Insurance and Liability

As I wrote about last time, the Uniform Law Commission recently passed the Uniform Automated Operation of Vehicles Act. Today, I want to focus on Sections 5, 6, and 7 of that Act, which are titled, respectively, “Vehicle Registration,” “Automated-Driving Provider,” and “Associated Automated Vehicle.” The three sections are meant to complement each other and the generally applicable rules regarding motor vehicle registration in a state. The Comments to Section 7 give a nice synopsis of the way these three sections interact:

Existing state law generally requires the registration of a motor vehicle that is operated on a public road. If an automated vehicle qualifies as such a motor vehicle, it too must be registered. The person seeking that registration—typically the vehicle owner—must comply with all conditions of registration under existing law. Section 5 of this act adds a further condition: For the owner of an automated vehicle to register the vehicle, an automated driving provider must have designated that vehicle as an associated automated vehicle. Section 6 specifies how an entity declares that it is an automated driving provider, and Section 7 specifies how that entity then designates its associated automated vehicles. These three sections work together with existing law to ensure that a properly registered automated vehicle has a legal driver when it is under automated operation. In general, only if an automated vehicle is associated with an automated driving provider may it be registered and operated on public roads.

The Act’s comments are fairly dense, but we can work through them section by section. Under current state law, the owner of a motor vehicle must generally register that vehicle with the state according to state registration rules. The Act retains that requirement for the owner of an automated vehicles, but also adds a new condition of registration. Under Section 5, an automated vehicle may be registered only if an entity has:

(1) declared itself to be an automated driving provider (ADP) (explained in Section 6) and

(2) designated that particular automated vehicles as one of its associated automated vehicles (explained in Section 7).

The vehicle owner and the ADP do not necessarily have to be the same legal person. The vehicle owner could be an individual, and the ADP could be an original equipment manufacturer (OEM) like Ford, Honda, or Tesla. The manufacturer, or some other entity like an insurer or fleet operator, would declare themselves to be the ADP to the state, and declare the automated vehicles to be one of its associated vehicles, but the individual would own and register the car. This has the effect of “compelling” vehicle manufacturers, or some other entity, to declare themselves to be the entity legally defined as the driver for any consequences that arise from the vehicles actions on public roads. The comments to Section 6 clarify:

To become an automated driving provider, an entity must make an affirmative declaration that includes specific representations. This means that, first, an entity does not become an automated driving provider against its will and, second, not every entity can become an automated driving provider. Subsection (a) identifies three basic qualifications, at least one of which a provider must satisfy, and subsection (c) identifies four key requirements, all of which the provider must satisfy.

To qualify as an ADP, an entity must have either participated substantially in the development of the system, submitted safety self-assessments with NHTSA, or be a registered manufacturer with NHTSA. The purpose of these sections is to require registration with the state, ensuring that every automated vehicle on a state’s roads has an entity associated with it, against whom the state can credibly enforce relevant provisions of the state vehicle code.

Manufacturers are not required to register as an ADP. But they will be incentivized to declare themselves as ADP’s for the simple reason that if they do not, their customers will be unable to register or use their vehicles in a state that has adopted the Act. If customers in one state were unable to register Ford vehicles but could register Honda vehicles, then everyone in that state would buy Honda automated vehicles and nobody would buy Ford. Under Section 7, once an ADP has designated an associated automated vehicle, the association remains until the ADP is not recognized by the state agency, ceases to exist under principles of corporate law, or affirmatively withdraws the designation.

This approach is a great way to allow manufacturers of automated vehicles to select the states in which they wish to be responsible for their vehicles. If they register as an ADP in Arizona, but not New Mexico, then their customers will be able to register and drive their vehicles on the public roads in Arizona, but not New Mexico. This can allow manufacturers to choose where they accept liability for the automated features of their vehicles.

However, this could cause problems. Assuming uniform adoption of the Act (which is unlikely), if manufacturers are selective with the states where they register as ADP’s then there could be adjacent states where a manufacturer is an ADP in state X, and not in state Y. If customers in state X drive their automated vehicle across the border into state Y, there could be legal questions if the manufacturer is liable for accidents that occur in state Y, especially if they specifically chose not to register there. This could lead to geofencing at state borders, requirements that shift control back to the human driver as they cross state borders, or a whole host of other potential solutions. These solutions could also cause problems. What if a driver is asleep as the vehicle crosses into a state where that manufacturer has not registered? What if the driver overrides and continues allowing the vehicle to drive? Has liability shifted from the manufacturer to the owner given the owner’s conscious choice?

Questions of tort liability, jurisdiction over manufacturers, and technological work-arounds could abound if OEM’s are selective with their registration as ADPs. But they should be allowed to select where they want to sell their automated vehicles if they will be required to legally be identified as the responsible entity. Sections 5, 6, and 7 of the Automated Vehicle Act will likely cause much debate in states that consider adopting the Act.

Earlier this week, Raphaël wrote about the role for no-fault insurance in an age of automated vehicles. The post raised several important questions about the future of the auto insurance industry as technology advances:

Who do we want to protect? Passengers, for sure. But drivers? There is no driver! Or rather, there are many drivers. To some extent, at least under a layman’s understanding of the term “driver,” all the actors along the supply chain are driving the AV. Or, to be more precise, it is difficult to pinpoint a single driver: the “operator”? The software designer? And that is already assuming that there is a single entity who designed the software or operates what may be a fleet of AVs. And there may be others, as the AV industry continues to evolve

While Raphaël’s questions will be extremely important for the industry going forward, I want to pose a slightly different question here: Should drivers of Level 4 or Level 5 automated vehicles be required to purchase insurance at all?

Today, every US state except New Hampshire and Virginia require vehicle owners to purchase auto insurance. In a world where human drivers are in control of the vehicle, this makes sense. While flawed product designs or other manufacturer errors may contribute to car accidents today, many if not most accidents also involve some element of operator error.

In a world of fully autonomous vehicles however, this will no longer be the case. The human sitting in what is today the driver’s seat will not be in control of the vehicle in any meaningful way. Every movement such a vehicle makes will have been designed and programmed by the manufacturer or some company along its supply chain.

Given the extent of manufacturer control, would it be reasonable to have accident payouts determined by the law of products liability? Or perhaps, as Michigan Law professor Kyle Logue argued in a recent article, liability should be apportioned through a scheme of strict enterprise liability, in which automakers would be “unconditionally responsible for the economic losses resulting from any crashes of their vehicles.”

Even with the heightened manufacturer liability that may accompany highly autonomous vehicles, some room may be left for owner error, and thus for traditional insurance. Owners may be held responsible for routine maintenance of their vehicles, for ensuring software is upgraded in a timely fashion, or for actions taken if the vehicle is ever taken out of autonomous mode. But even if some minimal level of owner responsibility leads lawmakers to reject Professor Logue’s theory of enterprise liability in favor of maintenance of insurance requirements for owner/operators, insurance’s role will surely be different than it is today.

Which brings me back to a version of Raphaël’s questions that I started with. Who is the true operator of an AV, and what are they responsible for? Answering these background questions will go a long way towards determining the future of auto insurance in the era of automated vehicles.

In a recent article published on Reuters Regulatory Intelligence, a DC-area lawyer said the following regarding the potential of implementing no-fault insurance “to” automated vehicles:

“Drivers have an inherent incentive to drive safely, so as not to be injured or killed on the roadways. That inherent incentive is what mitigates the “moral hazard” of a no-fault system. But in a no-fault model for autonomous vehicles, the incentives toward safety would be degraded given that manufacturers do not suffer the physical consequences of unsafe operation, as do drivers.”

Intuitively, this seems right. Yet, I thought: is there more to it? What does a world with an AV no-fault insurance scheme would look like?

This might be puzzling at first: what does one mean with no-fault AV insurance? In a more standard setting, a no-fault insurance system means that one gets the benefits of their own insurance without regard to the actual “fault” (such as negligence) and that civil suits on the basis of one’s fault are banned or severely restricted. No-fault systems are straightforward and predictable, although potentially less “just,” to the extent that a negligent driver may get away with nothing more than a deductible to pay or eventually a higher premium.

There was, and there still is, a good policy ground behind no-fault systems around car accidents: avoiding the social cost of civil litigation, and shifting the financial cost of such litigation towards the insurers, between whom things are more often than not settled out of court. Those we want to protect with no-fault insurance schemes are drivers (and passengers), and that is a majority of the population.

Now let’s consider AVs. Who do we want to protect? Passengers, for sure. But drivers? There is no driver! Or rather, there are many drivers. To some extent, at least under a layman’s understanding of the term “driver,” all the actors along the supply chain are driving the AV. Or, to be more precise, it is difficult to pinpoint a single driver: the “operator”? The software designer? And that is already assuming that there is a single entity who designed the software or operates what may be a fleet of AVs. And there may be others, as the AV industry continues to evolve; we can already see that various paths are taken by industry players, some acting for a form of vertical integration, others relying on a variety of suppliers, in a less streamlined way.

Do these all these industry players deserve extra protection? They are all corporate entities after all, and as the lawyer mentions in the above article, none of them are subject to physical injury in case of an accident. While expensive litigation can drive corporations to the ground, the case for shifting costs to insurers, when it comes to AV drivers, appears less clear. What is clear, though, is that human victims of an accident involving an AV ought to be as protected as if the accident did not involve an AV, and maybe even more.

The final answer will come from lawmakers. Moreover, one should not forget that no-fault insurance is mandatory only in a minority of US states, despite being prevalent in the rest of the world. Yet I believe there might be a case here to adopt a legal scheme which would both guarantee a litigation-free recourse to human accident victims, potentially in the form of an industry-funded guarantee fund, while giving the opportunity to the various players along the supply chain to fight it out, in court if need be, on the basis of the reality of their involvement in the cause of the accident; they are all sophisticated players after all and all share in the benefits of the risk they create. The stories of human victims, though, are what may “kill” the industry, if not enough care is taken to ensure a high level of legal protection.

A write-up of the afternoon sessions is now available here!

March 15, 2019 – 10:00 AM – 5:30 PM

Room 1225, Jeffries Hall, University of Michigan Law School 

In the case of automated driving, how and to whom should the rules of the road apply? This deep-dive conference brings together experts from government, industry, civil society, and academia to answer these questions through focused and robust discussion.

To ensure that discussions are accessible to all participants, the day will begin with an introduction to the legal and technical aspects of automated driving. It will then continue with a more general discussion of what it means to follow the law. After a lunch keynote by Rep. Debbie Dingell, expert panels will consider how traffic law should apply to automated driving and the legal person (if any) who should be responsible for traffic law violations. The day will conclude with audience discussion and a reception for all attendees.

(Re)Writing the Rules of the Road is presented by the University of Michigan Law School’s Law and Mobility Program, and co-sponsored by the University of South Carolina School of Law.

Schedule of Events

Morning Sessions 

  • 10:00 am – 10:45 am

Connected and Automated Vehicles – A Technical and Legal Primer

Prof. Bryant Walker Smith

Professor Bryant Walker Smith will provide a technical and legal introduction to automated driving and connected driving with an emphasis on the key concepts, terms, and laws that will be foundational to the afternoon sessions. This session is intended for all participants, including those with complementary expertise and those who are new to automated driving. Questions are welcome. 

  • 10:45 am – 11:15 am
Drivers Licenses for Robots? State DMV Approaches to CAV Regulation

Bernard Soriano, Deputy Director for the Califorina DMV and James Fackler, Assistant Administrator for the Customer Services Administration in the Michigan Secretary of State’s Office, discuss their respective state’s approaches to regulating connected and autonomous vehicles.

  • 11:15 am – 12:00 pm
Just What Is the Law? How Does Legal Theory Apply to Automated Vehicles and Other Autonomous Technologies?

Prof. Scott Hershovitz    

Human drivers regularly violate the rules of the road. What does this say about the meaning of law? Professor Scott Hershovitz introduces legal theory and relates it to automated driving and autonomy more generally.                  

Keynote & Lunch

  • 12:00 pm – 12:30 pm
Lunch

Free for all registered attendees!

  • 12:30 pm-1:30 pm

Keynote – Rep. Debbie Dingell

Rep. Dingell shares her insights from both national and local perspectives.  

Afternoon Sessions

(Chatham House Rule)

  • 1:30 pm – 3:00 pm
Crossing the Double Yellow Line: Should Automated Vehicles Always Follow the Rules of the Road as Written?

Should automated vehicles be designed to strictly follow the rules of the road? How should these vehicles reconcile conflicts between those rules? Are there meaningful differences among exceeding the posted speed limit to keep up with the flow of traffic, crossing a double yellow line to give more room to a bicyclist, and driving through a stop sign at the direction of a police officer? If flexibility and discretion are appropriate, how can they be achieved in law?

A panel of experts will each briefly present their views on these questions, followed by open discussion with other speakers and questions from the audience.

Featured Speakers:

Justice David F. Viviano, Michigan Supreme Court

Emily Frascaroli, Counsel, Ford Motor Company

Jessica Uguccioni, Lead Lawyer, Automated Vehicles Review, Law Commission of England and Wales

  • 3:15 pm – 4:45 pm
Who Gets the Ticket? Who or What is the Legal Driver, and How Should Law Be Enforced Against Them?

Who or what should decide whether an automated vehicle should violate a traffic law? And who or what should be responsible for that violation? Are there meaningful differences among laws about driving behavior, laws about vehicle maintenance, and laws and post-crash responsibilities? How should these laws be enforced? What are the respective roles for local, state, and national authorities?

A panel of experts will each briefly present their views on these questions, followed by open discussion with other speakers and questions from the audience.

Featured Speakers:

Thomas J. Buiteweg, Partner, Hudson Cook, LLP

Kelsey Brunette Fiedler, Ideation Analyst in Mobility Domain

Karlyn D. Stanley, Senior Policy Analyst, RAND Corporation

Daniel Hinkle, State Affairs Counsel, American Association for Justice

  • 4:45 pm – 5:30 pm 
 Summary and General Discussion                                     

Participants and attendees close out the day by taking part in wide discussion of all of the day’s panels.

Last week I covered the various companies who are seeking to use aerial drones to deliver goods to your door. Today, in the third part to my series on delivery (you’ll find Part 1 here, and an even earlier post on delivery, from December of 2018, here), I’m going to look at recent proposals to use automated vehicles to deliver consumer goods.

As an introduction, I’m going to include a paragraph from that December 2018 post as an introduction to some of the ways automated vehicles are being used to make deliveries :

The potential for CAVs as delivery vehicles is already being tested by companies like Domino’s and Kroger, among others. Earlier this year Toyota announced delivery partnerships with Amazon and Pizza Hut, and Waymo’s CEO recently highlighted it as an area of opportunity.  This week the New York Times profiled Nuro, the start-up working with Kroger to test robotic delivery cars in Scottsdale, Ariz. Nuro’s vehicles are designed in-house, and look like “toasters-on-wheels,” and are currently followed everywhere they go by human safety drivers in conventionally driven “shadow car.” When the vehicle stops for a delivery, customers enter a PIN code into a small touch pad to open the compartment containing their order. The current charge for same-day delivery using the system is around $6. Ford has also flagged the delivery market as an area they’d like to explore, citing projections that by 2026 the last-mile delivery market for CAVs will hit $130 billion.

Don’t Forget to Tip Your (Robotic) Delivery Driver – Dec. 21, 2018

Since that post, Domino’s has announced a partnership with Nuro as well, with plans to test in Houston at some point this year. Walmart has also jumped in on the action – partnering with another AV developer, Gatik. For now Walmart’s test is limited to a 2-mile route between two of their stores in the company’s hometown of Bentonville, Arkansas. Why the interest? In part because of the potential cost savings – a recent Ford estimate calculates AVs could reduce the cost per mile for deliveries from $2.50 to $1. No doubt the combination of lower costs and ever-greater demand for delivery is a powerful motivator, pushing companies to explore not only AVs, but also drones and delivery bots, as discussed in Parts I and II of this series.

Beyond last-mile deliveries, there is a great deal of interest in automating semi-trucks and other large delivery vehicles. One company, TuSimple, is working with both the U.S. Postal Service (USPS) and UPS to move packages between cities. Interestingly, in UPS’ case, the company only announced the partnership after TuSimple had already been delivering goods for months – which seems to indicate the program is not just a grab for positive PR. The USPS’ test was more limited, running for two-weeks and five round trips. All of the trips included a safety driver and an engineer, and both tests were carried out in the Southwest. Meanwhile, in Sweden, a completely driverless electric truck was deployed in May, a global first. Given a nation-wide shortage of truck drivers (a recent estimate puts the U.S. deficit at roughly 60,000 drivers), automated trucks present a solution that doesn’t overly disrupt a truck-heavy commercial delivery system.

But what would the wide-spread adoption of AVs as part of the delivery ecosystem mean? We can already see that the demand for faster and faster delivery is taking its toll. Recently, the NY Times and Buzzfeed News both published articles detailing the human cost of Amazon’s push for same or next-day delivery. Under-trained drivers pushed to the limit have killed people in seemingly avoidable accidents that don’t often happen with more highly-trained delivery drivers (like those used by the USPS, UPS, and FedEx). Amazon has avoided liability by using a number of third-party companies as contractors, making those companies, and not Amazon, responsible for accidents. AVs would certainly be safer for the public, as they wouldn’t fall prey to the pressures of human drivers, though that does nothing to alleviate the pressures on the human delivery people, who would still be needed to move goods from the vehicle to a door. At the same time, Amazon may continue to escape liability, if the AVs remain owned by third parties. There is also the greater question of the environmental impact of the growing number of delivery vehicles on the road (not to mention the waste created by packing materials and shipping boxes). I’ll leave a greater discussion about those issues to future posts and other forums, but those questions, among so many others (privacy, cybersecurity, and traffic management among them) are important to consider as automated delivery vehicles of all kinds begin to fill our streets and skies.

P.S. – In a follow up to last week’s blog, the USPS has stated to investigate the use of aerial drones, and is now seeking information from drone operators and developers.

Cite as: Raphael Beauregard-Lacroix, (Re)Writing the Rules of The Road: Reflections from the Journal of Law and Mobility’s 2019 Conference, 2019 J. L. & Mob. 97.

On March 15th, 2019, the Journal of Law and Mobility, part of the University of Michigan’s Law and Mobility Program, presented its inaugural conference, entitled “(Re)Writing the Rules of The Road.” The conference was focused on issues surrounding the relationship between automated vehicles (“AVs”) and the law. In the afternoon, two panels of experts from academia, government, industry, and civil society were brought together to discuss how traffic laws should apply to automated driving and the legal person (if any) who should be responsible for traffic law violations. The afternoon’s events occurred under a modified version of the Chatham House Rule, to allow the participants to speak more freely. In the interest of allowing those who did not attend to still benefit from the day’s discussion, the following document was prepared. This document is a summary of the two panels, and an effort has been made to de-identify the speaker while retaining the information conveyed. 

Panel I: Crossing the Double Yellow Line: Should Automated Vehicles Always Follow the Rules of the Road as Written?

The first panel focused on whether automated vehicles should be designed to strictly follow the rules of the road. Questions included – How should these vehicles reconcile conflicts between those rules? Are there meaningful differences between acts such as exceeding the posted speed limit to keep up with the flow of traffic, crossing a double yellow line to give more room to a bicyclist, or driving through a stop sign at the direction of a police officer? If flexibility and discretion are appropriate, how can this be reflected in law? 

Within the panel, there was an overall agreement that we need both flexibility in making the law, and flexibility in the law itself among the participants. It was agreed that rigidity, both on the side of the technology as well as on the side of norms, would not serve AVs well. The debate was focused over just how much flexibility there should be and how this flexibility can be formulated in the law.

One type of flexibility that already exists is legal standards. One participant emphasized that the law is not the monolith it may seem from the outside – following a single rule, like not crossing a double yellow line, is not the end of an individual’s interaction with the law. There are a host of different laws applying to different situations, and many of these laws are formulated as standards – for example, the standard that a person operating a vehicle drives with “due care and attention.” Such an approach to the law may change the reasoning of a judge when it would come to determining liability for an accident involving an AV. 

When we ask if AVs should always follow the law, our intuitive reaction is of course they should. Yet, some reflection may allow one to conclude that such strict programming might not be realistic. After all, human drivers routinely break the law. Moreover, most of the participants explicitly agreed that as humans, we get to choose to break the law, sometimes in a reasonable way, and we get to benefit from the discretion of law enforcement. 

That, however, does not necessarily translate to the world of AVs, where engineers make decisions about code and where enforcement can be automatized to a high degree, both ex ante and ex post. Moreover, such flexibilities in the law needs to be tailored to the specific social need; speeding is a “freedom” we enjoy with our own, personal legacy cars, and this type of law breaking does not fulfill the same social function as a driver being allowed to get on the sidewalk in order to avoid an accident. 

One participant suggested that in order to reduce frustrating interactions with AVs, and to overall foster greater safety, AVs need the flexibility not to follow the letter of the law in some situations. Looking to the specific example of the shuttles running on the University of Michigan’s North Campus – those vehicles are very strict in their compliance with the law. 1 1. Susan Carney, Mcity Driverless Shuttle launches on U-M’s North Campus, The Michigan Engineer (June 4, 2018), https://news.engin.umich.edu/2018/06/mcity-driverless-shuttle-launches-on-u-ms-north-campus/. × They travel slowly, to the extent that their behavior can annoy human drivers. When similar shuttles from the French company Navya were deployed in Las Vegas, 2 2. Paul Comfort, U.S. cities building on Las Vegas’ success with autonomous buses, Axios (Sept. 14, 2018), https://www.axios.com/us-cities-building-on-las-vegas-success-with-autonomous-buses-ce6b3d43-c5a3-4b39-a47b-2abde77eec4c.html. × there was an accident on the very first run. 3 3. Sean O’Kane, Self-driving shuttle crashed in Las Vegas because manual controls were locked away, The Verge (July 11, 2019, 5:32 PM), https://www.theverge.com/2019/7/11/20690793/self-driving-shuttle-crash-las-vegas-manual-controls-locked-away. × A car backed into the shuttle, and when a normal driver would have gotten out of the way, the shuttle did not.

One answer is that we will know it when we see it; or that solutions will emerge out of usage. However, many industry players do not favor such a risk-taking strategy. Indeed, it was argued that smaller players in the AV industry would not be able to keep up if those with deeper pockets decide to go the risky way. 

Another approach to the question is to ask what kind of goals should we be applying to AVs? A strict abidance to legal rules or mitigating harm? Maximizing safety? There are indications of some form of international consensus 4 4. UN resolution paves way for mass use of driverless cars, UN News (Oct. 10, 2018), https://news.un.org/en/story/2018/10/1022812. × (namely in the form of a UN Resolution) 5 5. UN Economic Commission for Europe, Revised draft resolution on the deployment of highly and fully automated vehicles in road traffic (July, 12, 2018), https://www.unece.org/fileadmin/DAM/trans/doc/2018/wp1/ECE-TRANS-WP.1-2018-4-Rev_2e.pdf × that the goal should not be strict abidance to the law, and that other road users may commit errors, which would then put the AV into a situation of deciding between strict legality and safety or harm. 

In Singapore, the government recently published “Technical Reference 68,” 6 6. Joint Media Release, Land Transport Authority, Enterprise Singapore, Standards Development Organization, & Singapore Standards Council, Singapore Develops Provisional National Standards to Guide Development of Fully Autonomous Vehicles (Jan. 31, 2019), https://www.lta.gov.sg/apps/news/page.aspx?c=2&id=8ea02b69-4505-45ff-8dca-7b094a7954f9. × which sets up a hierarchy of rules, such as safety, traffic flow, and with the general principle of minimizing rule breaking. This example shows that principles can act as a sense-check. That being said, the technical question of how to “code” the flexibility of a standard into AV software was not entirely answered. 

Some participants also reminded the audience that human drivers do not have to “declare their intentions” before breaking the law, while AV software developers would have to. Should they be punished for that in advance? Moreover, non-compliance with the law – such as municipal ordinances on parking – is the daily routine for certain business models such as those who rely on delivery. Yet, there is no widespread condemnation of that, and most of us enjoy having consumer goods delivered at home.

More generally, as one participant asked, if a person can reasonably decide to break the law as a driver, does that mean the developer or programmer of AV software can decide to break the law in a similar way and face liability later? Perhaps the answer is to turn the question around – change the law to better reflect the driving environment so AVs don’t have to be programmed to break it. 

Beyond flexibility, participants discussed how having multiple motor vehicle codes – in effect one per US State – makes toeing the line of the law difficult. One participant highlighted that having the software of an AV validated by one state is big enough a hurdle, and that more than a handful of such validations processes would be completely unreasonable for an AV developer. Having a single standard was identified as a positive step, while some conceded that states also serve the useful purpose of “incubating” various legal formulations and strategies, allowing in due time the federal government to “pick” the best one. 

Panel II: Who Gets the Ticket? Who or What is the Legal Driver, and How Should Law Be Enforced Against Them?

The second panel looked at who or what should decide whether an automated vehicle should violate a traffic law, and who or what should be responsible for that violation. Further questions included – Are there meaningful differences among laws about driving behavior, laws about vehicle maintenance, and laws and post-crash responsibilities? How should these laws be enforced? What are the respective roles for local, state, and national authorities?

The participants discussed several initiatives, both public and private, that aimed at defining, or helping define the notion of driver in the context of AVs. The Uniform Law Commission worked on the “ADP”, or “automated driving provider”, which would replace the human driver as the entity responsible in case of an accident. The latest report from the RAND Corporation highlighted that the ownership model of AVs will be different, as whole fleets will be owned and maintained by OEMs (“original equipment manufacturers”) or other types of businesses and that most likely these fleet operators would be the drivers. 7 7. James M. Anderson, et. al., Rethinking Insurance and Liability in the Transformative Age of Autonomous Vehicles (2018), https://www.rand.org/content/dam/rand/pubs/conf_proceedings/CF300/CF383/RAND_CF383.pdf. ×

Insurance was also identified as a matter to take into consideration in the shaping up of the notion of AV driver. As of the date of the conference, AVs are only insured outside of state-sponsored guarantee funds, which aim to cover policy holders in case of bankruptcy of the insurer. Such “non-admitted” insurance means that most insurers will simply refuse to insure AVs. Who gets to be the driver in the end may have repercussions on whether AVs become insurable or not. 

In addition, certain participants stressed the importance of having legally recognizable persons bear the responsibility – the idea that “software” may be held liable was largely rejected by the audience. There should also be only one such person, not several, if one wants to make it manageable from the perspective of the states’ motor vehicle codes. In addition, from a more purposive perspective, one would want the person liable for the “conduct” of the car to be able to effectuate required changes so to minimize the liability, through technical improvements for example. That being said, such persons will only accept to shoulder liability if costs can be reasonably estimated. It was recognized by participants that humans tend to trust other humans more than machines or software, and are more likely to “forgive” humans for their mistakes, or trust persons who, objectively speaking, should not be trusted.

Another way forward identified by participants is product liability law, whereby AVs would be understood as a consumer good like any other. The question then becomes one of apportionment of liability, which may be rather complex, as the experience of the Navya shuttle crash in Las Vegas has shown. 

Conclusion

The key takeaway from the two panels is that AV technology now stands at a crossroads, with key decisions being taken as we discuss by large industry players, national governments and industry bodies. As these decisions will have an impact down the road, all participants and panelists agreed that the “go fast and break things” approach will not lead to optimal outcomes. Specifically, one line of force that comes out from the two panels is the idea that it is humans who stand behind the technology, humans who take the key decisions, and also humans who will accept or reject commercially-deployed AVs, as passengers and road users. As humans, we live our daily lives, which for most of us include using roads under various capacities, in a densely codified environment. However, this code, unlike computer code, is in part unwritten, flexible and subject to contextualization. Moreover, we sometimes forgive each others’ mistakes. We often think of the technical challenges of AVs in terms of sensors, cameras and machine learning. Yet, the greatest technical challenge of all may be to express all the flexibility of our social and legal rules into unforgivably rigid programming language. 

By Emily Frascaroli, John Isaac Southerland, Elizabeth Davis, and Woods Parker

Cite as: Emily Frascaroli et al., Let’s Be Reasonable: The Consumer Expectations Test is Simply Not Viable to Determine Design Defect for Complex Autonomous Vehicle Technology, 2019 J. L. & Mob. 53.

Abstract

Although highly automated vehicles (“HAVs”) have potential to reduce deaths and injuries from traffic crashes, product liability litigation for design defects in vehicles incorporating autonomous technology is inevitable. During the early stages of implementation, courts and juries will be forced to grapple with the application of traditional product liability principles to a never before experienced category of highly technical products. Recent decisions limiting the use of the consumer expectations test in cases involving complex products prompted the authors to examine more closely the history behind and the future viability of the consumer expectations test in HAV litigation.

I.          Introduction

In 2016, more than 35,000 individuals died in vehicle crashes in the U.S. and the National Highway Traffic Safety Administration (“NHTSA”) estimated that 94% of these deaths were attributable to human error. 8 8. Automated Vehicles for Safety, NHTSA, https://www.nhtsa.gov/technology-inn ovation/automated-vehicles-safety (last visited May 2, 2019). × In 2017 and 2018, in their own self-driving safety reports, General Motors and Waymo also noted that approximately 1.2 million lives are lost worldwide each year due to car crashes. 9 9. Waymo, Waymo Safety Report: On the Road to Fully Self-Driving 3 (2018), https://storage.googleapis.com/sdc-prod/v1/safety-report/Safety%20Report%20 2018.pdf; General Motors, 2018 Self-Driving Safety Report 3 (2018), https://ww w.gm.com/content/dam/company/docs/us/en/gmcom/gmsafetyreport.pdf. × Each of these entities further agree that highly automated vehicle (“HAV”) 10 10. For purposes of this paper, the terms highly automated vehicle (HAV) or “self-driving” will refer to vehicles defined by SAE Levels 4–5. See SAE International, Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles J3016 (2018), https://saemobilus.sae.org/content/j3016 _201806. The SAE levels of automation are as follows: (0) No automation; the vehicle has zero autonomy, and the driver performs all tasks; (1) Driver Assistance: the vehicle is controlled by the driver, but some driver assistance features are included; (2) Partial Driving Automation: the vehicle has combined automated functions, but the driver must remain engaged with the driving task and monitor the environment constantly; (3) Conditional Driving Automation: the driver is necessary, but is not required to constantly monitor the environment—the driver must be ready to take control of the vehicle at all times; (4) High Driving Automation: the vehicle is capable of performing all driving functions under certain conditions, but the driver has the option of controlling the vehicle; and (5) Full Driving Automation: the vehicle is capable of performing all driving functions under all conditions, with the driver having the option of controlling the vehicle. × technology has the potential to reduce or remove human error from the equation. 11 11. See Waymo, supra note 2; General Motors, supra note 2. × Additional potential benefits include reduced traffic congestion; increased mobility options for currently underserved populations; and, increased comfort and a reduction in lost time during vehicle operation. 12 12. General Motors, supra note 2. × Put simply, the stage is set for HAV technology to revolutionize the mobile world.

During the implementation of HAVs, most sources agree that, due to their highly complex and technical nature, consumer education about the products will be key to successful and effective implementation. For its part, in the 2017 update, Automated Driving Systems 2.0, NHTSA stated that “[E]ducation and training is imperative for increased safety during the deployment of [HAVs],” and encourages the development of “consumer education and training programs to address the anticipated differences in the use and operation of [automated driving systems] from those of the conventional vehicles that the public owns and operates.” 13 13. NHTSA, Automated Driving Systems: A Vision for Safety 2.0 15 (2017), https://www.nhtsa.gov/vehicle-manufacturers/automated-driving-systems#automated-driving-systems-av-20. × General Motors and Waymo echoed these sentiments in their respective self-driving safety reports with Waymo, in October 2017, even helping to launch – Let’s Talk Self-Driving – which it describes as “the world’s first public education campaign about fully self-driving vehicles.” 14 14. Waymo, supra note 2 at 30; General Motors, supra note 2, at 32. × Taking this one-step further, in 2018, Ford provided its Voluntary Safety Self-Assessment Report – A Matter of Trust. 15 15. Ford Motor Company, A Matter of Trust: Ford’s Approach to Developing Self-DrivingVehicles, https://media.ford.com/content/dam/fordmedia/pdf/ Ford_AV_LLC_FINAL_HR_2.pdf. × In it, Ford makes clear “that the central challenge in the development of self-driving vehicles” is not the technology, but, instead, it is consumer trust in the “safety, reliability and experience that the technology will enable.” 16 16. Id.at 3. × Ford reiterates this point stating about consumer education and training that, “[B]ringing self-driving vehicles to market will require a thoughtful and sustained effort to teach customers how they work, why they’re safe and how to use them.” 17 17. Id. at 42 (emphasis added). ×

In light of this, questions remain, particularly with respect to liability, if, and when, an injury or death occurs in an HAV. The question of who is liable when a self-driving vehicle crashes has generated significant debate and conversation. Per NHTSA, “these are among many important questions beyond the technical considerations that policymakers are working to address before automated vehicles are made available.” 18 18. Automated Vehicles for Safety, supra note 1. × NHTSA also posits that questions of liability pertaining to HAVs are something within the purview of each state to manage. 19 19. NHTSA, supra note 6, at 24. × In the wake of some interesting opinions in 2017, this question, and others, prompted the authors to examine the historical development of product defect theories and, in particular, whether the consumer expectations test can reasonably be applied to determine liability in cases involving complex products.

II.         Adoption of Design Defect Tests in the Wake of Section 402A of the Restatement (Second) of Torts.

In 1965, the law of torts and the field of product liability were altered dramatically by the adoption of Section 402A of the Restatement (Second) of Torts. 20 20. See generally George L. Priest, Strict Products Liability: The Original Intent, 10 Cardozo L. Rev. 2301, 2301 (1989). × Section 402A sought to impose strict liability on the manufacturers or sellers of defective products, regardless of negligence, and became perhaps the most cited section of any Restatement of Law in legal jurisprudence. 21 21. See James A. Henderson Jr. & Aaron D. Twerski, Proposed Revision of Section 402A of the Restatement (Second) of Torts, 77 Cornell L. Rev. 1512, 1512 n.1 (1992). ×

A.         The Consumer Expectations Test

Section 402A provides that “[o]ne who sells any product in a defective condition unreasonably dangerous to the user or consumer or to his property is subject to liability for physical harm thereby caused to the ultimate user or consumer, or to his property . . . .” 22 22. Restatement (Second) of Torts § 402A (Am. Law Inst. 1965). × To guide courts in determining whether a product is unreasonably dangerous, the drafters of the Second Restatement included the following comment: “The article sold must be dangerous to an extent beyond that which would be contemplated by the ordinary consumer who purchases it, with the ordinary knowledge common to the community as to its characteristics.” 23 23. Id. at cmt. i. × This comment provided support for the pure consumer expectations test in product defect cases. In turn, this product defect test was embraced by courts in the years following the release of the Second Restatement. 24 24. See, e.g., Aller v. Rodgers Machinery Mfg. Co., Inc., 268 N.W. 2d 830 (Iowa 1978); Phipps v. General Motors Corp., A.2d 955 (Md. 1976); Estate of Pinkham v. Cargill, Inc., 55 A.3d 1 (Me. 2012) (citing Adams v. Buffalo Forge Co., 443 A.2d 932, 940 (Me. 1982)); Simonetta v. Viad Corp., 197 P.3d 127 (Wash. 2008). × Over time, courts across the country recognized that there were significant issues with the Second Restatement’s pure consumer expectations approach to defective design.

For example, in the 1967 case of Heaton v. Ford Motor Co., the Supreme Court of Oregon was faced with application of the consumer expectations test in the context of a design defect claim involving a motor vehicle. 25 25. See Heaton v. Ford Motor Co., 435 P.2d 806 (Or. 1967). × In Heaton, the plaintiff’s vehicle struck a rock in the roadway. After the accident, the rim of the wheel was found to have separated from the rest of the wheel assembly. The court utilized the consumer expectations test to determine design defect, stating:

In the type of case in which there is no evidence, direct or circumstantial, available to prove exactly what sort of manufacturing flaw existed, or exactly how the design was deficient, the plaintiff may nonetheless be able to establish his right to recover, by proving that the product did not perform in keeping with the reasonable expectations of the user. When it is shown that a product failed to meet the reasonable expectations of the user the inference is that there was some sort of defect. 26 26. Id. at 471–72 (emphasis added). ×

However, the court recognized that in Heaton, the jury could not possibly state from their own experience what the expectations of the average consumer would be. 27 27. See id. at 472–73. × After all, high-speed collisions with large rocks are not so common that the average person would know from personal experience how the wheel assembly would perform in such a situation. 28 28. Id. at 473. × As such, “[t]he jury would therefore be unequipped, either by general background or by facts supplied in the record, to decide whether this wheel failed to perform as safely as an ordinary consumer would have expected.” 29 29. Id. × Unfortunately, the Heaton court ultimately refused to acknowledge that the consumer expectations test simply did not apply in this situation, but instead seemed to suggest that expert testimony would be required to establish the consumer expectations. 30 30. See id. at 474. × The paradox is obvious: if an expert is required to tell the consumer what to expect, is that truly the expectation of an ordinary consumer?

Fortunately, courts have begun to recognize that utilizing the consumer expectations test in cases involving alleged design defects in technically complex products is simply not workable. 31 31. See, e.g., Montag v. Honda Motor Co., Inc., 75 F.3d 1414 (10th Cir. 1996) (citing Camacho v. Honda Motor Corp., 741 P.2d 1240, 1246–48 (Colo. 1987)). See also 2 Louis R. Frumer & Melvin I. Friedman, Products Liability§ 11.03 (Matthew Bender, Rev. Ed.). × However, there are courts that have found the consumer expectations test applicable, even where the requisite knowledge is not within the purview of lay jurors. 32 32. See, e.g., Bresnahan v. Chrysler Corp., 38 Cal. Rptr. 2d 446, 451–52 (Cal. Ct. App. 1995). See also 2 Frumer & Friedman, supra note 25. ×

B.         Risk Utility Test

As a result, many courts began to apply the test commonly referred to as risk-utility balancing. Under this test, to establish a prima facie case of design defect, the plaintiff must show that on balance, the utility of the challenged product design outweighs the risk of danger inherent in the design. 33 33. See, e.g., Thibault v. Sears, Roebuck & Co., 395 A.2d 843 (N.H. 1978). × Traditionally, under risk-utility, courts consider a multitude of factors to determine whether a defect exists, including the following factors identified in an influential article by Dean John W. Wade in 1973:

    1. The usefulness and desirability of the product;
    2. the safety aspects of the product;
    3. the availability of safer substitute products;
    4. the possibility of elimination of dangerous characteristics of the product without impairing its usefulness;
    5. the user’s ability to avoid danger by safe use of the product;
    6. the anticipated dangers inherent in the product due to general knowledge or the existence of warnings; and
    7. the possibility of loss-spreading by the manufacturer through price setting or insurance. 34 34. See John W. Wade, On the Nature of Strict Tort Liability for Products, 44 Miss. L.J. 825 (1973). ×

Further, “[t]he utility of the product must be evaluated from the point of view of the public as a whole, because a finding of liability for defective design could result in the removal of an entire product line from the market.” 35 35. See Thibault, 395 A.2d at 807. ×

In 1998, the element of a reasonable alternate design was written into the new Restatement (Third) of Torts. 36 36. See Restatement (Third) of Torts § 2 (Am. Law Inst., 1998). × Under § 2of the Third Restatement, a product is:

 “[D]efective in design when the foreseeable risks of harm posed by the product could have been reduced or avoided by the adoption of a reasonable alternative design by the seller or other distributor, or a predecessor in the commercial chain of distribution, and the omission of the alternative design renders the product not reasonably safe. 37 37. Id. The Third Restatement explicitly rejects consumer expectations as an independent standard for determining design defect. See id. at §2 cmt. g. ×

As noted in 2009, the “reasonable alternative design” standard of the Third Restatement ultimately came to embody the “risk-utility test” that is applied in the majority of United States jurisdictions today. 38 38. See Aaron D. Twerski & James A. Henderson, Jr., Manufacturer Liability for Defective Product Designs: The Triumph of Risk Utility, 74 Brook. L. Rev. 1061, 1065 (2009). Notably, even courts that continue to utilize the consumer expectations test exclusively often acknowledge that evidence of an alternative design is the most appropriate and useful means of showing that a product is unreasonably dangerous. See, e.g., Ford Motor Co. v. Trejo, 402 P.3d 649, 655 (Nev. 2017). ×

Significant developments since 2009, some of which are discussed more fully below, further exemplify the national trend towards applying risk-utility in complex design defect cases and moving away from the consumer expectations test in this context. Indeed, in 2017, the Ninth Circuit recognized that, “when the ultimate issue of design defect calls for a careful assessment of feasibility, practicality, risk, and benefit, the case should not be resolved simply on the basis of ordinary consumer expectations.” 39 39. See Edwards v. Ford Motor Co., 683 Fed. App’x 610, 611 (9th Cir. 2017) (quoting Soule v. General Motors Corp., 882 P.2d 298, 305 (Cal. 1994)) (emphasis supplied in original). ×

C.         Hybrid Test

Other jurisdictions utilize a dual-approach to design defect claims. California, for example, utilizes the consumer expectations test when consumers are capable of developing expectations about the characteristics of a product from everyday use. 40 40. See Soule, 882 P.2d at 310–311. × For more complex products, where the characteristics are outside the knowledge of an everyday consumer, courts apply the risk-benefit test. 41 41. See id. × Thus, the determinative issue in many cases in California and similar jurisdictions is whether a product is too complex or unfamiliar for average consumers to develop expectations, such that utilization of the consumer expectations test is improper. 42 42. See, e.g., Saller v. Crown Cork & Seal Co., Inc., 115 Cal. Rptr. 3d 151, 160–61 (Cal. Ct. App. 2010). × Making this determination in the context of autonomous technology should not be an issue.

