Predation, acceleration, and loss of control: a multilevel theoretical framework for algorithmic time politics and the occupational health of platform workers.

In platform labor, algorithms reshape workers' perception and control of time through mechanisms such as dynamic pricing, compulsory task assignment, time-limit compression, and real-time surveillance, giving rise to a novel power formation- "algorithmic time politics." Taking this concept as its analytical core, this article integrates Foucault's disciplinary theory, Rosa's theory of social acceleration, and Karasek's job demand-control model, supplemented by Bakker and Demerouti's job demands-resources model, to construct a multilevel theoretical framework linking algorithmic time control to the multiple health outcomes of platform workers. The core argument of the framework is that algorithmic time politics damages occupational health through three interconnected mechanisms-temporal predation, temporal acceleration and discipline, and temporal loss of control-which form a progressive chain from "the quantity of time" through "the quality of time" to "the sovereignty over time." Drawing on the relevant literature, the framework posits that temporal predation primarily damages physiological health-manifesting as cardiovascular strain and musculoskeletal injuries-through the mediating pathway of chronic fatigue. Temporal acceleration and discipline are theorized to undermine mental health, giving rise to anxiety and burnout via time panic and emotional exhaustion. Temporal loss of control, in turn, is expected to contribute to depression and to heighten occupational injury risk, with learned helplessness and the depletion of cognitive resources as key mediating processes. Interactive effects and dynamic vicious cycles exist among the three mechanisms: temporal loss of control amplifies the physiological effects of temporal predation, while temporal acceleration intensifies the psychological effects of temporal loss of control. The article further discusses moderating variables, including social security, algorithmic transparency, and alternative employment opportunities. It concludes with policy proposals to incorporate "algorithmic time politics" into occupational health risk assessments and to promote "health-friendly algorithmic design," and outlines directions for operationalizing the framework in future empirical research.