Regulators Should Think Twice Before Handing Out Pollution Credits for Self-Driving Cars

May 3, 2018 | 3:04 pm
Don Anair
Deputy Director & Research Director, Clean Transportation

A new report out by Securing America’s Future Energy (SAFE) suggests that automakers should get credits towards meeting emission and fuel economy standards for connected and automated vehicles (AVs) and related advanced driver assist systems—technologies that may or may not save any fuel. Doing so would not only increase pollution and fuel use, but would seriously undermine the integrity and enforceability of regulations that have delivered enormous benefits to our environment, our pocketbooks, and our national security.  The tens of thousands of traffic related fatalities every year in the U.S. demands that automakers and regulators must continue to make our cars safer.  But trying to encourage greater deployment of safety technologies by undermining pollution standards is the wrong approach.

Here’s why regulators should reject giving emissions credits to manufacturers for deploying safety and self-driving technologies.

Including emissions credits for safety and self-driving technologies in 2022-2025 vehicle standards would be a windfall for automakers, resulting in less deployment of proven efficiency technologies and more pollution.

There are more questions than answers about the potential impacts of various safety technologies and self-driving capabilities on vehicle and overall transportation system emissions, which I’ll get into more below.  But for now, let’s just take a big leap of faith and assume that some safety technologies actually do lower an individual vehicle’s emissions.

One example is adaptive cruise control.  This technology automatically adapts a vehicle’s speed to keep a safe distance from a vehicle ahead and theoretically could perform more efficiently than a human driver.  It is widely available and featured on vehicles like the Toyota Camry, Honda Accord and Ford Fusion.  One study examined this technology and found changes in efficiency could range from +3 to -5 percent during various types of driving. While there is some evidence that under certain conditions there might be a slight fuel economy benefit from this technology when it is in use, that same evidence indicates that increased fuel use and emissions are also possible.

In another recent study of self-driving cars, researchers found that while eco-driving capabilities could potentially provide savings, the increase in electric power demand, added weight, and aerodynamic impacts of sensors and computers would increase fuel use and emissions.  Both of these examples demonstrate the importance of testing and verifying any assumed change in emissions from the deployment of safety and self-driving technology as emissions reductions are anything but certain.

But even if credible testing and data were available, giving off-cycle credits for this technology within existing standards would be a giveaway to the auto industry.

Why? Adaptive cruise control is already being deployed on millions of cars – 1 in 5 new vehicles produced for the US market in model year 2017 were equipped with adaptive cruise control. Automatic emergency braking is another example, where automakers have already made commitments to make it standard on nearly all cars by 2022. Giving credits for these technologies would be a windfall for manufacturers and result in less deployment of proven fuel efficiency technologies.

The ICCT also identified this issue of providing credits for tech deployment that is already occurring in their review of the current off-cycle credit program and concluded that the program greatly reduces the deployment of other efficiency technology. They also identified the lack of empirical evidence to validate claimed fuel economy and emissions benefits from several technologies already included in the program as another big problem. And currently there is little empirical data to validate any efficiency benefits of safety and self-driving technologies.

Providing credits for emissions and fuel consumption impacts that are difficult to measure and not directly related to a vehicle – like possible impacts on traffic congestion—would increase pollution and undermine the standards.

Expanding the off-cycle program for safety technologies that might directly impact a vehicle’s emissions is just the tip of the iceberg.   The off-cycle credit program, like the vehicle standards in general, is limited to emissions directly related to the performance of a vehicle. But some automakers, and SAFE, are interested in allowing credits based on potential changes in emissions from the transportation system as whole. For example, automakers could earn credits toward compliance with vehicle standards for some future changes in traffic congestion that might result from the deployment of improved vehicle safety technologies. This would be a major change to the per-vehicle basis of the fuel economy regulations that were established in the 1970’s.

There are several serious problems with including speculative, indirect emissions impacts in existing vehicle standards.

1. Providing credits for emissions reductions that may or may not ever happen in the future will increase pollution in the short term and may never result in emission reductions in the long term

We only need to look back at the flex fuel vehicle (FFV) loophole to find an example of this kind of failed policy. Automakers were given fuel economy credits for selling cars capable of running on fuel that is 85 percent ethanol (known as E85), under the theory that this would help drive E85 to market and we would use less oil. Several automakers used it as a compliance strategy and avoided investing in other fuel efficiency technologies. But the cars almost never actually used E85, which means instead of getting more efficient vehicles, we got more oil use. The increased fuel consumption resulting from the FFV loophole is estimated to be in the billions of gallons.

Crediting future emissions reductions based on hopes and dreams has been tried before and doesn’t work.

2. Ignoring the potential negative impacts from self-driving technologies is a HUGE problem.

Self-driving cars have the potential for both positive AND negative impacts on pollution and energy use.

