Skip to content

Review of Simkins and Clarke: Structural Unemployment in Southern Africa

Literature › Works

D. C. Hindson (1978). Review of Simkins and Clarke: Structural Unemployment in Southern Africa.
Type other
Scope Economics and economic history of apartheid
Method synthesis review
Period studied 1960–1977
Regions South Africa, Southern Africa, Reserves/homelands, Mozambique, Rhodesia, Malawi (labour suppliers)
Themes inequality poverty unemployment, mining migrant labour, labour markets wages, bantustans removals influx, methods data sources
Citation key hindson1978

Argument

Hindson reviews Simkins and Clarke's book Structural Unemployment in Southern Africa. He argues that unemployment in South Africa is far larger and more long-standing than commonly recognised, that it is mainly structural rather than cyclical, and that it is likely to worsen. He explains how Simkins builds a single unemployment index from fragmentary data by using production, income and recognition approaches to employment, and he notes its limits, such as the omitted informal sector. Clarke's paper shows that unemployment must be understood regionally: since the early 1970s foreign migrant workers have been displaced from South African employment, which exports unemployment to the labour-supplying states. Hindson closes by asking whether unemployment is inherent in the political and economic system and calls for work on the subsistence sector and on accumulation.

Evidence and method

Secondary review of Simkins' and Clarke's papers, which draw on the 1960 and 1970 censuses, industrial and agricultural censuses, Reserve output estimates (Tomlinson Commission), and UN and ILO unemployment yearbooks for international comparison. Hindson adds footnoted references to Sen, Knight, van der Merwe and ILO studies.

Place in the debates

Findings

  • Simkins estimates that unemployment nearly doubled in 1960-77 and that the rate rose to 22.4 per cent. (p. 1)

    the level of unemployment in South Africa nearly doubled between 1960 and 1977 and the rate of unemployment increased from 18,3 per cent to 22,4 per cent in the same period.

  • South Africa's unemployment rate far exceeds that of the industrial countries for which comparable estimates exist. (p. 1)

    nowhere amongst the major industrial countries for which comparable estimates are available is there a recorded rate of unemployment even half that estimated for South Africa.

  • Simkins' estimates omit the urban informal sector and so understate employment. (p. 2)

    Simkins' estimates make no allowance for this and insofar as such employment has gone unrecorded, he understates the level of employment,

  • Employment growth slowed between the 1960s and the 1970s as the economy grew more slowly. (p. 3)

    Between 1960 and 1969 employment increased at an average rate of 2,34% per annum and between 1969 and 1976 at 2,06% per annum.

  • Unemployment is overwhelmingly structural, with cyclical unemployment important only after 1974. (p. 3)

    One of Simkins' most important findings was that cyclical unemployment has been relatively unimportant in comparison with structural unemployment, except since 1974.

  • Most unemployment is hidden in the Reserves and in neighbouring rural areas. (p. 3)

    Cyclical unemployment is merely the tip of an iceberg most of which is hidden from sight in the South African Reserves and in the rural areas of neighbouring countries.

  • Clarke dates the displacement of foreign workers: relative from c.1950, absolute after 1972. (p. 3)

    Since about 1950 relative displacement of foreign workers began to take place and after 1972, with the increase in unemployment in South Africa, this changed to absolute displacement.

  • The Chamber of Mines deliberately diversified its labour sources to weaken workers' bargaining power. (p. 4)

    deliberate policy of drawing labour from diverse sources in order to reduce the possibility of workers combining to push for higher wages and to reduce dependence on any one source.

  • Foreign African male employment in South Africa fell sharply between 1970 and early 1977. (p. 4)

    Clarke estimates that in 1970 the number of employed foreign African male workers in South Africa was 443 000 or 15,3% of permanent ('non-casual') employment in the country.

  • Workers displaced from South Africa account for a large share of wage-labour growth in supplier states, making them doubly vulnerable. (p. 4)

    workers displaced directly from employment within South Africa, who make up some 35-37% of the growth of the wage labour force in the supplier states.

