Institutionalisation, strike violence and local moral orders.
Article Type: Report
Subject: Iron and steel workers (Demonstrations and protests)
Iron and steel workers (Civil rights)
Strikes (Case studies)
Violence (Case studies)
Author: von Holdt, Karl
Pub Date: 01/01/2010
Publication: Name: Transformation Publisher: Transformation Audience: Academic Format: Magazine/Journal Subject: Social sciences Copyright: COPYRIGHT 2010 Transformation ISSN: 0258-7696
Issue: Date: Jan, 2010 Source Issue: 72-73
Product: Product Code: 9918320 Strikes
Geographic: Geographic Scope: South Africa Geographic Code: 6SOUT South Africa
Accession Number: 228435793
Full Text: Introduction

Strikes were frequently accompanied by violence during the 1980s in South Africa. Strikers were regularly beaten, arrested and shot by the police. Strike breakers were intimidated, beaten and sometimes murdered by striking workers. Labour analysts ascribed the high levels of worker violence to the conditions under which trade unions organised and engaged in collective-bargaining during the apartheid era--in particular, the failure to fully institutionalise industrial conflict, and, more broadly, the absence of political rights which imbued industrial action with a strongly political dimension (Von Holdt 1989 and 2003, Webster and Simpson 1990). The implication was that with the political incorporation of workers into a post-apartheid democracy, and with the full institutionalisation of industrial conflict in new post-apartheid labour legislation, strike violence and the high levels of mass militancy which sustained it, would decline.

This has not happened. Strikes have increasingly been accompanied by heavy-handed police action--beatings, shooting with rubber bullets, arrests --while intimidation, assaults and murders of strike breakers have been a persistent feature of many large-scale strikes. This article seeks to explain why violence remains so much a part of industrial action, and what this tells us about the post-apartheid social order. It argues that several factors continue to undermine the institutionalisation of industrial relations and that, as in the 1980s, these factors range from those that are specific to the field of industrial relations itself to those that arise from the nature of the broader political, economic and social transition. These broader factors dispose workers to draw on the repertoires of collective violence that have been established as a tradition of conflict and struggle in South Africa in order to strengthen their collective-bargaining position vis-a-vis employers.

This article begins by summarising a case study of strike violence in a steelworks during the apartheid period, and argues that political struggle played a considerably more significant role in undermining the institutionalisation of industrial relations than previous analysis had suggested. It then turns to two strikes in the public service, one in 1992 and the second in 2007. The first of these took place at a time when public service workers had no institutional or trade union rights and when, beyond the industrial relations field, society was marked by the heightened instability and extreme violence of the period of negotiated transition following the unbanning of the liberation movements. The strike was marked by extremely high levels of brutal violence. The 2007 strike, in contrast, took place at a time when public service workers had full trade union and collective bargaining rights, and when a constitutional democracy had been in place for more than a decade. Nonetheless, despite conditions that appeared propitious to the institutionalisation of industrial conflict, the strike was characterised by considerable violent intimidation of non-strikers.

The question is why this should be so. Analysis of interviews with workers suggests two reasons: firstly, steps taken by the employer acted to undermine what workers understood to be their collective-bargaining rights, reducing the strike to a naked power struggle; and, secondly, and more importantly, workers articulate a somewhat inchoate but nonetheless strong sense of grievance about their location in post-apartheid society which legitimates recourse to familiar tactics of struggle, including repertoires of collective violence. Two conclusions can be drawn from this. Firstly, a common thread that runs through the apartheid and post-apartheid period is that the degree of institutionalisation of industrial relations is dependent on broader political and social factors beyond the field of industrial relations itself. Secondly, the question of the industrial and social order in post-apartheid South Africa is not a settled matter: the authority of the state and of the law has a limited reach, and social hierarchy and the balance of power between social forces that underpins it remains contested. This explains the high levels of conflict and violence not only in industrial relations, but in many spheres of South African society.

The article concludes with the argument that the limits to institutionalisation cannot be understood only with reference to the dynamics within the field of industrial relations, but that institutionalisation implies acceptance by all parties of the underlying social order, of which industrial relations is simply one component. Where the underlying social order is unsettled or contested, institutionalisation is likely to remain partial and precarious. Research into institutionalisation, then, has to go beyond the analysis of formal institutions and procedures, and investigate the informal networks and rationales that underlie this, and what these tell us about the broader social order.

Strike violence under apartheid

Webster and Simpson argue that the institutionalisation of industrial relations requires both industrial relations institutions that enable the trade unions to exercise the power of collective action according to mutually agreed procedures, as well as the broader political incorporation that Geary (1985) termed the 'constitutionalisation of the working class' (Webster and Simpson 1990). In South Africa under apartheid, neither of these conditions existed. Although labour legislation introduced in 1981 established trade union and collective-bargaining rights for black workers for the first time, these rights were deficient. In particular, the right to picket--crucial for unions to exert the moral pressure of the majority against non-strikers--was not only absent from the labour legislation, but also effectively banned by political legislation. Nor did the law give unions the right to strike, since legally striking workers were not protected from dismissal. Politically, black workers had no vote and no citizenship rights, like all other black South Africans. Notwithstanding the new labour legislation, employers were frequently hostile in the face of union activity, and political repression undoubtedly exacerbated industrial conflict, with the incidence of murder of non-strikers escalating after the State of Emergency was imposed in 1986 (Von Holdt 1989, Webster and Simpson 1990).

An ethnographic case study of union organisation, conducted at Highveld Steel after the apartheid period, provided greater insight into the factors which undermined institutionalisation during the apartheid period (Von Holdt 2003). Highveld Steel is a large steelworks employing some 5,000 workers located 150 km east of Johannesburg in Mpumalanga. In the 1980s it was the scene for a series of extremely militant strikes by the metal workers union, NUMSA, marked by intense contestation between management and workers as well as a violent conflict between different factions of workers.

