African Migrant Labour Situation in Southern Africa

Literature › Works

Kanyenze, Godfrey (2004). African Migrant Labour Situation in Southern Africa. Paper presented at the ICFTU-AFRO Conference on 'Migrant Labour', Nairobi, 15-17 March 2004. Labour and Economic Development Research Institute of Zimbabwe (LEDRIZ).

Type	conference paper
Scope	Economics and economic history of apartheid
Method	synthesis review
Period studied	1850s–2003
Regions	South Africa, Mozambique, Lesotho, Zimbabwe, Botswana, Malawi, SADC region
Themes	mining migrant labour, origins segregation, labour markets wages, unions strikes, transition post apartheid, demography urbanisation, agriculture land
Citation key	kanyenze2004

Argument

The paper surveys cross-border migrant labour in Southern Africa from the 1850s to 2003, treating the regional system as a product of uneven capitalist development centred on mining and plantations. Taxes, land expropriation and forced labour fed the mines, and WENELA/TEBA recruitment and deferred pay kept migration temporary and oscillatory. After the mid-1970s, independence-era withdrawals by supplier states (above all Mozambique), the 1987 NUM strike and later retrenchments changed the national composition of the mine workforce. Supplier states stayed fiscally dependent on remittances, so regional attempts to end the system failed. The author argues that migrants were a docile, cheap and flexible labour source, that post-1990 policing has produced 'revolving door' undocumented migration, and that regional cooperation, tripartite policy and human-centred development are needed instead.

Evidence and method

Secondary and policy literature: Crush & Williams (2002), ILO/SAMAT (1998), Standing, Sender & Weeks (1996), Crush, Jeeves & Yudelman (1991), Davies (1995), Sachikonye (ed. 1998) and Matlosa (1998). Statistical tables on mine labour by source country 1920-2000, SAMP survey results (1997/98), census figures, deportation counts and work-permit data. Some union (NUM) positions are also reported.

Place in the debates

• Cheap labour thesis / migrant labour as a structured cost advantage (see Cheap labour thesis and the function of the reserves). Supports it: migrant labour is a 'structured cost advantage for South African industry'. Employers kept hiring and firing flexible at little cost, and host states did not bear the reproduction costs of labour, which fell on rural households in the supplier states.
• Causes of the post-1974 shift in the mine labour composition (see The mine migrant labour system: origins, recruitment, foreign labour and stabilisation). Attributes it mainly to the political decisions of newly independent supplier states and to South African retrenchment after 1987. It also stresses the mines' own policy of stabilising the workforce with more South African workers.
• Whether the institutionalised migrant system is eroding (see The mine migrant labour system: origins, recruitment, foreign labour and stabilisation). Cites ILO/SAMAT and Standing et al. that migration to the mines has 'stabilised' into career miners, with fewer TEBA recruiting stations and more contracting. It also stresses that supplier-state dependence on remittances persists.
• Policing versus development responses to undocumented migration (see Post-apartheid economic policy, governance and corruption). Argues that restrictive controls and deportations treat symptoms and fuel hidden migration. It favours addressing push factors through regional economic cooperation and a rejection of SAP market-driven approaches.

Findings

• The Witwatersrand gold discovery created demand for regional recruitment, which was organised through WENELA with exclusive rights. (p. 2)

  A recruiting agency in South Africa, the Witwatersrand Native Labour Association (WENELA) was created and given exclusive rights to recruit labour from the region.

• Cash taxes alone did not deliver enough labour, so land expropriation and forced labour were added as coercive supply measures. (p. 3)

  When it became clear that these taxes did not result in an adequate flow of labour into the mines and plantations, additional measures were used, and most importantly, the expropriation of land from the peasantry.

• South Africa's 1963 immigration controls ended much of the clandestine and family migration and narrowed legal entry mainly to the mines and agriculture. (p. 3)

  this was curtailed when South Africa imposed strict immigration control regulations in 1963 which adversely affected undocumented and family migration.

• Mozambique's post-independence disruption of the recruitment machinery caused a sharp fall in its supply of mine labour in 1976. (p. 4)

  the flow of migrant labour from Mozambique fell from 114,385 in January to 48,565 by end of 1976.

• Compulsory deferred pay served to tie earnings to the home economies and keep the system temporary. (p. 4)

  For the government, the system of deferred payment ensured that the migrant worker spent much of his earnings at home.

