Black Trade Unions and the Wage Gap in South Africa¶
W. S. Siebert (1987). Black Trade Unions and the Wage Gap in South Africa. Managerial and Decision Economics, Vol. 8, No. 1. John Wiley & Sons.
| Type | journal article |
| Scope | Economics and economic history of apartheid |
| Method | econometric |
| Period studied | 1947–1984 |
| Regions | South Africa, gold mining, manufacturing |
| Themes | unions strikes, labour markets wages, mining migrant labour, bantustans removals influx, manufacturing industrial policy, inequality poverty unemployment, methods data sources |
| Citation key | siebert1987 |
Argument¶
Siebert asks why the black-to-white pay ratio in mining and manufacturing rose after 1971. He builds a simultaneous demand-and-supply model of black labour in gold mining and manufacturing and fits it to 1947-84 data. He then uses the estimated coefficients to decompose wage growth in 1975-80 and 1980-84. He concludes that rising gold prices, falling foreign labour supplies and relaxed pass-law enforcement mattered in the 1970s. Black union density is the main driver of manufacturing wage growth after 1980, but has no measured effect yet in mining. Pass-law enforcement depressed black wages in both sectors. The employment cost of union wage gains is estimated to be small in manufacturing.
Evidence and method¶
Annual 1947-84 series built from Bulletin of Statistics, S.A. Statistical Yearbooks, censuses (geometric interpolation), SAIRR Surveys and Lipton (foreign mine labour). Series cover earnings, employment, capital stock, output and prices by race; pass-law prosecutions; estimated black union membership including unregistered unions; and white-to-black per pupil education spending. The Riekert and Wiehahn Commission reports supply institutional detail. Lucas (1985) supplies the demand specification.
Place in the debates¶
- Causes of the narrowing black-white wage gap after 1971 (gold price vs unions vs labour supply vs employer sensitivity) (see Racial wage gaps, wage convergence and wage determination). Multi-causal. Gold prices and reduced foreign labour mattered in the 1970s, but unions explain most of the real manufacturing wage rise after 1980. The 'employer sensitivity' explanation is dismissed as lacking a measurable metric.
- Whether black wages are market-determined or administered (Lucas's excess-supply view of mining) (see Racial wage gaps, wage convergence and wage determination). Treats this as an empirical question and estimates supply and demand. He finds an upward-sloping supply of labour to mining and downward-sloping demand, so market forces are relevant.
- Effect of influx control and pass laws on black wages (see Racial wage gaps, wage convergence and wage determination). Stricter pass-law enforcement lowers black wages in both mining and manufacturing, so relaxation or abolition would raise them. The mechanism is that mobility restrictions cause misallocation.
- Trade union effects on wages and employment (see Black trade unions: growth, wage effects, Wiehahn and strategy). Unions raised manufacturing wages after 1980 with small employment losses. Future effects in mining could be larger, with possible job shedding and downward pressure on informal and rural wages.
Findings¶
-
The black/white pay ratio rose in both sectors after 1971, bringing mining and manufacturing black pay closer together. (p. 2)
The increase in the case of mining has been from 5% to 19%.
-
Unions were too weak in the 1970s to explain the narrowing gap in that decade. (p. 2)
Thus we cannot logically credit black unions with reducing the wage gap for most of the period under discussion.
-
Foreign workers' share of the mining labour force fell sharply after 1974, shifting mines onto domestic labour with a less elastic supply. (p. 3)
there has been a marked reduction in the proportion of the mining labour force made up by foreign blacks-from 70% in 1974 to 30% now.
-
The estimated mining labour supply is more elastic than Lucas found. (p. 9)
the elasticity of labour supply here is greater than unity (1/0.76 = 1.32), whereas Lucas found less than unity.
-
In 1975-80 the gold price was the largest identified cause of mining wage growth, with a wage effect of 0.38. (p. 10)
In the period 1975-80 the most important cause of the rise in mining wages, after allowing for the effect of rising consumer prices, is the rise in the gold price.
