U.S. Fuel Economy: Past, Present and Future

03/16/2012 09:20 am ET | Updated May 16, 2012
  • Bill Chameides Dean, Duke University's Nicholas School of the Environment

Crossposted with TheGreenGrok.

To make it to 50 miles per gallon (mpg) by 2025, we'll have to put the pedal to the metal of improved fuel economy.

OK, amidst the discussion of rising gas prices, we have some good news on the car front. The average mpg for cars sold in the United States is on an upward climb, and has been for a while. (See figure below.) Between 2004 and 2011 the average mpg for cars sold in the United States climbed from 19.3 to 22.8.

MPG of U.S. cars 1975-2011, plus future estimates

Average fuel-economy trends for light-duty vehicles for model years 1975-2011 are from the Environmental Protection Agency [xls]. The data are for the final fleet-wide adjusted composite fuel economy for all light-duty vehicles released in the specified model year. (See report [pdf] for definitions.) The 2011 data are estimated. The data points in 2021 and 2025 are the combined average fleet-wide fuel economy that would be achieved by the proposed CAFE standards [pdf] for new model cars and light trucks in those years.

But we've got a long way to go to get to the 50-mpg target proposed [pdf] in the latest round of Corporate Average Fuel Economy (CAFE) standards for vehicle models manufactured from 2017 through 2025, standards that have yet to be finalized.

To meet that proposed goal, automakers will need to roll out new cars each year that, on average, beat the previous year's average by just under two miles per gallon. Is that possible? Well, in 1980 average fuel economy jumped by more than three miles per gallon over the previous year's average. (See chart.) But that was just a single year; to meet the 2025 standard under consideration, continued improvements in fuel economy that average out to almost two miles per gallon per year will need to be sustained over more than a decade. A sobering statistic: between 2004 and 2011, the average increase in fuel economy was only about 0.5 mpg per year.

So what are the odds for this fuel-economy future? It's a question that begets more questions.

Will the new standards be finalized in the coming months? Or in this contentious political climate, will Congress derail them before they're even on the books? Or, if the new standards are adopted, will some administration and/or Congress down the road hang a uey and roll them back? Or will they stay on the books for Detroit to use as a guidepost to get their mpg game on and maybe roll out a breakthrough technology that will make improved fuel economy a walk in the park?

Somewhere, someone has probably already set betting odds on whether we'll hit that goal of 50 miles per gallon or not. And for all I know there's an investment house packaging some kind of futures derivative-thingy that'll allow you to place your bet in a more respectable-sounding manner. Big opportunity to score some major bucks. So here's the final question: which way would you roll?


End Note

A 2008 study ("The MPG Illusion" by Duke's Richard P. Larrick and Jack B. Sol) pointed out that gallons per mile (gpm) provides a better metric for tracking fuel economy than miles per gallon (mpg). For example, suppose you drove 100 miles in a car that got 10 mpg; you'd have consumed 10 gallons. Now suppose you drove those same 100 miles in a car that got 20 mpg instead. You'd have needed only five gallons instead of 10. So a 10-mpg improvement in fuel economy saved you five gallons. But now suppose you drove 100 miles in a car that gets 30 mpg; this time you would have consumed 3.3 gallons. So another 10-mpg improvement in fuel economy saved you this time only 1.7 gallons of gasoline not five gallons. And if you went from 30 to 40 mpg, you would save only 0.8 gallons. Seems like a case of diminishing returns. However, if you use the inverse of mpg, that is gpm, the whole diminishing returns thing disappears and fuel-economy improvements and gasoline savings track one for one. What's going on? For those of you who are mathematically inclined it's a "one over x dependence" issue. For those of you who are not, trust me or better yet see my post on that study.