*Marc Scribner co-authored this piece.

The City of San Francisco is special. How special? In spite of having no evident expertise, the City’s mayor, Mark Farrell, has asked the manufacturers of “autonomous vehicles” to submit to local “safety assessment exercises” before deploying their vehicles. While there will be a strong urge among many Bay Area-based firms to go-along with the mayor’s demand, they should resist. Compliance may seem benign, but risks setting a dangerous precedent for the industry as it moves toward nationwide deployment.

The crux of the problem is twofold. First, local jurisdictions simply are not qualified to evaluate what is “safe” behavior when it comes to vehicle design, safety or performance. For this reason, state transportation regulators have smartly deferred to their federal counterparts at the National Highway Traffic Safety Administration (“NHTSA”) on such issues. And, to their credit, both NHTSA and Congress have been actively working to bring highly-automated vehicles (“HAV”)  into the existing regulatory fold.

For that reason it’s astonishing, if not unsurprising, that San Francisco’s political establishment seems to think it can master what state regulators recognize is beyond their core area of competency in the mere weeks before vehicles receive state deployment permits from the California Department of Motor Vehicles. To be clear, San Francisco cannot and will not be able to master anything of the sort.

What’s more, in the letter, the City provides a bogus rationale for the “safety assessment exercise.” The City claims that the exercise would provide manufacturers an opportunity to educate first responders and others about how to interact with highly-automated vehicles in the event of an emergency.

Yet, that requirement is already met by the filing of a “law enforcement interaction plan” at the state-level as a condition of being granted a permit. In other words, as soon as permits are granted, San Francisco will be notified of where vehicles will operate, how vehicles will operate, and the ways in which first responders should interact with them in the event of an emergency. Farrell’s “safety assessment exercise” is clearly redundant of existing state-level regulation.

Therefore, since none of this is actually about safety, it appears that the City is attempting to scramble for authority that more appropriately rests elsewhere. Which leads to the second major problem presented by Farrell’s request. Should manufacturers comply, other sub-state jurisdictions will be empowered to make similar demands.

It is not clear that Farrell’s request has much legal significance, since he cites no binding authority and notes that the state has – rightly – not given him the power to regulate the technology. However, an act of compliance would certainly be given precedential significance in the eyes of other jurisdictions. Sad will be the corporate counsel who is faced with the prospect of explaining to Los Angeles why her firm only intends to comply with San Francisco’s request.

Central to the development of HAV regulation has been the attempt to avoid a “patchwork” approach to oversight. While the danger of a patchwork at the state-level is significant, the danger of different requirements at the sub-state level is even greater. Consider, in California alone, there are 4,435 local governments. Each of those entities has the same legal, though not political, power of San Francisco. Therein lies the risk of compliance with Farrell’s demand. The sheer scale of the compliance demands created by taking such an approach could effectively grind HAV deployment to a halt.

To avoid the problems associated with a patchwork of local government regulations and requests, state legislators in Sacramento and around the nation should make clear the roles of various levels of government and work to prevent undue local discrimination against HAVs.

Several states have already enacted legislation preempting local authorities from meddling with HAV deployment, including Illinois, Nevada, North Carolina, Tennessee and Texas. These red, blue and purple states recognized that the best way to promote automated driving system development – and the resulting safety and mobility benefits – while ensuring effective government oversight, is to reinforce the traditional roles of federal, state and local vehicle and traffic regulators.

Just as it would be senseless for every state to attempt to reproduce NHTSA’s vehicle safety and performance efforts, having California’s 4,435 local governments attempt to reproduce authorities rightly possessed by state regulators runs counter to the goal of efficient, effective government oversight. Innovators would be strangled in red tape and inexperienced local regulators would be overworked on matters outside their areas of expertise. This is a recipe for less meaningful oversight of HAV testing and deployment, not more.


Image credit: Olivier Le Moal

Featured Publications