One major flaw in the analysis: The conclusion is based on the principle that most car-share users are non-car-owners who already drive infrequently–fair enough. It then poses the hypothetical about 20% of the country switching to car-shares, and still claims that the energy savings are a drop in the bucket.
The flaw, of course, is that if 20% of the population were to switch to car-sharing, you’d see significant number of car owners using car-sharing services. But why would a car-owner use a car sharing service, other than perhaps to rent a vehicle type they don’t have in the garage (say, a SUV to haul lots of stuff)?
Well, they just might if they got rid of their car.
This is the key potential win for car-sharing services: it permits people to reduce the number of cars they own. Many people live in relatively transit-rich areas, and aren’t opposed to biking or walking–but still have cars and use them for most trips. Why? Ignoring cultural reasons, many either don’t trust the availability of the transit system, or have need to make trips (such as hauling lots of stuff) that can’t be made easily without a personal vehicle. And so they buy cars. And once they have a car in the garage, it is often more economical to go ahead and use it for everything.
The whole point of car-sharing services, from an environmental point of view, is that it enables people to sell their cars. Once you no longer have one parked in the garage, driving is expensive. But if you already own a car, you’re already paying most of the costs for it (depreciation being the biggest expense) regardless of whether you use it or not; at that point the marginal cost of driving is small.