New Research: ‘Induced’ driving miles could overwhelm potential energy-saving benefits of self-driving cars

Originally published: 
April, 2019

Morteza Taiebat, a doctoral candidate in Resource Policy and Behavior (RPB) at SEAS and a Dow Sustainability Doctoral Fellow, is lead author of a new study on the energy-saving benefits of self-driving cars published April 17 in the journal Applied Energy. Here he shares insights on his latest published study and his team’s ongoing research.

What prompted you to study this dimension of self-driving cars?

We noticed that the existing estimates of energy savings for self-driving cars were all overly optimistic, and we, like many others, were concerned that the overestimated benefits might be used in future policy-making in energy and transportation sectors. The fuel savings of autonomous cars are countered because we are going to drive much more with them. This study tries to shed light on this “energy rebound effect,” showing the trade-off between energy savings and travel behavior change that stems from convenience and lower cost of travel with self-driving vehicles.

Before self-driving cars go to market, what can be done to mitigate the risk of people driving extra miles?

On the engineering and technology side, we have to push for maximum possible energy efficiency in the design of self-driving cars. In fact, we have to push the limits of energy efficiency and strive for average energy savings much higher than 20%.

On the policy side, travel behavior needs to be considered. Mechanisms can be designed to ensure that travel behavior doesn’t change drastically. There are different possible policy mechanisms to control induced travel, and a vehicle mile traveled (VMT) tax is one of the best options. But it should be taken with a grain of salt as it could lead to privacy concerns and adverse equity issues, especially for low income households.

Your study suggests that wealthier households are more likely to drive extra miles. Why?

Because wealthier households have the necessary affluence to be early adopters. In addition, their time cost savings is greater. We used a fraction of wage rate of household as a proxy for their in-vehicle time cost. This is the best available estimate of how people value their time and means that if you aren’t driving, you can recover your own time and use it to work or do other productive activities as you wish. Based on this widely accepted assumption, the wealthier you are, the higher your time cost rate and the more you will recover. Also, wealthier people are more elastic with respect to their travel and will respond more to the changing time cost. Overall, they benefit more from the time cost savings than low income groups do.

What kind of research might build on this study?

Our team’s next step is to try to understand how to design the VMT tax that is economically and socially optimal. The tax must be designed in a way that compensates travel behavior change of higher income households compared to low income households.

What is the key take-away from the study?

The main point of our study is that if you have an average 20% more efficient car and an average 38% time cost savings due to the self-driving feature of the car, all of the energy saving benefits go away. So we need to design self-driving cars as energy efficient as possible, make sure that automakers go toward full electrification of these cars to minimize their environmental impact—and, finally, start thinking about policy mechanisms to mitigate drastic travel behavior change in the age of self-driving cars.


  • Morteza Taiebat
  • Samuel Stolper
  • Ming Xu