In order to understand what was taking place inside the steelworks, it is necessary to have some information about what was happening outside of the workplace, in the communities surrounding the factory. In the early to mid-1980s the popular movement organised campaigns, marches, boycotts and stay aways. During 1986 confrontation became more intense as youths and activists targeted state property and black 'collaborators' for violent action. In response, the state made use of the national state of emergency to detain activists, ban organisations and launch frequent armed patrols through the township. Formal organisational structures became inoperable, and the struggle shifted to frequent clashes in the streets between armed police and stone-throwing youths, attacks on black collaborators such as policemen and town councillors, and attacks on government vehicles. Two off-duty police officers were killed, and councillors' houses were petrol bombed. The popular movement conceptualised its strategy as 'rendering apartheid unworkable or rendering ourselves ungovernable' and 'establishing people's power'. Although the security forces of the state were able to regain physical control of the township, the machinery of apartheid administration and coercion was broken, and rents boycotts bankrupted the local administration (Von Holdt 2003:90-96).

The nature of this broader political struggle shaped trade unionism in the workplace. The tactics of 'ungovernability', violent confrontation and intimidation that characterised community struggles were translated into wildcat strikes, symbolic confrontation and the subversion of order and institutional procedures in the workplace, generating a dynamic of disorder and violence. In part the prevalence of wildcat strikes and the flouting of procedure arose from the fact that the workers' rights established by labour law did not automatically translate into rights on the shop floor; the new procedures and rights remained highly unstable and fiercely contested throughout the 1980s; and this contestation was charged with racial tension:

'We were in war with management ... Our members did not even know whether they had rights or not, and the bosses were not happy about the rights extended to our members. We had to fight for the recognition of the rights that we already had, which were not recognized'. (Von Holdt 2003:76-77)

But the shop stewards' practices were deeply influenced by the broader political struggle. Union activists consciously rejected the tendencies towards institutionalisation which are inscribed in union recognition and negotiation:

'... the negotiations were not strictly speaking negotiations. Although it was unavoidable to articulate our demands, and put them before management for discussion, our approach was one of confrontation ... We would put forward demands and strike the following day. We were reluctant to use the Industrial Council or the procedures of the Labour Relations Act. We strike whether it's legal or illegal ... We were not only challenging the individual employers, we were even challenging the state'. (Von Holdt 2003:121)

The symbolic charge of challenging white control of the shop floor was potent. Sometimes white managers who tried to perform the tasks of striking blacks in dangerous areas of the works were injured. According to one shop steward:

'Then people are laughing, saying it's nice to see these white guys being involved in accidents. We have been slaves for such a long time, we want to see what happens when we strike ...'

This experience of collective power was more important than the instrumental ends of collective bargaining:

'Even if we are not paid for three days we don't care, now the whites can see what is going on, so they won't be able to come and talk the way they like'. (Von Holdt 2003:123)

The union was subject to contradictory tensions between the pressure to negotiate workplace order and procedures with management, and the pressure of radical militancy that rejected a workplace order which, like apartheid more generally, was structured by racial domination. This tension was deeply embedded in the internal structure of the union, and took the form of eventually violent conflict between the shop stewards, who were involved in negotiating and making daily use of workplace procedures, and the more militant strike committee. But even the shop stewards, involved in procedure and negotiation as they were, did everything they could to avoid recognising the legitimacy of white management. On one occasion they engaged in a lengthy battle with management over re-drafting the disciplinary procedure, and when management had finally conceded to their demands, they refused to sign, on the grounds that they did not recognise the right of management to punish workers and that they would refuse to be bound by the disciplinary procedures (Von Holdt 2003:142).

Integral to the intensity of workplace struggles was a degree of coercion in maintaining the solidarity of union members. The shop stewards established a 'strike committee' who would 'be our police' (Von Holdt 2003:129) for the purpose of enforcing discipline and punishing nonstrikers, and to begin with this appears to have had widespread legitimacy among the membership. The strike committee became known as the 'sjambok committee', and its purpose was to enforce the 'union law'. As strike action escalated in the workplace coercion intensified and dozens of workers were assaulted, sometimes severely. Members of the strike committee saw themselves as agents of order--'police'--and violence was an intensely ordering activity. However, an increasing number of workers, led by the shop stewards, objected to this violence and the fact that strike action was no longer discussed by the entire membership, but mandated by the strike committee, and eventually the union split.

This case study demonstrates how remote the prospect of institutionalising industrial relations under apartheid really was. Both managers and shop stewards had deeply ambiguous attitudes to the procedures through which worker and trade union rights were operationalised on the shop floor, although shop stewards in particular became master tacticians at making use of these procedures to build their organisational power and protect their members. Underlying the formal codes of procedures and agreements was a hostile struggle for power and control deeply shaped by racial and political dynamics. Informal rationales and an informal structure, the strike committee, were important factors in the disruption of agreements and procedures. Recognition of legal rights and workplace procedures implied recognition of management--and in the eyes of many worker activists management was an agent of the profoundly illegitimate apartheid regime.

The study demonstrates that the political factor, in the form of the national liberation struggle, was more significant in undermining processes of institutionalisation than the earlier studies acknowledged, because of the way it framed and gave meaning to union organisation in the workplace. And for a union organisation engaged in this kind of struggle, coercion and the violent imposition of collective discipline were legitimated by the notion of a war against apartheid--albeit that the extent and occasion of violence was subject to contention, a contention that became so sharp in the case of Highveld Steel that the union split in two, with leading shop stewards arguing that violence had no place in a trade union and simply subverted democracy.

Strike violence in the transition: the case of the public service

This section of the article and the next compares the strike violence that took place during two public service strikes which took place 15 years apart. The comparison is limited to the dynamics of the strike in the public hospitals, although in the later strike the withdrawal of labour took place more widely across the public service, and particularly in the schools. The first, which took place in 1992 in the middle of political negotiations over the constitutional future of South Africa, was marked by the absence of any institutional rights or procedures for public service workers. The rights won by industrial workers, and established in the Labour Relations Act of 1981 had never been extended to the public service. The apartheid state refused to recognize the unions which attempted to organise black public service workers in the last years of the apartheid regime. Black general workers, such as cleaners, porters, ward assistants and kitchen staff worked under particularly oppressive conditions: permanently classified as casual workers, they could be dismissed instantly, were not eligible for pension benefits, and women workers, when pregnant, had to resign their jobs.

