‘Not Wholly Justified’: The Deferred Pay Interest Fund and Migrant Labour in South Africa’s Gold Mining Industry, c.1970–1990

A little-known feature of the vast migrant labour system that supplied South Africa’s gold-mining industry was the Deferred Pay Interest Fund. For much of the 20th century, a portion of the wages owed to African mine workers was deferred and remitted to them only at the end of their contracts. This is well-known, but what happened to the interest that accumulated on these deferred wages remains virtually unknown. Mine workers did not receive this interest; it was, instead, deposited into a fund controlled by the mining industry. This article examines the operations of this fund in the Transkei in the context of the crisis in the migrant labour system precipitated by newly independent states refusing to supply further migrant labour to South Africa. This prompted the Chamber of Mines to reorient labour recruitment towards the South African bantustans, and the Transkei quickly became the most important source of labour for the mines in the 1970s and 1980s. Although the fund had a mandate to spend on welfare projects in labour-sending regions, we argue that patterns of spending clearly show how it was used to support the reproduction of the migrant labour system. Payments were used as patronage for local elites, upon whom recruitment depended, and for distributing propaganda for the mining industry. In contrast, payments were consistently directed away from education for able-bodied students, because education would reduce the pool of unskilled labour on which the gold industry relied. Money that, arguably, rightfully belonged to mine workers from the Transkei was used to perpetuate their dependence upon migrant labour to the mines.


Introduction
South Africa's gold-mining industry was at the centre of what Jonathan Crush, Alan Jeeves and David Yudelman memorably referred to as South Africa's 'labour empire', which stretched across the entire southern part of the continent. 1 At its peak, almost 500,000 migrant workers were recruited to the gold mines from as far away as Angola and Tanzania on temporary contracts and then, or so it was intended, returned to their regions of origin. All this required organisation on a grand scale, not only to serve the day-to-day labour requirements of the industry but to secure long-term labour supplies to mine South Africa's seemingly inexhaustible gold deposits.
The focus of this article is one seemingly small cog in this vast labour recruitment system: the Deferred Pay Interest Fund. We examine the operations of this fund in the Transkei, a region in what is now South Africa's Eastern Cape, during a time of crisis and reorientation in the gold industry's labour recruitment. We argue that this analysis offers insights into the functioning of the migrant labour system and the ways in which the mining industry sought to shape development in rural South Africa to support the perceived needs of the industry. The crisis was the loss, albeit temporary, of the main sources of African labour for the mines in the mid 1970s and the subsequent rapid emergence of the Transkei as the most important labour-sending region for the gold industry. How that industry sought to secure long-term labour supplies from the Transkei and elsewhere is an example of breathtaking corporate cynicism with few recent parallels.
Obtaining sufficient supplies of African labour was a persistent worry for South Africa's Chamber of Mines for much of the 20th century, as was the related problem of keeping those mine workers already recruited at work. 2 From early on, the mines faced a serious problem of desertion as African mine workers sought to escape harsh, often lethal underground workplaces. One response to this was the introduction of a system of deferred pay whereby a portion of mine workers' wages was disbursed only after completion of contracts. The deferred payment system is well-known and is briefly discussed below, but one important aspect of this system has been consistently overlooked: the Deferred Pay Interest Fund, which was established in 1918. 3 We show that part of establishing the monopsony of the Native Recruitment Corporation (NRC) from 1912that is, the NRC becoming the sole purchaser of mine labour in its jurisdictioninvolved the NRC wresting control of the rural credit market in the Transkei away from private recruiters and traders by legislatively reducing recruitment licenses, placing limits on advance payments and themselves financing advances. This connects to what some have called 'external judicial control', a legally based system in which creditors are favoured and debtors lose control over their financial actions. 4 In this way, we explore an historical aspect of the broader phenomena of what some have called 'credit apartheid '. 5 Although African mine workers were paid low wages individually, the sheer number of Africans working on the mines (see Table 1) meant that the total sum of wages was considerable. In 1980, for instance, African mine workers in the gold industry collectively received R868,422,347 in wages, equivalent to R25 billion in 2020. 6 Crucially, at the end of their contracts, African mine workers were paid only the cash value of wages owed to them and not the interest that had accrued on their withheld wages. The interest on deferred wages arguably rightly belonged to those mine workers but was never paid to them. Interest that accrued on wages was instead paid into the Deferred Pay Interest Fund. In a 1976 article on Anglo American, Martin Plaut noted the existence of the fund in passing and that this fund represented a significant source of cheap finance for the mining industry and thus that 'the question must be asked as to precisely what use the above funds are put'. 7 Broadly, these funds were spent in two ways. For mine workers recruited outside South Africathe majority of the mining workforce until the 1970s and for whom deferred pay was compulsorythe interest was paid to the colonies, and later states, from where they were recruited. These states regarded the interest payments as a kind of tax receipt. For mine workers recruited within South Africa, the NRC and its successor organisation, The Employment Bureau of Africa (TEBA), controlled payments from the Deferred Pay Interest Fund with assistance from the Native Affairs Department (NAD). 8 The NRC and TEBA were both established by the Chamber of Mines and paid careful attention to how the money was spent. Payments were allocated by the fund to projects under their mandate to finance 'welfare work'. Identifying spending patterns by the fund in the Transkei therefore offers insights into how the mining industry sought to shape life in rural South Africa. As will be seen, the fund spent generous sums on state-appointed chiefs and schools for the disabled but consistently refused to fund schools for able-bodied children or support local economic development.
