Introduction
Deviations from the standard employment relationships (SER), which provide full-time, permanent positions, have become increasingly prevalent under neoliberal pressures for greater profits and tighter budgets (Barchiesi 2007; Jordhus-Lier 2013; Lee and Kofman 2012; Standing 2011; Webster et al. 2008; Webster, Lambert, and Bezuidenhout 2008; Wilderman 2015). A key outcome of the SER is a workforce that is relatively easy to unionise (Godfrey et al. 2010). In South Africa, the end of apartheid and the election of the African National Congress (ANC) in 1994 saw contradictory processes evolve around employment. On the one hand, an opening up of the country’s economy exposed private and state-owned companies, along with municipalities, to neoliberal pressures to cut labour costs; while on the other hand, labour legislation, notably the Labour Relations Act (LRA) of 1995, strengthened employees’ individual and collective rights. The LRA focused primarily on employees in SERs (hereafter ‘permanents’) and on the rights of unions registered with the Department of Labour. Members of a registered union are permitted to embark on a protected strike, in which strikers’ jobs are safeguarded provided they follow a dispute resolution process managed by the Commission for Conciliation, Mediation and Arbitration (CCMA). The resulting tension, between neoliberal pressures and union-friendly policies, has resulted in a proliferation of employment forms differing from the SER (Benjamin 2006; Theron 2010).
There is a range of terminology used to describe deviations from the SER. Benjamin (2006) lists: casualisation, the use of temporary or part-time contracts; externalisation, employment regulated by a commercial contract in a triangular employment relationship between worker, employer and client; and informalisation, where work is unregulated – the informal economy. Taken collectively, these non-standard or atypical employments constitute precarious work, with limited employment security, low wages and limited benefits.
Precarious employment formats increase labour flexibility and decrease the ease with which a workforce can be unionised (Barchiesi 2011; Kenny 2004; Vosko, Macdonald, and Campbell 2009; Wilderman 2015). This case study looks at the use of labour broking, a form of externalisation, in the South African Post Office (SAPO or Post Office). Labour brokers provide workers (hereafter ‘casuals’, the term used by the SAPO labour broker employees themselves) to client organisations via a commercial contract. The client supervises the casuals. However, they are paid by the labour broker, who also remains responsible for disciplinary action. The resulting situation confuses the employment relationship with casuals employed by the labour broker but working for the client. The presence of multiple labour brokers providing casuals to a single client complicates union organisation and recognition. The lack of joint, legal liability of the client and labour broker for dismissal, in the pre-2015 amended LRA, meant that casuals could be fired at the request of the client company and the labour broker could justify this as a necessary commercial decision (Benjamin 2014). As Forrest (2015) points out, labour brokers’ profitability depends on preventing unionisation.
The creation of a vulnerable, unorganised stratum of the labour force threatens the position of unions as the key organisational form of the working class. In South Africa, Kenny and Webster (1998) described this as an ‘erosion of the core’. The Congress of South African Trade Unions (COSATU), the main union federation in South Africa, allied to the ANC, faced a dilemma: precarious work undermined their strength while methods of organising workers were inadequate when it came to casuals (Barchiesi 2011, 78). A protracted campaign by COSATU called for the banning of labour broking. However, on the ground, labour brokers entered COSATU-unionised companies resulting in divided workforces of permanent (unionised) and precarious (un-unionised) workers.
A range of strategies have been proposed by academics and activists for unions to shore up their position within a changing workforce and to act as champions of the wider working class, including explorations of different power resources. One prominent example is the moral or symbolic power of shaming client companies into abandoning labour broking or other forms of externalisation (Chun 2009; Webster 2015). There has also been attention given to alternative organisational forms that could be used by unions, or other organisations, to successfully mobilise precarious workers (Agarwala 2013; Fantasia and Voss 2004; Fine 2015; Lambert, Webster, and Bezuidenhout 2012; Ness 2014; Scully 2015; Theron 2009). The concept of social movement unionism, in which unions take on issues relevant to the working class as whole and not only the concerns of their members (Seidman 1994), is frequently cited as an approach that should be prioritised (Buhlungu and Tshoaedi 2012; Fine 2015; Kenny and Webster 1998; Webster, Lambert, and Bezuidenhout 2008).1
A smaller number of scholars have tracked how precarious workers have, in practice, organised themselves outside of unions. Sinwell (2015) has described how dissatisfaction with the National Union of Mineworkers (NUM), the dominant, COSATU-affiliated union in the platinum mines, led to the formation of workers’ committees and the unprotected strike of 2012, which cumulated in the Marikana massacre. Chinguno (2015) describes the ‘violent solidarity’ of this dispute in which intimidation, backed by violence, was used to ensure workers supported the strike. Sinwell (2015) outlines how these workers’ committees co-opted a rival union to NUM, the Association of Mineworkers and Construction Union (AMCU), which provided public leadership for a prolonged, protected strike in 2014 that resulted in substantial gains for workers. This co-option, Sinwell explains, resulted in an often uneasy co-existence of workers’ committees within the post-2012 AMCU. It also resulted in AMCU being an insurgent union; one supercharged by the anger of a newly incorporate militant rank and file with experience of extra-institutional struggles.