The strike took place at a time when waves of violence were rolling over South Africa, with the region around Johannesburg one of its epicentres. From 1990--from, that is, the moment of the unbanning of the liberation movements and the release of political prisoners--a series of massacres were perpetrated by a 'third force' in townships and on commuter trains, which was clearly intent on undermining and weakening the ANC, and perhaps destabilising the national negotiations over the outlines of a new democratic order. The negotiations themselves were tense, and marked by a series of stand-offs between the ANC and the National Party government. Much evidence points towards the 'third force' being linked to Inkatha, the Zulu ethnic nationalist organisation which was based in KwaZulu Natal and the migrant worker hostels of the Witwatersrand, and which had long been embroiled in violent conflict with the popular movement and the ANC, and being sponsored from within the apartheid state, whether with the tacit support of its national leadership or by rogue security force elements is not clear. For the National Party and the 'third force' the prospect of ANC-aligned unions strongly organised within the state was likely to have been an alarming one.

The National Education Health and Allied Workers Union (NEHAWU) was launched in 1989 through the merger of several different organisational initiatives over the previous few years. It was only able to win recognition through a wave of strikes that gained an added impetus from the unbanning of the ANC in 1990. The most important of these strikes nationally, and the one which finally won them recognition, took place in 1992, and started at Chris Hani Baragwanath Hospital It spread to 32 other hospitals in the old Transvaal province after the Chris Hani Baragwanath shop stewards were arrested early in the strike, and from there to public service workplaces in the rest of the country.

Workers who were in the leadership of the union at the time described the strike as a 'do or die' moment. Amongst themselves they discussed the fact that previous strikes had failed to achieve anything, and that at this time they would strike for as long as it took: Sibindi uyabulala, sibindi uyaphilisa ('Courage can make you do brave things; it can also make you do stupid things') , as one put it. The strike started at Chris Hani Baragwanath Hospital with a series of demonstrations culminating in a mass meeting at the recreation hall where workers decided to strike. The hall was surrounded, some say by the army, others say the police, tear gas was fired into it, and the shop stewards were arrested.

The strike was an extremely bitter and protracted one, lasting for almost four months. In the Transvaal six shop stewards and the husband of another shop steward were shot and killed by unknown assassins during the strike and in its aftermath, and at least six strike breakers were also killed (Fenichel 1992, NEHAWU nd). Quite early in the strike government employed strike breakers from the migrant worker hostels in the Witwatersrand to take the place of the strikers. The strike breakers were Zulu-speakers from KwaZulu Natal, and shop stewards believed they were part of the 'third force'. It was these strike breakers, they believed, who were assassinating shop stewards.

A clandestine committee was formed by workers from different hospitals to develop informal strategies to strengthen the strike across Gauteng. (2) Information was gathered about strike breakers, and strike breakers from a particular hospital would be attacked by workers from a different hospital, so that they could not be recognised. Petrol bombs and guns were used in some attacks:

Informants provided an example:

Workers in general seemed supportive of the strategy:

Non-strikers were also assaulted at taxi ranks and shebeens, and it seems likely that not all violence was coordinated through the clandestine committee. The majority of strikers were support workers, and there was considerable intimidation and assault of nurses. Asked where the ideas about using violence had come from, a woman worker explained:

Just as in the case of Highveld Steel, for these workers the trade union struggle was closely identified with the political struggle for national liberation, and the quasi-insurrectionary strategies developed within the ANC and the popular movement provided a reference point for the use of targeted violence against state property and agents as well as 'sellouts'. This violence was necessary in order to make the negotiating breakthrough that had eluded them in previous strikes:

The strike was successful. The union was recognised, and general workers won their demand for employment rights. This victory opened the way for inclusion of the public service in the new post-apartheid Labour Relations Act of 1996. Some 150 of the strikers accused of intimidation remained suspended from work until the conclusion of the disciplinary process. The strikers let it be known that anyone who gave evidence against them would be targeted:

Strike violence was not, however, only instrumental in the struggle to win negotiations. As in the Highveld Steel case study, to workers who worked under oppressive conditions and who had struggled for several years to win recognition, the strike and the violence associated with it provided a heightened sense of agency and power, and of connection to the broader political drama playing itself out around them during the transition. It was clear that the interviews evoked strong memories of an intense and dangerous time when they took their lives in their hands. In group discussions participants voices were at times loud and excited, and laughter marked the experiences as shared ones:

In these moments workers were recalling practices that bore no resemblance to the routines of institutionalised conflict resolution. However, in quieter moments some individual interviewees remembered the violence as having been deeply painful.

The 1992 strike provides an extreme case of the dynamics already identified in the wildcat strikes of Highveld Steel under apartheid. During the numerous strikes at the latter no workers were killed in strike violence, although many were severely beaten and at least one shop steward's house was petrol bombed. In contrast, over the four-month hospitals strike some 13 shop stewards, family members and strike breakers were killed across the Transvaal. The difference can be explained by the fact that, on the one hand, there were no organisational or institutional rights or procedures through which the conflict could be mediated, and on the other, the strike took place at a time of heightened political contestation and uncertainty regarding the outcome of negotiations, and was accompanied by extremely high levels of violence in the communities around the hospitals.

There are similarities, though. In both cases informal workers' committees were established with a mandate to make use of violence to strengthen the strike, in contrast to the role of the shop stewards committee, which was to lead the strike at a formal organisational and public level, and engage in negotiations with the employer. In both cases strike violence was infused with the practices and meanings attached to the broader political struggle for liberation. The repertoires of collective violence were by now deeply embedded in the social structure of the trade unions. The birth of NEHAWU was a particularly violent one, deeply marked as such in the memory of the union and its members.

Strike violence and the post-apartheid order

By the time of the 2007 public service strike, the full range of employee and trade union rights enshrined in national labour legislation applied to public service workers. Trade unions had the right to access members, represent them, organise, strike legally without dismissal and picket. Collective bargaining was institutionalised across the public service. Public service unions had overtaken many industrial unions in size, and NEHAWU was the biggest affiliate of COSATU. The negotiated transition had produced a constitutional democracy and the former liberation movement, the ANC, with which COSATU had a structured alliance, governed with a large majority. In other words, all the necessary conditions for the 'institutionalisation of the working class' appeared to be in place.

