Social Movement Unionism: the Case of South Africa¶
Karl von Holdt (2002). Social Movement Unionism: the Case of South Africa. Work, Employment & Society 16(2). SAGE / British Sociological Association.
| Type | journal article |
| Scope | Wider apartheid literature |
| Method | survey fieldwork |
| Period studied | 1980–1996 |
| Regions | Witbank (Highveld Steel), Transvaal, South Africa |
| Themes | unions strikes, resistance social movements, mining migrant labour, transition post apartheid, comparative theory, labour markets wages |
| Citation key | holdt2002a |
Argument¶
Through an ethnographic case study of NUMSA at Highveld Steel, the article argues that South African social movement unionism (SMU) was shaped from within by identities formed outside the workplace. These included national-liberation politics, race, Pedi migrant culture and hostel/township divisions. These identities produced both formidable solidarity and violent internal contestation, which earlier SMU analysis, focused on external alliances and class consciousness, treated as incidental. The transition to democracy, with its political incorporation and black class formation, fragmented union solidarity. The attempt to graft Northern 'strategic unionism' onto SMU failed and deepened the erosion. Von Holdt concludes that SMU should not be treated as a transferable model. Analysis should instead focus on the tension between a union's movement and institutional dimensions across historical and national contexts.
Evidence and method¶
63 in-depth interviews with 31 current and former trade unionists (shop stewards and rank-and-file activists) at Highveld Steel, snowball-sampled and interviewed one to five times from 1993 over three years. Also observation of union meetings, informal conversations, and union and management documents supplied by NUMSA. Management refused to cooperate. Draws on his PhD thesis (2000) and secondary literature.
Place in the debates¶
- What defines social movement unionism (Seidman, Webster, Lambert vs Moody, Waterman) (see Black trade unions: growth, wage effects, Wiehahn and strategy). 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.
- Alliance vs interpenetration of unions and community/political movements (see Black trade unions: growth, wage effects, Wiehahn and strategy). 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.
- Class consciousness vs popular and premodern identities in black unionism (see Black trade unions: growth, wage effects, Wiehahn and strategy). 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.
- Continuity vs erosion of SMU after the transition (Adler and Webster; Wood and Psoulis) (see Black trade unions: growth, wage effects, Wiehahn and strategy). 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.
- Transferability of union strategies across countries (see Black trade unions: growth, wage effects, Wiehahn and strategy). 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.
Findings¶
-
Union-community-political links were an interpenetration of movements, not external alliances. (p. 6)
an interpenetration of movements, a complex and dynamic network of political, community and workplace struggles woven together by a discourse of national liberation struggle
-
The workplace was racially structured, forming an 'apartheid workplace regime' that the union contested. (p. 6)
The occupational structure was racially defined, with whites occupying skilled and managerial positions and blacks occupying unskilled and semi-skilled positions.
-
Migrant workers made up a majority of the black workforce at Highveld Steel and held the toughest, lowest-paid jobs, creating a migrant/township divide. (p. 7)
In 1980, more than half of the 4000-5000 black workers employed by Highveld Steel were migrant workers accommodated in hostels in nearby townships.
-
Pedi migrant disciplinary practices were imported into union solidarity-building, a source of conflict with township workers. (p. 8)
The sjambok became a means for building black solidarity in the union.
-
Conflict between the hostel-based strike committee and the shop stewards led to a breakdown of the union's constitutional order in 1990. (p. 11)
Eventually, in 1990, the group in the hostels declared that the shop steward leadership had been suspended and elected new office-bearers.
-
Democratisation separated political and union struggle, reducing militancy and solidarity. (p. 12)
The disaggregation of political and trade union struggle diluted the militancy and solidarity of workers.
-
Black class formation after apartheid drew experienced shop stewards into careers elsewhere, undermining the shop steward role. (p. 13)
Increasingly, the shop steward committee was seen as a stepping stone to opportunities for promotion or careers outside the factory, which undermined its traditional role as representative of the workers.
-
Imported 'strategic unionism' failed at Highveld Steel because of technical complexity and the loss of officials to government. (p. 15)
The technical complexity of the strategy contributed to a growing gap between shop stewards and members
-
Unionism is defined by a persistent tension between movement and institutionalisation, which should be the focus of analysis instead of a fixed SMU model. (p. 17)
Trade unionism is characterized by a constant tension between movement and the institutionalization and routinization of industrial relations.
