Tech Worker Organizing in China
A New Model for Workers Battling a Repressive State
by Kevin Lin
In March 2019, an online mobilization by Chinese tech workers spurred a national discussion about the normalization of overwork and labour exploitation. The campaign-style, decentralized tech worker organizing, coordinated by workers across companies and regions, indicates the potential of a new type of labour organizing in China.
As the worker center organizing model is adapted to organizing industrial workers, it becomes increasingly unsustainable due to state repression. The tech worker organizing provides clues to a new model for worker mobilization, that may hold particular advantages under the current political environment in China.
A critical assessment of this new model is intended to advance the strategic considerations of organizers and their allies working to improve the conditions for workers in China. How do we understand the tech workers’ mobilization in the context of Chinese labour organizing? To what extent does tech worker organizing share elements of other recent labour organizing efforts, or does it offer a new model of labour organizing?
More broadly, what does this mean for the future of Chinese workers’ organizing and labour rights advocacy after the severe repression against labour groups and activists of the last several years? These questions are of central importance to the future of labor organizing in China.
Evolving Labour Organizing in China
Four decades of industrialization in China has led to the development of a working class growing in militancy. The past fifteen years in particular have seen intense labour organizing that is almost entirely independent of the labour officialdom, of the government-aligned All-China Federation of Trade Unions (ACFTU).
The ACFTU, an extension of China’s state bureaucracy, represents state interests at the grassroots level and monopolizes labour representation nationally. It serves to carry out the government’s labour campaigns and policies. It either co-opts—when it can—or displaces independent labour organizing, with top-down unionization campaigns to establish management-friendly workplace unions.
Yet led by rural migrant workers, the Chinese working class still organizes autonomously and has staged several thousand protests and strikes a year over the last fifteen years. This has been done without their own formal worker organization within the workplace. It has relied on organic worker leaders and hometown-based networks organizing on an ad hoc basis.
On the one hand, the government remains vigilant against independent unionism and disrupts any attempt at grassroots unionizing. But on the other hand, the transient nature of migrant labour, coupled with the constant capital relocation in search of cheaper labour across cities and regions, has militated against the formation of stable worker organization.
The independent labour organizing has taken advantage of China’s emergent labour law system to fight for the legal enforcement of basic workers’ rights. This has given rise to the largely rights-based labour advocacy primarily in the export-oriented manufacturing sector.1
It is further aided by the emergence of worker centers, also referred to as labour non-governmental organizations (NGOs). These are usually led by former workers whose primary roles are to educate workers on legal rights and support their organizing with strategic advice.
Despite their constrained role, the activities of worker centers and their staff have been increasingly criminalized since 2015. That year a number of worker center leaders were arrested and charged for their labour activism.
. . . [T]he . . . All-China Federation of Trade Unions . . . either co-opt[s]—when it can—or displace[s] independent labor organizing with top-down unionization campaigns to establish management-friendly workplace unions.
Since mid-2018, a new round of state repression against labour activism was ignited after workers and student activists rallied around a unionization drive at Jasic Technology in Shenzhen.2 Workers at Jasic organized to unionize their workplace. Management fired the organizers, and when they protested outside the factory gate, they were detained by police.
Workers, joined by college student activists, then demanded their release and called for union recognition. The government was alarmed by the participation and mobilization of college students in solidarity with workers. Over the next year, more than 100 workers, activists, and students were detained by the government either in direct connection to the Jasic organizing, or as part of the widening of labour repression that lasted into 2019.
It is therefore surprising that the tech worker mobilization took place amid this intense period of state repression, that has tightened in order to constrain labour unrest as the Chinese economic growth model runs out of steam. The mobilization therefore deserves a close look as to how it managed to break out and make a national impact, while other labour activists have less and less organizing space.