• Supplier states could not agree to withdraw their labour because their public finances depended on migrant remittances. (p. 5)

  For them, summary withdrawal from the migrant labour system would precipitate a severe fiscal crisis.

• Lesotho's dependence on migrant mine earnings is extreme at household level. (p. 6)

  Almost 70% of rural Basotho households depend on remittances, which constitute up to 80% of gross incomes.

• The foreign share of mine employment peaked in the 1970s and then fell steeply as supplier states withdrew workers. (p. 7)

  increased to 72% by 1970 (peaking at 78% in 1974), it fell to 42% by 1989, due to the withdrawal of cross border migrant workers

• The mines raised the share of South African workers in the 1980s to secure a stable labour supply when foreign supply was uncertain. (p. 7)

  reflecting the policy then of employing more South Africans to create stability at a time labour supply from some independent countries in the region was uncertain.

• Mine employment fell substantially from the late 1980s, with retrenchments starting after the 1987 NUM strike. (p. 7)

  more generally, employment levels declined from almost 422,000 in 1989 to about 231,000 by 2000.

• The mine workforce stabilised and aged, and sub-contracting grew as a means of bypassing unions. (p. 8)

  Whereas contractors accounted for only 3% of the workforce in gold mines, the level rose to 10% by 1994.

• After the 1987 strike, mines replaced militant South African workers with migrants, who were used as a docile source of cheap labour. (p. 16)

  more migrants from Mozambique and Swaziland were hired to replace militant South Africans dismissed after the 1987 strike

Evidence ledger

Sources used

Source	Kind	Period	How used
Crush & Williams (2002) ILO Tripartite Forum paper: Table 1 contract labour migration to South African mines, 1920-90	secondary literature compilation	1920-1990	Supplies the series of contract mine labour by country of origin used to show the regional reach of the contract system and the post-independence withdrawals.
ILO/SAMAT (1998) Policy Paper No.4 'Labour Migration to South Africa in the 1990s' and Crush & Williams (2002: 23): Table 3 sources of mine labour	secondary literature compilation	1920-2000	Used to trace mine employment totals, foreign share and the rise and fall of South African and foreign mine labour.
SAMP nationally representative migration surveys in Botswana, Mozambique, Zambia and Zimbabwe, 1997/98 (Table 2)	household or labour survey microdata	1997/98	Shows the share of adults whose parents or grandparents worked in South Africa, indicating the intergenerational depth of migration.
SAMP 1998 survey on reasons for entry to South Africa (Table 5)	author survey or fieldwork	1998	Used to explain differing patterns of cross-border migration by showing work, trade, shopping and family-visit motives by country.
SAMP 1998 survey of gender differences in migration (Table 6) and SAMP survey of legal status of African immigrants	author survey or fieldwork	1998	Establishes the male/female difference in purpose of visits and work history, and the legal status of immigrants.
Standing, Sender & Weeks (1996) Restructuring the Labour Market: The South African Challenge (ILO)	secondary literature compilation	1975-1995	Provides data on mine workforce stabilisation (age, experience), contracting, retrenchment and mine safety, and quotes mine managers.
ILO/SAMAT (1998) Policy Paper on undocumented migration estimates, wages, displacement and health	commission or official report	1990s	Cited for estimates of undocumented migrants, wage comparisons, displacement of South African labour after 1987, and occupational health problems.
South African deportation and overstayer statistics, 1988-1997	official statistical series	1988-1997	Used as a rough proxy for undocumented migration in the absence of actual figures.
Botswana 1991 population census	population census	1991 (with 1981 comparison)	Gives the number and origins of non-citizens and migrants in Botswana.
Botswana work permits by nationality, 1995-99 (Table 7)	official statistical series	1995-1999	Shows the composition of skilled migrant labour entering Botswana and its gender bias.
Malawi 1998 population and housing census	population census	1987-1998	Shows the decline in the foreign-born share after Mozambican refugees returned.
Namibia 1991 population census	population census	1991	Gives the number of persons born outside Namibia.
Zimbabwe 1992 census	population census	1992	Gives the foreign-born population and the composition of non-citizens by origin.
SIRDC (2003) An Analysis of the Cause and Effect of the Brain Drain in Zimbabwe	commission or official report	2002-2003	Source for the estimate and distribution of Zimbabweans living abroad.
Sachikonye (2003) The Situation of Commercial Farm Workers After Land Reform in Zimbabwe	commission or official report	2000-2003	Supports claims about farm-worker displacement and job losses after fast-track land reform.
Paton (1995) estimate of migrant worker-years in the region, cited via Sachikonye (1998)	secondary literature compilation	to 1990	Used to indicate the cumulative contribution of migrant labour to South African mining.
Matlosa (1998) on Lesotho's dependence on South Africa; Central Bank of Lesotho migrant wage survey	secondary literature compilation	1992-1998	Provides figures on SACU receipts, remittances and deferred pay to show Lesotho's dependence and the wages of Basotho miners.
Historical quotations and accounts via Manghezi (1998), Davidson and Mondlane on taxes, forced labour and shibalo	secondary literature compilation	1850s-1975	Used to establish the mechanisms of coerced labour supply (taxes, land expropriation, forced labour).
Davies (1995) Cross Border Migration in Southern Africa: Mission Report (IDASA)	commission or official report	1995	Source for estimates of undocumented migrants and the link between contract migrancy downturn and undocumented migration.