-
Rejects the idea of a transferable or global SMU. (p. 18)
This argument implies that globalization is unlikely to produce the conditions for a globalized SMU as advocated by Moody and Waterman. National reality counts.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| 63 in-depth interviews with 31 trade unionists and former trade unionists at Highveld Steel (NUMSA shop stewards and rank and file activists) | interviews oral history | 1993-1996 (three-year period starting 1993) | Primary evidence for union social structure, migrant/township conflict, strike committee, and the post-1994 fragmentation of solidarity. |
| Observation of union meetings and informal conversations with trade unionists | author survey or fieldwork | 1993-1996 | Supplements the interviews with ethnographic observation of union practice. |
| Union and management documents provided by NUMSA | union or movement records | not specified (1980s-1990s) | Supplementary documentation alongside interviews; not cited individually in the argument. |
| Von Holdt PhD thesis (2000), 'From Resistance to Reconstruction: A Case Study of Trade Unionism in the Workplace and the Community (1980-1996)' | author survey or fieldwork | 1980-1996 | Holds the extensive interview material underlying the article's arguments; the article refers readers to it. |
| Discussions with William Matlala (Pedi migrant worker, labour photographer, former shop steward and hostel dweller) | interviews oral history | not specified | Source of insights on the kgoro, the sjambok and Pedi notions of male discipline in union solidarity. |
| Maller (1992) Conflict and Cooperation: Case Studies in Worker Participation; Moodie and Ndatshe (1994) Going for Gold; Mercedes Benz sit-in strike 1990 (von Holdt, 1990, SA Labour Bulletin) | secondary literature compilation | c.1980s-1990 | Cited as supporting evidence of politicization, disorder and violent contestation in other workplaces beyond Highveld Steel. |
| Buhlungu (2001) survey of four COSATU affiliates (PhD thesis) | secondary literature compilation | 1973-2000 | Cited as evidence of erosion of SMU through upward mobility of union officials and new fracture lines. |
| Wood and Psoulis (2001) national survey of COSATU members | household or labour survey microdata | c.2001 | Contrary survey evidence that COSATU reaffirmed its social movement role, which the author discounts on methodological grounds. |
| SA Labour Bulletin articles (Mtshelwane 1994; 'Amplats: An Avoidable Tragedy?' 1996) and Business Day (21 Nov 1997; 20 July 1999) | newspapers periodicals | 1994-1999 | Provides a sample of episodes of rank-and-file 'ungovernability' during the transition as suggestive evidence of SMU erosion. |
New data¶
Author-generated ethnographic dataset: 63 in-depth interviews with 31 activists at Highveld Steel, Witbank (1993-1996), plus meeting observation and NUMSA-supplied documents; the interview material is held in the author's PhD thesis.
Inference¶
Quantification: None.
Single in-depth ethnographic case study of one workplace (Highveld Steel/NUMSA), tracing the union over time with snowball-selected key-informant interviews and quoted testimony. The author uses the case to challenge social movement unionism theory, supplemented by a handful of anecdotal episodes from other unions and secondary studies.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Political and racial identity forged in the national liberation struggle was central to union collective identity, so that unions and community organisations interpenetrated rather than forming an external alliance. (p. 287) | Highveld Steel interviews with shop stewards and workers, quoted testimony on the aim of overthrowing the government. | Illustrative or anecdotal |
| Migrant (largely Pedi hostel) and township-resident workers had distinct identities and practices, such as the kgoro and the sjambok, which generated tension and shaped union solidarity. (p. 289) | Interviews, discussions with William Matlala, and observation of hostel and union practice, with the occupational split cited for the 1980s. | Illustrative or anecdotal |
| Internal conflict between a migrant-backed strike committee and the shop steward committee split NUMSA at Highveld Steel in 1990 into 'a union of the township and a union of the hostel'. (p. 292) | Interview testimony from shop steward office-bearers and strike committee activists. | Illustrative or anecdotal |
| After the 1994 democratic transition, militancy and solidarity declined as political and union struggle were disaggregated and black class formation drew experienced shop stewards into new political and managerial careers. (p. 293) | Interviews at Highveld Steel, supported by the Buhlungu study and a sample of other union episodes. | Partial data |
| The 'strategic unionism' strategy borrowed from northern unions failed at Highveld Steel and contributed to the demobilisation of SMU. (p. 296) | Interviews with shop stewards, NUMSA documents and the author's earlier SA Labour Bulletin pieces. | Illustrative or anecdotal |
| Erosion of SMU nationally is attributable primarily to sociopolitical change rather than workplace restructuring, and the SMU concept should be reframed around the movement dimension of unionism. (p. 300) | Single case study generalised, with brief anecdotal episodes from other unions; restructuring dismissed by assertion. | Asserted |
The author's caveats¶
The author notes that Highveld Steel management refused to cooperate, so the study rests on union-side informants selected by snowball sampling. He acknowledges that the evidence is a single case and refers to the thesis for the full interview material. He also states that restructuring played a role but treats it as secondary.
Gaps in the evidence¶
The argument rests on one firm, with union-side informants only and no management view or systematic data on membership, strikes, turnover or promotions. Triangulation with company personnel records, NUMSA branch records, strike statistics and comparative cases at other plants would test how typical the migrant-township split and post-1994 fragmentation were. Such records and later surveys of union members likely existed or can be consulted now, but are not used here.
Replicability: Data described but not public.