Surge pricing and consumer surplus in the ride-hailing market: Evidence from China

Surge pricing leads to efficient operations of ride-hailing services, but its implications on consumer welfare are difficult to estimate owing to endogeneity biases resulting from the simultaneity between passenger demand and surge pricing. We address this issue and present a comprehensive study on consumer surplus estimation in the ride-hailing market using multi-day data of around two hundred thousand daily ride-hailing trips in Nanjing, China. We estimate a relationship between surge level and demand to obtain the elasticity of passenger demand relative to the surge level. Potential endogeneity issues are addressed by applying the dynamic panel generalized method of moments estimation. The price elasticity estimates are then used to compute the consumer surplus. Our results provide three important insights. First, the magnitude of the surge elasticity of demand during peak hours is the lowest compared to other periods. Second, the magnitude of elasticity is higher in urban areas during the off-peak than in suburban areas. Third, the consumer surplus per kilometer is lower in urban areas due to longer trip lengths at higher surge levels than in suburban areas. The consumer surplus generated by a ride -hailing company per yuan (Y) ranges between Y0.47 and Y0.89. Our findings are relevant for both ride -hailing companies and policymakers. While ride-hailing companies can decide areas, time of the day, and the extent of surge pricing based on the spatiotemporal heterogeneity in price elasticity, government agencies can use consumer surplus as a yardstick to devise pricing regulations.

Automated summary

We conducted a comprehensive study on consumer surplus estimation in the ride-hailing market in Nanjing, China, using multi-day data of around 200,000 daily ride-hailing trips. Our results show that the surge elasticity of demand is the lowest during peak hours, higher in urban areas during the off-peak, and the consumer surplus per kilometer is lower in urban areas compared to suburban areas. These findings are relevant for both ride-hailing companies and policymakers.