29. Why Would you Want to Eat Sushi in the Transkei?
The Economics of Apartheid
In December 1932, in the throes of a deep recession, South Africa left the gold standard. Britain had abandoned it the previous year – and a political battle within South Africa’s government had ensured a delay that severely hurt the economy.1 The decision to leave had an immediate effect; instead of having the currency backed by gold, the South African pound depreciated, making South African exports more attractive to foreign buyers. It proved a huge boon to gold-mining companies. Gold prices rose rapidly and mining output expanded, increasing the demand for inputs and workers, and as a consequence government revenues increased significantly. In 1936, only three years later, the Johannesburg municipality could begin construction of the South-Western Townships, or Soweto, on the back of windfalls from the mining industry.
These mining windfalls – profits for shareholders and taxes for government – depended on one important factor: paying cheap wages. Wage disputes were the main reason that mine owners frequently clashed with white labour unions, which wanted their members to be paid a ‘civilised’ wage. In fact, white miner wages were some of the highest in the world, both because of high living costs in Johannesburg and also because of the miners’ political importance as voters. One way white labour unions and their political supporters had managed to keep wages high was by introducing legislation known as the colour bar, reserving jobs classified as ‘skilled’ and ‘semi-skilled’ for white workers only.
To avoid paying these high wages, mine owners constantly looked for alternatives. One was simply to employ more black workers. But one concern of mine owners was that attracting more black workers to the cities would mean fewer workers on farms. This would lead to higher prices for staples such as maize and bread, which would result in demands for higher wages on the mines. Employing black South African workers could thus be counterproductive: the ‘gold–maize alliance’ between mine owners and farmers ensured that mines would not attract workers from the same labour catchment areas on which farmers depended.
The alternative that mine owners chose to solve this labour conundrum was to import workers. An early experiment to bring in 63,000 Chinese workers just after the Anglo-Boer War failed. They then set their sights on African labourers from neighbouring countries. The Chamber of Mines established the Witwatersrand Native Labour Association, a labour recruitment agency known as WNLA or simply Wenela, in 1900. Soon, young men from Lesotho, Eswatini, Malawi and, in particular, Mozambique began to arrive in Johannesburg. By the 1930s, when gold mining boomed, more than 300,000 labourers worked on the mines; at least half of them came from the Sul do Save district in Mozambique. These workers, as the historian Charles van Onselen explains, were delivered to the Highveld mines in train coaches that were often barred and locked, to live their working lives in enclosed, well-policed mine compounds.2 Worst of all, because of a deferred pay system, with half of all wages paid out only once the migrant workers returned to Mozambique, and because of theft due to corruption by those responsible for transporting the miners, it is estimated that miners received as little as 15 per cent of the wages they had earned on the mines when they finally arrived back in Mozambique.
Despite these conditions, African workers continued to make the journey to Johannesburg. There is some evidence that their wages, however meagre, did improve conditions for them and their families. In 1967 Malawi signed a new labour recruitment treaty that tripled the number of mine workers sent to South Africa’s mines. Seven years later a plane carrying Malawian mine workers crashed, which led to Hastings Banda, Malawi’s president, imposing a three-year ban on all labour recruitment in the country. The economists Taryn Dinkelman and Martine Mariotti have used the recruitment treaty and moratorium to estimate the effect of these income shocks on the children of those living close to mine-recruitment stations.3 They found that twenty years later, children whose fathers lived in recruitment areas (Wenela districts), and who were of school-going age in the seven-year period, were 5–7 per cent more likely to go to school. It makes sense: as more Malawian men migrated to Johannesburg to work on the mines, their wages increased, allowing their kids back home in Malawi to attend school.
In a separate paper, Mariotti considered children born in South Africa’s Transkei before and after 1974.4 After the plane crash and Malawi’s subsequent ban on labour recruitment, mine owners had to find alternative labour sources. The numbers of migrant workers from the Transkei rapidly increased. Mariotti found that children born in the Transkei in the years following their fathers’ move to the mines are taller today than those born in the years immediately preceding that move – further anthropometric evidence in support of a rise in living standards from mine work. It is a sad indictment of the time that the terrible conditions on the mines were often a better alternative than the poverty of the home, whether in Malawi or the Transkei.
The constant competition between white and black labour became acute after the Second World War. White women and black men had filled the semi-skilled or even skilled jobs of the more than 200,000 white men who had given up work to fight in the war. In fact, factory owners soon realised that they could pay lower wages to white women and black men and achieve the same efficiency as with the higher-paid white men.5 Yet, when the white soldiers returned after the war they expected their jobs back, only to realise that they now had to compete with men and women they had never considered their equals. They responded in the only way they could: at the ballot box.
