Who Should Regulate Autonomous Vehicles?
With the release of NHTSA’s Federal Automated Vehicle Policy (FAVP) and its proposed Model State Policy, the rubber has hit the road on the question of who should regulate autonomous vehicles, where the line between federal and state authority should be drawn, and how quickly regulators should act. Historically, states have regulated drivers, and the federal government has regulated motor vehicles. But when the car starts driving itself, does that require regulation of a driver or a motor vehicle? Should there be a single federal standard for autonomous vehicles or should states craft their own rules? How quickly should the federal and state regulators act to nail down formal rules governing autonomous vehicles as the technology continues to evolve?
NHTSA’s View
NHTSA has articulated its view in the FAVP that the “division of regulatory responsibility for motor vehicle operation between Federal and State authorities is clear” and that “NHTSA generally would regulate motor vehicles and motor vehicle equipment (including computer hardware and software that perform functions formerly performed by a human driver) and the States would continue to regulate human drivers, vehicle registration, traffic laws, regulations and enforcement, insurance and liability.”
In the FAVP, NHTSA states that the “Vehicle Safety Act expressly preempts States from issuing any standard that regulates performance if that standard is not identical to an existing Federal Motor Vehicle Safety Standard regulating that same aspect of performance.” Thus, “if NHTSA issued an FMVSS setting performance requirements for HAVs, then a State could not have its own performance standards on the same aspects of HAV performance unless they were identical to NHTSA’s standards.”
But the “if” in that statement is precisely the issue today. NHTSA has not issued new FMVSS regarding HAVs and has noted that while it intends to engage in formal rulemaking in the future, it issued this voluntary guidance in order to maintain flexibility as the technology develops. NHTSA is using what has been called an “agency threat” or “soft law” approach to regulation (see Adam Thierer and Tim Wu, among others, on this subject), hoping to postpone formal regulations until more is known about how this technology will develop. While the industry has generally applauded this approach, one downside may be NHTSA’s inability to preempt state regulation in the interim. States are actively looking to impose their own rules to protect public safety and address consumer concerns, as well as to compete for business. Moreover, as discussed below, states may enact regulations that turn NHTSA’s voluntary guidelines into mandatory rules, thus eliminating the benefit of NHTSA’s light touch approach.
Recognizing it cannot compel states to stand down, NHTSA’s Model State Policy in the FAVP provides a framework for state regulation, in the hope that states will at least adopt rules that are consistent and avoid conflict with federal policy. In the section titled “Application for Manufacturers or Other Entities to Test HAVs on Public Roadways” the framework provides “the application should state that each vehicle used for testing by manufacturers or other entities follows the Performance Guidance set forth by NHTSA and meets applicable FMVSS.” It also provides that “the application should include the [entity’s] safety and compliance plan for testing vehicles, which should include a self-certification of testing and compliance to NHTSA’s Vehicle Performance Guidance for the technology in the test vehicles under controlled conditions that simulate the real world conditions…to which the applicant intends to subject the vehicle on public roadways (e.g., a copy of the summary Safety Assessment submitted to NHTSA per the Vehicle Performance Guidance).”
After NHTSA issued its FAVP, numerous industry participants noted in public comments the ambiguity created by the Model State Policy and the concern that states would effectively make compliance with NHTSA’s Safety Assessment Letter process mandatory instead of voluntary. In response to comments, NHTSA put out an FAQ statement on its website “clarifying” that its FAVP is voluntary and that States should not codify “the Vehicle Performance Guidance or any of the specific elements of the Safety Assessment Letter.” The FAQ reiterates that “NHTSA strongly encourages States to allow NHTSA alone to regulate the safety and performance aspects of HAV technology and vehicles.”
NHTSA notes in the FAQ that “many State government officials expressed their views that while they were generally comfortable with the Federal Government having the responsibility for the regulation of vehicle and equipment safety, they would need confidence that vehicle manufacturers, testers and other entities were conforming to the safety approach envisioned in the FAVP.” Therefore, NHTSA notes that the Model State Policy provides that states “may request submission of confirmation that an automaker, tester or other entity has followed the principles set forth in the Vehicle Performance Guidance.” Since the Safety Assessment Letter process is voluntary, NHTSA notes that states “could request alternative documentation in the event an entity does not submit a Safety Assessment Letter.”
NHTSA acknowledged that “absent the implementation of a new Federal Motor Vehicle Safety Standard that would preempt State action, States have the legal right and power to develop their own vehicle safety and performance criteria.” In other words, until NHTSA takes formal regulatory action, states can enact regulations governing HAVs that may be different from the process NHTSA envisioned. California is one state that is moving forward with its own regulations.
California’s Approach
California has been an early testing ground for HAVs and enacted formal regulations in 2014. After NHTSA released its FAVP, California published new draft regulations in October 2016 governing testing and deployment of HAVs. California’s draft regulations have faced industry criticism for two primary reasons, both of which illustrate the problem with allowing states to regulate ahead of NHTSA formal rulemaking.
First, the California regulations have been widely interpreted as making NHTSA’s voluntary guidance into a mandatory rule. The draft regulations provide that manufacturers must obtain a permit to test HAVs and to deploy them. For a testing permit application, California requires “the manufacturer shall submit with the application a copy of the safety assessment letter submitted to NHTSA as specified in the Vehicle Performance Guidance for Automated Vehicles.” But what happens if no such safety assessment letter is submitted to NHTSA and thus none is submitted to California? The regulation does not provide for this possibility, suggesting that compliance with NHTSA’s guidance is required. Similarly, for a permit for post-testing deployment of HAVs, California requires the manufacturer to submit “a certification that the vehicle complies with the Vehicle Performance Guidance in NHTSA’s FAVP” and “a copy of the safety assessment letter submitted to NHTSA.”