We argue that the money disbursed by the Deferred Pay Interest Fund had a dual purpose. It was used effectively to bribe local elites in labour-sending regions and was deliberately directed away from education or economic development. Expanding education for rural Africans would ultimately reduce the pool of unskilled labour upon which the mines relied, while encouraging local industry or other enterprises in labour-sending regions would discourage labour migration by providing other economic opportunities for African men in these areas. The kind of 'development' promoted by TEBA helped to reproduce poverty in rural South Africa by deliberately directing investment into non-productive areas. In short, the Deferred Pay Interest Fund sought to perpetuate dependency upon the migrant labour system, and it did so by using workers' own money.
Harold Wolpe's classic argument on cheap labour power and apartheid asserted that 'the system of producing a cheap migrant labour force generated rural impoverishment, while at the same time it enabled extremely low wages to be paid to Africans in the capitalist sector' and that state policy towards labour reserves, critical for supplying African labour at a wage below its cost of reproduction, was centrally concerned 'with the control and supply of a cheap labour force '. 9 In this formulation, white industry is one of the intended beneficiaries of the policy, not an agent that shapes or enacts that policy. In this article, we stress that the mining industry itself was centrally involved in the reproduction of cheap labour in rural South Africa.
The focus of this article is on the Transkei and, in this sense, this article contributes to the rich historiography of the impact of the migrant labour system on the places that labourers came from. 10 However, this article concentrates on the recruitment agencies of the mining industry rather than the workers themselves or other residents of the Transkei, primarily because, as will be seen, labour migrants had no say over how this money was spent. There were never any African representatives on the board overseeing the expenditure of the Deferred Pay Interest Fund. Labour migrants themselves unwittingly shaped this system by depositing a portion of their wages with the deferred pay system, thereby providing the capital upon which interest was generated. The willingness of South African labour migrants, for whom deferred pay was not compulsory, to participate in this system and the benefits they perceived from this is intriguing but beyond the scope of this study. 11  the Transkei and other bantustans in the mid 1970s to meet the demands of the gold-mining industry, and on the insights that this generates for how the system functioned.
There are no public-facing reports on the Deferred Pay Interest Fund's expenditure or operations. Sources for this article are largely drawn from correspondence between the fund and district recruiters, the NAD and local organisations in the Transkei and minutes from NRC, TEBA and Chamber of Mines meetings. This material is housed in two places: the University of Johannesburg's Special Collections and the archives of TEBA itself, which still exists and has its head office in central Johannesburg. 12

Emergence of the Deferred Pay Interest Fund
South Africa's gold industry depended upon unskilled African labour, recruited in huge numbers to work on the Witwatersrand and later in the Free State. Indeed, such was the extent of this reliance that Jonathan Crush, Alan Jeeves and David Yudelman argued that, without high numbers of unskilled low-waged migrant labourers, 'there would never have been a deep-level gold-mining industry in South Africa'. 13 The mining companies required cheap labour to mine gold profitably largely because of the unusual cost structure of the industry. From 1902, at the conclusion of the South African War, all gold produced on the Rand was bought at a price fixed by the Bank of England. This had important consequences: it imposed a ceiling on costs and the mining companies could not pass increased production costs on to consumers. This left the industry highly vulnerable to cost inflation. 14 Mining companies had limited control over the cost of machinery, supplies and power, so the principal way to restrain production costs was to keep wages for African mine workers as low as possible.
The Chamber of Mines was therefore faced with the problem of obtaining the necessary number of recruits and ensuring that they completed their contracts without what might be regarded as the conventional mechanism of offering higher wages. Instead, the Chamber agreed to eliminate competition for labour between mineswhich could drive up wagesand established a monopsonic labour market. Before 1912when the Chamber established its own agencies to supply labour to the mines on standard contracts and wagesit had unsuccessfully attempted to create a monopsony on labour three times. 15 At this time, labour recruitment was largely in the hands of private recruiters, who offered credit and advances, including cattle advances, to prospective migrants. 16 There, prospective migrants were interviewed by a tax collector, and a year's or more of advances for tax payments was given to migrants upon signing mine contracts. 17 In Pondoland in mid 1904, total advances over £26 stood at £142 and, by the end of the year, the figure for total advances over £26 was £2,951. By mid 1905, the sum had risen to £17,264. 18 While private recruiters boosted the 12 The archives of TEBA, hosted at TEBA Head Office at 121 Eloff Street, Johannesburg, though in a good condition, have not been organised or catalogued by professional archivists, and so referencing material found there is challenging. We have in this article given as full a reference as possible as to the location of the material in this archive. supply of mine migrants, they exacted high prices, encouraged desertion and stimulated intra-industry competition. Traders and private recruiters were powerful and essentially controlled the rural credit system in the Transkei.
