To improve US fuel economy,
stop talking about MPG!

by Eric Drexler on 2009/03/05

Lamborghini Murcielago Roadster
Two seats, 12 gphm

I want a car that uses less gasoline — fewer gallons per mile — but car makers quote fuel economy in miles per gallon instead. The resulting numbers are counter-intuitive, leading to decisions that create needless waste, cost, and CO2 emissions.

For example, upgrading from a 20 mpg car to a 35 mpg car seems about the same as upgrading from 30 to 50 mpg (“better by 15 mpg”), but in reality, the upgrade from 20 to 35 saves not the same amount, but more than twice as much fuel per trip as the upgrade from 30 to 50. Do the math – a bit twisted, isn’t it?

To make sensible decisions as consumers, investors, and regulators, we need to think in terms of gas consumption, and that’s measured in gallons per mile, not miles per gallon. If fuel economy were measured this way, simple subtraction would give the right answer, and I’d bet that people would choose slimmer gas-hogs and use less fuel — millions and millions of gallons less, just by printing more meaningful numbers on EPA stickers at dealerships.

Some numbers:

To avoid fractions, let’s quote consumption in gallons per hundred miles: gphm. (Europe uses liters per hundred kilometers.) Consider four cars, chosen from the best and worst of their categories as listed at www.fueleconomy.gov:

Vehicle MPG
SUV #1 11
SUV #2 23
Best hybrid 48
Wondercar* 100

* Not on sale

Both SUVs look bad, but not greatly different, while the Wondercar looks like a vast improvement over the hybrid. These appearances are misleading, however. Here are the straightforward, right-side-up numbers:

Vehicle GPHM
SUV #1 9
SUV #2 4.3
Best hybrid 2.1
Wondercar 1.0

Turning the numbers right side up clarifies the situation: Upgrading from SUV #1 to SUV #2 saves 4.7 gphm; dumping SUV #2 in favor of the best hybrid would save an additional 2.2 gphm. Compared to the hybrid, the Wondercar saves 1.1 gphm. Going straight from SUV #2, to the Wondercar would save a total of 3.3 gphm, less than the savings of switching from SUV #1 to #2. (By the way, the best and worst SUVs are both Jeeps, the Grand Cherokee and the Patriot; the hybrid is the Toyota Prius.)

Discussions of fuel conservation would make more sense if we used units that made more sense. Let’s not talk about mpg: it’s confusing, and trading up to gallons per hundred miles costs nothing. Maybe actual gas consumption figures should be on EPA car stickers at dealerships, too.


16 March update: Here’s an article from Duke University on the advantages of using sensible units: Gallons Per Mile Makes More Sense. I’m pretty sure it was this work, reported last June, that prompted me to write on this topic.


{ 5 comments… read them below or add one }

Chris March 5, 2009 at 9:29 am UTC

I agree that the unit of gphm makes sense (as much as it makes sense to continue to use miles and gallons) but I don’t see mpg disappearing anytime soon without some serious effort. This kind of unit might be useful to the US automakers to capitalize on improving the fuel economy of their bad performers, instead of focusing on a single high mpg model (e.g. Volt). Small note – the table entry for the hybrid should read 2.1 gphm.

Eivind March 5, 2009 at 1:52 pm UTC

There’s a typo in your table, you seem to claim that the best hybrid uses more petrol than the SUVs, which is wrong.

Other than that — sensible suggestion. I expect it’ll happen any day now, the same week you actually start metric units, which also makes sense, for quite similar reasons.

Eric Drexler March 5, 2009 at 8:53 pm UTC

@ Chris, Elvind, re. typo — Thanks, fixed. Too much table-fiddling, too late at night.

I don’t expect to see mpg go away, of course, but there’s no great barrier to also quoting gallons per hundred miles, especially in policy-oriented discussions. Using gphm is just another way to use familiar English units, and makes it easier to think about the cost of what we do. Thinking in terms of mpg, people pay too little attention to numerically small differences in the low-mpg range. It’s hard to estimate how many more millions of barrels of oil are consumed as a result.

I wish the US used metric units (and have given up on it), but the advantage here is more fundamental. Neither pounds nor kilograms is actually confusing.

Erin March 7, 2009 at 4:39 am UTC

Chris, Eric, and all: Apart from the really fantastic cars and vehicles we could produce with advanced MNT, what sort of intermediate serious improvements could we see with early nanotech when it comes to cars and vehicles? I would say nanocomposite fullerene and ceramic materials enough would seriously transform autos, what else?

Jake Witmer April 21, 2009 at 10:57 am UTC

Dean Kamen just built a Mini Cooper that runs on a sterling engine, and uses junked food oil and garbage for fuel, and theoretically sells for $5K. I don’t know anything about science, but that sounds cool to me. I personally think that getting the brains where they need to be is more important than building a slightly-better-pre-nanotech car.

After all, there will be near perfect travel (forget cars! Paul Maccready is right, in the environment of the future! …And how long after advanced development will we even be traveling at slower than electron-simulation speeds?) after enough brains meet in the right conditions. So let’s just get the nanotech people meeting with each other, and who cares if they drive there in gas-guzzlers, so long as they can afford to?

…I also don’t care how much gas is burned. If we actually had a free market, we could correct for any problems caused, if evolution didn’t simply keep pace anyway. Heaping power on the already power-drunk and inept government is not a good idea. The EPA does nothing valuable that couldn’t be accomplished voluntarily (unless you think stealing farmers’ land is “valuable”).

Moreover, the problems attendant to too little liberty are unbelievably stifling, and tend to be underestimated by everyone who doesn’t naturally find themselves in a fight to the death with the government.

The fundamental principles serve as a good guide. Government requires the initiation of force for every function it currently serves. If it was voluntary, how much of it would go away? 99%

And we’d all be happier and not so bullied. Eric would be richer and more productive, as naturally good thinkers are rewarded in a rational environment.

I’ve always been surprised at the lack of free market thought on genius-level science boards. Is it true that Kurzweil and Drexler have thought out anarcho-capitalists Lysander Spooner, Carl Watner, and Marc Stevens’ statements and disagreed with them?

I don’t think so. I think rather that they always stop short of thinking them through fully, because of the emotional pain it brings their world view to do so: There is no immediate payoff, and the threat of long boring hard work (battling bureaucracies) and social ostracization. Moreover, there is the feeling of wrongness in not having acted sooner on a subject that it of primary moral importance.

(Who feels good about tolerating slavery? When literal chattel slavery existed in the South, mainstream Northerners and Southerners found bogus arguments to defend it. Of course, none of those arguments held up when honestly examined, but survived because there was no benefit seen to “fighting city hall”.)

Also: although Eric makes his living thinking about things vastly more complex than I am even capable of thinking about, it’s possible he simply has less experience with tyranny than I have. Moreover, it’s also possible to hold a concept in one’s mind abstractly, but not value its widespread implications, or to imagine that the negative implications are not as far-reaching as the abstract thought indicates they COULD be. Such as: Although the rule of law has broken down, and there are no more proper jury trials, per se, most people probably get a fair jury trial, anyway, because most people are decent people.

(Without examining the court system and prison system, and economic stagnation, –in relation to what it COULD be– that might appear to be true.)

Gallons per mile. Interesting. I’ll have to think about that.

BTW: Is Bjorn Llomborg (“The Skeptical Environmentalist”) wrong? Has Eric addressed Llomborg’s ideas elsewhere? I was under the impression Llomborg had some good arguments. But I admittedly have a defective wetware “math brain”.

{ 1 trackback }

Leave a Comment

Previous post:

Next post: