ARTICLE TRANSFORMATION 11 (1990) THE CONDITIONS FOR DEMOCRACY IN SOUTH AFRICA AW Stadler It is not inconceivable that the formal institutions of democracy, with universal adult suffrage in a unitary state at the centre, will be introduced in South Africa in the foreseeable future. It is also not beyond the bounds of probability that these formal institutions will be overtaken by a coup d'etat, or a one-party dictatorship which will subordinate or suppress opposition and dissent. This paper investigates the most propitious conditions for a stable future democracy in South Africa. Like other similar undertakings, the paper is primarily based on comparative studies rather than on South African politi- cal, economic or social material. The main conditions which will be investi- gated are social, or socio-economic; political; and institutional. The political conditions will be discussed in two intimately related contexts: the one focused on policy, and the other on political power. But it will also become clear that it disputes the assumption made in the earlier literature that it is possible neatly to separate political conditions from social ones, or to assume that the chains of causality can be arranged in a linear fashion. The paper hopefully contributes to political debates going on in this country. It therefore has a political purpose. It also makes the assumption that academic debate may contribute meaningfully to desirable political ends. The paper was written before the events of February, 1990. A postscript was added to consider briefly whether the De Klerk initiatives contribute meaningfully to the installation of a democracy. However, the paper does not, except in a superficial and piecemeal way, consider the extent to which South African conditions provide evidence of emergent properties condu- cive to democratic stability, or otherwise. Each of the conditions is the subject of great controversy in the literature, and much of the paper will be concerned to argue positions within the controversies. It ought not to be forgotten either that the very notion of democracy is itself a matter of controversy and polemic, and that the position which writers adopt in response to the substantive problem hinges on their understanding of, and attitude towards, democracy. The socio-economic conditions: wealth or equality? The orthodox position on this issue in political sociology was long domi- nated by SM Lipset, who made two important claims which have commanded strong support in non-Marxist political sociology. The first was that democ- 40 TRANSFORMATION STADLER racy was related to the state of economic development. The more well-to-do a nation, the greater the chances that it will sustain democracy ... only in a wealthy society in which relatively few citizens lived at the level of real poverty could there be a situation in which the mass of the population intel- ligently participate in politics and develop the self-restraint necessary to avoid succumbing to the appeals of irresponsible demagogues (Lipset, 1959:49). The second claim he made was a corollary of the first: much of the book elaborates the idea that the working classes were particularly vulnerable to authoritarian and extremist political movements. In sum, there are two biases in the argument. One privileges economic growth; the other extols limited or constrained participation, particularly working class participation, in poli- tics. The interest in these arguments for a student of democratisation in South Africa is that both are current in contemporary reformist debates. The thesis that economic growth provides the basis for democracy was developed in the well-known O'Dowd thesis which predicted the democratisation of South Africa in the wake of industrial growth, although O'Dowd modelled his thesis on WW Rostow"s 77ie Stages of Growth, not Lipset's work. It was thus more directly than Lipset's work based on a linear, determinist, stage-based conception of historical process. One of the implications of this argument, which was not foreseen, or not discussed, in the original formulations, was that conditions of economic stagnation or decline, such as set in from the early 1970s, might be unpropitious for the installation of democratic institu- tions. Embedded in the Lipset thesis about the economic bases of democracy were a number of assumptions about the meaning of democracy. It rested explicitly on the 'equilibrium' model developed by Joseph Schumpeter who defined democracy as competition for power between elites, which usurped and obliterated the older meanings held equally by its opponents and its advocates. These older meanings had embedded the concept of democracy in the struggles of the poor and the marginalised for political expression and power. They are virtually excluded in the Schumpeterian conception of democracy. So too are those conceptions of democracy as a process of political learning achieved through participation. These issues will be con- sidered in the discussion on political institutions. Like the thesis prescribing growth as the condition for democracy, the argument in favour of a notion of democracy as requiring limited participa- tion in government (and a limited state too) has resurfaced in the reformist debate in South Africa, in various forms, including monetarism, federalism, 41 STADLER TRANSFORMATION and in notions of 'divided sovereignty', etc. Jan Steyn, the former chairman of the Urban Foundation crystallised these ideas in an elegant combination of the Lipset thesis and the theory of overloaded government which became current during the 1970s: Some social scientists have observed that democracy is often most stable when most of the basic problems of the society have been solved and the population at large has been able to reduce its political fervour. In developing situations, including our own, our most basic problems of material inequality and de- privation, under the most favourable circumstances, will take many decades to alleviate to the extent that they no longer generate political passion. In societies with large masses of undifferentiated need, the popular demands on government are immediate and powerful ... and very