Skip to content

A century of migrant labour in the gold mines of South Africa

Literature › Works

J.S. Harington; N.D. McGlashan; E.Z. Chelkowska (2004). A century of migrant labour in the gold mines of South Africa. Journal of The South African Institute of Mining and Metallurgy (March 2004). South African Institute of Mining and Metallurgy.
Type journal article
Scope Economics and economic history of apartheid
Method descriptive quantitative
Period studied 1896–1996
Regions Witwatersrand gold mines, Transkei, Lesotho, Mozambique, Malawi, Zambia, Free State, Bophuthatswana, KwaZulu-Natal, Transvaal, Southern Africa
Themes mining migrant labour, labour markets wages, bantustans removals influx, demography urbanisation, origins segregation, macro fiscal finance debt, inequality poverty unemployment
Citation key harington2004

Argument

The article documents the size and origins of the black workforce on South African gold mines from 1896 to 1996, using Chamber of Mines and TEBA recruitment data to track roughly twenty source territories. The authors argue that the fluctuations reflect external events (wars, depression, the gold price) and internal politics and policy (the 1974 drive to replace foreign with South African labour, apartheid, law and order), together with rising costs, wages and tax against profitability. They treat migrancy as an entrenched, untenable system with severe social costs, especially in Lesotho and the Transkei, whose economies depend on remittances and are now devastated by mine retrenchment. The narrative leans heavily on Lang's history of the Chamber of Mines for the period to 1986. Historical explanation is organised around the lettered peaks and troughs in Figure 1.

Evidence and method

Chamber of Mines Statistical Tables (1994, 1998, 2001), TEBA recruitment database (annual figures by territory), Mining Statistics 1983, and Lang's 1986 history of the Chamber. Secondary works: Wilson (1972), Crush, Jeeves and Yudelman (1991), James (1992), Webster (1993), De Kiewiet (1978). Newspaper reports and the authors' own earlier cancer-epidemiology studies of miners. Presented as graphs (Figures 1-7) and a gold-price table (Table I).

Place in the debates

Findings

  • Mozambique was the dominant early source of labour. Recruitment shares changed over time, with the Cape and High Commission territories later rising. (p. 1)

    In 1899 two-thirds of recruited labour came from Mozambique, the second most important source being the Transvaal (18 per cent), and the Cape (5 per cent).

  • The black labour force fell sharply during the South African War, and a Chinese labour experiment failed. (p. 2)

    By 1903 the black labour force was down to half the 90 000 of 1899.

  • Mine employment rose through the Depression, thanks to the gold price and departure from the Gold Standard. (p. 2)

    The number of black workers in the mines actually rose in each of the Depression years and reached new peaks from 1932 to 1940

  • The foreign share of the labour force was deliberately cut after 1974, helped by Zambian and Malawian bans. (p. 3)

    TEBA then gradually reduced the foreign component of the total labour force, which fell from 37 per cent in 1966 to 16 per cent by 1979

  • The black gold-mine workforce peaked in the late 1980s. (p. 4)

    some 480 000 workers per annum over 1987 to 1988 (l, Figure 1).

  • The black workforce shrank by about 40 per cent between 1986 and 1994. (p. 4)

    had reduced its black worker numbers from 477 000 in 1986 (1, Figure 1) to 303 000 in 1994

  • Output value rose while employment halved, indicating labour shedding despite higher output. (p. 4)

    Between 1985 and 2000, the value of mine output had increased by more than 250 per cent, whereas employment had fallen by 50 per cent

  • Transkei supplied the largest single domestic share in the mid-1980s. (p. 4)

    South Africa itself contributed 254 000 (58 per cent), including the Transkei contribution of 133 000 (31 per cent)

  • Lesotho and Transkei are heavily dependent on mine remittances. (p. 6)

    In many rural areas in the Transkei or Lesotho, more than half of the local GDP came from remittances from absentee miners.

  • The system persisted because the alternative of settling families was unacceptable to owners and the state. (p. 6)

    The only alternative to this mass movement was the large-scale resettlement of the miners and their families along the Reef

