Introduction
South Africa’s negotiated transition to democracy hinged on bargains reconciling incumbent white economic elites with ascendant black political elites. Among these bargains the African National Congress's (ANC's) adoption of economic liberalisation would be accompanied by a tacit commitment to deracialisation of private sector ownership through black economic empowerment (BEE) (Von Holdt 2013). Becoming one of the ANC’s key redistributive strategies, BEE was institutionalised during the Mbeki presidency through legislation requiring larger companies operating in licensed sectors or seeking state tenders to pursue affirmative action in employment and procurement, alongside meeting targets for equity ownership by historically disadvantaged South Africans (HDSA).1 Mining has been pivotal to BEE as the industry with the largest BEE deal value, with an estimated R101 billion transferred through equity sales in established mining companies to HDSA between 2000 and 2015 (Intellidex 2015). However, BEE policy for mining has been subject to fierce and prolonged contestation. This has revolved around disagreements over the design and implementation of the Mining Charter, a landmark BEE policy document introduced in 2004 after years of fraught negotiation. The Charter has since gone through several iterations, the latest published in September 2018.
The article analyses the processes shaping the Charter’s successive iterations, focusing on ownership transfer, its most contentious aspect, and drawing on case study material from platinum-group metals (PGM) mining, the most important mining sub-sector by employment, and crucial to the Charter’s design and implementation (Capps 2012). Through this, the article provides new analysis of the outcomes of BEE for the political economy of South Africa. The ANC adopted BEE policy to provide a relatively market-friendly means of redistributing wealth and gradually deracialising control of the private sector, envisaging the formation of a ‘patriotic’ black capitalist class who would align industry with government at the economy’s ‘commanding heights’, around a progressive developmental agenda (Southall 2004). Extant literature on BEE is predominantly from the late Mbeki and early Zuma presidencies (Butler 2011; Freund 2007; Gumede 2008; Iheduru 2004, 2008; Mbeki 2009; Ponte, Roberts, and Van Sittert 2007; Tangri and Southall 2008; Taylor 2007). While taking different positions on its redistributive efficacy, this literature broadly agreed BEE had increased alignment between white economic and black political elites, though more around maintenance of the liberal economic framework implemented after 1996 than a progressive developmental agenda.
However, as this article argues, this process of alignment increasingly unravelled over the subsequent decade, due in part to the shortcomings of BEE. The downturn in the commodities cycle after 2008 exposed the inherent fragilities of conventional BEE asset transfer models, which the article argues represent a form of financialised redistribution. The model typically involved highly leveraged share purchases, with redistributive outcomes contingent on volatile equity returns and the vagaries of global commodity markets. It created in many instances a limited form of empowerment in which BEE partners were included as owners but structurally disadvantaged through their positioning at subsidiary level and in less viable mining assets. This approach, predicated on the creation of tradable minority stakes in the existing economic base, created some relatively successful BEE holding companies but few fully independent black-owned mining companies of significant scale (see also Zalk 2017), and has therefore not facilitated meaningful transformation of economic control in the manner envisaged. In a broader context of economic stagnation and persistently high unemployment and inequality, this contributed to increased factional and ideological fragmentation in the Tripartite Alliance over economic strategy and a deterioration in relations between the state and big business as ascendant nationalist factions in the late Zuma period pushed for a more confrontational, accelerated approach to ownership transfer. Ultimately, the prolonged slump in the mining sector and the accompanying concerns over falling employment and investment, alongside the fragmentation of power in the ANC, has enabled the mining industry to resist these pressures to a degree. The first Charter was a fractious but ultimately negotiated compromise between white business elites and a powerful multi-class, nationalist ‘empowerment alliance’ (Capps 2012, 325) comprising the ANC, trade unions, the South African Communist Party (SACP) and the nascent black business class. However, the intense political contestation around the third Charter reflects growing division within the Tripartite Alliance and an increasingly differentiated black business elite.
The article uses qualitative documentary analysis to trace the complex dynamics of both specific BEE deals and more general contestation of the Charter legislation. Sources include published materials from major mining companies, trade associations, the Department of Mineral Resources (DMR) and other relevant state agencies, and investment banks. Analysis of specific BEE deals is supported with analysis of company annual reports and annual financial statement.
BEE and the post-apartheid political compromise
Adopted by the ANC in the mid 1990s, BEE embodies the compromises of the post-apartheid transition. While the ANC would dominate electorally, the economy remained highly consolidated and white dominated. Six conglomerates controlled approximately 85% of the Johannesburg Stock Exchange (Chabane, Roberts, and Goldstein 2006), wielding enormous economic influence (Carmody 2002). Apartheid’s racial ownership restrictions meant the black business class was small, fragmented and lacking management experience (Iheduru 2004). Seeking to placate white big business and attract foreign direct investment, the ANC leadership underwent ideological conversion and abandoned socialist economic strategies, notably nationalisation, in favour of economic liberalisation (Carmody 2002; Marais 2011). Correspondingly, white big business tacitly committed to BEE as partial transfer of equity to HDSAs, to secure political legitimacy (Bassett 2008; Butler 2011). A fundamental tension underlies BEE. For white business elites, who had initially devised the model, BEE was a means of defensive self-preservation. For liberation movement adoptees, however, BEE was envisaged as a means to a more transformative end through the formation of a ‘patriotic bourgeoisie’ of black business owners and managers (Gumede 2008; Southall 2004). Cooperation with white big business and accumulation by a narrow, politically connected BEE elite was intellectually justified as a phase in the National Democratic Revolution (NDR), whereby formation of a progressive black capitalist class preceded more far-reaching social change (Hart 2014, 174–181). The patriotic bourgeoisie would align government with leadership of the economy’s ‘commanding heights’ around a shared developmental programme to unmake the colonial economic structure (Hart 2014; Southall 2004).
