Homelands, separate development, decentralisation and removals¶
Were the homelands and industrial decentralisation viable and effective, and what were their real purposes (labour reservoirs, dumping grounds, political legitimation)?
Works engaging the debate: 43
Positions¶
- Decentralisation spontaneous and incentives marginal (Bell)
- Incentive-driven or politically motivated (Wellings and Black, Pickles)
- Homelands as labour reservoirs vs dumping grounds for surplus people (Surplus People Project)
- Separate development economically unviable (Tomlinson debates)
Evidence base¶
The debate has been argued from a small set of official serial sources and a larger body of secondary estimates. On decentralisation, the core quantitative record is the industrial census and Census of Manufacturing tabulated by region or magisterial district. Bell (1973) built a regional concentration series from 1934-5 to 1959-60 out of Department of Statistics unpublished district records and Special Reports 240, 249, 250 and 257. Using the same reports he also constructed an ordinal divertibility index across 129 sectors. Bell (1986) extended the series through the censuses of 1956/7, 1961/2, 1963/4-1967/8, 1970, 1972, 1976 and 1979. It shows the four main industrial areas' share of manufacturing employment falling from 81.9% after 1965/6 to 76.9% in 1979, while border areas and homelands rose from 11.75% to 15.54%. Administrative programme data supplement this. Permanent Committee for the Location of Industry reports supply border-area employment and state assistance figures, including R79.7m in 1960-68 (Best (1971)) and about 23,000 induced jobs by 1970 (Bell (1973); Katzen (1974)). Board for the Decentralisation of Industry annual reports and licence applications cover 1970-92 (Tomlinson and Addleson (1986); Kerby (2016a); Kerby (2016b)); they underpin Kerby's figure of 214,394 RIDP jobs and 4,164 firms by 1992. Firm surveys provide the motive evidence: Bell's 1961 East London questionnaire, the 1983 postal survey of 13 growth centres (N of roughly 75-80 firms) (Todes and Watson (1984)), the Addleson et al. 1985 survey of decentralised industrialists, and interviews with Ciskei and Transkei textile managers (Tomlinson and Addleson (1986)). Spatial outcome tests rely on GGP by magisterial district for 1968-94 (McCarthy (2000a); McCarthy (2000b)) and on 1911 versus 2011 census place populations (Krugell (2014)). Wage evidence for homeland zones has had to be rebuilt from press clippings and promotional almanacs (Kerby (2014)).
On homelands as labour reservoirs or dumping grounds, the base is mainly demographic. Census-based indirect estimation (1936, 1946, 1960, 1970) gives 1.49 million male oscillating migrants in 1970 (Nattrass (1976b); Nattrass (1976c)). Census population distribution figures show the share of Africans living in their 'own' homeland rising from 41.8% to 47.8% between 1970 and 1980 (Christopher (1982)), and the de facto bantustan share rising from 39.5% in 1960 to 54% in 1980 (Hallett (1984)). Sabbagh's dot maps from the 1921-1960 censuses document 'stratified integration' (Sabbagh (1968)). Homeland economic viability is assessed through the Tomlinson Commission's figures (Sabbagh (1968); Best (1971)). Abedian rebuilt Transkei national accounts from Auditor-General records and a 1981 business survey, finding GDP of R547.5m against GNP of R1,115m in 1980/81 (Abedian (1982)). Regional case work draws on Ciskei agricultural and planning archives and interviews (Wotshela (2004)) and on the Botshabelo resident and employer interviews (Cobbett and Nakedi (1987)). Policy-purpose arguments rest on readings of Riekert, the Good Hope Plan and the SECOSAF manuals (Pickles (1988a); Pickles (1988b); McCarthy (1982)).