-
Manufacturing log wages rose 62% in 1980-84, and most of the real increase is attributed to rising union density. (p. 10)
In the period 1980-84 in manufacturing, log wages increased 0.62, or about 62%.
-
The union coefficient is small, so further union growth would add little to manufacturing wages. (p. 10)
Thus a movement to 36% unionization would only cause a further 2% real wage increase, ceteris paribus.
-
Unions show no measured effect on mining wages because they are new there, but the author expects future effects. (p. 11)
It will be noted that trade unionism is not recorded as having had any impact on mining wages, but this does not mean that future effects will be negligible.
-
Gold prices alone cannot explain the wage gains, since mine wages kept rising in the 1980s while gold prices were nearly static. (p. 11)
Though increased gold prices seem to have been important in the 1970s, mine wages continued to increase in the 1980s, even though gold prices were nearly static.
-
Abolishing influx control is predicted to raise black real wages in both sectors. (p. 11)
The abolition of controls might raise black real wages in manufacturing by 7% and in mining by a larger amount.
-
The union wage effect costs few manufacturing jobs, but unionisation of mining could shed far more. (p. 11)
These are rather small employment effects-20000 jobs out of a black manufacturing workforce of 700 000.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Bulletin of Statistics (Department of Statistics) – annual earnings and employment series, gold mining and manufacturing, by colour group | official statistical series | 1945-1984 | Source of average annual black and white earnings (wage bill divided by average monthly employment) used for wage-gap ratios and as the dependent variable in the supply equations. |
| S. A. Statistical Yearbooks and Bulletin of Statistics (employment, capital, output and price series) | official statistical series | 1947-1984 | Sector employment, capital stock, output and value-of-output series used to construct the labour demand variables and product-price deflators. |
| Population censuses 1946, 1951, 1960, 1970, 1980 (economically active population) | population census | 1946-1980 | Construct the non-farm workforce denominator for union density, with geometric interpolation between censuses; also occupational distribution of the black workforce. |
| Gross capital formation and fixed asset benchmark figures (mining and manufacturing) | official statistical series | benchmark years to 1984 | Construct real capital stock series by perpetual inventory using benchmark values and assumed depreciation rates, deflated by the wholesale price index. |
| Lipton, Men of two worlds (Optima, 1982) and SAIRR Surveys – foreign blacks in mining | secondary literature compilation | 1945-1984 | Provides the series of foreign black mine workers entering the supply equations and the claim that foreign labour fell from 70% to 30%. |
| SAIRR Annual Survey (various years, incl. 1979 p. 196; 1983) | secondary literature compilation | 1950s-1984 | Supplies Wage Board determinations dates, pass law prosecutions, union membership and unregistered union estimates. |
| Pass law prosecution series (SAIRR 1963; Bulletin of Statistics 1947-50; SAIRR Surveys 1975-84) | official statistical series | 1947-1984 | Spliced series of prosecutions used as the index of influx control strictness (PASSLAW) in the supply equations. |
| Black trade union membership series (Clack PhD thesis; SAIRR Surveys; 1984 Department of Manpower Report; Wiehahn Commission 1978 estimate) | secondary literature compilation | 1945-1984 | Constructs black union density (including unregistered unions) by linear interpolation between sparse observations, the key variable for the union-effect estimate. |
| SAIRR Yearbooks – per pupil education expenditure series | secondary literature compilation | 1945-1983 | Constructs white-to-black per pupil education expenditure ratio as an indicator of colour bar; ultimately dropped for unstable sign. |
| Riekert Commission (1979), Report of the Commission of Inquiry into Legislation Affecting the Utilization of Manpower, RP 32 | commission or official report | 1979 | Source for rural/urban labour numbers, farm employment, pass law definitions, evasion, mine contract lengths and the gold price explanation of wage rises. |
| Wiehahn Commission (1979), Report of the Commission of Inquiry into Labour Legislation, Part I, RP 47 | commission or official report | 1979 | Establishes recognition of black unions and the number of workers in unregistered unions, and rationale for registration. |
| Blacks (Urban Areas) Consolidation Act of 1945, Section 10; Industrial Conciliation Act | legislation gazette | 1945-1979 | Defines Section 10 rights shaping labour market segmentation, and the legal basis for black union recognition. |
| Siebert (1975), The Relative Pay of White and Non-White Workers in South Africa, unpublished PhD thesis, LSE | secondary literature compilation | to 1975 | Background on mobility restrictions and Wage Board pressure on unskilled pay in 1961/2. |
| Lucas (1985) 'Mines and migrants in South Africa' – demand specification and supply elasticity results | secondary literature compilation | 1985 | Demand specifications adopted and elasticity estimates compared against Siebert's results. |
New data¶
Annual 1947-84 time series for gold mining and manufacturing (black/white earnings, employment, capital stock, product prices) plus constructed series for pass law prosecutions, black union density including unregistered unions, and education expenditure ratios, assembled from published statistics and interpolated.