Early BEE deal-making was at the voluntary initiative of white big business, which was rapidly restructuring and internationalising and seeking to offload some assets in the process (Carmody 2002; Murray 2000). Widespread disillusionment with the slow pace and limited impact of this process within the nascent black business class (Murray 2000) propelled government legislative initiatives, resulting in the 2003 Broad-Based Black Economic Empowerment (BB-BEE) Act, the 2007 Department of Trade and Industry Codes of Good Practice, and Charter documents for sectors subject to licensing, such as mining and finance, that set targets for HDSA equity ownership, employment in management positions and preferential procurement (Hamann, Khagram, and Rohan 2008). Notwithstanding tense negotiations, these were ultimately welcomed by white business elites for their relatively conciliatory approach (Hirsch 2005; Tangri and Southall 2008). Earlier scholarship interpreted this as indicative of increasing alignment between white big business, BEE elites and a business-friendly ANC leadership, around maintaining a liberal economic framework then producing rapid growth (Iheduru 2008; Marais 2011; Mbeki 2009; Tangri and Southall 2008; Taylor 2007). As Ponte, Roberts, and Van Sittert (2007) argue, within this framework, BEE depoliticised redistribution, rendering it ‘technical, set within the limits of codification, measurement intervals and systemic governance’, beyond popular-political demands. Taylor (2007) argued BEE was a component of a durable ‘reform coalition’ comprising liberal ANC factions and white business elites, furthering liberalisation. Cooperative state–business relations, Taylor argues, were bolstered by the accumulation opportunities for black elites created by BEE, generating a ‘“tunnel effect” that insulates the state from popular pressures’ ( Ibid. , 191). For critics, this was at the expense of the poor, signalling a shift from racial to ‘class apartheid’ (Bond 2005), as black business elites were ‘absorbed into the existing white economic elite … adopting this group’s ways of conspicuous consumption, attitudes and norms’ (Gumede 2008, 296). The following sections examine how this process evolved in mining.
BEE in the mining industry
Mining has been a central focus of BEE policymaking. For the ANC, it constituted a strategic element of the economy’s ‘commanding heights’ alongside finance and state-owned enterprises (SOEs) (Southall 2007). Contributing 14% of GDP and 600,000 jobs in 1994, mining diminished relative to economic output in the preceding decades, to 8% of GDP in 2017. Nonetheless, mining has remained pivotally important economically, due to the rents it generates in boom periods – notably during 2004–08 – its linkages with other heavy industries and services, and its outsize contribution to export revenues, approximately 60% of which derives from minerals, metals and chemicals.2 Mirroring wider patterns, mining ownership during apartheid was concentrated among a handful of conglomerates. Despite complex restructuring processes, this concentration continued in the early democratic period, with mining’s capital intensity forming a formidable entry barrier. Voluntary BEE deal making created some significant new black-owned mining enterprises but undershot expectations, compelling legislation. The result was the 2002 Minerals and Petroleum Resources Development Act (MPRDA), and the 2004 Mining Charter (henceforth, MC1) deriving from it. PGMs were a particular focus for policymakers who saw them as having the best BEE potential (Capps 2012, 323). First, due to anticipated demand growth from the automotive industry as gold mining waned. Second, because of South Africa’s dominance of global production, with the country holding 80% of the world’s platinum reserves (Ibid.; Cawood 2004; HSBC 1999). However, here as for other commodities, concentrated ownership was an obstacle. Following mid 1990s conglomerate restructuring, PGM mining was dominated by three companies: Anglo-American Platinum (Amplats), created from JCI’s PGM assets and controlled by Anglo American; Impala Platinum (Implats), eventually unbundled from Gencor; and Lonmin, created from London-listed Lonrho’s mining subsidiaries. They accounted for 98% of 1997 output – 55%, 25% and 18% respectively – with similar dominance of resources.3 South Africa was among the few jurisdictions with private ownership of subsoil minerals, and during apartheid mining conglomerates acquired large portfolios of undeveloped mineral rights (Capps 2012).
Government accused mining companies of exclusionary, anti-competitive hoarding, which the MPRDA addressed by nationalising mineral resources and requiring companies to apply for ‘new-order’ mining licences ( Ibid. ). A ‘use-it-or-lose-it’ principle meant unutilised rights risked confiscation (Cawood 2004). These licences provided a means to compel transformation, and were contingent on fulfilling MC1’s BEE targets. This included HDSA ownership of 15% by 2009 and 26% by 2014, alongside targets for human resource development, employment equity, migrant labour, community development, housing, procurement and beneficiation (DME 2004). The ownership target, the most significant element in economic terms, was for ‘equity equivalent’ ownership. Besides selling shares to BEE partners, the target could be met by disposal of mining assets. In concessions to the Chamber of Mines (CoM), the mining trade association, transactions would be willing-buyer-willing-seller, at fair value, with government ‘facilitating’ a market-driven process ( Ibid. , 7). This compromise was hard-fought, with some within the Tripartite Alliance favouring tougher measures. A leaked June 2002 draft with 51% ownership requirements shook financial markets, resulting in the more modest 26% final target (Iheduru 2008, 355). Mining companies, especially Amplats, fiercely opposed resource nationalisation and BEE-ownership targets, threatening investment strikes (Amplats 1995, 1999). However, as Capps (2012, 325) notes, the CoM faced ‘an informal yet relatively cohesive “empowerment alliance”’ comprising the ANC, SACP, trade unions and nascent black business associations. Differing on details, they were unified on fundamental principles with ‘a strongly nationalist character [that could] confront monopoly capital with a popular front claiming to represent the black majority’, and thereby force concessions (Ibid.). The outcome was therefore much in keeping with the broader compromises of the post-apartheid transition discussed in the previous section (Ibid.).
Exchanging financial for political capital: implementing the Mining Charter
While generally perceived as a successful damage-limitation exercise, the implications of MC1 for industry were significant. The Charter introduced a range of obligations towards affirmative action in their workforce, procurement and community development through Social Labour Plans, a fuller exploration of which is beyond this article's scope (see Farrell, Hamann, and Mackres 2012; Rajak 2011). Most significant from a financial perspective and to the wider politics of economic transformation were the ownership requirements. This section analyses the responses to this aspect of the Charter. Using examples from PGM mining, it presents a brief typology of BEE deals to illustrate the range of BEE entrepreneurs and forms of ownership transfer.