When the NRC was formed in 1912, it had two strategies: reduce advances and limit the number of private recruiters. In 1910, advances were limited by law to £2, implying a strong reduction in rural access to credit, and this dramatically weakened the power of private recruiters. 19 A 1912 proclamation limited cash advances to £2 in Bechuanaland. 20 In the Transkei, the NRC started to give traders and recruiters the capital to pay advances. 21 Furthermore, the 1911 Native Labour Registration Act and the NAD reduced the proportion of recruitment licences in the Transkei by 75 per cent between 1912 and 1918. 22 In this way, the NRC in the Transkei, backed by legislation and the NAD, essentially wrested control of the rural credit market away from private traders and recruiters. The deferred pay system helped to reconcile traders to the new credit regime because deferred pay ensured that wages would remain in the local Transkei economy.
The Chamber of Mines' two recruitment agencies were the NRC, which recruited from Basutoland (Lesotho), Bechuanaland (Botswana), South Africa and Swaziland (Eswatini), and the Witwatersrand Native Labour Association (WNLA), which recruited labour from elsewhere in southern Africa. The formation of a centralised recruitment system was complete by 1920 and oversaw a steady increase in the numbers recruited over the next 50 years, primarily from outside South Africa. 23 The creation of this monopsonistic recruitment system was unthinkable without institutional support from the South African state. The Chamber had persistently called for state assistanceon the grounds that private business had constantly failed adequately to organise and supply sufficient labourand now forged close relationships 'at every level' of the NAD. 24 The NAD worked closely with the NRC within South Africa, and mining companies' control over African labour was greatly assisted by restrictive legislation. The essential mechanisms were the enforcement of pass laws that prevented the free movement of those racialised as black Africans, the Mines and Works Act of 1911 and the Native Labour Registration Act, which made contract-breaking a criminal offence and limited mine recruitment licences. It was in this context that the deferred pay system was introduced as another way of controlling African labour.
The idea was first raised to reduce the desertion of African miners from Mozambiquethe largest single source of labour to the mines at the timeand an agreement between the WNLA and the Portuguese colonial authorities over deferred pay was reached in 1912, though fully implemented only in 1928. 25 The perceived benefits of such a scheme for all migrant mine workers caught the attention of H.O. Buckle, Chief Magistrate of Johannesburg, in his 1914 Native Grievances Inquiry, in which he advocated a system whereby migrant workers would be compelled to receive a portion of their wages (at least 50 per cent of total) as a lump sum after completing their contracts. 26 Buckle argued that the deferred pay scheme would ensure that wages remained in the rural areas, because wages would be disbursed after contract completion.
However, many migrant workers sent a portion of their wages to rural areas as remittances. In 1918, the year the deferred pay system began, remittances totalled £99,469. By 1931, admittedly a depression year, remittances had fallen to £14,655. 27 Instead, under the deferred pay system, the money was placed under the control of the Chamber of Mines. The mining companies and government officials offered a paternalistic justification for the scheme, stressing that African mine workers could not be trusted to handle responsibly the money paid to them. They claimed that the system was popular among mine workers themselves, as it prevented relations in rural areas from spending their remittances before they returned, and for those relations it prevented mine workers from squandering their wages in urban areas. 28 The point, however, was that the Deferred Pay Interest Fund removed control over wages and financial decision-making from Africans and represented what some have called external judicial control.
