CONCEPTUALISING THE BLACK MIDDLE CLASS

Contemporary interest in the middle class in South Africa revolves overwhelmingly around the extent and consequences of black upward social mobility (with ‘black’ usually meaning ‘black African’). This is the sense in which the term ‘black middle class’ is used here, even though it is acknowledged that this can only be one dimension of the attempt to understand the wider dynamics of class formation and change across the spectrum of ‘race’ as it was constituted under apartheid.Footnote ¹

Since 1994, the profiling of the black middle class has most commonly been related to changing income and consumption patterns, these sometimes linked to changing patterns of behaviour (UCT Unilever Institute 2013; Visagie & Posel Reference Visagie and Posel2013). However, the problem with such consumptionist approaches is that while providing valuable information regarding the growth, living standards, spending power and lifestyles of a black middle class, which according to the last estimate of UCT's Unilever Institute, had grown to 4 million people (out of a total population of 52 million), they are largely descriptive. In contrast, what I term a productionist approach can assist us to explain the dynamics of class behaviour because it is based upon the notion that income, consumption and affluence are outcomes of some combination of work, occupation, education and/or wealth.

For my purposes, I start by referring to the class structure of contemporary South Africa as outlined by Jeremy Seekings and Nicoli Nattrass (Reference Seekings and Nattrass2006).Footnote ² Broadly, their analysis is based upon the notion that class is ‘consequential’, that is, it is strongly linked to what Weber termed ‘life-chances’ (so that, for instance, the better off in society are likely to enjoy better education and gain better jobs). From there, via a complex differentiation of advanced survey data relating to income, occupation and wealth, they sketch out a triangular, three-tiered class system, ranging upward from a numerous marginalised working class, through a narrowing tier in the middle, up to a small upper class at the top. Within each tier, there are subcategories of class: informal and casual workers alongside an impoverished and grant dependent underclass at the bottom; semi-professionals such as teachers and nurses alongside white collar workers and formally employed workers in the middle; through to the better off, composed largely of managers, professionals and the independently rich at the top. From their perspective, while semi-professional elements are privileged or semi-privileged, what they term the core working class (those contractually employed within the formal sector) is also located within the middle of the class structure in income terms. Within this overall context, they locate a new black elite as penetrating the upper classes, buttressed from below by a growing black middle class within the second tier. Providing fulsome statistical data, they conclude that the ‘accelerating growth’ of these classes provides ‘the most dramatic shift in the social landscape after 1994’.

Seekings and Nattrass have been criticised for their inclusion of the core working class within the middle tier. I will restrict myself here to merely noting that major work, from a variety of sociological traditions, has long sought to grapple with changing boundaries between the higher-skilled elements of the manual working class and white collar work. More important, for my purposes, is that while their approach is largely Neo-Weberian in inspiration, they simultaneously draw upon Neo-Marxist theorising, notably that of Erik Olin Wright, to relate the different categories and sub-categories of class to the exercise of workplace authority. Famously, Wright has argued the existence of contradictory class locations, as illustrated by the example of a property-less foreman exercising supervision over workers on behalf of capital (e.g. Wright Reference Wright, Giddens and Held1982). However, where Seekings and Nattrass fall short, in my view, is that while they berate the Neo-Marxist school in South African academia for having disdained the dirty work of mining quantitative data relating to changing patterns of race, income and inequality in favour of grandiose theorising, they themselves fail to link their class structure to ‘the big picture’ which, for all its flaws, the Neo-Marxists have generally attempted to present.

The problem is best illustrated by the blandness of their categorisation of those at the top of the class structure as simply the ‘upper class’. While clearly locating a small category of people at the top of the pile in terms of their income and wealth (even though relating this to the authority which managers and professionals typically wield in the workplace), this avoids dealing with the sources of its wealth, power and authority in post-1994 South Africa. Quite simply, the perspective here is that we cannot talk sensibly about ‘class’ without simultaneously linking it with the possession or otherwise of ‘power’ within the post-apartheid political economy.

