How NOT to Calculate Fuel Economy
By Michael Karesh
Sometimes it’s hard to believe what counts as news. The Economist had a blinding flash of the obvious recently and discovered that a used Civic is less expensive in terms of total purchasing and operating costs than a new Prius.
Aside from the odd logic of comparing a six-year-old used car with a new one, the author asserts that changing the oil and air filter boosted the used Civic’s highway fuel economy from 34 miles-per-gallon to 40. This set off my BS detector. Changing the oil should have a negligible impact on fuel economy, and changing the air filter should be worth, at most, two to three mpg if the old one was fairly dirty. So while changing out a dirty filter is definitely a good idea, a six mpg gain should not be happening.
When my BS detector goes off, I check the methods that yielded the results. Let’s assume that the 40 miles-per-gallon figure was properly calculated, though no method is stated. Problem is, the 34 mpg was calculated based on the following information, as stated in the article: “The previous owners told me the car could get from Austin to New Orleans on a single tank of gas. That implied 34 miles to the gallon on highways and about 30 in the city.”
I’ve come across some wildly inaccurate ways to calculate fuel economy, but this one beats them all. First off, the previous owners didn’t say the gas tank was bone dry upon reaching the Crescent City. I’ll wager that it had a gallon or two left in it. And where did that “30 in the city” come from?
The Economist is a highly respected magazine. (I know this because they used to send me subscription offers that stressed the status of the magazine and its readership.) So how could they publish an article that makes claims based on flimsy evidence and faulty logic? The answer seems clear: they take cars no more seriously than they would Hollywood gossip. Perhaps less seriously: no one’s likely to sue for libel as a result of this article.
This article demonstrates that not even a first-tier source like The Economist is worthy of your blind trust. It’s necessary to look beyond stated conclusions to the methods that yielded these conclusions, especially when the conclusions are surprising.
Michael Karesh is editor and publisher of TrueDelta, an auto research firm. You may visit TrueDelta at http://www.truedelta.com.