Balancing Efficiency and Safety: How and When Algorithmic Management Induces Gig Workers' Unsafe Behavior

Abstract

As the gig economy expands, millions of food delivery riders rely on gig platforms for their livelihoods, yet this growth has also been accompanied by rising traffic violations and accidents, posing risks to both rider and public safety. It is therefore critical to understand not only the mechanisms driving gig workers' unsafe behavior but also the factors that may mitigate it. Drawing on goal conflict theory and the job demands–resources model, we examine the mediating role of performance–safety goal conflict in the relationship between algorithmic goal setting and unsafe behavior, and further test a dual‐stage moderated mediation model in which algorithmic monitoring and conscientiousness function as boundary conditions. To test our hypotheses, we conducted four interrelated studies using a multi‐method approach: Study 1 employed LLM‐based text analysis ( N = 657), Study 2 adopted a video‐based scenario experiment ( N = 140), Study 3 implemented a three‐wave survey ( N = 242), and Study 4 incorporated objective behavioral data of unsafe behavior ( N = 151). Across these studies, the findings consistently demonstrate that algorithmic goal setting intensifies gig workers' performance–safety goal conflict, which in turn increases unsafe behavior. Moreover, algorithmic monitoring amplifies the effect of algorithmic goal setting on performance–safety goal conflict, whereas conscientiousness serves as a critical personal resource that mitigates the impact of performance–safety goal conflict on unsafe behavior. This study advances existing research by revealing how algorithmic management contributes to gig workers' unsafe behavior and offers practical implications for reducing such risks through both the optimization of algorithmic systems and the cultivation of gig workers' conscientiousness.