There are similarities (and differences2) between Sinwell’s description of industrial relations on the platinum belt and events within SAPO. In SAPO, casuals became increasingly dissatisfied with the dominant, COSATU-affiliated Communication Workers Union (CWU) and organised their own workers’ committees in Gauteng Province.3 After protracted attempts to resolve their inferior employment status through constitutional means (Dickinson 2017), the committees engaged in an unsuccessful unprotected strike in mid 2011 (hereafter ‘the mid-2011 strike’). This was followed in December 2011 by a protracted struggle waged by a minority of the casual workers, the Mabarete, using violent tactics to disrupt mail delivery in Gauteng townships (hereafter ‘the Mabarete strike’). The name Mabarete (The Berets [Sesotho], Amabarete in isiZulu) was a reference to the police’s paramilitary Tactical Response Team (nicknamed for their berets), which meted out corporal punishment when patrolling townships.
Having forced SAPO to the negotiating table in late March 2012, the Mabarete sought the assistance of officials from the South African Postal Workers Union (SAPWU), a registered union which had split from CWU in 2009. The result was an agreement to convert all SAPO casuals to permanent positions. The Mabarete then joined SAPWU, but in early 2014 established, with the SAPWU officials they had initially partnered, a new union: the Democratic Postal and Communications Union (DEPACU). DEPACU was registered with the Department of Labour and later recognised by the Post Office. A three-month period of industrial chaos at the end of 2014 (hereafter ‘the 2014 strikes’), largely the result of foot-dragging by SAPO over the 2012 agreement on conversions, exhibited features of insurgent unionism in which DEPACU, and rival unions, alternated protected strikes and negotiations with covert forms of action that had been developed in earlier struggles.
After describing the fieldwork on which this article is based, the history of labour broking in SAPO is outlined. The initial focus of this article is the workers’ committees and the unconventional and unconstitutional struggle of the Mabarete between December 2011 and April 2012. This is looked at through the lenses of several themes: liminality and networks; mutual mobilisation; secrecy; the application of (violent) labour; geographical limits and community embeddedness; and home visits and negotiations. There are four reasons for focusing on the Mabarete’s struggle. Firstly, it documents a moment in which casual workers organised outside of unions and outside of the industrial relations/constitutional framework. Their struggle provides, secondly, an example of industrial action based not primarily on the withdrawal of labour, but the application of labour. Thirdly, it was the phase of struggle that was finally successful in removing labour brokers from SAPO, even if other pressures contributed to this outcome. Fourth, the Mabarete’s success raises an important question: if the forms of organisation and power projection operated during this period were successful, why were these formats abandoned and formal union structures embraced when such organisations had previously failed them?
Following an assessment of the Mabarete victory, the article moves on to address this question. A foundation for this is provided by critically examining the benefits that the organisational form of workers’ committees may have over unions and the related issue of how power can be projected by these different organisational forms. Two models of industrial engagement are described: registered unions operating within a framework of institutionalised conflict and workers engaging in subaltern worker rebellions. Excluded from the constitutionally framed industrial relations system, casuals proceeded to subaltern worker rebellion, an approach that served them well. Once their goal of ending their casual status was achieved, it made sense to enter into the industrial relations system where their gains could be consolidated and advanced. However, insurgent unionism, as an oscillation between these two different models, was adopted as a third, hybridised method of projecting power in order to complete their conversion to full permanent status. The achievement of permanent status, with full salary and benefits, is key to explaining their transition from workers’ committees, via insurgent union, to registered union.
A conclusion reflects on how the case study deepens our understanding of: organisational forms, and their linked power resources available to precarious workers; contemporary industrial struggles in South Africa; and how this plurality of available organisational forms and tactics contributes to a potentially unconstrained proliferation of violent classification struggles.
