Lake Wobegon CAFE

In the course of writing yesterday's posting, it occurred to me that the efforts to increase US Corporate Average Fuel Economy Standards (CAFE) face a basic problem. Notwithstanding our concerns about energy security and climate change, some simple mathematical truths have been absent from much of the public discussion on the subject of CAFE, and they have something in common with the "Lake Wobegon Effect," in which all the kids in Garrison Keillor's fictional town are above average. Raising the average fuel economy of the country's light vehicle fleet by 10 miles per gallon doesn't sound very daunting, until you start looking at what the mathematics of averages implies for the type and number of cars consumers must actually buy.

It all starts with how you go about raising the average of any set of numbers to a higher target level. There aren't many options:

1. Add additional members that are above the new target.


2. Remove members that fall below the current average.


3. Various combinations of 1 & 2.

If this seems painfully obvious to you, please bear with me, because applying these simple ideas to the world of automobiles produces some awkward realities. Consider SUVs, which been a major factor in increasing US petroleum consumption in the last two decades. Since 2004, however, sales of large SUVs have fallen. At the same time, sales of "crossover vehicles"--smaller, mostly car-based vehicles that combine features of station wagons, vans, and SUVs--are up significantly. That suggests that consumers have taken the gas price signal to heart and are downsizing their old SUVs. But will that improve overall US fuel economy, let alone move us toward a new 35-mpg standard?

Suppose that John Customer has decided to trade in his family's 2001 Ford Explorer for a 2007 Ford Edge, a fairly typical crossover. The Explorer had an EPA rating of 15 mpg; the Edge gets 20 mpg overall, using the old EPA methodology. John might assume he's doing his small part to improve the US fleet mpg by 5 mpg, while saving some money at the gas pump, but there are two problems with his logic. Since his new car's mileage falls below the current average CAFE, it will actually drag it down further--or negate the efficiency gains from a higher-mpg car sold to someone else. Looking beyond the new car fleet, which is all that CAFE measures, unless John takes his Explorer to the scrap yard on the way to the dealership to pick up his new Edge, someone else will be driving his old 15 mpg SUV for years to come. On balance, then, John's switch to a crossover might look like a modest improvement, but it doesn't even contribute to maintaining the current fuel economy average, let alone approaching a stricter target.

Nor is this problem confined to SUVs. In 2004 I bought an Acura sedan, which provides the 5-star crash safety I was seeking and handily exceeds 30 mpg on the highway. But this car came with an EPA sticker estimate of 23 mpg, and because most of my driving is around town, I have only averaged 21 mpg, despite having a light touch on the accelerator and being very sparing of the brakes. This kind of performance is typical for other popular 6-cylinder sedans, such as the Chevrolet Impala or Malibu, or Toyota Camry.

What does this suggest about the kind of cars we're all going to have to start buying, if the US adopts a 35 mpg CAFE target, on the way to moving the entire US car fleet up to that level eventually? It will require a much bigger change than just switching from SUVs to crossovers. By the millions we'll need hybrid crossovers, diesel crossovers, and maybe even diesel hybrid crossovers. Unless the half of the market buying small and large SUVs does a lot better than its current 22 mpg, then in order to meet a 35 mpg overall target, the other half must achieve Prius-like efficiency of at least 48 mpg--probably requiring some of the passenger car fleet to exceed 60 mpg. And unless we want to wait until 2030 for the performance of the entire US light vehicle fleet to reach these levels, we may also need an accelerated-retirement program for the least efficient cars on the road, based again on the simple math of averages. The bottom line is that we must all shortly start buying cars that are truly more efficient than the average, or we will have to look elsewhere for our energy and emissions savings.