Black trade unions: growth, wage effects, Wiehahn and strategy¶
What drove black union growth, how large were union wage effects, and what strategies (workerism, populism, social movement unionism, corporatism) did unions pursue?
Works engaging the debate: 37
Positions¶
- Wiehahn as co-optation vs facilitation of union growth
- Organisation and agency vs structural and legal explanations (Maree, Baskin)
- Union premia large and compressing vs insider-outsider costs (Moll, Schultz and Mwabu)
- Workerist vs populist and SMU debates (Lambert, Webster, Von Holdt); corporatism and social contracts
Evidence base¶
The debate rests on four bodies of evidence. The first is a legal-documentary base that almost every work shares. It comprises the Industrial Conciliation Acts of 1924 and 1956 and their 1979-81 amendments, the Wiehahn and Riekert reports and White Papers, and the Botha Commission of 1951. These texts are usually read through secondary labour-law handbooks rather than archival files (Gould (1981)) (Budeli et al. (2008)) (De Clercq (1979)) (Webster (1984)) (Bendix (1989)). The second is a membership and density literature. Its core source is the Department of Manpower annual registered-membership returns divided by CSS non-agricultural or manufacturing employment. This yields roughly 702,000 registered members in 1979 rising to 2.72 million in 1991, with density moving from about 15% to 53% (Frost (1993)) (Macun and Frost (1994)). Researchers supplement it with SALDRU union directories (Hendrie and Kooy, Miller, Lundall), COSATU congress figures relayed through Baskin and SALB, and the 1995-96 SALFS firm survey (n=329/399 establishments, 86% with a recognised union) (Maree (1985)) (Macun and Wood (2001)). A rare mine-level source is Anglo American's company tabulations of recognised NUM membership by mine, source area and Paterson grade for February 1986 and June 1988 (Crush (1989a)) (Crush (1989b)).
The third body is econometric and narrow in its data. It consists of the 1985 Bureau of Market Research survey of urban African households, plus a small CSS white sample of 72 union and 235 non-union workers (Moll (1993a)) (Moll (1993c)). It also uses the 1993 SALDRU/World Bank PSLSD, with 2,364 African and 653 white male wage earners (Schultz and Mwabu (1998a)) (Schultz and Mwabu (1998b)). Hofmeyr applies a common set of four specifications to both surveys (Hofmeyr and Lucas (1998)). Siebert's annual 1947-84 Bulletin of Statistics series for gold mining and manufacturing take a different route. They combine earnings, employment, capital stock, gold price, SAIRR pass-law prosecutions and an interpolated union-density series (Siebert (1987a)) (Siebert (1987b)) (Siebert (1987c)). The fourth body is the strategy debate, which is argued almost entirely from case studies. These include seven plant agreements (Webster (1984)), three COSATU-linked firms (Maller (1994)), a single brewery (Poswa (1994)) and 63 interviews with 31 unionists at Highveld Steel (Von Holdt (2002a)) (Von Holdt (2002b)). Labour Monitoring Group employer surveys of the 1984-85 stayaways, including 71 unionised Transvaal factories, form a further source (Swilling (1985a)) (Swilling (1985b)). Other contributions rely on secondary syntheses, such as Cameron's 18-country study relayed second-hand (Maree (1993a)) (Maree (1993b)) and textbook aggregates (Hirschsohn (1998)). Genuinely archival work is confined to earlier periods: Department of Labour ARB files and union minutes (Lewis (1984b)), employer-association records (Coupe (1990)), and Botha Commission evidence with SATLC papers (Lichtenstein (2004)). A hand-coded file of 2,678 strikes from Work In Progress (1972-91), checked against ILO and GDELT data, is a new and still unfinished addition (Descriptive statistics on the strikes... (n.d.)).
