Working Paper: NBER ID: w28145
Authors: Fernando E. Alvarez; David O. Argente
Abstract: We use three quasi-natural experiments in Mexico and one in Panama to estimate the effects of having the option to pay with cash on Uber rides. The ability to pay in cash affects the demand for rides, which is reflected in large changes in the total number of trips, fares, miles, and number of users after Uber introduced cash payments, particularly in lower-income city blocks. On the other hand, the effects on prices, estimated times of arrival, and competitor pricing are negligible, consistent with the supply of trips being very elastic. Although cash payments naturally increase the fraction of users that pay exclusively with cash, more than half of the users have access to both cards and cash, and alternate between payment methods. We find evidence consistent with cash and card payments being imperfectly substitutable at both the intensive and extensive margins, which magnifies the impact of policies that restrict the availability of payment methods.
Keywords: cash payments; Uber; ride-hailing services; financial inclusion; payment methods
JEL Codes: E41
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| cash payments (J33) | demand for Uber rides (R22) |
| cash payments (J33) | trip quantities (C35) |
| cash payments (J33) | fares (R48) |
| cash payments (J33) | user behavior (D10) |
| ban of cash payments in Puebla (E42) | trip reductions (R41) |
| ban of cash payments in Panama City (F38) | trip reductions (R41) |
| cash payments (J33) | prices (P22) |
| cash payments (J33) | estimated times of arrival (C41) |
| cash payments (J33) | competitor pricing (L11) |
| cash payment bans (E42) | demand for Uber rides (R22) |