Your mileage may vary.
Those four words have been linked to the EPA’s fuel-economy estimates since the 1974 model year, when automakers began disclosing fuel efficiency of their cars and light trucks on window stickers in the wake of the OPEC oil embargo.
When the EPA issues its fuel-economy estimates for 2003 vehicles this month (www.fueleconomy.gov), critics say those words will be more of a promise than a warning because the numbers are based on assumptions that no longer apply to the way most Americans drive.
The EPA has not changed the way it calculates fuel economy since 1985, when it reduced mileage estimates generated in laboratory tests by “correction factors” of 10 percent for city driving and 22 percent for highway driving to make the numbers closer to what motorists achieve in the real world.
In the 17 years since, the number of vehicles on U.S. roads has increased 41 percent, to 221 million, according to the Polk Co. The Federal Highway Administration says the number of miles driven per year has grown 57 percent, to 2.78 trillion.
With so many more vehicles, logging more miles, traffic has slowed to a crawl on congested Chicago streets.
The EPA, however, still bases its city estimates on a simulated driving cycle developed nearly 30 years ago that averages 20 miles per hour, a pace many Chicagoans only fantasize about.
When the EPA created its highway mileage test, which averages 48 m.p.h., the speed limit on rural interstate highways was 55 m.p.h. Now, it is 65 in Illinois and several states, 70 in others and 75 in Montana and others.
The EPA advises that increasing highway speed by 10 m.p.h. can reduce fuel economy as much as 10 percent, yet there has been no change in its calculations to reflect the higher legal speeds.
The EPA has administrative authority to revise the fuel-economy numbers, as it did in 1985, but the agency says it lacks the funding to research what drivers get. The EPA also maintains that fewer than 100 people a year complain to the agency that their mileage doesn’t match the estimates, indicating its numbers are reasonably accurate.
“The honest answer is that we have not done a study since 1982. We probably should take a look at it, but it wouldn’t be cheap,” said Christopher Grundler, a deputy EPA administrator and head of the Ann Arbor, Mich., laboratory where the agency conducts fuel-economy and vehicle-emissions tests.
Grundler couldn’t estimate the cost of such a study, but added: “It would take time and money, but it would not save any fuel or reduce greenhouse gases, just give consumers better information.”
Even with variances from real-world fuel economy, the EPA estimates “allow consumers to compare relative differences in fuel economy among vehicles, and that’s what they are for,” he said.
“It would be interesting to find out what people actually get compared to the fuel-economy labels, but I don’t know if it would make a difference. I don’t know if, when the average person sees the numbers on the fuel-economy label, they actually expect to get that number.”
Tom Novak, a Batavia business owner, is one who did and is disappointed with his mileage.
Novak says he bought a Honda Civic Hybrid in April because of the EPA ratings of 48 city and 47 highway, but the closest he has come is 42 m.p.g. according to the trip computer on the gas-electric hybrid car.
If he divides the number of miles he drives by the amount of gas he pumps, his highest mileage at the pump is 40.3 m.p.g. from a mix of suburban and highway driving, and his overall average is 37.5 over more than 8,000 miles.
“My expectation was to get those [EPA] numbers, or at least get close to them,” Novak says. “I didn’t have any basis to expect anything different.”
In a recent report to Congress, the National Research Council said most consumers average less than the EPA estimates because the tests and correction factors do not reflect today’s urban congestion, higher speeds and driving patterns.
The Bluewater Network, a San Francisco-based coalition of environmental groups that has petitioned the federal government to change the fuel-economy tests, also contends that the EPA overstates fuel economy.
“They are so far off that this is false and deceptive advertising,” said Russell Long, executive director of Bluewater Network. “Consumers are given a vast underestimate of the real expense of operating a vehicle.”
The federal government uses the EPA estimates to calculate Corporate Average Fuel Economy, which fines manufacturers that don’t average at least 27.5 m.p.g. on cars they sell and 20.7 on light trucks. CAFE, however, is calculated before the EPA reduces the city estimates by 10 percent and the highway figures by 22 percent. Environmentalists contend that CAFE overstates actual mileage by 20 percent.
“Auto companies are getting the benefit of an easy test, and the American people are being sold a bill of goods about the kind of fuel economy they will get,” said Dan Becker of the Sierra Club, which advocates more stringent tests.
The EPA’s response is the one it has used since the 1970s: Everyone’s mileage can vary because of different driving styles, road surfaces, weather, traffic conditions, fuels and other such factors, and some drivers do better than the EPA estimates.
“It’s impossible for any single test to simulate all possible conditions,” said Eldert Bontekoe, senior manager of the testing program. “We have to come up with one number that meets everyone’s needs across the country. Our goal is to provide a number that the consumer can use to compare vehicles effectively.”