easily exploitable by democratic opposition to the point that no government can enjoy the security to pursue its longer-term priorities. Prompted by the many acute needs of the masses, a competitive democratic opposition can make facile promises, and thus create expectations that no government can meet.1 Within a few years of the appearance of Lipset's book, the publication of Barrington Moore's (1966) work cast doubt over the major assumption that there was a direct and uniform relationship between socio-economic condi- tions and political democracy, or indeed any specific form of government. Moore's case studies in 'historical sociology' provided the basis for the argument that there were three, and not one, routes to modernity, and that only one of them led to democracy. Yet in casting doubt on the validity of the Lipset thesis in one respect, Moore confirmed it in another. The major condition for democracy, he argued, was the prominence, or even predominance, of the bourgeoisie in the array of social forces seeking political change during the period preced- ing democratisation. Without this condition, the probable result would be a failure to modernise, or modernisation under the auspices of an aristocracy leading to fascism, or alternatively a communist revolution. This position has in turn been challenged and modified. John Stephens (1989) has shown that because Moore's study stopped too early, it under- played the role of the working classes in the achievement of democracy and overstated the role of the bourgeoisie. Stephens, like Therborn, whose influence he acknowledges, argues that an organised working class was the most important force in establishing democracy. Support for this argument comes from studies of democratisation in Latin America (Valuenzela, 1989). Stephens argued, however, that the working class needed allies from other 42 TRANSFORMATION STADLER classes in order to achieve the final push towards democracy. But, in contra- diction of Moore, he shows that these allies were to be found among the peasantry and the urban middle class, and not the bourgeoisie. Except in England and France, the bourgeoisie was a centre of resistance to the political incorporation of the working class. The democracy which the bourgoisie supported was democracy for the propertied. Stephens adds an issue which confirms one argument in Moore's study, but in a way which contributes a refinement which is absent from Moore's book: Stephens accepts that the bourgeoisie contributes towards the establishment of par- liamentary government, but insists that it is the working class which demo- cratises that form of government. A second way in which a refinement of the Moore thesis contributes to the understanding of the conditions of democracy lay in his argument that the major precondition was the need for a violent break with the past. But it is important to add that this was not elaborated in any systematic way, nor was it linked specifically to the conditions for democracy. Rather it provided one of the conditions for modernisation. Therefore Moore's work, though im- portant in disposing of the consensualist bias which characterised the Lipset thesis, did not clarify the political conditions fully, nor relate these to the social and economic ones. Another implication of Moore's study, and the criticism of it by Stephens, is that it raised the issue of the political conditions for modernisation, and specifically for economic growth, as well as the political conditions propi- tious for democracy. Although fuller discussion of this issue must be post- poned, there is at least one problem which needs airing now: the conditions most propitious for the formal establishment of democratic institutions were not necessarily the same conditions for the maintenance of stable democ- racies. Goran Therborn (1977) made an incisive contribution towards the devel- opment of this position in a paper which tilted the balance of plausibility even further away from the Lipset thesis. Therborn argued that far from democ- racy being established during periods of economic growth and social con- sensus, the most propitious periods for the establishment of democracy were periods of political and socio-economic upheaval and dislocation. The 'strik- ing absence in the history of bourgeois democracy is that of a steady, peaceful process accompanying the development of wealth, literacy and urbanisation'. He stressed the importance of defeat in war, and the attendant social upheaval, as major conditions for the installation of democratic govern- ments. Of eighteen cases which he cites, six (or seven) democracies had their origins in military defeat, and, in another eight eases, war was 'causally decisive' in installing democracies. Democracy, he argued, was a 'martial 43 STADLER TRANSFORMATION accomplishment'. Therborn's conclusion is interesting for students of the South African reform movement of the 1980s. It reinforces the case that the major pressures came from 'below* and 'without' rather than as an autono- mous process initiated within the political class. But Therborn did not pursue the issue of the stability or otherwise of democracies once they had been established. This issue was raised by Edward Muller (1988) in an important comparative study which confronts the Lipset thesis directly on its own terrain, that of the general socio-economic conditions conducive to democracy. Muller argued that the most important socio-economic condition for stable democracy was not growth but equality, and conversely, the conditions least favourable was inequality. The main lines of his argument may be summed up as follows. Democratic government produces egalitarian effects, but these effects are gradual rather than rapid: the egalitarian influence of democracy is a gradual process, resulting from a country's accumulated years of democratic experience, rather than a relatively immediate effect of the level of democracy in a given year. He argued that on the average approximately 20 years of democratic influence were required for these egalitarian effects to occur; 'a second generation of democratic experience enhances this egalitarian effect'. In the present