Evidence ledger

Sources used

Source Kind Period How used
TEBA (The Employment Bureau of Africa) database of annual recruitment records and home addresses, by up to twenty territories chamber of mines or recruiting records 1896-1996 Supplies the annual black labour force numbers by territory of origin that underlie Figures 1, 2, 4-7 and the analysis of the major supply areas.
Chamber of Mines medical records chamber of mines or recruiting records 20th century Mentioned as a firm basis for research on health and illness in miners; not directly analysed in this paper's labour-number argument.
Statistical Tables 1994, Chamber of Mines of South Africa employer association records 1984-1994 Source for black labour force totals in 1984, 1986 and 1994 and the South African/Transkei share, and for the classification of workers by category.
Statistical Tables 1998, Chamber of Mines, Johannesburg employer association records 1986-1998 Used for total employment in 1998 and its peak, world gold production share, the 1998 gold price, and Lesotho versus Transkei supply 1989-96.
Statistical Tables 2001, Chamber of Mines (Figure 3) employer association records 1990-2001 Reproduced to show the average number of employees in service on Chamber gold mines declining to 2001.
Mining Statistics 1983, Dept Mineral and Energy Affairs, Republic of South Africa official statistical series to 1983 Cited for the date of the gold discovery in the Transvaal.
Lang, J. Bullion Johannesburg: Men, Mines and the Challenge of Conflicts (1986) secondary literature compilation 1886-1986 The main source for the historical narrative, recruitment shares, strikes and events to 1986 that explain the fluctuations in Figure 1.
Wilson, F. Labour in the South African Gold Mines, 1911–1969 (1972) secondary literature compilation 1911-1969 Cited for migrant labour system consequences and 'acute urgency' of the situation; background rather than data.
TEBA presentation to Parliamentary Standing Committee on Mineral and Energy Affairs, 5 Oct. 1994 commission or official report 1994 Source for the history and role of recruitment organisations (NRC, WNLA, TEBA).
The Citizen, Johannesburg, comment on ILO Report (14 Oct 2002) newspapers periodicals 1985-2000 Supports the claim that mine output value rose by over 250 per cent while employment fell by 50 per cent between 1985 and 2000.
The Star, Johannesburg, Sept. 1998 and 29 Nov 2002 (Battersby and Terreblanche) newspapers periodicals 1998-2002 Source for the estimate that remittances are about 75 per cent of Lesotho's GDP and for rural land degradation and Eastern Cape conditions.
McGlashan, Harington and Chelkowska 2003, 33-year study of cancer among black gold miners 1964-1996 secondary literature compilation 1964-1996 Cited for the HIV/AIDS impact on the mine labour force and the scale of 12.8 million man-years of observation.
Annual average gold price per fine oz in SA rands and US$, 1903-1994 (Table I) international trade or finance data 1903-1994 Used to show the relationship of labour force numbers to gold price changes.

New data

The authors compiled, from TEBA and Chamber of Mines records, a series of annual black gold-mine labour numbers by territory of origin for 1896-1996 (up to 20 territories), presented as Figures 1-7 and Table I; the underlying tables are not published.

Inference

Quantification: Descriptive tables.

Descriptive time series of annual labour totals by origin territory (graphs), interpreted by matching peaks and falls to historical, political and gold-price events drawn mainly from Lang's history; no statistical testing. Social-consequence claims rely on secondary literature and press reports.

Key claims

Claim Basis Strength
The total black gold-mine labour force rose over the century to a peak of about 480 000 in 1987-88 and then fell sharply, with the 1985-96 fall linked to falling profitability and higher costs. (p. 4) TEBA and Chamber of Mines series in Figure 1 and Chamber Statistical Tables. Systematic data
Lesotho, Transkei/Cape and Mozambique were the dominant sources of labour, supplying over sixty per cent of the labour force in 1989-96. (p. 1) TEBA recruitment series by territory (Figures 1, 2) and Chamber Statistical Tables. Systematic data
Mozambique's share fell from two-thirds of recruits in 1899 to 48 per cent in 1912 and 37 per cent in 1930, while Cape/other South African shares rose and High Commission and foreign recruits grew. (p. 2) TEBA series and Lang's history, shares cited without table. Partial data
The Chamber's 1974 policy of replacing foreign with South African labour largely succeeded: foreign share fell from 37% in 1966 to 16% by 1979, and South Africans were 51% of the force in 1994. (p. 3) TEBA and Chamber of Mines figures (Figure 2, Statistical Tables 1994); political events (Kaunda, Banda bans) from Lang. Partial data
Employment fluctuations are explained by wars, depression, gold discoveries (OFS 1946), political events and the gold price, with falls consistently linked to the falling world gold price in US$. (p. 5) Visual matching of Figure 1 and Table I to events narrated from Lang; no formal test. Illustrative or anecdotal
The decline of mine employment has devastated labour-sending areas such as Lesotho and Transkei, where remittances made up a large share of local income. (p. 6) Secondary literature and press reports (Star, Crush et al., James, Webster); no new household data. Illustrative or anecdotal
The mines are slowly reducing reliance on migrant labour as the South African workforce stabilises near mines, but the migrant system remains entrenched. (p. 6) Secondary literature (Crush et al., James, Webster) and Figure 2 proportions. Asserted

The author's caveats

The authors note that categories of black workers were no longer classified separately from whites after 1994, and that Bophuthatswana is assigned to the Cape because about 80 per cent of its miners come from the former Cape Province. They also warn about the widely differing vertical axes of the minor-supplier graphs.

Gaps in the evidence

The causal explanations for employment changes rest on narrative and on visual correspondence with Figure 1 and the gold price, with no wage, cost, profitability or mine-level series and no statistical testing. The claims about remittance dependence and household hardship in Lesotho and Transkei lack survey or household data, relying on press estimates. Mine-level and wage data and recruitment records by district existed at the time, and TEBA and Chamber data could be combined with household surveys to test the links.

Replicability: Data described but not public.