Skip to content

The Narrowing of Wage Differentials in South Africa

Literature › Works

Jill Nattrass (1977). The Narrowing of Wage Differentials in South Africa. South African Journal of Economics, vol. 45, no. 4. Economic Society of South Africa.
Type journal article
Scope Economics and economic history of apartheid
Method mixed
Period studied 1960–1975
Regions South Africa, Homelands (KwaZulu), Durban, Johannesburg
Themes labour markets wages, inequality poverty unemployment, mining migrant labour, manufacturing industrial policy, unions strikes, bantustans removals influx
Citation key nattrass1977

Argument

Nattrass argues that between 1970 and 1975 Black average real wages in the modern sector rose much faster than White wages, narrowing the absolute wage gap for the first time and shifting the racial distribution of personal income from Whites to Blacks (White share 74% to 68%). She qualifies this: inequality within the Black population (skill groups, urban versus rural and migrant households) probably widened. A survey of manufacturing firms suggests Black employment is not strongly wage-elastic in the short run, and that market size and skilled labour, rather than Black wages, constrain growth. She concludes that redistribution need not conflict with growth, but needs sustained growth, skills upgrading, an informal sector and investment-goods strategy, and political reform.

Evidence and method

Published Department of Statistics and Bureau of Statistics wage and employment data (Statistical Year Book 1964, South African Statistics 1976); Bureau of Market Research household income studies for Durban and Johannesburg; homeland national accounts; her own earlier estimates of migrant numbers; comparison with Stadler, Spandau and McGrath income-share estimates; a 1973 University of Natal survey of 454 manufacturing firms (282 responses); Lenta's small township informal-sector survey.

Place in the debates

Findings

  • Black real wages accelerated after 1970 while White wage growth collapsed, and the absolute gap fell for the first time. (p. 1)

    the rate of growth of Black average real wages accelerated sharply to 6,6 per cent per annum, whilst the corresponding rate of growth of White wages fell to 1 per cent.

  • Despite Black occupational advance, Whites still dominated skilled and managerial jobs in 1975. (p. 2)

    in 1975 the White workforce still provided 60 per cent of all the professional and technical manpower; 95 per cent of the managerial and administrative manpower

  • The estimates imply a historically unprecedented redistribution of personal income toward Blacks. (p. 3)

    A redistribution which succeeds in reducing the White share from 74 per cent in 1970 to 68 per cent in 1975 is significant in an historical context

  • Black income grew substantially faster than White income in 1970-75, altering racial shares, in contrast to earlier estimates of constancy. (p. 2)

    there was a substantial redistribution of the increase in income to Blacks sufficiently large to alter the racial shares in total personal income.

  • Rising Black wages in manufacturing coincided with substitution of Black for White labour, so labour costs per man grew slowly. (p. 5)

    the ratio of the number of Blacks employed in the manufacturing sector to that of Whites increased from 2,75:1 in 1960 to 3,6:1 in 1976.

  • Black employment grew despite large real wage rises, suggesting low short-run wage elasticity. (p. 5)

    average real Black wage rates in the modern sector grew by 47 per cent. Black employment increased by 22 per cent over this period

  • Businessmen reported that wage increases did not reduce employment. (p. 5)

    84 per cent of these firms (which represents 74 per cent of the overall responses) said that employment levels had not altered as a result of the rise in wages.

  • Inadequate demand was the main perceived past constraint on manufacturing investment. (p. 6)

    Sixty-three per cent of the respondents (178 firms) listed inadequate demand levels as important and 32 per cent (91 firms) ranked them as being of prime importance.

  • South African manufacturing relies heavily on imported technology. (p. 7)

    Sixty per cent of the firms responding to the question (169 firms) stated that they used foreign technology exclusively

  • Remittances were a large fraction of homeland output, showing homeland dependence on migrant wages. (p. 3)

    the low estimate of remittances for 1975 was the equivalent of 62 per cent of the total Gross Domestic Product of the homeland areas in 1973/74

  • Black share of the growth in consumption rose much faster than their share of the stock. (p. 8)

    the Black share in total consumption grew from 26 per cent in 1970 to 32 per cent in 1975, the Black share of the increase in total consumption over the period grew from 26 per cent to 41 per cent.