Nonetheless, like a growing number of strikes through the mid-2000s, the public service strike was accompanied by considerable violent intimidation. This section of the article provides a brief overview of the strike, based mostly on participant observation at Chris Hani Baragwanath Hospital in Soweto, before exploring in some detail the views of worker activists regarding the strike violence. (3)

An alliance of public service unions launched the strike in mid-2007 in support of demands for a substantial real wage increase. The strike turned into a protracted test of strength and will between government and the unions, and lasted four weeks. At the level of industrial relations the right to strike became a matter of dispute in the public health sector, with government arguing that it was defined as an essential service in labour legislation and obtaining a court interdict against striking in the hospitals, which enabled the government to declare the strike in these workplaces illegal. The unions argued that this was an underhand tactic, as the legislation did not envisage all health sector workers being defined as essential service workers, but obliged the employer to meet with the unions in order to negotiate the scope of 'essential services' in the health sector. This had not been done. This dispute over the quite fundamental issue of the legality of the strike, at least in the hospital sector, qualifies the extent to which industrial relations procedures can be regarded as fully institutionalised, as will become evident below.

Politically, too, the strike was charged with considerable tension. Firstly, the ANC government under Thabo Mbeki had adopted a tough position on the public service unions, generally negotiating aggressively as well as imposing a programme of downsizing in the mid- to late-1990s. A bruising encounter for the unions had been their defeat in the 1999 strike, which was undermined by disunity and fizzled out, allowing the government to impose a unilateral wage increase. Secondly, a tacit political aim was to demonstrate labour's strength in the context of a leadership battle in the ANC, in which COSATU was allied with a range of forces committed to replacing President Thabo Mbeki with Jacob Zuma. For labour the stakes in the 2007 strike were therefore high, both in terms of the imperative of reasserting itself in the public service and in terms of the struggle against a 'neoliberal' leadership of the ANC.

The strike started on June 1, a Friday, with strong support, particularly in the public health and education sectors. On that day the heads of all hospitals were instructed to hand the unions a copy of the court interdict prohibiting essential service workers from striking. On the Sunday, government announced that any essential service workers still on strike by 10am on Monday would be dismissed. On Monday morning the Army medical corps was drafted into hospitals to run the kitchens and provide cleaning services. At Chris Hani Baragwanath Hospital, as the 10 am deadline approached, groups of strikers toyi-toyied between the wards, pulling out any nurses or support workers who were still working. On the Wednesday government announced that picket lines would be removed from hospitals and other institutions, on Thursday the COSATU general secretary addressed a rally of strikers at Chris Hani Baragwanath--always seen as one of the key strategic workplaces in the health sector--and called on workers across the country to shut government down, and on the Friday the police moved the picket line out of the hospital. Thus the first week of the strike was characterised by each side ratcheting up the pressure on the other. While government made use of legislative provisions and the court interdict to declare the strike illegal, the unions took a decision to ignore the interdict and rely on collective action to support their demands. Both parties continued to mobilise their resources over the following week.

On the second Monday, the 11th day of the strike, dismissal notices were posted in every hospital, listing between 30 and 40 strikers from that hospital who were declared to have been dismissed in light of the illegality of the strike. Every day a further 30 to 40 strikers per hospital were listed as dismissed. On the Wednesday, COSATU mobilised a national general one-day strike in support of the public service unions, an unprecedented display of solidarity for striking affiliates from the Federation as a whole. This was the high point of the strike. Over the following days the number of strikers gradually diminished in the public health sector and government administrative offices, but it remained extremely solid in the schools. The strike was sufficiently strong for the unions to hold out and force government back to the bargaining table with new offers. Eventually the strike was settled after four weeks, with government having revised its wage offer from 5.3 per cent to 7.5 per cent, and agreement to increase salaries by a further 1 per cent above inflation in 2008. There were also substantial increases in housing and medical aid benefits, and agreement to fast track the 'occupational specific dispensation' (meaning increased salaries) for underpaid and scarce-skill categories such as doctors, nurses and teachers. Although the increase was significantly below the 10 per cent demanded by the unions, they claimed the result as a victory for public service unions. (4)

Intimidation started even before the strike had begun. Two days before the strike groups of men appeared in both the adult and the paediatric ICU in the Burns Unit at Chris Hani Baragwanath Hospital, and forced the nurses on duty to accompany them to a meeting. Fortunately, two nurses were able to conceal themselves and remain behind, so averting a potentially hazardous crisis. The men were unknown, and there was speculation that they were from other hospitals or schools. By the middle of the first week of the strike there were increasing reports of intimidation of non-strikers, particularly nurses. On the Friday after COSATU's call for government to be shut down nurses in the surgical wards panicked and fled the hospital after a barrage of SMSs and phone calls to the wards threatened that their homes and children were known. Doctors and matrons spent the rest of the day preparing patients for removal to private hospitals, and by the end of the night most of the wards had been closed. By the Monday when the first dismissal notices were posted, strike violence was extensive in hospitals and in schools, as well as in communities where non-strikers lived, mostly in the form of verbal threats and assaults.

It is important here to draw attention to the specific character of the labour force and unionisation in the hospitals. The labour process combines a high labour intensity of high-level professional skills with a diversity of less skilled non-professional support workers. Thus at Chris Hani Baragwanath Hospital there were at the time of the strike almost 2,000 nurses and several hundred doctors, and perhaps 2,000 support workers, including those directly involved in the clinical process, such as clerks, cleaners and porters, as well as indirect support in the form of kitchen workers, laundry workers, drivers and, again, clerical workers. NEHAWU is most strongly organised amongst support workers, and to some extent among nursing assistants. Nurses are organised by DENOSA, a COSATU affiliate, and by non-COSATU union, HOSPERSA, which also organises clerical staff. While NEHAWU is the majority trade union, its base is among the militant and relatively unskilled support workers, not the professional workers who are central to the clinical labour process. The professional ethos of nurses means they are reluctant to strike even when strongly aggrieved, and strikes in the public health sector are generally led by NEHAWU and characterised by often bitter divisions between workers and unions. This also works against institutionalisation of industrial relations, since, while DENOSA and HOSPERSA prefer to make use of institutional channels and procedures to assert their professional concerns, NEHAWU and its members tend to make use of collective action to overcome a sense of marginalisation. Intimidation of non-strikers may be seen as the only way to disrupt the core labour process of the workplace.