Elsewhere I have outlined how, basing itself upon its political hegemony, the ANC has sought to bring public institutions which are decreed as politically independent under the constitution under the control of a hybrid ‘party-state’ (Southall Reference Southall2013a). Accession to state power has made the ANC the major fount of opportunity, for both employment and access to resources, for a black majority which hitherto was blocked from competing freely in the market. As a result, the party-state has become a – if not the – ;fulcrum around which upward social mobility, and chances for ‘private accumulation’, revolve for historically disadvantaged segments of the population. However, while the ANC has assumed political power, and wields considerable derivative economic power resulting from its control over state revenue, expenditure and parastatals, the economy remains overwhelmingly dominated by internationally and domestically owned large corporations and financial institutions. State power therefore co-exists and cooperates with corporate power, while power is simultaneously contested and shared by a ‘power elite’ which remains fractured along lines of race (Southall Reference Southall, Daniel, Naidoo, Pillay and Southall2013b).

The conclusion that follows is that any attempt to understand the social weight and political orientations of ‘the black middle class’ has to be located in this wider context, and that we need to link the class structure outlined by Seekings and Nattrass to the ANC's party-state. For the black middle class, the beneficiary of ANC policies such as affirmative action, the deracialisation of the education system, and BEE, it is the party-state as much as, if not more than, the private corporation or professional employment, which is the source of its upward mobility, income, wealth and authority. While, admittedly, all of this is difficult to operationalise, it is essential for us to discern the relationship of different elements of the black middle class to the party-state if we are to attempt to understand their political orientations and behaviour.

Let us now attempt to relate such a perspective to the three broad propositions concerning the politics of the black middle class which were listed above.

PROPOSITION 1: THE BLACK MIDDLE CLASS AND DEMOCRACY BEFORE 1994

The argument that the black middle class was overwhelmingly a progressive force in the struggle for liberation and democracy is writ large in the literature, and I have no intention of contesting it. In reaction to military defeat by colonising forces, a small but politically conscious black middle class emerging out of a handful of mission schools played the key role in forming the ANC in 1912 to protest against wide-ranging, highly oppressive impositions of imperial and settler colonial authority. Thereafter, although there is much debate about the relative political weight of different classes within the organisation, there is widespread recognition that the smallness of the African middle class, its limited access to property, and the denial to it of opportunities for upward social mobility by the various racial restrictions imposed by white regimes (especially after 1948) thrust it into alliance with a burgeoning urbanised working class, transforming the ANC into a mass movement during the turbulent 1950s.

While historical argumentation continues around the extent to which middle class elements led the ANC and influenced its strategies and tactics, with the South African Communist Party (SACP) emphasising the vanguard role played by the working class in the struggle for liberation, there is no denial of the leadership offered by middle class individuals such as lawyers Nelson Mandela and Oliver Tambo. During the subsequent years of the organisation in exile, the ANC's alliance with the SACP and the convergence of middle and working class interests in the overthrow of apartheid led to the elaboration of the theory of the National Democratic Revolution (NDR). In essence, this sketched out a two-stage historical trajectory in which, after the defeat of the white regime, a capitalist period of ‘national democracy’ would follow during which – overseen by the ANC – an African capitalist class would emerge as a ‘patriotic bourgeoisie’ and raise the productive forces of the economy, until somehow this would give way to a higher historical stage of socialism.

Even if we accept the broad thrust of this outline, there are important qualifiers to be made. A first is quite simply that, while we should endorse the black middle class as a major contributor to the struggle for democracy, the ardour with which it pursued that mission waxed and waned according to historical situation. Peter Limb (Reference Limb2010) has recently offered a strong critique of established views that the early ANC was dominated by a small African middle class elite which was divorced from its support base amongst workers and peasants. In contrast, he argues that the smallness of the middle class and its organic connections (via families and communities) to the impoverished masses ensured that – for all its limitations and periods of inactivity – the ANC in some form or other was constantly in touch with popular concerns, not least because it was subject to pressures and criticism from below. Even so, this does not contest the major narrative that the early middle class leadership of the ANC, while proclaiming the liberalism it had absorbed from the missions, was strategically and politically conservative. Unsurprisingly, given the sheer weight of white political domination, the goals and vocabulary of protest were largely defensive, and the means used overwhelmingly legalistic. It was only after the replacement of the conservative Xuma ‘old guard’ by the Youth League radicals in 1949 that the ANC launched out onto the road to becoming a mass movement, and that the nationalist cross-class alliance seriously took shape.

