Living like there's no tomorrow: Trade union growth in South Africa, 1979-1991¶

Literature › Works

Ian Macun; Andrew Frost (1994). Living like there's no tomorrow: Trade union growth in South Africa, 1979-1991. Social Dynamics: A journal of African studies 20(2). Routledge (Taylor & Francis).


Type	journal article
Scope	Economics and economic history of apartheid
Method	descriptive quantitative
Period studied	1979–1991
Regions	South Africa, Western Europe and North America (comparative)
Themes	unions strikes, labour markets wages, inequality poverty unemployment, methods data sources, comparative theory, transition post apartheid
Citation key	`macun1994`

Argument¶

The article explores why union membership, especially in COSATU-affiliated manufacturing unions, grew so fast after the Wiehahn reforms. It argues that existing explanations (legislative reform and the political character of the unions) are only explanatory hints. Using Bain and Elsheikh style union-growth models, it adds organisational factors (leadership, shop stewards, mergers and the formation of COSATU) and economic factors (prices, real wages, labour-market change). It concludes that legal reform and political conditions facilitated growth, but organisational and socio-economic factors drove it. Organisational character mattered more than political character. Growth is expected to slow in the 1990s because of saturation, shrinking employment and the changing political terrain.

Evidence and method¶

Department of Manpower annual reports, South African Labour Statistics 1991 and 1992 (Central Statistical Services), National Manpower Commission reports, COSATU secretariat reports and membership figures, and Von Holdt's and Hindson's secondary work. Comparative density data from Kelly and the Bain, Price, Elsheikh and Bean and Holden literature. Frost's unpublished Honours thesis (UCT 1993) is also cited. The analysis is descriptive, using tables and graphs of membership, employment and density, and does not run its own econometrics.

Place in the debates¶

What explains post-1979 union growth: legal reform and political unionism versus organisational and economic factors (see Black trade unions: growth, wage effects, Wiehahn and strategy). Treats Wiehahn legalisation and political unionism as facilitating conditions only. Gives greater causal weight to organisational features (leadership, shop stewards, mergers) and to the socio-economic status of black workers.
Business-cycle and econometric models of union growth (Bain and Elsheikh) versus multi-causal approaches (see Black trade unions: growth, wage effects, Wiehahn and strategy). Argues the econometric model is too narrow for South Africa and favours a broader multi-causal framework combining qualitative and quantitative factors. It also says cycle and trend must be separated.
Do unemployment and recession undermine union growth, as in advanced economies (see Black trade unions: growth, wage effects, Wiehahn and strategy). Holds that high and rising unemployment in the 1980s had no apparent effect on South African union growth, though adverse conditions are expected to constrain growth in the 1990s.

Findings¶

Registered union membership was nearly stagnant before 1979, so large absolute growth is a phenomenon of the 1980s. (p. 11)

Prior to 1979, there was no substantial growth amongst the racially based, registered unions, with total membership increasing from 400 000 union members in 1956 to 701 758 in 1979
South African union density rose faster in 1979-85 than in any of the 17 advanced capitalist countries compared. (p. 10)

the percentage change in union density in South Africa during the period 1979-1985 was higher (at 12,3%) than any of the advanced industrial countries shown in Table 1
COSATU's membership expanded sharply between its 1985 founding and 1991, accompanied by consolidation into fewer unions. (p. 13)

By the end of 1991, COSATU's membership stood at 1 258 853 in 14 affiliated unions (46% of total membership).
COSATU unionisation in manufacturing grew massively, especially relative to black employment. (p. 15)

Absolute growth rose from 147 672 in 1985 to 672 951 in 1991. Relative growth over the same period shows a massive 38% increase in unionisation for all employees (10% in 1985 to 48% in 1991).
Exclusively white unions declined as the progressive black unions grew. (p. 17)

the gradual decline of the exclusively white trade unions, whose membership dropped from 350 000 in 1981 to approximately 235 000 in 1990
Part of the rise in density reflects slowing membership growth and a shrinking pool of potential members. (p. 17)

the rate of growth has begun to slow down and the number of potential members has decreased, thereby inflating relative union growth.
Density changes did not track political mobilisation or state repression. (p. 12)

Fluctuation in the rate of change does not appear to correspond significantly with phases of major political mobilisation or state repression
Shop-floor leadership was key to winning recognition and attracting workers to the progressive unions. (p. 20)

Erwin and Morris argue that an active shop floor leadership was key to effective recognition battles
Union growth tracked price rises closely in 1985-91, consistent with workers joining unions to defend living standards. (p. 21)

From 1985 to 1991 the shape of the curve in the graph of union membership mirrors that for prices almost exactly.
Rising unemployment did not curb union growth in the 1980s, unlike in advanced economies. (p. 21)

While unemployment is certainly high and has been increasing, there has been no apparent effect on union growth.
Organisational factors outweighed the political character of the unions in explaining growth. (p. 23)

The organisational character of the progressive unions has, therefore, been more important in affecting their growth than the political character of these unions.
Data on South African union membership are limited and need improvement. (p. 23)

Simply stated, this data is limited and requires improvement if further meaningful research is to be possible in this area.

