No Union, No Problem: Anatomy of a Vietnamese Wildcat

An account of a recent wildcat strike in a Vietnamese shoe factory


August 16, 2018

Vietnam has hundreds of recorded wildcat strikes each year. This inquiry investigates one such recent event at Shoe Land factory.1 Friends of mine - workers at the factory - twice sat down with me for hours over coffee and shared their knowledge and experience of the strike. It is these friends who asked me to write about it, so more people can understand their circumstances. I also learnt from other workers who I didn’t previously know by waiting outside the factory gates and asking if they’d be willing to teach me about their strike experiences.

Vietnamese labour militancy: background and context

Vietnam’s transition from a planned economy to a market economy occurred during the 1980s, following a period of major economic crisis. Since then, millions have moved from smallholder farming into capitalist wage labour. The first major export industry was the textile industry, which is still one of the top two, along with the electronics industry. These industries employ large amounts of young migrant workers from rural provinces, many of whom often leave the industry, exhausted, by their mid-30s, and return to their home villages. The country’s integration into global capitalist production also coincided with the rise of neoliberalism, and working conditions in the so-called ‘formal economy’ have become much more precarious than they were for the small number of people who worked in state-owned industry during the planned economy period. Contracts are often short or non-existent, automation threatens many jobs, and foreign companies can flee the country with impunity.

Vietnam’s one-party “Communist” state allows only a single legal trade union federation, the state-led Vietnam General Confederation of Labour (VGCL). Its leadership is not made up of workers, but party cadres and government bureaucrats. In the past, the VGCL often took an explicit anti-worker position, persuading them not to strike or to go back to work if they did so. Now, however, the union is more neutral – it never organises strikes but also doesn’t generally pressure workers not to strike. Workers generally see the union as irrelevant – they do not think it is against them, but see it as largely doing nothing beyond giving workers small financial gifts if, for example, they have a child, or are taken ill.

Strikes in Vietnam were legalised in 1994. Since then there have been hundreds of recorded strikes every year. Not one of these strikes, however, has ever been legal. A legal strike requires an almost impossible set of bureaucratic procedures, which would give employers more than enough time to prepare for the strike and make sure it barely affected production. A legal strike must also be led by the state trade union, which does not happen. Consequently, every strike is a technically illegal, wildcat strike. These are, however, largely tolerated by the state. Strikes commonly begin with one major demand, with others added during the duration of the strike. They are often successful, with at least one demand usually being met by the companies concerned.2

Shoe Land: background and context

The factory in our investigation, Shoe Land, is a large factory employing over ten thousand workers, located in southern Vietnam (the country’s industrial heartland). It primarily produces shoes for a major international brand. Workers at Shoe Land work shifts. There are 3 shifts per day - morning, afternoon/evening, and night, ensuring that the factory continues operating for 24 hours a day. The exception to this is Sunday, when all workers have their weekly day off.

Workers work on production lines in a number of smaller ‘workshops’ inside the Shoe Land compound – each workshop has a few hundred to a thousand workers. Work at the factory has become more intense in recent years, with production targets increased each year. When there are no visitors or inspectors - such as quality controllers from the brands the factory produces for - the speed of the production belts is increased to force workers to work even faster. They are frequently verbally abused. In addition, hygienic practices are not properly observed. Water coolers, for example, are sometimes kept directly next to toilets, and the food in the workers’ canteen has been known to contain live bugs or worms. Workers are further threatened if they try to raise their voice about these issues.

Shoe Land is held up as an example of the ‘formal’ employment that the Vietnamese government wants to expand, but workers feel increasingly precarious. Automation is currently a hot topic in the country, and workers who are seen as working too slowly are often threatened with the sack and told that they can easily be replaced by robots. Another problem is ‘the shedding of the cicada skin’, as it’s known in Vietnam. Foreign factories sometimes just flee Vietnam, leaving weeks of unpaid wages and social security contributions. Shoe Land’s management sometimes threatens to do this if workers don’t work harder. In recent years, the management has attempted to force out many long-serving workers, replacing them with workers on one year contracts which are not then renewed (despite promises to the contrary at the start of their employment).

Shoe Land has an enterprise union, which is totally separated from workers’ lives. They have no say in electing the union rep, who is a member of the Shoe Land management. While some workers are not concerned about the enterprise union, viewing it as a nonentity except for occasional financial gifts, others consider it to be explicitly against them. They see it as another arm of management, and a tool for watching and controlling workers. The union never does anything to attempt to fight for or represent workers. They have slightly more sympathy for the local VGCL union office of the area in which Shoe Land is located. They don’t feel that it is under the cosh of management – that is, they don’t feel that this union office is against them. But neither do they feel that it is for them. Rather, they consider it to be irrelevant.