The 2007 strike was remarkable in that it was a joint initiative of most unions in the public service, both COSATU and non-COSATU. This held out the possibility that intimidation might be greatly reduced. However, a few days into the strike the support of DENOSA and HOSPERSA became increasingly lukewarm and, in practice, ceased supporting the strike. Once again, NEHAWU and the support workers were on their own.

In interviews some shop stewards and worker activists referred specifically to their sense that the employer had flouted legal and institutional procedures by declaring the strike illegal and dismissing workers:

But most of them focused on the fact that violence is intrinsic to strike action, drawing on their own experience and:

Workers, then, are disposed to use violence in strikes because it is intrinsic to the meaning of the strike and because it has proven its effectiveness. Violence is a necessary element in maintaining the solidarity of the union and strengthening negotiations. A strike is a power struggle and has its own morality, which is not affected by the fact that South Africa is now a democracy with the former liberation movement in government. In the words of a woman worker:

The use of violence explicitly breaks the law, but is justified by a different law, the law of the majority in the union; where they clash it is difficult to remain 'upright':

The authority of the law and the state has its limits when confronted by the imperatives of strike action. According to a woman worker:

What emerges from these interviews is that the collective action of a strike carves out a domain with its own laws and rationales, and its own codes of conduct, beyond the authority of the law or the moral codes that govern 'normal' individual interactions in society. These rationales and codes work against tendencies towards institutionalising industrial conflict. Several of the interviewees referred to broader social and economic injustices to explain their readiness to see the law and the authority of the state put aside; in the words of a woman worker:

There will be no peace until the revolution has been completed:

The position of workers in society is cause for anger, especially when compared to their aspirations and the promises made to them during the liberation struggle:

Democracy in itself is inadequate because it does not improve the lives of workers:

Thus broader socio-economic conditions and the continuing social injustice that workers have to bear frame the specific rationales and codes of the strike and of union organisation, providing a broader justification for strike violence.

Part of the explanation for the durability of strike violence and the continuing transgression of the laws and procedures through which conflict is institutionalised, lies at the subjective level, with the sense of empowerment and agency that workers who occupy dreary repetitive jobs experience when participating in collective action. This was evident both in the Highveld Steel case study and in the 1992 public service strike. Again, in 2007, a woman worker describes the sense of liberation and agency through taking part in strike violence:

Again, this feeling is associated with the special morality of the strike: when workers meet a strikebreaker with bags of groceries, even though they do not have food at home 'we just throw it into the streets for cars to smash it, we kick it, we will take nothing'.

This special morality of the strike is conveyed by one of the stories a shop steward told about the strike. A nursing assistant dressed up in bulky clothes and used ash to alter her complexion, and participated in the sjambokking of nurses at the taxi rank. The following day, seeing the nurses she had beaten toyi-toying and singing on the picket line, her comment was: 'See my patients over there--see my medicine has cured them!' In this inversion of the normal meanings of patient and medicine, non-striking nurses are regarded as ill, and beating with a sjambok is a medicine which cures them and brings them out on strike.

Collective action and violence is associated also with the visceral sense of oppression and marginalisation which the least skilled workers experience, particularly in a skill-intensive institution such as a hospital:

It is not that striking in the health services does not pose moral dilemmas. Indeed, it does.

This puts nurses, specifically, in a difficult position:

In the moral tension between the claims of patients and the duty nurses feel to them, and the moral imperatives of worker solidarity, solidarity comes first. However, as with the 1992 strike, an individual reflecting on her role in strike violence may feel a strong sense of guilt about actions that are unavoidable to those who have to act in the real world, on 'earth':

The local moral order constituted by collective action in the form of the strike, and by the union as the agency through which the strike is organised, should not be regarded as a coherent and all-encompassing counter-order of the same kind as the counter-order constructed by workers and community activists during the struggle against apartheid in the 1980s. Firstly, the 2007 strike remained limited to economic trade union demands, and never posed larger questions about the nature of the socioeconomic order. Secondly, the sense of the unfairness of the emerging post-apartheid socioeconomic order remains relatively inchoate, in contrast to the clear demands for democracy and the end of apartheid in the 1980s.

Thirdly, while a core of activists among the workers clearly regarded intimidation and strike violence as a legitimate practice, it is not clear how widely this view is shared. As we saw in the Highveld Steel case study, even at the height of the struggle against apartheid, for many workers there were limits to the amount of violence they found acceptable and were prepared to support, and the union split on this issue at the factory in the late 1980s. It is certainly true that in 2007 the majority of nurses rejected the idea that collective solidarity took precedence over their professional ethos of patient care, even while they supported the goals of the strike. Nonetheless, these practices had sufficient support for the union to retain its support and even increase its membership in the course of the strike.

Finally, it is important to avoid imputing to interviewees the kind of systematic worldview that sociologists attempt to construct through the process of theorising about social phenomena. The local moral order that cohered may be a quite transient construction, one that emerges during episodes of collective action and then becomes dormant again, rather than an ongoing stance in relation to state authority and moral codes. While the counter-order constructed in the struggle against apartheid may have shared some of this flexibility in its significance for the daily living of many individuals, it nonetheless seems to have been a much more durable and encompassing order, posing a real alternative to the apartheid order imposed by the state.

Concluding discussion

The interviews with workers who took part in the 2007 strike lay bare a number of factors that mesh together to propel unprocedural action and strike violence. Some of these factors belong to the industrial relations field, while others arise from the broader political and socioeconomic order beyond the industrial relations field, and some straddle the two. Located in the field of industrial relations are the disputed nature of some of the procedures themselves, the relatively weak bargaining position of the support workers and their union, and the historical repertoire of practices that are experienced as intrinsic to striking. Located in the broader socioeconomic order are continuing social and economic injustice, bolstered by the sense that the liberation struggle has not achieved its full ends, and the limited reach of the authority of the law and the state. Straddling the two are the sense of an alternative local moral order which coheres around the collective decision of the majority, and the subjective experience of agency and empowerment fostered by going on strike, transgressing laws and procedures, and deploying violence against non-strikers.