The deferred pay scheme won support from the South African government in 1914 and, having been endorsed by the NAD, was enshrined in law by 1917. 29 The system appealed to a wide variety of constituencies: families of migrants were guaranteed that some wages would return to the area, traders were assured a market for their goods and the Chamber of Mines did not have to risk deserting migrants defaulting on wages advanced to them. The system also bolstered South Africa's relations with its neighbours, as deferred wages became a major boost to the economies of neighbouring colonies. The deferred pay scheme neatly symbolises the deep institutional nexus in which the mine migrant labour system consisted. The system was operated at the behest of colonial governments in neighbouring countries such as Malawi, Lesotho, Swaziland and Mozambique for tax and domestic economic stimulation purposes. 30 The objectives and regulations of the deferred pay scheme were determined by the NAD. In 1917, the Secretary for Native Affairs directed that 'all monies accruing after allowance for a floating balance for working expenses [should] be invested in Treasury Bills', and, most importantly, that 'after meeting the costs of administration, any surplus will be devoted to matters connected with the welfare of Natives generally'. 31 Thus the deferred wages would provide a consistent, not inconsiderable source of funds for the new state by automatically purchasing state bonds, thus reducing the cost of South Africa's borrowing. The interest earned was allocated to 'welfare' projects. The Deferred Pay Interest Fund's spending decisions were determined by a Board of Control constituted by two NAD officials and three industry representatives. The Director of Native Labour was chairman, and the Minister of Native Affairs had to sign off payments personally. 32 The task of the Board was to allocate accrued interest to causes promoting welfare in labour-sending regions. It was the Board of Control's duty to determine what constituted 'welfare', and so, in studying these allocations, industry assumptions about its labour system emerge. This institutional environment proved remarkably stable for the next five decades. 33 It changed in 1977, when the NRC and WNLA merged and rebranded themselves. The words 'native' and 'recruitment' were dropped, and the new joint organisation became The Employment Bureau of Africa. The Deferred Pay Interest Fund became the TEBA Savings Trust, which held the interest from deferred pay, and disbursement was controlled by the Donations Appeal Committee, which first met on 23 June 1977. 34 What prompted this institutional change was a crisis in the migrant labour system in the mid 1970s, when the gold mines faced mass withdrawal of labour. The immediate triggers were a plane crash in April 1974 in Francistown, Botswana, which killed 74 Malawian mine workers and prompted Malawi's President Hastings Banda to suspend labour recruitment and, that same month, the overthrow of Portugal's Estado Novo and the installation of a new government willing to negotiate the independence of Portugal's African colonies, including Mozambique. Malawi and Mozambique were the two largest sources of labour for the gold mines at the time, and their simultaneous loss would have crippled the industry. The Chamber of Mines feared that an independent Mozambique would halt labour recruitment and, though this did not occur, the South African state sought to reduce labour recruitment from Mozambique as part of its general strategy of undermining the newly independent state. The wider context of African states achieving independence bears emphasising. A pattern emerged whereby newly independent states withdrew from the migrant labour system after independence, as occurred in Tanzania (1961), Zambia (1964) and Zimbabwe (1981). 35 Name-change notwithstanding, spending patterns and policies informing spending remained continuous and point towards long-term continuities in recruitment. The guiding policy to spend the interest on welfare stayed in place. 36 Substantial sums were paid out to institutions considered favourable to TEBA, and donations to anything that risked reducing future labour migration were ruled out.

The Transkei and the Migrant Labour System from the 1970s
In the 1970s, a confluence of circumstances threatened to plunge the gold-mining industry into crisis, namely the sudden withdrawal of migrant workers from outside South Africa, a surge of African labour militancy within South Africa and the transformation of the economics of the industry brought about when the USA abandoned the gold standard in 1971. By the 1970s, the gold industry had become increasingly reliant on workers recruited from outside South Africa. Persistently low wagesreal wages for African mine workers actually fell between 1889 and 1969 37meant that the mining industry could not compete with other sectors as South Africa's economy expanded rapidly in a post-war economic boom. Workers from South Africa had been a minority since the early 1940s, and their proportion shrank to a low of 20 per cent of the African mining workforce by 1974. At this juncture, the mining industry faced a sudden crisis with the withdrawal of Malawian workers and the threat that Mozambican workers would withdraw. The impact was dramatic. This abrupt departure forced a major reorientation of the migrant labour system. Faced with the loss of two of the major labour-sending regions, the Chamber of Mines turned to domestic sources for the first time in a generation. 39 Domestic labour sources were primarily the areas known as bantustans. Some mines experimented with recruiting African workers from urban areas, but this was unsuccessful, as urban residents had options other than the low wages and grim communal housing offered by the mining companies. 40 This was not the case for many rural South Africans. The apartheid state had forcibly resettled hundreds of thousands of black Africans into so-called homelands in the 1960s and 1970s, which were then granted fictitious independence and came to be popularly derided as bantustans. There were, by design, few resources, quantities of fertile land or economic opportunities in these homelands, and this was especially true of the Transkei. Instead, many inhabitants dumped or born into these regions had few options aside from engaging in long-distance labour migration. By 1970, 70 per cent of the Transkei's GNP came from migrants' wages. 