Of the major PGM mining companies, Amplats had most to lose from ‘use-it-or-lose-it’. Its spread of undeveloped rights, particularly on the Bushveld Complex’s Eastern Limb in the former Lebowa homeland, gave it ‘unique operational flexibility’ (Amplats 1995, 7), and almost double Implats and Lonmin’s combined reserves. Compelled to be proactive, it agreed with the Department of Minerals and Energy (DME, now DMR) to relinquish mining rights and establish BEE partnerships in advance of the MPRDA (HSBC 2000). Its early deals provide the typology. The first involved the debt-financed purchase of shares in an existing operation by BEE entrepreneurs. In 2000, Amplats underwrote former Gauteng premiere and Robben Island inmate Tokyo Sexwale’s Mvelaphanda Resources’ (Mvela) purchase of 22.5% of Northam Platinum, which Amplats majority-owned (RBC 2001). Mvela conducted further deals in platinum, gold and diamonds, at one stage controlling South Africa's fifth-largest PGM resource-base (Amplats 2014). This model of debt-financed, vendor-backed purchases of minority shareholdings by a politically connected cross-sectoral holding company was replicated throughout mining and elsewhere, notably with Cyril Ramaphosa’s Shanduka. For mining companies, this involved the implicit exchange of financial for political capital. However, the range of BEE entities is diverse, with important divergences in form and strategy. Amplats’ second transaction involved the BEE entrepreneur entering as a joint-venture partner. In 2001, Patrice Motsepe’s African Rainbow Minerals (ARM) took a 50% stake in Amplats’ undeveloped, Eastern Limb Modikwa project, building on earlier successful entry to mining services and acquisition of gold mines in 1990s BEE deals. Rather than an investment holding company, Motsepe created an operational mining company. Joint ventures enabled ARM to hurdle several entry barriers: access to reserves, smelting and refining capacity, and technical expertise. The model was, initially, favoured by the DME for these reasons. There were also benefits for Amplats. Besides sharing capital expenditure, it prevented dilution of existing shareholders (Macquarie 2007). ARM became an established, JSE-listed multi-commodity miner and an exemplar for successful BEE entrepreneurialism in mining alongside Exxaro in coal and iron ore.
Amplats’ third deal involved a corporatised traditional authority, the 2002 50:50 joint venture with the Royal Bafokeng Nation (RBN) for the Bafokeng-Rasimone Platinum Mine (BPRM) on communally owned land. Much of South Africa’s PGM deposits underlie densely populated, impoverished former ‘Homelands’, and several other traditional authorities have followed the Bafokeng in becoming BEE shareholders. For traditional authorities, equity ownership appeared more lucrative than existing royalty schemes, while for mining companies, the arrangement promised improved community relations and land access, with local political elites helping secure the operating environment. Developmental outcomes have been frequently disappointing, with a range of political conflicts over traditional authority governance, discussed extensively elsewhere (see Capps and Malindi 2017; Capps and Mnwana 2015; Manson and Mbenga 2014; Mnwana 2015). The RBN was the pioneer and most successful. The BPRM joint venture was in 2009 restructured into JSE-listed RBPlats, the first majority black-owned PGM-focused mining company, and one of the largest black-owned mining companies on the JSE. A 2007 royalty-to-equity deal made it Implats’ largest shareholder with a 13% stake. Finally, DMR (2009) concerns over disproportionate gains accruing to a handful of BEE entrepreneurs – discussed below – saw smaller community and employee share-ownership programmes (ESOP) widely introduced.
The ‘big three’ PGM miners used differing combinations of the above to meet BEE ownership targets. Details of the transactions, twelve for Amplats, five for Lonmin and seven for Implats, are too lengthy for inclusion (see Gqubule 2017), but according to CoM research, weighted by value, 49% of PGM BEE-ownership transactions have involved BEE entrepreneurs, 46% ‘communities’ (predominantly traditional authorities) and 5% ESOPs. This compares with 60%, 29% and 11% for the mining industry as a whole (CoM 2015b). BEE deals reflected strategic opportunism and convenience, particularly as deadlines loomed, with industry interviewees suggesting that securing BEE partners with the requisite financial capacity was challenging. Coinciding with the mid 2000s commodities boom, this made mining a key contributor to the formation of the BEE elite examined in earlier literature. As the next sections shows, however, in more recent years there have been intensifying disagreements over the measurement, quality and extent of BEE ownership transfers, while inherent weaknesses of the transfer mechanisms have been exposed by the cyclical downturn in commodity prices.
Financialised redistribution and the fragility of BEE
As a redistributive mechanism, BEE equity transfer was considered relatively market friendly because rather than direct state expropriation of assets or increased taxation, transfers of ownership would hypothetically take place through sale at fair value, on a willing-buyer-willing-seller basis. Share ownership would, hypothetically, align BEE partners’ interests with existing shareholders around corporate imperatives. Deals were typically highly leveraged, with BEE partners usually unable to purchase shares with cash, and redistributive outcomes contingent on equity returns. BEE was therefore a highly financialised form of redistribution. Figure 1 provides a stylised representation of the mechanism. BEE partners commonly purchased shares via a special purpose vehicle (SPV) with credit from commercial banks, development finance institutions or the vendor mining company. The shares purchased acted as collateral, with dividend payments enabling loan repayment and, subsequently, income for BEE beneficiaries.
This model functions when companies generate sufficient surplus to pay dividends and share prices are stable or rising, but it is inherently vulnerable to downturns. Without dividends debt repayment stalls, while falling share prices erode collateral value, entailing covenant breaches with some commercial lenders. Most BEE deals were struck during the 2004–08 boom years, entailing high purchase prices and accompanyingly high debt burdens (Figure 2). Post-2008, mining equities slumped, falling further still post-2014.
Mining BEE transactions and values for JSE Mining Index and JSE Platinum Mining Index, 2000–14. Sources: Chamber of Mines; Bloomberg.