Contested evidence¶
The Wiehahn dispute is mostly a disagreement over how to read one set of documents, the commission reports and the 1979 Act. Co-optation readings infer state intent from the commission's remit and from registration and control clauses (De Clercq (1979)) (Gould (1981)) (Webster (1984)). Facilitation readings infer the effect from the timing of later membership growth (Macun and Wood (2001)) (Frost (1993)) (Hirschsohn (1996)) (Hirschsohn (1998)) (Budeli et al. (2008)). Neither side tests the link directly. Frost concedes that legal reform was facilitative but 'parametric', on the strength of interviews reporting that workers 'knocked on the union door'. Membership figures are themselves contested because works mix incompatible universes. Registered DOM returns give about 702,000 members in 1979 (Frost (1993)). Claimed totals give 760,000 in 1979 and 1.8 million in 1985 (Moll (1993a)). Independent-union signed-up counts give about 70,000 in 1979 (Maree (1985)) (Lewis (1990)). Estimates of coverage likewise range from 8% of the Black labour force in 1983 (Maree (1985)) to 12% in 1984 (Webster (1984)). Frost and Macun state that both the DOM and union series are unreliable.
The agency-versus-structure dispute uses different evidence on each side. Maree relies on census plant-size data for the 1960s concentration of capital. Frost and Macun rely on visual juxtaposition of membership against CPI, unemployment and manufacturing employment, and reject cyclical theories without running the Bain-Elsheikh regressions they invoke. On wage effects, all authors draw on BMR 1985 and PSLSD 1993 yet reach different magnitudes. Moll finds a selectivity-corrected premium of about 24% in 1985. Schultz and Mwabu find about 60% at the mean in 1993, largest at the bottom decile. Hofmeyr shows that the premium rose under every specification, entirely through coefficients. He also shows that the level of the premium is unstable and that the two surveys measure membership differently, by dues paid in 1985 and by a direct question in 1993. Siebert's aggregate time series attributes most of the 1980-84 rise in manufacturing wages to unions. However, it finds only small job losses (about 20,000). Schultz and Mwabu's insider-outsider employment costs rest on just 14 regions in one cross-section. In the strategy debate the evidence bases barely overlap. SMU accounts generalise from single cases, while LMG employer surveys are non-random (Swilling (1985a)). Fine's and Hirschsohn's models are largely asserted, and arguments about corporatism transfer OECD correlations without South African data (Maree (1993b)). Strike evidence also diverges: the WIP counts show a 1981 jump that the ILO series lacks (Descriptive statistics on the strikes... (n.d.)).
Underused and missing sources¶
Several decisive sources are rarely or never used. The Industrial Registrar's LR1, LR2 and LR45 membership returns and the Department of Manpower's audited union files are closed according to Frost (1993), so no one has union-by-union, union-by-sector series. Internal union and COSATU affiliate records are also absent. Industrial Council agreements and coverage records, which would separate ergo omnes extension from plant-level bargaining, are used only through Williams's wage-rate tabulations (Moll (1993c)). Matched employer-employee or panel wage data do not exist for the period. Chamber of Mines and TEBA turnover and re-engagement records, needed to test Crush's 'career migrancy' argument, have not been consulted. The official Department of Manpower strike series is cited only in summary form (Bendix (1989)) (Maller (1994)). The Wiehahn evidence submissions and the relevant Hansard debates, which could test the co-optation thesis on state intent, are barely used. No study has worked with productivity and downtime data from participation schemes. Together with the ILO strike data and the WIP strikes file (Descriptive statistics on the strikes... (n.d.)), these permit a cross-check of registered membership and strike incidence against the DOM-based tables in Frost (1993) and Macun and Frost (1994).
Strength of key claims¶
| Basis | Share of key claims |
|---|---|
| Systematic data | 12% |
| Partial data | 50% |
| Illustrative or anecdotal | 24% |
| Asserted | 14% |
Comparative figures for all debates are on the Literature overview.
Best-evidenced works¶
Hofmeyr and Lucas (1998), Moll (1993a), Schultz and Mwabu (1998b), Frost (1993), Siebert (1987c), Crush (1989a)
Most-used primary sources¶
Holdings are matched by machine against the archive's file manifest; a dash means no match was found, not that the project certainly lacks the source.
