Editor’s Note:
For the labour movement, strikes occupy a special place, taken often as the most direct and visible demonstration of working-class power to disrupt the operations of capital. When workers secure immediate economic gains, the strikes are rightly celebrated. However, beyond individual strike actions and their immediate results, we are not always sure about the more medium and long-term outcomes of strikes and how and why the state responds to strike actions in certain ways.
Manfred Elfstrom, the author of Workers and Change in China: Resistance, Repression, Responsiveness, has studied strikes in China for many years, focusing the causes and consequences of worker protests during a particularly intense period of worker resistance. We spoke with Manfred about how he has studied strikes and state responses, the concepts and frameworks he has found useful, and how he manages to use imperfect data to illustrate trends.
ALR: Some of our readers may be aware of the strike wave in China in the 2000s and 2010s led by rural migrant workers in industrial workplaces. You have done a lot of research on worker resistance in China. Why do you think there were so many strikes in that period? And how did you become interested in studying them?
Manfred: China was going through a boom, and it had a tight labour market. Research in other contexts has shown that strikes are generally pro-cyclical, so they rise and fall with the economy. This feels a little counterintuitive. You imagine that strikes are about people having grievances, and you imagine that they’d have more grievances when times are bad. But when you think about it some more, it makes sense that when times are good, workers who might be laid off because of their activism can find alternative employment, so they’re not as afraid of taking action. There are also just more company profits to divide, and workers are aware of that.
Of course, there were other periods when activism in China was born more out of desperation. We saw that in the late 1990s and early 2000s, among state-owned enterprise workers in China’s northeastern rust belt, when restructuring threatened the world that the workers there had built up over decades. Maybe we’re going to enter a period of activism born of desperation again in China in the years ahead. But I think that in the 2010s, the strike wave came from conditions improving.
I became interested in studying the strikes because I had been involved with different labour non-governmental organisations with programming in China, both more grassroots, civil society-oriented programming, helping NGOs deliver legal training to workers, and more institutional programming in the form of labour law clinics at universities. After doing that for a period, I wanted to step back and see what it all added up to by taking a look at the big picture.
ALR: When it comes to China, sometimes it sounds like there is an undifferentiated category of “Chinese workers” on strike. In your research, you disaggregate “Chinese workers” both in terms of sectors and geographic locations with unique characteristics.
Manfred: I mainly look at manufacturing workers, and a lot of the activism among manufacturing workers in the period I study are light manufacturing workers, such as worked employed in electronics assembly and apparel. But I also look at construction workers, who have been a reliable source of protest over the decades in China, and I look at service workers, such as restaurant workers. There were strikes by people working in car dealerships for a while!
I have also looked in particular at taxi drivers. There was a big taxi strike wave in 2008 and 2009, and these drivers had different demands in different places. They were learning from each other, and they had different impacts in different places. An interesting thing about the taxi drivers is that they, in turn, have a lot of internal diversity. Some cities have municipal taxi monopolies that are set up under the government’s control, while other cities are a complete free-for-all. Some cities have a few companies controlling things or a limited number of medallions. They create different grievances and different openings for putting pressure on authorities. Another interesting thing about taxi drivers is that the government itself is often the centre of the claim-making, because it very much sets the terms of the market. That’s not just true in China, but also in other places.
So there’s a lot of diversity between and within sectors. While it’s sometimes useful analytically to just speak about “Chinese workers” and the “Chinese state”, you have to break those categories up.
ALR: This leads to my next question about how you analytically break it down. What are some of the concepts and frameworks that you have used to disaggregate things?
Manfred: My focus is on workers’ tactics and demands. I distinguish between workers making contained demands, boundary-spanning demands, and more transgressive demands.
Contained demands would be for the legal minimum that’s owed to workers ‘ compensation for work injuries and wage arrears. Boundary-spanning demands would be raising claims beyond those legal minimums. Then, the more transgressive demands would be for wage raises and also reforms to enterprise-level trade unions and other sorts of completely novel claims.
Similarly, with tactics, you could look at contained tactics like litigation and petitioning; boundary-spanning tactics like strikes, protests, and riots; and then more transgressive stuff, which would be especially disciplined or cross-worksite strikes.
Finally, you could divide up the kinds of organisations that workers are using along the same lines. You could look at contained organisations, which would just be certain networks within a particular work site; more boundary-spanning organisations, which would be legally oriented non-governmental organisations; and more transgressive organisations, which would be non-governmental organisations that have moved in a more movement oriented direction, and other associations of different kinds, engagement with people from other sectors and students.
What leads to these different forms of resistance in the first place? I think structural and demographic factors offer explanations. Some workers are housed and work together in large numbers in repetitive, dull, and dangerous work, and under a lot of pressure. Others have more training and work in smaller numbers. Some travel really long distances to work. Some work from home, or just in a province near where they grew up. All this shapes whether workers move in a more contained, boundary-spanning, or transgressive direction.