The slump has been particularly severe in PGMs, with falling metal prices hitting rising costs, profits declining and dividends halting (Table 1). These financial strains have contributed to severe labour and community conflict, which has been met with harsh repression, most significantly in the Marikana massacre of August 2012 (Bolt and Rajak 2016). Particularly in the case of traditional authority BEE partners, these dynamics intersect with the financial difficulties engulfing key BEE deals (Capps and Malindi 2017). This heightened the distributive pressure on companies’ dwindling cash flows, and increased socio-political tensions in mining areas as anticipated benefits from BEE evaporated. The following discussion examines some illustrative cases.
| Amplats | Implats | Lonmin | ||||
|---|---|---|---|---|---|---|
| Net profit margin | Ordinary dividends paid | Net profit margin | Ordinary dividends paid | Net profit margin | Ordinary dividends paid | |
| % | Rbn | % | Rbn | % | Rm | |
| 2002 | 28 | 5.9 | 39 | 2.3 | 37 | 0.6 |
| 2003 | 13 | 2.7 | 29 | 2.3 | 14 | 0.9 |
| 2004 | 13 | 1.6 | 25 | 1.5 | 17 | 1.1 |
| 2005 | 19 | 2 | 42 | 1.4 | 18 | 0.8 |
| 2006 | 30 | 4.9 | 25 | 5.5 | 23 | 0.7 |
| 2007 | 27 | 12.3 | 23 | 3.1 | 21 | 0.6 |
| 2008 | 29 | 13.8 | 47 | 6.1 | 25 | 0.8 |
| 2009 | 8 | 6 | 23 | 7.8 | −30 | 1.2 |
| 2010 | 22 | – | 19 | 1.9 | 8 | 1.4 |
| 2011 | 7 | 3.1 | 21 | 2.5 | 16 | – |
| 2012 | −16 | 0.5 | 16 | 3.4 | −34 | – |
| 2013 | −3 | – | 4 | 0.6 | 13 | – |
| 2014 | 1 | – | 0 | 0.4 | −21 | – |
| 2015 | 0 | – | −13 | – | −147 | – |
| 2016 | −17 | – | 0 | – | −36 | – |
Source: Company accounts, author’s calculations.
Lonmin’s has been the most high-profile example. Incwala Resources was formed in 2004 to purchase 18% of Lonmin’s operating subsidiary, Lonplats, from Implats (Lonmin 2004, 5). BEE investors, some closely linked to Mbeki’s government, purchased 53% of Incwala. Incwala struggled with indebtedness after 2008 (Investec 2010), and Lonmin facilitated Incwala’s takeover by Ramaphosa’s Shanduka with a US$304 million loan, to avoid repeat problems with inflexible commercial lenders (Lonmin 2010). Critics have presented this as the apogee of reciprocal rent distribution to a politically connected elite, particularly following controversy over Ramaphosa’s role as a Lonmin board-member at the time of the Marikana massacre (Alexander 2013). Lonmin’s unprofitability combined with the deal’s leveraging entailed limited economic benefit for BEE partners, such that Phembani, merged with Shanduka, stilled owed Lonmin US$376 million in 2016 due to accrued interest (Lonmin 2016). One of Amplats’ key BEE deals was the 2009 vendor-financed sale of 51% of the Bokoni mine to Atlatsa Resources (then Anooraq). For R3.6 billion, Atlatsa gained the country’s third-largest PGM resource-base, 24% of Amplats’ (2004) resources (Amplats 2010; Credit Suisse 2012). However, struggling with costs, Bokoni has been perpetually unprofitable with low PGM prices, and remained heavily indebted as a result. Repeated financial restructurings culminated in Amplats writing down their entire R3.2 billion debt and equity interests in the project (Amplats 2015, 52). At the time of writing, the mine is under care and maintenance.
This and similar instances produced many BEE vehicles that are essentially insolvent, but sustained to preserve vendor mining companies' BEE credentials. Elsewhere, BEE partners have exited, leaving companies with diminished BEE ownership. Notably, the aforementioned Mvela’s BEE shareholders sold their stake in Northam during 2009–12 as share prices plummeted, leaving Northam with only 4% BEE ownership (Northam 2015). Under DMR pressure, Northam subsequently re-empowered to reach 26%. This relates to a significant point of tension in the Charter’s interpretation, discussed in greater detail in the following section, concerning questions of ‘once-empowered-always-empowered’ (OEAE). That is, do companies retain empowerment credentials when BEE investors exit, or must they ‘re-empower’? Difficult economic conditions in mining after 2014 emphasised the importance of this question. For example, ARM’s sliding share price briefly put its BEE trust into financial difficulties, risking its black shareholding falling below 50%, and with it the empowerment credentials of joint ventures with Amplats and Implats (Creamer 2016). Similar issues accompanied RBH’s sale of 5% of Implats, shifting capital out of mining into more lucrative opportunities provided by financial services, but putting Implats below 26% BEE ownership (Seccombe 2016). Essentially, exiting investments is a key means of BEE beneficiaries translating ownership into income; indeed, during periods of low profitability when no dividends are paid, it may be the only means. However, this exit risks undermining the fundamental aim of BEE in deracialising control of the economy. This problem came to produce severe tensions between big business and government over the Charter.
Conflicting interpretations of change and deteriorating state–business relations
The challenges discussed in the previous section produced complex outcomes for ownership patterns. Interpretation of whether this represented compliance with the Charter and meaningful transformation has been a matter of considerable dispute. For the CoM, the industry exceeded MC1’s 2014 ownership targets, with the organisation claiming that 39.5% BEE ownership in PGMs weighted based on value transferred, and 38% for the mining industry as a whole (COM 2015). From one angle, the sector has diversified considerably. For example, in PGMs the big three’s wholly owned mines’ output share declined from over 90% in 1997 to 60% in 2015 (Figure 3). By 2016, nine additional companies were producing, with four more developing mines.4 Other commodity groups, notably coal, have also seen a wide range of new entrants.
However, for critics, most notably the DMR, the outcomes were unsatisfactory (see also Gqubule 2017). This relates partly to disputes over measurement explored in the next section. Significantly though, the piecemeal diversification of ownership did not produce meaningful transformation of control given the lack of fully independent black-owned mining companies operating at significant scale. In PGMs, while all companies came to include BEE shareholders, by 2016 there were only three significant majority black-owned producers, ARM, RBPlats and the moribund Atlatsa, collectively accounting for approximately 8% of output. The only serious challenger to their concentrated power has been Sibanye, a global top-10 gold producer and now the third-largest platinum producer after a recent PGM mine-acquisition spree. The company is in the process of acquiring Lonmin at the time of writing. Crucially, the big three retained control over refining, a strategic node in the PGM value chain with massive financial and technological barriers to entry. Typically, joint ventures or asset disposals were accompanied by offtake agreements to pass output through these refineries, giving the big three additional revenue and strategic control.
