MPG Is Stupid

A Closer Look at an Ill-Conceived Fuel Economy Standard

  • 2012 Toyota Prius C

    2012 Toyota Prius C

    The Toyota Prius C and standard Prius are famously rated at 50 mpg by the EPA. | July 31, 2013

18 Photos

When we talk about fuel economy in these parts we talk in miles per gallon, known to everyone as mpg. We may think we have it all figured out, but mpg is a stupid unit. It actually distorts the issue of fuel efficiency and leads to many misconceptions in the minds of consumers. We should all do ourselves a favor and stop using it.

Engineers know this. When they measure fuel consumption in the lab they talk in terms of gallons per mile. The CAFE standards that govern fuel economy targets and the EPA procedures that lead to window sticker numbers all operate behind the scenes in gallons per mile, too. The final results are only converted into mpg at the last minute for publication because consumers expect it.

What follows is why the engineers and the government rule makers actually have it figured out for a change. We as consumers should insist on the same sort of clarity and simplicity.

Efficient Cars Burn Less Gasoline
Ultimately, what most consumers really want to know is this: "How much money will I spend on gas?" For a given price of gasoline, that's the same as asking, "How many gallons will this car burn?" That's fuel consumption, and smaller numbers are better.

For some this question goes beyond spending less money and reducing dependence on imported oil because the number of gallons consumed also translates directly into CO2 emissions.

Gallons burned, therefore, is the quantity we want to measure, what your math teacher would call the "dependent variable." Said math whiz would go on to say that a dependent variable belongs on the top of the fraction. This means we should measure gallons per mile or gallons per hundred miles, not mpg.

Five Equals Two
Why is this important? Having gallons at the bottom of the fraction introduces a huge non-linearity. A single mpg has no fixed value; it isn't a tangible thing.

Take a car that is rated at 20 mpg combined: a 2014 Ford Mustang 5.0-liter V8 or a 2013 Hyundai Santa Fe, perhaps. After a week of less-than-patient driving, we fill it up to find we've gotten 18 mpg. We missed the rating by 2 mpg, which works out to 10 percent. Oh, well.

MPG Is Stupid

The next day we drive a 2013 Toyota Prius, which is rated at 50 mpg. And let's say we also underperform by 10 percent. Thing is, here that 10 percent works out to 5 mpg. But it's a hybrid. And it's off the mark by 5 mpg. That's outrageous!

Despite the difference in gut reaction, the 5 mpg miss in our Prius is no more significant than the 2 mpg miss in the Mustang. Five equals two. But people tend to think that the value of 1 mpg is fixed, that it has inherent meaning. Nope.

Actually, Ten Equals Two
Let's change both ratings into the gpm format, specifically gallons per 100 miles so the resulting numbers are greater than one.

Our 20-mpg Mustang consumes 5.0 gallons per 100 miles (gp100) and the 50-mpg Prius burns 2.0 gp100. Right away we can easily see how much gasoline each one will use over a given distance. And lower is better, which is how you want to think if your aim is to save money, use less fuel, emit less CO2 or all three.

Now let's see what happens when each underperforms by 10 percent as before. Starting at 5.0 gp100, the Mustang's 10 percent miss equates to an extra 0.5 gallon, bringing actual consumption up to 5.5 gp100. The same 10 percent offset brings the Prius' consumption up by 0.2 gallon to 2.2 gp100.

MPG Is Stupid

What this means is the Mustang's 2-mpg miss is actually more than twice as significant (0.5 gallon vs. 0.2 gallon) in pure gallon terms as the 5-mpg miss in the Prius over the same distance.

At a fuel cost of $4 per gallon, the Mustang's 2-mpg shortfall results in an extra 0.5 gallon and $2, but the Prius' 5-mpg miss represents just 0.2 gallon and 80 cents over the same 100-mile distance. Two is greater than five.

In fact, the 50-mpg Prius would have to underperform to the tune of 40 mpg, 10 mpg below expectations, to get to 2.5 gp100 where it'd burn the same extra half-gallon and have the same out-of-pocket impact as the 20-mpg Mustang does when it achieves 18 mpg. Ten equals two.

Hybrids Take It in the Shorts
This misdirection has everything to do with the bad-math nature of the upside-down mpg unit. And it gets worse the higher the mpg number gets.