Tech Workers Confront White Collar Exploitation
Tech workers are at the new frontier of labour organizing.3 Often seen as high-salaried and difficult to organize, tech workers in the United States have stepped up their activism in the last few years around issues such as sexual harassment, gender inequities, and military research, taking on giant corporations that wield enormous corporate power.4
In particular, some tech workers have actively been organizing through the Tech Workers Coalition, while unionizing remains a challenge.5
. . . [T]he activities of worker centers and their staff have been increasingly criminalized since 2015, when a number of worker center leaders were arrested and charged for their labour activism.
In China, the tech boom—with tech giants like Huawei, Tencent, Alibaba, Baidu, Xiaomi, and many others—has employed millions of young, college-educated workers. Overwork in China’s tech sector has plagued these workers for some years.6 Intense competition has given rise to this excessive and illegal work schedule that so transparently violates Chinese labor law.
Young workers have been willing to accept the culture of overwork as a trade-off for higher salaries in a prospering sector. However, the slowing down of the tech boom, led to a hiring freeze and layoffs in some tech companies. This fueled tech workers’ growing disaffection.
The stereotypical image of privileged white-collar workers persists, despite the changing employment reality. Their college degrees and technical skills are supposed to place them a favorable position in the job market, with high wages, end-of-year bonuses, generous benefits, and perks as well as a high degree of job security. This is despite the fact that white-collar salaries have been stagnating or declining.7
Although there have been small protests by white-collar workers against their companies over the years, the March 2019 virtual uprising by tech workers is the first widespread mobilization with a high level of public prominence. The existing network of labour organizers and worker centers in China, even during their peak in the late 2000s and early 2010s, had rarely dealt with issues affecting white-collar workers. None of them seemed remotely to have been associated with this type of mobilization.
The Chinese tech worker mobilization started in March 2019 when an anonymous user uploaded a repository named 996.icu on Github, a software development project hosting site owned by Google. The name refers to the extreme work regime of 9 a.m. to 9 p.m., six days a week that risks landing oneself in the ICU (intensive care unit).
It did not take long for tech workers to flock to the site to register their complaints, making the page number one trending on Github and sparking related complaints on other social media platforms.
. . . [A]n anonymous user uploaded a repository named 996.icu on Github . . . [which] refers to the extreme work regime of 9 a.m. to 9 p.m., six days a week that risks landing oneself in the ICU . . .
Despite the government’s close scrutiny of labour rights advocacy, the mobilization was not only widely discussed in public, but also elicited very sympathetic responses from China’s mainstream media. Even the state-run media published articles criticizing the exploitative conditions of the 996 work culture.
In response to the companies’ dismissal of workers’ concerns, one article in the state-owned People’s Daily commented, “Employees who object to 996 cannot be labeled as ‘slackers’ or ‘not fighters.’ Their real needs should be considered.”8
The focus on working hours and overtime by the tech workers is similar to issues manufacturing workers have organized around. The experience of the lengthening of working hours, including unpaid overtime, and work intensification, as well as job cuts and the hiring freeze has reached a boiling point for many tech workers in terms of job insecurity and work pressure.9
The growing number of college graduates in China, a result of the expansion of the higher education system, creates competitive pressure. Once highly prized jobs have now been increasingly downgraded, creating disappointment and resentment among tech workers.10
However, in contrast to manufacturing workers’ organizing, the tech workers’ mobilization did not involve work stoppages that most labour organizing in China has relied on to pressure management. The mobilization instead utilized public pressure to cause reputational damage to the companies. This is rare for labour activism in China, who rather exact direct economic costs by refusing to work. These are not mutually exclusive tactics, and labour organizing and campaigning could use both.
The tactic resembles other social movements such as Chinese feminist advocacy. In recent years, feminist activists have used social media to launch public campaigns. These focused on gender inequality and sexual harassment, to raise public consciousness, and to bring public pressure on the government and public institutions like the universities.11
While this tactic has garnered considerable interest, labour organizing often depends on workplace-based power. Hence building that power and sustaining it over time may need to combine both tactics.
The mobilization . . . utilized public pressure to cause reputational damage to the companies, something that is rare for labour activism in China . . .
