Skip to content

Comprehending Union Growth and Decline: The Case of the South African Independent Unions

Literature › Works

Ian Macun; Geoffrey Wood (2001). Comprehending Union Growth and Decline: The Case of the South African Independent Unions.
Type conference paper
Scope Economics and economic history of apartheid
Method descriptive quantitative
Period studied 1972–1996
Regions South Africa, Durban, East Rand, Natal
Themes unions strikes, labour markets wages, transition post apartheid, resistance social movements, manufacturing industrial policy, politics ideology elections
Citation key macun2001

Argument

The paper uses the South African independent unions to show that union growth and decline cannot be explained only by cyclical, structural or institutional forces. Cyclical, structural (racial fordism and its crisis) and institutional (Wiehahn reforms, 1995 LRA) factors all matter, but they interact in complex ways. Unions grew even under adverse institutional conditions (the 1970s) and recovered quickly from setbacks after 1976 and 1987. Strategic choices and internal organisation shaped outcomes: COSATU unions outperformed the Africanist NACTU unions. The authors argue against determinism and for closer study of union strategic decision-making. The argument rests on descriptive membership and density tables plus secondary literature.

Evidence and method

Descriptive tables of non-agricultural employment, union membership and density (CSS South African Labour Statistics, Hendrie & Kooy 1979, Department of Labour reports, SALDRU directories); sectoral density tables; macroeconomic indicators (Standing, Sender & Weeks 1996); ILO South African Labour Flexibility Survey 1995-6; secondary literature (Maree 1987, Friedman 1987, Kelly 1998, Nattrass 1992); statutes and commission reports.

Place in the debates

Findings

  • The Soweto uprising and banning of organisers caused only a small membership loss, suggesting the literature overstates their effect. (p. 6)

    the unions only lost roughly 8000 members between 1976 and 1977. This relatively small decline could point to an exaggerated effect, in the literature, of the Soweto uprising

  • The 1970s wage rises driven by the gold windfall and a desire to cut reliance on foreign migrant labour were weakest in Durban, where union growth was strongest (relative deprivation). (p. 9)

    led to a round of wage increases. In areas where this lagged, most notably Durban, union growth was strongest.

  • Wiehahn reforms were facilitative rather than protective: they excluded agriculture, public service and domestic work and imposed no duty to bargain. (p. 11)

    it did not apply to the public service, to agriculture, nor to domestic workers.

  • The 1980s were the take-off phase of union growth, with new members far outpacing job creation and density roughly doubling. (p. 13)

    an average of 191 378 new members were joining unions each year. This is far higher than the average number of jobs being created per year which stood at 58 946 for the 1980s

  • Union density rose very sharply in mining after the formation of the NUM in 1982 (from 6.81% in 1980 to 55.16% in 1990). (p. 13)

    Mining & Quarrying Employment TU membership Density 712 006 49 721 6.98 709 042 48 330 6.81 724 587 202 671 27.97 692 900 382 223 55.16

  • Structural change in the labour force (semi-skilled Africans, permanence of urban African workers) strengthened African workers' bargaining position in the 1980s. (p. 17)

    By the mid 1980s, an increasing proportion of African workers were at least semi-skilled (LRS 1988; 1990), reflecting the needs of industrial expansion

  • Hard-line employer responses after the 1987 State of Emergency and economic turnaround weakened unions such as the NUM, notably after the 1987 miners' strike. (p. 25)

    The latter culminated in the bloody 1987 miners' strike, which resulted in severe setbacks for the National Union of Mineworkers (NUM).

  • Union membership fell by about 400,000 in 1994, driven mainly by retrenchments in a period of worsening economic performance and tariff cuts. (p. 18)

    The main reasons underlying the drop in union membership was thus a loss of union members through retrenchments

  • Growth was uneven within the independent movement: NACTU's exclusive Africanism and reliance on key officials left it less successful than COSATU. (p. 29)

    The NACTU unions, which have pursued a particular set of organizational strategies – most notably focusing on exclusive Africanism, and a reliance on key officials and leadership figures - have generally been far less successful than their COSATU counterparts.