The second major qualifier is that black middle class support for the liberation project was never unanimous, and never unambiguous. A first theme in the literature is that, especially after the clampdown upon the liberation movements in the early 1960s, many if not most middle class Africans were, if not actually acceptant of defeat, then acquiescent in their racial subordination – a situation which was to be subject to a major generational challenge from the rise of black consciousness and the Soweto uprisings. Another theme is that middle class elements, in their ambition for improvement or advantage, served on various official structures, such as municipal advisory bodies, for limited financial or psychological gain, thereby accepting if not actually endorsing their subordinate racial status. In turn, this acceptance of subordination was to be writ large by the participation of middle class elements in the homeland project, devised by the NP government to undercut support for the liberation movements and to entrench white domination. Certainly, in retrospect, we may accept that much of this involvement was pragmatic rather than ideological for all but a particular coterie of leaders committed to ethnic nationalism, as indicated eventually by the covert assistance extended to the ANC by a number of homeland governments as the struggle for liberation moved to a climax. Even so, although middle class involvement in homeland governance was heavily subject to chiefly authority, the bantustan project – as a conservative bulwark against the forces of liberation – could not have succeeded as much as it did without it. This was to be emphasised by the popular upheavals which confronted various bantustan leaderships during the 1980s, as well as by the numerous popular campaigns waged against ‘collaborationist’ councillors in the urban areas when the regime sought to quell township rebellion by devolving authority to black municipalities. In short, middle class involvement in official structures under apartheid always demonstrated a marked political ambiguity towards the liberation project as envisaged by the ANC.Footnote ³

None of these qualifications refute the major narrative of the black middle class under apartheid as a leading component of the liberation movement in alliance with the working class. Nonetheless, they most certainly complicate it. A significant dimension of South African history is how white governments sought to capitalise upon the ‘respectability’ of the African middle class, recognising this as a basis for the forging of conservative alliances against radical elements and protest from below.

This was to gather major momentum from the mid-1970s. First, prompted by the increasing complexity of the economy, a changing racial division of labour featured an increased pace of black upward advance into occupational spheres previously dominated by whites, themselves upwardly mobile (Crankshaw Reference Crankshaw2002). Second, this in turn required a major increase in the provision of education to black children, the number of black African children in school leaping dramatically from 2·7 million in 1970 to over 7 million in 1988, as well as the opening of black universities and increased access of black students to ‘white’ universities and technikons, so that by 1989 there were around 110,000 black students in tertiary education (SAIRR 1971: 257, 289, 1989/90: 824,862, 872). Third, although it was to remain concentrated in historic spheres of the law, medicine, teaching and nursing, the black middle class increasingly began to penetrate the corporate sector as managers (Dreyer Reference Dreyer1987: 16 cites a total of 2860 blacks in managerial and administrative occupations by the mid-1980s). Fourth, relaxation of restrictions upon African business, first in the homelands, subsequently in urban areas, facilitated the expansion of a black entrepreneurial stratum, in what Kane-Berman (Reference Kane-Berman1991) has described as ‘a silent revolution’.

All this provided a more promising context for NP governments to step up the pace of counter-revolutionary reform as pressures upon the apartheid regime mounted. This shift was cautiously backed by large-scale capital, which envisaged the extension of property rights, training and professional employment opportunities to increased numbers of blacks as strengthening the stake of a subaltern middle class in the capitalist system. Eventually, of course, as the liberation movement gained traction, large-scale capital (alongside progressive elements within the white community) was to play a major role in the forging of a democratic transition which accepted the fundamentals of market economy. This registered a defeat for the more radical wing of the ANC, however much of this was to be wrapped up in the pretence that the implementation of the national democratic phase was to herald a transition to socialism.Footnote ⁴ In reality, the national democratic phase was to provide the foundations for a ‘middle class democracy’ under the leadership of a ‘party-state bourgeoise’.

PROPOSITION 2: THE BLACK MIDDLE CLASS AS THE PRODUCT AND PROPONENT OF THE ANC

Despite its more rapid development in the latter years of apartheid, the size of the black middle class at the dawn of democracy was still remarkably modest. In many ways, therefore, it is the ANC which has created the black middle class as we know it today. It has done so by using its ‘capture of state power’ to implement policies which, whether by intention or simply outcome, facilitate black upward mobility.

First, in line with the theory of the NDR, which requires that the vanguard party capture ‘the strategic centres of power’, the ANC has ‘deployed’ its chosen personnel to key positions in state and society. In theory, the ANC selects individuals for positions not merely according to their dedication to revolutionary duties but also according to their qualifications and abilities. In practice, however, as even the ANC has recently come to appreciate, the outcomes of deployment have become increasingly problematic – for at all levels of the state, a marked tendency to prioritise loyalty to the party (or often, factions within it) over merit has come to predominate, facilitating the rise of patronage-client relations constructed around political control over the allocation of state resources. Suffice it to say here, deployment has emerged as a major instrument whereby individuals have gained the opportunity to manipulate opportunities provided by the public sector, to accumulate and to ascend the class structure (Southall Reference Southall2013a: 134–6).