Evidence ledger¶

Sources used¶

Source	Kind	Period	How used
Department of Manpower annual trade union membership returns (Annual Reports), registered and many unregistered unions	official statistical series	1956-1991	Provides the annual union membership series (total, registered, white-only) used to measure absolute growth and, with employment, density.
South African Labour Statistics 1991 (Central Statistical Services)	official statistical series	1979-1991	Non-agricultural employment figures used as the denominator for union density in Tables 2 and 3.
South African Labour Statistics 1992 (CSS) and COSATU Secretariat Report 1992	official statistical series	1991	Sectoral employment combined with COSATU affiliate membership to estimate COSATU density by economic division (Table 4).
SA Labour Statistics '92 and COSATU membership data (manufacturing employment total/white/black; COSATU manufacturing membership and density, Figures 2-6)	official statistical series	1979-1991	Used to compute COSATU density in manufacturing for all and black workers and its rate of change.
COSATU, SALDRU and Department of Manpower data on CWIU and FAWU membership (Figure 7)	union or movement records	1986-1991	Compares growth rates of the two oldest COSATU manufacturing unions, CWIU and FAWU.
National Manpower Commission Annual Report 1991, RP46/1992	commission or official report	1991	Source for NACTU membership (327 000 in 23 affiliates) and federation membership figures for 1991.
Von Holdt (1991) 'COSATU Congress', South African Labour Bulletin 16.1, COSATU membership figures 1990-1991	newspapers periodicals	1990-1991	Source of the 36% COSATU affiliate membership growth for 1990-1991.
Kelly, Trade Unions and Socialist Politics (1988): union density in 17 countries 1979-85	secondary literature compilation	1979-1985	Comparative density benchmarks against which South Africa's 1979-85 density change is judged.
Hindson (1991) data on racial composition of employment and tertiary sector share, 1960-1988	secondary literature compilation	1960-1988	Used to show changing labour demand: tertiary share rising from 50% to 54%, shifts in racial employment shares.
Bain and Elsheikh (1976) and Bain and Price (1980) union growth models and cross-country data	secondary literature compilation	c.1900-1970s	Provides the econometric framework and variables (prices, wages, unemployment, density) used to structure the South African explanation.
Frost (1993), South African Trade Union Growth 1979-1991, unpublished Honours thesis, University of Cape Town	other	1979-1991	Source for the CWIU/FAWU comparison and underlying compilation of union growth data.
Personal communication, J. Baskin, ex-national co-ordinator, COSATU	interviews oral history	c.1992-93	Sole evidence that sub-contracting has negatively affected union growth in some manufacturing divisions.
South African Labour Bulletin 17.1 and The Cape Times 19/10/1992 reports of COSATU membership decline	newspapers periodicals	1992-1993	Cited as recent reports of COSATU membership decline, interpreted as short-term fluctuation.
Industrial Conciliation Act 1979 and Labour Relations Act 1981	legislation gazette	1979-1981	Legal reforms extending union rights to all workers, taken to improve reliability of membership data and to facilitate growth.

New data¶

Compiled series of union membership and density from Department of Manpower and CSS data (1979-1991), plus disaggregated COSATU manufacturing membership and density figures by affiliate, built from COSATU and official sources; Frost's honours thesis is the underlying compilation.

Inference¶

Quantification: Descriptive tables.

Descriptive time-series and sectoral comparison of union membership and density (tables and graphs), plus comparison with 17 OECD countries; causal explanation is by informal inference from the Bain-Elsheikh framework and secondary literature, with no regression.

Key claims¶

Claim	Basis	Strength
Union membership grew dramatically after 1979 and density rose from about 15% in 1979 to about 53% in 1991, partly because employment stagnated. (p. 76)	Department of Manpower membership and CSS non-agricultural employment, Tables 2 and 3.	Systematic data
Before 1979 there was little growth in registered unions; the large increase in membership is a phenomenon of the 1980s. (p. 75)	Department of Manpower registered membership series, Figure 1.	Systematic data
COSATU unions in manufacturing grew very rapidly, with density among black manufacturing workers rising from 13% to 60% (1985-1991), exceeding the OECD experience. (p. 79)	CSS labour statistics and COSATU membership figures, Figures 3-4.	Partial data
White union membership declined from about 350 000 in 1981 to about 235 000 in 1990, so growth was the preserve of black progressive unions. (p. 81)	Department of Manpower annual reports, Figure 8.	Systematic data
Organisational factors (leadership, shop stewards, mergers) and economic factors (prices, wages) drove growth more than political character or legal reform alone. (p. 87)	Secondary literature (Erwin and Morris, Baskin, Bain and Elsheikh) and visual correspondence of membership with CPI; no statistical test.	Illustrative or anecdotal
Union growth tracked price increases and rising black real wages, though this cannot be shown statistically. (p. 85)	Visual correspondence of membership and CPI curves and inference from Bain and Elsheikh.	Asserted
High and rising unemployment has had no apparent effect on union growth, but falling manufacturing employment is slowing growth and inflating density. (p. 85)	Employment series and membership trends; Hindson; Von Holdt reports.	Partial data

The author's caveats¶

The authors state that the data on union membership are limited. Department of Manpower figures only became reliable after 1979-81 and cover registered unions. COSATU data are not easily accessible, and data for other unions are of dubious accuracy. Data for all COSATU affiliates are not available, so sub-sector analysis is partial, and the effects of prices and wages cannot be illustrated statistically.

Gaps in the evidence¶

The causal claims about organisation, politics, wages and prices are not tested with data; a multivariate time series or panel on union-level membership, real wages, CPI and strike activity would be needed. Union-level membership by sector, age and gender, and audited COSATU affiliate returns, would test the claims about sub-sectoral unevenness. Some of this existed in union and Department of Manpower records, but it was not accessible to the authors. Density depends on employment definitions that exclude agriculture and many informal workers.

Replicability: Data described but not public.