Evidence ledger

Sources used

Source Kind Period How used
Central Statistical Services (CSS), South African Labour Statistics (non-agricultural employment series) official statistical series 1970-1996 Provides the employment denominators for calculating union density in Tables 1, 2 and 4.
Hendrie & Kooy (1979), Trade Unions in South Africa: Some Statistics, SALDRU Working Paper No. 22 secondary literature compilation 1972-1979 Supplies membership figures for the unregistered (independent) unions in the 1970s, from which the early growth and the 1976-77 decline are derived; 1978-9 figures are estimates.
Department of Labour, Annual Reports official statistical series 1980-1996 Source of registered union membership totals for the 1980s and 1990s used to calculate density in Tables 2 and 4.
SALDRU Trade Union Directories (with Hendrie & Kooy 1979) secondary literature compilation 1977, 1980, 1985, 1990 Supplies sectoral union membership and employment for Table 3, showing density by sector and concentrations of unionisation.
Standing, Sender & Weeks (1996), Restructuring the labour market: The South African challenge (ILO) – macroeconomic indicators secondary literature compilation 1980-1995 Source for GDP and private consumption growth rates (Table 5) used to link recessions to union growth and decline.
South African Labour Flexibility Survey (SALFS), rounds 1 and 2, 1995-1996 household or labour survey microdata 1995-1996 Shows union presence and unionisation of manufacturing firms by size (Table 6) and the Cosatu/Nactu share of main unions.
Wiehahn Commission (1979), Report of the Commission of Inquiry into Labour Legislation, RP 47/1979 commission or official report 1979 Cited for the recommendations that African unions be legally recognised and registered and for the new industrial court, shaping the institutional-change argument.
Industrial Conciliation Amendment Act (no. 94 of 1979) and Labour Relations Amendment Act of 1981 legislation gazette 1979-1981 Establishes the legal reforms that opened registration and collective bargaining to African unions.
Labour Relations Act no. 66 of 1995 and 1987 LRA amendments legislation gazette 1987, 1995 Used to show the 1990s legislative framework protective of unions, and the 1987 amendments as an attempted limit on union rights.
Black Labour Relations Act 1973 (liaison and works committees) legislation gazette 1973 Illustrates the limited concession to African workers in the 1970s institutional climate.
Public Service Labour Relations Act (1993) and public service organisation counts legislation gazette 1980s-1996 Establishes the timing of public service unionisation and the growth from one organisation to over 20 representing about 760 000 employees.
TUACC Constitution 1974 union or movement records 1974 Documents the aims of the first coordinating body of independent unions that led to FOSATU.
Labour Research Service, Operatives Wages 1987-88 and 1989-90 other 1987-1990 Cited for the rising proportion of semi-skilled African workers by the mid-1980s.
ASSACOM Business Confidence Indexes 1-12/1987 employer association records 1987 Supports the claim of an upsurge in business confidence following the 1986 State of Emergency.
IPSA Research 1988 on the 1987 Labour Relations Act amendments secondary literature compilation 1985-1988 Documents the 1987 amendments' ban on sympathy strikes and liability for damages.
FSA Contact wage data via Baskin (1996) and COSATU membership data via Baskin (1994) secondary literature compilation 1989-1996 Used for relative wage increases of lower-skilled versus white-collar workers and the fall in manufacturing's share of Cosatu membership.
South African Labour Bulletin and Maree (ed) 1987 as documentary record of independent unions secondary literature compilation 1972-1984 Relied on for the narrative of union origins, banning orders, defeats at Heinemann and Armourplate, and registration debates.

Inference

Quantification: Descriptive tables.

Narrative, periodised case study of the independent unions from the 1970s to the 1990s, supported by descriptive tables of membership, density and macro indicators. Claims about cyclical, structural and institutional drivers are inferred from visual comparison of time trends and sector tables against events and legislation; the authors state that source limitations rule out quantitative analysis and use Kelly's long-wave and mobilisation theory as a framework.

Key claims

Claim Basis Strength
Independent union membership grew from almost nothing in 1972 to about 87 000-95 000 by the mid/late 1970s, and the post-1976 repression produced only a small loss, suggesting the literature exaggerates the effect of Soweto. (p. 6) Hendrie & Kooy (1979) membership figures in Table 1, with 1978-9 estimated; interpretation from comparing 1976 and 1977 totals. Partial data
The 1980s were the take-off phase of union growth, with density roughly doubling, because average annual membership gains far exceeded job creation. (p. 13) CSS employment and Department of Labour membership series in Table 2, computed means. Systematic data
The Wiehahn reforms and registration were facilitative rather than protective, and were crucial to the growth of unions. (p. 11) Wiehahn Commission report, 1979 and 1981 legislation and secondary commentary; no direct test linking the legislation to membership change. Illustrative or anecdotal
Union growth slowed in 1982-84 and again in the early 1990s because of recession and retrenchments, but membership recovered quickly. (p. 18) Membership and employment in Tables 2 and 4 set against GDP growth in Table 5, compared by eye. Partial data
Union membership is highly sectorally uneven, with the initial concentration in manufacturing and later rapid growth in mining, public sector and services; sectoral figures are approximations. (p. 14) Table 3 from Hendrie & Kooy and SALDRU directories; classification of unions by main area of operation. Partial data
Unionisation in manufacturing in 1995-96 was very high, with 86 per cent of surveyed firms having a recognised union, rising with establishment size and dominated by Cosatu. (p. 21) SALFS rounds 1 and 2 (n=329/399) in Table 6. Systematic data
Strike waves, union growth, bargaining changes and mergers cluster together in 1973-4, 1979-81, 1985-7 and possibly 1995-6, and union strategic choices shape the pace of recovery. (p. 26) Figure 3 chronology and Kelly's long-wave framework; no strike data series presented. Illustrative or anecdotal

The author's caveats

The authors state that black unions were not legally recognised until 1979, so no official membership record exists, and that official procedures for calculating economic data changed several times. They say these limitations confine the analysis to descriptive rather than quantitative forms. They also note that 1978-9 membership figures are estimates and that sectoral membership figures are approximations based on classifying unions by main area of operation.

Gaps in the evidence

The argument that strategic choices explain differential growth lacks union-level membership series, internal union records, and interviews with organisers; these exist in archives and union records and could be gathered now. Causal claims about institutional reform, recession and mobilisation rest on visual comparison of aggregate trends, with no strike-day series and no controlled or panel analysis of unions or sectors. Membership estimates for the 1970s and sectoral breakdowns are weak, and better audited membership returns and later registrar data could test them.

Replicability: Data described but not public.