The second instrument of class formation has been the ANC's aggressive strategy of affirmative action (whose implementation, it needs to be noted, is often operationally difficult to separate from deployment). The outcome has been most dramatic in the public sector which the ANC has sought to render ‘demographically representative’ of the population as a whole. In so doing, it has massively widened access for its own political constituency to state employment on terms and levels of remuneration which, today, appear to be relatively generous compared with what the large majority of public servants might expect to receive within the private sector. Today, even the upper reaches of the public sector are black. Meanwhile, the private sector is similarly subject to pressures, legal and otherwise, to implement ‘employment equity’ (Southall Reference Southall2013a: 137–45).

Third, the ANC has massively increased educational opportunities for blacks at both secondary and tertiary levels, moving well beyond the expanding access availed in the late-apartheid era. Two key processes run contiguously. On the one hand, there has been an impressive expansion in the absolute number of places for blacks at both state secondary schools and institutions of higher education. On the other, the deracialisation of education has significantly widened access for blacks to schools and higher educational institutions which were previously reserved for the individual racial groups. Educational opportunity and resources under apartheid were severely skewed by race, massively favouring whites. Now in formal terms this is no longer so, the most notable feature of the transformation being increased black entry to the schools and higher educational institutions previously reserved wholly or overwhelmingly for whites. Entry into former white institutions massively increases chances of educational success, for overall standards and facilities in former black institutions at all levels remains much inferior (and success rates, too often totally dismal). This is in part because the educational advantages enjoyed by the former racially privileged institutions are buttressed by the right to levy and set fees, effectively restricting entry to those who can afford them. For all that rates of black entry may be increased by access to scholarships and loans, the class implications are so evident that black parents display enormous determination to meet the costs involved, and where necessary, to buck the lottery of the public school system by sending their children to a steadily expanding private sector in education. In sum, especially given the overall disappointing results of the public school system in areas formerly officially designated ‘black’, schools and higher education institutions which have been able to capitalise upon a legacy of racial privilege have emerged as key sites for the production and reproduction of class advantage (Hunter Reference Hunter2010).

Finally, the most explicit strategy of class formation pursued by the ANC has been the pursuit of Black Economic Empowerment. If the impact of affirmative action was to be most advanced within the public sector, the objective of BEE – as it was formally constituted during the years of the Mbeki presidency – was to challenge the domination of South Africa's corporate structure by whites and promote the development of a black capitalist class, thereby bridging the post-1994 separation of black-political and white-economic power. Suffice to say that the results have been mixed, and in many ways, for the government, disappointing (not to mention, often embarrassing), and share ownership of the majority of large- and medium-size firms still remains overwhelmingly white. Nonetheless, the parastatals – sites of Afrikaner power in years past – have become largely blackened, and have provided an important launch pad for ambitious individuals to transfer into the private sector. Concurrently, the lily-white racial hue of the topmost corporate ranks has been leavened by the politically leveraged development of a small, but enormously rich Black elite, alongside an increase in the proportion of black managers as companies have come under pressure to render BEE more ‘broadly based’. In short, even if BEE has in the view of many facilitated the rise of ‘crony capitalism’, it has played a major role in class formation (Southall Reference Southall2013a: 220–6; 263–6).

The argument that follows, broadly put, is that a black elite and middle class that has found itself in positions of power, privilege and profit is not likely to bite the hand of the party-state that feeds it. The more dependent they are upon the ruling party for their welfare, the more they are likely to support it. This finds endorsement in the work of Collette Schultz-Herzenberg, who in her detailed analysis of voter attitudes between 1994 and 2009, found that the proportion of ANC ‘partisans’ amongst black owner/employer and professional/supervisory occupational categories had increased as blacks had moved upwards: ‘core support for the ANC has begun to emerge from the material beneficiaries of a post-apartheid order, shown in the increases in levels and strength of party support among the black middle classes’ (Schulz-Herzenberg Reference Schulz-Herzenberg2009: 142).

