Effects of on-demand ridesourcing on vehicle ownership, fuel consumption, vehicle miles traveled, and emissions per capita in U.S. States

Abstract We estimate the effect of on-demand ride-hailing service market entry by Transportation Network Companies (TNCs) Uber and Lyft on per-capita vehicle ownership, energy use, travel distances, and emissions in U.S. states from 2005 to 2015 using a difference-in-difference propensity score-weighted regression model. We find evidence that TNC entry appears to cause a decline in state per-capita vehicle registrations by 3%, on average (95% confidence interval: 0.7–5.5%). Our results regarding travel distances, gasoline consumption, and several air pollutants are not conclusive, but we find evidence of a negative relationship with some EPA-estimated vehicle air emissions, representing a decline of $300 million to $900 million in externalities during the analysis period. However, these air emissions data are modeled, rather than directly measured, so we are cautious about causal interpretations, and uncertainty in the effects of TNC entry on travel distances and gasoline consumption suggests potential additional externality effects that could dominate the air emissions effect.