ALR: A lot of these concepts come from studies of contentious politics developed in very different contexts from China. When you bring those concepts and framework to bear on China, do you find they’re easily transferable?
Manfred: A lot of these concepts were developed to understand liberal democratic contexts, so they do need a bit of adaptation in an authoritarian setting. I think that’s particularly true when it comes to understanding movement outcomes. That’s always been a weak point of the study of social movements.
People find it more exciting to talk about why people protest in the first place than whether it works or not. It’s also just a little bit easier to study because you can dig down and look at the determinants of different tactics, or look at where unrest crops up and where it doesn’t. But pinpointing cause and effect, concessions or repression, is trickier. There are a lot of other things going on, and whether a movement is responsible for some change or something else is responsible for the changes, it’s just hard to say.
It is extra hard in an authoritarian context, because some of the traditional ways one might look at outcomes, like votes in a legislature and election results, don’t have the same meaning or are meaningless in an authoritarian context. In a way, the state is a little more front and centre in an authoritarian context, and people direct their demands more readily at the state rather than other parties. And at the same time, it’s got this kind of depoliticising effect. People try not to frame their demands in a way that touches on the state’s core legitimacy. So, all those things I think are a little bit different and take some thought when translating over.
ALR: If we can stay on the question of outcomes, for the labour movement, the outcome of a strike is often concerned mostly with whether the immediate demand has been met. It is less common to talk about the medium-term outcomes, such as whether it impacts state behaviours and policies, or long-term outcomes, such as whether it builds the strength of the movement.
Manfred: I think the immediate results of collective action are meaningful and worth studying. But I try to reach a little beyond that immediate result in my work in a couple of ways. In my book, I look at regional models of social control that have developed in response to the kind of contention governments are facing. And then I look more broadly at the state’s capacity to be responsive and capacity to repress activism, and how that’s shaped by rising and falling unrest.
So, at the level of regional models of control, I compare the Yangtze River Delta (YRD) and the Pearl River Delta (PRD). In the Yangtze River Delta, I say that authorities have faced contained to maybe boundary-spanning contention, and as a result, they’ve focused on preempting things as best they can. They’re cautious with the laws that they do pass, and they try to balance the social forces as best they can, nudging them into place.
In contrast, whether contention is more boundary-spanning to outright transgressive, like in the Pearl River Delta, officials have given up on preempting everything, and they mostly come in heavy when things get out of control. Instead, they’ve innovated to some degree, though this was more the case a decade ago than today, with respect to their labour laws and the policies of the official trade union. At the same time, they have come down really hard on individual organisers and high-profile incidents.
More broadly, if you have those sorts of regional models, which I describe as orthodox in the case of the YRD and risk-taking in the case of the PRD, you have this buildup of repressive and responsive capacity nationally. I measured that in my book with provincial-level spending on the People’s Armed Police, the paramilitary force that was boosted after Tiananmen to deal with routine social disorder, and pro-worker or split decisions in formally adjudicated employment disputes.
ALR: On the question of the state, for decades now, people refer to the Chinese state as a black box because of the lack of transparency and understanding of the decision-making mechanisms. Researchers and activists alike look at social unrest as something that goes into the system, and a policy or law comes out at the other end after some time. We don’t really know what happens in between. Is it still a black box?
Manfred: I think the central state still is a bit of a black box, and maybe more of a black box than it was under the previous administration of Hu Jintao and Wen Jiabao. But you can nonetheless try to think through the incentives of people at different levels of the system, and in particular, think through who people in different points of the system feel like they have to impress. Lower-level bureaucrats have to impress higher-level bureaucrats. One way they do that is how they handle social unrest. You might think that would mean a sort of uniform response everywhere, but my argument is that how they respond depends on the type of unrest they’re facing: contained, boundary-spanning, or transgressive.
Where unrest is transgressive, there’s an incentive for an official to take some risks, to show that they are taking the issue seriously. Maybe not all these policies stick around, maybe not all of them are sustainable, but at least they are trying different things out. As someone in the government in such situations, you know that unrest has already gotten away from you, and you can’t necessarily bring it to a halt in short order, but you can show you’re trying to do something right.
Where unrest is more contained to maybe a bit boundary-spanning, as an official, you don’t want to take risks. You don’t want to introduce problems where you don’t already have problems. You just want to kind of keep things humming along the way they are. Your task is to be orthodox, to keep your head down, and to show you’re just managing.
I think these different models of control you see in China are a result of the interplay between the pressures that are coming up from below and the incentives people at different levels of the bureaucracy feel to show that they’re competent in handling this stuff.
