Capitalist Agriculture: Accumulation and the Labour Supply - The Case of South Africa¶
Jill Nattrass (1976). Capitalist Agriculture: Accumulation and the Labour Supply - The Case of South Africa. SALDRU Farm Labour Conference, September 1976, Paper No. 8. SALDRU, School of Economics, University of Cape Town.
| Type | conference paper |
| Scope | Economics and economic history of apartheid |
| Method | descriptive quantitative |
| Period studied | 1911–1973 |
| Regions | South Africa, White farming areas, African homelands/reserves, Natal, Cape |
| Themes | agriculture land, mining migrant labour, bantustans removals influx, labour markets wages, demography urbanisation, business and state, comparative theory, inequality poverty unemployment |
| Citation key | nattrass1976 |
Argument¶
Nattrass tests the neo-Marxist view that South African development rested on a clean split between capitalist (White) agriculture, which supplies food, and the African reserves, which supply labour. She argues the split is too clear-cut. Because many Africans lived on White farms, capitalist agriculture was itself a major supplier of labour to the rest of the capitalist economy. Using census data and compounding population growth rates, she estimates large net and gross outflows of people from White farms between 1921 and 1973. These flows are linked to eviction, casualisation and growing use of migrant labour after 1950. She concludes that the farm sector failed to deliver adequate wage goods, but supplied about half the growth in the African non-agricultural workforce. This probably depressed African wages and shifted surplus population into the reserves. State labour controls after 1948 also strengthened farm capital against labour.
Evidence and method¶
Population censuses (1911-1970), Agricultural Censuses (1954, 1960, 1971, 1973), Union Statistics for Fifty Years, South African Statistics 1974, Reserve Bank capital stock estimates (de Jager), van der Horst's 1936-7 farm labour data, the Beaumont, Tomlinson, Marais and du Plessis commissions, and the author's own doctoral thesis estimates of temporary migrants.
Place in the debates¶
- Neo-Marxist reserve/labour-reservoir thesis (Legassick, Kaplan, Rex, Williams): capitalist agriculture as labour absorber and reserves as labour suppliers (see Cheap labour thesis and the function of the reserves). Agrees with the general thrust but argues the dichotomy is too simplistic. White farms were themselves major labour suppliers to the non-agricultural sector, and the reserves were filled partly by people evicted from farms.
- Push versus pull causes of labour outflow from agriculture (see Farm labour: mechanisation, labour shedding, wages and labour control). Argues for a combination. Pull from labour scarcity in non-agricultural jobs drew White, Coloured and Asian workers out. Push from farmer eviction of squatters and labour tenants, and substitution of casual or migrant labour, worked mainly on Africans.
- Purpose of apartheid-era state labour controls (stabilising farm labour versus supplying other sectors) (see Farm labour: mechanisation, labour shedding, wages and labour control). Rejects both simple readings. The controls let the state allocate African labour between competing fractions of capital, within the long-run constraint that farm employment must fall.
- Whether capitalist farming has settled or migrant labour (Wilson) (see Farm labour: mechanisation, labour shedding, wages and labour control). Supports Wilson's claim of increasing reliance on migrants once regional patterns are accounted for, using 1970 gross flow estimates.
Findings¶
-
Many Africans lived on White farms rather than in reserves, so the farm sector had to be a labour supplier. (p. 8)
estimated that in 1916 only 55 per cent of the African population were resident in the African reserve areas
-
Capital accumulation on farms before 1971 mostly widened rather than deepened capital, so it did not displace labour. (p. 13)
apparently used on a substantial scale to augment the labour supply of the sector
-
Between 1951 and 1970 a massive permanent exodus of Africans left White farms for towns and homelands. (p. 22)
During these 19 years more than 1,5 million Africans left the capitalist
-
Many of those pushed out were resettled in homelands that already could not support their populations. (p. 22)
and nearly 700 000 were resettled in the rural areas of the African homelands.
-
Gross flows reveal a much greater use of migrant labour on farms by 1970 than in 1936. (p. 24)
as much as 42 percent of the African labour force in the capitalist agricultural sector could have been composed of migrant labourers on the census date
-
Capitalist agriculture was an important source of labour for the non-agricultural economy across all race groups. (p. 25)
41 per cent of the Asian labour force and 20 per cent of the African labour force employed in the non-agricultural capitalist sector in 1970.
-
Demand for African labour outside agriculture barely outpaced population growth, so inflows from farms probably held down African wages. (p. 26)
The demand for African labour in the modern sector appears to have barely kept pace with the natural increase in population over this 19 year period.
