Collective Action in Autocracies: The Case of Workers and Strikes in China

Many collective action events occur in authoritarian regimes. While some are triggered by shocks and expand as participation snowballs, the majority emerge during unremarkable periods. These events draw participants primarily through personal social networks, since formal mobilizing organizations face systematic repression under authoritarian rule. Given the risk of repression, people rely on close contacts when deciding whether to participate. However, because the same repression constrains public space and associational life, individuals’ networks tend to remain fragmented, raising doubts about their mobilizing capacity. In such repressive and divided context, how can ordinary people still engage in collective action?

My book project, Collective Action in Autocracies, theorizes that ordinary people can adopt different roles and forge temporary ties to circumvent repression and overcome division. A few individuals can act as bridges by forming temporary ties across groups, while others can participate by mobilizing within their existing group networks. This mode of collective action, which I term networked mobilization, complicates state repression because authorities struggle to prevent individualized, temporary networking efforts, particularly as participation expands. Moreover, the creation of temporary ties generates a self-reinforcing feedback loop: as temporary ties connect groups, many people gain incentives to mobilize within their group networks, while the few individuals acquire further incentives to create temporary ties. These complementary incentives motivate ordinary people to differentiate their roles and coordinate collective action in repressive and divided contexts.

This project advances theories of collective action under authoritarian rule. Existing research rightly emphasizes the crucial role of social networks in collective action. However, under authoritarian rule, social networks are often fragmented because of constraints on public space and associational life. In this context, my model of networked mobilization shows how temporary ties connecting different groups can mitigate state repression and enable collective action. Moreover, it demonstrates that such temporary ties can emerge organically from people’s complementary incentives. This self-reinforcing feedback loop reduces collective action’s dependence on selective incentives or formal leadership, allowing it to persist despite state repression.

Part I of the book project formalizes networked mobilization in a game-theoretic model to identify the conditions under which ordinary people voluntarily differentiate their roles and forge temporary ties. Building on a global game, I compare networked mobilization—when temporary ties across groups are created—with spontaneous mobilization lacking such ties. I show that, as the self-reinforcing feedback loop unfolds, networked mobilization increases participation in collective action. The model’s extension further suggests that experiences of networked mobilization can diffuse and facilitate collective action in other settings. To develop the model, I conducted a factory ethnography in Guangdong, an industrial province of China, from May to November 2019. This participant observation helped me identify physical, organizational, and relational aspects of factory life that enabled workers to coordinate strikes under close employer surveillance. It also inspired me to extend the analysis to workers and strikes across China.

Part II analyzes the implication of networked mobilization for workers and strikes. First, in Guangdong, I find that worker organizers—especially those with more experience in engaging their peers—encouraged broader participation in strikes. They also inspired some workers to become organizers and expanded worker networks across additional cities. However, these trends were reversed after a government crackdown on organizers in 2015. Second, using a difference-in-differences design with a province-year panel of strikes across China, I find that crackdowns targeting organizers reduced the frequency and size of strikes both in the provinces where organizers were targeted and where targeted organizers came. The findings suggest that organizers’ mobilizing influence spread from their host to origin provinces, such that targeting organizers discouraged workers from striking in both places. However, the negative impact was weaker and shorter-lived in the organizers’ place of origin, revealing the limitations of government crackdowns. For these analyses, I constructed two original datasets: annual worker networks from 2011 to 2018 based on strikes in Guangdong, and a province-year panel of strikes between 2011 and 2019 across China.

Part III examines the distribution of strikes and other resistance tactics in China’s hinterland—the places of origin of most workers. I find that strike growth in the hinterland was positively correlated with the number of workers who migrated out and later returned, even after controlling for industrial growth. To understand how domestic labor migration relates to strikes in the hinterland, I continued my ethnographic study in Jiangxi, an inland province adjacent to Guangdong, working in a factory from November 2019 to January 2020. Most workers in the Jiangxi factory were locals, but they had extensive migration experience in industrial hubs such as Guangdong. These workers drew on their migration experience and local resources to challenge employers’ discriminatory narratives and exploitative practices, opening new fronts for labor struggle in China.