41 In the years that followed, dependence on remittances from migrant workers became more concentrated on the mining industry. In the late 1980s, Francis Wilson and Mamphela Ramphele calculated that wages from the mines were by far the most important source of income for households in the Transkei. In households with incomes less than R1,500, 66-71 per cent of income came from wage remittances from mining, compared to 11-15 per cent from wages from jobs in the Transkei. 42 One reason for this was that the long-term declining value of wages for African mine workers was reversed in the 1970s. Wage restraint was abandoned in the face of an upsurge in African labour militancy that began with mass strikes in Durban in early 1973. Consequently, wages rose throughout South African industry during the decade. Fortuitously for the gold industry, this occurred alongside a sharp increase in the price of gold after the USA ended the convertibility of US dollar holdings by central banks to gold at a fixed price. Gold prices on the London market underwent an 'unprecedented' increase, from US$36 per ounce in 1970 to US$613 in 1979. 43 In retrospect, the Chamber concluded that the 1970s had been 'a period of remarkable profitability and expansion', but the immediate effect of the end of gold-price stability was to render the market more volatile and unpredictable. 44 Together, industrial militancy and the soaring gold price led to unprecedented wage increases for African mine workers that significantly outpaced inflation. Wages increased by over 400 per cent between 1970 and 1987, and the increase was most rapid in the mid 1970s as the industry sought to attract domestic labour (see Table 2). Moreover, this pronounced upturn in the gold industry occurred alongside a downturn in the rest of the economy, as South Africa moved from labour shortages to structural unemployment, limiting alternative opportunities for African labour migrants to urban South Africa. 45 Estimated unemployment rose from 14 per cent in 1975 to almost 21 per cent in 1979, and many newly recruited mine workers in the 1970s had previously worked in construction or manufacturing. 46 Rising wages and the lack of alternative economic opportunities were the driving factors behind labour migration from the Transkei to the mines. The benefits for the gold industry were considerable. By the 1980s, the industry had a dependable and adequate source of African labour, something that had eluded the industry for much of the previous century. By the late 1980s, TEBA calculated that the mines could satisfy their labour demands from the bantustans alone. 47 Among these, the Transkei was by far the most important labour-sending region in South Africa, and no other region came close to assuming the same importance (see Figure 1). In 1981, more than one-third of the total African workforce on the gold mines was from the Transkei. Sources of labour outside South Africa never regained their previous importance. Malawi's boycott was short-lived, and recruitment for the mines resumed from 1976, but at a much reduced level. The number of Malawian migrants declined during the 1980s, and TEBA recruitment was curtailed completely in 1987. 48 The gold-mining industry was in a satisfactory situation and, as far as possible, intended to keep it this way. This is because the industry continued to rely upon the large-scale recruitment of unskilled workers and thought that this reliance would continue. In 1961, James Gemmill, then general manager of both the NRC and WNLA, publicly noted that the migrant labour system was crucial for the mining industry because this system was 'best suited to the type of unsophisticated Native on whom the Industry very largely depends'. 49 Twenty years later, the Chamber of Mines emphasised that the 'known expansion plans of the industry are expected to create nearly 100,000 additional jobs in mining, the overwhelming majority for unskilled workers'. 50 The Deferred Pay Interest Fund had an important role to play in securing uninterrupted supplies of unskilled labour. Unskilled or 'cheap labour' entailed various disadvantages for South Africa's longer economic trajectory, including relatively low levels of productivity in manufacturing, a small domestic market and a lack of human capital to drive growth and productivity. 51

The Fund's Spending Patterns and Chiefly Authority
The Deferred Pay Interest Fund started operating from 1918. With its theoretically broad remit about spending priorities, how did the Board decide what to finance? The fund's operations were deliberately opaque, and this may partly explain why they have attracted so little attention or have been treated cursorily or uncritically. 52 Francis Wilson, for instance, described the deferred pay system as an 'important service provided for blacks' as it allowed them to accrue savings. He claimed that the fund was partly used to pay for administering the deferred pay system, but 'the bulk of the fund is given to hospitals and medical missions in the areas from which the migrant workers come'. 53 From the late 1950s, however, the fund became more public, especially in the Transkei. It paid for sports stadia in Umtata and Basutoland and funded ballroom dancing events and tennis championships in Umtata. 54 As the fund became better known and the Transkei became more important to the mining industry, the process for deciding what to finance was as follows. Those seeking funding would prevail upon the NRC's district superintendent for  a specific areasay, Umtataand the superintendent would then write to the Board conveying or motivating the request. The Board would contemplate and then send their candidate requests to the Minister of Native Affairs for approval. Spending was always informed by the view that causes selected must be those, as NRC general manager James Gemmill later put it, 'with which the NRC was beneficially identified'. 55 In the Transkei, it was important for the NRC to be associated with the chiefs and headmen in recruitment areas because chiefs and headmen played a key role in recruiting mine workers for much of the 20th century. 56 By 1910, virtually the entire rural elite in the Transkei was already involved in mine recruitment and were major beneficiaries of the advance system. 57 To retain the rural elite's role after the NRC took control of the rural credit system, the NAD gave chiefs permission in 1913 to visit and tax mine workers in the compounds so that these rural elites could continue to secure material benefits from the migrant labour system.
