Black South African Unions: Relative Wage Effects in International Perspective¶
P. G. Moll (1993). Black South African Unions: Relative Wage Effects in International Perspective. Industrial and Labor Relations Review, Vol. 46, No. 2. Cornell University, School of Industrial & Labor Relations.
| Type | journal article |
| Scope | Economics and economic history of apartheid |
| Method | econometric |
| Period studied | 1924–1986 |
| Regions | South Africa, Johannesburg, Pretoria, Durban, Cape Town, Port Elizabeth, Vaal Triangle |
| Themes | unions strikes, labour markets wages, inequality poverty unemployment, comparative theory, education human capital, methods data sources, sanctions divestment |
| Citation key | moll1993b |
Argument¶
Using 1985 household microdata, Moll estimates the first union-nonunion wage differentials for South Africa and finds that black (emergent) unions had won wage gains comparable to unions in developed democracies, despite disenfranchisement, late legal recognition and repression. The overall union effect for black blue-collar workers is about 24%, in the high American range and above the European range, which he attributes to decentralised plant-level organising. Unions also compressed wages across skill levels: gains accrued mainly to unskilled workers while skilled and semi-skilled men showed no significant effect. A white-worker comparison is used to argue that this is not an artefact of industrial council 'ergo omnes' extension. The paper suggests union combativeness and multinational/foreign-ownership pressures as possible origins of the effects.
Evidence and method¶
Bureau of Market Research 1985 household surveys of urban blacks (N=4,963 blue-collar workers in the probit); Central Statistical Service 1985 household survey for whites (72 union, 235-237 nonunion observations, household heads in Johannesburg, Pretoria and Durban only); industrial council coverage data (TUCSA 1985, Williams 1990); strike data from the Department of Manpower; Friedman (1987) for union history; comparison with published international union wage studies.
Place in the debates¶
- Did apartheid repression keep black union wage effects small? (union power under racial oligarchy) (see Black trade unions: growth, wage effects, Wiehahn and strategy). Moll argues against the expectation of small effects: even in 1985, black union wage effects were similar to those in developed democracies.
- Centralised vs decentralised wage-setting and size of union premia (Blanchflower and Freeman; Calmfors and Driffil) (see Black trade unions: growth, wage effects, Wiehahn and strategy). Finds that large South African effects are consistent with the typology because emergent unions bargained at decentralised plant level, though he notes the typology is imperfect.
- Foreign ownership, disinvestment pressure and wages (see Black trade unions: growth, wage effects, Wiehahn and strategy). Suggests, without being able to test it, that multinational firms under Western pressure raised wages and recognised unions early, possibly with ripple effects.
Findings¶
-
Black union wage differential of 24% overall (19% men, 31% women) after correcting for self-selection. (p. 12)
According to this measure, the black union-nonunion differential was 19% for men, 31% for women, and 24% overall.
-
South African black union effects fall in the high 'American' range rather than the lower 'European' range. (p. 13)
The results for black South African unions thus fall in the higher "American" range rather than in the lower "European" range.
-
Union membership grew sharply after legalisation, from 760,000 in 1979 to 1.8 million in 1985 (claimed membership). (p. 3)
Total trade union membership soared, from some 760,000 in 1979 to 1.8 million in 1985.
-
Unskilled African workers gained most from unionisation; skilled gained little. (p. 13)
Taking all sectors together, the average union differential for unskilled Africans was 19% for men and 35% for women.
-
Black unions compressed wages by skill, owing to plant-level organising around a 'living wage'. (p. 13)
The unions appear to have compressed wages by skill level. This outcome is probably the result of the plant-level organization of the emergent unions
-
Strike activity by black workers rose steadily after legalisation in 1980, reaching levels comparable to Britain in the 1950s-60s. (p. 6)
following legalization in 1980, the number of days lost increased steadily, reaching 187 per 1,000 by 1985, which is comparable to the rate of days lost in Britain
-
The white skilled/semi-skilled union effect was substantial (about 13%), whereas for comparable black men it was negligible, so industrial council extension does not explain the black result. (p. 16)
The differential for white skilled and semi-skilled workers was about 13%.
