Your Mileage May Vary

"Your mileage may vary" is the government's way of tacitly admitting the fuel economy figures it publishes on new car window stickers may not reflect the actual miles per gallon your vehicle will deliver once it’s in your hands.

The figures are, at best, a rough guide -- not a definitive statement -- based on averages obtained during controlled testing. And as many of us have discovered, the government's estimates are often a tad on the optimistic side.

The reason for the disparity between what the government thinks we might get and what we actually get is pretty straightforward: The way we drive our vehicles is often very different from the way government testers do. We cruise at 70 to 75 mph on the interstate instead of 55 mph -- knocking down our potential highway fuel economy by as much as 5 to 10 percent. Or our typical city driving consists mostly of fairly abrupt (and very inefficient) 0 to 20 mph stop-and-go driving instead of steady-state 30 to 40 mph -- likewise reducing real-world city mileage by a significant margin.

The current testing procedures used to estimate fuel economy date back several decades -- and haven't been updated to reflect changing driving habits and traffic patterns.

For example, highway speed limits have gone up by 10 to 15 mph (or more) in many states, following the 1995 repeal of the 55 mph National Maximum Speed Limit. These higher speeds have a negative effect on real-world highway fuel economy -- but the current tests assume we still live in a "drive 55" world. Similarly, congestion has worsened markedly in and around most major urban/suburban areas -- increasing the time we spend accelerating and decelerating (which wastes energy and thus fuel).

Finally, the tests currently used are skewed because they assume "optimum" conditions -- even temperatures, intermittent use of power-drawing accessories such as air conditioning -- and so on. But cars burn more gas during "cold starts" on harsh winter mornings than they do at room temperature -- and constant cycling of the air conditioner on a 100 degree July afternoon will likewise cut into the "best possible" mileage you might otherwise get.

But new tests slated to go into effect beginning with model year 2008 vehicles should give consumers a substantially more accurate window into how much gas their new car or truck will use in real-world driving -- because the revised numbers will be based on more realistic driving patterns, including higher speed/higher load driving and more frequent use of power-using accessories such as air conditioning.

The expected result of the changes in testing methods will be an "on-paper" drop in fuel economy of as much as 10 to 15 percent relative to currently published figures -- but this won’t reflect any actual drop in efficiency (all else being equal).

There'll just be more truth in advertising.

For example, under current tests, an '05 Chevy Silverado is rated at 17.8 mpg in city driving. Under the proposed revised testing system, the same vehicle's published city mileage would fall to 13.9 mpg -- which is probably a lot more in line with what the typical owner is getting. The Toyota Camry's projected city mileage would likewise drop from the current 28.5 mpg to a more realistic 24.2 mpg.

One class of vehicles expected to take a major hit is hybrids -- whose "best case" fuel efficiency often founders on the shoals of real-world driving. The city rating of the Honda Civic hybrid, for instance, would drop some 12 mpg -- from the current 48.8 mpg to a much-less-spectacular 36 mpg. This will be a revelation (and reality check) for many people -- who drive their hybrids outside the envelope of the hybrid's optimal operating range.

This isn't to say it's impossible to realize the best-case mileage your vehicle is theoretically capable of. If you drive very moderately (no rapid acceleration; no high-speed driving) under ideal conditions (no sub-zero cold starts, etc.) you might equal (or even beat) the government's currently published guidelines.

But since few of us drive that way in the real world, it's probably a good thing the government is changing the way it does its testing. Better to know what we're likely to get rather than what we might get -- right?