The SELF DRIVE Act: 4 Questions left unanswered about federal regulation of autonomous vehicles

by on September 5, 2017 · 0 comments

Congress is poised to act on “driverless car” legislation that might help us achieve one of the greatest public health success stories of our lifetime by bringing down the staggering costs associated with car crashes.

The SELF DRIVE Act currently awaiting a vote in the House of Representatives would pre-empt the existing state laws concerning driverless cars and replace these state laws with a federal standard. The law would formalize the existing NHTSA standards for driverless cars and establish their role as the regulator of the design, construction, and performance of this technology. The states would become regulators for driverless cars and its technology in the same way as they are for current driver operated motor vehicles.

It is important we get policy right on this front because motor vehicle accidents result in over 35,000 deaths and over 2 million injuries each year. These numbers continue to rise as more people hit the roads due to lower gas prices and as more distractions while driving emerge. The National Highway Traffic Safety Administration (NHTSA) estimates 94 percent of these crashes are caused by driver error.

Driverless cars provide a potential solution to this tragedy. One study estimated that widespread adoption of such technology would avoid about 28 percent of all motor vehicle accidents and prevent nearly 10,000 deaths each year. This lifesaving technology may be generally available sooner than expected if innovators are allowed to freely develop it.

Are we preempting too early?

Currently twenty-one states have passed legislation related to driverless cars and 4 more have issued executive orders related to the issue. The content of these law varies from simple studies to requirements for testing to more general allowance of driverless cars on public roads. The requirements for operating driverless cars in these laws vary from highly restrictive in states like California to rather flexible and permissive in states like Florida.

While these laws may not be perfect, they provide a variety of examples to see the potential impact of regulation on driverless cars.  Additionally, they have shown general trends have started to emerge in how states are choosing to regulate the testing of driverless cars.

A federal intervention might be necessary were liability concerns to threaten to shutdown the potentially lifesaving industry over some early issues, but so far these threats have not surfaced. It also might be necessary if inconsistent standards about the use, design, and requirements for these vehicles become incompatible. In cases where pre-emption has shown to be necessary to allow the technology to develop, like Section 230 and the internet, traditionally we have at least seen what happens at common law first. This common law test has not yet occurred to know if such intervention would be necessary.

Are we preempting too much or too little?

The proposed federal legislation appears to preempt with a middle of the road approach. In this way the SELF DRIVE Act appears to risk preempting state laws that could lead to more rapid adoption of driverless cars. For example, some states, like Florida, have already passed laws that allow for private operation of highly autonomous vehicles. Other states, like Michigan, have also looked to address liability issues in their laws. The proposed federal legislation would supersede these more open standards for a middle of the road approach.

At the same time, the proposed federal law may not preempt enough. Driverless cars may change the traditional owner-operator connection that served as the basis for state regulation and licensing. The proposed legislation expressly does not pre-empt the state role in licensing and insurance. The law also continues to allow states to set higher standards for performance which could allow the more restrictive and precautionary regulations of states like California to survive while the more permissionless laws are pre-empted. If autonomous vehicles represent a fundamental change to ownership and a new least-cost complier, then maybe this relationship should also be reexamined.

Are we giving a federal regulator too much power?

Much of the proposed legislation delegates power to the National Highway Safety Transportation Administration (NHTSA). This raises concerns about the continued growth of the administrative state and the apparent preference of Congress to choose delegation over its own legislative authority.

While the bill adopts initial standards for driverless cars, the delegation of authority to NHTSA transfers the power for such regulation from Congress and the states to an administrative agency. The proposed legislation would place regulation of not only safety and design, but also cybersecurity for these vehicles with NHTSA. While NHTSA already regulates the automotive industry, this new delegation raises two questions. First, if the goal of delegation is to allow experts to make decisions, is NHTSA and expert when it comes to regulating cybersecurity of driverless cars? Second, how would NHTSA regulate and interact with those autonomous vehicle developers who are not currently automanufacturers and could this new delegation have spillover effects?

Does federal preemption provide legal clarity or not?

One of the arguments for federal preemption of state laws is that it provides legal clarity. This legal clarity, in turn, is seen to provide innovators with more certainty and more stability in which to pursue their innovation. In general, this is true, but the current patchwork of automotive, insurance, and tort law at a state level may not render certainty in this particular instance.

Insurance and tort law are primarily governed at a state level. Past attempts for federal preemption in these areas have failed. (For example a late 1990s push to preempt states from allowing no fault auto insurance.) As a result, states have the greatest common law experience in how to handle such issues when an accident involving a driverless car inevitably happens. If a federal law impacts how such vehicles are viewed, it may actually increase uncertainty regarding the existing common law for such issues.


There are still legal and regulatory questions to be worked out to both encourage and respond to driverless cars. The current situation at both a state and federal level raises questions as well as answers them. Still autonomous vehicles are poised to become one of the greatest innovative success stories of our lifetime.

Previous post:

Next post: