Disruption on the Streets : A Case Study on the Impact of Uber’s Entry on the Taxi Business

The disruptive effects of the sharing economy have received considerable attention from the press, researchers and policymakers. We empirically study whether and to what extent the entry of ridesharing programs (e.g., Uber) in a city changes the outcomes and behaviors of taxi drivers in that city. Using unique data that contain high frequency GPS and trip information from all taxis in a Chinese city, we examine the impact of Uber’s entry on taxi drivers’ work output and input behaviors, and importantly, their earnings. We use a “regression discontinuity in time” approach based on the temporal discontinuity in treatment (i.e., Uber entry date), and leverage high-frequency outcome data along with an extensive set of trend and control variables. We find that taxi drivers’ hourly earnings decreased by 6% after Uber entry. This net effect reflects the entry itself, as well as drivers’ responses to it. We show that a reduced number of trips accounts for the taxi drivers’ decreasing income. Further, we find that fulfilled trip distances are shorter after Uber entry, implying that trip characteristics have changed. Additionally, after Uber’s entry, taxi drivers take shorter driving paths for trips that have similar pick-up and drop-off locations as those before entry. We observe differences in the effects of Uber entry on taxi drivers’ behaviors between self-employed drivers (“entrepreneurial small businesses”) and fleet drivers. The former tend to respond differently to Uber entry by searching over longer distances for their next passengers, thereby mitigating the income lost due to the entry. We also find that trips after entry focus more on “parks and outdoor recreation areas” rather than on “tourism-related” locations. In terms of input factors, we find that taxi drivers increase their geographic coverage in the city to find passengers. We check the robustness of our results to various assumptions