Contested evidence¶
The sharpest clash is over the same regional manufacturing employment shares. Bell (1986) reads the post-1965/6 decline in the main industrial areas as market-driven. His grounds are timing: the decline predates the 1967 Physical Planning Act and large incentives. He also cites clothing and textile sector shares. Tomlinson and Addleson (1986) accepts the trend but argues three things. First, Bell's categories lump deconcentration points into 'rest of South Africa'. Second, enterprise-location and net-output shares of the core did not fall (Pretorius et al.; Bell 1985), which implies relocation of low-productivity, labour-intensive operations. Third, the Addleson survey ranks incentives above cheap labour for homeland locators. Hindson (1987) sides with Bell against McCarthy and Maasdorp. Attribution of induced jobs depends on counterfactuals. Bell's figure of about 5.2% of the 1960-66 national increase assumes a 12.2% (1959/60) baseline share (Bell (1973); Katzen (1974)), so the induced-job count, and the derived cost of about R1,500 per job, move with that assumption. Pickles treats the Wellings and Black statistic (decentralised share rising from 12.9% in 1978 to 19.3% in 1984) as evidence that incentives work, without any causal test (Pickles (1988a); Pickles (1988b)). Kerby (2016b), by contrast, finds Phase 2 incentive disbursements roughly flat and attributes growth to infrastructure. Kerby also argues that Taiwanese clustering (84% in 11 of 63 sites) reflects networks and female labour supply rather than rebate gaps, though only through ranked tables. McCarthy's GGP data show non-metro districts doing relatively better in the 1960s-70s, but the link to policy is asserted (McCarthy (2000a)). Krugell's two-endpoint Zipf exponent (0.783 to 1.356) cannot date any policy effect (Krugell (2014)). Removal totals rest on incompatible compilations: SAIRR's 1,820,000 for 1960-70, built from 'projections, induction and estimates' (Baldwin (1975)); a figure of about 2.1 million from Black Sash and SAIRR (Christopher (1982)); and SPP's 3.5 million for 1962-80 (Hallett (1984); Cobbett and Nakedi (1987)). These estimates carry overlapping categories and unexamined methods. The dumping-ground versus labour-reservoir reading also turns on inference rather than data. Mare (1981) deduces that removed people were surplus from the absence of business protest. Cobbett and Nakedi (1987) relies on one former administration-board director's account of Botshabelo's purpose. Wotshela (2004) uses DLA restitution figures, showing about a third of families moved without official transport, to argue that the state did not control resettlement.
Underused and missing sources¶
Mabin explicitly notes that census and agricultural-census series on Africans on white farms from 1946 went unused (Mabin (1991a)). Also missing are Permanent Committee reports, Department of Planning section 3 application records (used only by Bell (1973)) and homeland development corporation records. Sources rarely used by anyone in the debate include Department of Bantu Administration and Co-operation and Development removal records, Manpower Survey regional wage data (only (Kerby (2014)) uses it), Chamber of Mines and TEBA recruitment records for migrant estimates, and homeland budget and Auditor-General accounts beyond Abedian's Transkei reconstruction.
Strength of key claims¶
| Basis | Share of key claims |
|---|---|
| Systematic data | 13% |
| Partial data | 55% |
| Illustrative or anecdotal | 18% |
| Asserted | 15% |
Comparative figures for all debates are on the Literature overview.