This hybrid approach combines elements of both the consumer expectations test and the risk-utility test. One example is the “either-or” concept, which posits that:

[A] product is defective in design either (1) if the product has failed to perform as safely as an ordinary consumer would expect when used in an intended or reasonably foreseeable manner, or (2) if, in light of the relevant factors . . . the benefits of the challenged design do not outweigh the risk of danger inherent in the design. 43 43. Barker v. Lull Eng’g Co., 573 P.2d 443, 446 (Cal. 1978). ×

This approach allows courts more flexibility in applying the appropriate test based upon all of the relevant circumstances. For example, in Barker v. Lull Engineering, the plaintiff sustained injuries while operating a loader at a construction site and alleged that his injuries were caused by a defective design of the product because it was not equipped with a roll bar or seat belts. 44 44. Id. at 447–48. × The California Supreme Court rejected a pure consumer expectations test and a pure risk-utility test, instead articulating the two-prong test allowing a plaintiff to establish a design defect through either test. 45 45. Id. at 455–56. × In so holding, the court noted the benefits of the more flexible approach stating:

[I]t subjects a manufacturer to liability whenever there is something “wrong” with a product’s design – either because the product fails to meet ordinary consumer expectations as to safety or because, on balance, the design is not as safe as it should be – while stopping short of making the manufacturer an insurer for all injuries which may result from the use of its product. This test, moreover, explicitly focuses on the trier of fact’s attention to the adequacy of the product itself, rather than on the manufacturer’s conduct, and places the burden on the manufacturer, rather than the plaintiff, to establish that because of the complexity of, and trade-offs implicit in, the design process, an injury-producing product should nevertheless not be found defective. 46 46. Id. at 456. ×

Other courts have taken a different approach in formulating a hybrid consumer expectations and risk-utility test, incorporating risk-utility factors into the consumer expectation analysis, and vice versa. 47 47. See, e.g.,Potter v. Chicago Pneumatic Tool Co., 694 A.2d 1319, 1333–34 (Conn. 1997). × For example, in Potter v. Chicago Pneumatic Tool Co., the plaintiffs alleged that pneumatic hand tools manufactured by the defendant were defective in design because they exposed the plaintiffs to excessive vibration, resulting in injuries to the plaintiffs. 48 48. Id. at 1325. × Although Connecticut courts had long applied the Second Restatement’s consumer expectations test, the court recognized that “there may be instances involving complex product designs in which an ordinary consumer may not be able to form expectations of safety.” 49 49. Id. at 1333. × In recognizing this issue with the pure consumer expectations test, the Connecticut Supreme Court adopted a “modified consumer expectation test, provid[ing] the jury with the product’s risks and utility and then inquir[ing] whether a reasonable consumer would consider the product unreasonably dangerous.” 50 50. Id. × In determining a consumer’s reasonable expectations, the jury should consider various factors, including but not limited to the “relative cost of the product, the gravity of the potential harm from the claimed defect and the cost and feasibility of eliminating or minimizing the risk.” 51 51. Id. × In sum, under this approach, “the consumer expectation test would establish the product’s risks and utility, and the inquiry would then be whether a reasonable consumer would consider the product design unreasonably dangerous.” 52 52. Id. × The Connecticut Supreme Court’s approach was based, in part, on early drafts of the Restatement (Third) of Torts on Product Liability. 53 53. Id. at 1331. ×

III.       More and More Courts Are Recognizing the Limitations of the Consumer Expectations Test in Complex Design Defect Cases.

In March 2017, the United States Court of Appeals for the Ninth Circuit examined the question of whether the consumer expectations test or risk-utility balancing test should be applied to determine whether a design defect existed in a product liability case involving vehicle roof deformation. 54 54. See Edwards v. Ford Motor Co., 683 Fed. App’x 610 (9th Cir. 2017). × In Edwards v. Ford Motor Co., the plaintiffs claimed that the roof of their vehicle was defectively designed because it deformed inward eight inches into the passenger compartment during a multiple rollover event. The plaintiffs alleged that a properly designed roof should have resulted in less than three inches of deformation in the subject crash.

The Edwards plaintiffs sought to prove design defect by showing that the roof did not perform as the average consumer would have expected. Ford filed a motion contending that the jury should be instructed on the risk-utility test alone. Ford’s motion was granted and plaintiffs appealed. The Ninth Circuit held that the risk-utility test was the proper test to be applied, specifically recognizing the “lack of consumer expectations regarding the extent to which the [vehicle]’s roof would crush in a multiple rollover accident.” 55 55. Id. at 611. × The Ninth Circuit went on to note that “[d]rivers’ everyday experiences do not allow for the formulation of reasonable expectations as to the degree that a vehicle’s roof should crush during a rollover.” 56 56. Id. × The Ninth Circuit did not feel it necessary to state whether or not this product was too complex for the consumer expectations test to govern; instead, it was enough to know that consumers simply would not have expectations related to roof performance in a rollover. 57 57. See id.; See also Branham v. Ford Motor Co., 701 S.E. 2d 5, 13–14 (S.C. 2010) (finding that the consumer expectations test was not appropriate in design defect case after examining the issue in the context of an automotive rollover case involving an allegedly defective seatbelt design). × Thus, the risk-utility test was the appropriate test. 58 58. See id. ×

Another recent case decided by the Court of Appeal of California also limited the applicability of the consumer expectations test. The plaintiff in Trejo v. Johnson & Johnson contracted a rare condition known as SJS/TEN as a reaction to taking over-the-counter ibuprofen produced by Johnson & Johnson. Plaintiff sought to show that the drug was defectively designed through utilization of the consumer expectations test.

The Court of Appeal found the consumer expectations had no place in proving design defect under these facts, noting that “‘[t]he consumer expectations test is reserved for cases in which the everyday experience of the product’s users permits a conclusion that the product’s design violated minimum safety assumptions, and is thus defective regardless of expert opinion about the merits of the design.’” 59 59. Trejo v. Johnson & Johnson, 220 Cal. Rptr. 3d 127, 165 (Cal. Ct. App. 2017) (quoting Soule v. General Motors Corp., 882 P.2d 298, 308 (Cal. 1994). × The plaintiff essentially attempted to use consumer expectations to avoid having to confront the more difficult risk-utility standard or any showing of a reasonable alternative design, but also wished to introduce expert testimony to establish that the ibuprofen did not meet consumer expectations. The court found this fact alone sufficient to demonstrate the consumer expectations test was inappropriate for that case. 60 60. See id. at 168. ×

Succinctly explaining the problem with applying consumer expectations in the case of complex products or products with which consumers are unfamiliar, the court stated: “[I]t could be said that any injury from the intended or foreseeable use of a product is not expected by the ordinary consumer. If this were the end of the inquiry, the consumer expectations test always would apply and every product would be found to have a design defect.” 61 61. See id. at 167 (emphasis added). ×

As to a non-complex product, the Tenth Circuit’s examination of consumer expectations in Kokins v. Teleflex, Inc. is instructive. 62 62. See Kokins v. Teleflex, Inc., 621 F.3d 1290 (10th Cir. 2010). × Kokins involved the determination of what design defect test should be used under Colorado law in the context of a claim involving a metal marine cable, a seemingly simple product. The court initially noted that, under Colorado law, the risk-utility test and consumer expectations test are not mutually exclusive of each other and can sometimes even be applied in the same case. 63 63. See id. at 1297. × However, the Tenth Circuit held that in the context of this particular product, only the risk-utility test was proper, due to the technical and specific information related to metallic corrosion. 64 64. See id. × Quite simply, in cases where technical and scientific issues predominate, use of the consumer expectations test, alone or in conjunction with the risk-utility test, is inappropriate. 65 65. See id. ×

Finally, as recently as November, 2017, the Colorado Supreme Court determined that the “risk-benefit test is the appropriate test to assess whether a product was unreasonably dangerous due to a design defect when . . . the dangerousness of the design is ‘defined primarily by technical, scientific information.’” 66 66. See Walker v. Ford Motor Co., 406 P.3d 845, 850 (Colo. 2017) (quoting Ortho Paharm. Corp. v. Heath, 722 P.2d 410, 414 (Colo. 1986)). × In Walker v. Ford Motor Co., the plaintiff proceeded to trial against Ford for injuries sustained in a rear-end impact. 67 67. See id. at 847­–48. × The plaintiff alleged the seat in his vehicle was defectively designed, alleging theories based in both strict liability and negligence. 68 68. See id. at 848. × At the end of trial, the trial court instructed the jury that it could apply either a consumer expectation test or risk-benefit test, and the jury found in favor of the plaintiff. 69 69. See id. at 848. × The court of appeals reversed the jury verdict. 70 70. See id. at 849. ×

In affirming the Colorado Court of Appeals, the Colorado Supreme Court recognized that it had “stated repeatedly that the risk-benefit test, not the consumer expectation test, is the proper test to use in assessing whether a product like the car seat . . . is unreasonably dangerous due to a design defect.” 71 71. Id. at 850. × The Court further noted:

[P]roducts-liability law has developed in part to “encourage manufacturers to use information gleaned from testing, inspection and data analysis” to help avoid product accidents. Using the risk-benefit test . . . helps further this objective, as it directs the fact-finders to consider the manufacturer’s ability to minimize or eliminate risks and the effect such an alteration would have on the product’s utility, other safety aspects, or affordability. 72 72. Id. at 851 (quoting Camacho v. Honda Motor Co., 741 P.2d 1240, 1247 (Colo. 1987)). ×

While the authors recognize the debate about whether to apply the consumer expectations test or the risk-utility test continues to this day, and that some jurisdictions still apply the consumer expectations test, even in cases of complex products, the above referenced opinions illustrate the issues and concerns with asking jurors to determine the expectations of an ordinary consumer when evaluating a highly technical products in design defect matters.

IV.       The Consumer Expectations Test Is Not the Appropriate Test of Design Defect as Applied to Autonomous Vehicle Technology

The arrival of any new product technology will bring with it litigation, and along with that arguments for the legal standard that will place the lightest burden on plaintiffs in this new arena. Thus, it is likely that as lawsuits begin with autonomous vehicle technology, plaintiffs will argue that the consumer expectations test should apply to their claims for alleged design defects in autonomous vehicles. The argument will likely follow the reasoning employed by courts that refuse to adopt the Third Restatement approach, or that still strictly follow the consumer expectations test, i.e. that risk-utility balancing, especially when a reasonable alternative design is required, places too great of a burden on plaintiffs that do not have the resources to make showings that are so technical in nature. 73 73. See, e.g., Potter v. Chicago Pneumatic Tool Co., 694 A.2d 1319, 1332 (Conn. 1997); Vautour v. Body Masters Sports Indus., 784 A.2d 1178, 1183 (N.H. 2001). ×

A.         Highly Automated Vehicles Are Too Complex for Consumer Expectations to Govern.

The Society of Automotive Engineers lists six (6) levels of automation for HAVs. 74 74. See SAE International, supra note 3. × Currently, all vehicles on roadways are levels one and two, with Audi unveiling the world’s first production Level 3 vehicle in July 2017. 75 75. As reported in an article by IEEE Spectrum, Audi claims to have achieved level 3 through its “AI Traffic Jam Pilot” feature, which can only be activated when driving at less than 37 mph. See Philip E. Ross, The Audi A8: The World’s First Production Car to Achieve Level 3 Autonomy, IEEE Spectrum (July 11, 2017), https://spectrum.ieee.org/ca rs-that-think/transportation/self-driving/the-audi-a8-the-worlds-first-production-car-to-achieve-level-3-autonomy. × Further, even if fully autonomous vehicles were on the road today, the vast majority of consumers will remain unfamiliar with the technology for the foreseeable future. Drivers keep their vehicles on the road for over eleven years on average, 76 76. Reno Charlton, American Drivers Keeping Cars on the Road for Longer: Average Age Now 11.4 Years, Huffington Post (Aug. 9, 2013), https://www.huffpost.com/entry/american-drivers-keeping_b_3718301?guccounter=1. × so vehicles of lower automation levels will continue to be the predominant means of automotive transportation for years to come. 77 77. See Brian A. Browne, Self-Driving Cars: On the Road to a New Regulatory Era, 8. J. L., Tech. & Internet 1, 3 (2017) (Giving examples of the various lower level features many OEMs have planned for the coming years). ×

Further, NHTSA acknowledges the lack of consumer experience with autonomous vehicle technology, as well as how different these vehicles are from conventional vehicles on the roads today. In 2017, in Automated Driving Systems 2.0: A Vision for Safety, NHTSA pronounced that:

Proper education and training is imperative to ensure safe deployment of automated vehicles. Therefore, manufacturers and other entities should develop document, and maintain employee, dealer, distributor, and consumer education and training programs to address the anticipated differences in the use and operation of HAVs [highly automated vehicles] from those of conventional vehicles that the public owns and operates today. Such programs should be designed to provide the target users the necessary level of understanding to use these technologies properly, efficiently, and in the safest manner possible. 78 78. National Highway Traffic Safety Administration, Federal Automated Vehicles Policy: Accelerating the Next Revolution in Roadway Safety 24 (2016), https://www.transportation.gov/sites/dot.gov/files/docs/AV%20policy%20guidance%20PDF.pdf (emphasis added). ×

Essentially, NHTSA is recommending a completely new dimension of consumer education on how to use these products. Likewise, this education will be aimed at and received primarily by consumers who actually purchase and use autonomous vehicle technology and not automotive consumers generally.

On this point, in a 2014 survey conducted by researchers at the University of Michigan’s Transportation Research Institute, Americans were asked, “[h]ow interested would you be in having a completely self-driving vehicle . . . as the vehicle you own or lease?” The most commonly chosen answer, comprising 33.7% of responses, was “not at all interested” with another 22.4% of respondents answering that they would be only “slightly interested.” 79 79. See Brandon Schoettle & Michael Sivak, Public Opinion About Self-Driving Vehicles in China, India, Japan, the U.S. and Australia 16 (Univ. of Mich. Trans. Res. Inst. Report No. 2014-30, 2014), https://deepblue.lib.umich.edu/handle/202 7.42/109433 (emphasis added). × This information suggests that not only are most Americans personally unfamiliar with HAVs, but that a majority of Americans will not become familiar with such vehicles any time soon.

Another striking result of that survey was that, of Americans with Internet access, only 70.9% of respondents had even heard of autonomous or self-driving vehicles. 80 80. See id. at 5. × If these respondents were placed on a jury in a jurisdiction applying the consumer expectations test, roughly three of twelve jurors would be deciding liability based on the ordinary expectations of a consumer for a product about which they had never heard

Moreover, a study by various researchers in the MIT AgeLab suggests that naming conventions for autonomous or “advanced driver assistance systems” can influence the expectations that a consumer may have about these systems. 81 81. Hillary Abraham, et al., What’s in a Name: Vehicle Technology Branding & Consumer Expectations for Automation, AutomotiveUI ‘17 Proceedings of the 9th International Conference on Automotive User Interfaces and Interactive Vehicular Application 226-234 (2017), available at http://st.sigchi.org/publications/ toc/auto-ui-2017.html. × In particular, the authors of this paper observed that:

[D]rivers’ attitudes and beliefs about system capability and performance are known to influence their use of technology. Factors such as a driver’s prior experience with similar technologies, predisposed trusting tendencies, and attitudes formed from exposure to media and societal opinion might all contribute to a driver’s belief that a system can handle a task outside of its [operational design domain].” 82 82. Id. ×

Further, the authors found that “the name of a driver assistance system also has the potential to impact their perceptions of system capability. 83 83. Id. × These same perceptions or misconceptions developed by unfamiliar consumers simply from the name of a particular system are sure to carry over to these consumers ability to judge the systems if called upon in a legal setting.

This is important because, while the consumer expectation test is intended to be an objective test that is applied based on the ordinary consumer’s expectation, the gravamen of the test is that “the everyday experience of the product’s users permits a conclusion that the product’s design violated minimum safety assumptions . . . .” 84 84. See Edwards v. Ford Motor Co., 683 Fed. App’x 610, 611 (9th Cir. 2017) (quoting Soule v. General Motors Corp., 882 P.2d 298, 305, 308 (Cal. 1994)). × At least initially, and most likely for quite a period thereafter, the average juror will simply not possess the everyday experience necessary to properly assess the product in a consumer expectations analysis. Rather, it is much more appropriate and fair to aid a jury by allowing the greater body of evidence encompassed within a risk-utility analysis.

B.         Consumer Expectations of Autonomous Vehicle Technology are Inconsistent and Unrealistic at this Point.

Even when consumer expectations are drawn broadly (i.e., safe versus unsafe), instead of in terms of how a particular aspect of an autonomous vehicle should perform at a technical level, consumer expectations at this point in time have not reached any kind of meaningful consistency. For example, many consumers are highly skeptical of new HAV technology and believe that the technology is inherently unsafe. 85 85. Jeremy Hsu, 75 Percent of U.S. Drivers Fear Self-Driving Cars, But It’s an Easy Fear to Get Over, IEEE Spectrum (Mar. 7, 2016, 15:01 GMT), http://spectrum.ieee.org/ cars-that-think/transportation/self-driving/driverless-cars-inspire-both-fear-and-hope. × On the other hand, some organizations anticipate large reductions in automotive accidents and injuries as a result of this new technology and propound this message to the general public. 86 86. See, e.g., Mothers Against Drunk Driving, MADD Statement on Autonomous Vehicle Technology Legislation, (October 4, 2017), https://www.ma dd.org/press-release/madd-statement-autonomous-vehicle-technology-legislation/. × For its part, NHTSA helped promote the narrative that the promise of self-driving vehicles will lead to a marked increase in automotive safety, noting in their 2017 update that, “in the transportation sector, where 9 out of 10 serious roadway crashes occur due to human behavior, automated vehicle technologies possess the potential to save thousands of lives, as well as reduce congestion, enhance mobility, and improve productivity.” 87 87. NHTSA, supra note 6, at ii. × Some manufacturers are no different: in GM’s 2018 Self-Driving Safety Report, the manufacturer optimistically stated that as a result of self-driving technology, they “envision a future with zero crashes.” 88 88. General Motors, supra note 2 at 3 (emphasis added). ×

Further, HAV manufacturers, eager to explain the admittedly revolutionary technology their vehicles employ, may inadvertently present consumers with the impression that these vehicles truly can do no wrong. Consider the following language from Delivering Safety: Nuro’s Approach:

Our vehicle is engineered to be safer than nearly any other – it is lighter than a passenger vehicle, narrower and more nimble, and operates at lower speeds. This approach gives us more time to react, shortens our stopping distance, and provides an additional safety buffer to the side of the vehicle. Together, these advantages help prevent accidents that standard vehicles cannot avoid, such as someone jumping out from between parked cars or swerving across the road. 89 89. Nuro, Delivering Safety: Nuro’s Approach 8 (2019), https://tonnietal ler.files.wordpress.com/2019/03/d5d69-delivering_safety_nuros_approach.pdf. ×

It is certainly true that HAV technology will revolutionize automotive safety overall. However, these types of statements may lead many consumers to believe that autonomous vehicles should perform to the point of infallibility, which is simply not possible, especially at this early stage of development.

