the genesis of a vigorous and civilising civil society, lies not in synthesising or subverting formal representative institutions but in invigorating and expanding them. How might this be achieved? We have stressed the need for political reforms, of which the most important is the need for an electoral system which ensures that elected representatives are directly accountable to voters: an opportunity to introduce this for the 1999 elections was missed during the 1996 constitutional process. As indicated above, this is not simply a concern of liberal intellectuals worried about the neatness of the parliamentary system: it concerns all sections of society and should be as much a concern of the left as of liberals - witness repeated complaints by COSATU that its members cannot hold former unionist MPs to account. A strengthening of the parliamentary committee system, at provincial as well as national level, would also create enhanced opportunities for citizen participation in decisions. So too would a rethinking of the role of local government, one which recognised that its strength lies in its representational function, not in its ability to produce elected service deliverers. All these measures, however, do not guarantee effective representation. They merely create contexts in which it is more likely. Ultimately, as we have implied, much of the answer lies in the hands of political representatives and, more specifically, in their ability to rethink their role in such a way that they begin to take the representation of citizens, particularly grassroots ones, seriously. To some, this may sound alarmingly voluntaristic. But democratic rules can only TRANSFORMATION 29 (1996) 71 ARTICLE FRIEDMAN AND REITZES create the potential for effective representation. Whether the potential is realised depends on the willingness of representatives to be effective - and, of course, the ability of citizens to make them so. NOTES 1. An earlier version of this paper was commissioned by the Development Bank of Southern Africa. The assistance of Mark Shaw and Khehla Shubane is gratefully acknowledged. 2. The role of civil society in transitions is discussed in Schmitter, O'Donnel, Whitehead, 1986. 3. This argument is developed in Steven Friedman's contribution to the forthcoming Encyclopedia of Democracy, to be published by Congressional Quarterly, Washington, DC. 4. See, for example, speech by then ANC economic policy head Trevor Manuel to South African Clothing and Textile Workers Union conference. Business Day. June 21, 1995. 5. See Manuel's comments and then ANC education head John Samuel's remarks on the National Education and Training Forum, Business Day, Jan 28, 1994. 6. These views were, for example, expressed by civic activists at a Centre for Policy Studies conference on the future of the cities, held in 1990. 7. A resolution supporting these assemblies was adopted at an ANC local government conference in October, 1990. For an explanation of the proposal see Botha, 1992. 8. See comments by Max Sisulu, chair of parliament's RDP committee. Business Day, Nov 16, 1994. 9. For the classic statement of this position, see Dahl, 1961. 10. Calls by SANCO on township residents to resume payments were largely ignored: in November, 1994, its then PWV region resolved to collect money to meet the shortfall caused by non-payment, a tacit acknowledgement that it could not persuade non-payers to pay. Sowetan, Nov 2, 1994. REFERENCES African National Congress (1994) The Reconstruction and Development Programme : a policy framework, Johannesburg: ANC. Atkinson, Doreen (1994) 'Insuring the future?: the bill of rights', in S Friedman and D Atkinson (eds). The Small Miracle: South Africa's negotiated settlement, Johannesburg: Ravan Press. Botha, Thozamile (1992) 'Civic associations as autonomous organs of grassroots participation', Theoria, 79. Cathoun, Craig (1992) 'Introduction : Habermas and the Public Sphere', in C Calhoun (ed), Habermas and the Public Sphere, Cambridge: MIT Press. Cawson, Alan (1986) Corporatism and Political Theory, London: Basil Blackwell. Chamey, Craig (1995) Voices of a New Democracy: African expectations in the new South Africa, Johannesburg: Centre for Policy Studies. 72 TRANSFORMATIONS (1906) , FRIEDMAN AND REITZES ARTICLE Chazan, Naomi (1993) Contribution to discussion on 'Govemability and Compliance during the Transition', in R de Villiers (ed), Govemability During the Transition, Johannesburg: Centre for Policy Studies. Cohen, Jean and Andrew Arato (1992) Civil Society and Political Theory, Cambridge: MIT Press. Dahl, Robert (1961) Who Governs?, New Haven: Yale University Press. Fine, Robert (1992) "Civil society theory and the politics of transition in South Africa', Review of African Political Economy, 55. Friedman, Steven (forthcoming)'?', in Encyclopedia of Democracy. Washington DC: Congressional Quarterly. Friedman, Steven (1993) The Elusive 'Community': the Dynamics of Negotiated Urban Development, Johannesburg: Centre for Policy Studies. Gauteng RDP Commission (1995) Unpublished document. Johannesburg. Keane, John (1988) Democracy and Civil Society, London and New York: Verso. Narsoo, Monty (1991) 'Civil Society : a contested terrain', Work in Progress, 76. Narsoo, Monty (1992) Doing What Comes Naturally : A Development Role for the Civic Movement, Johannesburg: Centre for Policy Studies. Reitzes, Maxine (1994) 'Civil Society, the public sphere and the state: reflections on classical and contemporary discourses', Theoria 83/84. Schmitter, Phillipe, Guillermo O'Donnel, and Lawrence Whitehead, (1986) Transitions from Authoritarian Rule, Baltimore: John Hopkins. TRANSFORMATION 29 (1996) 73