Works and their positions¶
| Work | Position |
|---|---|
| Bendix (1989) | Adopts the pluralist perspective explicitly, while acknowledging the radical critique that pluralism gives workers no real power. |
| Bendix (1989) | Presents both as complementary, sets out the advantages and disadvantages of each, and expects a drift towards centralised bargaining combined with plant-level participation. |
| Bendix (1989) | Argues the legislation is even-handed, and that employers have under-used it and mistakenly brand the Court employee-biased. |
| Bendix (1989) | Sceptical: these are likely to be seen as co-option and to fail without rights and control at the workplace. |
| Budeli et al. (2008) | Favours the functional/collective view: the right exists to secure a more equitable distribution of power at work and is both individual and collective, not an individualistic right to an isolated existence. |
| Budeli et al. (2008) | Attributes reform to the shortage of skilled workers and international pressure, plus the unworkability of the dual system after the 1973 strikes. Does not weigh competing explanations in depth. |
| Coupe (1990) | Builds on Lewis, Webster and Crankshaw but says their approach leaves the wider political and ideological dimension relatively unexplored. Aims to fill that gap. |
| Crush (1989a) | Rejects the strong version as 'now clearly untenable'. Migrancy has changed form and compounds can aid organising, though migrancy still limits union cohesion and exposes it to dismissal and replacement. |
| Crush (1989a) | Argues the compound, long thought to inhibit solidarity, has been turned into an organisational asset for recruitment, meetings and strike enforcement. |
| Crush (1989a) | Holds that access and recognition depend on mining-house policy (reformist Anglo American vs reactionary Gencor and Goldfields) and are necessary but not sufficient for unionisation; the old docility thesis is called into serious question. |
| Crush (1989b) | Rejects the conventional view (Southall) as 'clearly untenable' for the mines. Migrancy has changed form and the NUM grew fast, but migrancy still limits cohesion and exposes the union to mass dismissal. |
| Crush (1989b) | Argues the compounds, once seen as inhibiting solidarity, have been turned to the union's advantage for recruiting, meetings and enforcing solidarity. |
| Crush (1989b) | Argues that the NUM's growth and the 1987 strike overturn the traditional picture of docile migrants. |
| Crush (1989b) | Holds that company policy (reformist Anglo American versus reactionary Gold Fields and Gencor) was the necessary precondition for union membership patterns. |
| Crush and Jeeves (1993) | Mather's study calls into question the long-established idea that migrancy is a fundamental barrier. Union organizing is hardest among workers housed on the farm. |
| De Clercq (1979) | Treats reform as a strategy of co-optation and control that does not envisage fundamental change in the apartheid framework. It responds to labour militancy, business pressure and international codes of conduct. |
| De Clercq (1979) | Holds that the leadership's wartime subordination to the war effort and SACTU's subsuming of workers' interests into nationalist politics weakened shop-floor organisation, alongside state repression. |
| De Clercq (1979) | Favours democratic, shop-floor-based independent unions. She is sceptical of TUCSA's 'parallel' unions and of registered unions' motives of control. |
| Fine (1992) | He credits the unions with a 'proletarian reformation' and the rebirth of civil society. He criticises the trade union left for its two-stage theory and for absorbing its politics into the ANC-SACP-COSATU alliance. |
| Frost (1993) | Holds that the 1979 legal reforms were necessary but not sufficient (facilitative). Organisation, worker need for protection and political struggle are also treated as important, and no single factor is prioritised. |
| Frost (1993) | Finds the cycle-based price and prosperity theories cannot explain growth amid structural unemployment and prolonged recession. Regression was dropped because the data are unreliable and short-run. |
| Frost (1993) | Favours a multi-causal, eclectic approach. Admits its own analysis is mostly structuralist, and treats agency and leadership tentatively. |
| Frost (1993) | Reports the debate (Baskin, Morris and Erwin, Maller, the CASE survey) as unresolved and ambiguous. |
| Frost (1993) | Accepts density as a useful measure but notes the caveat that union power depends on strategic location, not just density. |