A frequent criticism of the EPA’s methods is that the rubber doesn’t meet the road in its tests. They are conducted indoors on a dynamometer with a test driver accelerating and braking according to a procedure that is computer-monitored for consistency.
The drive wheels of a stationary vehicle spin a roller to simulate driving, similar to the emissions test in Illinois and other states. Because the dynamometer has only one roller, four-wheel- and all-wheel-drive vehicles are tested with the front or rear drive axles disabled or disconnected.
The EPA adjusts the dynamometer to reflect the vehicle’s weight, wind resistance, rolling resistance of the tires and frictional losses, all of which affect fuel economy. The manufacturers supply the data used to adjust the dynamometer, and the EPA verifies it on a spot basis.
Auto manufacturers generate most of the fuel economy estimates on their own by conducting tests using the same equipment and methods as the EPA.
The EPA verifies the industry’s estimates, each year testing about 250 of the more than 1,000 vehicles listed in its annual Fuel Economy Guide. The agency targets models that rank at or near the top of their class.
Consumer Reports magazine replicates the EPA’s city test by using the same 11-mile, 23-stop routine and same speeds but with three major changes. The magazine uses production vehicles purchased from dealerships instead of prototypes provided by the manufacturers, and it conducts the tests on its track and public roads in Connecticut.
Consumer Reports measures the amount of fuel a vehicle uses by splicing a flow meter into the fuel line to precisely measure consumption.
The EPA does not measure fuel consumption directly. Instead, it calculates consumption with a mathematical formula that compares the amount of carbon in the exhaust gas with the carbon content of the gasoline used for the test. The agency contends this method is more repeatable and accurate, capable of measuring consumption down to one ten-thousandth of a mile per gallon.
The magazine’s results from its city driving test are consistently much lower than the EPA’s.
For example, in a recent test of 2002 family sedans with V-6 engines, a Nissan Altima averaged 14 m.p.g. in Consumer Reports’ city driving test, 5 m.p.g. less than the EPA estimate, and a Toyota Camry came in at 13, 7 m.p.g. less than the EPA estimate. In a test of 4-cylinder family sedans, an Altima averaged 15 m.p.g. and a Camry 16. The EPA city estimate is 23 for both.
Consumer Reports also does a highway test at a steady 65 m.p.h. on a flat stretch of U.S. Highway 2 and consistently beats the EPA’s highway estimates. The magazine averaged 29 m.p.g. with the V-6 Camry, topping the EPA by 1, and 35 with the 4-cylinder, topping the EPA’s estimate by 3. With the Altima, the magazine’s highway test beat the EPA by 3 m.p.g. with the 4-cylinder and 4 with the V-6.
The magazine then does a 150-mile trip on public roads that encompasses city, suburban and highway driving. When it averages the mileage from all three tests, the result is usually close to the EPA’s city estimate.
The Camry V-6, for example, averaged 20 m.p.g. from all three Consumer Reports tests, matching the EPA’s city estimate. The 4-cylinder model averaged 24, one more than the EPA city estimate.
“We’ve always been much, much lower than the EPA on city mileage and tend to be higher on the highway,” said David Champion, Consumer Reports’ automotive testing director. Vehicle owners are more likely to see mileage similar to the magazine’s tests than the EPA’s laboratory simulations, he added.
“They’re a good comparison if you are looking at two cars and want to compare them side by side, but they’re not likely to be what you will get,” he said of the EPA estimates.
Road test undercuts mileage estimates
Our mileage varied.
The EPA maintains that most drivers come within in a few miles per gallon of the EPA’s fuel-economy estimates, and the Tribune did just that with two 2003 models.
A Lincoln Navigator averaged 11.3 m.p.g., essentially matching the EPA’s city driving estimate of 11, and an all-wheel-drive Pontiac Vibe averaged 25.5 m.p.g., coming within ounces of the city estimate of 26.
However, in both cases, nearly half the driving was at highway speeds on Chicago-area expressways, which the EPA says should yield better fuel-economy than running on surface streets. The EPA’s highway estimates are 16 m.p.g. for the Navigator and 31 for the Vibe.
The EPA estimates fuel economy for city and highway driving from indoor driving simulations conducted on a dynamometer, but the Kennedy Expressway was the main laboratory for the Tribune’s tests.
In rush-hour traffic on the Kennedy, speeds ranged from a walking pace to 65 m.p.h., but both vehicles also cruised for extended periods at 50 to 60 m.p.h.
Both also diced with cabs, buses and other vehicles on crowded downtown streets, ran short errands on suburban roads and cruised the Tri-State Tollway.
Compared with the EPA’s rigidly controlled tests, the Tribune’s was an unscientific sampling that could yield different results if attempted again because traffic might be lighter or heavier or the weather hotter or cooler.
But thousands of other Chicago-area motorists did the same type of driving at the same time as the Tribune, so their mileage probably varied, too.
— Rick Popely.