And so hybrids (and some diesels) tend to get criticized for underperforming far more than conventional cars. Hybrid offerings from Honda, Toyota and, most recently, Ford have taken heat in the form of lawsuits and local news exposés for gaming the system in the test but falling short in the real world.

But a good deal of the grief comes from a lack of understanding about how mpg works (or, more to the point, doesn't work) in the high-mpg rarefied air where hybrids and diesels live. Few people get too upset when their Mustang gets 18 mpg instead of 20 mpg; it's only 2 mpg. But take away 5 or 10 mpg from a 50-mpg hybrid and the world spins off its axis.

If we look at both in gpm terms, the real story emerges and the "problem" disappears.

It's All in Your Head
Finally, there's a practical reason to like the gpm format.

Say we're planning a trip. Our 2014 Chevrolet Silverado 1500 pickup is good for 18 mpg and we're going to drive about 200 miles. How many gallons? How much will it cost? Long division is involved. We'll need a calculator.

If our truck was labeled in the gpm format, its consumption rating would be 5.6 gp100. Multiply by 2 to get 200 miles' worth of fuel: 11.2 gallons. Multiply that by the price of fuel to get dollars. Sure, a calculator still helps, but the approach is more straightforward. Multiply, then multiply. And you're already starting with gallons, the thing you're buying.

MPG Is Stupid

And it makes budgeting and car shopping simpler, too. Many of us drive about 1,000 miles each month. It's ridiculously easy to take a gp100 miles rating on a window sticker and multiply it by 10 to get the number of gallons used during that month. Our 5.6 gp100 Chevy pickup would go through 56 gallons each credit card cycle.

You say you drive 1,500 miles a month? OK, tack on another 28 gallons. You don't have to be Rain Man.

The gpm format removes all the confusion and lays bare the information we really wanted to know in the first place: How much fuel does this thing use? We really ought to be using a gpm-based unit in all discussions of fuel consumption. It should be on the window sticker.

Oh wait, it is. The EPA snuck in the gpm format when a new window sticker design was rolled out for the 2013 model year. Today a vehicle's combined gp100 rating is right there, just below the prominent combined mpg number. It's also listed on the Web site.

They don't bother with city and highway versions of gp100, but that's fine with us because real-world driving is a combination of the two, anyway. The only thing more bogus than mpg itself is advertising that focuses strictly on highway mpg.

The gp100 rating on the new window stickers appears in small print, but the usefulness of the information it provides looms large. In actuality it's all you ought to be looking at.

For once, the fine print has it right. Find it, read it, use it, think it — and ignore MPG. It needs to disappear.


  • skw0123 skw0123 Posts:

    Sounds good. So, lead the way...put gp100 as the official measure in every review and put MPG in parentheses.

  • empowah empowah Posts:

    Excellent writeup. And I agree, Edmunds should lead the way.

  • Excellent explanation of why we should change to gpm. Lead the charge! If you go to Europe (or almost anywhere else in the world) they express fuel economy in terms of liters-per-100-kilometers. Same concept. (While you're at it, get the US to change to the metric system. This would REALLY simplify calculations.)

  • agentorange agentorange Posts:

    Dan, while what you say is true, you are fighting the long standing perception in the mind of public that bigger numbers are better, i.e., 50mpg is better than 20 mpg. Now you want to shove how much gas is being used in the faces of the mathematically challenged who have been in blissful ignorance all these years. Some of them are going to resent that. Then there will be car execs sending black helicopters after you because the LAST thing they want is Joe Public easily seeing how much money he is blowing on gas with most cars in the US market. No sir, we don't want those kinds of facts loose amongst the general public. Good luck with your campaign.

  • Spock would love this; it seems Logical! - The site is great. Really handy is the Compare Side-by-Side feature. You can even personalize it to your own personal mix of driving (highway vs city), and how many miles a year you drive...

  • chrgman chrgman Posts:

    Gee-It only took 28 paragraphs and a few sentences to explain a "simple" procedure.... Going to Las Vegas yesterday, a friends new 528 display read out the mpg as we drove(36.7), not sure I'd like to see .0272479 gallons per mile, if you know what I mean.