In the first half of the 20th century, local colonial governance was increasingly comprehensive and marked by a drive to a kind of colonial modernism, with bureaucratic rule central to how the Transkei's people were politically dominated. Magistrates ruled the Transkei like 'little kings'. 58 Especially after the 1927 Native Administration Act, traditional chiefly power was bypassed and local headmen instead were accorded power. 59 Chiefly power returned about two decades later.
The Bantu Authorities Act of 1951 ascribed greater authority to chiefs alongside apartheid state institutions within so-called African homelands. Thereafter, NRC correspondence contains myriad requests from district managers to fund or purchase gifts for chiefs' installation celebrations. Before the Act, chiefs preserved traditional social and cultural power, but, unlike headmen and magistrates, they did not hold office in the colonial administrations in the Transkei. 60 These chiefs, instrumental as indirect rulers and rejected by many as apartheid state functionaries, were granted extensive coercive powers. Chiefs became controllers of state resources, the bedrock of their power, in much the same way that they had previously controlled cattle and land. 61 The NRC wanted to leverage this power to its advantage.
Gifts offered by the NRCand subsequently TEBAshifted from cattle to providing instrumental articles of modernisation and funding public installation ceremonies. This occurred in the context of the Transkei's so-called self-governance in 1963, where a chiefly faction achieved political power, and became more pronounced when the Transkei became a nominally independent homeland. 62 In the 1970s, at chiefly 'installation[s]' and openings of new offices, Chief Ferguson Gwadiso of the Konjwayo Tribal Authority, Chief Douglas Ndamase and Chief Tutor Nyangilizwe each received a 'magnificent desk' with NRC-inscribed plaques, while the half-brother to Paramount Tembu Chief Sabata Dalindyebo, Chief Bambilanga Mtirara, received an office chair and a plaque. 63 When Chief Dumisa Meji was installed in the Manzamhlope Tribal Authority in 1976, he was gifted a loudhailer, and his and Chief Jackson Ndamase's installation ceremonies were partly funded by the Deferred Pay Interest Fund. 64 In deliberate view of subjects entering a chief's office, NRC-branded plaques signified the NRC's close association with apartheid's formidably empowered indirect rulers. Perhaps the clearest symbol of modernisation provided was the gift of an airport terminal by the Chamber for the Kaiser Matanzima Airport in 1976, when the Transkei was granted 'independence' by South Africa. 65 At the meeting where the gift was determined, industry representatives conceded that the WNLA and the NRC 'had enjoyed a degree of goodwill from these gifts which was not wholly justified'. 66 NRC officials were often VIP guests at chieftaincy installation ceremonies. In the 1980s, when Zwelivumile Mtwa was installed as Chief of the Xesibe Tribal Authority in the Mqanduli district by Sabata Dalindyebo, NRC representatives attended the ceremony and gifted the chief the 'usual clock'. 67 Unaware and without consent, mine workers were financing the very institutions that contributed towards their dependence on the migrant labour system. Spending on chiefs was explicitly unrelated to welfare concerns. From cattle to clocks, desks and office chairs, tools and equipment of modern administrative bureaucracies, these were bribes.

The Fund's Education Policy
Perhaps the most striking and illustrative feature of the Deferred Pay Interest Fund was its funding policy for education. In 1970, the Umtata superintendent was approached by the South African Red Cross Society, which asked for renewed financial support towards establishing an organisation in schools in the Transkei. First approached in 1963, the Board had agreed to donate R2,500 per annum for three years, an agreement that was renewed in 1966 but only after James Gemmill had expressed concern that 'the work of the Red Cross was concentrated upon schools, whereas the Industry obtains its labour from the uneducated classes'. 68 In relaying the Red Cross Society's renewed appeal to the NRC general manager in 1970, the superintendent revealingly pointed out that: our gold mineworkers are mainly raw tribal Africans, and not many of them have been to school. As soon as they acquire a little education, they tend to seek employment elsewhere, and this tendency is increasing, and is one of the reasons the Transkei labour supplies of the Gold Mining Industry have dwindled in recent years. 69 This was the period immediately prior to the reorientation of the migrant labour system towards the homelands, and the perceived need by the industry to limit the provision of education for Africans to preserve labour supplies was only heightened as the Transkei became the most important labour-sending region for the mines. This was greatly assisted by the apartheid state's neglect of African education in rural areas. In his 1968 review of 'Bantu Homelands', the economic adviser for Bantu Development, J.J.S. Weidemann, glaringly noted that, of the homeland children in school in 1965, only 3 per cent were in secondary education while 0.3 per cent were in technical education. In 1965, 71 per cent of homeland education spending came from the apartheid state. 70 By 1970, 93 per cent of rural Africans had either no formal education or were educated below grade 8. 71 Between 1970 and 1974 in the Transkei, less than 1 per cent of students were enrolled in grades 11 and 12, while 0.1 per cent received trade and technical training. The Transkei's Department of Justice, formed in 1963 and headed by Chief Kaiser Matanzima's notoriously cruel and corrupt brother, George Matanzima, received more money than the territory's Department of Education each year between 1964 and 1969. 72 Control, not education, was prioritised.
