Dear Car Talk:
My new 2020 Chevy Malibu promised on Chevy's website -- and on its MSRP sticker -- to get 29 miles per gallon in the city. It doesn't. The best gas mileage I have even gotten for a tankful of gas was 24 mpg.
I've returned to the dealer three times. They say everything is running as it should and according to spec. They showed little concern for my problem.
I live in San Diego, have 5,000-plus miles on it and feel cheated! I bought this car over two other cars due to its advertised exceptional gas mileage.
What gives? -- Craig
What gives? The large print gives, Craig. And the small print taketh away. The small print says "your mileage may vary." And if you read Car Talk regularly, you know that it almost always varies downward. Although some estimates are further off than others. The problem with mileage numbers is that they rarely represent real-world driving conditions or drivers.
The Environmental Protection Agency requires that car manufacturers test each model using very specific instructions. The vehicle is put on a dynamometer. What's that? It's the automotive equivalent of an exercise bicycle. The car sits still, and the driven wheels turn a roller, to simulate driving. Then the driver follows a very specific set of instructions to simulate a bunch of "trips." He or she will try to match the car's speed to the exact speeds on a screen, as it leads them through a "course" of stop and go trips. And as long as the test driver stays within 2 mph of the specified speeds, the test is valid, and the automaker has an EPA mileage number.
The city test reportedly covers 11 miles of driving in about 31 minutes and has a maximum acceleration rate of 18 seconds for a 0-60 time. That's the speed at which -- if that person were in front of you -- you'd assume they were driving that slowly because they were heading to a colonoscopy appointment.
Of course, hardly anybody drives exactly like that. Well, my late brother drove like that, but he was always afraid his cigar ash would fall off and land between his legs. More importantly, since the automakers know precisely how they'll be tested, they can design their cars to excel on that test. For instance, if they know the most common speeds on the test are 25-35 miles an hour, they can engineer their transmission shift points and engine parameters to be most miserly at that speed. Even if that's not how many drivers will use their cars.
For all these reasons, we always recommend using the EPA mileage numbers for comparison shopping, rather than as promised, real-world results. And we'd recommend to your dealer that he send a guy over to your house to add half a gallon of gas to your tank once a week. That'll raise your mileage just enough to keep you from coming back and complaining.