The Rationale of Apartheid

Literature › Works

Mats Lundahl (1982). The Rationale of Apartheid. The American Economic Review, Vol. 72, No. 5. American Economic Association.

Type	journal article
Scope	Economics and economic history of apartheid
Method	theoretical or conceptual
Period studied	1652–1970s
Regions	South Africa, Zimbabwe (Rhodesia)
Themes	origins segregation, bantustans removals influx, labour markets wages, agriculture land, mining migrant labour, comparative theory
Citation key	lundahl1982

Argument

Lundahl extends Richard Porter's (1978) heuristic model of a 'South African-type economy' by adding land alienation and skilled-job reservation. He also makes wages endogenous and divides history into three stages: settler intrusion up to the 1867 and 1886 mineral discoveries, the mining and industrial era to the end of WWII, and the postwar Bantustan and influx-control era. Comparative statics show how each white policy instrument (immigration, land alienation, job reservation, influx control) affects target variables for different white groups and African incomes. Land alienation created the reserves and a Lewis-type 'unlimited' labour supply, and policy instruments often set white groups against each other (landowners vs industry, skilled vs unskilled whites). He concludes that segregation's root cause may lie in white minority security fears rather than economic rationality, while pass laws and Bantustans also serve to cheapen unskilled labour by admitting only single male workers.

Evidence and method

No new empirical data. The paper is a theoretical comparative-statics model (full mathematics in a separate Lund working paper), illustrated with secondary literature on South African and Rhodesian history (Houghton, Wilson, Horwitz, Knight and Lenta, Rogers, Kane-Berman, Bundy).

Place in the debates

• Economic rationality vs white security as the root of segregation/apartheid (see Origins, coherence and ideology of apartheid and Afrikaner nationalism). Argues that segregation's root cause may lie in political security fears of a white minority rather than economic rationality. The Bantustan policy preserves white political dominance, though it also has a secondary economic role in cheapening labour.
• Cheap labour thesis / role of the reserves in determining African wages (see Cheap labour thesis and the function of the reserves). Holds that unskilled African wages in the white economy are tied to the marginal product of labour in the reserves, or to pass-law-limited migration. Land alienation and reserve over-crowding therefore reduce the supply price of labour.
• Racial wage discrimination and job reservation (Knight/McGrath, Porter) (see Racial wage gaps, wage convergence and wage determination). Treats white wages as endogenous, set by collective bargaining within restrictions on black labour supply. Job reservation creates conflicts between skilled and unskilled whites and between agriculture and industry.

Findings

• Land alienation and the creation of reserves expanded white agriculture and the supply of African labour to the white economy. (p. 3)

  the scope for white agriculture increased and the supply of African labor which was essential to the maintenance of the European standard of living in southern Africa could be increased in the white economy.

• Reserve populations grew on a shrinking land area, contradicting Porter's constant-returns assumption. (p. 4)

  As the African population grew, increasing numbers of people were concentrated on a shrinking land area.

• The reserve alternative sets a wage floor for African labour in the white economy. (p. 4)

  A policy that dictates a wage rate which falls short of what the Africans can obtain by staying in the reserves would never have been successful

• Mine owners tried to substitute cheaper African for white labour despite the colour bar. (p. 4)

  the mine owners did their best to substitute Africans for Europeans, since black laborers were willing to work for lower wages than whites.

• The 'poor white' problem put heavy pressure on rural whites to move to towns, driving demands for job reservation. (p. 7)

  The net result of all these factors was a tremendous pressure on some 200,000-300,000 whites to leave the countryside to find a job in the towns.

• In Stage II, land alienation benefits landowners without harming other white groups, while job reservation creates intra-white conflicts. (p. 8)

  It allows them to reach their goals without creation of adverse side effects on other white groups.

• In Stage III, a larger influx of African workers would benefit all white groups, which sits oddly with the rhetoric of separate development. (p. 10)

  an increase in the influx is beneficial for all white groups.

• Segregation may be explained by white minority security concerns rather than economic rationality. (p. 11)

  should not be sought in economic rationality, but in the fact that whites constitute a minority in South Africa

• Pass laws and Bantustans ensure only working migrants, not dependents, enter white areas, lowering labour costs and social overheads. (p. 11)

  Thus, the Bantustan policy and the pass laws can actually be employed to ensure that only African workers and not their dependents are allowed in white areas.