Methodology
I initially followed the SAPO casual workers’ struggles through informal discussions with workers in East Rand townships as a result of regular visits over 15 years. In late 2013, I approached the leaders of the Mabarete and SAPWU officials, then working together and later to form DEPACU, requesting research access. I interviewed eight of the ‘Top 9’ Mabarete leaders, along with others who had played a range of leadership roles, and the SAPWU officials. I also interviewed individuals with insights into labour broking in SAPO and the struggles that resulted. In total, I conducted 53 formal interviews of between one and two hours’ duration, with 33 individuals. Additionally, I held a group discussion with five SAPO supervisors on their experiences of the casuals’ strikes.
Engagement with the DEPACU leaderships also resulted in participatory observation, including the 2014 strikes. This provided me with the opportunity to conduct additional informal interviews with rank-and-file members of the old workers’ committees. Being ‘inside’ this strike deepened my understanding of what had happened in the Mabarete strike as well as the practices of insurgent unionism. CCMA and Labour Court files provided additional information, while a range of other documents, including handwritten minutes of worker committee meetings, provided by worker leaders, assisted in clarifying dates and processes.
Information drawn from informal discussions and observations is referenced as ‘Fieldnotes’. Ethical permission for the research project was granted by the Wits Human Research Ethics Committee. I avoided discussions in which individuals might render themselves or others vulnerable to legal action.
The South African Post Office and labour brokers
The post-apartheid government charged state-owned enterprises with a dual mandate: first, to roll out infrastructure that would support development goals, and second, to do this without burdening state finances. The Post Office’s key role in development was the establishment of ‘points of service’ (retail post office facilities) and the provision and servicing of physical addresses across the country. Mail delivery, particularly in townships where penetration of Internet-based communication has been slow, generates significant income for the Post Office from clients such as banks, municipalities, stores selling on credit and other companies that need to bill clients and provide accounts.
The Post Office delivered on much of its developmental mandate and it turned around its finances from an approximate R1 billion annual loss in the late 1990s to profits (excluding once-off income and government subsidy) from 2005 to 2010. However, this financial viability was achieved by an unremitting emphasis on cutting costs. A major mechanism in which expenses were contained was the use of labour brokers. In rounded, monthly 2011 figures, while permanent workers were paid R8000, labour brokers received R4000 to provide a casual worker who earned R2000. In addition to salaries, the lack of benefits, such as pension and medical aid contributions, negotiated by CWU for permanent employees, meant that the total cost to a company of a casual worker was approximately one third of a casual or, as explained by a SAPO manager, ‘Three for the price of one!’ (Interview, 2 December 2014).
Following a board decision in the early 2000s, recruitment of permanent workers into entry-level positions was frozen. This resulted in increasing number of casuals in SAPO. By 2011 there were some 8000 casuals, approximately one-third of SAPO’s workforce. Casual workers were particularly concentrated in SAPO’s mail delivery business, which, in the Gauteng region, largely employed African males. Labour-broking companies were typically owned by black entrepreneurs or, in some cases, Black Economic Empowerment fronts for established companies. Most took a militant anti-union stance and blocked any attempt by casual workers to organise. In 2005, CWU reached an agreement with SAPO to end labour broking within two years. However, minimal progress was made and subsequent annual agreements simply extended the deadline.
Liminality and networks
As a result of CWU’s failure to end labour broking in SAPO, casual workers began to establish workers’ committees. Although assistance was given by some sympathetic depot-level managers (who were also CWU members and sometimes officials), these committees came about through the efforts of casual workers themselves. Some committees proved to be ephemeral and the activities of all committees fluctuated, but eventually half a dozen, geographically based organisations were established in: Tembisa, the ‘Far East Rand’, Soweto, the ‘Far West Rand’, the Vaal and Pretoria. Organisational cultures differed between these committees, thus the Soweto group appointed office bearers and pursued a strategy of ‘double footing’ in which they maintained contact with CWU while continuing to mobilise independently. By contrast, the Tembisa committee rejected permanent office-bearer positions as divisive and rotated leadership. They were also the first committee to break decisively with CWU, arguing that casuals had to rely on their own resources.
The committees were liminal organisations without official status and minimal infrastructure. Meetings, typically held on Saturday mornings, took place in parks, waste land, car parks and on the edges of township football fields. At best, minutes of meetings were handwritten in a notebook to be read out at the following meeting. Funds were raised on an ad hoc basis from those attending meetings. Only the Tembisa committee managed to organise a levy that all casual workers in the six depots that it organised were expected to pay. The Vaal committee, when attempting to register as a union with the Department of Labour in 2010, wrote a constitution, hired an office in an Internet cafe and set up a bank account. However, their application was rejected and these symbols of official status lapsed.
