Déjà Vu: Shoddy Economic Study Touted by Automakers Flouts Facts

October 18, 2016 | 4:32 pm
Dave Cooke
Senior Vehicles Analyst

The mid-term evaluation of fuel economy standards is in full swing, and with the close of the public comment period on the Technical Assessment Report from the federal agencies, we now have a better understanding of the types of arguments being used by automakers as they try to weaken the federal passenger vehicle efficiency standards—to no one’s surprise, it relies on a lot of half-baked nonsense and fearmongering.

One statistic repeated over and over by automakers and their trade associations comes from a report that just happened to be released the same day industry was testifying at a House hearing on the standards—the claim is that up to 1.1 million jobs could be lost due to the 2025 standards, citing a report by the Center for Automotive Research (CAR).

This statistic is not just notable for its magnitude—it also runs counter to analysis released earlier this year which notes the positive impact these standards would have on automaker profitability in a variety of future fuel price scenarios.  Of course, it’s not the first time the Center for Automotive Research has used bad analysis to fearmonger around lost jobs…

Déjà vu all over again

This faulty analysis predicting 1 million jobs lost is really no surprise—five years ago, the same authors said that the standards adopted would lead to…1 million jobs lost in 2025.  So, at least they’re consistent? ¯\_(ツ)_/¯

They also predicted that while “this loss would happen by 2025 [it] would start to cumulate with the increase in standards in 2017.”  We’re basically at 2017 now…so how are we doing?  There are no hints yet of the massive increases in costs related to fuel economy standards, with estimates of the costs to meeting the standards actually dropping since the first CAR study in 2011.  Manufacturers are having absolutely no trouble meeting the standards, even exceeding them, despite gas prices having dropped significantly.  Automotive sales are booming, aiming for nearly 18 million units for the second consecutive year—about 5 million more vehicles sold than when the study was written, well exceeding the CAR projections.  And the automotive industry has added 450,000 jobs in the same timeframe.

So with all of the things that have gone right in the past five years for the industry, just how in the world does CAR think a catastrophe will strike this time?  To understand how the authors arrived at this ridiculous and incorrect assessment number, we have to dive deeper into the study.  There are a few basic assumptions on which the CAR analysis is based:  future gas prices, future technology costs, consumer response, and economic response—these latter three are the key sources of error in the study and deserve further detailing.

Cost assumptions are pulled out of thin air

The costs of improving fuel economy are a key parameter to any analysis of fuel economy standards—it goes directly toward the benefits to consumers.  The latest data shows that these standards continue to be cost-effective for consumers, even with low projected gas prices—so how did CAR come up with its evaluation of the cost of these standards?

Incredibly, the answer is pretty much that they pulled the values assumed for costs to the consumer out of thin air.  CAR assumed costs of $2000, $4000, and $6000 incurred by consumers to meet the fuel economy standards from 2017-2025.  The paper cited to support this was not based on any technology costs but rather an analysis from 1991 on pricing strategies used by automakers to shift consumer behavior, which is not pertinent to the question asked by CAR.

Furthermore, these costs represent significant overestimates for the expected costs.  The lowest assumed increase in cost was $2000 in 2025—accounting for the inflation assumed in the CAR report, this is equal to about $1500 in 2013$.  In the TAR, EPA’s projected costs of compliance were less than $1300 in 2013$.

In other words, the lowest cost scenario assumed by CAR in its analysis was actually higher than the agencies’ estimates!

CAR’s own data sources show that consumers value fuel savings more than assumed

The underlying principle behind the CAR analysis is that costly technology will not pay back in fuel savings for consumers, so consumers will buy less vehicles, resulting in less jobs.  However, in their assessment of how consumers evaluate future fuel savings, they ignore half the cited references and exclude anything published since 2011, including two papers which are actually updates to the cited work.  The authors do this by ignoring the studies which focus on a discount rate (rather than assuming that 100 percent of the fuel savings anticipated over the ownership of a vehicle will be achieved, consumers assume an implicit “discount” of those savings—this yields a discount rate).  In fact, when those studies are included in the estimate, the implied payback a consumer is looking for falls to within the life of the typical car loan based on the studies cited by CAR (5-6 years).

A recent consensus study by the National Research Council of the National Academies of Science, Engineering, and Medicine found that the literature shows significant uncertainty around how consumers value fuel savings:  “The results of recent studies find that consumers’ responses vary from requiring payback in only 2 to 3 years to almost full lifetime valuation of fuel savings” (Finding 9.3).

Once again, CAR used an estimate at the extreme end of a range of data in order to achieve a conservative, industry-favoring result instead of utilizing a value consistent with the literature.

Modeling is only as good as your inputs…and these are not very good

An age-old adage when it comes to models is “GIGO”—garbage in, garbage out.  Massive inconsistencies throughout the report and fundamental flaws in the work itself on a number of issues completely undermine the analysis.

The authors flip-flop repeatedly throughout the report on the use of “nominal” and “real” prices (nominal prices are the value of today’s goods in current dollars, while real prices include the impacts of inflation).  For example, in assessing how quickly the standards would pay back, they assumed gas prices in real dollars but costs in nominal dollars—this has the effect of overestimating payback time by nearly 25 percent!

As for the models themselves, there are a few primary flaws in the approaches taken by CAR.  Among other issues with the models, CAR conflates expenditure (which is a measure of both price and quantity) with sales—this leads to an overestimate of the impact on vehicle production.  Of greater concern, however, is the way in which the errors in assessing these questionable impacts are then compounded by inaccurate translation into job losses.

Rather than pursue any rigorous macroeconomic analysis, the authors choose to use a simplistic multiplier approach—i.e., there are x millions of vehicles being produced, y million direct jobs, and z indirect jobs, so changes in x result in a directly proportional change in y and z.  This is inappropriate under their assumption of a rosy economic picture due to the fungibility of jobs in a functioning economy.

As an example of why more rigorous analysis is needed, we can look at a previous macroeconomic analysis of the jobs resulting from these standards—while the standards would lead to significant job losses for the oil and gas sector, for example, those job losses pale in comparison to job creation in all other sectors, resulting in nearly 600,000 net jobs gained as a result of the standards.  This is exactly why in a fully-functioning economy you can’t just point to a single sector as the authors here have done but instead must look at the impacts on the entire economy.

I’m shocked—shocked!—to find industry up to its old tricks again

It’s no surprise the automotive industry would hang its hat on the Chicken Little posturing found in this report from CAR to engage in some good old-fashioned fearmongering. But it’s also pretty hard to take seriously given the numerous holes in the analysis and their record on previous predictions.

The standards are working, the automotive sector is doing well, and the latest data shows that we are well on our way to meeting the 2025 standards.  And we’ll continue reminding the agencies of those facts to counter the nonsense that industry tries to pass off as analysis.