Best-evidenced works¶
Bell (1973), Bell (1986), Kerby (2016b), Nattrass (1976c), Abedian (1982), Todes and Watson (1984)
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 |
|---|---|
| Abedian (1982) | He judges decentralised industry capital-intensive and subsidy-dependent, so it neither reduced migrant influx nor generated local income. Transkei is described as dependent on South Africa, not as an independent economy. |
| Baldwin (1975) | Rejects the view that rhetorical impracticality means the policy has failed. He warns that underestimating the system's pragmatism and resilience creates dangerous illusions about change. |
| Bell (1973) | Rejects the view that decentralisation is justified on economic grounds; the 'overriding' motive is racial separation, and policy should be judged against that aim and its growth costs. |
| Bell (1973) | Notes the government rejected in-Reserve development for political reasons (avoiding 'white spots') and that border areas are cheaper on infrastructure; presents the border policy as a political compromise. |
| Bell (1973) | Prefers market-based inducements, but argues controls are more effective at limiting African urban growth and cheaper in budget terms. Controls shift costs onto firms and workers and hurt confidence. |
| Bell (1973) | Labour cost is secondary to markets and materials; differentials are real and large (border areas 45-66% of southern Transvaal African earnings) and defensible, and cheap labour is decisive only in a few industries such as clothing and textiles. |
| Bell (1973) | Sceptical: agglomeration helped early growth, the Schumann Committee contradicts itself, and there is no evidence that border areas can reach self-sustaining growth. |
| Bell (1986) | Rejects the Tomlinson-era aim of stemming urbanisation and Hirsch's legitimation thesis as the main explanation of persistence, since the job numbers are too small. Suggests the primary current aim is a federal political structure, with containing Black unions as a possible secondary effect. Gives more weight to Lacey's economic interpretation than Hirsch does. |
| Bell (1986) | Argues that market and competitive forces, including labour-market segmentation, wage differentials and rising Black productivity, were the main cause. Incentives and controls played a smaller role than commonly thought, although the post-1982 incentive package probably matters. |
| Bell (1986) | Close to Davies in stressing competitive pressures as the fundamental force changing the racial division of labour, but adds emphasis on the changing inter-regional structure of production. |
| Bell (1986) | Doubts their explanatory power for the 1960s-70s trend and their long-term sustainability. Treats them as a pure income transfer from taxpayers to rural capitalists, citing the Ciskei cost of over R30,000 per job. |
| Best (1971) | Sceptical: the scheme yields marginal decentralization and negligible homeland development, is politically motivated, and the new centres are mere appendages of the metropolitan areas, in line with critics such as Nieuwenhuysen and Van der Horst. |
| Best (1971) | Sees the government's rejection of Tomlinson's homeland-based industry as a missed opportunity and the later 'agency basis' as a belated partial return to its recommendations. |
| Best (1971) | The scheme is rational only if total territorial segregation is the goal; the apartheid structure itself blocks real regional growth. |
| Bromberger (1978) | Reports the view of de Klerk and Bell that homeland neglect served the capitalist economy but says it is 'not entirely proven'. He calls for research on homeland economies and Lipton's 'two agricultures'. |
| Christopher (1982) | Skeptical: the partition policy is unresolved and contested, the 'independent' states are unrecognised and economically peripheral, and meaningful consolidation is doubtful. The author treats it as a political-geographical phenomenon worthy of study rather than a solution. |
| Christopher (1982) | Holds that influx control, group areas and administrative transfer of black townships to homeland areas allowed large-scale 'repatriation', but that this did not stop the demographic trend toward a black majority in white areas. |
| Cobbett and Nakedi (1987) | Rejects the state's claim that people moved voluntarily or fled ethnic conflict; argues removal was planned, with violence orchestrated at Thaba'Nchu and farm labour expelled. |
| Crush and Wellings (1983) | Rejects dismissal of 'independence' as mere fraud and adopts the radical/Marxist view (Legassick-Wolpe, Innes-O'Meara, Molteno) that the Homelands are redefined for accumulation and control; adds tourism capital as an overlooked element in Pretoria's strategy. |
