The Dual Effects of Algorithmic Management on Platform Workers: An Attribution Perspective

ABSTRACT Existing research often highlights the negative consequences of algorithmic management (AM) for platform workers. By contrast, less is known about what, how, and when AM may produce both positive and negative outcomes. Drawing on attribution theory, this study examines the dual effects of core AM dimensions (i.e., algorithmic recommending, restricting, evaluating, and rewarding) on platform workers' perceptions of work overload and customer‐oriented service behavior. A two‐wave survey of 213 online platform workers in China reveals that algorithmic recommending and rewarding improve customer‐oriented service behavior and reduce work overload through AM commitment attributions. However, AM control attributions link algorithmic restricting, recommending, and evaluating (the latter two at low algorithmic transparency) to increased work overload. Algorithmic transparency moderates these effects, reducing the negative impacts of AM through AM control attributions. These findings contribute to a more nuanced understanding of the dual effects of core AM dimensions and provide practical insights for platforms seeking to enhance service quality while supporting worker well‐being.