In this case, the tactic centered on public pressure worked remarkably well. The 996.icu website listed several labour rights laws related to working hours and overtime pay, and blacklisted companies such as Alibaba, JD.com, and Huawei that are among the worst offenders.12
The public and the Chinese media - both those operating as a mouthpiece of the government and more commercialized media - reacted quickly. They expressed condemnation of the practice and support for workers’ rights, something not seen in recent years regarding a labour rights issue.
In general, the media and the public do not discuss labour rights issues in such a public manner. In no small part, this is because reporting strikes and labour actions, or social protests of any scale, are not allowed. Print, television media and social media are under a highly restrictive censorship regime.
The lack of reporting on labour rights was also enforced through the arrests of citizen journalists. These publish on non-traditional media platforms, such as Lu Yuyu and Li Tingyu in 2016 who maintained a database on protests; and Zhen Jianghua in 2017 who documented human rights abuses.
The Chinese government has tightened control over once popular social media platforms such as Weibo, and increasingly WeChat, with some users being visited by the police for their posts.13
This environment has given rise to efforts to take the action offline. The 996 action responded to this heightened repression, with activists organizing public actions that targeted both the tech companies and the government.
The first action took place on April 21, 2019 when these activists launched a freedom of information campaign by sending letters to local government departments. They requested information regarding 2018 to 2019 steps, that the government had taken since 996.icu made public the violation of the labour law regarding work hours.
The public reacted with severe backlash on social media against comments made by Alibaba’s founder, Jack Ma, who called working 996 a “huge blessing,” while JD.com’s CEO, Richard Liu said, “Slackers are not my brothers!” On May 4, 2019, activists sent Jack Ma copies of the labour law, but photos of the actions posted on social media were quickly censored.
The mobilization appeared to have hit a hard limit when it tried to move off-line. Due to censorship and the news cycle, once the public eye was no longer on the mobilization, it was not able to build on its early success.
Networked and Decentralized Labour Mobilization
Networked and decentralized mobilization has both familiar and new features. Networked organizing has happened many times before in China, including taxi drivers’ strikes for many years, the truck drivers and crane operators’ strikes in 2018, and Walmart store workers’ strikes in 2016.
In each of these cases, workers from more than one workplace or one location were part of a network that served to transmit information, discuss strategy and, to varying degrees, coordinate protest and strike actions. This often is born out of necessity due to their decentralized workplaces. Or, it is sometimes because the same company policy or change in employment conditions would have a very similar impact on workers across locations.
There are fewer such cases of coordinated worker organizing across factories. When it happens, it is usually because the factories have the same employer or parent company, or there is a concentration of one industry such as garment manufacturing or electronics in one locality where strikes often become contagious.
It is worth considering the example of Walmart worker organizing in China over a number of years. The campaign to improve working conditions and better treatment of workers at Walmart retail stores in China has been built by a group of activists, all of whom were former or current employees, around the issue of flexible scheduling which creates extreme uncertainty of work hours for workers.14
A Walmart employee association (not an actual registered organization) was autonomously organized as an online hub for sharing information and as an organizing platform. In some ways, it is similar to OUR Walmart, a group created in the United States by the United Food and Commercial Workers (UFCW) union aiming to organize Walmart employees. In both cases, the organization operates outside the workplace and aids activist employees to do advocacy.
Both are also born out of Walmart’s hostility to labour organizing. But one difference is that, while Walmart in the United States remains union-free, the Chinese government forced Walmart to set up a union. However, the management of Chinese Walmart outlets manipulated and obstructed union elections, resulting in the firing of the activists who stood for elections as union representatives of the Walmart store-level union.
The (autonomous) Association helped coordinate the strikes at four retail stores across as many cities in 2016, a rare case of coordinated actions across several regions. This is important because the Chinese government has guarded against workers and activists linking up outside of their own workplaces, and the case of Walmart strikes countered this.
Efforts continue to organize Walmart store workers. However in addition to greater government repression against organizing, the fact that the organizers do not have organized presence within the stores poses a further challenge.
In the case of 996.icu, the more sophisticated use of technology not only aided the mobilization across multiple workplaces, but it managed to shift broader public consciousness. It also suggests a new basis of international solidarity.