-
Farm output failed to keep pace with demand, so food prices rose relatively and became a constraint on development. (p. 28)
Output levels have not grown sufficiently to prevent relative increases in food prices and this failure in the supply of the critical component of wage goods is still a major constraint
-
Up to half of recent African workforce growth in the non-agricultural sector may have come from farms. (p. 30)
in recent years as much as 50 per cent of the increase in the African workforce employed in the non-agricultural capitalist,sector may well have been drawn from capitalist agriculture
-
After 1950 farmers moved from settled labour to casual and then contract migrant labour, strengthening capital against labour. (p. 30)
capitalist farmers substituted first casual labour for full~time African employees and then contract migrant workers for their settled workforce
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Union Statistics for Fifty Years (employment in agriculture by race, population census data 1911-1951; output and employment series) | official statistical series | 1911-1960 (employment series 1911-1951) | Supplies agricultural employment by race for census years 1911-1951 used in Tables 1, 2 and A1 to estimate labour flows. |
| South African Statistics 1974 | official statistical series | 1921-1973 | Source of agricultural output, 1960 and 1971 employment, and the population growth rates by race used to compound employment forward and estimate migration. |
| Agricultural Census Nos. 34, 44 and 46 (1959-60, 1970-71, 1972-73), plus Census No. 28 (1953/54 labour on farms) | agricultural census | 1953/54-1973 | Provides farm employment by race (including casual workers), capital investment in improvements, and African farm population for 1960, 1971, 1973 used in Tables 1-3 and A1. |
| Monthly (Quarterly) Bulletin of Statistics, December 1975 | official statistical series | 1951-1975 | Supplies economically active population outside agriculture, African manufacturing wages, and urban African counts for growth rates in Table 6 and the 1951-70 job estimates. |
| Statistical Year Book 1964 | official statistical series | 1950-1964 | Used with South African Statistics 1974 for numbers of permanent African farm employees in Table 3. |
| South Africa 1974, Official Year Book | official statistical series | 1974 | Cited for population growth and for subsistence output per worker in the homelands (immiserisation claim). |
| 1936 Population Census (Sixth Census) Vol. 1 and Ages and Marital Condition of the Bantu Population | population census | 1936 | Gives the share of Africans resident on White farms (33 per cent) and the 1936 settled population for the permanent-migration estimates. |
| Population Censuses 1946, 1951 (Vol. V Ages), 1960 (Vol. 7 No. 1) | population census | 1946-1960 | Actual population counts by area and age used as the 1951 and 1960 end-points in Table A2 migration estimates. |
| Report 02-02-02, Population Census 6 May 1970, Sample Tabulation Bantu Age, Occupation, Industry, School Standard and Birthplace | population census | 1970 | Basis for estimates of gross temporary migrant flows by age, language group and industry in 1970 (Tables 5, A3-A7). |
| Beaumont Commission (Native Land Commission) Report UG 19/1916 | commission or official report | 1916 | Estimates that only 55 per cent of Africans lived in reserves in 1916, used to argue the dichotomisation was incomplete. |
| Tomlinson Commission Report UG 61/1955 (cited as 1961) | commission or official report | 1950s | Evidence that African rural areas could not support their populations by the early 1950s, so resettled people were unlikely to make a viable living. |
| Marais (Second) and du Plessis (Third and Final) Reports of the Commission of Enquiry into Agriculture, RP 84/1970 and RP 19/1972 | commission or official report | 1970-1972 | Taken as reflecting official policy shift toward capitalising agriculture and sharply reducing non-White farm labourers. |
| Report of the Commission on Conditions of Trade and Industries UG 10-12/1912 | commission or official report | 1912 | Cited for rising Johannesburg living costs and for the official view that agriculture must develop first. |
| de Jager, 'The Fixed Capital Stock and Capital Output Ratio of South Africa from 1947-1972' | secondary literature compilation | 1946-1972 | Supplies the agricultural capital stock series for 1946-1972 used for capital per man and capital-output ratios. |
| Sheila van der Horst, Native Labour in South Africa (1936-37 farm worker data) | secondary literature compilation | 1936-37 | Data on regular farm workers resident on farms and migrant share (11 per cent) in 1936 as benchmark against 1970. |
| Nattrass, The Migrant Labour System and South African Economic Development 1936-1970 (PhD thesis); Nattrass 1976 SAJE article | secondary literature compilation | 1936-1970 | Source of the estimated numbers of temporary migrants by area used in Table A2 and the rural out-migration estimates. |
| Annual Reports of the Department of Bantu Administration and Development RP 41/1972, 41/1973; Transkei Statistical Labour Report 1971 | commission or official report | 1971-1972 | Used to estimate the industrial distribution of migrants by language group (Table A7). |
| Hansard, House of Assembly, 5 March 1971 (Minister of Bantu Affairs reply), via Survey of Race Relations 1971 | parliamentary debates | 1971 | Source for the distribution of foreign African migrants by industry (Table A7 column F). |
| J.L. Sadie, Projection of the South African Population 1970-2020 | secondary literature compilation | 1970 | Source of female population by age used to estimate homeland male migrants (Table A4). |
| Frankel, Capital Investment in Africa (state assistance estimate 1910-1936) as cited via Kaplan | secondary literature compilation | 1910-1936 | Estimate that state aid to White agriculture exceeded R113 million, comparable to mining tax receipts. |
New data¶
Author-constructed estimates of net permanent African and other race out-migration from White agriculture 1921-1973 (Tables 2, 3, A1-A2), gross temporary migrant flows for 1970 (Tables 5, A3-A7), and an agricultural capital stock/employment index (Table 1), derived by compounding census employment at population growth rates.
