By on September 10, 2009

Yes, well, Car And Driver‘s Dave Vanderwerp kind of stole our headline for his takedown of EPA testing practices. Still, he earns it by stripping away any illusions you may have harbored about the rigor of the EPA testing process. Did you know that the EPA only tests 15 percent of all new vehicle models, taking automakers at their word for the other 85 percent? Surely you were aware that the EPA once had to convert a Bugatti Veyron to two-wheel drive for testing because it didn’t have four-wheel dynos at the time? No? Hell, the EPA test can’t even tell whether the BMW M5’s +100hp “M” button is on or off. No wonder the Volt is going to get 2,347 mpg! Read the whole thing over at MSN Autos.

Get the latest TTAC e-Newsletter!

26 Comments on “The Reality Behind EPA Testing...”


  • avatar
    GS650G

    Your tax dollars hard at work. And we are about to give this agency even more power and authority to regulate even more of our lives. We should be able to vote for the leadership of these agencies since they have direct influence over our lives and frequently the leadership survives past elections.

  • avatar
    Paul Niedermeyer

    I’m becoming increasingly suspicious that some (GM and Ford) are aggressively hyping the the EPA test for higher numbers. The Fusion hybrid never seems to deliver better economy than the Camry hybrid in comparison tests, yet is rated substantially higher. And GM’s optimistic highway ratings on a lot of their recent offerings ssems suspect.

    So if GM were to get caught cheating, we, the taxpayers would be paying the fine??

  • avatar
    psarhjinian

    Did you know that the EPA only tests 15 percent of all new vehicle models, taking automakers at their word for the other 85 percent?

    I believe the European testing cycle ups the figure to 100%.

  • avatar
    psarhjinian

    And GM’s optimistic highway ratings on a lot of their recent offerings ssems suspect.

    This is a long and storied game with General Motors. They’ve always been very, very good at gaming the EPA highway number on cars that would otherwise be gas guzzlers. It’s less of an issue than it used to be, if only because everyone else does it, too.

    The Corvette is a classic example: people quote the EPA number like it’s gospel, and Corvette fans are quick to point to it, ignoring the fact that no one drives their Corvette in such a manner and b) that figure is actually not all that easy to acheive.

    Even worse are people who quote the European cycle mileage figures for diesels, but that’s another rant entirely.

  • avatar
    johnthacker

    Did you know that the EPA only tests 15 percent of all new vehicle models, taking automakers at their word for the other 85 percent?

    Yes, and that part at least doesn’t bother me. They select most of the 15% they do test randomly, and the others for specific reasons. If they do that, can automakers really afford to lie on their models, not knowing which ones will be tested?

    You might as well complain that the IRS doesn’t audit every individual taxpayer, instead trusting most returns but auditing some randomly and some for various reasons of looking suspicious. The EPA’s auditing procedure is much like the IRS’s.

  • avatar
    johnthacker

    Even worse are people who quote the European cycle mileage figures for diesels, but that’s another rant entirely.

    As far as diesels go, there’s also the problem of the mileage figures making sense if you’re talking about miles, but being a poor measure if you’re talking about energy/CO_2 use, since a gallon of diesel has more energy and thus releases more CO_2.

  • avatar
    johnthacker

    This from the article, though, was pretty amusing:
    The EPA’s Wehrly describes a hypothetical, but typical, discussion where an automaker wants to get around activating a particular feature that degrades fuel economy during the testing. Luxury automaker: “We have this feature, but no one uses it.” EPA: “Then why do you have it?” Automaker: “Because Mercedes has it. If we don’t have it, it looks bad.”

  • avatar
    wsn

    The ultimate “EPA” happens at the pump, just tax gasoline at $5 per gallon and we will quickly know which vehicles are gas guzzlers by looking at the sales chart.

    And BTW, that will allow us to pay off the national debt quickly. (Yes, we do have a national debt in Canada, just like the US.)

  • avatar
    redrum

    …he earns it by stripping away any illusions you may have harbored about the rigor of the EPA testing process.