Despite the committees’ liminality, networks of personal connections maintained through personal contact and cell phones were established. These networks linked committees across Gauteng. However, this did not result, despite several attempts, in a stable coordinating structure. Indeed, the problem of committees working in silos was widely recognised as resulting in parallel initiatives and a lack of coordination. A critical problem was that casual worker leaders were locked into rivalries over strategies, such as whether to remain in contact with CWU, and over leadership positions. These differences presented within as well as between committees. One result was parallel networks of connections between like-minded individuals that straddled the committees.
Mutual mobilisation
These parallel networks played an important role in initiating strike action outside the industrial relations system. Without a registered union, obtaining a strike certificate from the CCMA was all but impossible. Those participating in an unprotected strike risked dismissal. In August 2008, TAS, the largest SAPO labour broker in Gauteng, dismissed 18 casuals en masse at the Edenvale depot, part of the Tembisa committee structure, after they delayed mail delivery in support of a colleague who had been transferred to another depot (CCMA 2008). In October 2009, 11 TAS employees at the Springs (Far East Rand) depot were dismissed for leaving work early to attend the Department of Labour’s public hearing into labour broking (CCMA 2009). However, through trial and error, the committees worked out how to protect an unprotected strike using mutual mobilisation. This used staged intimidation by one group of workers on another. One group would arrive at a depot and demand that the workers there join them. Casual workers could then claim they had no option but to join; this meant that they were not, at least immediately, vulnerable to disciplinary action.
At the beginning of the Mabarete strike, the Soweto committee initiators sent two representatives to Tembisa requesting support. They were sent back, by a key Tembisa leader, and told to return the next day to his depot ‘in numbers’. Embarrassingly, at that stage the Soweto strikers totalled only around 20. Nevertheless, helped by a piece of pre-arranged theatre, they brought the depot out. The Tembisa leader arranged to leave the depot on an errand and returned just as the strikers were arriving. The strikers then attempted to capture him in full view of the depot, giving the impression that, if caught, he would be beaten. He, however, managed to reach the depot unscathed. The anger of the strikers now demonstrated, those in the depot elected to join the strike. After that, matters became easier; numbers swelled with each depot visited in the Tembisa area.
However, even with a growing group of workers marching from depot to depot, it was not automatic that their arrival would see casuals join a strike. Mutual mobilisation required not only numbers but coordinated internal and external activity. Casuals in the depot would already know that a group was coming to take them out as a result of activists being in contact by phone. The internal activist would prepare their workmates who would be expecting the strikers and would have decided whether to join or not.
The limited and patchy mobilisation beyond Tembisa during the Mabarete strike illustrates that such intimidation could be ignored. Knowing that they were about to be mobilised, casuals in some depots locked the depot gates and kept out of sight until the strikers moved on. And if such open defiance was too risky, they could join the strike for a day and then drift back to work. At most, some 500 casuals in Gauteng joined the Mabarete strike, far less than the mid-2011 strike. This number was gradually whittled down, despite some latecomers joining the struggle, to fewer than 300.
Secrecy
The liminal status of the workers’ committee had always limited their visibility, but the visibility of the Mabarete leaders was increasingly and deliberately curtailed. In part, this was a result of using tactics that rendered them vulnerable to criminal charges. It was also a result of seeing what happened to leadership who put their head above the parapet. After the mid-2011 strike, which had ended with a Labour Court interdict that named and bound over six strike leaders, three of the six had written to E. T. Mpai, the Post Office’s Regional Manager, pointing out that their grievances were not resolved and requesting a meeting. They were charged by TAS with, inter alia, gross misconduct for requesting a meeting with a SAPO manager and were dismissed.
During the Mabarete strike the leadership remained all but invisible. SAPO again obtained an interdict, but was unable to specify on whom it should be served. Leaders who were already known to management didn’t participate in the disruption of postal deliveries. Centralised communication was facilitated by an ‘office phone’ which was rotated between leaders. Strikers knew that calls and SMS messages from the number were instructions, but didn’t know who was issuing them.
The application of (violent) labour
Although a range of tactics were used, the ho tsoma (Sesotho: to hunt) tactic was the Mabarete’s stock weapon against postal delivery. This involved groups of strikers patrolling townships looking for amagundwane (Zulu: rats, i.e. strike breakers). When groups went out ho tsoma they operated in areas in which they neither worked nor lived to avoid being identified. After the leadership had made a decision as to where groups would be operating, strikers were instructed to board a Metrorail train but not told their final destination, to avoid information being leaked. If necessary, they would be accompanied by a striker from the area who acted as a guide, particularly over where and when postmen4 would be delivering, but who would remain out of sight should a gundwane be captured.