For example, on May 7, 2016, a driver of a Tesla Model S was killed when the driver collided with a tractor-trailer who was crossing an uncontrolled intersection. 90 90. NHTSA Office of Defects Investigation Report, available at https://static.n htsa.gov/odi/inv/2016/INCLA-PE16007-7876.pdf. × The vehicle’s data resulted in three important findings:

  1. That the Tesla was being driven in autopilot mode at the time of the accident;
  2. the automatic emergency braking (AEB) system did not automatically brake or warn to avoid the collision, and;
  3. that the driver did not take any preventive steps, i.e. braking or steering, to avoid the collision. 91 91. NHTSA Office of Defects Investigation Report, supra note 83, at 1. ×

Because of the accident, both the National Transportation Safety Board (“NTSB”) and the National Highway Traffic Safety Administration (NHTSA) through their Office of Defects Investigation (“ODI”) conducted investigations. 92 92. See NHTSA Office of Defects Investigation Report, supra note 83;see also NTSB, NTSB/HAR-17/02, Collision Between a Car with Automated Vehicle Control Systems and a Tractor-Semitrailer Truck (2017), available at https://www.ntsb.gov/investigations/AccidentReports/Reports/HAR1702.pdf. ×

For example, the ODI investigated: (1) the AEB system design and performance; (2) human-machine interfaces related to operating in autopilot mode; (3) additional accident data regarding Tesla’s autopilot and AEB systems; and, (4) the changes if any Tesla has made to such autopilot and AEB systems. 93 93. See NHTSA Office of Defects Investigation Report, supra note 83, at 1. × The result of the investigation was that there were no defects in the design or performance of the autopilot or AEB systems in the vehicles studied – nor was there a situation to which the systems did not perform as designed. 94 94. See id. at 12. ×

Given the situation, is it reasonable to task an “ordinary consumer” with properly determining whether the AEB and autopilot systems are in fact functioning properly or improperly? Compare the reported results of the investigations by NHTSA’s ODI and the NTSB with the statements by Forbes contributor, Brad Templeton, in his article, “Tesla Autopilot Repeats Fatal Crash; Do They Learn From Past Mistakes? 95 95. Brad Templeton, Tesla Autopilot Repeats Fatal Crash; Do They Learn From Past Mistakes?, Forbes (2019), https://www.forbes.com/sites/bradtempleton/2019/05/2 1/tesla-autopilot-repeats-fatal-crash-do-they-learn-from-past-mistakes/#400f773f2f2e. × To wit, Templeton posits, “Even so, most would hope the Tesla Autopilot would have detected the truck crossing in front of it, which appeared not to happen. No braking or evasive actions were taken. The Autopilot was engaged just 10 seconds before the collision.” 96 96. Id. × He further opines:

As such, having already had a fatality from (the old system’s) failure to identify the broad side of a transport trailer, that would have to be very high on the list of the sort of thing they would want their fleet to find and identify for them, so they can confirm it never fails to perceive a crossing truck. Somehow, it still failed. Of all the things you would expect Tesla to identify, these few things which resulted in fatal accidents, like a truck side and a highway crash attenuator, should be at the very top of the list. 97 97. Id. (emphasis in original). ×

Although Templeton is likely more informed than the ordinary consumer, the opinions expressed in his article and the conclusions reached by the NHTSA and NTSB are in clear contradiction of one another. Thus, when considering the expectations of the everyday consumer, it is clear the necessary information is simply not available to conduct investigations such as the one carried out by the ODI or the NTSB, which can take months of analysis and result in sixty-three-page accident reports, and ultimately determine what actually occurred.

Similar to the California Court of Appeal’s reasoning in Trejo that the consumer expectations test could lead to virtually unlimited liability in cases of complex products, the current climate of high expectations regarding HAVs would likely mean that a HAV manufacturer would lose every time when the consumer applications test is applied. Consumers will expect that HAVs should avoid accidents one hundred percent of the time, so any time one of these vehicles is involved in an accident, it has already failed the consumer expectations test. This type of res ipsa loquitur conclusion undermines the concept of design defects in products liability law and would allow plaintiffs to completely sidestep the requirement of a showing that an HAV was in fact defective, effectively making manufacturers of HAVs insurers of those products’ safety. 98 98. See Funkhouser v. Ford Motor Co., 736 S.E.2d 309, 314–15 (Va. 2013) (noting that in failure to warn cases, as well as in products liability cases, removal of the defect requirement could allow plaintiffs to attribute any generalized danger to a manufacturer without any showing of defect in that product). × In essence, plaintiffs would no longer bear the burden of making a showing of product defect.

Further, much of an individual consumer’s expectations about the way a vehicle should perform in an accident scenario are shaped by the behavior of other drivers. 99 99. See Michael Sivak & Brandon Schoettle, Road Safety with Self-Driving Vehicles: General Limitations and Road Sharing with Conventional Vehicles 5 (Univ. of Mich. Trans. Res. Inst. Report No. 2015-2, 2015). × Without the traditional feedback from other drivers to which consumers are accustomed, these expectations are wholly lacking to describe how autonomous vehicle technology will perform in an accident situation. 100 100. See id. × As noted in a report issued by the University of Michigan’s Transportation Research Institute, “[t]he degree of importance of both driver expectations and feedback from other drivers, and the consequent effects on the safety of a traffic system containing both conventional and self-driving vehicles, remain to be ascertained.” 101 101. See id. (emphasis added). ×

V.        Policy Reasons for Not Applying the Consumer Expectations Test to Autonomous Vehicle Technology.

As noted, autonomous vehicle technology has the potential to decrease traffic injuries and deaths. 102 102. See, e.g., Mothers Against Drunk Driving, supra note 79. × By applying the consumer expectations test, in which unknowledgeable consumers are not required to take into account the utility of a product, or the possibility of a feasible alternative design, courts could expose manufacturers to significant uncertainty in product liability litigation. If the standard by which a product will be judged is on the unpredictable expectations of consumers in such a complex and changing technology, rather than by demonstration of the product’s utility, the threshold for deployment by a manufacturer may change:

Thus, even though an autonomous vehicle may be safer overall than a conventional vehicle, it will shift the responsibility for accidents, and hence liability, from drivers to manufacturers. The shift will push the manufacturer away from the socially optimal outcome—to develop the autonomous vehicle. 103 103. See Gary E. Marchant & Rachel A. Lindor, The Coming Collision Between Autonomous Vehicles and the Liability System, 52 Santa Clara L. Rev. 1321, 1334 (2012). ×

To the contrary, under a risk-utility analysis, particularly one that requires proof of a safer, practicable alternative design, automotive manufacturers will be able to show that the societal benefits from the use of HAV technology as opposed to other technologies outweigh the risk of individual malfunctions in individual cases. 104 104. See id. ×

Consider the following example that illustrates the possible effect of unbridled consumer expectations on the introduction of beneficial new technology:

Suppose . . . that a particular type of “autobrake” crash-avoidance technology works to prevent crashes 80 percent of the time. The other 20 percent of the time, however, the technology does not work and the crash occurs as it would have in the absence of the technology. Victims in those crashes may sue the manufacturer and argue that the product was defective because it failed to operate properly in their crashes. Under existing liability doctrine, they have a plausible argument: The product did not work as designed . . . . A manufacturer facing the decision whether to employ such a technology in its vehicles might very well decide not to, purely on the basis of expected liability costs. 105 105. James M. Anderson et al., RAND Corp. Autonomous Vehicle Technology: A Guide for Policymakers 125 (2016). ×

Without any balancing of the utility of these vehicles or the requirement of a reasonable alternative design, it would be possible, even reasonable, for juries applying the consumer expectations test to find defective design every time. This will be especially true in situations such as those involving self-driving vehicle technology, since consumers tend to have unrealistic expectations about the benefits of this new technology as a whole. 106 106. See id. at 125. ×

Further, the consumer expectations test will not allow for consideration of non-safety related societal benefits that HAV technology provides, since the only consideration will be on whether the product performed as expected in that one instance. The average American commuter spends about one week of his or her life in traffic each year—a statistic that HAV manufacturers have set their sights on reducing. 107 107. See General Motors, supra note 2, at 3; see also David Schrank, Bill Eisele, Et Al., The Texas A&M Transportation Institute & INRIX, 2015 Urban Mobility Scorecard 1–2 (2015), https://static.tti.tamu.edu/tti.tamu.edu/documents/mo bility-scorecard-2015.pdf (noting that, as of 2014, the American commuter spends an average of approximately 42 hours per years in traffic). × The potential time saved by commuters on the whole is not a factor that would be considered under the consumer expectations test.

Another benefit of HAV technology outside of the realm of safety is the potential for added mobility for those who cannot currently drive. 108 108. See generally Waymo, supra note 2, at 6. × According to a report from NHTSA, 3 million Americans are blind or suffer from poor vision. 109 109. NHTSA, DOT HS 811 304, Quieter Cars and the Safety of Blind Pedestrians: Phase I 6 (2010), https://www.nhtsa.gov/DOT/NHTSA/NVS/Crash%20Avoidance/Technical%20Publications/2010/811304rev.pdf. × Further, 79 percent of Americans over the age of 65 live in car-dependent communities. The independence these communities could gain with the widespread use of HAV technology would be yet another consideration the jury could not take into account when utilizing the consumer expectations test.

VI.       Conclusion

Courts should reject the consumer expectations test as grounds for determining design defect in cases involving autonomous vehicle technology. This technology is simply too complex and unfamiliar for consumer expectations to have developed enough to have any real meaning or reasonable application. Utilization of risk-utility balancing is a more appropriate means of establishing whether or not a design is defective and will encourage manufacturers to continue to develop and implement this important technology, which stands to have a truly revolutionary impact on automotive safety.


  John Isaac Southerland is a partner at Huie Fernambucq & Stewart LLP in Birmingham, Alabama. Mr. Southerland’s practice areas include automotive product liability, personal injury, heavy equipment product liability, trucking litigation, and towing and recovery liability. He also serves as national coordinating discovery counsel for a major automotive client. Mr. Southerland is a frequent lecturer at various industry conferences and has written and spoken about the emergence of highly automated vehicles and technology on numerous occasions.  He also serves as a Barrister and the Programs Chairperson in the James Edwin Horton Inn of Court at Cumberland School of Law. 

Elizabeth Davis is an associate at Huie, Fernambucq & Stewart LLP in Birmingham, Alabama. Ms. Davis concentrates her law practice in the areas of automotive litigation, product liability and discovery practice and procedure, including serving as national coordinating discovery counsel for a major automotive client. Ms. Davis is an active member of Alabama Defense Lawyers Association, Birmingham Bar Association, and Defense Research Institute.

Woods Parker is an associate at Huie, Fernambucq & Stewart LLP in Birmingham, Alabama. Mr. Parker concentrates his law practice in the areas of automotive litigation, product liability, trucking litigation, consumer lemon law, and discovery practice and procedure, including serving as national coordinating discovery counsel for a major automotive client. Mr. Parker is an active member of Alabama Defense Lawyers Association, Birmingham Bar Association, and Defense Research Institute.

Earlier this month, the Journal of Law and Mobility hosted our first annual conference at the University of Michigan Law School. The event provided a great opportunity to convene some of the top minds working at the intersection of law and automated vehicles. What struck me most about the conference, put on by an organization dedicated to Law and mobility, was how few of the big questions related to automated vehicles are actually legal questions at this point in their development.

The afternoon panel on whether AVs should always follow the
rules of the road as written was emblematic of this juxtaposition. The panel
nominally focused on whether AVs should follow traffic laws. Should an
automated vehicle be capable of running a red light, or swerving across a
double yellow line while driving down the street? Should it always obey the
posted speed limit?

The knee-jerk reaction of most people would probably be
something along the lines of, “of course you shouldn’t program a car that can
break the law.” After all, human drivers are supposed to follow the law. So why
should an automated vehicle, which is programmed in advance by a human making a
sober, conscious choice, be allowed to do any differently?

Once you scratch the surface though, the question becomes
much more nuanced. Human drivers break the law in all kinds of minor ways in
order to maintain safety, or in response to the circumstances of the moment. A
human driver will run a red light if there is no cross-traffic and the car
bearing down from behind is showing no signs of slowing down. A human will
drive into the wrong lane or onto the shoulder to avoid a downed tree branch, or
a child rushing out into the street. A human driver may speed away if they
notice a car near them acting erratically. All of these actions, although they
violate the law, may be taken in the interest of safety in the right
circumstances. Even knowing they violated the law, a human driver who was
ticketed in such a circumstance would feel their legal consequence was
unjustified.

If automated vehicles should
be able to break the law in at least some circumstances, the question shifts –
which circumstances? Answering that question is beyond the scope of this post.
At the moment, I don’t think anyone has the right answer. Instead, the point of
this post is to highlight the type of moment-to-moment decisions every driver
makes every day to keep themselves and those around them safe. The rules of the
road provide a rough cut, codifying what will be best for most people most of
the time. They could not possibly anticipate every situation and create a
special legal rule for that situation. If they tried, the traffic laws would
quickly grow to fill several libraries.

In my view, the question of whether an AV should be able to
break the law is only tangentially a legal question. After arriving at an
answer of, “probably sometimes,” the question quickly shifts to when, and in
what circumstances, and whether the law needs to adapt to make different
maneuvers legal. These questions have legal aspects to them, but they are also
moral and ethical questions weighted with a full range of human driving
experience.  Answering them will be among
the most important and difficult challenges for the AV industry in the coming
years.

The common story of automated vehicle safety is that by eliminating human error from the driving equation, cars will act more predictably, fewer crashes will occur, and lives will be saved. That future is still uncertain though. Questions still remain about whether CAVs will truly be safer drivers than humans in practice, and for whom they will be safer. In the remainder of this post, I will address this “for whom” question.

A recent study from Benjamin Wilson, Judy Hoffman, and Jamie Morgenstern at Georgia Tech found that state-of-the-art object detection systems – the type used in autonomous vehicles – demonstrate higher error rates in detection of darker-skinned pedestrians as compared to lighter-skinned pedestrians. Controlling for things like time-of-day or obstructed views, the technology was five percentage points less accurate at detecting people with darker skin-tones.

The Georgia Tech study is far from the first report of algorithmic bias. In 2015, Google found itself at the center of controversy when its algorithm for Google Photos incorrectly classified some black people as gorillas. More than two years later, Google’s temporary fix of removing the label “gorilla” from the program entirely was still in place. The company says they are working on a long-term fix to their facial recognition software. However, the continued presence of the temporary solution several years after the initial firestorm is some indication either of the difficulty of achieving a real solution or the lack of any serious coordinated response across the tech industry.

Algorithmic bias is a serious problem that must be tackled with a serious investment of resources across the industry. In the case of autonomous vehicles, the problem could be literally life and death. The potential for bias in automated systems begs for an answer to serious moral and legal questions. If a car is safer overall, but more likely to run over a black or brown pedestrian than a white one, should that car be allowed on the road? What is the safety baseline against which such a vehicle should be judged? Is the standard, “The AV should be just as likely (hopefully not very likely) to hit any given pedestrian?” Or is it “The AV should hit any given pedestrian less often than a human driven vehicle would?” Given our knowledge of algorithmic bias, should an automaker be opened up to more damages if their vehicle hits a black or brown pedestrian than when it hits a white pedestrian? Do tort law claims, like design defect or negligence, provide adequate incentive for automakers to address algorithmic bias in their systems? Or should the government set up a uniform system of regulation and testing around the detection of algorithmic bias in autonomous vehicles and other advanced, potentially dangerous technologies?

These are questions that I cannot answer today. But as the Georgia Tech study and the Google Photos scandal demonstrate, they are questions that the AV industry, government, and society as a whole will need to address in the coming years.

In the coming decades, advancing technology is likely to strain many tried-and-true legal concepts.  The tort law cause of action for design defects is likely to be among the most impacted. This post will explore the current understanding of design defect claims, and highlight areas where autonomous vehicles and other highly complex technologies will likely lead to a rethinking of the doctrine.

As outlined in the Third Restatement of Torts, design defect claims can be brought against a manufacturer when “the foreseeable risks of harm posed by the product could have been reduced or avoided by the adoption of a reasonable alternative design . . . and the omission of the alternative design renders the product not reasonably safe.” Essentially, plaintiffs who bring a design defect claim after being harmed by a product bear the burden of showing that the product was designed unreasonably for its intended use, and that an alternative design would have been safer for the user and reasonable for the manufacturer to adopt.

Traditionally, courts have adopted one of two tests to determine the reasonableness of a product design under these claims. Under the consumer expectations test, the key question is whether a product performed up to the level at which an ordinary consumer would expect. Because this test is based on the expectations of an ordinary consumer (one who is presumed to not have any special knowledge about the product), a claim can be successful without any expert testimony about the alleged design failure, as in McCabe v. American Honda Motor Corp.

Alternatively, many courts have adopted the risk-utility test for proving a design defect. The risk-utility test is more akin to cost-benefit balancing. The South Carolina Supreme Court in Branham v. Ford Motor Co. noted that the risk-utility test balances “numerous factors . . . including the usefulness and desirability of the product, the cost involved for added safety, the likelihood and potential seriousness of injury, and the obviousness of danger.” As this test requires testimony on cost of the current design and any proposed alternatives, it will require expert testimony and specialized knowledge, unlike the consumer expectations test.

Some have made the case that the consumer expectations test will be inadequate to address claims of design defect in complex technologies such as autonomous vehicles. After all, the argument goes, how could an ordinary consumer possibly have a realistic expectation of how an autonomous vehicle is supposed to perform in a given situation? Any given action by an AV is the result of a series of algorithms that is being constantly updated as the car gathers new information about the world around it. Should a consumer expect the AV to act just as a human would act? Should it be more cautious? Or perhaps even take actions that would seem overly risky for a human driver, because the AV system was certain of what every step in its maneuver would look like going in? How could a human passenger know? If courts are persuaded by these concerns, they will likely need to address them by adopting the more expert-reliant risk-utility test.

On the other hand, some scholars argue that the consumer expectations test is perfectly adequate to handle claims involving advanced technology such as AVs. In a recent article, NYU Law Professor Mark Geistfeld notes that consumers need not understand the intricacies of how a technology works in order to have “well-formed expectations of the product performance.” Under Geistfeld’s approach, a consumer either should have such a well-formed expectation or, in the case where they have yet to develop one, should be warned by the manufacturer or dealer in such a way as to make them aware of the risk they are taking on.

It remains to be seen how design defect claims will be forced to evolve as autonomous vehicles come on the scene. Like many areas of law though, this is a field that will be stressed, and potentially forced to evolve, by the advent of this revolutionary technology.

Cite as: Kyle D. Logue, The Deterrence Case for Comprehensive Automaker Enterprise Liability, 2019 J. L. & Mob. 1.