    Did we read the same article? I think this quote from the article pretty much sums it up: Measuring fuel economy during the tests is likewise hugely complex, which is why the automakers and the EPA both follow precisely the same protocol.

    I trust their testing methodology over the anecdotal rants that have been published on this site with regards to fuel economy.

    The only “illusion” being stripped is that there is one definitive way of measuring fuel economy. In reality it is a complicated (and highly politically charged) process that relies on many variables that cannot be perfectly accounted for.

  • avatar
    thetopdog

    psarhjinian :

    Who cares if nobody drives a Corvette in that manner? All cars should be tested on equal footing. Plus, most Corvettes are driven by old guys that actually DO drive that way

    And it is exceedingly easy to get 26+ highway mpg in the Corvette. You basically just put it in 6th and cruise. 26mpg is easy enough to attain at any speed up to about 85mph, which is well above the speed limit almost anywhere in the country. There is no special trick to it

    The city mileage might be a different story. I have no idea how well a Vette fares in a normal city, but in Boston I struggle to get about 12mpg.

  • avatar
    PeteMoran

    Even worse are people who quote the European cycle mileage figures for diesels, but that’s another rant entirely.

    Hello!? Metric system anybody???

  • avatar
    AG

    Oh snap, I used to drive by that sign every day.

  • avatar
    Autosavant

    “ignoring the fact that no one drives their Corvette in such a manner ”

    Wrong. see plenty of drivers that take their corvettes on long trips, when I take such trips myself, and I notice they do not drive like maniacs, none of them I saw ever did, they are not too “stealth” (esp. if the corvette is fire-engine red), so they tend to drive with the cruise on 65-75 and should get fantastic MPG (28, even 30) esp. in 6th gear.

  • avatar
    CarShark

    “We have a guy that’s literally made a career out of specializing in rounding,”

    That line really stuck out to me.

  • avatar
    Eric Bryant

    Face it – unless your driving habits precisely match the EPA’s test cycles, the numbers on the window sticker are largely irrelevant. Sure, maybe they provide some approximate and relative indication of fuel economy, but not nearly so precise as to merit the sort of hand-wringing over single-digit differences of the sort in which we typically engage.

    Since I spend a lot of time on the highway traveling for work, the real-world fuel economy achieved at an average pace of ~75 MPH (with occassional deviation from that due to traffic and what-not) is most important to me. The window sticker fuel economy of my old Mercury Grand Marquis is rated at 23 MPG highway (“revised” EPA numbers). My Chrysler 300C SRT8 gets a rating of 18 MPG. Both pull down about 20-21 MPG in my real-world highway driving, despite the Merc’s supposed 27% advantage (also note that one handily beats the EPA’s rating, while the other falls a bit short). That, of course, doesn’t take into account each vehicle’s senstivity to driving habits – the 170 HP Grand Marquis will hardly deviate from my observed number with any possible variation in driving style, while the SRT8’s observed economy greatly depends on the driver’s level of socialpathy. That doesn’t show up in the rigid EPA tests, though.

    As I said – the window numbers are meaningless for most of us, and unfortunately, the EPA tests and not our real-world driving behavior is being used by the government and OEMs to make product decisions for us.

  • avatar
    dean

    Eric: your last statement is exactly why CAFE should be scrapped and fuel taxes implemented that attempt to add the cost of externalities into the price of fuel. And it is exactly why every “freedom-loving” American should support it. You would be free to buy and drive any damn vehicle you wanted provided you were prepared to pay the price. Under CAFE, however, your choice is restricted by an arbitrary and ridiculously complicated/expensive regulatory regime. (And it still won’t have the same effect on national gas consumption that a tax would.)

    But of course, “freedom-loving” Americans want the best of all worlds: buy whatever they want, and have everybody else contribute by subsidizing the external costs and keeping fuel prices low.

  • avatar
    PeteMoran

    @ Eric Bryant

    If you drive a Mercury Grand Marquis and then back-to-back a 300C SRT8 in the same manner, over the same course you will see the approximate relative difference the EPA suggests.

