Some Geographical Characteristics of a Plural Society: Apartheid in South Africa

Literature › Works

M. Ernest Sabbagh (1968). Some Geographical Characteristics of a Plural Society: Apartheid in South Africa. Geographical Review, Vol. 58, No. 1. American Geographical Society.

Type	journal article
Scope	Economics and economic history of apartheid
Method	mixed
Period studied	1904–1966 (with projections to 2000)
Regions	South Africa, Bantu Reserves, Transkei, Witwatersrand, Border region of the Eastern Cape, Natal (Tugela basin)
Themes	bantustans removals influx, mining migrant labour, demography urbanisation, manufacturing industrial policy, agriculture land, labour markets wages, politics ideology elections, business and state
Citation key	sabbagh1968

Argument

Sabbagh argues that pluralism, meaning racial separation under apartheid, is the main force shaping South Africa's geography. It produced a depleted Reserve economy beside a thriving White economy that depends on Bantu labour. Demographic, migration and urbanization evidence shows 'stratified integration' deepening over decades, so Separate Development is an attempt to reverse an established trend. He examines the Tomlinson Commission programme, the government's cuts to it, the Border Industries scheme and the Reserve villages. He concludes that the political and economic motives are incompatible and that economic feasibility is repeatedly sacrificed to political desirability.

Evidence and method

1904-1960 population censuses (the Bantu census is acknowledged to be unreliable), the Statistical Year Book, the Tomlinson Report (1955) and the 1956 White Paper, and secondary studies (Houghton, Leistner, Mayer, Nieuwenhuysen, Fridjohn, Fair and Green). Population distribution maps, one life history of a migrant worker, and Bantu Investment Corporation loan figures.

Place in the debates

• Integration versus separation: can apartheid reverse the economic integration of Africans? (see Homelands, separate development, decentralisation and removals). Sees Bantu as already integrated into the White economy and argues that Separate Development runs against that trend and cannot succeed.
• Political versus economic motives for apartheid policy (see Homelands, separate development, decentralisation and removals). Holds that the two are incompatible, and that where they conflict (Border Industries, Tomlinson cuts) politics prevails over economic rationality.
• Sanctions and boycotts effectiveness (see Sanctions, disinvestment and the anti-apartheid movement). Briefly suggests in a footnote that boycotts have not worked because they hurt the Bantu they are meant to help.

Findings

• The Reserves are small, fragmented and economically marginal. (p. 2)

  The Bantu Reserves, some 260 fragmented areas, occupy 13 percent of the land area of the country

• Migrancy is near-universal among Reserve men, which makes the Reserves dependent on absentee wage labour. (p. 14)

  At any one time probably 40 percent of the working-age males of the Reserves are absent, and practically every able-bodied male leaves the Reserves at some time in his life, temporarily or permanently.

• Most Africans already lived in White areas by 1960, which undercuts the logic of territorial separation. (p. 15)

  Of the 1960 Bantu population of eleven million, about 60 percent (6.5 million) had settled in White areas, about half of whom were living and working on farms and half in cities and industries.

• Urban African numbers grew rapidly between 1911 and 1960. (p. 17)

  From 1911 to 1960 the urban Bantu population increased from less than half a million to almost 3.5 million

• Africans remained concentrated in unskilled work despite long industrial employment. (p. 19)

  less than 1 percent of the male Bantu labor force were classified as skilled (professional and technical).

• The Tomlinson plan required removing huge numbers from Reserve land to create viable farm units. (p. 21)

  the Commission planned the removal from the land of about half of the present population-that is, some 300,000 families.

• The government accepted Tomlinson in theory but gutted its funding and barred White capital, undermining Reserve industrialization. (p. 21)

  The investment funds were reduced by nearly two-thirds; any prospect of a successful urban-industrial development was virtually eliminated

• Border industries such as the Good Hope Textile Mill succeeded commercially but did not develop the Reserves, because workers were migrants living in locations. (p. 24)

  The first border industry, the Good Hope Textile Mill, was established in 1946 in King Williams Town. By 1964 it had become the largest textile mill in South Africa, employing some 3500 Bantu.

• Low African wages restrain the domestic market, and Separate Development will not fix this. (p. 27)

  Clearly the economy would be given a boost if Bantu wages were upgraded sufficiently that the Bantu could play a consumer role.

• Industry concentrates in a few metropolitan nodes, so decentralization conflicts with the economy's geography. (p. 28)

  The major economic islands of South Africa (focused on Johannesburg, Durban, Port Elizabeth, and Cape Town) employ 80 percent of all industrial workers in the country

• The central conclusion: economic feasibility gives way to political goals. (p. 29)

  economic feasibility is consistently sacrificed in favor of political desirability.