Furthermore, ‘black-owned’ means only formal majority ownership. In practical terms, the black-owned PGM miners have remained tightly entwined with the big three. In Atlatsa’s case, Amplats owned 23% of the company, 49% of its Bokoni mine, and extended nearly R2 billion of loans (Amplats 2016). While significant, ARM’s PGM production is via joint ventures with Implats and Amplats, which, analysts argue, allow the latter to halve risky mining capital and operational expenditure but still capture downstream refining revenues, ‘utilising more fully [their] vast beneficiation machine at a marginal cost’ (JPMorgan 2005, 3). Critics, notably the DMR (2009), argued BEE partnerships at subsidiary rather than group level restricted BEE partners’ strategic influence. For RBPlats, up until a 2018 sale, Amplats owned 33% of its operating subsidiary and 12.6% of the holding company – a substantial portion of incomes alongside a refining offtake agreement (Amplats 2010).
The faltering BEE progress in PGMs is reflected across other mineral commodity groups. According to Gqubule’s (2017) calculations, gross black ownership of the JSE’s top-25 mining companies’ South African assets fell from 16% to 12% between December 2014 and December 2016. Other commodities have been worse than PGMs for the emergence of sizeable majority black-owned firms ( Ibid. ). The arguable exception is coal, where entry barriers are lower and Eskom’s increasing BEE procurement a crucial enabler for smaller companies. Additionally, Exxaro was a standout BEE success as the largest majority black-owned mining company and one of the big-five coal producers. However, it lost its majority black-owned status as its BEE-SPV unwound to settle debts (Exxaro 2016), and the emerging black-owned mining companies remain small-scale relative to the five coal majors which account for 85% of output.5
It is important to note in addition that DMR policymakers’ focus on BEE asset transfers as the key strategic challenge for mining corresponded with relative neglect of alternative approaches to redistribution and transformation, such as more progressive fiscal measures or minerals-based industrial policy (see also Zalk 2017). While mining as a proportion of GDP decreased from close to 20% in the early 1980s to 8% in 2013, mining as a proportion of state tax revenue declined from 29% to just 2.5% in 2013–14 (Davis Tax Committee 2015; IMF 2015). Corporate income tax rates declined from 50% in the 1980s, to 28% in 2007, while the industry benefits from redemption of much of its capital expenditure against taxable income (Davis Tax Committee 2015, 58–59). Consequently, the IMF (2015, 9) observed South African mining ‘receives generous tax treatment relative to other sectors’. As such, effective taxation rates remained low, even at the peak of mining sector profitability in 2008 (Table 2).6 Since 2010, a royalty has been levied in fulfilment of the MPRDA,7 but this was too late to capture the spectacular resource rents of the 2000s boom (ANC 2012, 36–37). Proposals for a resource rent tax by a 2012 ANC study group have not been adopted ( Ibid. ). Meanwhile, minerals-based industrial policy initiatives to spur enterprise development and create manufacturing employment in mineral value chains have received relatively little serious attention, with key parts of the mining capital goods sector suffering increased import penetration (ANC 2012; Jourdan 2014; Morris, Kaplinsky, and Kaplan 2012, 180–182).
| Coal | Platinum | Gold | Iron ore | |
|---|---|---|---|---|
| 2008 | 18.7 | 7.1 | 14 | 18.9 |
| 2009 | 15.3 | 11.4 | 17.1 | 15.8 |
| 2010 | 7.3 | 6.3 | 2 | 17.5 |
| 2011 | 11.6 | 12.4 | 8.9 | 15.8 |
Source: Davis Tax Committee.
In summary, the first decade of legislated BEE in mining failed to produce the anticipated outcomes. Industry compliance with the Charter produced a patchwork of highly leveraged minority shareholdings and a handful of smaller companies that have remained closely tied to dominant players, but nothing quite resembling the envisaged patriotic bourgeoisie. The financialised model of redistribution produced a predominantly fragile form of empowerment that proved vulnerable to the inevitable volatility of mining industry investment. As the next section shows, these shortcomings had political ramifications as the government during the latter years of the Zuma presidency sought to adopt a more assertive approach to BEE.
Contesting transformation: the politics of reforming the Mining Charter
Alongside the processes described in the previous section, the Charter and mining companies’ compliance record thereof received escalating criticism. This derived not only from the DMR, but also nationalist ANC factions and elements of the black business community. Besides specific responses to the Charter, this also reflected wider processes of factional and ideological realignment during the Zuma administration. Zuma’s election as ANC leader drew support from a complex coalition comprising leftist factions and trade unionists seeking a more interventionist economic strategy, alongside nationalists and black business organisations’ interests frustrated by the slow pace of transformation (Southall 2009). As early analysis anticipated, the durability of the gradualist, market-oriented approach to intra-elite redistribution embodied by conventional BEE policy hinged on accelerated economic growth and employment generation (Butler 2011; Freund 2007; Ponte, Roberts, and Van Sittert 2007). However, following the end of the commodities supercycle, the South African economy entered a prolonged stagnation with rising unemployment. Real-terms GDP per capita increased by just 1% between 2008 and 2017, while unemployment grew from 23% to 28%, or 36% by its expanded definition.8
Mounting popular redistributive pressures became an increasingly pertinent challenge for the maintenance of ANC hegemony. Mining has featured prominently at various stages, initially in demands articulated by the then ANC Youth League President, now Economic Freedom Fighters leader, Julius Malema, who called for wholesale mine nationalisation and denigrated existing BEE models as white controlled (Serino 2009). Nationalisation was ultimately rejected as ANC policy (ANC 2012) with Malema subsequently expelled and the business-friendly 2012 National Development Plan reaffirming, rhetorically, commitment to a liberal economic framework. However, this heralded growing ideological and factional division over the pace and method of economic transformation which would define the latter years of the Zuma presidency.