Depending on circumstances, striker breakers who were caught were subject to warnings, stripping and beatings. Beatings became more severe as the strike dragged on. In a somewhat macabre lesson drawn from the mid-2011 SAPO Labour Court interdict, it was deduced that damaging Post Office property provided easily presented evidence. As a result, when working postmen were caught, their bicycle and letters were carefully put to one side before they were taken to a secluded place where they could be dealt with. Once a gundwane (strike breaker) had been caught and punished, all groups in the area were told, via cell phones, that a group had ‘prayed’ (i.e. dealt with a strike breaker). They would then disperse and make their way to the nearest station and leave the area.
This activity was not, however, without its risks. Numerous arrests were made, one estimate being 70 (Interview, 2 August 2014). The need to appear in court was an irritant, but in the end no one was convicted, generally because there was a lack of evidence or charges were withdrawn by complainants, the latter being something that the Mabarete could influence. At one point, replacement workers in Tembisa organised themselves to deliver mail in groups, carried weapons and fought back when confronted. This led to the development of the ‘mob squad’: a larger group able to overwhelm such resistance.
Through a combination of mobility and secrecy it was all but impossible to know where the Mabarete were operating on any particular day. Over large parts of Gauteng, postal deliveries ground to a halt.
Geographical limits and community embeddedness
Townships were affected most by the casuals’ strikes. Here, groups of African men could move about without attracting undue attention. It was a very different matter in suburbs where such groups would be quickly reported to police or private security companies. The presence of security cameras, common in city centres, also limited where the Mabarete could ho tsoma.
Within townships, the Metrorail system determined which areas were accessible to strikers. Unable to afford taxi fares, they were reliant on this much cheaper form of transport. When travelling as a group, there was no need to buy tickets as long as they didn’t attempt to enter or exit at larger stations. The group would approach the station security guards, who would open a gate for them. These guards, invariably Africans and themselves outsourced workers, were sympathetic to the Mabarete’s struggle. Even if they weren’t, armed only with a baton they were not going to argue with 20, 50 or perhaps 200 strikers. Delivery areas close to Metrorail lines were therefore much more vulnerable.
The police approach to the Mabarete was ambiguous. On the ground they were often supportive. As a highly unionised organisation, they often took a dim view of strike breakers. Amagundwane reporting assaults at township police stations were likely to be dismissed. While the police were required to take action when crimes were reported by Post Office management, their response was frequently dilatory and sometimes subversive. In one incident, two strikers had, in an unplanned encounter, confronted a group of replacement workers who were delivering early on a Saturday morning, one of the tactics used by some SAPO supervisors attempting to maintain mail delivery. The group had panicked and fled abandoning their bicycles, two of which the pair of Mabarete then stole.5 Later in the day they were tracked down by a SAPO supervisor with a police escort. They cooperated with the police, showing them the bicycles and explained, in detail and with passion, why they were on strike. The supervisor insisted that they be charged with theft, but the police refused on the grounds that the bicycles had been recovered and that no theft had therefore occurred (Fieldnotes, 11 February 2012).
The police’s attitude reflected commonly held values within townships: that strikes are a legitimate way for workers to pursue their interests and that strike breaking is a moral infringement. This is combined with a view that corporal punishment is an appropriate way of imposing discipline and teaching people right from wrong. Thus, in the township where the bikes were stolen, suspected criminals would periodically be caught and beaten by groups of residents. When this happened, police would generally wait for the punishment to be completed before stepping in, if necessary to transport the suspects to hospital (Fieldnotes, 19 January and 22 February 2013).
Home visits and negotiations
In March 2012 the Mabarete paid a ‘home visit’ to Mpai, the SAPO Regional Manager. It was not their first home visit, though it did complete a symbolic trinity. The first, spontaneous, visit was to the mother of TAS’s owner, who lived in Tembisa. The second, planned, visit was to the home of Clyde Mervin, the chair of CWU’s Gauteng provincial structure. A couple of hundred strikers arrived and delivered their message to terrified family members: the CWU leader should stay out of the casuals’ struggle. The visit to Mpai’s house was, in part, out of desperation; much as they were able to stop postal delivery they were getting nowhere. Using beaten amagundwane as messengers was not getting though and the Post Office’s strategy appeared to be to wait them out. The Mabarete were on the verge of collapsing as an organisation.
