Is EPA playing a game to favor local products?
It seems that we are 'selectively' being taken for a ride.
In many cases the consumption obtained by CR is 10% to 20% higher than reported by EPA, specially on Ford and GM products?

OTOH, Honda's and Hyundai's products often did better?

Using CR figures, the average may be closer to 22 mpg than the reported 24.6 mpg?

The Ford CR results consistently fall short.

Hyundai admitted to reporting false MPG last year. They probably have an incentive to report accurate number snow.

I own a VW 1.8T and a VW 2.0T, and love them. But they appear to be very susceptible to individual driver habits. If I drive like an old lady, I get 15% better mileage than if I drive like a red-blooded American leadfoot. I suspect that the big 2.5L fours in Japanese mid size sedans are less variable, and that is what CR is observing.

OTOH, I looooooove spooling up the turbo.... so to hell with Consumer Reports.

Stoichiometry is the big dog, and charging extra money for forced induction is the cat's meow.

Is Ford 'polishing' the numbers? Their hybrids seem to suffer the same gap between real world fuel economy and EPA ratings.

How can we assume that CR drivers would selectively drive a few trade marks differently than others?

Harvey,

Here's a different way of stating the same thing: a car with a V6 or a big 4 will always have poor fuel economy, no matter how you drive it. You can try driving slow, but you'll be using extra fuel because of the inefficiencies inherent in the design. Most non-turbo mass-market cars are really unpleasant to drive fast (noisy, no top-end power, jerky gear changes), so people won't bother. Ever try to drive a Camry in a lively way? There's really no point...

A smaller turbocharged engine can have very good fuel economy, or very poor economy, depending on how you drive it. The key difference is that turbo cars get more entertaining the more you rev the engine and the faster you go. The power just builds-up with the revs.

Most of us know that there's a time and a place for that kind of stuff. Automotive "test pilots" don't. That's why you read about stuff like "understeer at the limit," which is code for "nearly crashed." That's not the kind of stuff that regular drivers want to experience every day in cars that they own, with kids in the back.

@Harvey: Most EPA testing is done by the car companies themselves with results submitted to EPA. The auto companies have a strong incentive to get good results, and are not unbiased testers. I would expect that the companies vary in how much they are willing to fudge the tests. One can imagine analyzing the test and tuning the ECU to the test to some degree. I have no specific knowledge about such things, but having worked in large corporations for many years, I can imagine just the sort of things that might go on, depending on the corporate culture involved.

Having an single, independent entity run the testing for all cars might give a more consistent picture, or help reveal flaws in the test methodology at any rate.

FWIW, I have found CR's mileage ratings to be on target.

Yes NL... this type of self-testing-monitoring is probably as suggestive and biased as the self-testing-monitoring done by Tar Sands Developers and Operators to determine the pollution created by their current and future operations.

Most car specialists-testers will 'selectively' add 20% to EPA's and/or Big-3's claims.

Most Environmental Specialists will 'selectively' add 50+% to claimed-declared pollution from Tar Sands Operations.

Well protected (incorruptible) independent monitoring entities (if they still exist) should be called upon to overview the testing-monitoring done by manufacturers-operators and governmental agencies?

The EPA sticker fuel economy is derived from a weighted average of 5 driving cycles: city, highway, city with cold start, US06 (more aggressive driving), and SC03 (hot ambient with air conditioner set for maximum cooling rate). In the recent past, auto companies had the option of just using the city and highway driving cycles and using a linear correlation to get "real world" driving numbers for the EPA sticker. EPA created the linear correlation to produce reasonably accurate estimates for 2008 vintage vehicles. I'm not sure if auto makers now have to use the 5 tests, or if they can still use the linear correlation with the city and highway tests. Detailed test result data is available at http://www.epa.gov/otaq/testdata.htm and for many of the cars there are only city and highway test results.

Given this background, there are several possible reasons why the EPA sticker numbers might not be accurate for turbos and for the Ford Fusion hybrid. Perhaps the weighted average of the 5 tests does not work very well for these vehicles. Or perhaps the linear correlation for the city and highway tests does not work very well for these vehicles. Of course, if drivers are much more aggressive in terms of acceleration and maximum speed, then their fuel economy will be worse than the EPA sticker would predict. Similarly, the Consumer Reports tests may or may not do a good job of predicting real world driving for these vehicles.

I also find it amusing that CR is ignoring the difference in cost of fuel per mile... some of the turbo-charged examples that get better mileage like VW and Audi require premium fuel to do so.
The cost of fuel per mile may actually be better in the 1.8T Cruze than the 2.0T BMW. CR doesn't point it out so I have to assume they didn't account for it.
I see this same failure when comparing the mileage of gas hybrids to diesels. Different fuels have different costs and energy content per volume so why would you directly compare their MPG ratings? Doing so is comparing apples to oranges.