With this context in mind, the article examines in closer detail government’s criticism of the Charter and attempts to accelerate transformation through reforms to it. A starting point is the DMR’s (2009) Charter review, which highlighted diverging perspectives of big business and government not simply around measuring BEE, but over questions of what constituted meaningful empowerment. The DMR argued BEE ownership of the mining industry was ‘at best’ 9%, below the interim 15% target set for that year. Besides this, however, the DMR criticised companies for ‘insidiously’ integrating BEE partners at subsidiary level, and thereby limiting their strategic influence, and for concentrating deals among ‘a handful of black beneficiaries’ contrary to MC1’s ‘spirit and aspiration’ ( Ibid. , 17–18). High deal leveraging meant, it said, that many BEE partners realised limited tangible benefits because cash flows were diverted to debt repayment, in the manner described in the previous section ( Ibid. ).
Subsequent negotiations between the DMR, the CoM, trade unions and the South African Mining Development Association (SAMDA) – representing the interests of black-owned junior mining companies – produced a moderately revised 2010 Charter (MC2). This was accompanied by a stakeholder declaration acknowledging the ‘unsatisfactory’ progress on transformation (DMR 2010a). MC2 advanced ownership requirements, defining ‘meaningful economic participation’ to include ESOPs and communities, and ensuring a percentage of cash flows reached BEE partners before debt repayment (DMR 2010b). Additionally, all BEE beneficiaries would have full shareholder rights regardless of ownership structure ( Ibid. ). Overall, the reform was incremental, and proved short-lived in terms of its ability to generate consensus.
The DMR’s (2015) Charter review reopened reform disputes. Contrary to the CoM’s assessment, the DMR claimed only 79% of companies met BEE ownership requirements. While weighted (problematically, by employment) average BEE ownership was 32.5%, only 20% (weighted) of right-holders met the DMR’s interpretation of ‘meaningful economic participation’ ( Ibid. , 13). Calculating ‘economic benefit’ as share value minus debt, plus dividends, the DMR argued 64% of right-holders had provided no economic benefit to their BEE partners ( Ibid. , 14). The CoM contested the DMR’s figures, though neither fully disclosed their methodology. Alongside definitions of ‘meaningful’ empowerment, disagreement centred on OEAE, with the Charter equivocal enough to support multiple interpretations. While government asserted OEAE for post-2004 deals was void, the CoM argued the opposite.
Applying in 2015 for a High Court declaratory order, the CoM argued that removing OEAE ‘undermines the sustainability of the sector’, since the prospect of repeated re-empowerment deals, with attendant financing costs and dilution of ordinary shareholders, would deter investment (COM 2015a). The resort to legal recourse exemplified deteriorating relations between the state and big business in mining, which intensified dramatically with the DMR’s efforts to produce a third Mining Charter. A first draft was published in April 2016, with a finalised version gazetted in June 2017 (henceforth, MC3-1), only to be suspended following CoM legal challenges. After a further year of fraught negotiation, a final version (henceforth, MC3-2) was published in September 2018. Rather than simply technical and legal disputes between mining companies and the DMR over the design of BEE policy, this turbulent political process reflected widening factional power struggles and ideological fissures in the ruling Tripartite Alliance over transformation.
The DMR’s handling of MC3-1 typified a changing approach to BEE from government. ANC factions aligned with President Jacob Zuma argued for an acceleration of economic deracialisation, and a more adversarial approach to big business, as part of a broader critique of the post-apartheid constitutional settlement and earlier BEE policy approaches (Bhorat et al. 2017). Protagonists encapsulated this with the term ‘radical economic transformation’. Adopted after the ANC’s (2012) Policy Conference, this denoted a second phase of the NDR in which, having established political control, the Tripartite Alliance should more aggressively pursue economic transformation (Jonas 2015). The increasingly populist articulation of radical economic transformation (RET) by Zuma and his allies called for a more overtly confrontational use of state power against ‘white monopoly capital’, in particular through increased preferential procurement by state agencies and amendments to constitutional safeguards on private property around land restitution. Zuma’s factional support base also came increasingly to be defined by proximity to the Gupta family, Indian businessmen at the centre of multiple recent corruption scandals involving the subversion of SOE procurement (Bhorat et al. 2017). Given this, critics, including within the ANC, have argued this use of RET functioned as a disingenuous legitimation narrative for processes of individual enrichment and clientelism that have come to be discussed as ‘state capture’ (Jonas 2016). However, while entangled with these processes, this adaptation of RET echoed broader grievances (Bhorat et al. 2017). Notably, from elements of the black business community frustrated by the perceived lack of opportunities afforded by a highly consolidated private sector, the narrow range of actors benefiting from conventional BEE ownership transfer processes, and the instrumental approaches to BEE compliance from white businesses (Bell et al. 2018; Von Holdt 2018).
MC3-1 exemplified these tendencies. The Charter was promulgated by Mosebenzi Zwane, a controversial 2015 ministerial appointment whose most senior prior position was the Free State provincial government’s executive council as Member of the Executive Council for Agriculture. The appointment drew criticism from industry, opposition parties and some within the ANC (Mail&Guardian 2015). Besides questions over his inexperience, this related to Zwane’s links to the Gupta family through the Vrede dairy scandal, a Free State agricultural development project involving multiple financial irregularities relating to Gupta-linked companies. In one of the most high-profile ‘state capture’ scandals, a Public Protector investigation found Zwane to have improperly facilitated the sale of Glencore’s Optimum coal mine to Tegata Resources, a company co-owned by the Guptas and Jacob Zuma’s son, Duduzane (Public Protector 2016). Leaked emails revealed Zwane’s CV to have been sent to the Guptas prior to his appointment, while his predecessor, Ngoako Ramatlhodi, claimed he was sacked shortly after refusing to cancel Glencore mining licences to expedite the sale (Sole and Comrie 2017).