I.         Introduction

Automobiles are much safer today than they used to be. Perhaps the best illustration of this fact is the decades’ long decline in the number of auto-related deaths per-mile-driven. 110 110. General Statistics: Fatality Facts, INS. INST. FOR HIGHWAY SAFETY: HIGHWAY LOSS DATA INST., https://www.iihs.org/iihs/topics/t/general-statistics/fatalityfacts/over view-of-fatality-facts (last visited Nov. 13, 2018) (showing motor vehicle crash deaths per 100 million miles driven has declined from 3.35 in 1976 to 1.18 in 2016). × And yet motor vehicles—including cars, trucks, and SUVs— continue to be among the most dangerous products sold anywhere. Automobiles pose a larger risk of accidental death than any other product, except perhaps for opioids. 111 111. In 2016, which is the most recent year for which the Centers for Disease Control has final data as of the time of this writing, there were 58,335 deaths attributable to accidental poisoning, which includes accidental deaths from drug overdose (which, of course, includes accidental opioid overdose). JIAQUAN XU ET AL., CTRS. FOR DISEASE CONTROL & PREVENTION, NATIONAL VITAL STATISTICS REPORTS 34 (vol. 67, no. 5 July 26, 2018) [henceforth, CDC, 2016 Final Death data]. The CDC estimates that roughly 42,000 deaths in 2016 were attributable to opioids, the vast majority of which would presumably be considered accidental deaths. Drug Overdose Death Data, CTRS. FOR DISEASE CONTROL & PREVENTION, https://www.cdc.gov/drugoverdose/data/statedeaths .html (last visited Nov. 13, 2018). Note also that firearms are involved in more deaths per year than motor vehicles, but the vast majority of those deaths are caused intentionally, either suicide (22,938 deaths in 2016) or homicide (14,415). The other leading causes of accidental deaths that year included the following: falls (34,673), firearms (495), and drowning (3,786). CDC, 2016 Final Death data, at 50. × Annual auto-crash deaths in the United States have never fallen below 30,000, reaching a recent peak of roughly 40,000 in 2016. 112 112. For 2016, there were 40,327 motor-vehicle-related accidental deaths. CDC, 2016 Final Death data, at 52. × In addition to these tens of thousands (internationally, millions 113 113. Internationally, the number of annual fatalities attributable to motor vehicle accidents is in the millions. Number of Road Traffic Deaths, WORLD HEALTH ORG., http://www.who.int/gho/road_safety/mortality/number_text/en/ (last visited Nov. 13, 2018) (estimating worldwide auto accident deaths in 2013 to be roughly 1.25 million). × ) of deaths attributable to motor-vehicle crashes, there are many other social costs. Victims of serious auto accidents, for example, often incur extraordinary medical expenses both to provide treatment immediately after the accident and, sometimes, to provide treatment for the rest of their lives. 114 114. LAWRENCE BLINCOE ET AL., NAT’L HIGHWAY TRAFFIC SAFETY ADMIN., THE ECONOMIC AND SOCIETAL IMPACT OF MOTOR VEHICLE CRASHES, 2010 (REVISED) 5 (May 2015) (finding medical costs responsible for $23.4 billion of the total economic cost of motor vehicle crashes in 2010). × Those crash victims whose injuries render them unable to work can experience weeks, months, even years of lost income, which, from their employers’ perspective, is lost productivity. 115 115. Id. (finding $77.4 billion in lost productivity as a result of motor vehicle crashes in 2010). × Auto accidents also cause non-trivial amounts of property damage, mostly to the automobiles themselves though also occasionally to highways, bridges, or other elements of transportation infrastructure. Finally, serious motor vehicle accidents often cause severe noneconomic injuries—that is, severe “pain and suffering”—as a result of accident victims’ painful and debilitating physical injuries. According to some estimates, such noneconomic harms, in the aggregate, amount to more than twice the magnitude of the aggregate economic damages caused by auto accidents. 116 116. DANIEL SMITH, NAT’L HIGHWAY TRAFFIC SAFETY ADMIN., OVERVIEW OF NHTSA PRIORITY PLAN FOR VEHICLE SAFETY AND FUEL ECONOMY, 2015 TO 2017 at 2 (June 2015) (“In addition to the terrible personal toll, these crashes have a huge economic impact on our society with an estimated annual cost of $242 billion, which is an average of $784 for every person in the United States. These crashes also result in $594 billion in societal harm from loss of life and the pain and decreased quality of life due to injuries.”). ×

All of this may be about to change. According to many auto-industry experts, the eventual transition to driverless vehicles will drastically lower the economic and noneconomic costs of auto accidents. 117 117. See, e.g., Adrienne LaFrance, Self-Driving Cars Could Save 300,000 Lives Per Decade in America, THE ATLANTIC (Sept. 29, 2015), https://www.theatlantic.com/ technology/archive/2015/09/self-driving-cars-could-save-300000-lives-per-decade-in-america/407956/ (“Researchers estimate that driverless cars could, by midcentury, reduce traffic fatalities by up to 90 percent.”). × Why might this be so? Because humans are so bad at driving. When it comes to operating motor vehicles, people have bad judgment, slow reflexes, inadequate skills, and short attention spans. They drive too fast. They drive while intoxicated. They drive while sleepy. They drive while distracted. In fact, according to the National Highway Traffic Safety Administration, roughly 94 percent of auto accidents today are attributable to “driver error.” 118 118. NAT’L HIGHWAY TRAFFIC SAFETY ADMIN., CRITICAL REASONS FOR CRASHES INVESTIGATED IN THE NATIONAL MOTOR VEHICLE CRASH CAUSATION SURVEY 2 (Mar. 2018), https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812506. Another 2 percent of accidents are attributable to vehicle component failure, 2 percent to environmental conditions (such as slick roads), and 2 percent to “unknown.” Id. What precisely these statistics mean, however, is not entirely clear. Specifically, it is not obvious how NHTSA’s statistical categories (such as “vehicle component failure”) would relate to analogous legal concepts (such as “defective product”). × Computers can do better. At least that is the hope: that machine-learning computer algorithms, in combination with state-of-the-art sensors and advanced robotics, will be better—much better—drivers than humans are. 119 119. For an extended argument for why driverless cars are better drivers than humans, see, e.g., HOD LIPSON & MELBA KURMAN, DRIVERLESS: INTELLIGENT CARS AND THE ROAD AHEAD (MIT Press 2016). × Whether this will in fact be true is still unproven, but is most likely to be true with respect to so-called fully driverless “Level 5” vehicles, 120 120. See infra note 12. Level 5 vehicles are also sometimes called autonomous vehicles, though that use of the term ignores the distinction between connected vehicles and truly autonomous vehicles. This Article will larger ignore that distinction as well. × which are those autonomous or connected vehicles that are capable of operating on any road and under any conditions that a human driver can handle but with no input from a human passenger other than the choice of destination. 121 121. It is these Level 5 vehicles that hold the real promise for substantial accident-risk reduction. SAE INTERNATIONAL, TAXONOMY AND DEFINITIONS FOR TERMS RELATED TO DRIVING AUTOMATION SYSTEMS FOR ON-ROAD MOTOR VEHICLES J3016 (2018), https://saemobilus.sae.org/content/j3016_201806 (explaining levels 0, no autonomous features, through level 5, where a computer is “operating the vehicle on-road anywhere that a typically skilled human driver can reasonably operate a conventional vehicle”). This is in large part because Level 5 vehicles do not have the “handoff problem,” which occurs at that moment (with levels 1 through 4) when control of the vehicle must be transferred from the algorithm to the human driver. Alex Davies, The Very Human Problem Blocking the Path to Self-Driving Cars, WIRED (Jan. 1, 2017), https://www.wired.com/2017/01/human-problem-blocking-path-self-driving-cars/ (discussing how existence of handoff problem led Google, in 2012, and other companies more recently, to commit to developing level 5 autonomy). × Level 5 vehicles, because they would not suffer from the problems that plague human decision making in the driving context, do hold the promise to be substantially safer than the fully or even partially human-driven alternative. 122 122. One can certainly imagine the possibility of driving algorithms going haywire or sensors failing in ways that cause terrible accidents. But presumably, regulators will not permit Level 5s to be sold until they prove themselves in large numbers of test miles to be substantially safer than human drivers. Some commentators have suggested that regulators not approve Level 5s unless and until they are shown to be twice as safe as human drivers. Mark A. Geistfeld, A Roadmap for Autonomous Vehicles: State Tort Liability, Automobile Insurance, and Federal Safety Regulation, 105 CALIF. L. REV. 1611, 1653 (2017). ×

As promising as a world of highways filled with computer-driven vehicles might be, from an accident-reduction perspective, 123 123. If the advent of autonomous and connected vehicles means more vehicles on the road, it could be bad news for efforts to combat climate change and improve air quality. × such a high-tech world is still only a possibility. And even if it happens, it will not be for a number of years. There continue to be major technological hurdles, as well as potential consumer resistance to actually riding in a driverless vehicle. 124 124. American Drivers Grow More Afraid of Driverless Vehicles, INS. J. (May 22, 2018), https://www.insurancejournal.com/news/national/2018/05/22/490014.htm (noting that 73% of American drivers report being too afraid to ride in a fully automated vehicle). × Therefore, the introduction, spread, and eventual dominance of Level 5s will take some time. 125 125. Many experts think consumers will not even be able to purchase fully autonomous vehicles for another decade. Justin Gerdes, Not So Fast. Fully Autonomous Vehicles are More than a Decade Away, Experts Say, GREEN TECH MEDIA (Feb. 6, 2018), https://www.greentechmedia.com/articles/read/fully-autonomous-vehicles-decade-away-experts (reporting results of informal poll of 300 industry experts). × During that transition, most automobiles will continue to be driven mostly by humans. Indeed, even in the long run, when Level 5 vehicles have been perfected and are available to the general public either through individual purchases and leases or through some ride-sharing arrangement (via Uber or Lyft or some similar web-based platform), we should still expect to see a substantial number of fully or partially human-driven vehicles traveling alongside them. 126 126. See, e.g., Background On: Self-Driving Cars and Insurance, INS. INFO. INST. (July 30, 2018), https://www.iii.org/article/background-on-self-driving-cars-and-insurance (“According to the Insurance Institute for Highway Safety, it is anticipated that there will be 3.5 million self-driving vehicles on U.S. roads by 2025, and 4.5 million by 2030. However, the institute cautioned that these vehicles would not be fully autonomous, but would operate autonomously under certain conditions.”). ×

If I am right about this picture of the automotive future, what should the role of auto tort law be, now and going forward? More specifically, if we conceive of auto tort law—including both automaker product liability and driver negligence liability (and the insurance that covers both types of liability)—as a system of ex post auto-crash deterrence, what would the optimal or efficient auto tort/insurance regime look like? 127 127. For this Article, I assume that the primary role of auto tort law is efficient deterrence. That means, creating incentives that induce all relevant parties—drivers, automakers, even pedestrians—to take efficient or cost-justified steps to minimize the probability and severity of accidents. On this view, the goal is not necessarily zero accidents, because the cost of accident avoidance eventually renders additional investments in accident prevention inefficient and socially undesirable. This is a standard type of normative analysis of accident law. It is, of course, not the only way to evaluate an accident law regime. For example, if the primary function of auto tort law were instead merely compensation for the harms caused by auto accidents, or were to achieve corrective justice (in the sense of reversing wrongfully caused harms), some system other than the one proposed in this Article might make more sense. × Further, how should such an optimally designed auto tort/insurance regime take into account the emergence of Level 5 vehicles?

These questions are the subject of this Article. Specifically, this Article lays out the potential (at this point purely theoretical) deterrence benefits of replacing our current auto tort regime (including auto products liability law, driver-based negligence claims, and auto no-fault regimes) with a single, comprehensive automaker enterprise liability system. 128 128. The term “enterprise liability” has long been used to stand for the idea that “business enterprises ought to be responsible for losses resulting from products they introduce into society.” George L. Priest, The Invention of Enterprise Liability: A Political History of The Intellectual Foundations of Modern Tort Law, 14 J. LEGAL STUD. 461, 463 (1985) (describing intellectual history of enterprise liability idea). See also Gregory C. Keating, The Theory of Enterprise Liability and Common Law Strict Liability, 54 VAND. L. REV. 1285, 1287 (2001) (“[E]nterprise liability expresses the maxim that those who profit from the imposition of risk should bear the costs of the accidents that are a price of their profits.”). The concept of enterprise liability was much discussed in the 1980s and 1990s among tort scholars. See, e.g., Priest, supra; James A. Henderson Jr. The Boundary Problems of Enterprise Liability, 41 MD L. REV. 659 (1982) (discussing line-drawing issues that arise in connection with adopting enterprise liability regimes); and Kenneth S. Abraham & Paul C. Weiler, Enterprise Medical Liability and the Evolution of the American Health Care System, 108 HARV. L. REV. 381 (1994) (applying enterprise liability concepts to medical system). I, together with my colleagues and friends Jon Hanson and Steve Croley, started writing about enterprise liability around this time. See, e.g., Jon D. Hanson & Kyle D. Logue, The First-Party Insurance Externality: An Economic Justification for Enterprise Liability, 76 CORNELL L. REV. 129 (1990); and Steven P. Croley & Jon D. Hanson, Rescuing the Revolution: The Revived Case for Enterprise Liability, 91 MICH. L. REV. 683 (1993). × This new regime would apply not only to Level 5 vehicles, but to all automobiles made and sold to be driven on public roads. 129 129. For a more recent proposal to create a special auto-manufacturer responsibility regime, which has similarities to the one I am describing here, but that—critically—would be limited to accidents involving fully automated vehicles, see Kenneth S. Abraham & Robert L. Rabin, Automated Vehicles and Manufacturer Responsibility for Accidents: A New Legal Regime for a New Era, 105 VA. L. REV. (forthcoming 2019). × Because such a system would make automakers unconditionally responsible for the economic losses resulting from any crashes of their vehicles, it would in effect make automakers into auto insurers as well, although such a change will likely lead to some restructuring in how automobiles are insured and sold. Or so I will argue.

My basic argument is that a comprehensive automaker enterprise liability regime may have previously unexplored, or at least forgotten, deterrence benefits. 130 130. In an article published in 1985, Professor Howard Latin outlined an automaker enterprise liability proposal similar to the one I am describing in this Article. Howard A. Latin, Problem-Solving Behavior and Theories of Tort Liability, 73 CAL. L. REV. 677 (1985). That article, brought to my attention by Professor Stephen Sugarman, also makes a deterrence case for adopting an automaker enterprise liability regime, emphasizing some (though not all) of the same arguments I make here. See also Bryant Walker Smith, Regulation and the Risk of Inaction, in AUTONOMOUS DRIVING 584 (Markus Maurer, et al. eds., 2016), https://link.springer.com/chapter/10.1007/978-3-662-48847-8_27 (exploring (very briefly) the idea of using auto enterprise liability as a means of encouraging automotive safety innovation). × First, it could greatly simplify our existing auto tort regime by replacing all of automaker liability law (including product design defect claims) and driver liability law (as well as existing no-fault regime) with a single enterprise liability regime under which all auto-accident victims could seek recovery. Second, it could encourage automakers to design and manufacturer safer vehicles, whether that means safer human-driven vehicles (with automated features) or Level 5 vehicles. Third, it could incentivize automakers to provide better warnings and instructions with their vehicles, including better ways to deal with the “hand off” problem that occurs when vehicles switch from semi-self-driving mode to human-driven mode. 131 131. Steven Ashley, Level 3 “Hand Off” is Challenging AI Researchers, SAE INT’L (Jan. 17, 2017), https://www.sae.org/news/2017/01/sae-level-3-hand-off-is-challenging-ai-researchers. × Fourth, enterprise liability could result in automobile prices that better reflect the actual costs of driving, leading to more optimal levels of auto sales and miles driven. Fifth, enterprise liability could induce auto companies to coordinate (in a way they are not presently coordinating) with the one industry that has more information than the auto companies have about how the specific driving patterns of individual human drivers affect the risk of auto accidents: namely, the auto insurance industry. Finally, a comprehensive automaker enterprise liability regime would provide an implicit subsidy for the development and deployment of driverless technology, but only to the extent that automakers actually expect such technology to reduce accident costs. All of these points will be developed below.

The argument will proceed as follows. Part II evaluates existing auto tort law—including automaker liability law and driver liability law—from the perspective of optimal deterrence. Part III outlines one plausible version of a comprehensive automaker enterprise liability regime and summarizes the primary deterrence advantages of such a regime. Part IV briefly concludes with a discussion of caveats, concerns, and a list of questions for future research.

II.         Evaluating the Deterrence Implications of Current Auto Tort Law

Automaker Liability Law

To understand the deterrence benefits of an auto enterprise liability regime, it is necessary first to understand the deterrence consequences of the current auto tort regime. To that end, this Part describes the current auto tort system—both automaker liability law and driver liability law—and, drawing on well-known insights from deterrence theory and economic analysis of liability rules, explores what the general deterrence consequences of that regime might be. This is an entirely theoretical discussion. The ultimate question—which auto tort regime comes closes to minimizing the costs of auto accidents—can of course only be answered with empirical research that is beyond the scope of this short paper.

Current automaker liability law, like manufacturer liability law generally, is primarily a negligence-based regime, by which I mean the following: Under current law in most U.S. jurisdictions, individuals who suffer harm caused in an automobile crash can recover from the automaker in tort if they can prove that the harm resulted from negligence (or a lack of reasonable care) on the part of the automaker in designing or constructing the vehicle. 132 132. See, e.g., Larsen v. GM, 391 F.2d 495, 504 (1968) (holding, among other things, that auto manufacturers have a duty to use reasonable care in design and construction of vehicles). × Alternatively, auto accident victims can invoke modern products liability doctrine and argue that a “defect” in the vehicle’s design, manufacturing process, or warnings caused the harm. 133 133. RESTATEMENT (THIRD) TORTS: PROD. LIAB. §§1 & 2 (AM. LAW INST. 1998) (setting forth general rules for liability resulting from product defects). × This latter approach also typically requires some showing of automaker negligence. This is because, in the bulk of U.S. jurisdictions, important aspects of the product defect law are equivalent to negligence law. 134 134. A majority of jurisdictions apply a risk-utility version of the design defect test, which is similar to the common cost-benefit formulations of negligence. RESTATEMENT (THIRD) TORTS: PROD. LIAB., §2, cmt. D (AM. LAW. INST. 1998). Moreover, the adequacy of product warnings is often evaluated according to a negligence-based reasonableness test. RESTATEMENT (THIRD) TORTS: LIABILITY FOR PHYSICAL AND EMOTIONAL HARM, § 3 (AM. LAW. INST. 1998). Also, in the jurisdictions that define design defect according to “reasonable consumer expectations,” there is an obvious reliance on negligence-based principles as well, such as the concept of reasonableness. Current products liability law with respect to warnings is also essentially a negligence-based regime. RESTATEMENT (THIRD) TORTS: PROD. LIAB., §2, cmt. I (AM. LAW. INST. 1998). (“Commercial product sellers must provide reasonable instructions and warnings about risks of injury posed by products.”). × For this reason, auto products-liability, despite sometimes being labeled a form of “strict liability,” 135 135. For an example of lawyers characterizing auto products liability generally, including design defect and warning defect claims, as a form of “strict liability,” see, e.g., DEREK H. SWANSON & LIN WEI, MCGUIREWOODS, UNITED STATES AUTOMOTIVE PRODUCTS LIABILITY LAW 7 (2009), https://www.mcguirewoods.com/news-resources/publications/us-automotive-products-liability.pdf. × is in fact largely a form of negligence liability. 136 136. I am of course not the first person to observe that modern “strict” products liability operates in practice largely as a negligence regime. See, e.g., DAVID G. OWEN, PRODUCTS LIABILITY LAW 38 (3rd ed. 2015) (“These two propositions—that manufacturers must guard against risks only if they are foreseeable, and that manufacturers must guard against those risks only with precautions that are reasonable—are the two major pillars of modern products liability law in America.”). This is not to say that there are no aspects of strict liability in the current auto products liability system. For example, manufacturing defect cases approximate true strict liability. That is, when the product’s design satisfies the risk-utility standard (or is, in a sense, reasonable or non-negligent) but the particular product that caused the harm in the case did so as a result of some sort of malfunction in the vehicle that is not a result of negligent maintenance on the part of the vehicle owner (e.g., the brakes or the steering mechanism simply fails), the automaker is strictly liable. With respect to Level 5 vehicles, presumably a much higher percentage of the accidents would be a result of vehicular malfunction than is the case with human-drive vehicles; thus, there would be a much larger domain of true strict liability if existing product liability doctrine were applied to Level 5 accidents than is currently the case with automaker liability cases. Further, in design defect jurisdictions that apply a consumer expectations test, strict liability also seems likely, assuming courts conclude that consumers reasonably expect Level 5s not to crash. Still, there would be some negligence-based liability with respect to the design of the vehicles and the algorithm that drives them. See, e.g., Bryant Walker Smith, Automated Driving and Product Liability, 2017 MICH. ST. L. REV. 1 (2017) (noting that, applying existing law, automaker liability for Level 5 accidents will likely turn on some version of “unreasonable performance” by the vehicle, which sometimes will approximate strict liability and sometimes negligence); and Mark A. Geistfeld, A Roadmap for Autonomous Vehicles: State Tort Liability, Automobile Insurance, and Federal Safety Regulation, 105 CAL. L. REV. (forthcoming 2018) (observing that applying existing products liability law to Level 5s will sometimes result in strict liability and sometimes negligence-based liability). ×

