As the world begins a move to electric vehicles, we are working out how to quote their fuel economy. Remember when GM announced their new electric Chevy Volt was going to get 230 MPG? What they did was consider a trip of a certain length (set by the EPA fuel economy driving cycle test) and only count the gas used. Because the first part of the trip could use electricity from the battery, the first miles were "free" from the point of view of the gas tank. Of course that is very misleading, because the battery is still providing energy.
It looks like the EPA is going to regulate how automakers must report electric car fuel efficiency. I see that we are going to need it ... the automakers have already shown they are not responsible enough to manage this on their own. They have let the marketing department take the driver's seat and publish information that is not useful to consumers.
It seems that the EPA is still thinking about the best way to present electric car fuel economy data. One suggestion is to report two numbers: one giving the gas (normal internal combustion engine) fuel economy, and the other the energy use of the electric engine. At the root, I think this shows us that we should move away from thinking about gallons and towards thinking about Joules, the physical unit for raw energy. The most useful number would be energy used per distance traveled. In modern international scientific units this probably would be expressed in Joules per meter. It could be given as millions of Joules per 100 miles or something else a little more familiar to the American public.
At the end of the day what is being spent to move the vehicle is energy. In most circumstances the distance we want to go is the fixed given information. The natural measure combining these two is energy per distance traveled. Simple and it works for all types of vehicles: electric, hydrogen fuel cell or regular gasoline.