  • g3ngo g3ngo Posts:

    I don't know a single person that struggles to understand the concept or math behind MPG. Even if they did, most understand that more MPG means more efficiency and less money spent on fuel. I don't think this argument is sufficient to warrant a change, nor does it protect consumers from companies inflating their fuel efficiency ratings. As another poster mentioned, the best thing we could do is make the EPA tests better reflect real-world driving conditions and consumption.

  • ampim ampim Posts:

    Preaching to the choir. An example I've been using that blows people's minds is that regardless of the price of gas, and how many miles you drive, you'd save the same amount of money going from a 10 mpg truck to a 16 mpg SUV as you would going from that 16 mpg SUV to a 40 mpg hybrid. Or in the 'math' in this case would be a 60% improvement equals 150% improvement. Or 6 mpg equals 24 mpg.

  • magius39 magius39 Posts:

    I just don't understand why it took a 20+ paragraph article to make the point that numbers are different if you divide them differently. The most appropriate unit is the one that answers the question you are asking. "How many gallons will it take to drive 100 miles?" And "How many miles can I drive on one gallon of gas?" Are different questions, needing different units. I just think its a stupid argument/waste of an article to say, "no, the inverse of this number is clearly the better number." When it really depends on the questions being asked. Not to mention the fact that these numbers are derived from archaic/gamed systems. The fact that people are "up in arms" over hybrids/turbo cars not hitting their sticker values for fuel usage is based on the fact the test is wrong, and not representative. Changing the number derived from that test (be it mpg or gpm) won't make it any more accurate, in the same way that changing fonts won't change what is actually written.

  • bc1960 bc1960 Posts:

    Because as the article alludes, when you divide large numbers by small numbers the result exaggerates the significance of a small delta in the denominator, especially when you crudely round to the nearest whole number as the EPA does. 300 miles on 10 gal is 30mpg. 300 miles on 9.5 gal is 32mpg. People tend to assume that the mpg difference is important when it's arguably not--if you rounded to the same significance they're both 0.03 gal/mi; even if you keep the same significance as currency it's 0.0333 vs. 0.0317. Gal/mi also directly translates into $/mi by multiplying by the pump price; you are correct that division and multiplication are equivalent operators but humans psychologically find addition and multiplication easier to do mentally than subtraction and division, by a large margin. I disagree with emphasizing the combined number on the current sticker; it assumes a specific percentage city/hwy split that may or may not be relevant--you could be hitting the city and hwy figures exactly and miss the combined if your split is different. At least the sticker continues to publish both numbers so you can figure your own approximate combined number. Edmunds bloggers in the long-term section make a big deal about getting 23.5mpg (which should be rounded to 24) combined instead of 26mpg combined without noting what the driving circumstances were. The sticker also eliminated any mention of the expected variation in the averages (which was previously cited as plus or minus 6 mpg for 30mpg, 8 mpg for 40mpg, e. g.), which only reinforces the public tendency to expect that the sticker numbers are a promise and that 2mpg differences are always a big deal.

  • duck87 duck87 Posts:

    The only thing stupid about the system is that it's in imperial units. Come over to the dark side... start using metric units like liters and kilometers... I promise it won't hurt and will make calculations easier.

  • magius39 magius39 Posts:

    This is a stupid article making a foolish point poorly. The crux of the article focuses on the idea that - miles / gallons is somehow less accurate and leads to more misconceptions than gallons / mile. Its the same number. Just divided differently. Particularly annoying is the section "It's All in Your Head." - You're attempting to make the point that it is difficult to calculate the amount of gallons used for a 200 mile trip because it involves doing division. Whereas, multiplication is so much easier, if only! Its the same basic mathematical function. It really is. Gallons used = 200/18 (if expressed in miles per gallon) or Gallons used = 5.6/100 * 200 (if expressed in gallons per mile) I fail to see your point that the latter is somehow "more straightforward" than the former. If anything, its more complex if your trip length is anything other than a multiple of 100. The key point is that the "best" unit is useless discussion unless you know the primary question. What the EPA is saying by keeping MPG on window stickets is that the most common question on car buyers minds is, "How far can I go on 1 gallon of gas with this particular vehicle?" Isn't it easier to just answer that question in 1 glance? If you want to make a point about window stickers being inaccurate/uninformative, write an article describing the city/highway testing cycles and how they are/are not similar to modern day car usage and/or the leeway car companies have in the production of those numbers.