| Giliomee and Schlemmer (1985) | Greenberg and Giliomee argue the homelands no longer 'cheapen' labour via subsistence but serve as dumping grounds and underbelly of a stratified order of privilege; Nattrass and Giliomee acknowledge the earlier reservoir role and the radical challenge. |
| Giliomee and Schlemmer (1985) | McCarthy, Maasdorp and Oppenheimer argue decentralisation is politically motivated, costs jobs overall and cannot stop metropolitan growth; the Good Hope proposals modestly better but limited. |
| Giliomee and Schlemmer (1985) | Giliomee judges rural revolt unlikely (migrancy, cleavages, state control) but sees rural misery being exported to cities, with squatters and unions as potential destabilisers. |
| Hallett (1984) | Hallett holds that the SPP reports show no real abandonment of apartheid's basic tenets; removals continue with less overt force and more 'voluntary' rhetoric. |
| Hallett (1984) | Hallett endorses the SPP framing: the economy requires fewer unskilled workers, so those removed are 'surplus' and dumped in the bantustans. |
| Hallett (1984) | He sees the state as powerful but not monolithic: community resistance, liberal NGO support and financial constraints can force it to hold back locally. |
| Hindson (1987) | Criticises McCarthy and Maasdorp for ignoring Bell's work questioning whether state inducements were the primary cause of industrial dispersal. |
| Horrell (1968) | Reports official admissions (BIC head Adendorff) that homelands cannot support even natural increase, and border industries absorb only 4.4 per cent of African labour-market entrants; sceptical of success. |
| Horrell (1970) | Records the government's own position: Africans were in white areas only for their labour, and apartheid is presented as 'differentiation' of 'multi-national' groups. No analytical stance of its own. |
| Horrell (1970) | Reports the United Party's estimate that full development of the Reserves would cost about six times current spending, and the government's reluctance to give a total cost. |
| Houghton and Dagut (1973) | Presents Hoernle's three alternatives, the Fagan gradual integration view and the Tomlinson recommendation of separate development; later Oppenheimer argues for African advancement. |
| Katzen (1962) | Argues removal is administratively possible but economically costly, producing slower growth, lower investment and higher costs. This is an early liberal-economist cost-of-apartheid argument. |
| Katzen (1974) | Reporting Bell, the review holds that neither market imperfections nor external diseconomies justify decentralization in South Africa; agglomeration economies are more likely, and any case must rest on socio-political grounds. |
| Katzen (1974) | The review reports that the policy was costly and ineffective at diverting jobs, and that controls over African employment encouraged more capital-intensive production. |
| Kerby (2014) | Cites Bell (1986, 1987) that international competition, rising costs and political pressure also drove relocation. Hypothesises that lower unskilled wages raise a zone's attractiveness for labour-intensive industry. |
| Kerby (2016a) | Declares some truth on both sides. Policy aimed at anchoring Africans in the homelands (political motive) and at cheap-labour competitiveness, as exposed by Taiwanese inflows. |
| Kerby (2016a) | Finds Phase 2 incentives roughly flat and infrastructure spending more important, supporting Bell on that phase. For Taiwanese firms argues geography, incentives and networks interacted, with incentives not the primary driver. |
| Kerby (2016a) | Argues economic factors (MFA quotas, NT$ appreciation, rand weakness, labour costs, incentives) and ethnic networks, not only diplomacy, explain investment and clustering. |
| Kerby (2016a) | Suggests, non-causally, that Taiwanese migration and investment raised and diversified exports to Taiwan, while acknowledging sanction-busting and re-export as alternative explanations for homogeneous exports. |
| Kerby (2016a) | Concludes RIDP largely failed to generate labour-absorbing employment (214,394 jobs by 1992) and mainly relocated jobs, though a Taiwanese production-network agglomeration emerged. |
| Kerby (2016b) | Argues some truth lies on both sides; the RIDP was a political tool to anchor Africans in the homelands but later phases were also driven by economic pressures, and Taiwanese investment exposes both logics. |
| Kerby (2016b) | Sides with Bell that incentives were less decisive than commonly thought in Phase 2 (incentive outlays flat, infrastructure expanded), while in Phase 3 incentives rose sharply; for Taiwanese firms he argues geography and networks interacted with incentives rather than incentives alone explaining clustering. |