The strike brews

At the beginning of 2018, Shoe Land announced a change in their salary policy. The company had 24 salary levels. Each year, a worker would move up to the next level and get a pay rise. Shoe Land wanted to reduce this to just 8 salary levels. This meant that every worker who was currently on levels 9 to 24 would receive a pay cut (and a very substantial pay cut for the longest-serving workers). Workers currently on level 8 would stay there, and other workers would max out at level 8.

At the end of 2017, Shoe Land announced a consultation period regarding the new policy. It was a total sham. Management conducted a survey to see what workers thought of the policy, but there was no opportunity to disagree, or to say that they did not want it to change. Rather, they were presented with various options, all of which involved cutting the salary levels. The most generous option was a salary system with level 10 as the highest. In addition to the survey, workers were asked to contribute their own ideas. These were not treated confidentially or seriously, but were used as a way to identify and keeps tabs on those who disagreed.

Workers were at first unclear about the new policy. They began talking to each other about what it really meant, beyond the management rhetoric. They discussed it when eating lunch or having coffee, and through a Facebook group for all of the factory’s workers. Initially, workers wanted to analyse and understand the implications of the new salary policy; once these became clear, the discussion turned to striking.

The role of the Facebook group was very important here. Workers started building solidarity using posts, photographs, and GIFs recalling previous outbursts of militancy at Shoe Land, 3 urging each other to act with ‘one heart’, and calling on each other to bring Vietnamese flags to work around which they could unite. Some wealthier workers or their families offered to buy food and water for striking workers. Over a week or so, calls and support for a strike grew palpably stronger, although nobody knew exactly when it would occur.

The strike happens

The strike eventually started on a Friday morning. It began in the workshop where workers make the base of the shoe and where the production of a shoe begins. Workers in this workshop are respected by their colleagues, as they have a harder and more strenuous job than those in other workshops. If these workers stop working, the rest of production has to stop too.

Workers from this workshop went outside, calling other workers to come out, by shouting and by posting calls to join the strike on the Facebook group. This quickly gathered momentum, as workers from more and more workshops joined the strike. The Facebook group became filled with posts such as “Workshop X, come on down!”, or “Workshop Y, where are you?” The supervisor of one workshop tried to lock the workers in, but quickly gave up that idea. Workers inside the workshop posted on Facebook saying that they weren’t allowed to leave, so workers outside surrounded it and forced the boss to open the door. Within an hour or two, the factory was at a standstill, and the workers headed outside the factory gates. Later, workers arriving for the afternoon shift joined the strike – many, of course, knew through the Facebook group that the strike had begun so came with the intention of taking part, rather than going to work.

Small groups of striking workers took autonomous actions. One worker led a group of around 20 others to try and stop scabs. They carried mắm tôm – an incredibly pungent shrimp paste – which they threw at people attempting to enter the factory to go to work, and threatened to throw it at others. This is a standard way to stop scab labour during strikes in Vietnam; mắm tôm stains and leaves a strong, unpleasant smell on clothes that is very difficult to remove. Other groups of workers organised themselves to do the same or similar.

Shoe Land is located on a main road leading into a big city. Many buses, vans, and motorbikes traverse this road every day. Some groups of workers took it upon themselves to block the road, in order to try and force the state to put pressure on factory management to concede to workers’ demands. The striking workers had naturally spilled out onto the road, as the pavement outside the factory gates was not wide enough to hold all of them. Workers taking the autonomous action used these physical bodies as a roadblock. They stopped traffic from passing, although most drivers decided to avoid the road. Only vehicles carrying, for example, schoolchildren or the elderly were allowed through, but the rest were stopped. The road soon became empty of traffic.

The strike had no leaders. Logically, there would have to be one worker who mentioned going on strike before the others, or who downed tools before the other workers. Nobody, however, knew who this person was (perhaps they did not even know themselves). Certainly no one was trying to control the direction of the strike. And nobody declared themselves to be the strike representative, wanting to negotiate with the company.

Nobody from Shoe Land’s management came down to the factory gates to talk with the workers, and neither were workers invited inside to negotiate. This would not have worked anyway – the strike had no leadership or official representatives. Even if a worker had declared themselves to be the workers’ representative, there is no guarantee that other workers would have accepted this. Rather, the factory communicated by printing announcements, which were pinned to noticeboards and passed around the workers. The workers in turn communicated with each other by talking to other workers nearby, but most importantly through the Facebook group. Here, workers would ask what the latest offer from the company meant and discuss whether to accept it or not. Discussions and debates could be started by anyone on the Facebook group and workers would respond to them. There were no formal procedures for taking votes or making decisions. The general position of most workers regarding whether to accept the company’s offer or not would become clear through this process.

