This dissertation examines the impact of labor unrest under authoritarianism. It uses evidence from China to explore the possibility that autocracies, especially state socialist and post-socialist ones, are uniquely vulnerable to worker resistance and therefore react to it in a dual manner, at once repressive and responsive. Drawing on an original dataset of strikes, protests, and riots by Chinese workers, I find that increases in unrest are correlated with both increases in public security spending (repression) and pro-labor rulings in formally adjudicated employment disputes (responsiveness). Using a “most similar” case comparison informed by field theory, I then show how in Jiangsu’s portion of the Yangtze River Delta, moderate industrial contention is paired with governance that can be characterized as preemption, caution, and nudging, while in Guangdong’s portion of the Pearl River Delta, high contention is paired with reaction, experimentation, and crackdowns. Thus, consistent with the dissertation’s quantitative analysis, repression and responsiveness are stronger where resistance is more widespread, but governance is also qualitatively different. I argue that, at the level of local governments and local officials, there is a logic to this divergence between the cases: militant workers make a liability of the state’s commitment to stability, thereby threatening the careers of officials, who must, as a consequence, demonstrate grit and creativity. Two issues remain: the role of regional political elites and diffusion between regions. I find that the ideologies of particular local politicians do not alter the protest-policy relationship. However, even as protest tactics diffuse outward from the country’s hotspots of contention, official counter-measures against protests also spread. Worker resistance thus profoundly influences authoritarian rule, but not in a linear manner.