Evidence ledger

Sources used

Source	Kind	Period	How used
Mines and Works Regulation Act of 1911 (color bar in mining)	legislation gazette	1911	Cited as the legal origin of the color bar and job reservation for whites in mining, illustrating how white wages were set.
Horrell, The African Reserves of South Africa (1969), maize yield evidence as quoted by Porter	secondary literature compilation	c. 1940s-1960s (over 30 years)	Average maize yield per morgen is reinterpreted to argue that Porter's constant-returns-to-labor assumption for the reserves is unrealistic.
Knight and Lenta (1980), 'Has Capitalism Underdeveloped the Labour Reserves of South Africa?' - farm production per household in the reserves	secondary literature compilation	c. 1920-early 1970s	Used as evidence that reserve farm production per household fell steadily, supporting diminishing returns to labor in the reserves.
Houghton, The South African Economy (4th ed., 1976), on reserves and poor whites	secondary literature compilation	c. 1900-1970s	Source for population pressure and 'desert conditions' in the reserves, and for the account of backward white farming before WWII.
Houghton (1971), Wilson (1971) and Horwitz (1967) on the poor white problem	secondary literature compilation	post-WWI to 1940s	Relied on for the account of the poor white problem and white pressure for job reservation, including the 200,000-300,000 figure.
Horwitz, The Political Economy of South Africa (1967), summary of strikes and Rand Rebellion	secondary literature compilation	1911-1924	Source for the narrative of mine owners substituting Africans, white strikes, the Rand Rebellion and the 1924 change of government.
Rogers, South Africa: The 'Bantu Homelands' (1972), and Kane-Berman (1979) on urban African population composition 1960-1970	secondary literature compilation	1960-1970	Supports the claim that the urban African population shifted toward men of active age and that dependents were removed.
Knight and McGrath (1977), 'An Analysis of Racial Wage Discrimination in South Africa'	secondary literature compilation	1960s-1970s	Used for the upward movement of the formal color bar in the 1960s-70s and for the view that wages are determined by demand and institutionally determined supply.
Lundahl and Ndlela (1980), 'Land Alienation, Dualism, and Economic Discrimination'; Ndlela (1981)	secondary literature compilation	1652-1914	The author's earlier historical work underpins the claim that land alienation created the 'unlimited' labor supply.

Inference

Quantification: None.

Theoretical, comparative-static general-equilibrium modelling: an extension of Porter's heuristic model into three historical stages (pre-1867, 1867-WWII, post-WWII) with sign-pattern tables from totally differentiating the equations. Historical realism is argued through secondary literature and qualitative narrative; no empirical estimation or testing.

Key claims

Claim	Basis	Strength
Land alienation created the 'unlimited' African labor supply and increased white incomes while lowering African wages and reserve per capita incomes, if European agriculture is more land-intensive. (p. 1170)	Theoretical comparative statics (Table 1) plus the author's earlier historical work with Ndlela; the land-intensity assumption is simply posited.	Asserted
Porter's constant-returns assumption for the reserves contradicts the historical pattern of falling per capita incomes and diminishing returns as population concentrated on shrinking land. (p. 1171)	Horrell's maize yield evidence reinterpreted; Houghton 1976; Knight and Lenta's household production decline.	Partial data
White wages were endogenously determined by collective bargaining and restrictions on black labor supply; the color bar followed white strikes and the Rand Rebellion. (p. 1172)	1911 Mines and Works Regulation Act and the historical narrative from Horwitz.	Illustrative or anecdotal
In Stage III, an increased influx of African workers benefits all white groups, so the root of segregation lies in white minority security concerns rather than economic rationality. (p. 1178)	Model results (Table 3) combined with Porter's analysis and argument by interpretation.	Asserted
Pass laws and Bantustan policy let the white economy admit only male workers and not dependents, lowering the supply price of unskilled labor and shifting social costs to the reserves. (p. 1178)	Urban demographic change 1960-70 per Rogers and Kane-Berman; Knight and Lenta; theory of diminishing returns in reserves.	Partial data

The author's caveats

The author states the model is a stylized approximation; stage dates are 'only suggestive' and the stage concept should not be interpreted too literally. No attempt is made to explain transitions between stages. Behavioral assumptions such as full white employment and fixed capital stocks are acknowledged.

Gaps in the evidence

The argument rests on assumed factor intensities (European agriculture more land-intensive than African) and on fixed capital stocks, none tested with data. Evidence such as agricultural census data on land and labor per farm, reserve production and income surveys, and wage series by race and skill (including Chamber of Mines and industrial census data) existed then and would allow testing the sign predictions. The urban demographic claim relies on secondary sources instead of the 1960 and 1970 population censuses.

Replicability: Not applicable.