Chunna Shi, Di Zhang, Junping Ma, Jianping Hou, Li Cao, Yong Yang
Algorithmic management is central to platform gig work, but its distinct forms of control may relate differently to worker strain. We examine these relationships in food delivery, a typical form of platform gig work. In this setting, algorithms continuously allocate tasks, track performance, evaluate workers, and shape behavior. The Stressor-Strain-Outcome (S-S-O) model provides the overarching framework. The Job Demands-Resources (JD-R) perspective explains why the three control dimensions may show different associations, while cognitive appraisal theory provides a supplementary lens. Specifically, the study links standardized guidance, tracking evaluation, and behavioral constraints to work engagement through role stress and emotional exhaustion, with overall social support as a moderator. Survey data from 608 food-delivery workers were analyzed using partial least squares structural equation modeling (PLS-SEM) with 5000 bootstrap resamples. Standardized guidance was negatively associated with role stress and emotional exhaustion, whereas tracking evaluation and behavioral constraints were positively associated with both. Role stress and emotional exhaustion each carried significant indirect associations with work engagement; the serial indirect associations were significant but small. Overall social support strengthened the positive associations of tracking evaluation and behavioral constraints with both strain variables. For standardized guidance, however, higher support strengthened, rather than weakened, the negative associations with strain. Predictive assessment, control-variable models, and covariance-based SEM sensitivity analyses produced broadly consistent findings. These results show that distinct forms of algorithmic control relate differently to strain. Because the data are cross-sectional and were collected from food-delivery workers, causal and temporal inferences, as well as generalization across occupations, remain limited.