A negligence-based automaker liability regime can in theory have certain deterrence advantages, if one makes particular assumptions. Those assumptions, however, are keys to the analysis—and do not always apply. For starters, a negligence-based automaker liability regime can create efficient incentives with respect to automaker care levels. Automaker “care levels” are the precautions taken by automakers—in the design, production, and warnings with respect to their vehicles—that reduce the probability or severity of auto accidents. 137 137. See generally STEVEN SHAVELL, ECONOMIC ANALYSIS OF ACCIDENT LAW 73–85 (1987) (comparing theoretical deterrence benefits of negligence liability rule and strict liability rule, in terms of care levels and activity levels); WILLIAM M. LANDES & RICHARD A. POSNER, THE ECONOMIC STRUCTURE OF TORT LAW 54–84 (1987) (same); and A. MITCHELL POLINSKY, AN INTRODUCTION TO LAW AND ECONOMICS 113–123 (3rd ed. 2003) (discussing choice of optimal products liability rules in particular). × Efficient automaker care levels occur when the automaker has made all available investments in care—in crash-risk reduction—that reduce expected auto-accident costs by more than the marginal costs of the additional care. 138 138. Put differently, when an automaker is investing efficiently in care, there are no additional investments in accident reduction that could be made that would reduce expected accident costs by more than the costs of accident avoidance. × Thus, an efficient negligence-based tort liability rule would hold an automaker liable for the harms resulting from a given auto accident only if that automaker failed to take efficient care. For example, if there was an alternative automotive design or alternative warning that the automaker could have used that would have reduced expected accident costs by more than the marginal costs of that design or warning change, failing to deploy that alternative design or warning in their vehicles would constitute negligence on the part of the automaker, and would therefore be potential grounds for tort liability. 139 139. The plaintiff must also demonstrate causation. ×

This sort of efficient negligence-based liability rule would induce automakers to take efficient care if we assume the following to be true: (a) that automakers are aware of the law and respond rationally to it, and (b) that courts applying a negligence-based automaker liability rules perform a thorough and accurate cost-benefit analysis (for example, judges and juries do not tend to make systematic errors in their determinations regarding what constitutes automaker negligence or what counts as a design defect). Under those assumptions, the negligence-based regime would incentivize efficient automaker care levels. Why? Because automakers would under those assumptions realize that they can avoid negligence-based liability entirely if they merely make all cost-justified investments in auto safety (e.g., all cost-justified design and warning changes). Knowing this, they would have a strong legal and financial incentives to do just that. 140 140. This insight is simply an application of a standard conclusion regarding the effects on injurer care levels of a perfectly applied negligence rule. See generally supra sources cited in note 28. ×

In addition, a negligence-based automaker liability regime can also create incentives for efficient driver care-levels—incentives for drivers to drive reasonably carefully—even in the absence of a defense of contributory negligence or comparative fault. 141 141. See generally supra sources cited in note 28. × This is because a negligence-based regime, by its nature, leaves accident costs on victims and their insurers when the automaker is not negligent. That fact will induce drivers to drive carefully, so as to minimize their own risk of uncompensated accident losses. Again, however, this conclusion holds only if certain key assumptions are also true. Specifically, we must assume the following: a) that drivers, like automakers, are knowledgeable about tort law and respond rationally to the potential of tort rules to apply to their future conduct; and b) that drivers actually bear these costs and do not externalize them to someone else.

To put all of this together, according to standard deterrence theory, an efficiently and accurately applied negligence-based automaker liability rule can produce efficient incentives for both automakers and drivers to take care to avoid auto accidents. 142 142. See generally supra sources cited in note 28. This conclusion also assumes that automakers are well informed about and respond rationally to tort liability rules. ×

But there are obvious problems with this rosy picture. First, consider the effects on automaker care levels if we relax the assumption that courts accurately apply negligence-based standards. If judges and juries are not very good at doing the complex and information-intensive analysis necessary to determine what particular automotive designs, warnings, or instructions are cost-justified or reasonable (or not defective), the outcomes of courts’ negligence determinations become highly uncertain. This can in turn produce incentives for automakers both to over-invest and to under-invest in auto safety. 143 143. See generally Mark Grady, A New Positive Economic Theory of Negligence, 92 Yale L.J. 799 (1983); and John Calfee & Richard Craswell, Some Effects of Uncertainty on Compliance with Legal Standards, 70 Va. L. Rev. 965, 982 (1984). ×

The incentive to over-invest in auto safety can arise when manufacturers expect courts to set the standard of reasonable care (or a non-defective design) inefficiently high—that is, when manufacturers expect that courts may find a design defect notwithstanding the fact that the automaker’s design decisions were consistent with an accurate, objective, comprehensive risk-utility test. If that is the expectation, then automakers would have an incentive to satisfy the inefficiently high court- or jury-imposed design standard (or warning standard) in order to avoid liability. The incentive to under-invest in safety can arise if courts rely too much on custom within the industry as their source for what constitutes reasonable care, or a non-defective design or warning. This is because industry custom can (famously) lag behind what is truly efficient levels of safety. 144 144. The T.J. Hooper, 60 F.2d 737 (2d Cir. 1937) (“a whole calling may have unduly lagged in the adoption of new and available devices”). In product liability design defect cases, of course, courts do not generally permit compliance with industry custom to be totally exculpatory; however, it can be considered relevant to the risk-utility negligence-based balancing test. See, e.g., Carter v. Massey-Ferguson, Inc., 716 F.2d 344 (5th Cir. 1983). However, if a defendant in a negligence-based product liability regime has adopted a design that is the safest in use at the time of manufacturing, it may be difficult for the plaintiff to prevail. See RESTATEMENT (THIRD) TORTS: PROD. LIAB., §2, cmt. D (AM. LAW. INST. 1998). × It is not a surprise, then, that commentators have argued that custom-based standards of care, like those that currently apply to automaker liability, can inhibit innovation. 145 145. See, e.g., Gideon Parchomovsky & Alex Stein, Tort and Innovation, 107 Mich. L. Rev. 285 (2008). ×

A second problem with a negligence-based auto products liability regime has to do with driver care levels. For a negligence-based regime to efficiently incentivize drivers to drive carefully (by imposing on drivers the risk of accidents that are not cost-justifiably preventable by the manufacturer), recall that we assumed that drivers are well informed of both accident risks and how those risks are allocated according to the specific rules of auto tort law. Those assumptions are obviously unrealistic. Drivers simply are not aware of the tort law rules that apply to them or the product liability rules that apply to automakers. Moreover, even when drivers do know about accident risks and legal rules, there are reasons to believe (discussed below) 146 146. See infra, discussion at notes 44–48 and 51. × either that drivers will not respond rationally to that information or that they will externalize those risks to insurance companies. If I am right about that—about drivers’ lack of information about driving risk and auto tort law, and about their cognitive biases and cost-externalization—then the ability of a negligence-based auto products liability regime to optimize driver care levels is substantially undermined. Legally imposing costs on drivers would not, or at least may not, have the desired deterrence effect on driver care levels. 147 147. By contrast, the assumption that automobile manufacturers—with their teams of expert engineers, lawyers, and accountants—are fully informed of the torts liability regime in which they operate and how those rules are likely to affect them. This claim—that auto manufacturers are likely to be better informed (both about the risks of auto accidents and about the relevant liability rules) and more classically rational in their decision-making than drivers—is not new. See, e.g., Latin, supra note 21, at 692–93 (arguing that, because drivers are much less likely to know about and respond rationally to having auto-accident losses imposed on them than auto manufacturers are, a regime of auto-manufacturer enterprise liability could produce an overall improvement in social welfare, through a reduction in overall auto accidents). ×

The final deterrence problem with a negligence-based auto products liability regime would exist even if judges and juries were good (accurate and unbiased) at applying risk-utility or cost-benefit standards. In fact, this problem results because automakers would expect accurate application of the negligence-based rules. The problem involves the effect of a negligence-based automaker liability rule on the number of vehicles sold, or, in the language of deterrence, the effect on automaker “activity levels.” 148 148. See supra sources cited in note 28. × Even an efficiently safe car (one with no defects whatsoever) that is driven carefully by its human or algorithmic driver poses some residual or irreducible risk of crashing. This residual risk will have a tendency to be ignored or externalized by automakers under a negligence-based product liability regime because automakers can virtually insulate themselves against liability by merely complying with the liability standard. 149 149. See supra sources cited in note 28. Of course, this automaker activity level effect is mitigated to some extent when automakers except to be held liable by courts despite having taken reasonable care. These effects are unlikely, however, to be perfectly offsetting. × The result of this externality is that the scale of operation in the auto industry—the number of cars sold—may be higher than the social-welfare maximizing level, even ignoring the effect of automobile emissions on the environment, because the price of vehicles does not include this cost of unpreventable auto accidents.

To summarize, given how our current negligence-based automaker liability regime is applied in practice, there are reasons to be concerned that automaker and driver care levels may be too low and activity levels too high. What’s more, this concern would apply not only to human-driven vehicles, but to Level 5 vehicles as well. That is, there is nothing about the nature of Level 5 vehicles that would suggest these problems are less likely to be present than would be the case for human-driven vehicles. 150 150. With Level 5 vehicles, if there were a problem with “driver care levels,” it would be a problem automaker care levels. That is, with Level 5s, driver care levels are, by definition, included as a part of manufacturer care level. × This activity-level inefficiency associated with current automaker liability law has been totally ignored by those who have argued in favor of applying existing product liability standards, or revised but still negligence-based versions of existing product liability standards, to Level 5 vehicles.

Driver Liability Law

In a majority of states in the U.S., if someone is injured or suffers property damage as a result of a driver’s negligent operation of an automobile, rather than as a result of automaker negligence, the victim may recover from the negligent driver under standard common-law principles of tort. 151 151. For a summary of emergence of fault-based and no-fault auto liability/insurance systems, see JAMES M. ANDERSON, PAUL HEATON, SEPHEN J. CARROLL, THE U.S. EXPERIENCE WITH NO-FAULT AUTOMOBILE INSURANCE: A RETROSPECTIVE 19–61 (2010), https://www.rand.org/content/dam/rand/pubs/monographs/2010/RAND_MG8 60.pdf. × The victim must demonstrate that the harm to her was a result of the driver’s failure to do something that a reasonable driver would have done under the circumstances, or the drivers’ doing something that a reasonable driver under the circumstances would not have done. 152 152. RESTATEMENT (SECOND) OF TORTS § 284 (defining negligent conduct in terms of what reasonably prudent person would do or not do); see also RESTATEMENT (THIRD) OF TORTS: LIABILITY FOR PHYSICAL AND EMOTIONAL HARM, § 3 cmt. h (Am. Law. Inst. 1998) (“Many cases say that negligence consists of “the failure to do something which a reasonably careful person would do, [or] the doing of something which a reasonably careful person would not do.”“) (citations omitted). × Accident victims who can recover include pedestrians, cyclists, passengers, or other drivers—anyone who is harmed as a result of driver negligence.

Because driver liability is also a negligence-based regime, it has similar potential to provide efficient deterrence as does a negligence-based automaker liability regime. Specifically, negligence-based driver liability law can have beneficial deterrence effects on driver care levels, if we make the following assumptions:

  • Drivers are well informed about accident risks (and how their behavioral changes affect those accident risks),
  • Drivers are well-informed about the rules of tort law,
  • Drivers internalize those risks (do not externalize them to insurers, for example), and
  • Drivers process the information about those risks rationally (without any systematic cognitive biases), and we assume again that
  • Courts are good at applying cost-benefit-type negligence-based liability rules.

If all of those assumptions are true, then, for the same reason that automakers would be incentivized by a negligence-based automaker liability regime, drivers too would be incentivized to drive with efficient care—in terms of driving speed, safe braking and passing practices, smart-phone usage (or non-usage), and the like. This is so because, by taking efficient care in driving, drivers would avoid liability for the accidents that nevertheless occur. Again, under a negligence-based regime, driving with efficient care can be seen as a type of insurance for drivers, a fact that—if all of the above-listed assumptions are true—would incentivize safe driving.

The reasons that this vision of negligence-based driver liability law do not describe reality should be clear at this point. The assumptions listed above on which the analysis depends almost certainly do not hold in the real world. While drivers may be generally aware of the broad outlines of the driver liability regime in their state (whether it is fault-based or no-fault), they likely do not understand what the precise implications of that fact are on their chances of being found liable in court for unsafe driving. What’s more, the average driver, while generally and vaguely cognizant of the risks of driving, is almost certainly uneducated about the precise levels of risk associated with various aspects of driving—for example, precisely how much the chance of a crash is increased by texting while driving or changing lanes abruptly with no signal. In fact, there is a good chance that most drivers underestimate those risks.

Why would drivers tend to underestimate such risks? First, there is the long list of well-documented cognitive biases that affect how individuals process information generally. 153 153. The sources here are many. A decent place to start would be the classic essay by Amos Tversky & Daniel Kahneman, Judgment Under Uncertainty: Heuristics and Biases, in JUDGMENT UNDER UNCERTAINTY. HEURISTICS AND BIASES 3 (Daniel Kahneman, Paul Slovic, & Amos Tversky eds., 1982). More recent summaries of the literature include DANIEL KAHNEMAN, THINKING, FAST AND SLOW (2011); and RICHARD H. THALER, MISBEHAVING: THE MAKING OF BEHAVIORIAL ECONOMICS (2015). A classic article summarizing the application of behavioral insights to law and economics is Christine Jolls, Cass R. Sunstein & Richard H. Thaler, A Behavioral Approach to Law and Economics, 50 STAN. L. REV. 1471 (1998). × One famous example is the tendency of individuals to ignore the risk of very low probability events and underestimate the likelihood of some high probability events. 154 154. For a summary of the relative behavioral literature as it relates to products liability, see Jon D. Hanson & Douglas A. Kysar, Taking Behavioralism Seriously: the Problem of Market Manipulation, 74 N.Y.U. L. REV. 630, 643–87 (1999). For early applications of behavioral insights to products liability law, see Latin, supra note 21; and Howard A. Latin, “Good” Warnings, Bad Products, and Cognitive Limitations, 41 UCLA L. REV. 1193, 1194–95 (1994). For a discussion of the findings on very low probability and high probability events, see Hanson & Kysar, supra note 45, at 716–20. Note that there is also research showing that consumers sometimes overestimate the risks of merely “low probability” events—those that fall between very low probability and high probability. Id. (discussing research summarized in Enterprise Responsibility for Personal Injury, 1 A.L.I. 230 (1991)). × Auto-crash risks may similarly be ignored or underestimated. 155 155. Whether auto risks are more likely to be very low, merely low, or high probability events is not entirely clear. However, the most important insight of the Hanson & Kysar article is that, because product manufacturers—including automakers–have considerable influence over how consumers perceive the risk of their products (and because product manufacturers—including automakers–have a strong market incentive to ensure that consumers underestimate the risks of their products), there is every likelihood that consumers on balance underestimate the risks of auto accidents. Id. Moreover, Hanson & Kysar, in a follow up article, provide considerable anecdotal evidence of actual market manipulation of consumer risk perceptions by manufacturers. Jon D. Hanson & Douglas A. Kysar, Taking Behavioralism Seriously: Some Evidence of Market Manipulation, 112 HARV. L. REV. 1420, 1466 (1999). × Also, drivers are especially prone to overestimating their own driving ability and thus their own ability to avoid crashes. 156 156. This finding has proven robust over many years. See, e.g., Ola Svenson, Are We All Less Risky and More Skillful Than Our Fellow Drivers?, 47 ACTA PSYCHOLOGICA 143 (1981); Timo Lajunen & Heikki Summala, Driving experience, personality, and skill and safety-motive dimensions in drivers’ self-assessments 19 PERSONALITY AND INDIVIDUAL DIFFERENCES. 307 (1995); A.F. Williams, Views of US drivers about driving safety 34 J. SAFETY RES. 491 (2003). See also Jolls et al., supra, at 1537–38 (discussing problem of “overoptimisim” among drivers); Hanson & Kysar, Taking Behavioralism Seriously: A Response to Market Manipulation, 6 ROGER WILLIAMS L. REV. 259, 354–55 (1999) (same). × Moreover, drivers not only underestimate their own likelihood of a crash relative to the average driver (which they do), they also overestimate their own likelihood of a crash relative to the actual probability. 157 157. Christine Jolls, Behavioral Economics Analysis of Redistributive Legal Rules, 51 VAND. L. REV. 1653, 1660 (citing Richard J. Arnould & Henry Grabowski, Auto Safety Regulation: An Analysis of Market Failure, 12 BELL J. ECON. 27, 34–35 (1981) and Colin F. Camerer & Howard Kunreuther, Decision Processes for Low Probability Events: Policy Implications, 8 J. POLY. ANALYSIS & MGMT. 565, 566 (1989)). × For all of these reasons, a negligence-based driver liability regime, which relies on assumptions of informed and rational drivers to produce optimal driver care levels, may not produce the deterrence benefits that are predicted by deterrence theory. 158 158. The evidence on whether negligence-based driver liability law reduces auto accidents, or the harms resulting from auto accidents, is probably best characterized as inconclusive. See generally Nora Freeman Engstrom, An Alternative Explanation of No-Fault’s Demise, 61 DEPAUL L. REV. 303, 332–333 (2012) (“[R]oughly half of the studies published thus far claim that no-fault coverage increases fatal accidents, while the other half find no effect, and the notion that no-fault reduces fatalities has been seemingly put to rest.”) (footnotes omitted). The hope that a shift away from a negligence-based driver-liability regime would not substantially reduce auto accidents, along with the overall desire to lower auto insurance rates, was one of the original justifications for the movement towards auto no-fault regimes in the 1970s and 1980s. Id. × In addition, it is commonly argued that drivers have many powerful incentives to drive carefully even in the absence of a negligence-based regime that left on them the uninsured costs of auto accidents, incentives such as the desire to avoid a traffic fines or, more importantly, a crash that could be painful or even fatal to them or their loved ones. 159 159. Latin, supra note 21, at 690–91; Engstrom, supra note 49, at 330. ×

How does this pessimistic picture of driver liability law as a system of incentivizing good driving change if we introduce auto insurance? The answer to that question turns out to be complicated. On one hand, automobile insurance has the potential to correct some of these deterrence-related problems. 160 160. See generally Omri Ben-Shahar & Kyle D. Logue, Outsourcing Regulation: How Insurance Reduces Moral Hazard, 111 MICH. L. REV. 197 (2012) (discussing ways that insurance companies help insureds reduce risk). × Here’s why. Auto insurers are, unlike most drivers, extremely well informed about the intricacies of accident law. They employ teams of lawyers whose job is to understand how driver liability laws in each state affect the liability risks of their customers. Indeed, their profitability and their survival as going concerns depend on this expert understanding of the auto liability laws of all sorts. In addition, auto insurers have unparalleled access to enormous amounts of detailed information regarding the crash-risk characteristics of millions of drivers and automobiles. This is the result of decades of experience providing auto insurance coverage to hundreds of millions of drivers and vehicles, which in turn means pricing millions of auto insurance policies and adjusting millions of auto-crash claims over the years. No other institution or organization would have the same amount of driver-specific, automobile-specific data, as would the auto insurance industry.