  • Despite redistribution, per capita racial inequality remained very large. (p. 10)

    the ratio of average White and Black per capita income was still of the order of 11:1 in 1975

Evidence ledger

Sources used

Source Kind Period How used
Statistical Year Book 1964 (Bureau of Statistics) and South African Statistics 1976 (Department of Statistics), wage and employment series official statistical series 1960-1975 Source of modern sector average real wage rates by race, wage bills and employment, from which the Black-White wage gap and sectoral wage income decompositions are estimated.
Manpower Survey No. 8 (1969) and Manpower Survey No. 11 (1975), Department of Labour official statistical series 1969, 1975 Used to compute growth rates of Black and White employment in professional, managerial, clerical and sales categories.
National Accounts of the Bantu Homelands, 1969/70 to 1973/74, Report 09-17-01 national accounts macro 1969/70-1973/74 Provides subsistence production growth (11,6% p.a.), homeland GDP and subsistence income used for Black income estimates and remittance comparisons.
Bureau of Market Research, Income and Expenditure Patterns of Urban Black Households in Johannesburg, Research Report No. 50.3 (1976) author survey or fieldwork 1970-1975 Shows similar growth in money incomes across African household heads' skill categories and falling remittance/support payments by single households.
Bureau of Market Research, Income and Expenditure Patterns of Urban Black Households in Durban, Report No. 50.10 (1976) author survey or fieldwork 1972, 1975 (also 1970 expenditure) Compares single vs multiple households' incomes and remittance rates, and food expenditure shares/elasticities.
Bureau of Market Research Report No. 34, Market Potentials of Consumer Goods and Services for Non-White Population Groups in the Five Main Metropolitan Areas, 1970 author survey or fieldwork 1970 Establishes that 37% of African expenditure (55% of foodstuffs) occurred at African business outlets, supporting informal sector demand argument.
Population Census 1960 (Vol. 6) and 1970 Population Census Reports 02-02-02, 02-06-01, 02-05-11 (Occupations) population census 1960, 1970 Domestic service employment and African population/activity data used to estimate domestic wage income and migrant numbers.
Report 11-03-11 Statistics of Houses and Domestic Servants, October 1975 official statistical series 1970-1975 Index of cash servants' wages used to escalate domestic service wages for 1973 and 1975.
J.J. Stadler, Die Nasionale Rekeninge van Suid Afrika (1973) and Stadler unpublished D.Com. thesis (1962) on Gross Domestic Product 1911-1959 secondary literature compilation 1911-1960 Benchmark for White share of income in 1960 (76,6%) and basis for the 95% White non-subsistence income share justifying the assumption.
A. Spandau, Income Distribution and Economic Growth in South Africa (unpublished D.Com. thesis, UNISA, 1971) secondary literature compilation 1925-1960 Comparison estimate of White income share (73% in 1960) and finding of constancy of racial shares over 1925-1960.
M.D. McGrath, Racial Income Distribution in South Africa, Black-White Income Gap Research Report No. 2 (1977) secondary literature compilation 1917-1970 Comparison estimate (White share 72% in 1970) and review of earlier estimates showing no previous redistribution.
Nattrass, 'Migrant Labour and South African Economic Development', SAJE 44(1), 1976 secondary literature compilation 1960-1970 Source of migrant worker numbers for 1960 and 1970 and the estimate that over half of economically active African men in White areas were migrants.
Nattrass, Migrant Labour and Underdevelopment: The Case of KwaZulu, Black-White Income Gap Research Report No. 3 (1977) secondary literature compilation 1960-1970 Source for KwaZulu remittances as a share of subsistence income and the claim that migrancy lowers rural productivity.
Report of the Commission for the Socio-Economic Development of the Bantu Areas (Tomlinson Commission), U.G. 61/1955; Keiskammahoek Rural Survey Vol. II commission or official report 1950s Cited to show the heavy dependence of rural areas on remittances (46,5% and 50% of incomes).
Studies of migrant remittance rates (Leistner 1964; BMR Report 34; van der Merwe, Lesotho National Accounts; Department of Statistics 19% assumption) secondary literature compilation 1954-1974 Justifies assuming remittances equal 20% of migrant earnings in Table 2.
Nattrass/Duncan/Morris 1973 survey of manufacturing firms (454 firms contacted, 282 responding) author survey or fieldwork June-July 1973 Provides businessmen's views on Black wage increases and employment, growth constraints, capital intensity and technology dependence.
Nattrass and Duncan, A Study of Employers' Attitudes Towards African Worker Representation, Income Gap Research Report No. 1 (1975) author survey or fieldwork 1975 Documents the sample and sampling procedure of the manufacturing survey.
B.L. de Jager, Fixed Capital Stock and Capital-Output Ratio of South Africa 1946-1972 (SARB Quarterly Bulletin, June 1973) with South African Statistics 1976 Table 7.4 national accounts macro 1950-1972 Computes the manufacturing capital-labour ratio's growth (2,5% p.a.) to test the claim of rising capital intensity.
SARB, A Statistical Presentation of South Africa's National Accounts 1946-1970 (Supplement to Quarterly Bulletin, June 1971) national accounts macro 1946-1970 Source of personal consumption expenditure for Table 4 and of net national income comparison in Table A.
G. Lenta, 'The Economic Structure of KwaZulu' (unpublished PhD thesis, Natal, 1977), township informal sector survey author survey or fieldwork 1975 Small impressionistic survey illustrating the composition of the informal sector (food and liquor trading).
N.C. Morris, 'Determinants of Investment in the South African Manufacturing Sector' (M.Com thesis, 1977) and R.P.C. Brown, 'Capital Intensity in South African Manufacturing' (1976) secondary literature compilation 1973-1977 Analyses of the firm survey results on investment constraints and capital intensity.