In 1948 the National Party – a small, populist party that had only come into existence a decade earlier and that catered mostly to lower-income Afrikaners – won the national elections with 37 per cent of the vote. Their opponents had no clear plan about how to deal with the race issues of the time. The National Party, in contrast, had a very clear vision of a country in which white and black people would live separately. The set of laws and policies they began to implement immediately after their victory – including the Group Areas Act (1950), which defined where South Africans could live and own property; the Population Registration Act (1950), which classified all South Africans into four race groups; and the Bantu Education Act (1953), which legislated differences in curriculum and funding between white and black schools – was known as apartheid. These discriminatory and repressive policies – many of them an extension or intensification of the segregationist policies introduced in the first half of the century – were condemned in 1973 by the United Nations as a crime against humanity.
One key tenet of apartheid was the creation of Bantustans or homelands. Although these had no international standing, all black South Africans were considered inhabitants of one or other of these ‘countries’ – and could therefore only move to the South African cities temporarily. For mine and factory owners in ‘white South Africa’, this created the problem of a labour shortage. This was particularly true for semi-skilled and skilled occupations. The first two decades after the National Party victory were a period of rapid growth in the economy: this was true not only in South Africa, but for much of the Western world. New technologies (often created for war purposes) improved productivity in many industries, leading to mechanisation and a demand for skilled jobs in the secondary and tertiary sectors.
But apartheid’s masterminds had envisioned a very different society, one where better education for whites would allow them to fill the well-paid, skilled occupations while a large pool of cheap black labour would fill unskilled jobs. This was the model that had in effect been in place for the first half of the century. In fact, the Bantu Education Act was specifically designed to fulfil this objective: preparing black workers for unskilled work and white workers for skilled jobs. These actions show that the National Party policies were not only immoral but also economically disastrous: the economy was transforming away from unskilled work just when the National Party hoped to produce more unskilled labourers. Something had to give.
Martine Mariotti, in a 2012 paper, shows that, despite legislation preventing black workers from taking up semi-skilled jobs, there were by the 1970s many black workers filling these types of occupations.6 White trade unions were, of course, heavily opposed to any relaxation of the colour-bar legislation, and threatened to shift their support to the right-wing Conservative Party should the National Party consider doing so. An alternative plan was devised. If black workers could not move to the cities, the policymakers asked, why not take factories to the workers?
The plan was to build ‘border industries’ – in other words, factories on the edge of the Bantustans. The attraction to investors was the cheap labour available in the homelands. Here is Albert Luthuli, the 1961 Nobel Peace Prize recipient, in his book Let My People Go, on the poor economics of the plan:
White industrialists are invited to place factories on the edges of the destitute Reserves. The bait is cheap labour. Nothing but cruelly underpaid labour could counteract the economic effect of putting factories out of reach of railheads and sources of power. We are told that these factories will enrich the Reserve population. We shall never have had it so good. We are even told what proportion of the populace will be employed in border industries, solely as unskilled workers. Quite arbitrarily, it is arranged that 40 per cent of the Reserve people will remain farmers, while 60 per cent will become industrial workers. The industrialists’ response to the Government invitation so far is a few clothing factories and a few peanut factories … Border industries cannot conceivably employ 60 per cent of the Reserve people. Economics will not obey racial blueprints.7
Luthuli was right, of course. For most firms, locating to the countryside was simply not an option; it was too expensive there to access production inputs and skilled workers, who wanted to live in cities within reach of the services and amenities that large agglomerations of people offered. There were exceptions, however. From the late 1970s large labour-intensive manufacturing from Asia began drifting towards these special economic zones in the rural homelands of South Africa.8 With extensive ‘bamboo networks’, which connected them to supply chains in Taiwan, Hong Kong and South Korea, these firms provided rare success stories. At one stage the Transkei was the world’s largest producer of chopsticks.
But despite some infrastructure investment (the IDC estimated that the costs of constructing this infrastructure equated to 1.5 per cent of GDP in each year from 1972 to 1979) and generous incentives, there was very little domestic interest in these rural homeland zones. When the tax and wage incentives ceased in 1996, so did the businesses.
Elsewhere in South Africa, instead of moving their location, factories mechanised by using (expensive) capital equipment to substitute for labour. It was not only manufacturing that mechanised. Even agriculture – the sector that traditionally employed unskilled labour – began to transform. While the amount of farmland in cultivation declined between 1946 and 1976, the number of harvester combines, a machine used for harvesting crops and a substitute for labour, increased more than tenfold from 1,722 to 23,767. This mechanisation across the economy had an important consequence, one that still haunts South Africa today: by reducing the need for unskilled labour, it gave rise to high levels of unemployment. When black labour unions were unbanned and understandably began to push for higher wages, mechanisation and unemployment were further bolstered. If one includes both formal and informal employment, the unemployment rate increased from just above 10 per cent in 1976 to more than 50 per cent in 2000.
