Influx control, pass laws and African urbanisation¶
What were the purposes of influx control and the pass laws (labour allocation vs political control), how effective were they, and how did African urbanisation proceed?
Works engaging the debate: 23
Positions¶
- Labour-supply and allocation function (Hindson)
- Political control and exclusion (Posel, Savage)
- Ineffective vs significantly slowed urbanisation (Simkins)
- Reform as restructuring rather than abolition
Evidence base¶
The debate rests mainly on official statistics re-tabulated from published sources, statutory and commission texts, and a small number of local archival, survey and interview studies. The quantitative core is the national prosecution series compiled by Savage from the Annual Reports of the Commissioner of the South African Police, Official Year Books, Union Statistics for 50 Years and Hansard question replies. It counts more than 17.25 million pass and influx-control prosecutions over 1916-1982, extended to 17.7 million for 1916-1984, peaking in 1967/68 (Savage (1984); Savage (1986)). Savage sets this series against Simkins's corrected census distribution of the African population, which shows the Bantustan share rising from about 40 to 53 per cent after 1950. He adds Hansard arrest figures for 11 urban centres, 1974-82/84, broken down by enforcing agency and sex. Earlier snapshots come from Simons's census-office table of convictions under pass and allied laws, which rise from 217,387 to 337,603 a year over 1950-55 (Simons (1956)). Davenport supplies a census table for six cities, 1911-1960 (Davenport (1969)). For Cape Town, West and the 1983 collection build tables from SAIRR Surveys and Hansard on arrests by sex (1978 to mid-1981) and on fines, plus Koornhof's 1981 reply that 85,436 of 199,600 Africans in greater Cape Town were 'illegal' (West (1982); (ed.) et al. (1983)). The 1983 collection also tabulates Black Sash Athlone Advice Office intake forms for 673 Nyanga East squatters ((ed.) et al. (1983)). Most of the remaining literature is legal-doctrinal, reading the Urban Areas Act and its Section 10, the 1952 Abolition of Passes Act, the Black Labour Act 1964 and its 1965 Regulations, Proclamation R74 of 1968 and the homeland Status Acts (Chaskalson and Duncan (1984); Frankel (1979); Davenport (1969)). It also draws on the Stallard, Fagan, Tomlinson and Riekert commission reports.
Archival and field evidence is local and thin. Maylam uses Durban Town Clerk's and Native Affairs Department files plus the 1936 and 1946 censuses; Durban's African population rises from about 71,000 in 1936 to about 150,000 in 1949 (Maylam (1990)). Bonner reuses a pass-record survey of some 20,000 Johannesburg workers (1936-44), East Rand municipal reports and oral histories (Bonner (1995)). Posel reads Federated Chamber of Industries memoranda and surveys, Johannesburg Non-European Affairs Department files and IANA congress proceedings for the 1950s (Posel (1990a)). She uses Fagan and Tomlinson evidence for Afrikaner organisational positions (Posel (1987)). For the 1980s, Greenberg and Giliomee interviewed labour bureau and TEBA officials in four Bantustan areas in April 1982 (Greenberg (1987); Giliomee and Schlemmer (1985)). The Up Against the Fences volume adds Centre for Applied Social Sciences attitude surveys (for example 676 migrants and a quota sample of 478 contract workers), Urban Foundation interviews and the 1970/1980 censuses via Smit and Kok (Giliomee and Schlemmer (1985)). Cobbett et al. use a resident questionnaire of unstated size and interviews with 12 firms at Botshabelo (Cobbett and Nakedi (1987)). Manona offers two genealogical case histories from Grahamstown (Manona (1984)). The only post-apartheid individual-level evidence is in Kok et al.: the 2001-02 HSRC Migration Survey (3,618 completed interviews), and the 1980 and 1996 census migration questions (Kok et al. (2006)).