-
Returns to schooling and experience for blacks were about a quarter of those for whites, attributed to poor schooling quality and discrimination. (p. 15)
Black schooling returns were low due to low-quality schooling, discrimination in education spending, and labor market discrimination
-
The 1973 Durban strikes and employer pressure on government preceded the Wiehahn Commission and the 1979 reform allowing Africans to join registered unions. (p. 3)
A wave of spontaneous strikes erupted in Durban in 1973.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Bureau of Market Research (BMR, UNISA) 1985 household survey of urban black households (data set; BMR Research Report 130.9 on Johannesburg/Soweto) | household or labour survey microdata | 1985 | Provides the microdata (N=4,963 black full-time blue-collar workers) for the union membership probits, wage equations, and union-nonunion differentials for black workers. |
| Central Statistical Service 1985 household survey (white households) | household or labour survey microdata | 1985 | Used to estimate the union-nonunion wage differential for white male skilled and semi-skilled workers (72 union and 235 nonunion observations) as a comparison for the black results. |
| BMR (1986) Income and Expenditure Patterns of Urban Black Multiple Households in Johannesburg, 1985, Research Report 130.9 | official statistical series | 1985 | Cited for the survey response rate in the Soweto survey (only 14 of 440 households refused). |
| Trade Union Council of South Africa (1985) Official Trade Union Directory and Industrial Relations Handbook, pp. 267-81 | union or movement records | 1984/85 | Source of data on the number of industrial councils, their extension to non-parties, and workers covered (Table 1). |
| Williams (1990) Industrial Council Wage Rates in South Africa (1978-1988), SALDRU | secondary literature compilation | 1978-1988 | Supplies industrial council counts and coverage data for Table 1. |
| Central Statistical Service (1989) economically active population, p. 20 | official statistical series | 1985 | Denominator for industrial council coverage rates in Table 1. |
| Central Statistical Service (1990), South African Labour Statistics, Table 2.2.1.1 (non-agricultural employment by race) | official statistical series | 1970-1986 | Employment denominators for person-days lost in strikes per 1000 employed in Table 2. |
| Department of Manpower Annual Reports (various years) | commission or official report | 1970-1986 | Source of strike days lost by race used to document the rise in black strike activity after 1980. |
| US Department of Labor (1990) Handbook of Labor Statistics 1989; ILO Yearbook of Labour Statistics | international trade or finance data | 1950s-1980s | Comparative strike rates for the US, UK and Japan in Table 2. |
| Industrial Conciliation Act 1924 and Industrial Conciliation Act of October 1979 (and amendments) | legislation gazette | 1924-1984 | Describes the legal framework: industrial councils, exclusion of Africans from the definition of employee, and legalization of African unions in 1979. |
| Wiehahn Commission (appointed 1977) | commission or official report | 1977-1979 | Mentioned as the government's response to unrest that led to the legalization of African unions. |
| Union membership figures (claimed by unions), 1979 and 1985 | union or movement records | 1979-1985 | Documents growth of union membership from 760,000 to 1.8 million and split between registered and unregistered unions. |
| Friedman (1987) Building Tomorrow Today: African Workers in Trade Unions, 1970-1984 | secondary literature compilation | 1970-1984 | Historical evidence on union recognition, FOSATU bargaining rights, strikes and state violence used to frame hypotheses. |
| Published union wage differential studies (Blanchflower and Freeman 1990; Lee 1978; Duncan and Leigh 1980; Lewis 1986; Shah 1984 etc.) | secondary literature compilation | 1967-1987 | Compiled in Table 6 as the international benchmark for the South African union wage effect. |
Inference¶
Quantification: Econometric (causal design).
Cross-sectional Mincerian log-wage regressions on 1985 household microdata, estimating union-nonunion differentials by a single dummy, separate union/nonunion equations (Oaxaca-style) and Lee's selectivity correction with jackknife standard errors. Results are compared with published international estimates and with a separate white-worker sample. Skill compression is inferred from skill- and sector-specific union dummies; the institutional background is narrative.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| The union wage effect for black blue-collar workers in 1985 was about 24% overall (19% men, 31% women) after correcting for self-selection, in the high US range and above European estimates. (p. 255) | BMR 1985 microdata, Tables 4-5 (union/nonunion regressions with selection correction), compared with the literature compiled in Table 6. | Partial data |
| Black unions compressed wages across skill levels: large gains for unskilled workers, and none for skilled workers. (p. 256) | Skill- and sector-specific union dummy regressions on African workers in BMR data (Table 7), plus smaller skill coefficients in the union sector (Table 4). | Partial data |
| White skilled and semi-skilled union workers had a substantial ~13% union differential, so the lack of a black skilled effect is probably not due to ergo omnes extension of industrial council rulings. (p. 259) | CSS 1985 white household sample (72 union, 235 nonunion) regressions in Tables 8-9, compared with a matched BMR black subsample. | Partial data |
| Black union strike activity rose sharply after legalization in 1980, reaching levels comparable to Britain in the 1950s-60s. (p. 249) | Department of Manpower strike days data divided by CSS employment (Table 2), compared with US, UK, Japan from ILO and US DOL. | Systematic data |
| The origin of the union effects (militancy, foreign ownership pressure) cannot be determined with the data; they are offered as possible explanations. (p. 260) | Strike rates (Table 2) and narrative of foreign firms recognizing unions; no tests. | Asserted |
The author's caveats¶
BMR data are household samples rather than worker samples; hours worked are unavailable (though all workers are full-time); the BMR data cannot distinguish industrial council coverage from plant agreements; the white CSS sample is small, nonrandom, limited to household heads in three cities; union membership figures are claimed and probably exaggerated; male-female differences should be treated with skepticism.
Gaps in the evidence¶
The argument would be better tested with establishment-level or matched employer-employee data including firm size, ownership (foreign vs local) and industrial council coverage, which would separate plant-level agreements from council rulings and test the foreign-ownership and militancy explanations. Hourly wage data, panel data on workers moving in and out of unions, and a worker-based rather than household-based sample would reduce bias; such data did not exist in 1985 but could be built from later labour force surveys. Union-density claims rest on unions' own claimed membership, and archival records of recognition agreements would help verify them.
Replicability: Data described but not public.