| Gould (1981) | Sceptical: reform is limited and aims to control black unions and impress the West, preserving white leadership and segregation. |
| Gould (1981) | Registration risks lending credibility to a control system and favouring parallel unions, but participation is understood; recognition should not depend on registration. |
| Hirschsohn (1996) | Holds that the reforms backfired on their architects; any view that policymakers deliberately used them as a trial run is an ex-post rationalization. |
| Hirschsohn (1996) | Treats COSATU's move to corporatist policy-making (NEF, NEDLAC) as a natural outgrowth of its founding mission, noting that it was hotly debated within the movement. |
| Hirschsohn (1998) | Builds on both, but argues SMU also depends on members' 'cognitive liberation' and grassroots pressure, not only leaders' strategic choices. |
| Hirschsohn (1998) | Treats the debate as resolved into a pragmatic 'strategic compromise' blending shopfloor strength with national democratic alliance. |
| Hirschsohn (1998) | Argues Marx glosses over the mechanism; worker control and grassroots pressure translated repression into union action. |
| Hirschsohn (1998) | Admits some bureaucratic tendencies but argues rank-and-file majorities in structures kept grassroots mobilisation alive. |
| Hirschsohn (1998) | Holds that the reforms aimed to control unions but had the opposite effect, opening space for rapid union growth. |
| Hofmeyr and Lucas (1998) | Finds premia highly sensitive to specification, with the switching model preferred on nesting tests. Sample selection corrections are judged probably relevant but not robustly detected. All methods agree the premium rose. |
| Hofmeyr and Lucas (1998) | Composition changes reduced the premium. The rise is attributed to changes in pay structure, i.e. higher rents or undetected composition shifts. |
| Hofmeyr and Lucas (1998) | Finds wage compression across experience, education and occupation among union members by 1993, consistent with union egalitarian aims. |
| Lewis (1984b) | Against Horrell, Walker and Weinbren and the Simons, who stress nationalist penetration. The decisive blow was the 1950 disaffiliation of the English-speaking craft unions, caused by deskilling and the shift to demarcation by race, with anti-communism as the precipitant. |
| Lewis (1990) | Argues that labor history became tied to the narrow 'workerist' survivalist tradition. He sees political unionism (Lambert) as a needed corrective that links unions to national liberation. |
| Lichtenstein (2004) | Rejects the view (Herbert Hill, Nelson and others in the US; sceptics in South Africa) that unions were simply instruments of white supremacy. He argues the CIO and the SATLC left made real, if limited, efforts at interracialism that the state and anti-communism undercut. |
| Lichtenstein (2004) | Grants white racism a role but holds that the political and legal context, set as much by the state as by the unions, was equally significant. |
| Lichtenstein (2004) | Accepts the critique (Witz, Berger) that parallel branches were subordinating, but says the later attack on parallelism was too quick to impugn the motives of every SATLC element before 1954. |
| Macun and Frost (1994) | 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. |
| Macun and Frost (1994) | 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. |
| Macun and Frost (1994) | 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. |
| Macun and Wood (2001) | Accepts that cyclical, structural and institutional forces matter but stresses that union strategy and internal organisation shape the speed and extent of recovery and growth. |
| Macun and Wood (2001) | Argues the literature exaggerates the damage, since membership fell by only about 8,000 in 1976-77 and recovery was quick. |
| Macun and Wood (2001) | Sees the reforms as facilitative, not protective: they created a stable terrain and bargaining access, but no duty to bargain and no right to strike, and unions grew largely on their own initiative. |
| Maller (1994) | Maller rejects both pure boycottist opposition and management-led participation without a union. She argues for a creative synthesis of collective bargaining and codecision-making, with union autonomy retained. |
| Maller (1994) | She argues that most management-initiated schemes (quality circles, suggestion systems) share information but do not extend workers' decision-making, and often undermine unions. They are only meaningful if unions are placed at the centre. |
| Maree (1985) | Cites Lambert, who blames SACTU's subordination to the Congress Alliance and premature mass political campaigns. Maree adds that SACTU leaders sought shortcuts around slow workplace organisation, and stresses that the 1970s unions built shopfloor power first. |
