The Rise in African Wages: 1975-1985¶
J. F. Hofmeyr (1994). The Rise in African Wages: 1975-1985. South African Journal of Economics, 62(3).
| Type | journal article |
| Scope | Economics and economic history of apartheid |
| Method | econometric |
| Period studied | 1975–1985 |
| Regions | South Africa, urban areas (Johannesburg, Cape Peninsula and other main urban areas) |
| Themes | labour markets wages, origins segregation, bantustans removals influx, education human capital, inequality poverty unemployment, methods data sources, mining migrant labour |
| Citation key | hofmeyr1994a |
Argument¶
Average real African wages in the modern sector rose from the 1970s even though labour demand was slack and unions were weak until about 1984. Hofmeyr argues that the explanation lies in institutional change: the dismantling of apartheid labour-market restrictions and the expansion of African education unified a labour market that had been segmented by sector, region and gender. Using BMR household survey data for 1975 and 1985, he fits human-capital wage regressions and decomposes wage growth into changes in the structure of earnings and changes in workers' characteristics. The rise in average wages came almost wholly from Africans moving up the occupational and educational ladder. Real wages of unskilled workers actually fell, so occupational differentials widened.
Evidence and method¶
Previously unexploited household survey data from the Bureau of Market Research (UNISA) for urban Africans in 1975 and 1985 (the small 1980 survey is omitted). Employed workers aged 16-65 are used, with variables for occupation, gender, sector, area, education, accommodation, marital and household status, age and job duration. These are compared with South African Labour Statistics national averages and with Knight and McGrath (1987).
Place in the debates¶
- Causes of rising African wages in the 1970s-80s (labour scarcity vs unions vs institutional reform) (see Racial wage gaps, wage convergence and wage determination). Rejects general labour scarcity and trade-union pressure, since employment growth lagged population growth and mass unionisation came only after 1984. Attributes the rise to labour-market unification after reform and to upward occupational mobility.
- Labour market segmentation under apartheid and its erosion (see Racial wage gaps, wage convergence and wage determination). Finds significant sectoral, regional and gender segmentation in 1975 that weakened or disappeared by 1985. Sees this as consistent with the erosion of influx control and other labour laws.
- Real wage trends for unskilled workers vs average wages (compared with Knight and McGrath) (see Racial wage gaps, wage convergence and wage determination). Finds unskilled real wages fell sharply and occupational wage rates for males mostly fell, whereas Knight and McGrath found rising real occupational wages. Hofmeyr attributes the difference mainly to sampling.
Findings¶
-
Real wages of non-primary sectors rose over 1975-90 while White wages fell, a puzzle given weak labour demand. (p. 1)
Between 1975 and 1990 real wages rose by 2,9 per cent per annum on average for the non-primary sectors, while those for Whites actually dropped
-
Unionisation cannot explain the wage rise because mass African unionisation came late. (p. 1)
Since mass unionisation for Africans really took off only in 1984, union wage pressures do not provide a plausible explanation either.
-
Education expansion in the 1970s was rapid, raising African mobility up the occupational ladder. (p. 1)
Expenditure on this grew rapidly from 1972, and enrolment grew at more than 20 per cent per annum
-
The gender wage gap narrowed sharply between 1975 and 1985. (p. 5)
women were paid significantly less than men (some 48 per cent) in 1975 after accounting for differences in the distribution of characteristics
-
Unskilled workers in the reference categories suffered a large real wage fall. (p. 5)
The rate of increase in wages for unskilled workers whose characteristics fall into the reference categories was approximately 8,6 per cent per annum over the decade, representing a fall of no less than 3,3 per cent per annum in real terms!
-
Regional wage differences, linked to influx control and proximity to homelands, narrowed by 1985. (p. 5)
Turning to the results for 1985, there is no longer any significant difference between Johannesburg and its hinterland.
-
Nearly all the growth in average wages came from changes in characteristics, not in the structure of earnings. (p. 6)
Overall, the change in the earnings functions contributed practically nothing to the increase in average wages, while changes in characteristics were responsible for almost all the growth observed.