In addition, recent innovations in “telematics” (which combines telecommunications, data science, and automotive technology) have increased auto insurers’ ability to gather and analyze risk-relevant driver and vehicle data. 161 161. Background On: Pay-As-You-Drive Auto Insurance (Telematics), INS. INFO. INST., https://www.iii.org/article/background-on-pay-as-you-drive-auto-insurance-tele matics (last visited Nov. 21, 2018). × With this new and emerging technology, not only do insurers have access to information regarding how drivers’ past auto-claims and traffic-ticket histories affect their riskiness as drivers; they also have the ability to gather information on the effects of a range of specific driving behaviors on auto-crash risks. 162 162. Id. See also Ben-Shahar & Logue, supra note 51. × For example, a number of insurers currently gather information about drivers’ braking, acceleration, speeding, turning, and cornering behaviors and then send that information back to the insurers for analysis. 163 163. Yuanjing Yao, Evolution of Insurance: A Telematics-Based Personal Auto Insurance Study, U. CONN. HONORS SCHOLAR THESES, 590, 598 (2018), https://opencommons.uconn.edu/srhonors_theses/590/. × Once this driver-specific data is combined with data gather by insurers and others (including NHTSA) about what factors cause auto accidents generally, it becomes possible for auto insurers to link specific driving behaviors of particular drivers with premium discounts. 164 164. Id. ×

All of this information is to varying degrees already being taken into account by many auto insurance companies in the pricing of their insurance policies. For example, policy discounts are offered to drivers with good safety records 165 165. Most auto insurers give discounts for being accident free for a given period of time. Car Insurance Discounts, VALUEPENGUIN, https://www.valuepenguin.com/car-insurance-discounts (last visited Nov. 21, 2018). See also Yao, supra note 54 (discussing use of behavioral driving discounts among insurers). × as well as for vehicles with particular safety features. 166 166. One survey of the leading car insurers, found the following additional vehicle-safety-related discounts: passive restraint (25% to 30%), new car (10%), daytime running lights (around 3%). Id. Some insurers are starting to offer discounts for semi-autonomous features such as adaptive cruise control, collision avoidance systems, and lane departure warnings. Cherise Threewitt, What Car Insurance Discounts Can I Get?, U.S. NEWS & WORLD REP. (June 29, 2018), https://cars.usnews.com/cars-trucks/car-insurance/car-insurance-discounts. × In addition, insurers are now offering discounts if drivers will improve their driving ability—for example, if they will take defensive driving classes. 167 167. Car Insurance Discounts, supra note 56 (reporting insurers giving discounts of 10% to 15% for completion of defensive driving courses). × Because of telematics revolution, auto insurers are even able to adjust premiums on the basis of the specific driving behavior of individual drivers. For example, some insurers give discounts for a range of driver-care-level factors such as wearing seatbelts, driving at moderate speeds, limiting late night trips, and avoiding aggressive braking. 168 168. Id. See also Barbara Marquand, Comparing Drivewise, Snapshot and Other Usage-Based Insurance Plans, NERDWALLET (Feb. 8, 2016), https://www. nerdwallet.com/blog/insurance/comparing-drivewise-snapshot-usage-based-insurance/. × Also, the advances in telematics have made “pay as you go” auto insurance, under which premiums are a function of the number of miles driven, more accurate—and thus more prevalent—than ever before. 169 169. Usage-Based Insurance and Telematics, NAT’L ASS’N OF INS. COMMISSIONERS (July 26, 2018), https://www.naic.org/cipr_topics/topic_usage_based_insurance.htm. × Driving-behavior-sensitive auto insurance premiums—which could take into account both good and bad driving choices (i.e., driver care levels) and, critically, the number of miles driven (i.e., driver activity levels)—hold the promise of incentivizing risk-reducing driving behavior in a way that even the most sophisticated government regulator could not hope to do. 170 170. See generally Ben-Shahar & Logue, supra note 51 (discussing potential risk reducing benefits of high-tech auto-insurance pricing); Hanson & Logue, supra note 19, at 192–93 (suggesting reasons why auto insurance is better at risk-segregating than other types of first-party insurance). What little empirical research has been done on the subject tends to confirm that incentive-based insurance pricing tends to alter driving behavior in a risk-reducing direction. See Mark Stevenson et al., The effects of feedback and incentive-based insurance on driving behaviours: study approach and protocols, 24 INJ. PREVENTION, 89, 93 n. 27–30 (2018), https://injuryprevention.bmj.com/content/24/1/89; see also Telematics Helps Reduce collisions and Claims, AUTOMOTIVE FLEET (Nov. 10, 2017), https://www.automotive-fleet.com/157806/telematics-helps-reduce-collisions-and-claims. ×

But here is the problem: Under current law and given existing market conditions, auto insurers do not have strong incentives to make full use of their comparative advantage at gathering risk-relevant information and pricing their insurance on the basis of that information, or at least there is reason to be concerned about their incentives to do so. The reason for concern is that the amount of coverage currently being provided by auto insurers presently represents only a fraction (in many cases a small fraction) of the total risks of auto crashes. This is true of first-party auto insurance coverage, which tends to cover only a fraction of the accident risks that any driver faces. 171 171. This assertion requires some explanation. There are no good studies on this particular question. So my claim is derived from circumstantial evidence of a sort. First, note that, whereas all states require some amount of liability insurance coverage for anyone who drives on public roadways, only a small minority of states require drivers to purchase auto insurance that provides any sort of first-party medical or disability benefits. See, e.g., Background on: Compulsory Auto/Uninsured Motorists, INS. INFO. INST. (April 16, 2018), https://www.iii.org/article/background-on-compulsory-auto-uninsured-motorists. What’s more, even when there is mandated medical or disability coverage, the amount of required coverage is almost always far less than would potentially be recoverable under an auto tort claim, whether it be an automaker liability claim or a driver liability claim, and far less than the potentially enormous total costs (in the millions) of any given auto accident. Id. (for example, noting mandatory amounts of bodily injury liability coverage ranging from $15,000 to $50,000 per person). Karl Eisenhower, Personal Injury Protection: How PIP Insurance Works in Your State, WALLET HUB (Jan. 9, 2015), https://wallethub.com/edu/pip-insurance/9248/ (noting PIP mandates ranging from $2000 per person in Utah to $50,000 in New York.) The only state that requires unlimited PIP coverage is Michigan. Id. Further, even when drivers do purchase first-party medical or disability coverage through their auto-insurance policy, that coverage is often secondary to the victims’ other forms of first-party health or disability insurance. For example, in some states requiring PIP coverage in auto policies, the insured can elect to make auto PIP coverage secondary to other first-party health and disability insurance. This is sometimes called the “coordination” option. MICH. DEPART. INS. & FIN. SERV., YOUR GUIDE TO AUTOMOBILE INSURANCE: FOR MICHIGAN CONSUMERS 10 (Sept. 2017), https://www.michigan.gov/documents/ difs/Auto_Insurance_Guide_448003_7.pdf. Because making auto health insurance secondary lowers the insured’s auto insurance premiums, and has little or no effect on her first-party health and disability insurance premiums, most insureds choose the coordination option, which means most insureds choose to make their non-auto first-party insurers primary. In sum, most auto health and disability risks end up being borne by non-auto first-party insurers—such as health insurers and disability insurers. × It is also true of auto liability coverage, owing in part to the fact that the mandatory minimum amounts in most states are far less than the maximum harm threatened by an auto accident that results in even one serious injury or death. 172 172. State mandated minimums for liability coverage for personal injuries to a single person range from a low of $10,000 (Florida) to a high of $50,000 (Alaska and Maine), and by for the most common minimum is $25,000. Car Insurance Laws by State, FINDLAW, https://injury.findlaw.com/car-accidents/car-insurance-laws-by-state.html (last visited Nov. 19, 2018) (gathering links to state laws). A single accident resulting in serious bodily injury or death could easily produce economic losses alone in excess of $1 million. × As a result, many of the costs of auto accidents are currently being externalized to non-auto first-party health and disability insurers who—unlike auto insurers in the telematics age—do not tailor premiums at all based on their insureds’ driving decisions. 173 173. Most first-party health and disability insurers make no effort to price their coverage in a way that reflects the riskiness of insureds’ driving choices—such as how they drive, how much they drive, or even what type of vehicle they drive. See generally Hanson & Logue, supra note 19 (using fact that most non-auto first-party insurers do not price-differentiate on basis of consumer product use to argue for enterprise liability for product accident risks). There is a perfectly sensible reason for this fact: the risks of auto-related health or disability claims are only a small fraction of the overall health and disability risks covered by any given first-party health or disability insurer. It is not worth the insurers’ while to tailor their insurance premiums on the basis of any particular behavioral choice of their insureds, other than perhaps the choice to smoke or not. The result of all this: that portion of auto crash risks that are ultimately born by non-auto first-party insurers get externalized (or largely ignored) by drivers, with obvious deterrence consequences. × Moreover, to the extent auto insurers do attempt to charge individualized, behaviorally- and risk-adjusted auto insurance rates (which, as I noted above, they are increasingly trying to do), this incentive is undermined by the fact that auto insurers cover only a fraction of the risks of auto accidents. 174 174. Because auto insurers do not bear all of the risks of auto accidents, the premium discounts they are willing to offer to induce safer driving habits may not be adequate. The point can be illustrated with a simple example. Suppose there was some investment in driver care that cost $50 but would reduce expected accident costs by $80. Say it would reduce a chance of a $200,000 loss from .001 to .0006. If the auto insurer bore the full $200,000 risk, it would have an incentive to offer a premium discount to cover the cost of driver care, with an additional discount perhaps up to a total just short of the $80 saved by the additional investment in driver care. But what if the auto-insurer bore only, say, $40,000 of the $200,000 potential loss? Then the largest discount it could offer without losing money would be $16, which would be the amount of the savings in going from a .001 to a .0006 risk of, now, $40,000 in covered losses. But that discount would not be enough to induce the consumer to make the investment in care, assuming the other $160,000 in expected accident cost is externalized either because of drivers’ under-estimation of risk or because of non-adjusting non-auto first-party insurance coverage. ×

It should also be noted, however, that there are important ways in which the allocation of auto-accident risks to non-auto first-party insurers has cost-reducing advantages. This may seem incongruous with the argument in the previous paragraph, but it is not: While auto insurers are in a good position, through premium discounts, to help optimize driver care and activity levels, auto insurers are not necessarily in a good position to minimize some other costs associated with providing insurance benefits. For example, primary health care coverage provided through auto insurance companies is almost certainly much more expensive than primary health care provided through regular non-auto first-party health insurers. This would be because, although auto insurers, in a sense, specialize increasingly in reducing driver ex ante moral hazard, it is non-auto health insurance companies who specialize in reducing ex post medical moral hazard—that is, excessive or wasteful use of the healthcare system. 175 175. The claim that first-party health insurers specialize in trying to hold down health care costs may seem controversial, at least for fee-for-service policies. My claim is only that health insurers—especially ones that use managed care tools—are probably better at holding down ex post health care costs than are auto insurers. This is one of the reasons that auto no-fault regimes which make auto PIP coverage primary over non-auto first-party health coverage are so expensive, and why auto-no-fault regimes have not led to the cost savings that were expected. See Engstrom supra note 49. Professor Engstrom notes that “[m]edical insurers . . . reduce costs via discounts and fee schedules, and the limit patient treatment using any number of mechanism, including deductibles, co-payments, utilization controls, and medical protocols . . . , [a]uto insurers . . . tend to pay almost any bills that a victim incurs . . . dramatically increasing . . . the cost of care. Id. at 341 (citations omitted). × My point here is only that the current division of auto-accident costs, allocating so little to auto insurers, may be non-optimal, given auto insurers potential ability to incentivize better (and less) driving. 176 176. In Part III below, I suggest that, by adopting an automaker enterprise liability regime, automakers will have an incentive to strike the efficient balance between amount of auto-crash costs allocated to auto insurers and amount allocated to non-auto insurers. ×

To summarize, because of drivers’ lack of accident-risk information and understanding of auto tort law and their susceptibility to cognitive biases, and because of the presence of cost-externalizing insurance coverage, there is reason to be doubtful that the current negligence-based auto tort laws—automaker liability laws as well as driver liability laws—work to optimize driver care and activity levels. As discussed in the next Part, the adoption of an auto enterprise liability regime could in theory create incentives for automakers, together with auto insurers, to provide better driver-side incentives, as well as better automaker safety incentives.

III.         The Automaker Enterprise Liability Alternative 177 177. The arguments in this section draw on prior work done by me and a number of other scholars on the deterrence benefits of enterprise liability in various contexts. See, e.g., Hanson & Logue, supra note 19; Steve P. Croley & Jon D. Hanson, Rescuing the Revolution: The Revived Case for Enterprise Liability, 91 MICH. L. REV. 683 (1993); Kyle D. Logue & Jon D. Hanson, The Costs of Cigarettes: The Economic Case for Ex Post Incentive-Based Regulation, 107 YALE L.J. 1163 (1998); and Jon D. Hanson & Douglas A. Kysar, Taking Behavioralism Seriously: Some Evidence of Market Manipulation, 112 YALE L.J. 1420, 1553 (1999). In addition, as mentioned in an earlier footnote, the argument here has some overlap with a proposal made by Howard Latin. See generally Latin, supra note 21 (making a deterrence case of automaker enterprise liability). ×

The Basic Proposal

As an alternative to our current negligence-based auto tort regime, consider the possibility of a comprehensive automaker enterprise liability regime. Under such a regime, anyone who suffers a physical injury or property damage in an automobile accident—whether driver, passenger, or pedestrian—would be legally entitled to recover, from the manufacturer of the vehicle involved, compensation for the losses sustained as result of the accident. 178 178. This could be done through the existing court system or through specialized courts or agencies set up to handle auto-crash disputes. × Thus, to recover under this enterprise liability regime, accident victims would not be required to show negligence on the part of manufacturer or anyone else. Nor would accident victims have to prove that the automobiles, or any of the warnings or instructions accompanying the automobiles, are in anyway defective or unreasonably dangerous. Rather, crash victims would need only to prove that the harms for which they seek compensation “arose out of the use of” a vehicle that was designed and built by the manufacturer from whom compensation is sought. Each automaker, therefore, would be financially responsible for the losses resulting from any crash arising out of the use of that automaker’s vehicles. 179 179. The “arising out of the use of” analysis would replace a causation determination. This phrase is used now in standard auto insurance policies. Thus, an automaker enterprise liability regime would be a particular type of cause-based no-fault compensation regime, modeled after similar programs that have been adopted outside of the auto context, such as workers’ compensation laws at the state level or the vaccine compensation program at the federal level. See generally Jon D. Hanson, Kyle D. Logue & Michael S. Zamore, Smokers’ Compensation: Toward a Blueprint for Federal Regulation of Cigarette Manufacturers, 22 S. ILL. L.J. 519 (1998) (discussing deterrence benefits of cause-based no-fault compensation regimes). In the Workers’ Compensation context, claims against employers are limited to injuries or illness that “arise out of the course of employment.” Professors Abraham and Rabin have proposed a similar regime—that would also use the “arising out of” standard—but that would apply exclusively to accidents involving Level 5 automated vehicles. Rabin & Abraham, supra note 20. My idea is to make such a regime comprehensive, to apply to all motor vehicles, subject to transition rules discussed below. Some scholars have expressed concern about the use of enterprise liability regimes that rely on boundary-maintaining doctrines such as the “arising out of” concept used in workers’ compensation regimes, among other places. See, e.g., Henderson, supra note 19. This is of course a reasonable concern, although the evidence suggests that programs such as workers’ compensation have found relatively effective ways to police the borders of their programs. DON DEWEES, DAVID DUFF & MICHAEL TREBILCOCK, EXPLORING THE DOMAIN OF ACCIDENT LAW: TAKING THE FACTS SERIOUSLY 393-394 (1996) (reporting administrative costs for workers’ compensation regimes that are low relative to those of the tort system). ×

That is the most basic picture of the proposal. Now consider a few possible details of such a program. One important initial question is who exactly would fall within the class of “automakers” to whom the enterprise liability regime would apply. The most obvious class of defendants/payers would be the original equipment manufacturers (OEMs) of the vehicles involved in the crash. They are the ones who generally make the key automotive design choices, have control over the manufacturing processes, and decide on the terms of any warning or instruction manual; and they are also the ones with the greatest expertise on such questions. Auto manufacturers also determine the pricing of their vehicles and the number of them to produce, subject of course to the constraints of supply and demand. Given that manufacturer care levels and activity levels are key auto-accident deterrence variables, making OEMs responsible for the auto-crash costs associated with their vehicles has obvious deterrence benefits, discussed further below.

Liability under an enterprise liability regime, however, would not necessarily be limited to auto manufacturers. Liability could also be extended, on a joint and several basis (or on a several basis), to a range of other enterprises that fall within the design, production, sale, and distribution chain of any given vehicle. 180 180. Thus the concept of “automaker” in an automaker enterprise liability regime could be similar to the concept of a “seller” in existing products liability law. × In most cases, it is likely that the crash victim would bring the claim against the manufacturer, and then the manufacturer would either implead the other parties in the chain of production into that suit or would sue them separately in a contribution action. Precisely how the responsibility for the costs of any accident would be allocated among the various parties on the automaker side of the ledger is beyond the scope of this Article. That allocation of responsibility, however, would presumably be determined mostly by contracts among the counter-parties, which contracts should be enforced so long as the cost of auto accidents is not allocated to parties who are insolvent or judgment proof, which if permitted would undermine the deterrence benefits of the regime. 181 181. See Steven Shavell, The Judgment Proof Problem, 6 INT’L REV. OF L. & ECON. 45 (1986) (explaining how presence of insolvent defendants undermines incentive effects of liability law). ×

The types and amount of compensation recoverable under an automaker enterprise liability regime would probably be limited to economic losses—medical expenses, lost income, and property damage. There is of course a deterrence argument for including noneconomic or pain-and-suffering damages as well, since failing to include noneconomic damages could produce a serious externality. 182 182. Noneconomic damages are generally not covered by first-party insurance policies, which means, insofar as drivers (and consumers generally), because of the cognitive biases already discussed, ignore or underestimate the risks of auto accidents, they will externalize noneconomic damages as well. Including noneconomic damages, therefore, has the potential to improve care levels and activity levels. See Hanson & Logue, supra note 19, at 186–89 (describing the “unambiguous deterrence benefits of nonpecuniary-loss damages”). × However, some have argued that individuals do not desire to purchase insurance against non-economic losses (as evidenced by the dearth of pain-and-suffering insurance observed in the marketplace), and therefore should not be forced to purchase such coverage through a mandatory compensation regime. 183 183. Examples of this sort of argument can be found in George L. Priest, The Current Insurance Crisis and Modern Tort Law, 96 YALE L.J. 1521, 1546–47; and Alan Schwartz, Proposals for Products Liability Reform: A Theoretical Synthesis, 97 YALE L.J. 353, 362–67 (1988). For a powerful set of counter arguments, providing arguments why consumers might—and evidence that they in fact do—demand insurance for noneconomic losses, see Steven P. Croley & Jon D. Hanson, The Nonpecuniary Costs of Accidents: Pain-and-Suffering Damages in Tort Law, 108 HARV. L. REV. 1785 (1995). × In any event, limiting compensation to economic losses, and thus not providing compensation for noneconomic harms, is a common and reasonable political compromise that is often made when no-fault cause-based compensation regimes are adopted. 184 184. See generally Hanson, Logue & Zamore, supra note 66, at 556–62 (reviewing arguments for limiting damages in no-fault cause based compensation regimes to economic damages). ×

It is worth emphasizing again that the compensation regime I am imagining is a comprehensive automaker enterprise liability regime. In other words, it would apply to all automobiles (sold after the effective date of the enacting legislation), whether driven by humans, computer algorithms, or any combination of the two. Thus, unlike some other proposals for manufacturer-funded vehicle compensation regimes, my proposal would not apply exclusively to Level 5 vehicles. 185 185. For an interesting proposal to create a special enterprise liability-type regime, similar to the one I am describing, but that would be limited to accidents involving automated vehicles, see Abraham & Rabin, supra note 20. × Which is not to say that the regime would not have special rules for autonomous and connected vehicles. For example, whereas Level 5s may be in fewer accidents, or fewer accidents involving serious physical injuries or deaths (that’s the hope anyway), Level 5 vehicle accidents may involve much higher auto-repair costs than accidents involving human-driven vehicles, because of the expense of repairing high-tech sensors as well as computer hardware and software. 186 186. Because of the higher repair costs, some in the auto insurance industry have proposed mandatory minimum auto repair coverage for self-driving vehicles. TRAVELERS INSTITUTE, INSURING AUTONOMY: HOW AUTO INSURANCE CAN ADAPT TO CHANGING RISKS 11 (2018), https://www.travelers.com/iw-documents/travelers-institute/Final-Digital-2018-0710-AV-White-Paper-No-SAE.pdf. Under enterprise liability, of course, there would indeed be mandatory minimum auto repair coverage, as well as mandatory minimum personal injury coverage, but the coverage mandate would be imposed on automakers instead of auto purchasers. ×