I have elsewhere identified the black elite and middle class as composed of four major segments (Southall Reference Southall2005). The first is constituted of an upper layer of ‘state managers’ comprising the president, premiers, members of the Cabinet and provincial governments, senior civil servants and senior executives in parastatals and other public institutions. State managers include the key political decision-makers, are relatively few in number (there were just 10,598 state employees in national and provincial governments classified as ‘senior management’ in 2010) (SAIRR 2010/11: 246), and it can be assumed the large majority of them are united by their loyalty to the ANC (even if they adhere to the different factions of the party).

A second segment, a ‘corporate black bourgeoisie’ is composed of those who have significant shareholdings in major companies and/or who are employed at executive or managerial level in the large corporations which dominate the private sector, significant numbers of them having previously served within government or parastatals. Overall, despite official lamentations that the private sector lags behind in its enthusiasm for ‘transformation’, this segment has grown significantly since 1994. According to the Commission for Employment Equity, the proportion of ‘top managers’ in the private sector who are black (African, Coloured and Indian) has increased from 12·7% in 2000 to 30·8% in 2011, this matched by an increase in the proportion of ‘senior managers’ who are black from 18·5% in 2000 to 38·4% in 2011, with indications that around two-thirds of those who are black are African (SAIRR 2012: 240–1).Footnote ⁵ While the black senior executive and managerial corporate element may regard itself as autonomous of the ANC, the steady growth of this segment is a product of the democratic era. At one level, this records commercial common sense. At another, it reflects the desire of large corporations for ‘political connectivity’ alongside their need to transform their demographic profile in conformity with official demands.

The black middle class is further composed of a third segment, what I have termed a ‘civil petty bourgeoisie’, comprised of members of parliament and provincial legislatures, middle management within the public sector (20,996 of them in national and provincial governments in 2010), along with a much larger mass of lower-level public employees in non-manual occupations, as well as black professionals and semi-professionals, both within the public service (notably teachers and nurses) and beyond it (for instance, independently employed accountants and lawyers). Overall, this is clearly an extremely heterogeneous category. Nonetheless, as with ‘state managers’, it is likely that the majority of this grouping is loyal to the ANC, having benefited variously from strategies of ‘demographic representivity’ and deployment, notwithstanding the fact that many of them may have many discontents with their terms and conditions of employment.

There is, finally, a black business and trading bourgeoisie, combining a mix of owners and managers of medium and small businesses, the diversity of this grouping indicated by the fact that, at its lower levels, black operators merge into the lower regions of the informal sector of the economy, their activities often characterised by their taking place outside the tax net and sometimes (according to popular legend) spiced up by involvement in protectionism and criminality (the taxi industry being a particular example). Historically, this grouping was heavily constrained by apartheid limitations imposed upon black activity in business, and even today, given lack of capital, skills and training, it continues at a major disadvantage relative to white-owned small businesses. Unsurprisingly, therefore, since 1994, black business has become increasingly vocal and active in lobbying the ANC to promote its interests, while generally its competitive disadvantage appears to render it heavily dependent upon its relations with the party-state, notably in terms of its reliance upon concessions and contracts from government at all three levels, and from public sector entities more generally.

It is straightforward to locate state managers and the higher ranks of the ‘corporate black bourgeoisie’ as falling into the upper class (as depicted by Seekings and Nattrass) and merging into a more narrowly defined ‘power elite’ by virtue of the wealth it enjoys and the powers it deploys. Likewise, the lower ranks of the ‘corporate black bourgeoisie’ join the remaining two segments as falling into the middle class proper. All these various elements of the black elite and middle class have been beneficiaries of ANC rule,Footnote ⁶ and for the moment at least, the majority of their ranks are likely to continue to align with it politically. Yet things do change, and as the black middle class becomes more consolidated, its political orientations are likely to become more diverse. It is not wholly without reason that the ANC is becoming somewhat nervous about the reliability of black middle class vote support.