Contested evidence¶
The sharpest disagreement is over how to read the same enforcement series. Prosecutions fell steeply after 1967/68. Frankel and Savage hold that this marks no relaxation, attributing it partly to Aid Centre diversion and mainly to housing, contract-labour and workplace controls (Frankel (1979); Savage (1984); Savage (1986)). Their support is an ad hoc Aid Centre counterfactual and housing backlog figures, and Savage concedes he cannot separate enforcement intensity from reporting changes or from the shift of arrests to Administration Boards. Greenberg uses Savage's figures alongside official interviews to argue instead that control was chronically breaking down (Greenberg (1987)). On effectiveness, the 'failure' position rests on stocks of 'illegal' residents, the return of deportees and rising urban totals (West (1982); (ed.) et al. (1983); Davenport (1969); Giliomee and Schlemmer (1985); Chaskalson and Duncan (1984)). The 'significant slowing' position rests on Simkins's census-based finding that African urbanisation rates declined in the 1960s-70s and on his counterfactual urbanisation estimates (Hindson (1987); Greenberg (1987); Savage (1984)). The two sides are partly answering different questions: absolute urban growth versus deviation from an unobserved counterfactual. Kok et al. add a third reading from UN residual estimates and the 1980/1996 census migration questions (Kok et al. (2006)). They find net rural-urban migration was very low, did not surge after 1986, and the migrant share fell from about 13 to 11 per cent; but the 1975-80 base excludes the TBVC states. The 'displaced urbanisation' argument turns largely on classification. Cobbett et al. and Simkins treat commuter settlements such as Botshabelo, and the homeland parts of metropolitan areas (about 80 per cent of net metropolitan in-migration, 1980-85, on unreferenced figures), as functionally urban (Cobbett and Nakedi (1987); Simkins (n.d.)). Census urban/rural and homeland boundaries would score the same people as rural, so verdicts on effectiveness depend on the spatial unit chosen. On purpose, Posel cites the 1956 FCI survey (38 per cent of firms used the bureaux) and a NEAD manager's interview to argue that the urban labour preference policy was inoperative, contrary to Hindson's allocative reading (Posel (1990a)). That claim rests on selected employer documents, and both sides lack placement records. Savage's labour-channelling-to-exclusion periodisation is inferred from the timing of prosecution peaks against legislation (Savage (1986)). Maylam and Davenport read the same statutes and commissions differently on whether 1937 and 1948 were breaks, with Maylam stressing municipal continuity (Maylam (1990); Maylam (1995); Davenport (1969)). Posel uses Fagan and Tomlinson evidence to contest O'Meara's single-blueprint reading (Posel (1987)). The categories overlap, and none is a direct count.
Underused and missing sources¶
Almost every evidence ledger flags the same gaps. No study uses Commissioner's Court registers, endorsement-out and admission-of-guilt records, or individual-level prosecution files (West (1982); Savage (1984)). No study uses labour bureau registration, requisition and placement returns, or Administration Board and Department of Bantu Administration files on illegal residents and deportee returns (Greenberg (1987); Posel (1990a); Chaskalson and Duncan (1984)). Firm-level hiring and wage data by worker origin, which would test the Hindson-Posel dispute, are also absent. So are agricultural census series on farm employment, tenancy and mechanisation for the farm-labour and farm-to-town arguments (Jeeves and Crush (1990); Manona (1984)). Consistent census time series of urban African population, sex ratios and birthplace by town or magisterial district are largely missing as well; most authors take Simkins's corrected figures second-hand rather than re-deriving them (Hindson (1987); Bonner (1995); Maylam (1990)). The Chamber of Mines/TEBA recruiting records are unused for the apartheid period (Kok et al. (2006)).
Strength of key claims¶
| Basis | Share of key claims |
|---|---|
| Systematic data | 11% |
| Partial data | 49% |
| Illustrative or anecdotal | 20% |
| Asserted | 20% |
Comparative figures for all debates are on the Literature overview.
Best-evidenced works¶
Savage (1984), Savage (1986), (ed.) et al. (1983), Giliomee and Schlemmer (1985), Posel (1990a), Kok et al. (2006)
Most-used primary sources¶
Holdings are matched by machine against the archive's file manifest; a dash means no match was found, not that the project certainly lacks the source.
Works and their positions¶
| Work | Position |
|---|---|
| (ed.) et al. (1983) | Holds that the policy failed to reverse African migration to the Western Cape. Rural poverty makes illegal urban work worthwhile even with repeated arrests, and the state conceded on housing and Crossroads when pressed by economic forces. |
| Baldwin (1975) | Official policy of reducing the urban black population has been softened and permanent urbanisation partly acknowledged. Nonetheless the thrust of legislation is to convert workers into migrants, including daily Bantustan-border commuters. |
| Bonner (1995) | Critiques them as abstract and unable to explain why migrants settled rather than oscillated; stresses women's agency, housing, wage-payment forms and culture. |
| Bonner (1995) | Disputes this: the new Nationalist government decided to stabilise the existing urban population in family units while tightening influx control against new arrivals, partly to serve farmer and mining labour demands. |
| Bonner (1995) | Rejects the view that first-generation immigrants are merely transitional and that urbanisation has a clear point of completion; urban commitment often remained conditional. |
| Chaskalson and Duncan (1984) | Treats the system as a mechanism that ties urban presence to white labour demand and channels workers into low-paid unpopular jobs. It also argues that attempts to curb migration were unenforceable against economic forces. It makes no formal engagement with the liberal/revisionist literature. |
| Chaskalson and Duncan (1984) | The 1968 policy of restricting family housing and promoting commuting and hostels failed to reduce the urban black population and produced overcrowding, illegal residence and contempt for the law. |