The economist Nicoli Nattrass has calculated the profitability rates of firms during the apartheid period.9 She finds that low black wages during the 1940s and 1950s did indeed allow for growth in white wages and a rise in firm profits. But by the 1960s manufacturing profits began to decline. Mining profits followed the downward trend by the 1970s. Instead of opening up the economy to encourage exports, the South African government chose the route of import substitution because of international pressure and internal politics. This hurt not only workers, but also shareholders. Nattrass notes that during the 1970s and 1980s profitability ‘continued to fall because of declining capital productivity. This, in turn, almost certainly stems from apartheid policies that encouraged greater capital intensity, notably racial restrictions on the labour market and subsidies to capital.’10 Whereas average annual GDP growth had been 5 per cent between 1948 and 1974, it fell to just 1.5 per cent in the two decades that followed.11
Why, one has to ask, did it take so long for political change to come about when the economic situation was so dire? One, perhaps counterintuitive, answer is that it took so long because of the poor economy. This was a hot topic of discussion by the 1980s: would international sanctions that worsened the economic conditions – of both white and black South Africans – achieve their aim of a faster transition to democracy? In one sense, it seems obvious that they should: if you hurt someone, they ought to change whatever they have been doing in order to reduce the pain. But economist Mats Lundahl explains why the sanctions might not have had the desired effect: ‘Sanctions have different impacts on different groups, as does the apartheid system itself. Some groups gain while others lose. Thus, there is scope for conflicts of interests within the white community, and if those groups that stand to gain if sanctions are imposed are politically influential, the sanctions weapon may work in the completely wrong direction.’12
The international trade and investment boycotts, as the economist Merle Lipton found, strengthened the South African state vis-à-vis other sections of society.13 It prolonged rather than put an end to apartheid. The same may have been true of Robert Mugabe’s Zimbabwe. Sanctions gave him political legitimacy – he had an outside enemy to blame for the woes that Zimbabwe experienced after the land reform programme of 2000.
Ultimately, apartheid ended because, as Albert Luthuli had remarked in 1961, ‘economics will not obey racial blueprints’. The costs of repressing black resistance, influx control and labour market policies that prevented black workers from filling semi-skilled and skilled jobs, education policies that created a mismatch between what firms demanded (skilled workers) and what was supplied (unskilled workers), and macroeconomic policies such as exchange control came to exceed the benefits to white South Africans, and left the last apartheid leaders with little alternative.14 The system collapsed, ushering in a new, democratic era with the political freedom that so many black South Africans had been fighting for.
Eichengreen, B. (2021). Gold and South Africa’s Great Depression. Economic History of Developing Regions, 36(2), 175-193.↩︎
C. van Onselen, The Night Trains: Moving Mozambican Miners to and from the Witwatersrand Mines, circa 1902–1955 (Cape Town: Jonathan Ball, 2019).↩︎
Dinkelman, Taryn, and Martine Mariotti. "The long-run effects of labor migration on human capital formation in communities of origin." American Economic Journal: Applied Economics 8, no. 4 (2016): 1-35.↩︎
Mariotti, Martine. "Fathers’ Employment and Sons’ Stature: The Long-Run Effects of a Positive Regional Employment Shock in South Africa’s Mining Industry." Economic Development and Cultural Change 63, no. 3 (2015): 485-514.↩︎
Clark, N. L. (2001). Gendering production in wartime South Africa. The American historical review, 106(4), 1181-1213.↩︎
Mariotti, Martine. "Labour markets during apartheid in South Africa." The Economic History Review 65, no. 3 (2012): 1100-1122.↩︎
A. Luthuli, Let My People Go (London: Fontana, 1982), 180.↩︎
E. Kerby, Bamboo shoots: Asian migration, trade and business networks in South Africa, Studies in Economics and Econometrics, 42 (2), 2018, 103–37.↩︎
N. Nattrass, Deconstructing profitability under apartheid: 1960–1989, Economic History of Developing Regions, 29 (2), 2014, 245–67.↩︎
Ibid., 264.↩︎
W. H. Boshoff and J. Fourie, The South African economy in the twentieth century, in Business Cycles and Structural Change in South Africa, edited by Willem Boshoff (Cham: Springer, 2020), 49–70.↩︎
M. O. Lundahl, Apartheid in Theory and Practice: An economic analysis (London: Routledge, 2019), 218.↩︎
M. Lipton, Capitalism and Apartheid: South Africa, 1910–1986 (Cape Town: David Philip, 1986).↩︎
A. D. Lowenberg, Why South Africa’s apartheid economy failed, Contemporary Economic Policy, 15 (3), 1997, 62–72.↩︎