It is clear from this that industrial conflict is only partially institutionalised in post-apartheid South Africa. On the one hand, trade unions and employers make regular, indeed daily, use of the institutions and procedures for resolving conflict between individual employees and employers, as well as for collective dispute resolution and collective-bargaining. On the other hand, procedures may be disputed or regarded as unfair, and collective action not infrequently involves violence and intimidation, including violence by the police. Explaining the partial quality of institutionalisation needs to take account both of the industrial relations factors as well as the factors located in the broader society.

In the field of industrial relations

Despite the fact that post-apartheid labour legislation was shaped by extensive consultation and at times tough negotiations between labour, business and government, the fairness and meaning of some of its provisions remains contested between unions and employers. This was clearly evident in the dispute over essential services and the legality of the strike in the hospitals in the public service strike. Declaring the strike illegal fuelled trade unionists' determination to use illegal methods themselves, firstly by ignoring the interdict, and secondly by stepping up intimidation.

More important, however, is the fact that many sectors of the labour force are denied an effective voice even though they have formal access to substantial trade union rights. To some extent this was the case in the public service strike. Semi-skilled and unskilled support staff have full access to trade union rights and have seen their basic wages increased substantially in the years of democracy; nonetheless, they do not form a majority of hospital staff and feel marginalised in an institution that values the ethos and voice of professionals relatively highly, and even though they form a majority of the unionised workforce, their impact on the clinical labour process is relatively marginal. Striking, and using violent and disruptive strike tactics, is one way of making their power felt, as the 'illiterate' worker quoted above makes clear, and of course the subjective sense of agency becomes even more important in these circumstances.

Marginalisation and disempowerment are experienced by workers in many sectors of the economy. Indeed the employers' strategy of casualising and outsourcing workers is explicitly designed to undermine union power and reduce workers' access to labour legislation; this effectively denies large sections of the workforce substantial access to employee and trade union rights, despite the existence of these rights at a formal level (von Holdt and Webster 2005). In these circumstances, industrial relations cannot be institutionalised. One would expect workers in this position more readily to resort to violence in order to enhance their weakened bargaining position, and indeed much of the violence in the past few years has taken place when relatively vulnerable workers have engaged in strike action--such as in the security guards' strike of 2006, which was the most violent strike in the post-apartheid period, in which 57 people died (Ehrenreich 2007, Jansen 2006, Makgetla 2006), and in the construction sector strike in 2009. The growing tendency for heavy-handed and violent policing of strikes--which was particularly obvious during the security guards strike--is a factor which straddles the industrial relations field and broader political and social factors. In the field of industrial relations, it suggests an underlying failure on the part of the state to accept the institutionalisation of industrial relations enshrined in legislation, since police action frequently violates the rights to gather, picket and march.

There is, in other words, a persistent ambivalence towards institutionalising industrial relations procedures on the part of unions, employers and the state. At the COSATU 10th congress, held in 2009, the secretariat report criticised the use of violence and the trashing of streets in strikes, arguing that this delegitimised the strike weapon in the eyes of the public. Many of the affiliate leaders were clearly uncomfortable with this view, and argued that provocative and brutal police action and the use of scab labour to undermine strikes drove workers to use such tactics. These rationales illustrate that industrial relations procedures and the industrial relations order (employment of scabs, the actions of the police) remain matters of deep contention.

The institutionalisation of industrial relations implies an acceptance of the underlying industrial order and the balance of power between workers and employers. The ambivalence of all parties to industrial relations institutions expresses an underlying contestation over the industrial order and the relative position of workers and employers in it.

Beyond the field of industrial relations

But strike violence also takes place where workers occupy a relatively powerful position in the labour market and are not subject to the kind of de facto exclusion or marginalisation described above. During the public service strike in 2007 there were persistently high levels of intimidation and violence in schools, often directed against school principals, despite the fact that the strike was extremely strong and that teachers can in no way be described as vulnerable or insecure workers, but have a similar professional status to nurses. Municipal workers in their national strike in 2009 engaged in widespread disruption and trashing of towns and cities, despite occupying a relatively strong labour market position with full access to the rights enshrined in labour legislation (in this case police 'provocation' in initiating violence was cited by trade union leaders). In order to understand violence in these conditions, one cannot rely on explanations that referred only to problems within industrial relations itself, but has to explore the factors beyond this field.

Here the prevailing sense among organised workers, reflected in the interviews, that the current social order is unjust, a view which is fuelled by the legacy of radical rhetoric that was current during the anti-apartheid struggle, has to be the pre-eminent factor. The struggles COSATU has engaged in to neutralise the 'neoliberal' leadership of the ANC and shift its policies towards the left, and its calls for a more socialist orientation, reflect the same sense that the legacy of apartheid continues to blight workers' lives. This stance implies a lukewarm and conditional acceptance of institutions that may be held to buttress an unfair social order, including the institutions and procedures that define the industrial relations field. The outlines of the post-apartheid social order are, therefore, provisional and contentious.

Such attitudes are shared quite broadly in the trade union movement. It is not only COSATU and organised workers that engage in this kind of contention; virtually every matter of public policy, public morality and public interest--urban policy, land policy, public transport initiatives, policing, housing policy, the meaning of HIV/Aids--all are contentious, all are matters of dispute, and involve interest groups in heated argument, conflict and, sometimes, violence. The relative power of different interests, hierarchies, authority, winners and losers, are subject to contestation, negotiation, definition and redefinition. The nature of South Africa's society and the post-apartheid social order are not settled matters.