Inference¶
Quantification: Econometric (causal design).
Time-series estimation (1947-84) of simultaneous labour demand and supply equations for gold mining and manufacturing using instrumental variables and lag specifications, followed by a decomposition of wage changes in 1975-80 and 1980-84 into contributions from exogenous variables via the solved structural system.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Black trade unions were the main factor behind the real black manufacturing wage rise in 1980-84. (p. 63) | Manufacturing supply equation (post-1974 union density coefficient 0.083) multiplied by a 2.2 increase in log union density, in the decomposition of Table 3. | Partial data |
| Unions had no measurable effect on mining wages in the period examined, though future effects are expected. (p. 64) | Insignificant union coefficient in the mining supply equation (Table 2); extrapolation using the manufacturing coefficient. | Partial data |
| Rising gold prices were the most important cause of mining wage rises in 1975-80 but not the full explanation, since mine wages kept rising in the 1980s when gold prices were static. (p. 64) | Gold price coefficient in the mining equations and the decomposition in Table 4. | Partial data |
| Reduced foreign labour supply after 1974 raised mining wages. (p. 62) | Foreign labour coefficient (-0.199, t=1.94) in the mining supply equation using Lipton/SAIRR foreign labour series. | Partial data |
| Relaxation of influx control (fewer pass law prosecutions) raised black wages in both sectors, and abolition would raise them further. (p. 62) | Negative pass law prosecution coefficients in both supply equations, with a linear extrapolation of abolition effects; prosecutions are a crude proxy. | Partial data |
| Union-induced wage increases cost few manufacturing jobs (about 3.4% of employment, around 20,000 jobs) but could displace more in mining and informal sectors. (p. 64) | Estimated manufacturing labour demand elasticity (-0.22) applied to the union-induced wage rise; mining effect projected by assumption. | Partial data |
| The black-to-white pay ratio rose markedly after 1971 in both mining (5% to 19%) and manufacturing (16% to 26%). (p. 55) | Bulletin of Statistics earnings series charted in Figs 1 and 2. | Systematic data |
The author's caveats¶
Author states results should be read cautiously given data and specification problems; union series quality is too poor for yearly rates of change (though secular trends are reliable); the model fits mining less well than manufacturing; wage data for agriculture and informal sectors are sparse; the education variable was unstable and dropped; unionization and white employment are treated as exogenous for simplicity.
Gaps in the evidence¶
Union density is built from sparse, interpolated membership figures, and pass law prosecutions are a crude proxy for enforcement, so the key causal attributions rest on weak measures with only ~38 annual observations and many trending series. Firm- or mine-level wage and union data, or strike and bargaining records, would test the union effect better; some of these exist in company, Chamber of Mines and union archives. Wage data for agriculture and informal sectors, and direct measures of Wage Board determinations and employer pressure, were lacking then and would be needed for the spillover and employment claims.
Replicability: Data described but not public.