-
Occupational and educational upgrading was the main driver of average wage growth. (p. 6)
the major factors were upward movement through the occupational and educational structure, which contributed 1,4 per cent per annum to overall
-
Mining wages for Africans caught up with manufacturing between 1975 and 1984. (p. 9)
average wages for Africans in mining rose from 70 to 89 per cent of those in manufacturing between 1975 and 1984
-
Segmentation was confirmed in 1975 and eroded by 1985, while unskilled male wages fell and occupational differentials widened. (p. 8)
the results of the analysis do indeed confirm that in. 1975 there was significant segmentation of the labour market for urban Africans along sectoral, regional and gender lines
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Bureau of Market Research (BMR, UNISA) household surveys of urban Africans, 1975 and 1985 (1980 survey omitted) | household or labour survey microdata | 1975, 1980, 1985 (1975 and 1985 analysed) | Raw microdata are the basis for the wage regressions and the decomposition of wage growth by structure and characteristics. |
| South African Labour Statistics 1990 (SALS 1990), Table 4.1.1 | official statistical series | 1975-1990 | Establishes the rise in real African wages (2.9% p.a.), the fall in White wages, and mining wages rising from 70 to 89 per cent of manufacturing. |
| South African Labour Statistics 1987, Table 4.1.1.1 (Central Statistical Services employer sample surveys) | official statistical series | 1975 and 1984 | National average wage benchmark against which the BMR sample averages are compared. |
| Current Population Survey based on the 1985 census, Statistical News Release P27.3, 6 March 1987, Table 4.3 | official statistical series | August 1986 | Used to argue that hours-of-work differences are unlikely to explain much of the male-female wage gap. |
| Knight and McGrath (1987) private-sector wage survey data (nineteen job-grading levels), 1976-85 | secondary literature compilation | 1976-85 | Results from this survey-based study are used as a comparison with the author's decomposition results. |
| Knight and McGrath (1987) on education expenditure and enrolment | secondary literature compilation | 1970s | Supports the claim of rapid expansion of African education from 1972. |
| Van Niekerk (1983) PhD dissertation on education statistics for Black education | secondary literature compilation | c. 1983 | Supplies average school-leaving ages by standard, used to convert age into working life. |
| Hofmeyr (1990), Occasional Paper No. 22, Economic Research Unit, University of Natal | author survey or fieldwork | 1975-1985 | Holds the full regression results, the 1980 survey, and the methodology on which the article's summary relies. |
Inference¶
Quantification: Econometric (causal design).
Semilogarithmic human-capital (Mincer-type) regressions on pooled BMR urban African household survey data for 1975 and 1985, with gender interactions and dummies for area, sector, occupation, and accommodation. An Oaxaca-type decomposition splits the growth in mean wages into earnings-structure and characteristics effects, and the results are read as consistent with a labour-market unification hypothesis. The results are compared with Knight and McGrath (1987), and the inference is by hypothesis consistency rather than a identified causal test.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Real average African wages rose in the modern sector despite slack labour demand and weak unionisation until 1984, so the rise needs an institutional explanation. (p. 198) | SALS 1990 Table 4.1.1 national wage series, plus employment and unionisation context that is asserted. | Partial data |
| Real wages of unskilled reference-category workers fell by about 3.3 per cent a year over 1975-85, and occupational differentials for males widened. (p. 208) | BMR 1975 and 1985 regressions and decomposition (constant term -3.29), deflated by the low-income inflation rate. | Systematic data |
| Almost all growth in average real cash wages came from changes in characteristics (upward occupational and educational movement), with the earnings structure contributing practically nothing. (p. 210) | Decomposition of BMR wage change, Table 3 (structure 0.04, characteristics 1.28 per cent p.a.). | Systematic data |
| The female wage penalty fell from about 48 per cent in 1975 to 29 per cent in 1985, consistent with decreased gender segmentation. (p. 207) | BMR regressions with gender interaction terms; hours worked unavailable, so only partly checked with the Current Population Survey. | Partial data |
| Regional and sectoral wage differentials, significant in 1975, had largely disappeared by 1985, indicating a unifying labour market as influx control and related restrictions eroded. (p. 208) | Area and sector dummy coefficients from the BMR regressions, with SALS 1990 mining wage ratios as a supplementary check. | Partial data |
| Migrant and living-in domestic workers were paid significantly less in 1975, and this gap largely vanished by 1985 as influx control fell into disuse. (p. 208) | Accommodation-type dummies in the BMR regressions, with and without payments in kind. | Partial data |
| The author's finding that occupational wages fell in real terms contrasts with Knight and McGrath's rise, which the author attributes to differences in sampling. (p. 212) | Comparison of the BMR-based results with the Knight and McGrath (1987) private-sector survey results. | Illustrative or anecdotal |
The author's caveats¶
The author notes that published South African wage data give only sectoral averages. The BMR samples are not random for the country as a whole and the areas covered vary by year. The 1980 survey is too small and was omitted. There are no hours-worked data, and wages are weekly or daily rates. Sectors are coded at one digit only, and there were only 44 miners in 1975. BMR and national averages are not strictly comparable, and the data exclude Whites and the self-employed.
Gaps in the evidence¶
The argument ties wage trends to labour-market reform but offers no direct measures of enforcement of influx control or of the timing of reforms, so causation is inferred from consistency. Panel or repeated random national samples with hours worked, firm-level data and union wage settlements would test the unification and skill-upgrading claims better. The BMR microdata are an unpublished and unrepresentative sample, which restricts external validity, though more representative sources such as the October Household Surveys only came later.
Replicability: Data described but not public.