Abstract: Asymmetric information about market participants’ valuations and costs plays a crucial role in the efficiency of a platform’s design. Using novel data from a ride-hailing platform called inDriver, I examine whether decentralizing the pricing mechanism improves market efficiency. Unlike its competitors, inDriver requires riders to offer a price for their requested trips, and allows drivers to either agree to the offer, ask for a higher price, or ignore the request. Under this mechanism, a rider with a high willingness to pay for a trip can offer a higher price to increase her chances of being matched. At the same time, under decentralized pricing riders might shade their valuations, which can result in lower average prices on the platform. To understand welfare implications for riders and drivers, I develop an equilibrium model of a decentralized ride-hailing market and estimate its parameters using user-level data on the universe of ride requests in a single city. The obtained estimates are then used to compare welfare under a decentralized mechanism to an alternative mechanism in which prices are chosen by the platform. I find that decentralized pricing significantly improves efficiency in the studied market.
"Work Flexibility and Labor Supply: Quasi-experimental Evidence from a Ride-Hailing Platform" (draft available upon request)
Abstract: The platform’s fee structure alters participants’ incentives and shapes market functioning. I examine the effects of a change in the fee structure implemented by a ride-hailing platform, where drivers’ fees were changed from an upfront charge for a fixed amount of time on the platform to a per-ride commission. The upfront participation fee leads to a less flexible arrangement, which is likely important for part-time workers. Exploiting the fact that the change was implemented as a city-wide staggered rollout, I study how much the switch affected the number and types of participating drivers. I find that the switch to a per-ride commission system led to a sharp increase (18.5%) in the number of drivers participating daily on the platform. However, the rise in the number of drivers does not lead to market expansion, does not affect equilibrium prices, and is paired with lower matching rates for the requests (2 pp). Following the switch from the upfront fee to commissions, significant reallocation of the requests occurs: each driver completes fewer requests and experiences lower earnings after the switch. Using granular data on drivers’ behavior, I then explore the mechanisms responsible for the observed reallocation and lowered matching rates. I find that the crucial factor is the readjustment of incumbents’ labor supply. This readjustment happens due to the change in monetary incentives rather than the change in the competitive landscape. I also find that the new drivers are different from the incumbents and cannot fully substitute for lost labor hours due to the change in the compensation.
Work in Progress
"Enhancing Market-Thickness through Batching: An Experiment in a Two-sided Market"