| Cobbett and Nakedi (1987) | Argues influx control was restructured rather than abolished: pass-law offenders are now evicted under squatting and trespass laws, and 'urban' is redefined as daily access to the urban labour market. |
| Davenport (1969) | Argues policy was not always monolithic: the 1923 Act was mild, the Stallard doctrine was opposed, and the Smuts government considered abandoning influx control in the 1940s. |
| Davenport (1969) | Rejects the view that influx control had been abandoned in favour of undermining rights. Removal of 'redundant' Africans intensified after 1964, notably in the western Cape. |
| Frankel (1979) | Holds that, whatever its economic origins, it now serves primarily political control, independent of other functions, while recognising that it also serves labour-supply interests of agriculture and mining. |
| Frankel (1979) | Sceptical: the changes are cosmetic, the basics are unchanged, and the 1978 legislation threatens Section 10 rights. |
| Giliomee and Schlemmer (1985) | Bernstein, Schlemmer, Relly, Oppenheimer and Rosholt argue influx control is ineffective, costly and contrary to free-market growth; business is portrayed as no longer supporting it, though Schlemmer notes industrialists hesitate to demand its abolition. |
| Greenberg (1987) | Rejects the critical consensus (Magubane, Budlender, his own earlier work, and the state's self-image) that the system effectively abrogated the market. He argues it was incomplete, unevenly enforced and self-undermining. |
| Hindson (1987) | Holds that influx control and removals slowed and reshaped urbanisation (citing Simkins), so the 'illusory' label is unfounded for earlier decades, though the long-term aims of territorial apartheid were never met. |
| Jeeves and Crush (1990) | Notes that research shows influx control failed to protect low-wage employers' labour supplies. The authors want to know why this did not force wage improvements or the end of migrancy. |
| Kok et al. (2006) | Chapter 4 says oscillating migration kept mine labour cheap and externalised social costs, and influx control reflected employer demands for state regulation of the labour market plus the state's unwillingness to grant political rights. This is broadly a cheap-labour view. |
| Kok et al. (2006) | Wentzel & Tlabela say labour bureau controls were largely ignored and urbanisation escalated. Anderson says rural–urban migration was far below what it would otherwise have been until the 1990s. Together they say controls were only partly successful but still suppressed formal urban growth. |
| Kok et al. (2006) | Gelderblom and De Jong & Steinmetz criticise income-maximisation models; they emphasise immobilising structures, networks, norms and life satisfaction, and seek a synthesis. |
| Kok et al. (2006) | Anderson and Wentzel & Tlabela find no immediate jump after 1986 and little change in aggregate mobility, partly because of displaced urbanisation and peri-urban commuting. |
| Manona (1984) | Argues that without restrictive 'homeboy' groups, kin sponsorship gives migrants a wider choice of adjustment, and that the inflow has 'ruralized' the city. |
| Maylam (1990) | Neither was a stark alternative. Following Hindson, policy moved toward a system of 'differentiated labour-power'. Durban leaned Stallardist in the 1920s and accepted stabilisation in the 1940s. |
| Maylam (1990) | Economic imperatives mattered, but the explanation must also include white middle-class mentality, police and ratepayer pressure. Segregation was often not in employers' interests. |
| Maylam (1995) | Notes both modify earlier determinism. Hindson stays materialist while Posel stresses labour supply plus limiting urbanisation. Calls for more attention to local variation and the central-local relationship. |
| Posel (1987) | Challenges the conventional wisdom: the SAAU accepted continuing, controlled African urbanisation as an economic necessity. |
| Posel (1990a) | Rejects the functional-compatibility claim. She argues the state's administrative categories cut across employers' skill-based demand, and that industrialists could have raised wages themselves rather than rely on state barriers. |
| Sapire and Beall (1995) | Backs Bonner's challenge to the image of a unilinear, irreversible family move to town. Urbanisation was incremental, and rural ties were retained until the 1940s. |
| Savage (1984) | Treats Riekert-style reform as a tightening of controls plus privileges (not rights) for a select group of urban 'insiders', leaving the distribution of power unchanged. |
| Savage (1984) | Rejects the view that they indicate relaxation or deterrence from higher fines. Attributes the fall partly to Aid Centres and mainly to new enforcement through housing, labour bureaux and contract labour, and deportation of 'foreigners'. |
| Savage (1984) | Holds that the laws serve both an inclusionary labour-supply function and an exclusionary security function, with the balance shifting after 1948 and especially in the 1960s. |
| Savage (1986) | Rejects the view of imminent abolition. Arrests rise whenever reform is announced, and control is shifting to housing, contract labour and deportation. |
| Savage (1986) | Holds both functions operate in tension, with the balance shifting from inclusion (to c.1950) to exclusion (after 1950). |
| Simkins (n.d.) | Notes that abolition will open non-homeland metropolitan areas to black immigrants and that population growth there is already accelerating. Argues that rural populations may still grow, and thinks the Urban Foundation projections are probably too low on urbanisation. |
| Simons (1956) | Holds that both aims run throughout, with labour supply and control central to the survival of the system and bound up with the colour-class structure. |
| Simons (1956) | Rejects it: the reference book consolidates passes, makes control more efficient, and leaves the residence, work-seeking and contract restrictions intact. |
| Simons (1956) | Argues that Africans will remain and multiply in towns as the only large labour reserve, while the state insists on migrancy to protect the master-servant relationship. |
| West (1982) | Argues it has failed to stem urbanisation. People and deportees simply return, so the policy yields enormous hardship but no durable control. |
| West (1982) | Sees it as a cornerstone of apartheid. It sustains the claim that blacks are not permanent urban residents, and it also distributes black labour. |