Partition and Population in South Africa¶
A. J. Christopher (1982). Partition and Population in South Africa. Geographical Review, Vol. 72, No. 2. American Geographical Society.
| Type | journal article |
| Scope | Wider apartheid literature |
| Method | descriptive quantitative |
| Period studied | 1913–1980 |
| Regions | South Africa, Bantustans/black states (Transkei, Bophuthatswana, Venda, Ciskei, KwaZulu, QuaQua), Witwatersrand, Natal, Cape Province |
| Themes | bantustans removals influx, demography urbanisation, politics ideology elections, agriculture land, origins segregation |
| Citation key | christopher1982 |
Argument¶
Christopher uses the preliminary 1980 census to assess the government's policy of territorial partition into separate black 'national states'. He argues that the policy rests on the de facto land division created by white settlement and on the white minority's wish to keep power. It has made some demographic progress: a larger share of blacks now live in the state of their own language. But half of the black population still lives outside its own state, 'white' South Africa remains ethnically mixed, and the economic and urban cores stay in white hands. Meaningful consolidation is therefore doubtful. The argument is built from census tables, maps and the legislative and commission history.
Evidence and method¶
Preliminary 1980 census statistics (South Africa and Bophuthatswana, with estimates for Transkei and Venda), unpublished 1980 census returns, government commission reports (Tomlinson 1955; 1960 commission on European occupancy of rural areas), land legislation, and secondary literature on homeland consolidation and population transfers (Lemon, Bundy, Tatz, Smit and Booysen, Waite/Black Sash, SAIRR).
Place in the debates¶
- Viability and meaning of separate development / homeland independence (see Homelands, separate development, decentralisation and removals). 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.
- Population removals and urban 'repatriation' as the instrument of partition (see Homelands, separate development, decentralisation and removals). 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.
Findings¶
-
Native reserves were a small share of the land when entrenched in 1913, and were already inadequate for the black population. (p. 3)
The land in native reserves was only 7.8 percent of the total area of the country but was already inadequate for the needs of the black population
-
The 1936 Act added land but mostly ratified the existing situation. (p. 3)
In 1936 the Natives Trust and Land Act provided for an additional 6.2 million hectares to be purchased by the government for the expansion of black reserves.
-
Consolidation gained little economic territory, and major urban centres stayed white. (p. 5)
Significantly, few areas of any economic importance have been added to the black states, and the major urban centers adjacent to them have remained in white hands.
-
Administrative transfer of urban black housing to homeland areas achieved 'repatriation' without physical movement. (p. 6)
Thus four-fifths of the black population in the Durban metropolitan region live in housing estates in KwaZulu.
-
Large-scale physical removals took place in the 1960s-70s, mostly from white farms. (p. 7)
It is estimated that as many as 2.1 million persons were transferred to the black states between the mid-1960s and the mid-1970s.
-
The removals targeted labour tenants and squatters on white farms, following the official alarm about 'blackening' of rural areas. (p. 7)
The majority constituted rural residents, either labor tenants (750,000) or rent-paying squatters (750,000) on white-owned farms.
-
The white share of 'official' South Africa rose but fell within the 1970 boundaries. (p. 7)
(from 17.5 percent to 18.7 percent white) may be taken by the government as some evidence of successful pursuit of the policy. However, in the area of the 1970 boundaries, the white proportion of the population continued to decline to 16 percent.
-
The white rural ratio worsened despite removals. (p. 8)
In 1980 the ratio was 1:10.6, white to other groups, in the white rural areas.
-
Policy had partial success on its own terms: more blacks lived in their own-language state. (p. 8)
In 1980 almost half, or 47.8 percent, of the black population of South Africa resided in the state of its own language in contrast with 41.8 percent a decade earlier
-
The largest homelands are peripheral and fiscally dependent on white South Africa. (p. 8)
They are essentially peripheral to the South African spatial economy and are dependent on financial support from the white areas.
