Julian A. Gomez-Gelvez

Ph.D. Candidate

Agricultural and Resource Economics

University of Maryland, College Park

jgomezge@umd.edu

Congestion charges under market power: An application to ride hailing in Bogotá, Colombia

Job Market Paper

The rapid growth of ride-hailing services over the last decade has caused concerns about their potential to worsen traffic congestion. Economists usually prescribe a Pigouvian tax or congestion charge, equal in size to the marginal external cost of congestion, to treat this illness and contain excessive growth. However, ride-hailing markets suffer from another ailment in that they are usually concentrated in the hands of very few digital platforms like Uber. Platforms can then exert market power and raise prices above competitive levels. Under these two conditions (negative externalities and market power), the size of the optimal congestion charge is less than marginal external cost and may even turn negative. In this paper, I build a structural model of ride hailing to compute the optimal congestion charge for a ride-hailing market monopolized by a digital platform. I calibrate the model to the morning peak period in Bogotá, Colombia, in 2019 and find that the optimal congestion charge corresponds to 62% of the marginal external cost of congestion, 17% of the rider fare or COL$240 per kilometer. This optimal charge takes into account the price reduction that the platform executes as a response to the charge, which causes an incomplete pass-through of the charge to riders. The structural model has four components: (i) a demand model where travelers choose between ride hailing and other modes based on price, in-vehicle time and wait time, (ii) a supply of drivers that adjusts to achieve a fixed revenue per hour, (iii) a matching process between riders and drivers in which riders can observe the wait time to the closest idle vehicle before deciding to hail a ride, and (iv) an empirical estimate of the marginal effect of additional vehicles on traffic congestion.