Partly arising from this, and partly exacerbating it, is the limited reach and authority of the state. As Tilly points out, the scope for contentious politics and collective violence is expanded in societies characterised by low-capacity states. Where the state has limited powers to enforce the rules of the game and punish transgressors, non-state actors are able to push back the limits of contention and transgress laws and deploy violence with impunity (Tilly 2003:41ff). This description can be applied quite convincingly to South Africa. In the public service strike the unions and their members flouted the court interdict against illegal striking with impunity, and no workers have been charged with intimidation or violence. Nor has anyone been charged with any of the murders arising out of the security guards strike. There is, in other words, very little cost attached to subverting the institutionalisation of conflict. The violence of the police in response to legal and peaceful strikes--as well as to other conflicts, such as service delivery protests in communities--is an aspect of the same problem: where order and the authority of the state are continually placed in question, the police themselves tend to behave in an unpredictable and even arbitrary fashion. Uncertain of their authority, they too can transgress the law and the bounds of legitimacy, frequently with the same impunity as their adversaries.

These factors beyond the field of industrial relations heighten the salience of the factors that straddle industrial relations and the wider social order, and serve in some ways to connect the wider disorder to industrial disorder: limited state authority can be eroded or supplanted by local moral orders which legitimate alternative authorities, in this case, the authority of the collective decision made by union membership which is buttressed by the sense that the prevailing order is unjust and unsettled. Likewise, the subjective potency of the experience of agency fostered by industrial action or collective violence is enhanced by the sense that it is part of a wider contestation. It is also the wider reality of contestation and instability that heightens contestation within the industrial relations arena, not only over the ends prescribed by the industrial relations system, but over the very legitimacy and acceptability of elements of the system itself.

As was the case in the apartheid era and the transition, the broader dynamics of South African society have a profound impact on trade unions and industrial relations. Although there is a greater degree of institutionalisation than before, it is only partial, and it is likely to remain in this state for as long as the broader social order remains unsettled and contested. The institutionalisation of a specific field like industrial relations depends on an acceptance by all parties of the broader socio-economic order and the balance of power it stabilises in society--what is suggested by the idea of 'constitutionalisation'. The 'constitutionalisation of the working class'--as well as the broader constitutionalisation of society--proves to be a more contradictory and precarious process than imagined.

References

Ehrenreich, T (2007) 'Interview: security strike violence not in keeping with COSATU's values', South African Labour Bulletin 30(5).

Fenichel, J (1992) 'Most NEHAWU workers to be reemployed', South African Labour Bulletin 16(8).

Hassen, E-K (2007a) 'Crouching comrades, hidden dialogue: interpreting the public service strike', South African Labour Bulletin 31(3).

--(2007b) 'Public service strike: what was finally agreed?', South African Labour Bulletin 31(4).

Jansen, M (2006) 'Security strike violence: union leadership found wanting', South African Labour Bulletin 30(4).

Makgetla, T (2006) 'Security strike: why the violence?', South African Labour Bulletin 30(3).

NEHAWU (no date) 'NEHAWU history, the unfinished story: the history of the National Health Education and Allied Workers Union 1887-1997'. Johannesburg: NEHAWU.

Pardesi, S (2007) 'Interview: negotiations in the public sector--what went wrong?', South African Labour Bulletin 30(3).

Tilly, C (2003) The Politics of Collective Violence. Cambridge, New York: Cambridge University Press.

Von Holdt, K (1989) 'Violence in the workplace: introduction and overview', South African Labour Bulletin ('Special focus on violence in the workplace') 14(3).

--(2003) Transition from Below: forging trade unionism and workplace change in South Africa. Pietermaritzburg: University of Natal Press.

--(2007) 'Strikers, professionals and patients: negotiating essential services on the ground', South African Labour Bulletin 31(5).

Von Holdt, K and E Webster (2005) 'Work restructuring and the crisis of social reproduction: a southern perspective', in E Webster and K Von Holdt (eds) Beyond the Apartheid Workplace: studies in transition. Pietermaritzburg: University of KwaZulu-Natal Press.

Webster, E and G Simpson (1990) 'Crossing the picket line: violence in industrial conflict--the case of the Afcol strike', Industrial Relations Journal of South Africa 11(4).

Karl von Holdt

karl@yeoville.org.za

Notes

(1.) A rough draft of this paper was presented at the International Sociological Association (ISA) 1st Forum of Sociology, Barcelona Spain, 5-8 September, 2008, and at the Comprehending Class Conference, University of Johannesburg, organised by the Centre for Sociological Research, University of Johannesburg, and the History Workshop, University of the Witwatersrand, 23-26 June, 2009. I would like to thank participants at both conferences, as well as colleagues at SWOP and Wits University, for comments on earlier drafts. I would also like to thank Jacob Dlamini for assistance in translating Xhosa/Zulu proverbs.

(2.) The information presented here is culled from interviews and discussions with numerous workers and shop stewards across Gauteng. The author was unable to interview anyone who had participated personally in the committee.

(3.) At the time the author was an adviser to the Surgical Division Transformation Project at Chris Hani Baragwanath Hospital. As with the 1992 strike, the views of workers were solicited through interviews across Gauteng. (See also Hassen 2007a and 2007b, Pardesi 2007, Von Holdt 2007).

(4.) This article does not claim to assess the effectiveness of the public service strike as industrial action, but rather to investigate the strike violence which accompanied it.
They used to go and destroy. The army was surrounding the hospitals
   so they couldn't get access. They went to their homes. Homes were
   destroyed and people have been killed. Many people died. There was
   a third force, and some of our shop stewards died. They were
   assassinated. Shop stewards and known activists didn't sleep in
   their homes, there were strange cars parked outside watching.


The spirit of workers becomes cruel even if they are wearing
   Christian clothes. One of the matrons said on TV that the strike
   can continue, the hospital was okay, they didn't need the strikers.
   Within an hour her house was bombed, even though the police were
   watching. We told her she must resign, and she did.


You just feel good when the enemy is shaking, you hear the news and
   you say good. You don't ask who did it. Workers were targeting a
   monster at the time, so afterwards they would just say mission
   accomplished. Otherwise we would still be in darkness.