-
A black urban population with common economic participation has little in common with the rurally based national states. (p. 13)
produced another black urban group with little in common with the rurally based black national states that are being established.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Preliminary Census Statistics 1980 (South Africa, Department of Statistics, 1980) | population census | 1980 (with comparisons to 1970) | Provides the 1980 population by race and language group, the share of each black group living in its own state, and the white share of the population, used to assess the policy of separate development. |
| Preliminary Census Statistics 1980 (Bophuthatswana Department of Statistics, Montshiwa, 1981) | population census | 1980 | Supplies census figures for Bophuthatswana, supplementing the South African preliminary results; Transkei and Venda figures were estimates. |
| Unpublished census returns for 1980 from the South African Department of Statistics | population census | 1980 | Used for the 1980 ratio of whites to other groups (1:10.6) in white rural areas, compared with the 1921-1960 trend. |
| Report of the Commission of Inquiry into the European Occupancy of the Rural Areas (1960) | commission or official report | 1960 | Cited for the official alarm over the 'blackening' of white rural areas that preceded action against labor tenancy. |
| Summary of the Report of the Commission for the Socio-Economic Development of the Bantu Areas (Tomlinson Commission, 1955) | commission or official report | 1955 | Establishes the commission's role in changing the terms of debate on the black areas and its proposal to consolidate the reserves in seven blocks. |
| Report of the Ciskei Commission (Government of Ciskei, 1980) | commission or official report | 1980 | Cited for Ciskei's move toward independence within a confederal framework. |
| National Accounts of the Black States, 1972 to 1976, Report 09-17-04 | official statistical series | 1972-1976 | Cited in the discussion of QuaQua's overpopulation and population shifts. |
| Survey of Race Relations in South Africa for the Year 1978 (SAIRR) and B. Waite, South Africa - A Land Divided against Itself (Black Sash, 1977) | secondary literature compilation | mid-1960s to mid-1970s | Underpins the estimate that 2.1 million persons were transferred to the black states, including 750,000 labor tenants and 750,000 squatters. |
| Natives Land Act 1913, Natives Trust and Land Act 1936, Natives (Urban Areas) Act 1923, Group Areas Act 1950 | legislation gazette | 1913-1950 | Statutory framework for reserve land (9.5 million ha in 1913, plus 6.2 million ha in 1936) and for urban segregation and 'repatriation' of blacks. |
| P. Smit and J. J. Booysen, Urbanisation in the Homelands (University of Pretoria, 1977) | secondary literature compilation | 1970s | Supports the claim that urban segregation allowed administrative transfer of urban blacks to the black states (e.g. 4/5 of Durban's blacks in KwaZulu). |
| Maps and proposals for consolidation (Lemon 1976; Lombard 1980; Best and Young 1972) | secondary literature compilation | 1972-1980 | Source for the 1973 and 1975 consolidation proposals and for KwaZulu's territorial fragmentation (48 blocks to 10 units). |
Inference¶
Quantification: Descriptive tables.
Descriptive political-geographical assessment: 1980 census figures are compared with 1970 and earlier proportions, supported by maps and a historical narrative of legislation and commissions. There is no statistical testing.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| The share of the black population living in the state of its own language rose from 41.8 percent in 1970 to 47.8 percent in 1980, so the policy made gains, but half of blacks still live outside their own state. (p. 133) | 1980 preliminary census (Table I) compared with 1970 figures. | Systematic data |
| Official South Africa became whiter between 1970 and 1980, but within the 1970 boundaries the white share continued to decline to 16 percent. (p. 132) | 1980 preliminary census compared with 1970 census figures. | Systematic data |
| In the white rural areas the ratio of whites to other groups worsened from 1:3 in 1921 to 1:8.5 in 1960 and 1:10.6 in 1980, despite removals of blacks. (p. 133) | Census series 1921-1960 and unpublished 1980 census returns. | Partial data |
| About 2.1 million people were transferred to the black states in the 1960s-70s, mostly labor tenants and squatters from white farms. (p. 132) | Estimate drawn from Black Sash and SAIRR secondary compilations. | Partial data |
| Urban segregation laws let the state administratively 'repatriate' urban blacks to the black states, except on the Witwatersrand and in Port Elizabeth, where the states are too distant. (p. 131) | Smit and Booysen's study and the legislation cited; no systematic data. | Illustrative or anecdotal |
| The black states are peripheral to the South African economy and dependent on the white areas, and few economically important areas were added to them in consolidation. (p. 133) | Asserted, citing Butler, Rotberg and Adams; observations on town transfers. | Asserted |
| The Witwatersrand and Johannesburg are ethnically heterogeneous, producing an urban black group with little in common with the rurally based black states. (p. 137) | Table II (1980 census Johannesburg counts by group). | Partial data |
The author's caveats¶
The author notes that Transkei and Venda census statistics were delayed and estimates were used, that the results are preliminary, that Ciskeian and Transkeian citizens are not separated in the census, and that enumerator bias may affect group classification in marginal cases.
Gaps in the evidence¶
The argument leans on preliminary census figures and secondary estimates, particularly the 2.1 million removals figure, whose basis is not examined. Final 1980 census returns, de jure and de facto population series, and administrative removal records would test the claims better. Such records exist, in part in state archives, but were largely not accessible then. Economic data on the black states' dependence are cited from secondary works rather than analyzed.
Replicability: Data described but not public.