Feds Should Preempt State Regulation for Self-Driving Cars
A technology once thought to be decades away is now being tested on public roads and clashing with state and local governments. Ford Motors has announced plans to release a full fleet of driverless cars by 2021. Google has been testing fully autonomous vehicles since 2014. In Pittsburgh, one can summon an autonomous Volvo using the Uber app. Other companies including Apple, General Motors, and Chrysler are also involved.
The rapid development of a technology that has great potential to save lives, increase mobility, and reduce pollution also creates a challenge for regulators. Current vehicle regulations assume that a human driver is in control of the vehicle. Driverless cars operate within the realm of soft law.
The National Highway Traffic Safety Administration (NHTSA) has attempted to fill this regulatory void with its Federal Automated Vehicles Policy. These nonbinding guidelines recommend that state regulations only deal with licensing, registration, insurance, and other roles that traditionally belong to states. A 15-point safety assessment would be conducted voluntarily by companies and submitted to NHTSA.
Similar to current vehicle regulations that allow manufacturers to self-certify compliance, these guidelines provide basic standards without capping innovation. Self-certification has proved remarkably successful in protecting consumers and improving technology for 50 years.
NHTSA needed to act because, instead of embracing the benefits of innovation, many state and local governments are stifling innovation through overly-burdensome regulations. Unfortunatly, suggestions from NHTSA fail to correct this censorship of innovation.
For example, California state law uses legislation to dictate which companies are allowed to test self-driving cars. While there are currently 33 corporations developing autonomous vehicle technology, the state has granted permission to only 15 companies for testing. The state’s regulatory code requires that a driver is present to monitor the vehicle and that the car still has brake pedals and a steering wheel. Since technology is now being developed that will have no need for a human driver, this regulation is already outdated. In attempt to keep up with innovation, California has granted only one company, Easymile, the right to test shuttles without drivers, brake pedals, steering wheels, or accelerators on a single strip of public road.
Allowing regulators to grant de facto monopolies will significantly hinder the introduction of this technology to market. The unprecedented rate at which this technology has been developed can partially be credited to fierce competition – an autonomous vehicles arms race. Affordable self-driving vehicles will become a pipe dream if half of the competition is excluded from Silicon Valley.
Even worse, Chicago’s mayor Rahm Emanuel recently proposed a ban on autonomous vehicles. In a press release, city alderman Ed Burke defended the proposed policy by writing:
“We do not want the streets of Chicago to be used as an experiment that will no doubt come with its share of risks, especially for pedestrians. No technology is one-hundred percent safe.”
Burke’s fear of technology does not reflect reality. Of the 35,000 U.S. vehicle fatalities in 2015, 94 percent resulted from human error. In contrast, a May 2016 Tesla crash is the only death in an autonomous vehicle. Furthermore, in a study conducted by the University of Michigan Transportation Research Institute, every recorded accident involving an autonomous vehicle – 11 per million miles – was the fault of a human driver. Burke is right—nothing is one hundred percent safe. However, only one recorded fatality when moving at highway speeds is close.
Given these current regulatory trends, NHTSA needs to take a more assertive approach to regulating, which it will likely do after a public comment period on its non-binding guidance. And NHTSA has reason to act—since vehicle automation is essentially a safety feature, it falls under federal jurisdiction.
When moving forward with finalized regulations, NHTSA needs to keep in mind that fully autonomous vehicle technology is still developing. Any comprehensive regulatory code that mandates certain design features will restrict innovation. As for the best way to move forward, the non-profit organization Securing America’s Future Energy, recommends developing performance-based standards. An example of this system already in use is the Federal Aviation Administration’s requirement that software components in small planes be “highly reliable” without specifying what type of software must be used. This allows companies to constantly develop better software to meet this standard.
NHTSA should work to clarify its 15 point safety assessment so it can be written as a national regulation. Once defined, outcome-based rules for priorities such as “crashworthiness”, “digital security”, and “data recording and sharing” give auto manufacturers and innovators the freedom to determine how to best achieve these goals.
While allowing states to set their own regulatory policies is usually desirable, federal preemption is necessary to establish regulations for product standards when there are coordination problems between states. This is necessary, as discussed in an American Enterprise Institute book edited by Richard Epstein and Michael Greve, in the rules for drug labeling, cell phone regulations, and financial market regulations. Similarly, in the market for self-driving cars, NHTSA admits that 50 different sets of regulations and rules would constrain the development of the product.
Car manufacturers would not create multiple variations of a technology in order to meet different state standards; instead, they would just adapt to the biggest market. The market power of California makes it likely that their regulations would become de facto law for the rest of the United States – similar to how the state’s zero-emission requirement prompted every major auto-maker to bring an electric car to market.
Any smaller states that impose radically different regulations on self-driving cars would prevent residents from accessing a potentially life-saving technology. When Vermont passed a mandatory GMO labeling law, food manufacturers pulled 3,000 products out of the state rather than comply with the costly law.
One can always hope that voters will choose not to reelect officials who passed detrimental bans, but this is not always the case. Many incumbent industries benefit from government stifling of competition and are willing to spend to remain protected. The proposed ban on self-driving cars in Chicago was heavily financed by the city’s taxi industry in order to block autonomous Ubers similar to those in Pittsburg.
Innovation from private companies has brought about almost every great invention of the past decade. With proposed regulations in California that would require companies to report the data collected on every vehicle to the state and force companies to get permission for testing from each locality, national performance-based guidelines are necessary to prevent the stifling of beneficial innovation.
Allie Howell is a contributor to Economics21.
Interested in real economic insights? Want to stay ahead of the competition? Each weekday morning, E21 delivers a short email that includes E21 exclusive commentaries and the latest market news and updates from Washington. Sign up for the E21 Morning Ebrief.