  • jimveta jimveta Posts:

    Let me air my pet peeve about people conflating efficiency with economy. With all due respect this article also does the same and misses the point about real efficiency. GPM is still a measure of economy, or frugality, and not efficiency. Efficiency is measured as a ratio between input and output. For cars it's BSFC (brake specific fuel consumption) or Torque per unit-fuel @ rpm used. That it measure the efficiency of converting between input energy and output energy. Look up any automotive engineering book and this--BSFC--is the definition of efficiency you'll see, which is not a single figure but varies over rpm, not MPG or GPM. You don't measure an AC/DC power supply's efficiency just by how much electricity it consumes. That's frugality or economy. You measure by how much it consumes (which can be over time or instantaneous) vs how much DC power produces. You don't measure a refrigerator's efficiency just by how much electricity it consumes. That's frugality or economy. You measure by how much it consumes (which can be over time or instantaneous) vs how much cooling power it produces.

  • klinkerkc klinkerkc Posts:

    Magius39, I think you're missing the point of this article. You are correct in that there's nothing technical wrong with the current miles/gallon value advertised in the US. Whether you use mpg or gal/mile, you are correct that it's all the same data, just presented in a different format. The point of this article is that mpg is MISLEADING to consumers, completely unintentionally but misleading nonetheless. By using mpg, you might think that 14 mpg vs. 15 mpg isn't that big of an economy gain, but it's actually a helluva big gain compared to 49 vs. 50 mpg. I just happened to run this analysis a couple years ago when I was buying a new vehicle and comparing vehicle cost vs. total monthly cost inclusive of fuel economy. : Assumptions: 12000 mi/yr mpg $3/gal 10 $300 15 $200 20 $150 25 $120 30 $100 35 $86 40 $75 45 $67 50 $60 You'd save $100/month in gas cost if you ditched that 10mpg bubba truck for a 15mpg SUV. Huge change there. However, if you ditch that 45mpg Hyundai hybrid for a 50mpg Prius, you'd only save $7/month and hence might be considered an idiot if you did so purely to save on gas costs. Same incremental mpg gain between the two vehcile scenarios but radically different fuel cost changes. The net change in gas cost for a certain mpg change completely depends on the reference mpg. It's a crappy indicator because the amount of gas used does not response linearly to mpg change. On the other hand, gallons/mile allows for a linear response between gas used and gpm. If your gpm decreases by 10%, you're going to spend 10% less on gas, and visa-versa. It's simple. That's the point of this article.

  • cowichan cowichan Posts:

    This is really just the world standard, liters/100 kilometers, Americanized for local tastes.

  • compressor compressor Posts:

    GPM is fine for a comparison of fuel efficiency given that most people in this country apparently don't know what a reciprocal is (don't get me started as to why we should always appease the lowest common denominator - as your math "whiz" would say) That said, your example of planning a road trip is why, given peoples poor math skills, mpg is better. Very few plan on how much gas it takes to take a road trip. For those that do, the math is easy (not to mention very few trips are rounded to the 100 miles) What people do need to figure out while driving is how much farther I can go with my remaining gas. Say I have a 16 gallon tank in my 18 mpg car. I am at the 1/4 mark. Sans reserve I know I have about 4 gal of fuel. At 18mpg, I can go approximately 72 miles. Now, tell the average person how much farther they can go with a 1/4 tank of fuel (4 gals) in a car that gets 5.5 gp100 and I think you will have some very confused people - remember, these people think, as you have displayed, that gpm is different than mpg, and not just a simple reciprocal.

  • compressor compressor Posts:

    magius39 - thank you!

  • jl335i jl335i Posts:

    Agreed, mpg does not convey a direct correlation. However, it still is a measure of fuel efficiency. It measures the following: "how many miles one can travel per gallon of fuel used." A more fuel efficient vehicle will travel farther, or more miles per gallon. It's not as direct as gpm, but it provides a metric for efficiency just the same. gpm hasn't caught on because, for the standard consumer, "who cares"? No one is wasting time measuring if their vehicle is under performing by 10% and the "true" meaning of that number.