I was involved in politics, I was involved in the community, I was
   a member of the UDF (United Democratic Front). I started working at
   the hospital, I was involved in the strike in 1985. I was part of
   that school of thought where we had strategies for fighting
   government, we had all the ideas of how to use force if necessary,
   we were taught those things, that we could never submit. Meetings
   took place under duress [from the police], so we knew how to fight.
   I nearly went to Robben Island.


You know what is frustrating? You are on strike. Others are
   comfortable, they are eating. But they will also benefit even
   though they are afraid. We called ourselves the hard-hearted.

   All tactics were to make the negotiators achieve--we didn't do it
   for fun, but for frustration.

   Few people would not achieve, we needed many--therefore force was
   required. It was necessary in order to achieve what we have
   achieved.


It worked, the story spread in the hospital, and people saw if you
   give evidence you will die, they will burn your home and hurt your
   children. It was successful, because there was no evidence.


When we strike we kicked everything, like patients' food, we
   destroyed everything. They called the army in to run the kitchen
   etc. We called it 'days of chaos'. But strikes are enjoyable. It's
   enjoyable, it is bad, it's exciting. I remember escaping in the
   boot of a car, after I found myself next to a meeting where I heard
   they were planning the assassination of Oupa, one of our leaders. I
   told him but he did nothing and indeed he was assassinated. It was
   painful but exciting.


All proper procedures were followed, but they declared the strike
   illegal, stopped our salaries, served us with dismissal letters.
   That is where everything broke out of proportion. We decided we
   were going to smash everything.

   The LRA and the Constitution seem not to bind the employers--they
   can do anything, they don't follow procedure, but the unions are
   expected to follow procedures.

   They are above the law, the seniors in government break the law.


Since I was born, I have seen all strikes are violent. There are no
   such strikes as peaceful strikes. Some workers do not join a strike
   because of fear. We must develop a mechanism for all these workers
   to participate. By force they must join the strike. Otherwise
   anybody would do their own thing.

   You know you are holding a peaceful strike, but practically you do
   not do it peacefully. You defend what is yours. My job is my job,
   it belongs to no one. I know what I am fighting for. If you don't
   use force, problems won't be resolved speedily. This puts pressure
   to the management or government to act.

   Violence sends a message to the whole country, those responsible
   will quickly realise they must resolve things. So the violence
   assists to wake up the entire country, that the innocent will
   suffer.


There's no sweet strike, there is no Christian strike.... a strike
   is a strike. You want to get back what belongs to you. You want the
   response must be positive and quick. You won't win a strike with a
   Bible. You do not wear high heels and carry an umbrella and say
   1992 it was under apartheid, 2007 is under ANC. You won't win a
   strike like that. You need to wear takkies and jeans.


I do not think the law is wrong as such. Law is supposed to defend
   the right to strike and the rights of those not on strike. But how
   can we follow that law? Thina, how are we going to be successful in
   winning our demands? We can't always be upright. Umthetho
   oyaphulwa, oyenzelwe oko phulwa ('Laws were made to be broken'). We
   must follow the majority. The majority vote for a strike. We must
   find ways to make those others join the strike and the decision of
   the majority. You are working, we are on strike. You must be afraid
   for your safety.


Everybody has rights. You have rights, I have rights. That's why
   when workers do it they hide it. They know they are not supposed to
   do that ... A sjambok is during the night, but during the day they
   sing. Everyone has the right to be covered [by the law]. That's why
   they assault at night, not daylight. Yes, they know it's illegal to
   assault. They are breaking the law. Yes, when it comes to a fight
   they don't care.


When you're in a strike you are in a war. Even if the ANC is the
   government, it is not yet uhuru. Workers issues and demands have
   not yet been met, although we are compromising.


We know according to law it is an injustice to sjambok or assault
   somebody. You cannot say because of anger, an individual must
   assault somebody. I do not say that law must be amended, because we
   will be living in a lawless country. But we know that during a
   strike there will be lots of casualties. Since our revolution,
   until we try to live according to its true nature, there won't be
   peace.


Workers are still poor, most workers are still not educated, most
   workers are still suffering Transparency is not practised.
   Agreements are reached but they are not implemented. There's lots
   of corruption in the health department, money is misallocated. This
   brings frustration, when salaries come workers have hope, but when
   workers do not get what they want they get frustrated and violent.
   Service delivery is not properly made, even in local authorities.


The struggle continues. We have democracy, but the struggle is an
   economic one--you need to maintain your family. It is a democratic
   country, but you cannot sit back and just drink water.


During the strike it's an exciting life, after the action we will
   be laughing--'did you see how she jumped?' We will be excited that
   we made the rat busy. In the past we used to call it Ratex, now we
   call it Hale Phirimi (different makes of rat poison), after eating
   that it won't return. It's nice to deal with a rat. It's exciting
   to deal with a rat. We go on strike voluntarily because we know
   what we want. We know why we are on strike. To be in a strike is to
   be a leader, you are not a leader because you benefit. In fact
   being a leader you lose a lot.


When you fight with an illiterate, you must be ready to fight. I
   might start thinking you take advantage, or you do not respect me
   because I am not educated. Even our members have an inferior
   complex, when others start speaking English we start thinking laba
   abafundile baqalile. ('the educated ones are starting again'). We
   use all our force, we pull all the masses. You will never defeat
   us.


It's very painful, because you see patients suffer. But you can't
   serve your patient while ... they must serve my interests first, so
   that I can serve my patient.


It's difficult to be a nurse during a strike. We take them as
   amagundwana ('rats', that is, strikebreakers). But deep down we
   know they are not amagundwana. We understand they have to be there
   to take care of patients, but there is a strike. If they are not
   there patients will die. We feel pity for them, but there's nothing
   we can do.


You pray. God, please forgive me, I know I have sinned, but this is
   earth. You console yourself that you destroyed property, you did
   not kill. Please help me in resolving this strike. I remember in
   1992 we held prayers, we went to the mountains and held a night
   vigil. We prayed. But you motivate your prayer that at least I did
   not kill somebody. Even my kids are sleeping with an empty stomach,
   their children are sleeping with a full stomach. When you are many
   you feel better and enjoy [the strike], but when you are alone you
   think about the situation.
Gale Copyright: Copyright 2010 Gale, Cengage Learning. All rights reserved.