Inference

Quantification: Descriptive tables.

Synthetic review paper: assembles tables and figures from secondary compilations (Crush & Williams 2002, ILO/SAMAT 1998, Standing et al. 1996, SAMP surveys), tabulates trends over time and across countries, and combines them with historical narrative and policy argument. No original data or statistical testing.

Key claims

Claim	Basis	Strength
The foreign share of mine employment rose to 72% by 1970 and fell to 42% by 1989 after regional states withdrew their labour, while South African mine employment rose and later fell sharply. (p. 7)	Tables 3 and 4 built from ILO/SAMAT (1998) and Crush & Williams (2002), supported by Table 1.	Systematic data
Mozambican migration to South African mines fell sharply after independence because the new government dismantled the WENELA recruitment machinery. (p. 4)	Figures on migrant flows from Mozambique in 1976 and Table 1, drawn from secondary sources.	Partial data
Sub-contracting and the stabilisation of the mine workforce are rising, with a shift away from institutionalised recruitment. (p. 8)	Standing et al. (1996), ILO/SAMAT (1998) and Sachikonye (1998) on contractor shares, workforce age and WENELA/TEBA recruitment stations.	Partial data
Undocumented migration to South Africa has increased markedly since 1990 as legal immigration fell, although its true scale is unknown. (p. 10)	Deportation and overstayer figures and ILO/SAMAT estimates ranging from 2.5 to 12 million.	Illustrative or anecdotal
Mozambican migrants are predominantly labour migrants, whereas migrants from Zimbabwe, Lesotho and Botswana mostly enter for trade, shopping or family visits. (p. 9)	SAMP 1998 survey, Table 5 (Crush & Williams).	Systematic data
Migration to South African mines was highly gendered, with men dominating legal contract and cross-border work. (p. 11)	SAMP Table 6 and the statement that 99% of mine workers are male.	Partial data
Supplier states, especially Lesotho, depend heavily on migrant remittances and deferred pay, which blocked collective withdrawal from the migrant system. (p. 5)	Crush et al. (1991), Matlosa (1998) and deferred pay figures.	Partial data
Foreign migrants occupy the worst jobs and bear the health and safety costs, and these costs are shifted to supplier countries. (p. 17)	ILO/SAMAT (1998), Standing et al. (1996), and mine accident statistics for 1993-94.	Partial data

The author's caveats

Notes that analysis is hampered by the dearth of adequate information, especially on undocumented migration. States that the extent of undocumented migration is not known and that deportations are only a rough proxy. Says the regional summaries are only a thumbnail based on Crush & Williams (2002), and that some estimates (e.g. Zimbabwe emigrants) are likely gross underestimates. Table 3 omits some countries' figures.

Gaps in the evidence

The paper relies almost entirely on secondary compilations and offers little primary evidence on the apartheid-era recruiting system, such as Chamber of Mines/WENELA records, wage series, or bilateral agreements. The causal claims about push factors, displacement of South African labour and cost advantages to mining capital are asserted without wage, employment or cost data. Mine-level employment and wage series, TEBA/Chamber of Mines archives and national censuses exist and could test these claims. Undocumented migration cannot be quantified from deportation counts, though household survey data such as SAMP and later South African censuses exist.

Replicability: Data poorly documented.