  • vroom10 vroom10 Posts:

    magius has said it best. He spoke exactly what I was thinking. The first paragraph into the article I was rolling my eyes. The author thinks he has it all figured out by playing a semantics game with fuel consumption. Math is math people! You can approach an equation several different ways, but if calculated correctly, you will arrive at the same conclusion. Needless to say, I didn't bother finishing the article. No wonder i stopped coming to this site for my car news.

  • bankerdanny bankerdanny Posts:

    mspeedmusicman said: "While you're at it, get the US to change to the metric system. This would REALLY simplify calculations." How so? 5 litres/100km is no more simple to work with than 5 gallons/100 miles.

  • all of this is somewhat pointless for the people like the guy at the gas station back when gas prices were closing in on $5 a gallon. He asked how much it just cost me to fill up my little Tacoma and I said $50. He responded with "that's what we need, my wife's car costs $70" I tried to point out that I still only got around 20 mpg and that it just held less gas; he waved his hand in the air and said "doesn't matter, just needs to be cheaper to fill up" and walked away.

  • mboily mboily Posts:

    The metric systems already uses this calculation, rated liters per 100 kilometers. It is more logical.

  • mboily mboily Posts:

    @ chrgman It's not per miles, it is per 100 miles; So it would read 2.7ghm

  • compressor compressor Posts:

    mboily - it is only more logical if you don't understand math. That said, it is a clearer comparative indication of fuel usage in some ways. But a less clear indicator of fuel consumption in other ways. Both methods have pros and cons for those that do not like to do math As for metric - all our maps, road signs, speed, etc are in miles. All our pumps are in gallons. The cost to change to liters and km would be huge for absolutely no gain whatsoever.

  • greenpony greenpony Posts:

    Most people are lousy at math anyway; any change to how fuel efficiency is expressed will likely just lead to confusion. Still, I think you're skirting about the REAL problem - that's the EPA tests themselves. The tests are antiquated and based on driving conditions that were slower, with less traffic, and with less-capable vehicles, and rely too much on calculations and fudge factors to arrive at today's window sticker estimates. Diesels are underrated, hybrids and turbo DI motors are overrated. Fix the rating system, then we can talk about which reciprocal is better.

  • compressor compressor Posts:

    bc1960 - but if you want to figure out how far you can go on a tank of gas (or any quantity of fuel) you have to do division with gpm. People that find math hard, with find it hard no matter what. Giving them a reciprocal wont help. My apologies, but they are just stupid and seem to be operating at the math level of a 2nd grader, proven by your point that people find addition and multiplication easier - since subtraction and division are technically addition and multiplication, respectively.

  • grover432 grover432 Posts:

    Welcome to the way Canada measures fuel economy: liters/100 km or how many liters of gas it takes to drive 62 miles. The only problem is Transport Canada hasn't kept up with realistic mileage testing numbers like the US has so no Canadian car can hit it's sticker mileage.

  • serezhkin serezhkin Posts:

    Great article, US is one of the few countries still stuck with MPG, and the thing just doesn't make sense. Especially when the laws use it to set economy standards: they say "raise efficiency by 5mpg" without realizing that 5mpg raise not only gets the more expensive the more efficient a car is, but it also has way less impact than making the trucks and other gas-suzzlers be only 2mpg more efficient.

  • jvoelcker jvoelcker Posts:

    What a great headline! No, seriously, it's awesome. It just sounds sorta familiar .... best, jv.

  • bobert2013 bobert2013 Posts:

    To make another point MPG can't be calculated on a vehicle that's coasting while in gear, ie not consuming any fuel while moving. Infinite miles per gallon. A car that's coasting while using the GPM figure will get 0 gallons per 100 miles. To the person who wrote "I just think its a stupid argument/waste of an article to say, "no, the inverse of this number is clearly the better number." I point to the above. Why so upset about it? Is it really fair to call something "stupid" when you don't understand it?

Leave a Comment

Featured Video




Get Pre-Approved for a Loan

Credit Problems?
We can help you get Financing!

Have a question? We're here to help!
Chat online with us
Email us at
*Available daily 8AM-5PM Pacific