Zwane’s governing style was more adversarial and unilateral in its approach to big business, with his legislative agenda centred on radical reform to the Charter. In contrast with the extensive dialogue – albeit tense and fractious – leading to MC1 and MC2, the MC3-1 draft was published unexpectedly in April 2016, with limited prior industry consultation and a month for submission of responses (DMR 2016). The draft proposed a significant expansion of ownership requirements, most notably by terminating the OEAE principle. It was opposed by industry, and caused considerable alarm among investors and industry analysts (Deloitte 2016). However, the finalised MC3-1, gazetted in June 2017, pushed further. BEE ownership targets were raised to 30%, with mandatory 8% shares for communities and workers (DMR 2017a, 7–8). Existing right-holders below the 30% level would have one year to re-empower ( Ibid. , 10), with a requirement that this level be maintained following restructurings or share offerings ( Ibid. , 8). New prospecting rights, meanwhile, would have to be majority black owned ( Ibid. , 7). Additional measures sought to increase the financial benefits deriving from ownership. Notably, a minimum 1% of revenue would have to be paid ‘prior to and over and above’ ordinary dividends, alongside write-offs of all outstanding BEE partners’ debt after 10 years. Additionally, BEE shareholders would take ‘direct’ control over ‘transportation as well as trading and marketing of the proportionate share of the production’ ( Ibid. , 9). The Charter, Zwane argued, gave ‘practical expression to the meaning of radical economic transformation’ (Zwane 2017).
The CoM’s response was to pursue immediate legal recourse, precipitating a near unprecedented deterioration in state–business relations in the mining industry as big business mobilised to obstruct the ANC’s legislative agenda. A fortnight after the Charter’s publication, the CoM lodged an application at the High Court for an urgent interdict preventing the Charter’s implementation pending a judicial review. In its founding affidavit, the CoM argued the Charter was ‘unlawful’ and should be set aside. The reasoning is too complex to outline in the space available here, but relates to Zwane exceeding his powers as minister in the procedures leading to publication, contested interpretations of the MPRDA and numerous definitional ambiguities around ownership requirements (see Chamber of Mines vs Minister of Mineral Resources 2017a). Calling it an ‘unmitigated disaster’ for employment, the CoM threatened that mining companies would effectively freeze investment ( Ibid. ). These warnings around investment repercussions were echoed by key international capital market actors, notably credit rating agencies who surmised that ‘government is prioritising radical transformation even if this leads to weakening of the business climate’ (Fitch 2017; see also Moodys 2017).
The DMR’s answering affidavit attempted to refute the Chamber’s legal arguments, with Zwane calling the interdict ‘an attempt to block effective and meaningful participation of black persons in the mining and minerals industry’, and claims around economic impacts ‘overstated’ (Chamber of Mines vs Minister of Mineral Resources 2017b). This was accompanied by a threat to suspend all mining rights applications, halting new investment in the sector (DMR 2017b). However, the concessions ultimately came from Zwane, with an agreement in September 2017 to suspend implementation of the new Charter prior to completion of a judicial review, followed by the Chamber withdrawing its interdict. The approach taken by Zwane did not reflect a position of strength. As discussed above, when the CoM negotiated the MPRDA, it faced a broad-based, multi-class ‘empowerment alliance’ (Capps 2012). For MC3-1, however, no such alliance existed. Instead, to the contrary, the increased factional division and fragmentation that defined elite politics in the latter years of the Zuma presidency ultimately undermined this bargaining power. A range of groupings within the Tripartite Alliance came to oppose the Zuma-aligned faction’s approach to RET at a general ideological level, and in the practical implementation of MC3-1 more specifically. This included the more moderate, pro-business elements of the ANC seeking to defend the liberal economic framework and orthodox interpretations of good governance (Von Holdt 2018). In the approach to the ANC’s (2017) elective conference, this faction coalesced around then Deputy President Cyril Ramaphosa and Pravin Gordhan, following his sacking as finance minister in early 2018, and were joined by the Congress of South African Trades Unions and the SACP as alarm grew over the economic and electoral impact of state capture ( Ibid. ).
The Charter dispute became integrated with this intensifying factional and ideological contestation. Zuma, Zwane and his core ANC allies attempted to present MC3-1 as non-negotiable, while the Black Business Council called the document ‘effective and practical radical transformation’ (Solomons 2017). However, this was directly contradicted and ultimately undermined by Zuma’s opponents within the Tripartite Alliance. The Economic Transformation discussion document for the ANC’s (2017) Policy Conference, written by Ramaphosa allies, argued that the Charter should be rewritten to improve investment competitiveness ( Ibid. ). Ramaphosa himself subsequently suggested ‘both parties must go back to the drawing board’ to reach a consensus around shared interests (Creamer 2017). Then ANC Secretary General Gwede Mantashe circumvented Zwane to speak directly with the CoM about their concerns with MC3. Reflecting the significant differentiation within the black business community over transformation issues, public opposition to MC3-1 also stemmed from prominent black business leaders in the major mining companies constituting the CoM membership, notably the then CoM President and Exxaro CEO, Mxolisi Mgojo, and Steve Phiri, the CoM Vice President and RBPlats CEO. The latter characterised the Charter as transformation ‘at the expense of the downfall of the economy’ with the risk that ‘capital will be chased away’ (Seccombe 2017).
This approach ultimately prevailed following the ascendancy of Ramaphosa as ANC president in December 2017 and to the South African presidency in February 2018. Ramaphosa sought to mark a departure from Zuma’s RET with conciliatory gestures to domestic big business and international investors, and one of the first significant moves to this effect was to scrap MC3-1. With Zwane swiftly fired and replaced by Mantashe as Minister of Mines – a move welcomed by big business – Charter negotiations were reopened. The redrafted document released in June 2018 significantly weakened the key ownership provisions of MC3-1 and removed several of its more contentious elements, including around the OEAE principle. However, it was rejected by the Minerals Council of South Africa (MCSA) – the rebranding of the CoM which took place in May 2018 – on the basis that key ownership requirements would prove too onerous for its members and repel investment (DMR 2018a; MCSA 2018a). However, responses from other key stakeholders such as the National Union of Mineworkers, mining community representatives and SAMDA were much the opposite. Further phases of renegotiation included complex consultations with a range of community groups and civil society actors to bolster procedural legitimacy. However, these took place in a context of heightened concerns over the severe deterioration of economic conditions, and the financial strains afflicting mining as a result of continued low commodity prices. As shown in Table 3, recent years have seen steady declines in mining employment, with over 70,000 jobs lost since 2012, alongside severe declines in investment.
| Mining employment | Mining gross fixed capital formation | Net foreign direct investment | |
|---|---|---|---|
| 000s | Rbn (2017 prices) | Rbn (2017 prices) | |
| 2010 | 499 | 93 | 40 |
| 2011 | 513 | 95 | 46 |
| 2012 | 525 | 95 | 17 |
| 2013 | 510 | 100 | 20 |
| 2014 | 496 | 100 | −24 |
| 2015 | 480 | 84 | −58 |
| 2016 | 458 | 71 | −35 |
| 2017 | 465 | 81 | −80 |
| 2018 | 454 | NA | NA |
Source: SARB Quarterly Bulletin; DMR.