PROPOSITION 3: THE BLACK MIDDLE CLASS AND THE CONSOLIDATION OF DEMOCRACY

Prior to the May 2014 general elections, there was widespread suggestion that the political alignment of the black middle class to the ANC was under stress, and that opposition parties were targeting it as an attractive constituency. This was argued in particular relation to the Democratic Alliance (DA) which having since 1994 gobbled up remnants from the formerly ruling National Party, has been particularly eager to rebut attacks that stereotype it as little more than a front for white interests. In recent years, it has vigorously promoted its non-racial credentials in a manner which, although designed to appeal to black voters more widely, seem particularly geared to woo the black middle class. In particular, under Helen Zille, the DA has proclaimed its support for ‘broad based’ BEE in order to address the legacy of racial inequality, while indicating that it is highly critical of the manner in which the ANC has implemented it (Mail & Guardian 2013). Unsurprisingly this has created considerable controversy within its ranks, its right wing complaining that any embrace of policies which involve recognition of racial categories constitutes an abandonment of the party's liberal heritage; Zille responding that the social engineering of racial redress to counter the apartheid legacy is fully compatible with the party's mission.Footnote ⁷ For some observers, this debate implied that for all its efforts to bridge the racial divide, the DA was unlikely to significantly expand its support amongst black voters (McKaiser Reference McKaiser2014). This seemed to be the view of Mamphela Ramphele, who – deeply critical of the ANC's performance in power – chose to launch her own party, Agang SA, rather than throwing in her lot with the DA with which ideologically she had little difference. Agang SA was thought by some to constitute an attractive rival for the affections of disaffected members of the black middle class (Zibi Reference Zibi2013), until it more or less self-destructed as a result of its leader's spectacularly inept politicking in the lead up to the election.Footnote ⁸

The view that the black middle class was becoming more politically diverse was backed by the findings of a survey conducted by Lawrie Schlemmer in 2003. These acknowledged that ‘the race-based transformation polices of the ANC have created an exceptionally strong bond between it and the African core middle class’ (Schlemmer Reference Schlemmer2005: 10). Nonetheless, they also indicated that the black middle class expressed numerous misgivings about ANC performance in power, with some respondents indicating support for other parties. This, Schlemmer (Reference Schlemmer2005: 11) concluded, indicated that their political attitudes were pragmatic, and exhibited ‘anything other than a consistent, solidly based, and … durable preference’. Even so, because of the social distance between the black middle class and their white counterparts (expressed by lack of social contact between them outside the workplace), ‘the area of least convergence between the African and white middle classes (was) that of party politics’. This was linked to a widespread belief amongst members of the black middle class that white wealth was the product of a racist past, and that consequently the position and privileges of the white middle class were illegitimate.

Schlemmer's findings are now dated. However, their thrust finds support in later work. Robert Mattes (Reference Mattesforthcoming), interpreting AfroBarometer data, reports widening intra-black class differences with regard to electoral participation. Individuals within the black middle class were less likely to have voted in the 2009 election, to have identified with any political party, and to have identified with the ANC specifically (this being most marked amongst the young). The black middle class also proved less likely than other black survey respondents to have contacted elected or government officials, or to have taken part in community action or protest.

Similar findings were reported by David Everatt (Reference Everatt2012) after conducting focus group discussions amongst black middle class participants. These indicated that black voters with relatively high incomes were looking for a party that would defend their middle class status. They recognised that the ANC had played a major enabling role in their lives. Nonetheless, they complained about high levels of corruption, broken promises, failures of service delivery and so on, while simultaneously worrying about how the growing militancy of the poor might impinge upon their security. However, while guarded approval was extended to the DA for demanding accountability from the ANC, few would consider voting for it nationally, and if they did so, it would be primarily to register discontent. There was resentment at being regarded as ‘voting fodder’ by the ANC, and a growing willingness, as one participant put it, to ‘jump ship’ (Everatt Reference Everatt2012: 37). However, without any serious political choice but the DA on offer, the major alternative was simply that of abstaining from the polls.

Focus groups with black middle class participants were similarly conducted by the present author with Amuzweni Ngoma. The latter's interpretation suggests that as ‘professionalism’ (centred around a set of ethical values and practices) takes hold amongst the black middle class, inclusive of the public sector, individuals are more likely to claim independence from the ANC.Footnote ⁹

Mattes (Reference Mattesforthcoming: 12) concludes that there is little to support the argument that the ANC has created a party-state bourgeoisie:

South Africa's emerging black middle class are no less loyal … to the new, democratic South Africa. While they are less likely (than the poor) to voice their dissent, at least in terms of active protest action, they also show some limited signs of exiting from the electoral process and the larger democratic process between elections.

However, note that both Mattes and Everatt refer only to what is termed here the ‘core’ (or upper) black middle class, and make no reference to the more extensive, lower black middle class whose politics may be different, influenced by high levels of trade union membership. In contrast, it is argued here that while, indeed, the ANC has good reason to be worried about the potential for greater loss of black middle class support, this is most likely to be occurring amongst those elements of the class who are increasingly distant from the party-state: amongst what I have termed the ‘civil petty bourgeoisie’, that is, those working in the private sector, the media or civil society.