Rural and urban education, however, was distinct and non-static, and education for black South Africans comprised not only the reproduction of unskilled migrant labour. Although rural education was dramatically limited in order to reproduce an unskilled labour force, urban schools, by contrast, 'played a distinctive role'. 73 Jonathan Hyslop has shown that, within the urban working class, urban education under the Bantu education policy reproduced unskilled, semi-skilled, technical and clerical workers for discrete parts of the urban labour market. Within the urban education section, 'across time there were important shifts in the weight given within the education system to the reproduction of different forms of labour power'. 74 Money generated by the interest on deferred wages was used to perpetuate the striking educational profile shown in Table 3. The Deferred Pay Interest Fund's position on education remained resolute and was reiterated in numerous letters between school principals and NRC superintendents and general managers into the 1980s. 75 Notably, only able-bodied schools were refused funding, while disability schools and rehabilitation centres received grants. The distinction is a crucial one. Disabled persons would not constrain labour supplies, and the mines were persistently repatriating injured mine workers.
These continuities are evident in the refusal of the request of F.L. Matyumza, the inspector of religious education in Transkei. In 1976, Matyumza appealed for R300 to host a conference on religious education, arguing that it 'would strengthen the Transkei Independence if the conference met in Umtata'. 76 The Umtata district manager then wrote to the NRC general manager noting that he informed Mr Matyumza that 'it is not normally the policy of our Company to donate monies for educational purposes' but argued that a donation was in the NRC's interests. 77 The general manager reiterated that donations were not for educational purposes, maintained that providing the funds would 'create a precedent' and refused to authorise the payment. 78 Similarly, in response to an application to fund repairs to the Dinizulu JS school in Elliotdale, the district manager contended that, although 'the school is in need of considerable repairs', this was not recommendable as 'this would create a precedent', and the NRC would be 'inundated' with requests from other schools in the Transkei. 79 The district manager wrote a terse letter to the Dinizulu principal informing him that the NRC's '[d]onation policy does not extend to schools'. 80 This was the NRC's consistent response. The NRC's notion of 'welfare' did not regard education as constitutive of welfare. The continuity in policy was clear: even after the Deferred Pay Interest Fund became the TEBA Savings Fund Trust, it was explicit policy not to fund education.
It was not the case that the NRC and TEBA refused to fund any educational facilities or projects. Instead, in an apparent link between expenditure and labour recruitment priorities, funding was refused only for educational projects for able-bodied persons. The case of the Efata School for Blind and Deaf children highlights this nuance. In 1976, R28,000 for a swimming bath was donated to the Efata School from the fund, and, in 1977, the TEBA Savings Fund Scheme granted R5,250 for school house renovations at Efata. 81 In mid 1978, another R30,000 (equivalent to R1,033,235.29 in 2020) was donated for the construction of Efata's school hall. 82 While on principle refusing to donate trifling sums to schools for ablebodied persons, TEBA readily spent comparatively large sums for schools and institutions supporting disabled persons.
Funding for such schools enabled the NRC, and subsequently TEBA, to construct an image of itself as benevolently concerned with disabled Transkei residents, despite externalising the costs of thousands of injured and sick mine workers back into laboursending regions. 83 The Ikwezi Lokusa Rehabilitation School and Sheltered Employment Society of Umtata, which offered pottery classes to disabled persons, provided for TEBA an opportunity to present itself as magnanimous. R3,000 was donated in 1976 for the maintenance of Ikwezi's school for children with cerebral palsy, another R4,000 for the same purpose the following year, an additional R3,000 for general running costs, R30,000 for essential services like lights, water and sewage removal, and R6,000 for the purchase of pottery kilns in 1977. 84 Similarly, in 1978, R30,000 was donated for a pottery workshop and R5,000 for a playroom. 85 In the context of refusing to donate even minor sums to education facilities for able-bodied persons, donations to schools for disabled persons were large.
An industry in which, according to a 1991 estimate, a person who works for 20 years faces a 1 in 30 chance of dying in an occupational accident has an image problem. 86 To counter this, TEBA began spending money accumulated from deferred pay interest on company publicity to portray TEBA as a favourable employer and an agent of development. This signalled a move towards a stronger emphasis on shaping perspectives and neutralising criticisms of the mine migrant labour system, again effectively paid for by mine workers themselves. By 1983, TEBA was distributing hundreds of thousands of educational magazines called TEBA Topics, paid for by the TEBA Fund, in rural schools. 87 TEBA now presented itself as concerned with 'disadvantaged' children, but it is more convincing to see this as a mass advertising campaign for the next generation of workers in the mining industry. 88 By the 1980s, TEBA was concerned to publicise its alleviation of hardships in laboursending regions. A draft document for its magazine TEBA Times noted that the industry's reaction to the droughts of the early 1980s was to allocate 'some millions of Rands of aid from the surplus in the TEBA Savings Fund'. The following phraseology was suggested for the magazine: '[o]ur board did this from a sense of responsibility and community of interest with the "little people"'. 89 No longer was TEBA branding office furniture and using it to bribe chiefs, supporting ballroom dancing or building stadia; the fund had, in the corporate fantasy of its desired public relations, become 'responsible', engaged and humanitarian, even helping out during a Transkei drought.