| Maree (1985) | Combines factors: skills shortages, industrial and political conflict, international pressure, regional decolonisation, and a shift in the class base of the government. Recognition was meant to co-opt and control unions, and was the 'lesser of two threats'. |
| Maree (1985) | Argues that dependence on foreign capital and the disinvestment lobby helped the unions survive, since they targeted subsidiaries of multinationals. Their international campaigns for recognition brought few formal wins by 1979. |
| Maree (1993a) | Maree argues unions should participate and rejects Callinicos's view that social contracts weaken socialism and workers; he accepts the risks of bureaucratisation and dualism and says they must be guarded against. |
| Maree (1993a) | Follows Cawson that corporatist theory is open-ended: unions need not be junior partners, contrary to Taylor's finding of union subordination in advanced countries. |
| Maree (1993a) | Drawing on Cameron, he sees a trade-off between militancy and employment, and stresses that corporatism cannot alone solve mass unemployment. |
| Maree (1993b) | Maree is cautiously in favour. He treats participation as a question of 'how' rather than 'whether'. He rebuts Callinicos and Vally-critics' claims of incorporation and bureaucratisation, while acknowledging risks. |
| Maree (1993b) | Maree follows Cawson in treating union power as variable and open-ended. He cites Cameron and the Australian Accord against Taylor's finding of union subordination, and holds that outcomes depend on class struggle. |
| Maree (1993b) | Following Cameron, he holds that wage restraint under centralised unions is traded for employment and a higher social wage, without lowering labour's income share. He concedes this is hard to apply where workers earn poverty wages. |
| Maree (1993b) | He adopts Schmitter's distinction and restricts the analysis to societal corporatism, arguing that union autonomy and mobilisation survive within it. |
| Mohamed (1991) | Presents Wiehahn as the state's and capital's response to economic and political crisis and to black worker militancy. It also holds that reform partly undermined itself by allowing independent unions to consolidate. |
| Moll (1993a) | Argues black unions achieved wage gains comparable to those in developed democracies by 1985, rejecting the expectation that the racial oligarchy's power would keep them small. |
| Moll (1993a) | Partly supports Blanchflower and Freeman's centralisation typology (large South African effects reflect decentralised plant-level bargaining) but notes the typology is inadequate, since the UK and Switzerland do not fit. |
| Moll (1993a) | Finds black unions compressed skill differentials, in line with standard US and Canadian findings, and attributes this to the living-wage strategy of the emergent unions. |
| Moll (1993a) | Suggests multinational pressure to recognise unions and raise wages may have contributed to the union effects, but cannot test this with the data. |
| Moll (1993c) | Moll argues against the expectation of small effects: even in 1985, black union wage effects were similar to those in developed democracies. |
| Moll (1993c) | 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. |
| Moll (1993c) | Suggests, without being able to test it, that multinational firms under Western pressure raised wages and recognised unions early, possibly with ripple effects. |
| Poswa (1994) | Argues that South African participation has been dominated by the management perspective, aimed at productivity, consent and bypassing unions, and calls for a shift toward real industrial democracy. |
| Poswa (1994) | Notes that unions long rejected participation as co-optation, but argues they must now engage with the schemes and shape them to serve workers' interests. |
| Schultz and Mwabu (1998a) | Mixed. Unions compress wages among members and narrow the racial gap within the union sector. Because most African workers are non-union, the authors say the premium may disadvantage them and the marginalised poor. |
| Schultz and Mwabu (1998a) | Regional estimates give small negative employment effects (elasticity of about -.06 for young men), but unemployment falls where the premium is larger, the opposite of the Harris-Todaro-Calvo prediction. Higher premia seem to push youth out of the labour force instead. |
| Schultz and Mwabu (1998a) | Industry controls absorb much of the union premium, which the authors take to reflect Industrial Councils or spillover of administered wages. Their data cannot measure firm size or council enforcement, so they suggest the union effect is probably understated. |
