Judging fuel efficiency using MPG apparently leads most of us to misjudge the benefits of increasing fuel efficiency for the least efficient cars.
Miles per gallon tells us how many miles we can drive on one gallon of gas. But who has a sense of their gas usage?
What we do have a clear sense for--what we most clearly experience--is how many miles we drive. So the measurement that best matches how we intuitively measure our fuel use is how many gallons it takes to drive a certain number of miles (gallons per mile rather than miles per gallon).
The economists who conducted these experiments found that the MPG measurement is misleading because people don't realize that trading in a car that gets 18 mpg for one that gets 28 mpg is a bigger jump in fuel efficiency (twice as big!) than trading in one that gets 34 mpg for one that gets 50 mpg. The implication for policy is that increasing the fuel efficiency of behemoth gas guzzlers is more effective than incrementally tweaking hybrids.
This would be a great problem for a critical thinking course!