Several domestic companies including Alibaba, Tencent, and Xiaomi reportedly blocked access to the 996.icu repository. This led to the concern that the Chinese government may attempt to pressure Microsoft and Github into censoring the project. In response, Microsoft software developers based in the United States released a letter in solidarity with tech workers in China.15
The letter called on the company to ensure the project remains uncensored, and was signed by Chinese and American tech workers alike. It is important to highlight how rare this type of international solidarity is for labour. Despite years of both high-level union exchanges and people-to-people discussions, actual communication and solidarity actions in support of one another remain uncommon.
This is an important step in forging international labor solidarity between American and Chinese workers. Github remains open in China, creating the possibility of a space for discussions surrounding social justice issues.
. . . Microsoft software developers based in the U.S. released a letter in solidarity with tech workers in China.
Tech worker organizing poses important strategic questions. Networked and decentralized organizing is the approach that many labour activists and other civil society activists used in the early to mid-2000s before they formed their NGOs from the mid-2000s onward. Prior to taking on organizational form, activists relied on networked advocacy both on the internet and through offline personal and activist networks.
This was the period before an emerging civil society developed apace, and activists, even if they wanted, would have difficulty finding a rights-advocacy organization to join. Then throughout the 2000s, the activists fought for the political space, usually with international funding pouring into China to support civil society, to consolidate their activist networks into rights-based non-profit NGOs.
For many activists, these NGOs—as sustainable organizations employing paid full-time staff to coordinate and implement projects—amplified their advocacy impact. While most of the organizations that emerged remained small, they multiplied and played increasingly significant roles in providing direct services, trainings, and rights and policy advocacy.
One of the most successful examples in terms of organizational development and impact of their rights advocacy is the network of anti-discrimination organizations, Yirenping. It established a set of independent but networked branches across China that advocated against discrimination against people living with HIV/AIDs or hepatitis B or C, and extended their work to include combating gender-based discrimination.
Their growth took advantage of the relative political openness from the mid-2000s to early 2010s, and became effective as an advocacy network. For labour activism, the loose network of worker centers in Guangzhou, which were the target of a crackdown in 2015, is another example of labour activists consolidating into organizations. In South China, Hong Kong–based labour organizations have played an instrumental role in this consolidation.
The professionalization of rights advocacy has its downsides. They have become easy targets of criminalization, and are unavoidably financially dependent on international funds from North America and Europe. Both the Yirenping network and the worker centers were harshly criminalized.16
The professionalization of rights advocacy has its downsides. They have become easy targets of criminalization, and are unavoidably financially dependent on international funds . . .
In contrast, the advantages of the networked model are many but with some important caveats. There is no organization, or center of organizing to be the target of a crackdown. The users who started the mobilization remain anonymous, though the government may still be able to identify and harass them.
While this does not mean a crackdown cannot be effective in silencing the mobilization. The arrest of a few active organizers could still very well threaten others into inaction. But it is nevertheless much harder than closing down organizations.
The tech worker organizing model has another advantage: it needs much fewer resources, compared to maintaining offices and full-time staff at worker centers. And this is important, given the fact that China’s Foreign NGO Management Law (in place since 2017) has very successfully limited the acquisition and use of foreign funding.17 While the new law hinders NGO-based activism, it may have the unintended consequence of encouraging more embedded activism in the workplace.
Which Path for Labour Organizing?
Should labour and other activists consider this decentralized and networked model of organizing as an alternative to the civil society organizational model? This debate is certainly not limited to China. The non-profit sector in the United States and Europe has been subject to critiques, among many things, for not being effective vehicles of social change.
However, in the Chinese context, the rise of worker centers is in part the result of the state blocking independent union organizing, making NGOs an important alternative means of protecting workers.
Still, in China itself, criticisms of worker centers are easy to come by, ranging from inefficiency and mismanagement to their inability to build worker power and solidarity. While they may contain the seeds of mass worker organizations, on their own they have not built a mass base of workers.