| Schultz and Mwabu (1998b) | Unions compress wages among members and narrow the racial gap within the union sector, but widen the union-nonunion gap. Most African workers are nonunion and may lose relatively. |
| Schultz and Mwabu (1998b) | Finds a larger union wage effect associated with lower employment and participation among African youth, but also with lower unemployment, contrary to the model's prediction. Concludes mainly that the distortion pushes people out of the labour force. |
| Schultz and Mwabu (1998b) | Cannot separate these because union status is treated as exogenous. The premia may overstate or understate union rents, and the authors flag this as a limitation. |
| Siebert (1987a) | 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. |
| Siebert (1987b) | Sceptical. Strikes reflect unexpected changes in the bargaining environment and the many new bargaining pairs after 1979, so they are not a reliable index of militancy. Union density is used instead. |
| Siebert (1987b) | Predicts small job losses in manufacturing (about 20,000), but possibly around 100,000 jobs shed across mining and manufacturing. This would put downward pressure on informal-sector and rural wages. |
| Siebert (1987c) | Union wage gains are real, but the estimated employment loss in manufacturing is small (about 20,000 jobs). Larger losses are possible if mine unionisation proceeds, with downward pressure on informal and rural wages. |
| Swilling (1985a) | Sympathetic to the community organisations. He describes the unions' 'workerist' reluctance as shaped by their defensive shop-floor position in the recession and finds that pure working-class politics cannot be organised outside the workplace. He also warns that populism risks dividing employed from unemployed workers. |
| Swilling (1985b) | Sees the Eastern Cape split between 'workerist' FOSATU unions and populist community organisations as damaging. In the Transvaal he sees the union-community alliance as a source of strength, and he does not endorse the unions' claim that the Port Elizabeth stayaway was against working-class interests. |
| Von Holdt (2002a) | Critiques existing SMU analysis (Lambert, Webster, Seidman, Moody, Waterman) as too externally focused on alliances, assuming a new class consciousness, and ambiguous. Proposes studying the movement dimension of unionism as 'contentious challenge' in tension with institutionalization, not a fixed model. |
| Von Holdt (2002a) | Sceptical: strategies are embedded in social structure, so strategic unionism failed when grafted onto SMU at NUMSA. Rejects the Moody/Waterman view that globalization produces a global SMU ('National reality counts'). |
| Von Holdt (2002a) | Argues for erosion, against Adler and Webster (1995), whom he criticises for assuming overarching continuity, and against Wood and Psoulis (2001), whose survey finding he regards as weak. |
| Von Holdt (2002b) | Argues the term has become ambiguous through over-broad application. Rejects a prescriptive SMU model in favour of analysing the movement versus institutional dimensions of unionism in specific contexts. |
| Von Holdt (2002b) | Holds that the union-community-ANC link was an interpenetration of movements bound by a national liberation discourse, not an alliance between autonomous organisations, so it shaped the union's internal structure. |
| Von Holdt (2002b) | Argues South African SMU rested as much on popular, ethnic and rural identities as on class consciousness, and that internal conflict and violence were integral rather than anomalous. |
| Von Holdt (2002b) | Argues SMU eroded after 1994 owing to democratic incorporation and class formation. Criticises Adler and Webster for assuming continuity, and Wood and Psoulis for loose usage of SMU and a quantitative method. |
| Von Holdt (2002b) | Argues strategies are embedded in social structure and do not travel easily, citing the failure of 'strategic unionism' at NUMSA; globalisation is unlikely to produce a global SMU. |
| Webster (1984) | Takes the Hyman/Marxist side: industrial relations is a power struggle over control, and grievances are settled by relative power rather than by rules. |
| Webster (1984) | Sees Wiehahn as a contradictory restructuring of repression plus reform. Recognition aimed to control unions from above but also opened space for shop-floor organisation. |
| Webster (1984) | Finds the major encroachment on discipline and dismissal, but little on directing tasks or evaluating performance. |