Evidence ledger

Sources used

Source	Kind	Period	How used
Population censuses 1904, 1911, 1921, 1951 and 1960 (Table I demographic structure; 1921-1951 and 1951-1960 Bantu density maps)	population census	1904-1960	Supplies racial population numbers, shares and growth rates, and the district-level maps of total and Bantu population distribution and density change.
Statistical Year Book 1964, Pretoria	official statistical series	1951-1960	Compilations of the 1960 census used for the population data and the 1963 immigration figures.
Statistical Year Book 1966 (urban White and Bantu population, 1904-1960, Table II)	official statistical series	1904-1960	Source of the urban population totals and White and Bantu shares showing the rising Bantu urban share.
Tomlinson Report: Summary of the Report of the Commission for the Socio-economic Development of the Bantu Areas (1955)	commission or official report	1955 (data to 1954)	Source of the population projections A and B, absenteeism estimates, economic farm unit and cost estimates, the Reserve boundary map, and the program that Separate Development builds on.
Government White Paper F (1956), Memorandum of Government Decisions on the Recommendations of the Tomlinson Commission	commission or official report	1956	Documents the government's reduction of investment funds and its rejection of white capital and freehold tenure.
Report of the Native Economic Commission (1932) and Social and Economic Planning Council Report No. 9 (1946)	commission or official report	1932, 1946	Cited as earlier official investigations preparing the ground for Separate Development.
Official immigration figures for 1962-1963	official statistical series	1962-1963	Shows net White immigration exceeding the Tomlinson assumption, which bears on the realism of the projections.
G. M. E. Leistner sample survey of urban Bantu labour, Pretoria metropolitan area (1964)	secondary literature compilation	c. 1960s	Provides the five categories of urban Bantu by permanence, tribal job preferences, and job turnover by age 23.
D. Hobart Houghton, Men of Two Worlds (1960), migrant life history and manpower estimates	secondary literature compilation	1906-1953 life history; 1960 estimates	Supplies the illustrative migrant case (John Mvalo) and estimates of two million migrants, unused manpower and 370 million man-miles.
Philip Mayer studies of East London Xhosa migrants and townsmen	secondary literature compilation	c. 1950s-1962	Used for the social cleavages between migrants and townsmen, and between Red and School Xhosa.
D. J. Gouws, Labour Resources of the Southern Transvaal (1964)	secondary literature compilation	1960	Source for the 66 percent Bantu share of the labour force and the under 1 percent skilled.
Houghton, Economic Development in the Reserves (1962) and Kayaletu survey; King Williams Town district figures	secondary literature compilation	c. 1959-1962	Evidence on the shortfall of economic farm units (3400 for 11,400 families) and household incomes in Kayaletu.
Nieuwenhuysen articles on economic policy in the Reserves (1964)	secondary literature compilation	1960-1964	Source for the government's departures from Tomlinson, the Border Industries policy, rural villages and the manufacturing in the Reserves.
Fridjohn, Border Industries (Race Relations Journal, 1962)	secondary literature compilation	1962	Source for the investment requirement of R1000 million and for the critique of the border industries.
Fair and Green, Development of the Bantu Homelands (Optima, 1962)	secondary literature compilation	c. 1960s	Supports the claims on industrial concentration (80 percent of industrial workers in four centres) and the Tugela basin failure.
Survey of Race Relations in South Africa 1960 (SAIRR) and Bantu Investment Corporation data	official statistical series	1952-1966	Data on general merchants in the Reserves (1199 to 2373) and on Bantu Investment Corporation loans and purchases of White-owned stores.
South African Summary, Vol. 5, No. 32, 1966	newspapers periodicals	1966	Source for the Bantu Investment Corporation loan figures by 1966 and the purchases of White trading stations.
Mortimer, Border Industries and the Bantustans (African World, 1964)	newspapers periodicals	1964	Source for the Good Hope Textile Mill's employment of about 3500 Bantu.

Inference

Quantification: Descriptive tables.

Descriptive, interpretive geographic analysis. It reads census tables and dot maps of population distribution and density change, then draws on official reports (above all the Tomlinson Commission) and secondary studies. Claims about integration versus separation and the economic infeasibility of Separate Development are argued by synthesis and reasoning, not by statistical testing.

Key claims

Claim	Basis	Strength
The racial proportions of the population were nearly constant from 1904 to 1960, so no demographic swamping occurred, although the Bantu share is projected to stay high. (p. 6)	Census totals 1904-1960 in Table I and the Tomlinson projections A and B.	Systematic data
Most Bantu live in White areas, not the Reserves, and the Bantu population map pattern closely mirrors the total population pattern, showing stratified integration that Separate Development opposes. (p. 10)	1921, 1951 and 1960 census maps (Figs. 2-5) and Tomlinson Report estimates of resident and absentee Reserve population.	Partial data
Migrancy is pervasive and structural: about 40 percent of working-age Reserve men are absent at any time and virtually all leave at some point. (p. 13)	Tomlinson Report p. 53 estimates, Houghton's studies, and an illustrative life history.	Partial data
The urban Bantu population grew rapidly and Whites became a minority in every South African city by 1960. (p. 16)	Statistical Year Book 1966 (Table II) and 1960 census city figures (Pretoria, Witwatersrand, Johannesburg).	Systematic data
The government gutted the Tomlinson program by cutting funds by two-thirds, refusing white capital and rejecting freehold tenure, leaving agricultural stabilization and Reserve industrialization hopeless. (p. 20)	Tomlinson Report, White Paper F, and Nieuwenhuysen's summary of the departures, with Houghton on the farm units.	Partial data
Border industries and Reserve rural villages do not create viable Reserve economies; they perpetuate migrancy and low wages. (p. 24)	The Good Hope Textile Mill case, Fridjohn's critique, Nieuwenhuysen on villages, and Bantu Investment Corporation loan data.	Illustrative or anecdotal
Economic concentration in a few nodes makes dispersal to the Reserves contrary to economic trends, and economic feasibility is sacrificed to political desirability. (p. 27)	Fair and Green's data on industrial employment shares and the Tugela basin, plus the author's argument.	Partial data

The author's caveats

The author warns that census data for the Bantu, both inside and outside the Reserves, are unreliable because of the transiency of the Bantu and the methods of enumerating, and notes that survey counts do not agree with the census. Reserve population trends cannot be assessed in detail from the census figures outside the Transkei.

Gaps in the evidence

The argument relies mostly on one official report (Tomlinson) and on secondary studies, with no direct archival evidence on how policy was made or enforced. It would be better tested with district-level census data on Reserve outmigration and wages, together with Bantu Investment Corporation and Border Industries records on firms, subsidies, and jobs. Much of this existed then, and the archives and census microdata are accessible now, but the author does not use them.

Replicability: Data described but not public.