Every American knows what “fuel mileage” is—especially these days. Whether it’s the EPA test results on the window sticker, or the actual mileage you figured out for your car, we’re all asking: How many miles can we go on a gallon of fuel?

In Europe and Asia, though, they flip it around, and say “fuel consumption.” That means: How much fuel does it take to go a set distance? There, it’s liters per 100 kilometers—we’d say gallons per 100 miles.

Frankly, this method makes far more sense when you’re talking about saving money on fuel, reducing CO2 emissions, or working to displace petroleum for national security reasons.

A new study, released yesterday by Duke University’s Fuqua School of Business, demonstrates the damage done by discussing mileage rather than consumption. Experiments by management professors Richard Larick and Jack Soll show that consumers believe fuel consumption is reduced at an even rate as mileage increases. For instance, most people ranked an improvement from 34 to 50 mpg as saving more gas over 10,000 miles than an improvement from 18 to 28 mpg.

In fact, that’s exactly backwards. Using rough numbers to make the math easier, here’s how it works: If a Chevy Tahoe improves from 10 mpg to 20 mpg, your fuel usage over 100 miles drops from 10 gallons to 5, saving you five gallons. Switching from a
Toyota Corolla
at 25 mpg to a Prius at 50 mpg, on the other hand, saves only two gallons for that same 100 miles, down from 4 gallons to 2.

Looking at the amount of gas (or diesel) needed to go a set distance makes it easier to calculate the effects of any consumption-reducing change. If window stickers showed how many gallons it would take a car to go 10,000 miles a year, buyers could readily calculate and compare the car’s annual fuel cost.

And in fact, that’s exactly what happens. “When you use that measure,” says Professor Soll, “people get all of our questions right.”

Among other effects, this means that for the greenest of car owners, trading in an already high-mileage Prius for a plug-in hybrid may save relatively little petroleum, depending on how the car is driven—although electricity is far cheaper per mile than is gasoline. Getting that Texan rancher to take the mileage on his big SUV or pickup from terrible to less-terrible, however, saves quite a lot.

And as Honda’s John German points out, the battery advances necessary to make plug-in hybrids practical will trickle down to regular hybrids—and they’re likely to improve the economics considerably. This leads to the conclusion that if we persuade 10,000 less-advanced buyers to buy hybrid versions of their current vehicles, we’d get better results—at a far lower cost in batteries—than from selling 10,000 (very pricey) plug-in hybrids to the greenest drivers.

Unfortunately, this argument isn’t (yet) working in the market. Dealer stocks of Toyota Priuses are measured in “hours, not days,” according to one dealer recently quoted, while GM has managed to sell fewer than 1,500 of its full-size Tahoe and Yukon Hybrid SUVs.

In reality, carmakers will likely try pursue all of these alternatives. But the U.S. definition of “mileage” sure doesn’t make the math any easier for consumers—and it may actually impede progress.