Inference¶
Quantification: Descriptive tables.
Residual (balancing) estimation: employment at one census date is compounded forward at population growth rates and compared with actual employment to infer net migration, supplemented by 1970 census-based gross flow estimates and descriptive trend tables; conclusions are then interpreted in a neo-Marxist/dual-economy framework by comparing time periods and race groups.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Capitalist agriculture has been a major supplier of labour to the rest of the capitalist sector, providing 24% of Whites, 33% of Coloureds, 41% of Asians and 20% of Africans employed outside agriculture. (p. 25) | Residual migration estimates from census employment and population growth rates (Table 2, A1-A2), compared with non-agricultural economically active counts. | Partial data |
| More than 1.5 million Africans left capitalist farming in 1951-1970, about 1 million to urban areas and nearly 700,000 to the homelands. (p. 22) | Table 4/A2 estimates using 1936-1970 census counts, assumed natural increase rates, and the author's temporary migrant estimates; 1960-70 outflow is a balancing figure. | Partial data |
| Outmigration from agriculture accounted for almost 50 per cent of the increase in the African labour force outside agriculture 1951-1970, likely depressing African wages. (p. 27) | Comparison of estimated job growth (1.5 million) with supply components from Table A2 and Bulletin of Statistics data; wage effect inferred. | Partial data |
| Capitalist farmers substituted casual for permanent African labour in the 1950s, and by 1970 relied heavily on migrant contract labour (up to 42% of the farm labour force). (p. 22) | Table 3 casual/permanent estimates and 1970 census-based gross migrant estimates (Table 5) compared with van der Horst's 1936 data. | Partial data |
| Capital accumulation in agriculture to 1971 was mainly capital widening (capital-output ratio constant, capital per man grew slowly), augmenting rather than replacing labour; transformation began only after 1971. (p. 13) | Table 1 indices built from de Jager capital stock, census employment and output series. | Partial data |
| Agricultural output has not grown fast enough to prevent relative food price increases, so wage-good supply remains a constraint and the non-agricultural profit rate may have fallen. (p. 12) | Illustrative growth-rate arithmetic with assumed elasticity and growth parameters; compared with output series in Table 1; no food price data presented. | Illustrative or anecdotal |
| The state's attitude is shifting toward forcing a sharp reduction of non-White farm labour, as reflected in the Marais/du Plessis Commissions and a farm labour levy. (p. 27) | Commission reports RP 84/1970 and RP 19/1972 quoted; levy mentioned in a footnote. | Illustrative or anecdotal |
The author's caveats¶
Acknowledges that the aggregated data may conceal the nature of the transformation, that the residual method yields only an 'impact' of migration (natural increase included), that temporary migrants could be allocated between farms and homelands only for 1970 (otherwise arbitrarily), biasing the permanent outmigration estimates, that the 1960-70 White farm outflow is a balancing figure, that substitution of casual for permanent labour cannot be converted to labour input units, that foreign migrants are all treated as migrants (upward bias), and that the mean-rate versus intercensal-rate choices change results.
Gaps in the evidence¶
The migration estimates are residuals that depend on assumed uniform growth rates and census undercounting, with no direct measure of eviction or flows such as farm-level or labour bureau records. Evidence on food prices, marketed surplus and profit rates is not presented to support the wage-goods and profitability claims, and the 'pull/push' causal story is inferred rather than tested. Farm-level records, labour bureau registrations, and 1970-type census cross-tabulations for other dates (partially available then and in archives now) would test the argument better.
Replicability: Data described but not public.