It is interesting to relate the above discussion to the course and outcome of the recent general election. There was substantial indication during the campaign that the ANC was worried by the threat to its support among the black middle class presented by the DA and Agang SA. Concerned that many traditional supporters were embarrassed by the image of President Jacob Zuma, party leaders in Gauteng appealed to former President Thabo Mbeki to re-enter active politics and to campaign amongst black middle class voters (Mthembu Reference Mthembu2013; Munusamy Reference Munusamy2013). Although Gwede Mantashe, Secretary-General of the ANC, expressed the national leadership's displeasure at the affront to Zuma (Marrian Reference Marrian2013), it was widely accepted that the ‘grumblings’ of the black middle class (especially in Gauteng, where its core elements were concentrated) over governmental inefficiency and corruption were taken particularly seriously (Merten Reference Merten2013).

This should not surprise us. When, after Mbeki's ejection from the state presidency in September 2008, some of his supporters formed the Congress of the People (COPE), it was widely predicted that it would attract the black middle class. This persuaded the ANC to deploy huge resources to counteract it. Even so, COPE still took some 7 per cent of the national vote (a not unworthy achievement for a newly launched party). Subsequently, however, COPE fell victim to perpetual leadership squabbles, and failed to pack any significant electoral punch in May 2014, gaining less than 1 per cent of the national vote. However, its performance in 2009 underlined the threat which dissent within ANC ranks represents to the party's hegemony, especially if this were to be channelled into a wider coalition pact. Although the majority of the electorate are poor and black, the ANC's electoral hegemony since 1994 has been founded upon its claim to represent a home for all across race and class. Accordingly, the drift of black middle class support to parties of opposition would represent a major symbolic defeat, of far wider import than merely the loss of votes that it would entail.

With Agang SA virtually imploding before the election, the major threat to the ANC's black middle class vote remained the DA. Helen Zille's anger with Ramphele when she rejected their agreement that she stand as the DA's presidential candidate was made embarrassingly public. Not only had Zille herself gone ahead of many within her party's top echelons, but she made the DA vulnerable to ANC jibes that it was resorting to tactics of ‘rent-a-black’. Nonetheless, the DA recovered to mount a vigorous campaign to attract black voters. Carefully choreographed rallies with party leaders flanked by enthusiastic black supporters, all clad in the party's bright blue t-shirts boasting the party logo, were matched by loud insistence that the DA's predecessors had played a major role in the liberation struggle. Interestingly, too, the DA contrasted the liabilities of the Zuma era with the more favourable records of the Mandela and Mbeki presidencies. It also came up with a remarkably effective black leader in the form of Mmusi Maimane as its front in Gauteng, where he ran an Obama-like revivalist campaign (Jolobe Reference Jolobe, Schulz-Herzenberg and Southall2014). However, while the DA was to improve its overall proportion of the vote (up nearly 8 per cent from 2009 to a total of just over 22 per cent in the national election), its performance amongst black voters remains difficult to assess.

The DA claims that it secured some 760,000 black votes out of its total of 4,089,215. It improved its showing in the provincial contests in each and every province, its worst improvement being in Kwazulu-Natal, where its proportion of the vote increased from 9·1 to 12·7 per cent, its best performance, relatively, being in Northern Cape, where it leapt from 12·5 to 23·9 per cent. Above all, it could be pleased with the leap in its proportion of the vote in Gauteng from 21·8 per cent in 2009 to 30·7 per cent in 2014, with the DA making some significant inroads into more prosperous areas within Soweto such as Diepkloof.Footnote ¹⁰ In turn, there were across the board increases in the DA's shares of the vote in the metropolitan areas, these ranging from 18·8 per cent in Buffalo City (East London) to 59·3 per cent in Cape Town. It was particularly gratified by its gaining 40·1 per cent of the vote in Nelson Mandela Bay (Port Elizabeth), a result which signalled that the ANC stands in danger of losing control of the city council in the forthcoming 2016 elections.