| Kerby (2016b) | Challenges the diplomatic-only reading; stresses Taiwanese offshoring, MFA quotas, exchange rates, financial liberalisation, property-rights protection and low-cost labour as complementary economic drivers. |
| Kerby (2016b) | Finds exports to Taiwan growing fastest, partly consistent with re-export/conduit suspicions for homogeneous minerals, but differentiated goods growth cannot be explained by this and is associated with Taiwanese investors (not causal). |
| Kerby (2016b) | Argues all three interacted; first-mover Taiwanese production networks, market access and female labour pools explain clustering in four of sixty-three zones, with incentives amplifying rather than explaining it. |
| Krugell (2014) | Argues these policies failed to stop agglomeration; border industries and homeland growth points had limited impact, and wealth did not spread to border regions. |
| Mabin (1991a) | Traces it to liberal Native Affairs ideas but argues that as implemented it deprived the majority of land and cattle, benefiting only a few, and was fiercely resisted. |
| Mare (1981) | Treats relocation as enforcement of influx control that moves the unemployed and potentially unemployed out of white South Africa into the bantustans. |
| McCarthy (1982) | Sceptical of the social-cost case for decentralization, which he says is unproven. He allows that taking work to the workers has merit if overall employment and income growth is not sacrificed and locational economics is respected. |
| McCarthy (1982) | Argues that the separate-development ideology underlies or transcends the other obstacles. He holds that the 1982 reforms only partly escape it. |
| McCarthy (1982) | Sees the proliferation of growth points, and the 49 development points in the new policy, as undermining the agglomeration advantages needed for self-sustained growth. He is sceptical that infrastructure can be politically concentrated on a few. |
| McCarthy (1982) | Favours small-scale industry for its stronger regional linkages and lower investment per job. He doubts that the incentive package can really foster the informal sector. |
| McCarthy (1988) | Critical on economic grounds. Heavy infrastructure spending at remote points, while metropolitan infrastructure such as the Western Cape's is under-utilized, makes job creation costly. |
| McCarthy (2000a) | Finds only a modest effect. Non-metro areas kept 89-95% of metro growth in the 1960s-70s but 84-85% in the 1980s-90s. Overall spatial patterns persisted, and policy rarely deflects historically formed growth patterns. |
| McCarthy (2000b) | Sceptical: patterns are path-dependent and policy-induced deviations are small. The non-metro share of growth was higher in the 1960s-70s (89%, 95%) than in the 1980s-90s (84%, 85%), but the underlying hub-and-rim structure persisted. |
| Nattrass (1976b) | Says the imbalance resulted largely from market forces but was strengthened by institutional elements, so the market alone cannot correct it and government intervention is needed. |
| Nattrass (1976c) | Holds that the imbalance 'has largely resulted from the operation of market forces but has been strengthened by certain institutional elements', so government intervention is needed rather than markets alone. |
| Nattrass (1976c) | Supports continued decentralization, Homeland urban housing and cautious Homeland industrialization, and phasing out migrancy over the long run. This is reformist and accepts the separate-development framework. |
| Pickles (1988a) | Draws on the 'surplus reserve army' framing and treats regional policy as serving the labour-supply and accumulation needs of capital and the state. |
| Pickles (1988b) | Describes homelands as a 'surplus reserve army' and sees recent policy as integrating them into functional labour-market regions. |
| Pickles (1988b) | Notes earlier failure even by government accounts, but points to some changes in location decisions and rising decentralised manufacturing share, while remaining cautious about the evidence. |
| Rex (1975) | He sees the question of whether border industries are seriously pursued as an open and important argument. Political sociologists differ on how far the Bantustans threaten stability, and Rex leans pessimistic. |
| Sabbagh (1968) | Sees Bantu as already integrated into the White economy and argues that Separate Development runs against that trend and cannot succeed. |
| Sabbagh (1968) | Holds that the two are incompatible, and that where they conflict (Border Industries, Tomlinson cuts) politics prevails over economic rationality. |
| Schmidt (1975) | Sceptical: argues that weak spread effects and the flow of labour to the core make viable Homeland growth centres unlikely, so separate development is internally inconsistent. |
