Approach and avoidance mechanisms in the effects of algorithm transparency on the service performance of food-delivery riders

The advent of the digital era has seen algorithmic decision-making become ubiquitous in online service platform management. However, the lack of algorithm transparency has been widely reported as a barrier to successful human-machine collaboration. Drawing on self-regulation theory, this study explores the effects of algorithmic transparency on the approach and avoidance behaviours of 265 food-delivery riders on online platforms in China. Results from our two-wave survey show that algorithm transparency can improve the service performance of food-delivery riders by inspiring their task crafting along an approach path and reducing their negative work rumination along an avoidance path. The extent to which riders perceived the algorithmic allocations of their work tasks as just or unjust have an effect on whether riders engaged with task crafting or negative rumination. Specifically, perceiving low algorithmic injustice can strengthen the effect along the approach pathway and weaken the effect along the avoidance pathways. This study advances empirical research on algorithmic transparency in the fields of platform economy and service management, while providing practical insights for platform companies to develop, design, and apply algorithmic systems in ways that mutually benefit themselves and their workers.