    EPA tests are all relative, it’s a GUIDE, that’s it, that’s all.

    EPA tests and not our real-world driving behavior is being used by the government and OEMs to make product decisions for us.

    Eh? As if that was practical. Decisions need to be made Choice A relative to Choice B. EPA tests, being relative, are a perfectly satisfactory way to do it, with similar systems used around the world.

    @ dean

    What you’re effectively saying is that each consumer has acquired ALL the information to make the correct decision. That’s never true.

    “I’m free” and “I want it” so “it’s my RIGHT to have it” are paths to a worse civilisation.

  • avatar

    @PeteM-

    How bout specifications in g/mile…GRAMS per MILE!!! Can ya sit on the fence with your thumb up your butt any better? Yeah we need to teach our future scientists and engineers hell everybody(that means you teachers with TWO Masters degrees) the METRIC system. Is there a bigger waste of time than multiplying fractions? Can ya even buy an Fing metric/inch tape measure at Home Depot in this dumass country? If GE makes a godam thing that requires an 11/16 wrench, WE are done.

  • avatar
    PeteMoran

    Wow, grams per mile. That’s truly f’ked up. I don’t think I’ve seen that yet (luckily). Perhaps they use that at NASA.

  • avatar
    brandloyalty

    I’m under the impression EPA recently changed their tests to more accurately refect real world mileage. This had the effect of generally lowering the numbers. Since Eric Bryant’s old Mercury’s number probably is from the old system, and the Chrysler’s number is from the new system, a direct comparison of their EPA numbers would not be useful.

    Another thing to watch out for is the difference between 4-cyl and 6-cyl versions. People commonly quoted the EPA numbers for the 4-cyl current generation Rav4 when they were otherwise either talking about the 6-cyl version, or comparing the Rav4 to 6-cyl cuv’s.

  • avatar
    redrum

    As I said – the window numbers are meaningless for most of us, and unfortunately, the EPA tests and not our real-world driving behavior is being used by the government and OEMs to make product decisions for us.

    You mean the EPA test is meaningless for YOU and does not reflect YOUR real world driving. Why is it so many people assume their style of driving is the “real world”? No two drivers are exactly alike. The article pretty clearly explains the challenge of coming up with a standardized method of measuring fuel economy when so much is dependent on how the car is driven.

    I’ve personally found the new EPA estimates to be pretty accurate with regards to my own personal driving style, but I realize it is only an estimate. But I recognize that their methodology is scientific and repeatable. “I get xx MPG in this car and xx MPG in that car” is not.

  • avatar
    niky

    Well… when the EPA changes its testing paradigm to “better reflect real-world driving”, it gives consumers the idea that it actually should. Which is wrong.

    I’m of the mind that economy testing should not be “highway/urban”, but a matrix, showing “aggressive highway, normal highway, eco-highway” and one “urban”.

    This eliminates the artificial advantage automatics have at the EPA, where they’re allowed to self-shift.

  • avatar
    Robstar

    With the newer revised ratings, they pretty much match what I get in my STi (16/22) unless I drive really slow or take really short trips, or drive aggresively (not common).

    The revised numbers for the Neon are nowhere close. I barely get the city highway mileage (22) and almost always get 35+ on the highway (rated 28).

    Then again, skewing the numbers on the highway is pretty easy unless you are stuck in traffic.

    STi in the right lane at 52-53mph (6th gear)= 32mpg
    STi in the left lane with the turbo spooled (3rd-4th gear)up closing in on 80mph = 8mpg.

    I’ve also averaged a tankful on the neon (all highway) at 40mpg.

    I have done a full 35 (29 hwy, 6-7 city) mile commute with an average over 29mpg in the STi (rated 22 in the new numbers). I have also done 19’ish on the same commute coming home in heavy traffic.

    I use a real-time obd2 meter (scangauge) to get mileage…it’s a couple years old and works well.