The biggest X-factor is how drivers will respond to these new technologies, which make vehicles safer, but also makes them easier to drive (or not drive at all as the case may be). A paper by Wadud et. al examined a range of direct and indirect impacts self-driving vehicles could potentially have on emissions.  And there are several possibilities, some of which could reduce emissions while others could increase emissions dramatically (see figure).   Increased emissions could result from higher highways speeds enabled by increased vehicle safety, increased vehicle size or features as drivers expect more features in their vehicles while their car drives them around, and most importantly, increases in the amount of vehicle travel overall.  Combined, these effects could increase emissions by more than 100% according to the study.

Automated vehicles could have both positive and negative impacts on energy consumption and emissions. Wadud et al.

We’ve already experienced increased highway speeds as vehicles have become safer with seatbelts, air bags and a host of other safety technologies.  And it’s not hard to imagine increases in vehicle miles traveled as cars take over the task of driving so we can do other things.  Just think about for a minute—what different choices might you make if you didn’t have to drive your own car?  Living farther from work or taking that extra trip during Friday rush hour might not seem so bad anymore when you can read a book or watch a movie while your car chauffeurs you to wherever you want to go.

Based on the current scientific literature, SAFE’s estimate of potential efficiency improvements from automated vehicles is misleading at best. Their analysis ignores any possible disbenefits, like increased vehicle travel, even while specifically acknowledging AVs “can also give drivers one thing of tremendous value to most Americans – an increase in personal or productive time”. The analysis also uses the upper range of efficiency benefits from a handful of studies estimated over limited driving situations, and inappropriately applies them to all driving.  The conclusion that a handful of safety technologies could reduce emissions 18-25%  across the entire vehicle fleet is not supported by current evidence, ignores any other effects of self-driving cars, and is not a sound basis for policymaking decisions.

My point isn’t that we should prevent self-driving technology and the many potential benefits it could deliver if done responsibly.

But vehicle standards aimed at reducing emissions and fuel consumption shouldn’t include credits for potential positive changes to transportation system emissions while ignoring the negative ones.

3. Finally, regulatory enforceability and accountability—the key to the success of today’s vehicle standards—would be severely undermined

The effectiveness of vehicle standards, any standards for that matter, is having effective enforcement which ensures regulated entities are all participating on a level playing field and that the actual benefits of the standards are realized.  We’ve seen the importance of enforcement over the decades as automakers have been held accountable for the performance of their products. Think ‘VW diesel scandal’ for one, and the numerous examples of erroneous fuel economy labels (Ford and Hyundai-Kia to name just two). These enforcement actions have one important thing in common: regulators were able to perform tests on the vehicles to determine if they were performing as the automakers claimed, and demonstrate that they were not.

Current vehicle standards are robust because they are predicated on direct emissions and fuel savings benefits that are verifiable on a vehicle level. An automaker makes a car, it’s tested, and they are held accountable for the results. How might a regulator, or an automaker, test and verify the congestion impacts of an individual Cadillac STS with Super Cruise?

Providing credits to automakers for emission reduction benefits that cannot be verified or attributed to an individual manufacturer, nevermind an individual vehicle make or model would be a massive change in approach to the program introduced through a mechanism – the off-cycle credit provisions – which was never intended to be more than small part of automaker compliance.

Where’s our insurance policy?

SAFE makes the case that giving away credits to automakers now, even without proof that these technologies reduce fuel use and emissions, is worth it because it would allow EPA and NHTSA to run a research program to understand the impacts on fuel economy of self-driving technology. But why should we accept increased pollution for collecting information? A better path forward for regulators is to indicate their intention to consider the direct vehicle emissions and fuel economy impacts of safety and self-driving technology in setting post-2025 vehicle standards and implement a testing program now to collect the necessary data to see whether giving credits for these technologies is appropriate. This would motivate automakers to do their own testing and to work with EPA and NHTSA to develop appropriate test procedures for ensuring the claimed benefits are actually occurring.

If safety and self-driving technology off-cycle credits are a proposed solution to the current impasse over 2022-2025 vehicle standards between federal regulators, the auto industry, and California, then we all need to be clear about the costs. They would provide windfall credits to auto companies for something they are already doing, while stalling deployment of proven efficiency technologies and increasing emissions.  If indirect changes in transportation system emissions and fuel consumption are included, such as some theoretical impacts on congestion sometime in the future that may or may not happen, the move would risk undermining the foundation of the standards themselves.

We should not be forced to make a choice between improving vehicle safety and reducing emissions. We need to protect the public from vehicle crashes and protect the public from pollution. If there is proven safety technology that is saving lives, automakers should deploy it and safety regulators should require it. But moving from a regulatory structure that is built on verifiable and enforceable emission reductions to one that is based on speculation and indirect impacts is a dangerous move that should be avoided.