| Schmidt (1975) | Reformist: holds that greater spatial integration and co-optation of black counter-elites would reduce conflict and promote development; apartheid is likely to prove transitional. |
| Simkins (n.d.) | Critical of past homeland 'development' for providing infrastructure where incomes are low and out-migration high. Sceptical that decentralised-manufacturing incentives can be replaced easily, given vested interests. |
| Smith (1972) | Homelands cannot support their existing populations, let alone returning migrants. The urban African population's status as 'foreign' homeland citizens is a major flaw in the policy's logic. |
| Todes and Watson (1984) | Argues the growth centre model has failed on all three counts (self-sustaining growth, quality of life in the centre, spread to the region), as in studies elsewhere, and that failure is not a teething problem. |
| Todes and Watson (1984) | Partly sceptical of Bell. The authors say industries were sometimes forced out of metropolitan areas and that incentives were a reason for location in almost half of cases, though labour supply matters. They hold that incentives mainly distort location and encourage inefficient location. |
| Todes and Watson (1984) | Holds the policy was never simply developmental. It was governed by contradictory imperatives (curbing migration, bolstering homelands, white constituency politics), and homeland unemployment is the direct result of state policy. |
| Todes and Watson (1984) | Inconclusive. Non-independence shows no clear negative or positive effect in the data, because local linkage is low everywhere. |
| Tomlinson and Addleson (1986) | Partly sceptical of Bell. They accept that a trend exists and that much of the employment shift cannot be counted as policy-induced jobs. They hold that his case is not proven and that incentives, the Environment Planning Act and labour costs together shaped the outcome. For firms locating in the homelands they would reverse his ranking, putting policy first and labour second. |
| Tomlinson and Addleson (1986) | Bell's implication is that the policy was ineffective and the concessions a needless cost. The authors question this, because the post-1982 incentives now direct location decisions, though they concede that pre-1982 decentralization largely occurred without them. |
| Tomlinson and Addleson (1986) | They note Bell's labour-discipline argument and question its narrowness. They argue the state is responding to high labour costs by relaxing influx control, promoting 'orderly urbanization' just inside homeland borders and favouring deconcentration points. |
| Venter (1979) | Critical of South African land policy: argues the 1936 land allocation and unresolved consolidation undermine the credibility of separate freedoms, though the author accepts the homeland framework and supports developing the homelands' core areas. |
| Venter (1979) | Rejects the puppet view. Argues Matanzima used separate development as a stepping-stone to black nationalism. |
| Venter (1979) | Cautiously optimistic about industrial growth points and agricultural potential, while stressing subsistence-economy constraints and dependence on South Africa. |
| Wilson (2011) | Argues that reserve poverty was generated by the dynamics of white-controlled industrialisation (labour drain, declining rural investment, forced removals), not pre-existing or incidental. |
| Wittenberg (2003a) | Decentralisation aimed to preserve central control by fragmenting opposition and coercing Africans; partition failed because land and economic resources were niggardly. Independence was 'not a complete sham' because it complicated direct intervention. |
| Wotshela (2004) | Accepts that the state drove the policy but argues that local actors (homeland politicians, chiefs, planners, relocatees) shaped its implementation and outcomes. Moves beyond the Surplus People Project's focus on removals as events to their long-term consequences. |
| Wotshela (2004) | Notes the shift from the 1960s onward to social control and housing surplus people. Stresses that land was dispensed for political ends, with planners making only limited provision for viable agriculture. |
| Wotshela (2012) | Holds that homeland industrial units and irrigation schemes were costly, ineffective showcases that collapsed with the administration. It also stresses the agency of local organisations that rebuilt community and authority. |
| Wotshela (2012) | Treats betterment as having left a lasting legacy in land-use ideas and settlement layouts. 1990s land invaders were successive generations of those who resisted betterment. |