Mpai lived in a gated community outside of the Mabarete’s normal territory and a long way from the Metrorail. When the long line of strikers finally reached the complex gates, they posed as workers come for unpaid wages from their employer. The police were called, but eventually the group was allowed to enter. The format followed that of the visit to Mervin’s house though the message was different; they were hungry and would be back ‘to eat’ with Mpai if he did not deal with their grievances. By chance, a postman was delivering in the complex. His bicycle was left at Mpai’s gate and the man taken with the Mabarete.
They had not even reached the Metrorail station when they received a call on the office phone, brokered by an intermediary, that Mpai wanted to meet with them. Mpai could no longer go along with the organisation’s denial over labour broking; his family was now on the front line of an industrial war. The ho tsoma tactic had made it clear how far the Mabarete were willing to take matters. He confronted senior management and it was agreed to talk to the Mabarete.
The Mabarete, despite their long campaign, were unprepared for talks and turned to SAPWU, then still seeking recognition from SAPO. The SAPWU team helped shape their demand for permanent positions into a form that SAPO could take on board. The initial challenge, however, was management’s need to know that they were, in fact, talking to the strikers. The Mabarete had covered their tracks well. None of the Top 9 knew anything about the visit to Mpai’s house. Even the person who had taken the call on the office phone was unknown to them. To resolve this problem the Mabarete marshalled their members to stay in Germiston’s Golden Walk car park for a week. No incidents were reported and the management knew it was talking to the right people.
After that, agreement came quickly. There were two clauses, though nothing was put on paper to avoid provoking CWU, still at that point the only recognised union in SAPO. The first clause was that the Mabarete, all long since fired from employment by the labour brokers, would be employed directly on short-term contracts by the Post Office, paid at a rate equal to the fee labour brokers had been paid. This was, more or less, cost neutral for SAPO, but it meant an approximate doubling of the workers’ previous salary. It was also agreed that the Mabarete would be converted to permanent positions within three months. The second clause was that all labour brokers’ employees would follow a similar process. The Mabarete returned to their depots victorious in early April 2012.
Assessing the Mabarete victory
Not everything worked out in line with the agreement. Indeed, it was not until July 2016 that the conversion of all casual workers to permanent employment was fully completed. Despite this slow finalisation, the Mabarete were, arguably, pushing on an open door. The SAPO board had decided, in September 2010, to stop the utilisation of labour brokers. This, however, was never communicated or acted upon. Indeed, no one whom I interviewed was aware that this decision had been taken until it was revealed in the Public Protector’s report on SAPO released in February 2016 (Public Protector 2016). However, assuming that this decision was known to senior SAPO managers, it would have forced their hand once the Mabarete had finally got the issue into negotiations. There was also the longer-term process of legislative reform for labour broking that COSATU unions, including CWU, had been campaigning for over a decade and which the ANC publically supported in its 2009 election manifesto. This eventually bore fruit, with the LRA amendments which regulated, but did not ban, labour broking, on 1 January 2015.
Nevertheless, despite caveats, the magnitude of the Mabarete’s victory should be recognised. They succeeded in doing what other, much more powerful, forces had long been unsuccessfully campaigning for. They ended labour brokers in SAPO, immediately doubled the salary of casual workers and, over the next four years, achieved equal terms and conditions of employment for all SAPO workers.
Worker committee organisational form
The organisational form of the Mabarete emerged as a result of their exclusion from the industrial relations system. They were, in effect, denied the right to join or set up their own union; a right supposedly guaranteed by the South African Constitution. Denied this right, alternative means of prosecuting demands were developed.
Given dramatic changes in the composition of the workforce globally, with increased casualisation, externalisation and informalisation, it can be argued that workers’ committees present an alternative organising model that could be adopted by precarious workers. Such forms of organisation allow divided workforces, for example one fragmented by multiple labour brokers, to organise in ways optimal to their needs. Bypassing the legal requirements of formal unionisation also removes the need for officials and office bearers, along with the vested interests that such individuals soon establish. The absence of a union hierarchy keeps the leadership of workers’ committees close to the ground and close to the issues that concern casuals. This form of organising relies on voluntary contributions of time and money from members, along with the potential to mobilise, rather than demobilise, the rank and file. Finances are collected by hand and for specific projects, rather than through automated stop orders that are swallowed up by union salaries and overheads. Outside of the legally framed industrial relations system, casuals are forced into activism that explores alternative ways of projecting power.