The final published Charter gazetted in September 2018 contained a series of further significant climb-downs from the ownership requirements tabled in MC3-1 and Mantashe’s June 2018 draft. OEAE was partially reaffirmed, with mining right-holders achieving their ownership targets and then subsequently losing empowerment partners still to be recognised as compliant (DMR 2018b, 13–16). Albeit, these ‘continuing consequences’ would, significantly, no longer be recognised for the renewal or transfer of mining rights. BEE ownership levels were retained at 26% for existing mining rights, and raised to 30% only for new or renewed mining rights ( Ibid. , 14). Requirements for majority-BEE ownership in new prospecting rights were dropped, alongside the requirement for 1% revenue to be paid to BEE beneficiaries prior to over and above ordinary dividends, the requirement for BEE beneficiary debt write-off after 10 years, and the requirement for direct control by BEE beneficiaries of the trading and marketing of their attributable shares of production ( Ibid. ). These all represented major concessions to industry demands. Nonetheless, MC3-2 did go further than MC2 on ownership in retaining the contentious mandatory percentage shareholdings for workers and communities, albeit lowered from 8% to 5% ( Ibid. , 14–16). This was as ‘carried interest’, essentially meaning these BEE partners would not be required to purchase their shares up-front. Furthermore, this level would have to be maintained in any subsequent equity capital raising. This and the requirement to re-empower on renewal or transfer of mining rights remains at the time of writing the major concern for the mining investors, with negotiations ongoing as to how MC3-2 will be implemented and a further challenge from the MCSA mounted in 2019. Both Mantashe and the MCSA have sought to represent the new Charter as a return to compromise and consensus-building which will bring stability to the mining industry. However, with the underlying causes of recent years contestation of BEE in mining largely unresolved, this may prove temporary.
Discussion and conclusions
The financialisation of the South African economy (Ashman, Fine, and Newman 2011), the internationalisation of domestic big business and consequent exposure to international capital market sentiment have shaped and constrained the ability and inclination of the state to pursue radical redistributive measures. Reflecting concerns about the mobility of mining investment, the concessions ultimately obtained from the DMR by the MCSA around the third Mining Charter represent a significant exercise of power by big business which re-emphasises the saliency of these constraints. Combined with the readoption of pro-business rhetoric in the earlier phases of the Ramaphosa administration, it is a departure from the late Zuma era’s rendering of RET. Nonetheless, this does not necessarily mark a simple return to the state–business relations characterising earlier phases of BEE legislation in the Mbeki era. Instead, unresolved tensions remain over the trajectory of transformation in the mining industry, and indeed across the wider South African economy, both between established big business and government, and within the Tripartite Alliance, which remains riven with factional fragmentation.
The contestation around MC3 is markedly different to the processes anticipated by earlier political economy literature on BEE discussed in Section 1. This posited an increasing alignment between white economic and black political elites fostering cooperative state–business relations, with a deracialised elite coalition created by BEE defending the liberal economic framework. As the above analysis shows, however, the relatively coherent empowerment alliance which existed around MC1 has given way to a much more complex fragmentation of competing interests and ideological positions. This fragmentation corresponds with the emergence of intense factional contest within the Tripartite Alliance and increased differentiation within the black business class, and mounting dissatisfaction over the pace and extent of economic transformation. As has been argued above, this is in part a consequence of the shortcomings of conventional BEE policies as a form of financialised redistribution and the changing forms of inequality and exclusion it created. Rather than the envisaged ‘patriotic bourgeoisie’ aligning political and economic power, this has, despite ostensibly large transfers of wealth, predominantly created a fragile patchwork of black business interests in mining, holding highly leveraged minority stakes in established enterprises, with ultimate redistributive outcomes hinging on volatile equity market returns.
This generated a relatively small group of wealthy politically connected individuals among those benefiting from opportune timing in the commodity cycle for entry and exit from investments. However, in many instances inclusion of BEE partners has not translated into meaningful economic empowerment, either in terms of income streams or the ability to exert strategic control of the industry’s trajectory (Gqubule 2017). Independent black-owned mining enterprises operating at sectorally significant scale are few in number, notwithstanding some success stories such as RBPlats and ARM. The industry remains characterised by racialised inequalities, with high barriers to entry for prospective entrants. Mining is an extreme case in some regards given the unusual volatility of the industry and its enduring socio-political resonance deriving from its historic role in shaping patterns of inequality, but nor is it entirely unique. Economic and political power remain broadly misaligned here and in other parts of the South African society, and as such the underlying tensions which gave rise to the severe conflicts around MC3 remain unresolved.
Correspondingly, so too do the internal contradictions and ambiguities around the ANC’s overarching approach to transformation of private sector ownership. This has been defined by an uneasy balance between imperatives to accelerate deracialisation of private sector ownership in a manner which necessarily involves a degree of confrontation with established economic power, and an orthodox growth model predicated on maintaining a favourable investment climate (Tangri and Southall 2008). As the article has shown, there have been significant swings in this balance during recent years. Nonetheless, the fundamental mechanics of BEE policy have remained as the basis of the state’s attempts to reconcile these imperatives. As Zalk (2017) has argued, alternative approaches to economic transformation have been caught in this pincer between narrow nationalist approaches to BEE focused on ownership transfer, and the strictures of an orthodoxly liberal economic framework. Attempts to leverage South Africa’s still-vast mineral wealth for economic diversification, the development of new industrial capabilities and labour-intensive re-industrialisation have received limited policy attention relative to efforts to produce new tradable claims in the existing economic base ( Ibid. ; Bell et al. 2018; Jourdan 2014).