How Do Platform Participants Respond to an Unfair Rating? An Analysis of a Ride-Sharing Platform Using a Quasi-Experiment

The rise of peer-to-peer marketplaces has led to many previously unrated commercial transactions being rated. Of course, online rating systems can lead, on occasion, to reviews that are unfair or unrepresentative of the true quality provided. On the one hand, receiving an unfairly low rating once, might induce participants to exert more effort and receive a better rating the next time. On the other hand, it might dispirit participants and make them exert less effort. We use data from a ride-sharing platform in India where driver ratings were made particularly salient to the driver after each trip. We use instrumental variables to isolate the causal effect of receiving an unfairly bad rating. As our exogenous shifter of a driver’s rating, we use whether or not that driver was a replacement for another driver who had previously canceled that particular customer’s ride. We show that if a customer experiences a ride cancellation, they are more likely to unfairly blame the replacement driver. We show that drivers are more likely to respond negatively to a bad rating and receive subsequently bad ratings if they were blameless for the previous negative rating. We show that this effect is larger in contexts where there is a higher potential for an emotional response and when there is a greater need for driver skill in the subsequent ride. Finally, we show that these potentially unfair ratings can lead drivers to leave the platform, suggesting a broader negative effect of unfair negative ratings on platform participation.