This outline of the advantage of workers’ committees is, however, simplified and sanguine. For example, collecting money by hand not only is only time-consuming but brings additional problems, especially in the context of precarious workers’ financial situation. Activists who, given their employment status, are likely struggling to make ends meet and may be in debt, are tempted to divert collected money. The result is that they may avoid meetings where they would have to account for missing funds. While workers’ committees which ignore the legalities of employment contracts can bring unity to a workforce divided between different labour-broking companies, as with the SAPO, casuals’ political and personal differences can sometimes be impossible to contain within workers’ committees’ more informal structures.
It is easy to romanticise workers’ committees, especially from a distance. They may well form a viable organising alternative, but this is not automatic and any evaluation needs to involve a sober analysis that incorporates the complexities of social relationships and human weaknesses that may be particularly prominent within precarious workforces.
Power projection: two models of engagement
In analytical terms, the power resources available to worker committees are facilitated by their liminal status and freedom from the institutional constraints of the industrial relations system. As seen in the case of the Mabarete, they are able to project power in ways that are impossible for a registered union. They are also freed from the psychological constraints of industrial partnership in which ‘unreasonable’ demands are impossible to put forward. Union and management are tied together by projected financial viability that takes account of the needs and demands of external stakeholders. However, if we are to fully understand how workers’ committees can operate, we need to see them not as a deviation from the prescribed industrial relations system, but as an alternative.
In the prescribed form of industrial disputes, institutionalised conflict, workers organise in registered unions which then put forward demands to management. When the gap between union demands and management’s response cannot be bridged, the union declares a dispute. If the dispute cannot be resolved internally it moves into the state-sponsored industrial relations system, which will, if necessary, sanction a protected strike. The key effect is a withdrawal of labour where management suffers lost production and workers lose pay. There is a relationship between these two effects, governed by the numbers of strikers and their position within the labour process. The strike is a controlled trial of strength, or endurance, which ends with the reopening of negotiations. Such negotiations may well reach an agreement different from any original offer, but it will still remain a compromise between the original starting positions of union and management.
The alternative method of industrial engagement, illustrated by the Mabarete, subaltern worker rebellion, starts at the point where rank-and-file workers come to a realisation of what, they believe, will resolve their condition while also realising that they do not have mechanisms that will lead those with authority to concede this to them. At this point organising structures alternative to those prescribed by official frameworks emerge, such as workers’ committees. Demands may be presented, but this is not to open negotiations. Negotiations are viewed with deep suspicion, since negotiators are vulnerable to being bought off. Communicating their demands may, in fact, be difficult given the absence of established channels, the liminal nature of these workers’ structures and the shielding of leadership to avoid organisational decapitation. The pace and intensity of industrial conflict with which workers’ objectives are pursued is dependent on the workers organisations’ ability to project power and management’s ability to counter. Such conflicts are frequently conceptualised as wars, often with millennial-type hopes of prosperity. The outcome sought by workers is the conceding of their demand, in toto. In the absence of an agreed framework, limits to the conflict are self-imposed or result from practical considerations. Typically, the form of the disruptive action undertaken is likely to escalate to the level that matches the cost of conceding the demand or, alternatively, the struggle ends when the operational ability of the combatants is broken.
This description of subaltern worker rebellion, like that of institutionalised conflict, is an ideal type that will vary by circumstances. Insurgent unionism, addressed in the following section, comprises an oscillation between the two forms of engagement.
If it ain’t broke why fix it?
The question arises as to why the Mabarete, having prevailed in a David and Goliath conflict, then changed tack. This change first occurred when they brought in seasoned unionists from SAPWU to assist them in the April 2012 negotiations that the Mabarete alone were responsible for bringing about, and, then, secondly, and on a more permanent basis, when establishing DEPACU.
Calling on SAPWU officials was partly out of desperation. The Mabarete were at this point exhausted, destitute and dwindling in numbers, but following the visit to Mpai’s house had suddenly made a breakthrough to negotiations. It also, however, illustrates the structural constraint of the subaltern worker rebellion engagement strategy. Other than achieving total victory or being totally defeated, negotiations are inevitable. This was the situation in late March 2012. The Mabarete leadership recognised that they had reached their limits and did not have the necessary skills or knowledge to secure agreement with SAPO management.
The continuation of the relationship between the Mabarete and SAPWU officials was, in part, due to a power struggle within SAPWU. The SAPWU officials saw the Mabarete as a power resource that could be tapped. In a candid moment, one of the SAPWU officials explained how joining with the Mabarete had given him a ‘private army’ which, with careful handling, could be used in disputes with rival unions and rivals within the union. However, despite the Mabarete joining SAPWU their limited numbers curtailed the decisiveness of this power resource; one of the reasons why the Mabarete–SAPWU officials’ alliance then set up DEPACU.