While this language largely reflects the intent of the NHTSA model policy, it does appear to make the NHTSA process mandatory. This is problematic because it appears to provide back door enforcement of rules that NHTSA otherwise claimed were voluntary. NHTSA is required to hear public comment and go through specified steps under the Administrative Procedure Act before it can make mandatory rules, which it has not done. Moreover, the industry may contend the regulation violates the California Administrative Procedure Act because California has not engaged in a full rulemaking process on the specific vehicle safety and performance standards set forth in the FAVP either. California Government Code Section 11340.5 provides that “no state agency shall issue, utilize, enforce or attempt to enforce any guideline…or other rule, which is a regulation…unless the guideline…or other rule has been adopted as a regulation and filed with the Secretary of State pursuant to this chapter.” In other words, agencies cannot avoid the rulemaking process by enforcing a guideline or practice instead. Allowing California to sidestep the rulemaking process in this way seems unlikely to survive scrutiny. An easy fix would be to revise the draft rules to make the submission voluntary. However, that would leave California with no verification of safety standards if a manufacturer declines to follow the guidance. NHTSA’s decision to postpone formal regulations leaves open this possibility that states will try to fill the void either by making the FAVP mandatory or by enacting their own substitute provisions. Either result would defeat NHTSA’s rationale for using voluntary guidance at this stage.
Second, California’s draft regulations deviate from the FAVP in other ways, including specifying that deployment vehicles will need to be equipped with certain “autonomous technology data recorders” to collect data, and that manufacturers must release such data to law enforcement within 24 hours of a request. This requirement has been criticized as creating unworkable standards that may conflict with federal rules. Data recorders and data collection are items discussed in NHTSA’s Vehicle Performance Guidance, but NHTSA has not yet settled on the specifics of what data should be collected, when data recorders should engage, how the data would be stored and shared, and many other details it intends to work out with the industry moving forward. If California enacts its regulations now, manufacturers will need to comply with these requirements in addition to, and perhaps in conflict with, any voluntary guidelines set by NHTSA. This would be precisely the patchwork of rules that NHTSA was trying to avoid.
California’s draft regulations illustrate that NHTSA’s attempt to encourage consistent state rules without forcing the issue through federal preemption is a difficult path, especially as additional states enact more statutes and regulations.
The Road Ahead
Should NHTSA abandon its voluntary guidance approach and move more quickly to adopt formal regulations to avoid a patchwork of state rules? Or is there some benefit to allowing exploration of different approaches at the state level?
At least one industry participant has taken the position that state level regulations will allow for more innovation while the technology develops. In its public comments to NHTSA, Uber contends that “regulating drivers who operate vehicles has always been the purview of individual states, and outside the federal authority of NHTSA.” Uber contends that states should be allowed to continue “to investigate and pilot different approaches as they deem appropriate.” Given how quickly the technology is evolving, Uber notes “it is premature to try to mandate a nationalized framework and shift responsibilities away from the states.” Uber has obtained the right to test its vehicles in certain states, and presumably does not want NHTSA to limit what it has been able to negotiate at the state/local level. If the new administration changes its tune on the FAVP, this concern could escalate and this view might gain additional proponents.
While HAVs are being developed and tested, there may be some advantage to allowing states to try different approaches to maximize innovation, as Uber suggests. This could result in some states allowing more aggressive testing regimes in order to attract new technology business. While this could backfire if safety concerns develop, it could also accelerate development of the technology. On the other hand, states may also enact anti-competitive rules designed to protect existing businesses in the state at the expense of innovation by newer technology companies. Several states have already proposed such laws. Thus, allowing different state rules will not necessarily promote greater innovation, and could be used to impose other political agendas such as preserving jobs, favoring existing manufacturers and their labor unions, or deterring foreign manufacturers.
Even if there is some advantage to allowing state-level innovation while HAV technology is being developed, once HAVs are ready to be manufactured and deployed it seems clear that a single federal standard will be required, which is why NHTSA regulates motor vehicle safety standards today. Presumably even Uber will be in favor of federal standards when it is ready to deploy fleets. NHTSA clearly intends to preempt the vehicle safety standards for HAVs at some point in the future, so does it make sense to allow states to create different rules now? These rules will be irrelevant once preempted and in the meantime will impose unnecessary costs on manufacturers as they are required to monitor and comply with many different state rules.
NHTSA has recognized this issue and tried to use the Model State Policy and its work with state governments to encourage states to defer to federal policy rather than enacting different rules. If that approach ultimately is unsuccessful, NHTSA is faced with two unappealing choices — rush to enact federal regulations and risk setting in stone policies that may limit innovation, or use voluntary guidance to evolve the rules but face a bevy of conflicting state laws in the interim imposing additional costs and confusion. NHTSA and other agencies will continue to face this regulatory dilemma as they try to fit the square peg of nascent industries and fast-moving technologies into the round hole of the current regulatory process regime, particularly where consumer safety concerns motivate states to step in and regulate. Ultimately this dilemma, and the inability of agencies like NHTSA to keep up with new technologies using outdated regulatory processes, may lead to creation of a new paradigm for regulating cutting edge technology. Perhaps if NHTSA had an on-ramp to regulations that permitted preemption but preserved its ability to make changes with input from industry in a timely manner, then both the regulators and industry could better focus on creating the right framework for autonomous vehicles to bring this new technology to market.