  • avatar
    Autosavant

    “Author: dean
    Comment:
    Eric: your last statement is exactly why CAFE should be scrapped and fuel taxes implemented that attempt to add the cost of externalities into the price of fuel. And it is exactly why every American should support it. You would be free to buy and drive any damn vehicle you wanted provided you were prepared to pay the price. Under CAFE, however, your choice is restricted by an arbitrary and ridiculously complicated/expensive regulatory regime. (And it still won’t have the same effect on national gas consumption that a tax would.)

    But of course, Americans want the best of all worlds: buy whatever they want, and have everybody else contribute by subsidizing the external costs and keeping fuel prices low.”

    Absolutely true. Dean sure is Econ Literate. And so important, I wanted to quote the whole thing. However, the cowards in Wash DC (COngress especially) will NEVER accept the TRUTH and impose a huge tax on gas (even if it is 100% offset, as I would prefer, with an income or other tax cut of equal magnitude)

    And this is true whether you are “freedom loving” , “Oprah loving”, or whatever, the laws of Econs do not change.

    Now as a consolation, one positive reault I expect from the new CAFE rules, is that finally these clowns in Germany and Japan that make luxury cars will offer them in the USA AS WELL as overseas with the more efficient diesels, with smaller gas engines, etc. There is no need for a 4.3 lt v8 on the S class. “4 out of 5 dentists” that are dumb enough to buy them new, would not notice it if it had the 2.8 6 instead, or even this new 4 supercharged!

  • avatar
    JMII

    This is why YMMV so much. Consumer Reports actually DRIVES the cars and reports back the results, this gives much better results then any EPA test done on a dyno, in a lab, while looking at a monitor.

    The whole system is flawed, everyone knows it, this is why model specific fan forums always have a sticky thread at the top called “Whats your MPG?”. It would be really interesting if all cars downloaded their fuel use/miles driven yearly into some official database that is published for all to see.

    Prime examples of YMMV: my Dodge Dakota has always come 2-3 mpg short of its EPA city/highway #s while driving on the ruler flat roads of Florida, yet my VW Passat gets nearly 3 mpg BETTER then its ratings under the same conditions (and same driver).

  • avatar
    DweezilSFV

    The EPA #s have rarely been reflective of what actually is achievable or what drivers actually get.Those are a broad estimate from a lab test. They were always an off shoot of emissions testinin the first place. I guess if I drive in a lab I will achieve those exact #s.

    And the new stats for the earlier cars are simply derived by taking the old #s and subtracting 10%.

    My ION is rated at 32 HWY under the old testing [2.2, auto]. Now it’s rated at 29. In all of my freeway driving I have never seen 29.It’s generally between 34 and 37. And I have recorded every drop of gas since the first tank.

    In the city it’s always gotten better than the EPA #s old and new. Los Angeles traffic and commuting, sea level, hot, cold, mountain driving it’s given better mileage.

    But in the uproar everyone forgets the disclaimer: “your mileage may vary” and it does.

    It’s never been a secret that the EPA #s were essentially irrelevant. That’s why there’s such a wide range in the classification on the window sticker. Which no one reads. And when the car doesn’t get the “claimed” range, there’s something “wrong’ with their car. This has been going on since the 70s. Nothing new here.

    @ brandloyalty: Eric Bryant clearly stated he was using the new recalculated EPA #s of the Mercury Marquis when he was comparing the 300’s mileage.


Back to TopLeave a Reply

You must be logged in to post a comment.

Subscribe without commenting

Recent Comments

New Car Research

Get a Free Dealer Quote

Staff

  • Contributing Writers

  • Jack Baruth, United States
  • Brendan McAleer, Canada
  • Marcelo De Vasconcellos, Brazil
  • Vojta Dobes, Czech Republic
  • Matthias Gasnier, Australia
  • W. Christian 'Mental' Ward, Abu Dhabi
  • Mark Stevenson, Canada
  • Cameron Aubernon, United States
  • J Emerson, United States