The author of the MPG illusion on why we misunderstand the meaning of miles per gallon (and why we shouldn’t completely hate the hybrid Cadillac Escalade)

By Richard Larrick

Jack Soll and I recently looked at people’s intuitions about car fuel efficiency when expressed as miles per gallon (mpg), which is the common measure used in the United States. We realized that the mpg scale was not intuitive. If you are buying or trading in a car, what mpg increases are worthwhile in terms of reduced gas consumption and carbon emissions? Certainly a car that gets 50 mpg looks great compared to one that gets 33 mpg. But many other trade-ins for small improvements didn’t seem worthwhile. Why bother trading in a 16 mpg car for a 20 mpg one? Why bother putting hybrids on huge SUVs (like the Chevy Tahoe or the Cadillac Escalade), increasing their mpg from 12 to 14? What’s the environmental payoff?

Surprisingly, however, for the same distance driven, each of the improvements listed above is equally beneficial in reducing gas use. They all save about 1 gallon over 100 miles and 100 gallons over 10,000 miles (with a little rounding). Without question, 50 mpg is the most efficient level and ideally everyone would strive for it. But, if we are simply considering changes to existing vehicles, 16 to 20 mpg can help save as much gas as 33 to 50 mpg.

In short, although mpg always tells you which car is most efficient it obscures the value of improvements as fuel efficiency improves, leading people to undervalue small mpg improvements on inefficient cars, and overvalue large jumps between efficient cars. We call this effect the “MPG Illusion,” and published a paper about it in the journal Science last week. So don’t dismiss that hulking Escalade. Gallon-per-gallon, it’s a big improvement.

The Math Behind the MPG Illusion

Of course, none of our conclusions are too surprising once you apply a little math. Miles per gallon is a ratio. Gas consumed is an inverse of that ratio. A ratio and its inverse do not have a linear relationship. They have a curvilinear one, as shown in the graph below, plotting gallons used per 10,000 miles driven (10,000 divided by mpg) by mpg.

We suspected that most people would not spontaneously think in terms of this curvilinearity and would, in fact, think that fuel consumption decreases as a linear function of MPG. In our paper, two studies confirmed that people reason in a linear fashion; a third study showed that their misperception could be corrected by using gallons per 100 miles as the measure of fuel efficiency. (For a full description of all three studies, click here.)

A Nudge for Car Shoppers

The mpg illusion suggests an obvious nudge: Start expressing fuel efficiency as gallons per mile (gpm) for some standard (meaningful) distance. We prefer gallons per 10,000 miles because it makes clear that seemingly small mpg improvements are valuable. Here’s an abbreviated table of what gpm numbers would look like:

10 mpg = 1,000 gallons per 10,000 miles
15 mpg = 667 gallons per 10,000 miles
20 mpg = 500 gallons per 10,000 miles
25 mpg = 400 gallons per 10,000 miles
30 mpg = 333 gallons per 10,000 miles
35 mpg = 286 gallons per 10,000 miles
40 mpg = 250 gallons per 10,000 miles
45 mpg = 222 gallons per 10,000 miles
50 mpg = 200 gallons per 10,000 miles

Gallons per mile should help consumers recognize the value of replacing inefficient cars in the 10 to 20 mpg range. Ten thousand miles not only makes the difference look large, it is close to the national average for yearly driving. It’s a realistic estimate of fuel consumption for many drivers. And, because the number is round, it is pretty easy to adjust up or down for actual driving. The perfect gpm expression, however, would be to a personal distance. We’ve created an excel worksheet like this that is available on the MPG Illusion web site.

We prefer putting gas consumed in the numerator, rather than cost, because gas prices can change substantially in a short period of time. Also, a direct measure of gas consumption is just one step away from calculating greenhouse gas emissions from burning gas (one gallon equal a little over 20 pounds of carbon dioxide).

The gpm measures listed above also point to a clear policy implication: Replacing the most inefficient vehicles, and improving their mpg by a seemingly small amount, can make a big difference in total gas consumption. In fact, it can make a larger difference than replacing a Honda Civic with a Toyota Prius.

Clearing Up Public Misconceptions

A few misconceptions have come up when people have reacted to this proposal:

1) We are not proposing to replace mpg, just to supplement it. Miles per gallon is very useful once you own a car and need to know the range of the gas tank. However, gpm is more useful when deciding about replacing a car or choosing between two cars. Both measures are useful at different times.

2) The metric system does not solve the mpg illusion. India uses kilometer and liters but expresses efficiency as kilometers per liter. Because the ratio is distance over volume, it creates a parallel illusion to mpg. See this blog for a nice translation to the Indian context.

Many countries currently use liters per 100 kilometers, which has the right numerator and denominator. However, some people living in those countries have questioned how helpful it has been. We think that the base distance should be larger so that differences between efficiency levels are clearer and involve fewer decimals.

3) Displaying the percentage increase in mpg does not solve the MPG Illusion either. Many people look at a 50 percent mpg improvement, such as 33 to 50 mpg, and assume that it will save more gas than a 30 percent mpg improvement from 10 to 13 mpg over the same distance. A quick check of the math will show that 10 to 13 saves 230 gallons over 10,000 miles; 33 to 50 mpg saves only 100 gallons over the same distance. Percentage increase in mpg, like linear increases, is a fallible indicator of improvements in efficiency. The problem with percentage reasoning is that the percentage change has to be applied to an initial level of gas that is being consumed.

One of the beauties of gpm is that private organizations can take the lead in promoting awareness about the mpg illusion. If Consumer Reports, Edmunds, and car manufacturers started using gpm to express fuel efficiency, the actual value of replacing inefficient cars with more efficient cars would be clear. Car buyers could make better decisions to their own benefit, as well as ours.

Addendum by Richard Larrick: I will add that the Cadillac Escalade is acceptable only if it is replacing identical miles from a worst car! Otherwise, forget it.

Addendum by Nudge blog: Mahesh Sethuraman thinks cars should come with a cost meter that calculates the cost of each mile travelled based on the cost of gas and the vehicle’s current mileage.

Imagine a “cost meter” (which takes the cost of petrol/diesel and the mileage of the vehicle as inputs) connected to all the motor vehicles which displays the cost of every meter of travel just like a “speedometer” and as an add on it could have a sensitivity meter which shows what would be the cost of every meter of travel if the petrol/diesel price rises by another 10%.A similar device for cooking gas and for that matter for usage of any non-renewals sources of energy. Maybe just maybe this might instill a greater discipline and judiciousness in consumption of scarce, non-renewable sources of energy. If it does, then this would be a very effective way for disciplining consumption to a level which would have otherwise taken another 5rs hike in prices. And even if it doesn’t, this experiment isn’t very expensive!