context, one of the most important of Muller's findings was that countries with 'less than a generation of democracy are no less inegalitarian, on the average, than nondemocracies...' It is these countries which display the features which he describes as 'unstable democracies': those in which the formal institutions of democracy are vulnerable to being replaced by auth- oritarian regimes. Muller's study confirmed precisely the broad thesis which Robert Dahl (1971) asserted: In a society that already has a regime with public contestation, extreme inequalities increase the chances that competitive politics will be displaced by a hegemony. Polyarchies are par- ticularly vulnerable to the effects of extreme inequalities. Ex- treme inequalities in the distribution of key values are unfavourable to competitive politics and to polyarchy because this state of affairs... is equivalent to extreme inequality in the distribution of key political resources and is likely to generate resentments and frustrations which weaken allegiance to the regime. Muller's (1988) statistical analysis confirmed this exactly: 'AH democracies with high income inequality ... were unstable. These very inegalitarian democracies were highly susceptible to military coups, which were respon- TRANSFORMATION STADLER sible for four of the six instances of instability*. Muller's broader conclusion about political strategies is also worth quoting at length: If a democratic regime is inaugurated in a country with an extremely inegalitarian distribution of income, high inequality is likely to undermine the legitimacy of the regime and cause democratic institutions to be replaced by authoritarian rule. ...It seems the only way to break out of this vicious circle is for a strong political party with redistributive goals to develop during the early years of democracy and then hold office for a sufficiently long period of time to implement policies that significantly reduce income inequality. The political conditions: strong party and unions An earlier paper by Christopher Hewitt (1977) was more explicit about the kind of political party, and the sort of strategies it would require to reduce inequality as rapidly as possible. He argued that political democracy was not a sufficient condition for the achievement of a more equal society. The crucial matter is what the mass electorate does with the franchise and other democratic procedures. Only if the lower classes use their votes to elect socialist governments will democracy result in more equality, since non-socialist govern- ments will not be concerned with redistribution and social equality. The conclusion that equality and not wealth was the crucial socio-economic condition for stable democracy ought not to obscure the fact that growth in a relatively poor country like South Africa will in fact remain an important priority in the future. The fact that the case for economic growth is often argued with an anti-democratic purpose in mind, or without consideration of the social or political costs or consequences, ought not to minimise its importance for a democratic society. Radicals are sometimes understandably dismissive of the emphasis conser- vatives place on the need for growth. But the case to be met is not the need for growth (at least in a poor country), but the claim, sometimes implied in conservative arguments, that growth is incompatible with strong democratic action and control over policy-making institutions. Specifically conservatives often argue that growth is inimical with the presence of strong unions and with policies of public welfare. Unions push for wage increases, thus inhibit- ing investment; welfare services consume resources which might otherwise contribute to growth. A study by Peter Lange and Geoffrey Garrett (1985) showed that these arguments were not necessarily true. They also emphasised the importance 45 STADLER TRANSFORMATION of political structures in providing conditions for growth. The impact of 'domestic political-economic structures' were just as important in affecting the economic performance of national economies as their positions in the international economy. In particular, they suggested that the organisational structure and political power of labour were key determinants of economic performance under the stagflationary conditions prevailing during the 1970s. Centralised and dense unions and powerful political parties of the Left appear conducive to the development of the politics of concertation between unions, employers' associations and government... Concertation, in turn, is associated with a 'vir- tuous circle' of government political economic policy, which both promotes and is promoted by regulation of workers' behaviour in the market... Where this virtuous circle is estab- lished, it results in better aggregate economic performance. Their study emphasised that the presence of both a powerful Left-wing party and a strong union movement were indispensable in securing economic growth. 'Labour strength only in the market or in politics may not be conducive to economic growth'. They add an interesting rider: where 'labour is weak, both organisationally and politically, it is possible that this general weakness will be associated with economic performance levels similar to those where the labour movement is very strong'. This rider may tempt support for a union-bashing policy from those for whom this has anyway ideological and political attractions. In fact, there would seem to be econ- omic advantages to capital, particularly large-scale capital which may offset these ideological and political disadvantages. Lange and Garrett (1985) argue that associations like unions can advance the interests of their members through one of two strategies: redistributional or collective gains. They would be more likely to support the second strategy - for instance by supporting growth through imposing wage restraints on their members, if they could be certain that their members would benefit from such a strategy. Because capital, not labour, determines how profits are distributed, there is in fact considerable uncertainty that this will happen. There is nothing... to stop capitalists from spending