The DA's performance in the metros suggests that it gained increased support from black voters whose personal or familial circumstances have improved materially and otherwise since 1994. It also benefited from the collapse of COPE, and it may even have gained grudging support from some who would have voted for Agang SA had it not run its campaign so spectacularly badly. However, qualitative analysis suggests that black professionals continued to remain shy of backing the DA (Ngoma Reference Ngoma, Schulz-Herzenberg and Southall2014), this implying that if they did not remain with the ANC, they might have chosen to abstain. In turn, this suggests, intriguingly, that the bulk of the DA's improved black vote may have come from the lower rather than the core segments of the black middle class.

The only other significant alternative on offer for black voters (bar the declining Inkatha Freedom Party in KwaZulu-Natal and the usual smattering of small parties) was the Economic Freedom Fighters (EFF), established by former ANC Youth Leader Julius Malema after his expulsion from the ANC in 2012. Centring its support amongst striking mineworkers on the platinum belt in North-West and in Malema's home province of Limpopo, the EFF put forward a militantly radical-nationalist programme designed to appeal, notably, to the youthful unemployed, marginalised and economically disposed. However, once the election campaign began in earnest, it cast its net much wider, appealing amongst others to professionals frustrated with being unable to rise in the workplace. However, while it swiftly acquired the support of certain student elements (some of them migrants from the Youth League),Footnote ¹¹ and while it gained the backing of a handful of high-profile professionals, its support seems to have been overwhelmingly rooted amongst its original target group (Robinson Reference Robinson, Schulz-Herzenberg and Southall2014).

More detailed analysis is required to confirm these ruminations about how the black middle class voted in the 2014 elections. Meanwhile, it is important to stress that the question of whether the black middle class will abandon the ANC is logically quite separate from that of whether its expansion and greater diversity will contribute to the consolidation of democracy. Far too often these questions are confused, with the implication being that only when black middle class voters cut their links to the ANC will they mature as democrats. While this may be a defensible postulate, it is only that, based upon the conventional supposition that the emergence of a genuinely competitive party system requires the erosion of ANC dominance and the genuine prospect of regular alternations in power. However, it will take far more than the defection of the black middle class from the ANC to promote a more democratic polity, not least because democracy requires much more than just competition between parties. Consequently, in moving towards a conclusion, I want to put forward two further propositions.

First, it may well be that the extensive alignment of the black middle class to the party-state and the continuing social distance between the black and white segments of the black middle class throws doubt upon the potential for convergence within the context of party politics. Nonetheless, this does not mean that the black middle class is not in line to make a significant contribution to democracy. Most visible (and audible) today are the growing legions of the black middle class who contribute energetically to the public discourse about anything and everything as editors, journalists, anchormen, commentators, letter writers, bloggers and so on through the established and social media. Many are as virulently critical of the ANC, if not more so, than the DA or other parties of opposition, and offer more punch precisely because of where they come from.

Second, although surveys indicate that the black middle class does not as yet display high levels of involvement in independent civil society, it would be reasonable to surmise that this will grow. Blacks are making their way forward in key occupations such as law and accounting, and playing a more prominent role in their professional associations. Consequently, just as poor communities across the country are choosing to make their voices heard via ‘service delivery’ protests, so it is likely that members of the black middle class will find common cause with their white counterparts in registering protests about discontents – whether about schooling, transport, municipal service, environmental issues or whatever. In short, even if black middle class voters choose to stop short of voting against the ANC, we can be assured of their growing demands for accountability – hopefully from the private sector as much as from government.

Third, it is often suggested that there is greater prospect of a fundamental challenge to the ANC resulting from a split within the Tripartite Alliance than from a migration of black voters to the DA. Were this to happen, various commentators envisage a significant re-arrangement of the party system, this often viewed as the product of a divergence between ‘nationalist’ and ‘socialist’ tendencies within the ANC. This is often said to open up the prospect for coalition between ‘nationalist’ elements of the ANC and pro-market opposition parties, and for ‘socialists’ to move into a new party of labour. If such a scenario were to come about, the further suggestion is that the large body of the black middle class would stay with the ‘nationalist’ wing of the ANC, or to coalesce with it under whatever label it might choose to operate.

Present struggles inside the Alliance (between rival elements within COSATU, between COSATU dissidents and SACP-aligned union leaders and so on), suggest that a realignment of the party system is on the cards, perhaps as a result of the 2014 election. However, whereas during the economic crisis in Zimbabwe between 2000 and 2008, significant elements of the middle class joined up with workers to form the MDC, in contemporary South Africa it seems unlikely that core elements of the black middle class would abandon the ANC for a militant party of labour.