The restricted organizing space in China necessitates the exploration of an alternative model of organizing, such as the networked model, which could be adapted to the specific configuration in which organizers find themselves. The tech worker mobilization should encourage labour organizers to explore a new flexible and networked structure of mobilization, while feminist and anti-discrimination advocacy may also offer useful lessons for labour organizing.
The tech worker mobilization should encourage labor organizers to explore a new flexible and networked structure of mobilization . . .
There are further questions to consider related to the fact that decentralized organizing may be well-suited to mobilizing, but less able to maintain gains. In social movements, there are moments of mobilization and consolidation: What happens when the mobilization for a particular campaign ends? What is left of the impressive mobilization and lessons learned? How is the organizing to be consolidated and sustained beyond single mobilizations? Can this be done without civil society organizations? And if not civil society organizations, what other forms of organizations could emerge? These are perennial questions for labour organizing in China.
It was already challenging when independent labour organizations in the form of worker centers were tolerated to operate and consolidate organizing experience. It has become much harder now to conceptualize and envision how labour organizing could consolidate and build organizations.
The decentralized organizing model cannot answer all these questions. Indeed, the model can only be sustainable over time when it is based on and is able to provide further momentum for deep workplace-based organizing, instead of becoming a substitute for it.
In the short term, the current political context makes any labour organizing exceedingly difficult, and we should not at all underestimate the power of state repression to curtail labor organizing in China. But political space is not given but fought for, and organizing is key to create more space.
The decentralized organizing model, especially if it is built on workplace organizing, could help create the much needed space and more national attention for worker activism. Ultimately, any solution will depend on workers’ and activists’ ingenuity to experiment with different forms of organizing and to learn from those experiments.
Kevin Lin is a researcher at Hong Kong University. He studies the labor movement and civil society in China. This article first appeared on the New Labor Forum website. Reproduced with the author's permission.
Footnotes 1. Kevin Lin, Rising Inequality and Its Discon-tents in China, New Labor Forum, December 8, 2016.
2. The Jasic Workers’ Struggle in China, Labor Notes, November 5, 2018.
3. Julianne Tveten, Daniel in the Lion’s Den: Platform Workers Take on Tech Giants in the Workplace and the World, New Labor Forum, January 24, 2019.
4. Employee Activism Is Alive in Tech. It Stops Short of Organizing Unions, The New York Times, July 8, 2019.
5. J Chen, Tech Workers Need to Keep Organizing, Jacobin, November 15, 2018.
6. ‘996’ Is China’s Version of Hustle Culture. Tech Workers Are Sick of It, The New York Times, April 29, 2019.
7. China’s White-Collar Workers Earned Less in First Quarter of 2019 Despite Signs of Economic Recovery, Survey Finds, The Guardian, April 30, 2019.
8. Working 9 to 9: Chinese Tech Workers Push Back against Long Hours, The Guardian, April 15, 2019.
9. China’s Slowdown Already Hit Its Factories. Now Its Offices Are Hurting, Too, The New York Times, accessed March 14, 2019.
10. Li Xiaotian, The 996.ICU Movement in China: Changing Employment Relations and Labour Agency in the Tech Industry, Made in China Journal, June 18, 2019.
11. Bin Wang and Catherine Driscoll, “Chinese Feminists on Social Media: Articulating Different Voices, Building Strategic Alliances,” Continuum 33, no. 1 (2019): 1–15.
12. 996icu/996.ICU, GitHub.
13. Chinese Court Sentences Activist Who Documented Protests to 4 Years in Prison, The New York Times, August 4, 2019.
14. Kevin Lin, In China, Walmart Retail Workers Walk Out over Unfair Scheduling, Labor Notes, July 1, 2016.
15. Kari Paul, Microsoft Workers Decry Grueling ‘996’ Working Standard at Chinese Tech Firms, The Guardian, April 23, 2019, sec. Technology.
16. China Raids Offices of Rights Group as Crackdown on Activism Continues, The New York Times, March 26, 2015.
17. Clampdown in China Restricts 7,000 Foreign Organizations, The New York Times, accessed April 28, 2016.