profits in ways which only inefficiently contribute to national economic growth, such as distributing large portions to stock-holders or deciding that overseas investments are more attractive than domestic ones (Lange and Garrett, 1985). . These uncertainties might be reduced by labour taking control over the "westment process - that is to say, by the transformation of the political economy into a socialist state. But an alternative - a more feasible alternative * Way of reducing uncertainties would be to use state policy to reduce the 46 TRANSFORMATION STADLER risks involved in a 'collective gains' strategy. They argue that two political conditions are likely to 'induce capitalists to spend profits in ways efficient for domestic growth, while also assuring that workers will share proportion- ately in the fruits of that growth and will be protected as well as possible should the expected growth not materialise...: an historically strong political Left, and prospects for direct control of government by a party of the Left, closely linked to the union movement'. The authors argued that an 'encom- passing' union movement would be inclined to pursue a collective gains strategy (such as growth) rather than a redistributionist one; while the presence of strong Left governments would reduce the risks that this strategy would not be translated into economic gains for workers. The remainder of their paper consisted of an empirical verification of this hypothesis. It will not be repeated here. It is worth quoting one final remark from the study. Our findings also make clear that the economists' prejudice ... that politics just gets in the way of optimal economic perfor- mance ... can only be sustained under certain circumstances. The conditional relationships we have uncovered certainly suggests that where unions are very weak or are stronger but fragmented, governments of the Left or the Right will find it difficult and probably fruitless to attempt to use political means to achieve desired market outcomes through labour self-regU- lation. The desired outcomes are more likely to be more effec- tively and efficiently achieved by letting the market do its work. In fact, where the unions are fragmented but strong enough to disrupt market outcomes, governments which are perceived to support labour may actually contribute to poorer growth per- formance (Lange and Garrett 1985). In 1988, Alexander Hicks undertook a re-examination of the study. He confirmed that a combination of powerful unions and powerful left-wing governments are conducive to economic growth. Contrary to the assump- tions embedded in the 'overloaded government' thesis, he argued that strong unions and left-wing governments contribute significantly to economic growth, to job security, and to income redistribution. He thus tends to strengthen the case made by Hewitt. One of the reasons for this is the capacity of strong unions to use wage restraints to stimulate growth (Hicks, 1988). The studies by Lange and Garrett were based on advanced industrial countries. It could be argued that South Africa does not exhibit comparable features in its demographic, social, political, economic and infrastructural make-up. This may be true in many respects. But it ought to be stressed that South Africa is a mature industrial society, albeit one with special features. 47 fcTADLER TRANSFORMATION ^Despite these features, it should be recalled that by 1960, 30% of its work- force was employed in industry - the same as Japan. (Subsequently, the figure t o r Japan has increased to 37%, while it has remained about constant for SJSouth Africa.) In two respects, the relative maturity of the South African industrial economy suggest that one of the structural conditions for concer- ttation is present, and the other at least emergent. Concertation can be defined a» as the mutual integration of labour and "capitalist organisations to a high degree. The first condition of concertation $is that feature of the South African economy so widely criticised by left and bright that it has become something of a ritual to condemn it - the heavy •concentration of control over the political economy located in a handful of -companies. It ought to be noted that this is a feature too of the Swedish "economy which was reinforced by Social Democratic tax policy which bene- fits the richest and the poorest income groups at the expense of the two intermediate ones. After the protracted repression of black labour, it is possible to see developing the second condition for concertation: an increasingly coherent sand well-organised union movement with a highly competent leadership. But '•of course it would require the establishment of universal political rights, including the right to organise politically before the potentialities of this latency could be realised. True, the South African union movement is not - or not yet - a movement *n Which any significant level of concertation has taken place, and it is also none which remains highly vulnerable to attack by the state. What is absent of course is a powerful left-wing party in power or in a position of influence •rover government. This is precluded at present by the battery of anti-leftist legislation, by the state's systematic attacks on the left, and by the absence of •common political rights which would enable the majority of workers to achieve political organisation. This situation/o/ce? the union movement to opt for strategies of'redistribu- tion' rather than of 'collective gains', simply because the risks that the benefits •of wage restraints will not benefit workers, or contribute significantly to growth, are very high. From the perspective of capital, interested in growth, •the situation is also unsatisfactory. It is difficult to imagine turning the clock Iback to the days when growth could be achieved because the weakness of the union movement made it difficult for them to prevent controls over wages Ibeing imposed from outside. In the circumstances, then, capital and labour cannot collaborate on a strategy of growth. Such a strategy could only be attained on the basis of the accession to power of a left-wing government, which capital shows little sign