Drivers of disruption? Estimating the Uber effect

A frequent belief is that the rise of so-called “gig work” has led to the displacement of workers in a wide range of traditional jobs. This paper examines the impacts of the flagship of the gig economy-Uber-on workers employed in conventional taxi services. Our analysis exploits newly collected data on the staggered rollout of Uber across metropolitan areas in the United States and a difference-in-differences design to document that incumbent taxi drivers experienced a relative earnings decline of about 10 percent subsequent to Uber's entry into a new market, while there are no significant effects on their labor supply. Additional evidence from a battery of placebo tests, event study estimates, and specifications using Google Trends data to capture differences in treatment intensity underlines these findings. A triple-differences design that compares changes among taxi drivers relative to bus, tractor, and truck drivers that were unaffected by the arrival of Uber, provides further supporting evidence that the diffusion of Uber has reduced the earnings potential of incumbent drivers in conventional taxi services in the United States.