Asymmetric Taxation, Pass-through and Market Competition: Evidence from Ride-sharing and Taxis

When deciding whether to tax a product, policymakers need to consider the effects of the intervention on price and output. While the existing literature focuses on “symmetric taxes,” i.e., on schemes imposing the same tax amount on all products in the market, in practice, some taxes are “asymmetric”, in the sense that they levy different amounts on competing products, or even affect only a subset of them. Since the equilibrium responses to asymmetric taxes are going to affect the relative prices of different products in the market, asymmetric taxes can have large effects on competition and market outcomes. The goal of this paper is to study empirically the effects of asymmetric taxes on market outcomes. In January 2020, nominally motivated by a desire to reduce congestion and raise revenues for the city budget, the city of Chicago taxed ride-sharing (e.g., Uber and Lyft) but did not impose any surcharge to riders taking rides with traditional taxis. By taxing ride-sharing up to $3.00 per trip, this policy imposed the highest surcharge faced by ride-sharing companies in the US, providing an informative case study to investigate the consequences of asymmetric taxes. First, I investigate the pass-through rate of the tax on ride-sharing. Ride-sharing companies are peer-to-peer marketplaces which connect riders to drivers. This generates additional insights to the analysis of pass-through because the reduction in demand following a tax levied on riders can reduce drivers’ willingness to supply rides, which, in turn, increases the equilibrium price of ride-sharing. This means that, in a two-sided market, indirect network effects make it easier to rationalize tax overshifting than it would be in a conventional one-sided firm analysis. Second, I quantify the extent to which the tax shifts demand back to traditional taxis, and how this effect varies across different areas of the city. This sheds light on the substitutability between traditional taxis and ride-sharing and on its determinants. Third, I consider the effect of the policy on congestion. In fact, the exacerbation of congestion is a critical issue in many large and growing metropolitan areas, and congestion taxes have been a widely used tool to reduce traffic. Finally, I study the impact of the tax on consumer welfare, by developing and calibrating a logit demand framework that accommodates asymmetric tax schedules. The results of my analyses provide evidence that the tax on ride-sharing had economically significant effects on markets outcomes. Consistently with the existence of indirect network effects, I estimate pass-through rates exceeding unity on single rides, while for shared rides I find incomplete pass-through. This pattern can be explained by a higher elasticity of demand for shared rides as compared to single ones. The change in prices of ride-sharing shifts demand back to traditional taxis but only for downtown rides, where the number of equilibrium taxi pickups increases by 3.63%. This is consistent with a more intense competition between ride-sharing and traditional taxi services in the downtown area. Overall, the tax reshapes the demand for rides in Chicago without reducing the total daily number of trips across ride-sharing and taxis. Consequently, there is no discernible impact on congestion. Finally, the reallocation of market share to taxis comes at the cost of reducing consumer surplus, with an estimated loss of $1.74 and $0.49 per trip downtown and non-downtown,