Lee Schipper contemplates recent trends in automobile fuel economy. Photo by EMBARQ.

This talk reports on several recent analyses of important trends in automobile fuel economy and usage as well as other modes of travel. The work reported on was supported by grants from EMBARQ and the Climate and Energy Program of the World Resources Institute. Among recent trends we note are these:

* Improvement in both new vehicle test fuel economy and on-road fuel economy in the US from 2002 through 2008 after nearly a decade of stagnation, improvements driven both by tighter fuel economy standards for light trucks and higher fuel prices; Slower but longer-term improvements were observed in Europe and Japan, driven by voluntary agreements on fuel economy of new cars as well as a trend towards mini-cars in Japan.

* Clear stagnation or even slight declines in per capita travel of all modes (passenger-km) in Japan, Australia, N. America and Europe. The drop in car use per capita is striking, and only partly compensated by greater transit use.

* Further analysis of new diesel cars bought and diesel utilization and on-road fuel economy shows only marginal energy and CO2 savings have resulted from a large shift of new vehicle purchases in Europe from gasoline to diesel.

* In the U.S. the Cash-For Clunkers program yielded tiny fuel savings compared to trends in new car purchases from before the C4C program sprung to life. New cars bought under C4C were only 7% more fuel economic than those bought earlier in the year, and they replaced older cars that had only a few years left to lift. Compared to the average subsidy of $4000, then C4c was a very expensive way to save oil or CO2. (Read blog posts about the inefficiency of the Cash For Clunkers program on TheCityFix.com, here and here.)

* GMS has built a simple tool to depict both how US passenger transport uses fuel and emits CO2, and how proposed new policies might affect future emissions. We modeled a variety of programs, including C4C. There we found that even giving the most possible savings to each C4C transaction, the program savings were very small compared to the expected tightening of new car fuel economy standards. And while the C4C results go to zero in a few years, by the time the junked cars would have stopped running anyway, the impact of the new fuel economy standards increases with every year’s new vehicles.

The work led to a surprising finding about diesel cars in Europe. That despite much ballyhoo, new diesel cars in Europe or the on-road fleets are barely more fuel economy than their gasoline counterparts. In fact, new gasoline car fuel economy while still lagging that of diesel (measure in energy / kilometer to take account of the greater energy content of diesel fuel) has improved more than that of diesel since the 1990s. Accounting for the greater CO2 content of a unit of energy of diesel fuel (with or without the life-cycle emissions from refining and other stages of fuel preparation) narrows the “advantage” of diesel even more. Diesel cars are driven 50-100% more than gasoline cars. Some of that surprising difference is because of the kinds of people who selected diesel cars. Yet one must conclude that whatever hoped for energy and CO2 savings diesel cars were supposed to deliver in Europe, little or none is actually observed.

The diesel work has implications for programs that reward purchasers of low emission cars with tax credits or deductions, access to high-occupancy vehicle lanes (HOV) or toll free access to toll roads, as in the SF Bay Area. It is far harder to control for the overall fuel use or emissions from vehicles by rewarding only the purchase choice, but not behavior.

This work was used as background for analysis of the “Cash For Clunkers” (C4C) program from the summer of 2009. Most analysts compared the fuel economy of cars bought with those junked to estimate fuel and CO2 emissions saved. We show that this approach is wrong. Rather the savings from purchasing the new car depend on what other cars the household owned and how they were deployed once the new vehicle “bumped” others and the oldest was junked. The impact of C4C depends on comparing what would have happened without C4C with what actually happened. By that criteria, C4C had almost no impact, even if only evaluated over the nearly 700 000 vehicles exchanged.

While final results await release of 2008 data on US household car ownership and use, a number of surprising findings emerged:

* The average fuel economy of vehicles purchased under C4C was only 7% better than all others bought in the first half of model year 2009. Using a good model of monthly new car fuel economy, one researcher predicted an even higher fuel economy that would have been

* Since the “clunkers” were going to be junked within a few years anyway, any savings from this program can only be credited against the remaining lifetime of the junked vehicles. Since many eligible vehicles were otherwise road-worthy and worth more than the junking premium offered by C4C.

* Since the Administration has announced even stricter new vehicle fuel economy standards, any extra cars bought now may be less fuel saving than if they had been bought five years from now. However, Honda of North America reports that the C4C month was their best sales month in history; the following month was the worst. Thus it remains to be shown whether C4C stimulated any car buying beyond what would have occurred anyway.

* If the cars junked were the oldest and least roadworthy, i.e., worth less than the C4C premium for junking, then the savings from the program were even smaller, since these vehicles had the least number of miles or years left among all those eligible.

* Thus the impact of C4C is minor, at best the difference between the fuel economy of cars bought and those junked over the number of miles the junked car would still have run. More realistic is that the savings are only the difference between C4C fuel economy and the average of all other fuel economy bought in the year.

* The CO2 emissions from manufacturing are “wasted” when a car is junked. However, this was a small consideration because these are typically only 10% or less of the emissions from using the car. Ironically, junking a relatively new very low MPG car would save far more emissions, even including those of manufacturing, then junking a very old car, because the relatively new car would have used more fuel at a lower MPG than the old car.

The real lesson from C4C is that it is very difficult to use policies to “declare winners” by rewarding consumers for behavior they should be doing (and may have done) anyway.

Lee Schipper, Ph.D.
Global Metropolitan Studies, University of California, Berkeley
Precourt Energy Efficiency Center, Stanford University

Additional Resources and Information: