Yuxiao Fu (Peking University)
Dandan Zhang
This paper studies how gig workers adjust their labor supply in response to unanticipated negative income shocks, and what these responses reveal about the underlying behavioral foundations of labor supply. Using transaction-level administrative data from Chinas second-largest food delivery platform, we examine how delivery riders react to transitory income losses caused by low ratings, complaints and cancellations based on a discrete stopping-choice framework that models both extensive and intensive margins of labor supply within work shifts. The results show that unanticipated shocks induce significant loss-aversion effects among gig workers, while riders responses to anticipated fluctuations are fully consistent with predictions of the neoclassical model. Using historical complaint and low-rating data from merchants and customers as instruments, the results continue to show a significant extension in labor supply. Further analyses document that shock responses are strongest when losses occur recently and when riders are close to their income targets, suggesting dynamically adjusting reference points. Heterogeneity analyses indicate that full-time and less-experienced gig riders are particularly sensitive to negative shocks. These findings carry implications for platform incentive design and labor regulation in gig work environments.