As the strike progressed, new demands were added. Again, this was largely through the Facebook group, and groups of workers would go to banner shops to get banners printed with further demands, or simply write them on pieces of paper. The three demands which caught on the most were: the general, and generally hated, manager must go; sick leave should be increased to its previous level; 4 and the 14 days per year previously allowed off for ‘private affairs’ (a term in Vietnamese labour law, which is a mix of compassionate leave and important personal events such as weddings or appointments with government departments to make official documents such as ID cards or passports) should be reinstated. 5

Shortly after the strike broke out on Friday, Shoe Land’s management put up their first announcement. It said that they had been merely looking for the views of employees regarding the new policy, and that employees pay policies were still in place. The announcement finished with the hope that everyone would go back to work with ‘peace of mind’. Unsurprisingly, the workers rejected this.

The second announcement, distributed on Saturday morning while the strike was still ongoing, finally made some concrete commitments. This announcement said that the company had cancelled and would no longer spread the ‘reform plan’ for changing the salary level policy. It also committed to keeping the current policies, including salary policies. There were many discussions about this announcement on the Facebook group whilst the strike was continuing.

On Saturday afternoon, the company management printed a third announcement. This time, they added that the sick leave policy would be increased to its previous level. They also committed to not docking workers’ pay for the 2 days of strike action – stressing that they did not have to do this according to the law – as long as everyone returned to work as normal on the following Monday.

The strike ends

At the end of Saturday, workers were considering the most recent printed announcement from the company. Nobody was sure whether they would accept this offer, or whether the strike would continue the following week. All workers have a day off on Sunday; throughout the day there was keen discussion and debate on the Facebook group regarding what to do. By the end of the day, it was clear to all – even those who disagreed – that most workers in the group wanted to return to work. On Monday, this is what they did.

Some militant workers were disappointed with this, thinking that had the strike gone on longer they would have been able to extract more concessions from management. They felt that if workers had not had a day off after the second day of the strike, it would have continued; the Sunday had acted as a ‘cooling off’ period, which meant that the strike somewhat lost its momentum. These workers wished that the strike could have started on a Monday, so that there would have been 6 days before the day off, rather than starting on the Friday, which meant only two days before the rest day. They knew, however, that it would have been almost impossible for any one small group of workers to fix the start day of the strike in advance. Overall, however, workers considered the strike to be a success. The main demand was met, along with some other demands - they regained 30 days of sick leave and were paid for the time they were on strike.

Implications

The most obvious point to make about this strike is that the union was quite literally nowhere to be seen. They were absent for the entire duration of the strike. Workers, however, were not demanding that the union get involved. Some workers dryly commented that the union is supposed to help workers, but only as a sarcastic joke, accompanied by laughter. They did not actually want the existing union to step in.

Instead, the strike bypassed the union completely, and took a leaderless form, with workers communicating, discussing, and debating with each other primarily through the Facebook group. Small groups of workers took extra, autonomous actions. The strike worked. The main demand was met very quickly, as were some extra demands. Nobody was disciplined or fired as a result of the strike. But there were also limitations. As mentioned above, some workers wished there was a way to control the day the strike began, so the action would go on longer and secure further concessions, which is very hard to do given the form the strike took. Since the strike, the company has also increased production targets for all workers by 5%.

Vietnamese workers are rejecting the existing union structure, of the big, bureaucratic “Communist” type. Successful worker actions instead take alternative, grassroots forms, frequently bypassing the union altogether. Such strikes are successful at frustrating capital at immediate sites of accumulation. The challenge, however, is whether and how this can develop into struggles and movements which force more sustained concessions from capital, or which challenges broader structures of capital accumulation.

These questions are relevant beyond Vietnam. The garment industry is one of the largest points of surplus value extraction worldwide. Vietnamese garment sector workers are part of a massive class bloc, both in terms of the millions of other workers in the industry across the world, and in terms of workers throughout the commodity chain: agricultural workers gathering the raw materials; manufacturing workers making the garments; retail workers selling the clothes in shops around the globe; and transport workers, moving the commodities between these different stages. We need to understand existing militancy in all of these places, to consider how local struggles could develop into transnational struggles, linking and building solidarity with other workers across the sector and across the commodity chain.


  1. A fake name to protect the workers involved in this inquiry. 

  2. This is a very brief overview of the Vietnamese labour landscape. For more on the macro context of wildcat strikes, see Eric Bell’s Chuang blogpost, Đình công tự phát: wildcat strikes in post-socialist Vietnam

  3. Prior to this strike, the most recent had been in 2016. 

  4. Previously, workers were entitled to 30 days sick leave per year. A few months before the strike, however, this was changed. Workers were now still “entitled” to 30 days per year, but if they took more than 15, their Lunar New Year bonus would be cut. Workers rely on a 13th month pay packet for the Lunar New Year, so they can buy tickets home and gifts for family. The workers were demanding a return to the 30 day allowance with no cut to the year-end bonus. 

  5. Again, the company’s regulation had become that workers were allowed to take these days off, but that their year-end bonus would be cut if they did so. Workers were demanding a reversal of this punitive policy. 


author

Joe Buckley

Joe Buckley is a PhD candidate in International Development at SOAS, researching changing working conditions and labour militancy in Vietnam’s garment and textile industry.

read next