The latest paper by the Center for Automotive Research is not what it thinks it is

Last month, the Center for Automotive Research (CAR) released a report on the potential economic effects of EPA/NHTSA’s 2017-2025 light duty rule. Although CAR receives funding or research support and contributions from the Alliance of Automobile Manufacturers, Global Automakers, and individual auto manufacturers, the organization says that this report “was internally funded by the Center for Automotive Research through its Internal Research and Development funds so as to maintain objectivity of the analysis.”

Unfortunately, it’s hard to see much objectivity in the analysis. The report’s analysis hinges on a set of inflated cost estimates that ignores both accumulating data on real technology costs to automakers under the new fuel-economy standards and the extensive empirical analysis of technology costs that EPA and NHTSA just released. The result is an inherent bias toward the conclusion that the standards are overwhelmingly likely to be economically harmful.

The analysis in the report proceeds as follows:

  1. Estimate the required improvement in vehicle efficiency from MY2016 to MY2025 (12.7mpg, or 46.5%, from a baseline of 27.3mpg)
  2. Assume the increased cost to consumers of the improved technology at three levels ($2000, $4000, $6000)
  3. Estimate the consumer valuation of the savings due to the fuel efficiency technology using a payback period of 3 years, three fuel price scenarios ($2.44/gal, $3/gal, and $4.64/gal, from EIA’s 2015 Annual Energy Outlook), and about a 9% increase in VMT for MY2025 vehicles (20% rebound effect)
  4. Estimate vehicle own-price elasticity, using data from 1953-2013 (-0.79 in the short-run, -0.61 in the long-run)
  5. Assume the full cost of technology is passed along to consumers, and convert the net costs into reductions in vehicles sales
  6. Using economic data from another CAR report (table 2.1), correlate the decrease in vehicle sales to drops in direct and indirect employment in the U.S.

It’s a reasonable approach—except for those grossly inflated cost estimates in step two that skew the end results.

How grossly inflated? The high estimate, $6000, would, on average, pay for every car, SUV, and pickup to be hybridized and, as my colleague John German puts it, “blow the doors off the standards.” Even the low estimate, $2000, is much higher than the $1565 that EPA and NHTSA estimate it will cost automakers on average, per vehicle, over the remaining life of the regulation, 2016 to 2025. And it’s more than twice as much as the $894 average per vehicle the agencies estimate it will cost automakers for the improvements needed over the final four years of the rule, 2021 to 2025—which is the specific question being addressed under the midterm evaluation right now. (See the draft TAR, Table 12.44).

How, you might ask, did CAR come up with these cost numbers?

The report relies on a 1991 study by David Greene, which found that automakers could improve their CAFE fuel economy level by increasing the sales price of less fuel efficient models (engine-transmission combinations within carlines) while simultaneously decreasing the price of more fuel efficient models. Greene concluded that this pricing scheme is effective in the short-run for fuel economy improvements of up to 1 mpg, and would cost $100–$200 (in 1985 dollars). But, Greene also found, for fuel economy improvements greater than 1 mpg, pricing out less-efficient vehicles generates increasing losses for automakers. He concluded that improved technology and design changes are by far the more cost-effective solution for long-term, large fuel economy improvements—like those required for meeting the 1978–1985 standards (18 mpg to 27.5 mpg) or 2017–2025 standards (35.5 mpg to 54.5 mpg). That is, the pricing scheme is useful if, for example, an automaker falls shy of meeting its CAFE standard by a fraction of 1 mpg one year, because it could raise its fleet average by slightly increasing the price of less efficient vehicles and slightly decreasing the price of more efficient ones. But if the automaker needs a bigger, sustained increase, that strategy won’t work.

CAR’s report converts Greene’s estimated cost for a 1 mpg fuel economy increase to 2010 dollars, from $100–$200 to $200–$400, then multiplies that by 12.7 mpg—the difference between 2016 and 2025 “real world” mpg, defined as 80% of the lab-tested fuel economy. Presto! A cost range of $2540–$5080.

In other words, they do exactly what their own source says you cannot do with his conclusions about the short-term pricing tactic, and they ignore his analysis about the long-term costs for large improvements in fleet average fuel economy.

Worse, the CAR report does not consider, in any way, the actual costs of recent fuel efficiency technology available to meet the standards, as opposed to the costs of what’s basically a twenty-five-year-old retail-price manipulation strategy. The draft TAR per-vehicle cost of compliance is lower than in the rulemaking (which is 4-5 years older; see EPA RIA Table 5.1-8). The agencies base their calculations on extensive and detailed empirical information, including teardown studies, as well as modeling, to accurately assess technology costs. So their reduced cost estimates accurately reflect declining technology costs. There is ample evidence that costs for certain technologies may be lower still than in the draft TAR, and that technology costs will continue to decline due to constant improvements in computer simulations and controls, and more efficient assembly lines. In fact, CAR has its own reports on powertrain and lightweighting technology costs, which could have served as a better basis for estimating costs in the report under discussion. It is an egregious oversight that the authors ignored all real technology cost estimates, especially the thoroughly researched ones in the TAR, which was released two months before the CAR report.

What would happen if CAR’s report used the technology costs from the TAR, $1565?

With these technology costs, and all other assumptions the same, the break-even future fuel price is $2.97/gal; anything greater leads to net savings over the 3-year payback period used in the study. Net savings then results in greater vehicle sales and higher employment. At $2.44/gal, consumers may face net costs of $278 over the 3-year payback period, which is one third lower than the least net cost calculated in the CAR report. Thus, even under the least expensive fuel price scenario, any economic effects of the cost of fuel efficiency technology due to reduced vehicle sales alone are extremely small.

Finally, the report barely acknowledges other important factors that impact the net costs and savings to consumers and benefits to the economy. Consumers may value fuel efficiency technologies for reasons other than fuel savings, such as better performance due to weight reduction, more low-speed torque and acceleration with a turbocharged engine, quieter operation with more transmission gear ratios, and better cabin thermal management. Resale value of more efficient vehicles is higher than that of less efficient vehicles. Reduced petroleum usage benefits the U.S. economy and energy security. Reduced GHG and other pollutant emissions have large environmental benefits. Less spending on fuel beyond the 3-year line means much more money in consumers’ pockets to spend on other items, with increased jobs throughout the entire economy.

But the basic problem is that the whole report rests on a false premise about the costs of meeting the standards. And whether that reflects a lack of “objectivity” or something else, it renders the report completely unreliable as an analysis of the employment impacts of the CAFE/GHG standards.

Engineering & manufacturing