ALR: I want to ask more about state responsiveness because state repression is a constant factor and something the labour activists in China expect. But we talk much less about state responsiveness, where the state tries to accommodate workers’ demands to some degree.
Manfred: I think what’s interesting about China is that you can get repression and responsiveness at the same time. Governments around the world repress labour, and governments around the world make conciliatory gestures toward labour. And if you look at, say, U.S. history, you have these kinds of ebbs and flows over time. Even if you zoom into something like the New Deal era, you have periods when the government was very responsive to labour demands and moments when they moved back into a neutral position and even turned a blind eye to brutal crackdowns on labour. But in the U.S., and the history of a lot of other places, you don’t generally see the government going all out with carrots and sticks at the same time.
I think the reason that the Chinese government tries everything at once is that the Chinese labour doesn’t have a strong institutionalised voice. There’s no one across the table to bargain with, and so the authorities are flying somewhat blind and just sort of seeing what sticks.
Each part of the bureaucracy does what it can to show that it takes the issue seriously. That means that the coercive organs coerce, and the welfare- and justice-oriented organs dole out some welfare or dispense some semblance of justice. It all happens at the same time.
Over the long term, the argument of my book is that this all-of-the-above approach isn’t really sustainable, and kind of warps the Chinese state: a gesture to labour alienates business, and a crackdown on labour alienates workers. Money for the People’s Armed Police, especially now with the economy slowing, means less money for the social programs that China desperately needs, and vice versa.
ALR: Can you give some examples of how this responsiveness works in terms of specific measures or actions?
Manfred: At a provincial level, the Pearl River Delta and its province, Guangdong, at the height of labour unrest in the late 2010s and early 2010s, put through a bunch of different labor laws that were inadequate but gestured at what was needed: language about workers being involved in the governance of enterprises to some degree, language that seemed to come close to recognising the normality of striking, and language about reforms to the trade union system. Nationally, you had a trio of laws that went into effect in 2008: the labour contract law, the employment promotion law, and the labour dispute mediation and arbitration law. So, those are all conciliatory gestures. And on the ground, you can have the government leaning on employers to pay workers what they’re owed, or if the employers clearly can’t do it, the government stepping in itself to pay part of the difference. You can have local unions being a little more proactive within the constraints of the Chinese system.
ALR: What I find interesting when I read your book is your use of data, especially expenditure on public security and the people’s armed police, which you found in published government yearbooks.
Manfred: It’s all imperfect. I think the people’s armed police is the best barometer of state reactions to popular pressures, because that’s its remit. The data for that stopped being available in the yearbooks I was looking at after 2009, so after 2009, you have to fall back on this broader category of Public Security spending, which also includes spending on the judiciary and other things, and is a less perfect measure. I also pull in anecdotal data about spending on surveillance systems, procurement for anti-riot gear, but it’s fairly patchy.
ALR: For data on strikes, the Chinese government has long ago stopped releasing any data on strikes and protests, but there have been other attempts to document strikes. You wrote about how to count contention and the problems arising from some of the efforts. And now, globally, there are several projects to collect data on strikes, which have come together to produce international strike reports.
Manfred: I think it’s really important to collect that kind of data because it gives you a sense of trends. But it’s always difficult if you’re not dealing with government statistics, and maybe even when you’re dealing with government statistics, to distinguish between what’s a rise or fall in strikes in reality and what’s just a rise or fall in reported strikes. And that’s particularly an issue in China.
The strike wave in the 2000s and early 2010s was real, and the reported relative quiescence of labour nowadays is probably also real. But it was also the case in the 2010s, especially, that you had a lot more data to work with. You had the people involved in the Wickedonna project constantly drawing down social media reports on actions of all kinds that were happening. You had China Labour Bulletin’s Strike Map that drew on the Wikidonna effort to some degree, and you also had a more active state media at the local level. It used to be that you’d have a fair bit of official reporting on incidents that were big enough or disruptive enough that local reporters felt like they had to say something about them.
The challenge is always disaggregating that data. Taking things back to a topic we discussed earlier, I think that the outcomes are often what’s most important to the people who are actually involved. But outcomes are tricky to parse out if you do it at the level of an individual news report of the sort that makes up a protest dataset. Usually, the report is written before the full response has been made. If you take a longer-term perspective on strikes, then you have to start exercising a lot of personal judgment about what was effective and what wasn’t, and what added up to something down the line and what didn’t. This is fine with a couple of reports, but it becomes a problem when you’re dealing with hundreds or thousands of them.
So, those are the two issues: one is the availability of data, and the other is reporting outcomes in a reliable and meaningful way. I think the more you can capture that in the big picture, the more valuable your findings are for people in the labour movement. And if you can’t capture it with quantitative data, it speaks to the need for really in-depth qualitative studies that document the advances and setbacks of a particular campaign in more detail.
