Skip to content

Mechanisation in South African Agriculture 1936 - 1974

Literature › Works

Reggie Africa (1976). Mechanisation in South African Agriculture 1936 - 1974. SALDRU Farm Labour Conference, September 1976, Paper No. 35. SALDRU, Division of Research, School of Economics.
Type conference paper
Scope Economics and economic history of apartheid
Method descriptive quantitative
Period studied 1936–1974
Regions South Africa (white farming areas), Transvaal, Orange Free State, Cape Province, Natal, Homelands (as labour source)
Themes agriculture land, labour markets wages, bantustans removals influx, manufacturing industrial policy, demography urbanisation, inequality poverty unemployment, methods data sources
Citation key africa1976

Argument

The paper traces the commercialisation of South African agriculture and its intense post-1945 mechanisation, using census and official statistics on implements, capital formation, output and farm labour. It argues that mechanisation was driven mainly by technological availability, high postwar prices and the pull of expanding secondary industry. Labour shortage was not the prime cause. Unlike in the US or West Germany, mechanisation did not cut the regular farm labour force, which stayed roughly constant. The author concludes that this implies considerable underemployment, that farm wages and conditions stayed very poor, and that machine operators were hardly trained. Influx control is suggested as the main reason African workers were not pushed off farms. The question of underemployment is left open.

Evidence and method

Official statistics: agricultural censuses, Abstract of Agricultural Statistics 1975, Commission of Enquiry into Agriculture (R.P. 84/1970), Commission on implement cost/profit margins (R.P. 10/1962), Agrekon articles, Popping (1959) on capital formation, Financial Mail (1970, 1975), a S.A.A.U. labour survey, and the ILO and secondary histories (Goodfellow, Wilson, Tinley).

Place in the debates

Findings

  • Combine harvesters spread very rapidly after the war, which shows the scale of postwar mechanisation. (p. 6)

    The number of combine harvesters increased from 581 during 1936/37 to 8,940 in 1955 and in 1973 it stood at a figure of 19, 577

  • The machinery boom was concentrated in the 1940s and lost momentum by the mid-1950s, after which investment was largely replacement. (p. 7)

    The expansion in machinery which was the pher.lomenon of,'the 1940s, lost much of its impetus from about the middle of the 1950s.

  • Mechanisation was neither a response to labour shortage nor associated with a fall in the farm workforce. (p. 11)

    has,not resulted in a diminution of farm workers nor did mecha,nisation appear to have been stimulated by labour shortages.

  • Regular farm employment was roughly constant over the long run, using Commission data on employees to 1969. (p. 11)

    the number of regularly employed farm labourers has over a long period r,emained more or .less constant.

  • The reported 1969 farm labour shortage is treated as regional, with its causes unclear. (p. 12)

    However, the shortage seems to be of a regional nature and the reasons behintj .it are still obscured.

  • The farm population grew strongly, mostly African, in the 1960s. (p. 13)

    The total farm population increased from 2,955,000 in .1960 to 3,989,000 in 1969.

  • Capital invested in farm machinery and vehicles rose roughly tenfold between 1937 and 1968. (p. 16)

    Capital assets'in agricultural machinery and vehicles increased from R66 million in 1937 to R650 million in 1968.

  • Mechanisation was tied to the growth of agro-processing and secondary industry. (p. 20)

    The proportion of agricultural produce processed in some form or other increased significantly from 36% annually during 1947/48

  • Operators of the new machinery were almost entirely untrained, raising maintenance costs. (p. 21)

    established that during 1968 fewer than lO,OOO'of the approximately 180,000 African tractor operators in the Republic had received any sort of formal training

  • Mechanisation did not raise farm wages, which stayed extremely low. (p. 22)

    Mechanisation has not resulted in a general rise in wages for farm'workers. They still rank amongst the lowest paid employees in the country.

  • The average cash wage of farm labourers was very low in 1973. (p. 22)

    The average cash wage of farm labourers in 1973 was about Rll .50 a month.

  • Influx control is offered as the reason African farm labour was not pushed to the cities. (p. 22)

    Influx control s seem to have been the main factor in stemming the outflow of African labourers from the

Evidence ledger

Sources used

Source Kind Period How used
Agricultural censuses (Union/RSA), 1911–1950 and later, as reproduced via Agrekon Vol. 9 No. 2 April 1970 agricultural census 1911–1950 Supplies counts of ploughs, threshing machines, tractors and lorries and fertiliser tonnage to show growth in mechanisation and commercialisation.
Agrekon Vol. 9 No. 2 April 1970 (tables on volume of agricultural production, fertiliser, implements) secondary literature compilation 1910/11–1950/51 Source of the volume-of-production index in Table 4 showing slow growth to 1945/46 and faster growth thereafter.
Agriculture in the Economic Development of S.A. (Department of Agricultural Technical Services, 1963) commission or official report to 1963 Cited for the share of implements in agricultural capital assets and for the link between mining, secondary industry and agriculture.
Commission of Inquiry into Cost/Profit Margins in Respect of Agricultural Implements and Components thereof, R.P. 10/1962 commission or official report to 1962 Source for the implement numbers in Table 5 and, with agricultural census data, for the 1937–1973 implement stock figures.
Second Report of the Commission of Enquiry into Agriculture, R.P. 84/1970 commission or official report 1937–1969 Provides farm labour, casual labour, tractor/morgen and capital expenditure series (Tables 8, 10, 11) used to argue that the regular labour force stayed constant despite mechanisation.
Abstract of Agricultural Statistics 1975 official statistical series 1937–1974 Source for the number of farms, area and irrigation (Table 7), gross capital formation (Table 15) and export volume indexes (Table 17).
Supplementary Data to the Abstract of Agricultural Statistics, Jan. 1967 and 1966-67 official statistical series 1947/48–1965/66 Gives the indexes of agricultural production volume (Table 13) and tractor imports (Table 16).
State of South Africa Yearbook 1971 official statistical series 1961 Source of the distribution of farms by size group and value in 1961, used to show land concentration.
Report on Agricultural and Pastoral Production 1972-73, Agricultural Census No. 46 agricultural census 1973 Gives holdings by size group at 31 August 1973 to show concentration of land area.
F. Popping, 'Some features of Capital Investment in Agriculture in the Union of South Africa, Exclusive of Land Purchase', S.A. Journal of Agricultural Science Vol. 2 No. 3, 1959 secondary literature compilation 1936–1959 Supplies the net capital formation estimates (Table 14) used to show a post-war investment surge that faded in the 1950s.
Financial Mail, 18 December 1970; 4 April 1975 (and 13 Oct 1972 in bibliography) newspapers periodicals 1969–1975 Source for the farm population table (Table 12), the S.A.A.U. farm labour survey, the labour shortage, housing conditions and wages.
South African Agricultural Union (S.A.A.U.) survey of farm labour, 1969 author survey or fieldwork 1969 Gives the numbers of regular male workers by race, total farm population and the regional labour shortage in 1969.
Survey by the Division of Agricultural Engineering, Department of Agricultural Technical Services, 1968; J.J. Bruwer figures author survey or fieldwork 1968 Shows that few African tractor operators had formal training and that most workers on machines were untrained.
Agrekon Vol. 2 No. 4 Oct 1963; Vol. 4 No. 1 Jan 1965; Vol. 8 No. 2 April 1969; Vol. 14 No. 1 (articles on agriculture's role, processing, farmers) secondary literature compilation 1950s–1970s Supplies the farm labour population count, the processed-produce table (Table 18), the machinery cost shares and the farm entrepreneur and size statistics.
Annual Report of the Secretary for Agricultural Economics and Marketing, July 1973 - June 1974 commission or official report 1973–74 Source for agriculture's falling share of GDP (about 7.7% in 1973).
Crops and Markets Vol. 54 No. 636, July 1975, Division of Agricultural Marketing Research official statistical series 1958/59–1975 Source for expenditure on machinery, tractors and vehicles and other capital spending.
S.A.I.R.R. Survey of Race Relations 1968 secondary literature compilation 1968 Cited for the extent of machinery training of African farm labourers.
ILO, Mechanisation and Employment in Agriculture (1973); J.L. Meij (ed.), Mechanisation in Agriculture secondary literature compilation 1973 Used for international comparison (US, West Germany) and the effects of mechanisation on yields and land saving.

Inference

Quantification: Descriptive tables.

Descriptive time-series analysis of official statistics. Tables of implement stocks, land use, labour, capital formation, production and exports are read side by side. Reasons for mechanisation are inferred by correlation and argument rather than formal testing. There is some comparison with the US and West Germany.

Key claims

Claim Basis Strength
Intense mechanisation, especially of tractors, combines, trucks and electric motors, occurred after the Second World War, notably from the 1940s. (p. 6) Agricultural census counts and the R.P. 10/1962 and R.P. 84/1970 series (Tables 2, 5, 6). Systematic data
Despite heavy mechanisation, the regular farm labour force did not fall and tended to remain constant or rise. (p. 11) Commission of Enquiry into Agriculture R.P. 84/1970 series (Tables 8 and 10) and S.A.A.U. survey figures. Partial data
Mechanisation was driven by technological availability, high post-war prices and demand from secondary industry rather than by labour shortage. (p. 19) Price and terms-of-trade reasoning, tractor import and export tables and processed-produce shares from Agrekon; the correlation is inferred. Illustrative or anecdotal
Agricultural output volume rose strongly from the early 1950s, helped by mechanisation and by seed, fertiliser and other inputs. (p. 14) Index of production volume from the Supplementary Data to the Abstract of Agricultural Statistics and Agrekon. Systematic data
Capital investment surged after the war but lost impetus in the 1950s as replacement demand dominated, and running costs exceeded new investment. (p. 16) Popping's net capital formation estimates, Abstract of Agricultural Statistics 1975 and R.P. 84/1970 expenditure data. Systematic data
Land is concentrated in large farms and farm numbers fell while average size rose. (p. 9) Size group tables from the State of South Africa Yearbook 1971 and Agricultural Census No. 46 (1972-73). Systematic data
Farm workers' wages and living conditions remained very poor, with little training for machine operators; Coloured and Asian workers left farms while African workers were held by influx control. (p. 22) Financial Mail 1975, Agrekon figures, the Agricultural Engineering survey and the 1960/1970 shares of Coloured workers in agriculture; the influx control explanation is asserted. Partial data

The author's caveats

The author notes that statistics are missing for some years (marked with a dash). The text states that the extent of underemployment and the socio-economic effects of mechanisation 'still have to be determined.' The paper is a preliminary draft.

Gaps in the evidence

The argument never uses farm-level or household microdata, so it cannot test whether mechanisation displaced labour within farms, or whether it displaced seasonal workers whose numbers varied. The labour series stop at regular workers and the 1960s and cover only White-owned farms, with no data on wages over time. Farm-level surveys, the census manuscripts and Department of Agricultural Economics farm cost studies would have allowed this, and some of them exist in archives today.

Replicability: Data described but not public.