When TEBA's migrant labour system was criticised by Archbishop McCann of Cape Town, TEBA responded with a volley of assertions that are noteworthy. The most relevant include the following from TEBA's responding letter: [a]t the regional level, TEBA is a development agency … We believe the TEBA acronym could just as well stand for Training and Enterprise Build Africa … Many thousands of men owe their literacy to free lessons on the mines … A total of R305.8 million was distributed in 1982 by TEBA in deferred pay, remittances and savings. This money does a lot of good in Southern Africa. 90 The monetary, social and political costs accompanying mine work, borne by workers, were not factored into the sums reported. Quite what training and enterprise was being referred to is unclear, but there was clearly a pronounced effort from the 1980s to appear to value education, human well-being and human development. TEBA's publicity endeavours from the late 1970s were driven by a desire to change or neutralise perceptions of the migrant labour system rather than changing or mitigating those aspects of migrant labour which generated the negative perceptions. 91 Conclusion South Africa's gold industry has long had a well-deserved poor reputation among historians of the region. The appropriation of interest from wages was by no means the worst injustice and indignity inflicted on those who laboured in South Africa's gold mines, but it does give us additional insight into how these mine workers were regarded by the state and industrial officials in charge of the Deferred Pay Interest Fund, and this has wider applicability. Just as African mine workers could not be trusted to spend their own wages 'responsibly', Africans could not make decisions about their own welfare, so decisions about what would benefit them were made by others, as was the case throughout colonial Africa. The Board of Control directing the fund's spending never had any African representatives on it. Even in the 1980s, African mine workers and other Transkei residents were regarded as beneficiaries of benevolent corporate policies, neither entitled to the full value of the wages they received nor to have any say over how this money was spent.
This was not, however, simply another instance of how the interests of Africans were ruthlessly disregarded by the mining industry. We argue that the spending patterns of the fund reveal one important way in which the mining industry sought to secure continued labour supplies. When the crisis in the migrant labour system in the mid 1970s prompted the gold industry to reorient recruitment towards the homelands, the fund directed spending ostensibly earmarked for 'welfare' to bribe local elites and promote the industry's public image and consciously avoided spending on education for able-bodied children. Payments from the fund were not designed to enable education and skills training precisely because this would reduce the pool of unskilled labour upon which the industry relied. Funding for the education of able-bodied persons was consistently withheld on principle. To ensure a steady supply of labour to the mines, alternative means of making a living in rural areas had to be suppressed or discouraged. This assumption has had significant implications for the areas affected and played a part in shaping rural poverty.
Expenditure from the Deferred Pay Interest Fund certainly supported state policy aims in the bantustansand we discussed the fund's role in bolstering chiefly authority, mentioned by Wolpe as a way in which the state sought to maintain labour reserves 92but it is important to note that this spending was discretionary. Although expenditure was aligned with the state, it was not controlled by the state. In this, the fund's activities, particularly by the late 1980s, has some relation to modern-day 'corporate social responsibility' in the mining industry, whereby mining companies provide funding for development initiatives in the communities that provide the mines with labour, or broader philanthropic spending. In such schemes, control over the money and how it is spent is not in the hands of the supposed intended beneficiaries or their representatives.
We do not wish to suggest that those in charge of the Deferred Pay Interest Fund or of the mining industry were omniscient and prescient in designing these systems to limit alternative livelihoods in rural South Africa. Indeed, in some ways they were remarkably short-sighted, as workers from the Transkei played a prominent role in the strikes in the industry during the 1980s. After the huge 1987 miners' strike, some companies began pressuring the Chamber to reduce recruitment from the Transkei. 93 Moreover, the labour requirements of the industry changed soon afterwards. Gold-mining employment peaked in 1987, and thereafter fell sharply and more or less continuously to the present. In 2019, only 95,130 people were directly employed in South Africa's gold mines, a small fraction of what it had been 30 years previously. 94  required large numbers of unskilled workers, and anxieties that had plagued mining companies for the previous century over how to recruit sufficient numbers quickly dissipated from the early 1990s. The operations of the deferred pay scheme outside South Africa are worthy of further investigation, as is the effect of using mine workers' deferred wages to buy state bonds. Millions of African men through successive generations laboured on South Africa's gold mines, and the sums of money involved in deferred pay were huge. The uses to which this money was put merits further investigation.