Step on It: Approaches to Improving Existing Vehicles’ Fuel Economy

In this paper we attempt to understand how policy interventions could improve the fuel economy of drivers without changing the vehicles they drive or the trips they take. To do this, we use high-resolution driving data to document the large variation in actual on-road fuel economy achieved by drivers of identical vehicles. We analyze this variation using a model that links driver behavior to vehicle fuel consumption. The physical vehicle model, estimated econometrically from data on acceleration and speed choices, predicts observed fuel consumption extremely well. We combine this physical model with a model of optimizing behavior by drivers, in which drivers tradeoff higher fuel consumption for shorter trip times. Using the variation in value of time for the drivers in our data set, we simulate route choice and driving behavior on a stylized set of trips. We find that drivers have little incentive to change their driving behavior to improve fuel economy, since the time cost of driving more efficiently generally outweighs the value of fuel savings. This means that gasoline taxes, while internalizing environmental externalities, are not as effective at reducing overall fuel use as policies that allow for smoother traffic flow (fewer accelerations and decelerations) such as stoplight timing, infrastructure improvements, or the increased use of vehicle automation. ∗Department of Economics, University of Arizona. Email: [email protected] †Department of Economics, University of Michigan. Email: [email protected].