Evidence ledger

Sources used

Source Kind Period How used
Simkins and Clarke, Structural Unemployment in Southern Africa (1978), table 20: Simkins' unemployment estimates for South Africa 1960-1977 secondary literature compilation 1960-1977 Supplies the estimates of unemployment level and rate (18,3% to 22,4%) and comparative rates used in Table 1 that anchor the review.
South African population censuses of 1960 and 1970 population census 1960, 1970 Provide the definition of the economically active labour force and labour-supply data; their unemployment counts are judged unreliable, so unemployment is derived as labour supply minus employment.
Tomlinson Commission Report (1955), Reserve output and minimum income assumption commission or official report 1955 Used to set the minimum income/output level for full-time agricultural employment so that Reserve output can be converted into full-time employment equivalents.
Recorded Reserve agricultural output statistics official statistical series c. 1955-1970s Total recorded Reserve output divided by the minimum output level gives full-time employment equivalents in the Reserves.
Agricultural and population censuses (counts of casual farm workers and wages) agricultural census 1960-1970s Provide counts of casual workers and total annual wages, used with regular workers' wage rates to compute full-time equivalent casual employment.
South African industrial censuses industrial mining census 1960s-1970s Cited to show that informal-sector activity is mostly excluded from official statistics, so Simkins' employment estimates understate it.
Department of Statistics employment statistics (recognition-based) official statistical series 1960-1976 Source of employment statistics for the capitalist sector; the reviewer judges them reliable there but not in the non-capitalist sector.
United Nations, Statistical Yearbook, 1970, table 22 international trade or finance data 1961 Source of 1961 comparator unemployment rates in Table 1.
International Labour Office, Yearbook of Labour Statistics, 1976, table 22 international trade or finance data 1976 Source of 1976 comparator unemployment rates in Table 1, based on sample surveys of each country's labour force.
Clarke's estimates of foreign African workers in South Africa and labour-supplying states' employment growth secondary literature compilation 1969-1977 Supplies estimates of foreign workers (443 000 in 1970 to 290 000 in 1977) and displaced workers' share of wage-labour growth in supplier states.
J. B. Knight, 'Labour Supply in the South African Economy and its Implications for Agriculture', Saldru Farm Labour Conference, 1976 secondary literature compilation 1976 Cited as evidence that structural unemployment has been chronic since the war and probably through most of the century.
ILO, Employment, Incomes and Equality: A Strategy for Increasing Productive Employment in Kenya (1972) commission or official report 1972 Cited for the finding that informal sectors elsewhere in Africa employ many people at incomes often above rural levels.
S. Duncan, 'The Central Institutions of South African Labour Exploitation', NUSAS Labour Week, May 1977 other 1977 Cited as evidence that ministerial closure or restriction of rural recruitment districts has been implemented in some areas.

Inference

Quantification: Descriptive tables.

Book review: summarises and evaluates the constructed unemployment estimates of Simkins (residual of labour supply minus employment, with proxies for Reserve and casual farm employment) and Clarke's argument on displacement of foreign workers, supported by a simple cross-country comparison table. The reviewer's own arguments are conceptual critique, with a few reasoned generalisations.

Key claims

Claim Basis Strength
Unemployment in South Africa nearly doubled in level from 1960 to 1977 and its rate rose from 18,3% to 22,4%. (p. 1) Simkins' constructed estimates (table 20), built from censuses and Reserve output; the reviewer notes wide margins of error. Partial data
South African unemployment is far higher than in major industrial countries; none has a rate even half that of South Africa. (p. 1) Table 1 comparing UN and ILO yearbook survey-based rates for 1961 and 1976 with Simkins' estimate. Partial data
Structural unemployment has outweighed cyclical unemployment since 1960 (except since 1974) and has been chronic since the war and probably most of the century. (p. 3) Simkins' trend-employment versus labour-force estimates for 1960-77; earlier period supported only by reference to Knight. Partial data
Simkins' method may over- or under-estimate unemployment: Reserve employment is inferred from output, informal-sector employment is omitted, and casual farm employment is estimated by wage ratios. (p. 2) The reviewer's reasoning about the construction of Simkins' estimates; no independent data. Illustrative or anecdotal
Employment of foreign African workers fell from 443 000 (15,3%) in 1970 to 290 000 (8,7%) by early 1977, indicating displacement of foreign by local workers, especially on the mines. (p. 4) Clarke's estimates as reported by the reviewer. Partial data
Supplier countries' employment growth fell short of labour-force entrants in 1969-1976, with displaced workers from South Africa making up 35-37% of wage-labour growth, so structural dependence makes them doubly vulnerable. (p. 4) Clarke's provisional estimates as summarised. Partial data

The author's caveats

Acknowledges that Simkins' estimates are open to wide margins of error given the limits of South African labour statistics, that census unemployment figures are entirely unreliable, that comparisons with semi-developed countries are impossible, and that the Reserve-output, casual-labour and informal-sector assumptions are approximations.

Gaps in the evidence

The review rests entirely on secondary summaries of Simkins' and Clarke's estimates and gives no access to the underlying series, so it cannot test them independently. Direct household or labour-force surveys of Reserve and informal-sector work, plus mine recruitment records on foreign labour, would test the arguments better. Some of these (Chamber of Mines/WNLA records, later labour force surveys) existed or exist, but were not used here. The claim that unemployment was chronic since the war or earlier is supported only by one cited conference paper.

Replicability: Data poorly documented.