Further, the combination of experienced union officials and worker committee militants provided a potent structure able to operate both within and without the industrial relations framework when in dispute with management. What was evident in the 2014 strikes was an oscillating between constitutional process and covert disruption. Coded messages instructed a relatively small number of strikers to disrupt mail delivery using a repertoire of tactics that included those developed by the Mabarete. Union leaders then rushed in offering to calm the situation and enter into negotiations. To an extent, insurgent unionism allows you to ‘have your cake and eat it’ in regard to rights and extra-legal power.
Nevertheless, such advantages that insurgent unionism might present should be placed in context; some level of oscillation between wildcat action and orderly negotiations is part of most industrial struggles. Being able to operate in and outside the law was never the objective of the Mabarete. What casual workers wanted was to ‘enter the House of the Post Office’. Or, in other words, they wanted to stop being employees of the labour brokers, a state of permanent casualisation, and become permanent employees of SAPO.
Once this was under way there was every reason to transform into a form of organisation more appropriate to their new classification. If registered unions had served the permanents well when the Mabarete were casuals, unions would serve the Mabarete well after they had become permanents. Of course, on occasions when DEPACU was unable to deliver, members of the rank and file would hark back to the days of the Mabarete and the effectiveness of their methods. Sometimes that call was partly heeded, as with the oscillating practices described above, but in general, as permanent workers they were better off with a recognised union representing their interests.6
Conclusion
The account of casual workers in SAPO adds to our understanding of the forms of struggles that precarious workers can undertake. Considerable academic effort has gone into exploring constitutional, or otherwise procedural, ways of resolving the power imbalances of precarious employment. These are valuable, but are not balanced by similar levels of scholarship documenting the full range of struggles taking place. There are understandable methodological challenges to such studies, but, as the work of Chinguno (2015), Von Holdt (2012) and others demonstrates, it can be done. This study of the Mabarete illustrates how violence can be used in different ways within industrial settings. In Chinguno’s study, violent solidarity was aimed at enforcing strike participation as a means to stop production. The Mabarete brought out only a minority of casual workers, but found a way of using violence to prevent SAPO conducting its business. The Mabarete struggle also demonstrates that there is no necessary link between classification struggles and them being pursued using symbolic or moral power. Rather, what has been outlined is an alternative form of industrial engagement when the institutionalisation of conflict is abandoned: subaltern worker rebellions to which covert and violent forms of disruption are integral.
The way in which the Mabarete’s struggle was sculpted spatially and morally by community geography and values has also been outlined. But their struggle was not about addressing social or class concerns. Rather, it was narrowly focused on advancing their position within the working class: from casual to permanent workers. Once this was achieved, they exchanged their subaltern worker rebellion for a seat within the industrial relations system. The transition phase (the 2014 strikes) illustrates insurgent unionism that combines features of both subaltern worker rebellion and the institutionalisation of conflict. The way in which the SAPO casual workers’ actions shifted, from the constitution, to the extra-constitutional and back, is worth reflection. It provides sobering analysis for those looking to new social formations to bring about a radical transformation of society. But it also should sound warning bells for those still committed to South Africa’s constitutional democracy.
The Mabarete struggles provide a lens to view conflict in South Africa (and elsewhere). A prominent feature of contemporary South Africa is increasing levels of violence and intimidation. The struggle waged by the Mabarete was clearly violent, although we need to contextualise this within a longer time frame that includes the protracted initial attempts to resolve their classification struggle within constitutional parameters. We also need to scrutinise what, at first sight, may appear to be intimidation, such as the processes of mutual mobilisation that have been described. A significant component of conflict is theatre, in which a range of props are used to create illusions. Nevertheless, despite this caveat, it is clear that the Mabarete conflict had scenes in which blooded bodies were real enough.
Framing intimidation and physical violence within a context of structural violence is important, especially in the case of precarious workers for whom proclamations of equality as citizens mean, in practice, little for them as workers. However, as illustrated in this case study, technologies of struggle, once developed, can be used by other groups. This happened with the tactics developed by the Mabarete being appropriated by other groups in the 2014 strikes. In the context of South Africa, such processes contribute to the increasing disjuncture between constitutional, democratic procedures and an expanding repertoire of violent forms of action. If we take a wider view of classification struggles to embrace all groups unhappy with their relative position in society, the vast and visible inequality in South Africa provides fuel for violent forms of social action as groups jockey with each other to get ahead.