New data

Author-constructed estimates of the racial distribution of personal income for 1960, 1970, 1973 and 1975 (Table A), migrant earnings/remittance estimates (Table 2), sectoral wage-income decompositions (Table 3), and a Natal 1973 survey of 282 responding manufacturing firms.

Inference

Quantification: Descriptive tables.

Descriptive reconstruction of racial income shares from published wage and employment statistics under a strict assumption that all non-wage income accrues to Whites, supplemented by shift-share style decompositions, comparison with earlier estimates, scenario estimates for migrant flows, and the interpretation of a firm opinion survey; policy conclusions are drawn by theoretical argument.

Key claims

Claim Basis Strength
Black real wage growth accelerated to 6,6% p.a. in 1970-75 while White growth fell to 1%, so the absolute gap fell for the first time (1973-75). (p. 1) Wage and employment series in the Statistical Year Book 1964 and South African Statistics 1976. Systematic data
The Black share of personal income rose from 26% in 1970 to 32% in 1975, the first redistribution from Whites to Blacks recorded since 1917. (p. 3) Author's Table A estimates from official wage, employment, domestic service and homeland national accounts data, compared with Stadler, Spandau and McGrath. Partial data
The distribution of Black income became more unequal, with a widening urban-rural gap, as migrant remittance rates fell. (p. 2) Bureau of Market Research Johannesburg and Durban household studies, with Nattrass's migrant estimates. Illustrative or anecdotal
Rising Black wages did not reduce employment; 74% of sampled firms reported no employment change after wage rises, and Black-to-White employment ratios in manufacturing rose, suggesting substitution of Black for White labour. (p. 5) Wage-bill and employment series for manufacturing, plus the 1973 manufacturing firm survey. Partial data
Manufacturing growth was constrained mainly by market size, and firms rely heavily on imported technology. (p. 6) 1973 firm survey responses (63% listing demand, 60% using foreign technology exclusively) and the capital-labour ratio series (de Jager). Partial data
Redistribution raised the Black share of the increase in consumption from 26% to 41% by 1975, probably shifting demand toward informal sector goods. (p. 8) Table 4 applying income shares to SARB personal consumption expenditure; BMR 1970 metropolitan expenditure data; Lenta's small survey. Illustrative or anecdotal

The author's caveats

Acknowledges the strict assumption that all non-wage income other than homeland subsistence income accrues to Whites, no allowance for payments in kind or government transfers, the likelihood that the estimates understate redistribution, general inaccuracy of such estimates, a survey sample biased toward large firms and conducted in an optimistic period, the "suggestive, not conclusive" remittance evidence, and that Lenta's survey cannot be generalised and that survey results are not yet fully published.

Gaps in the evidence

The argument lacks direct data on non-wage (property) income by race, on actual unemployment by race, and on wage elasticity of employment beyond stated business opinions. A panel of firm-level wages and employment or labour force survey microdata would better test the substitution and employment claims; such data were scarce in the 1970s, though later surveys exist. Income distribution within the Black population rests on a few BMR city studies and cannot support national conclusions.

Replicability: Data poorly documented.