If an automaker enterprise liability regime were adopted, there would be no need for either the existing automaker liability laws (i.e., products liability as applied to automobiles), driver liability laws, or state auto no-fault laws. All of those auto tort regimes would be replaced by a single comprehensive automaker enterprise liability regime. 187 187. Tort liability for non-economic and potentially punitive damages could be retained for especially egregious behavior, such as recklessness or intentionally harmful actions, on the part of defendants. × Further, if a motor-vehicle crash were to involve two or more vehicles made by different auto manufacturers, the enterprise liability regime would handle the accident as follows: First, the victims would simply file claims for their covered economic losses, naming the automakers of all of the vehicles involved in the accident. After a factual determination was made of whether in fact all of the named vehicles contributed in some way to the accident, the victims’ crash costs would be split between or among the automakers (or the auto insurers covering the losses for each automaker). The split among the automakers could either be equal (each vehicle manufacturer bearing its pro rata share of the crash costs) or according to any other reasonable allocation formula that the industry agrees upon. 188 188. The deterrence benefit for automaker and driver care and activity levels would largely remain, without the need for individualized fault determinations in multi-vehicle crashes. The crash data gathered from all the payouts under the program would provide information as to which types of vehicles and which drivers tended to be in accidents, how much damage resulted from those accidents, and under what circumstances. This information would be combined with the data gathered by automakers and auto insurers regarding individual vehicle and driver behavior in contexts not involving accidents. There would be little additional deterrence benefit to investing in the costly judicial determination of which driver, if any, was at fault or which vehicle, if any, was defective. ×

One result of the adoption of a comprehensive automaker enterprise liability regime would be an increase in the apparent (and the experienced or internalized) price of most newly purchased automobiles, relative to vehicles purchased before the effective date of the enacting legislation. 189 189. This assumes that the new law would have a grandfather provision exempting vehicles built and sold before new law’s effective date. As discussed in the conclusion, such complete grandfathering is not the only conceivable approach to handling the transition to the new regime. × This would happen because the cost of auto accidents that had been hidden in non-auto first-party insurance coverage prior to the enterprise liability regime would, with the adoption of the new system, be brought into the open through increases in automobile and auto-insurance prices. Because such a shift would be a significant change in the automotive marketplace, it would probably be prudent (and politically necessary) to institute a delayed effective date and/or an extended phase-in period over which the law would take effect. 190 190. If there is indeed deterrence value to shifting these costs from non-auto first-party insurers to automakers, as argued in the next section, then the overall price of vehicles (including the costs covered by various forms of first-party non-auto-specific insurance) should eventually go down, especially if the pace of the transition to driverless technology is hastened. Indeed, a significant result of adopting a comprehensive automaker enterprise liability regime is that Level 5 vehicles, if they provide as big an advance in safety as many are expecting, would be substantially less expensive overall than conventional human-driven vehicles. Cf. Bryant Walker Smith, Automated Driving and Product Liability, 2017 MICH. ST. L. REV. 1 (2017) (discussing the potentially massive reduction in overall vehicle prices resulting from the shift to automated vehicles under existing product liability law). A bigger concern with the price increase is the effect on low-income drivers. I address this concern briefly in Part IV below, though the cost of mobility generally is a topic worthy of special consideration. ×

The Theoretical Deterrence Benefits

Under a comprehensive automaker enterprise liability regime, because automakers would be responsible for all of the economic costs of auto accidents associated with their vehicles, they would be forced to internalize those costs. As a result, there would be beneficial deterrence consequences for automaker, and potentially driver, care and activity levels. This section explores those consequences.

First, automakers would have a strong legal and financial incentive to develop and implement cost-justified auto-safety innovations, whatever those might be. That is, if an automaker determined that there was some new brake design (such as a new computer-assisted automatic braking system) or some new guided cruise control mechanism that would reduce overall accident costs relative to its costs of development and implementation, then enterprise liability would reward them implementing those innovations, and punish them for not doing so. What’s more, there would be no inefficient incentive to stick with existing industry customs or consumer expectations if such customs or expectations were lagging behind proven safety innovations. Likewise, there would be no incentive to over-invest in safety features that are likely to impress a court or jury in a negligence-based lawsuit (such as a design defect lawsuit) but that, in actuality, provide less additional accident-risk reduction than they cost to produce. 191 191. Obviously, automakers already have some incentives to develop such safety technology, in part because of consumer tastes for safer vehicles and perhaps because of the threat of potential liability under existing tort law. See, e.g., Press Release, NHTSA, NHTSA-IIHS Announcement on AEB, (Dec. 21, 2017), https://www.nhtsa.gov/press-releases/nhtsa-iihs-announcement-aeb (“Twenty automakers pledged to voluntarily equip virtually all new passenger vehicles by September 1, 2022, with a low-speed AEB system that includes forward collision warning technology proven to help prevent and mitigate front-to-rear crashes.”). Consistent with this fact, it is common knowledge that safety innovation has been happening for decades without the presence of an automaker enterprise liability regime. My argument is that, according to a nuanced application of standard deterrence theory, safety-enhancing automotive innovations would be more likely to be adopted and would be adopted more quickly under an automaker enterprise liability regime. See, e.g., Latin, supra note 21, at 691 (making similar argument with respect to airbags, suggesting that adoption of automaker enterprise liability would have hastened the industry’s move to include airbags). ×

Second, enterprise liability would force the price of automobiles to reflect the full-expected costs of auto accidents. That cost-internalization, in turn, could result in a scale of automotive manufacturing and sales that would be closer to the social optimum than is currently the case, as drivers would—in deciding whether to purchase a vehicle—be more likely to take into account something closer to the full social costs of that decision. In other words, auto enterprise liability could push us in the direction of optimal manufacturer activity levels: the optimal number of vehicles being sold. If that were to happen, it would be a clear improvement—in terms of overall efficiency—over the existing negligence-based automaker liability regime.

It is worth pausing here to emphasize the potential effects of a comprehensive automaker enterprise liability regime on the development of and transition to Level 5 vehicles. Because it to would be a comprehensive regime, it would apply to both driverless and human-driven vehicles. Assuming automakers expect Level 5s to bring a dramatic reduction in expected accident costs relative to human-driven vehicles, then Level 5s, when they eventually are available for sale to consumers, would have a substantially lower enterprise liability “tax” relative to human-driven vehicles made and produced after the new regime is adopted, since the human-drive vehicles would have much higher expected accident costs. 192 192. For now, I am assuming that “old” vehicles, produced before the enactment of the new regime, would be totally grandfathered and thus exempt from the new law. I return to this assumption in the conclusion. × As a result, there would be a natural enterprise liability subsidy in favor of the production of Level 5 vehicles; and this subsidy, in effect, would be funded by a relatively high enterprise liability tax on human-driven vehicles, again, assuming such vehicles are not nearly as safe as Level 5s. Thus, the adoption of a comprehensive automaker liability regime would, under present assumptions, strongly incentivize and reward auto manufacturers to proceed, as quickly as is feasible, with the development and distribution of Level 5 vehicles. By contrast, if a special liability regime were adopted just for Level 5s, that increased their potential accident liability relative to human-driven vehicles, there would be a disincentive to move to Level 5s in the absence of a separate subsidy regime, perhaps funded by federal income taxes. 193 193. Abraham and Rabin make such a proposal, including the need for a subsidy for Level 5 vehicles. Abraham & Rabin, supra note 20, at 45. ×

There are efficiency reasons to prefer a Level 5 vehicle subsidy that is funded through an enterprise liability tax on auto sales, with the amount of the cross-subsidy depending on the relative risk of vehicles (and drivers), over a subsidy funded by federal income taxes. The main advantage has to do with information. Under the direct subsidy, the regulator—or whatever government body would be asked to determine the amount and structure of the subsidy—would have to determine which particular safety technologies to subsidize and which not to subsidize and how much the subsidy should be. This would require an enormous amount of information and expertise that is not within the government’s comparative advantage relative to the auto industry. By contrast, under the subsidy structure inherent in a comprehensive enterprise liability regime, it is the auto industry who would calculate the appropriate amount of the subsidy ex ante, based on their educated guesses about (a) the amount of costs to be imposed on them under the regime for accidents involving human-driven vehicles, (b) the amount of costs that would be imposed on them if they make the investment necessary to develop and implement Level 5 vehicles, and (c) the R&D, design, manufacturing, marketing, training, and other costs that would be necessary to get Level 5s fully up and running. 194 194. It is at least possible that the expected accident costs associated with Level 5 vehicles will not be, overall, a lot less than that of human-driven vehicles, once the cost of repairing the vehicles is taken into account. That is, while Level 5s are expected to reduce frequency collisions and the number of auto-related deaths and serious bodily injuries, they may result in increased repair costs, due to the expense of repairing or replacing the damaged technology in a Level 5 vehicle. In that sense, we might be trading one sort of cost for another, which of course can be socially desirable. This fact, however, might counsel in favor of including at least some noneconomic (or pain-and-suffering) damages in the enterprise liability regime, to make sure that such costs get included in the enterprise liability subsidy for Level 5 vehicles. ×

If an enterprise liability regime is likely to have deterrence benefits on the automaker side, what about its deterrence effects on driver behavior? How an auto-enterprise liability regime would affect the driving behavior of human drivers is of course an especially important question, given that, with non-Level 5 vehicles, human drivers make most of the important operational decisions. In fact, enterprise liability could actually help with driver care and activity levels in a number of ways. First, enterprise liability would create strong legal and financial incentives for automakers to develop and adopt the most cost-effective ways of warning drivers about crash risks and of instructing drivers about how best to avoid certain types of accidents. 195 195. It is a standard conclusion of deterrence theory that enterprise liability would provide strong incentives for manufacturers to develop effective warnings. Croley & Hanson, supra note 68, at 786–792. × This effect flows from the fact that enterprise liability makes automakers’ responsible for all the economic costs of their vehicles’ accidents: If an automaker could actually reduce the frequency or severity of accidents in its vehicles by altering the wording, design, or placement of warnings or instructions, it would have an incentive to do so. On the other hand, if some new or revised warning would be more likely to confuse or annoy drivers than to educate them, the automaker would be incentivized under enterprise liability not to add that sort of unhelpful warning—even if it would have gotten the automaker “off the hook” under a more traditional negligence-based warning-defect standard. Automakers would do whatever works best to reduce accident costs, which would redound to their benefit as reduced auto-accident claim payouts over time.

In addition, enterprise liability could incentivize automakers to restructure the ways that automobiles are insured and sold in order to improve driver care and activity levels. First, consider how an enterprise liability regime might affect how auto insurance is provided. Note that under an enterprise liability regime automakers would have an incentive to shift contractually much of the expected costs of auto accidents to auto insurers. This somewhat counterintuitive result flows from the fact that auto insurers’ have a comparative advantage with respect to monitoring and regulating driver care- and activity-levels. If automakers could get auto insurers to take on somewhat more of the risk of auto accidents, the insurers would have a strong incentive to help drivers reduce expected accident costs. That is, because of competition for customers in the insurance industry, auto insurers would be incentivized to use the tools at their disposal—including individualized, driving-behaviorally-sensitive, risk-adjusted insurance premiums—in ways that would tend to encourage better driving habits and perhaps less driving, especially by high risk drivers. 196 196. Ben-Shahar & Logue, supra note 51, at 220–223. ×

What does this mean for how auto insurance would be sold? Auto insurance under an enterprise liability regime might be sold in the same way that it is today. An individual auto purchaser, in other words, might pay the automaker one price for the vehicle itself and then purchase a separate auto insurance policy at the same time from a separate auto insurance company. However, given that automakers would be ultimately responsible legally (through the doctrine of subrogation) for the auto-accident losses paid by the auto insurers, there would be strong incentives for contractual coordination between automakers and auto insurers. Individual auto manufacturers might even be induced to partner with particular auto insurers in an effort to offer the best, most competitively priced, combined product of vehicle and vehicle-insurance coverage. 197 197. Why are automakers and auto insurers not incentivized now, without the adoption of enterprise liability, to coordinate in the way described in the text? This is a fair question, one that I have put to representatives of both industries and to which I have yet to get a good answer. I suspect that under current legal rules and market conditions, those incentives are dampened. Automakers can largely avoid liability by complying with the largely negligence-based product liability rules, and auto insurers make profits from insuring the residual accident risk. Neither industry—car makers or car insurers—are being forced to bear the full losses of auto accidents. In addition, because much of the risk of auto accidents are externalized by drivers to their non-auto first-party insurers, as discussed above, there is little demand-side incentive for either industry to coordinate with the other. ×

Another way that enterprise liability could improve driver care and activity levels is through its effect on how automobiles are sold. For example, the introduction of an enterprise liability regime might push the automotive industry in the direction of lease transactions rather than outright sales. This is because leasing would make it easier for automakers to enforce the terms of the auto insurance policies, which again might be sold by an insurer who was contractually partnered with the automaker. Under a lease arrangement, for example, if a driver became uninsurable (because of bad driving behavior and/or increased claim payouts), or if the driver simply stopped paying her premiums, there might be a provision in the lease empowering the automaker to reclaim the vehicle.

In addition to favoring leasehold arrangements, the introduction of enterprise liability might create market pressure on auto manufacturers to sell vehicles to commercial purchasers rather than individual consumers. That is, automakers under enterprise liability might be incentivized to sell to commercial entities—fleet operators—who would agree contractually to indemnify the manufacturer for any enterprise liability payments made to victims harmed by vehicles in their fleets. These commercial purchasers, in turn, would either lease the vehicles to individual drivers or perhaps make them available through ride-share arrangements. Automakers in turn could be incentivized to choose commercial purchasers who are financially responsible and would be incentivized to purchase efficient auto insurance contracts to cover the enterprise liability payouts. Such a trend toward commercial fleets would be consistent with already existing market trends towards ride-sharing companies, which trends are expected to accelerate with the advent of Level 5 vehicles. 198 198. Andrew G. Simpson, If They Try It, They’ll Like It: How Ridesharing, Autonomous Cars Will Win Over the Public, INSURANCE JOURNAL (Jan. 23, 2018), https://www.insurancejournal.com/news/national/2018/01/23/478073.htm. ×

I am not suggesting that comprehensive automaker enterprise liability would necessarily result in auto lease arrangements replacing individual sales, or ride sharing replacing driving. Rather, the point is that, once automakers are made legally responsible for the cost of auto accidents (or for most of those costs), they will have an incentive (and the ability) to structure automobile distribution markets in ways that are closer the social optimal.

IV.         Caveats, Concerns, and Conclusions

The description I have given here of an automaker enterprise liability regime is necessarily only a rough outline of an idea, a jumping off point for further discussion. The actual design of such a program would require empirical research into a range of topics, including whether shifting to enterprise liability would actually, and not just theoretically, produce substantial deterrence benefits. Among the other questions that would need to be answered include the following:

Under any real-world version of an automaker enterprise liability regime, there is the question of how long the automakers’ responsibility for insuring their vehicles would remain in effect. Would it be for the useful life of the vehicle or for some set period of time, say, ten years? If for some set period of time, who then would be responsible for covering the accidents arising out of the use of the vehicle? Also, what would the precise relationship be between an automaker enterprise liability regime and state mandatory insurance/financial responsibility laws? Presumably, rescission of coverage by the insurer due to excessive accident experience or failure to pay premiums would result in a suspension of driving privileges, but how would that be enforced? All good questions.

Similarly, if an auto enterprise liability regime were adopted, would it in fact have a grandfather provision perhaps exempting all vehicles manufactured and sold before a given date, as suggested above? Or would older vehicles made before the new law goes into effect be transitioned into the new regime over time? If older vehicles were fully exempted from (or grandfathered out of) the new regime, how would we deal with the resulting, potentially large, price differential between new vehicles (which would be priced with full accident costs internalized into the purchase price) and used vehicles (which would not be)? What role could increased mandatory minimum levels of auto insurance play in assisting with that transition? 199 199. An alternative to a comprehensive automaker enterprise liability regime would be simply to increase auto insurance mandates to provide coverage closer to what would be provided under an enterprise liability regime. Such a regime could be made comprehensive, in the sense that it would apply both to human-driven and computer-driven vehicles, with perhaps higher repair cost minimum coverage for Level 5 vehicles. A comprehensive mandatory auto insurance regime could also replace all of existing automaker liability and driver liability law in much the same way as I propose for auto enterprise liability and would also likely result in contractual coordination between the automakers and the auto insurers to provide the best combination of auto safety and auto-crash risk coverage. A full discussion of this auto-insurance-mandate alternative to automaker enterprise liability is of course beyond the scope of this Article. ×

There is also a whole range of question regarding how an automaker enterprise liability regime would deal with the threat of auto crashes (or stolen or destroyed data) resulting from criminal hacking of a connected system. Existing and growing markets in cyber insurance coverage might be able to handle the risks of posed to data stored in the vehicles, but the market may have more difficulty covering cyber risks to life, limb, and property. 200 200. Most cyber polices include exclusions for physical damages to persons or property. John Buchanan, Dustin Cho, and Patrick Rawsthorne, When Things Get Hacked: Coverage for Cyber Physical Risks, 2018 A.B.A. LITIG. SEC., INS. COVERAGE LITIG. COMM. HOT TOPICS FOR ICLC’S 40TH—THE COVERAGE BATTLES OF 2028 TUCSON, ARIZONA—MARCH 3, 2018, 2. × Solutions range from expanding the role of the federal government as a reinsurer of last resort to limiting liability for cyber-related physical risks to the amount of mandatory liability insurance coverage. All of these details, and many others, would need to be addressed before any comprehensive automaker enterprise liability regime could seriously be considered.

The final concern raised by the idea of an automaker enterprise liability regime involves the cost. The concern is not that the “experienced” price of autos would rise, although that would certainly be true in the short run. As already noted, such a price increase would be the source of much of the deterrence benefit of an enterprise liability regime, the mechanism through which deterrence would work, incentives for accident-avoidance optimized. Rather, the concern has to do with the problem of affordability. For some households, owning an automobile is already unaffordable, which is a source of hardship and an obstacle to social mobility. For those households, a program that raised the price of autos, even in an effort to make them safer, may not be a welcome change without some form of compensating subsidy. My own view is that some type of taxpayer funded transportation subsidies for the low-income drivers may indeed be desirable (from a social justice perspective), whether or not an automaker enterprise liability regime were adopted. But that topic too must await another day.


   Douglas A. Kahn Collegiate Professor of Law, University of Michigan Law School. Professor Logue received helpful suggestions on the ideas in this Article from numerous participants at a conference on Traffic Accident Liability and the Future of Autonomous Vehicles held at Wake Forest University School of Law and from his Michigan Law colleagues at a Fawley Lunch Workshop in Ann Arbor. Thanks also to Bryant Walker Smith for his comments and notes.

 

Tesla’s enthusiastic marketing of its Autopilot feature may be landing the company in legal hot water. Last week, a Florida man sued the car manufacturer after his Model S crashed into a stalled vehicle at high speed. The driver, who allegedly suffered spinal and brain injuries, claims that Tesla’s “purposefully manipulative sales pitch” had duped him and other Tesla owners into the mistaken belief that their vehicles can travel on the highway almost without supervision. The outcome of the case may carry key lessons not only for Tesla, but for all automakers as they develop more autonomous features.

This isn’t the first time Tesla has faced legal challenges related to the Autopilot feature. In May, the company paid $5 million to settle a class action suit claiming its Autopilot 2.0 upgrade was unusable and dangerous. This case, while currently only involving one plaintiff, could have even broader ramifications. The plaintiff’s products liability suit claims that the company has systematically duped consumers through a “pervasive national marketing campaign.” If successful, this suit could open the door to recovery for others who crash while using Autopilot.

While Tesla has typically been more grandiose in their advertising techniques than more traditional automakers, their legal challenges do serve to highlight the struggles that auto manufacturers will face in the coming years. This year alone, Ford has packaged its driver assist features into a system called Co-Pilot 360 and GM has called its Super Cruise system “the world’s first true hands-free driver assistance feature for the freeway.” In the near future, other car manufacturers are expected to join these companies in developing ever more autonomous features.

As the auto industry collectively drives toward the creation of truly autonomous vehicles, there will be an understandable temptation to hype up every new technological feature. Arguably, many of these features will increase auto safety when used properly. Certainly, road testing such features is a key step on the path towards fully driverless cars. The challenges facing Tesla should serve as a warning though. Companies need to be cautious in describing their driver-assist technologies, and ensuring that customers understand the limits of such new features. Doing so will have the dual benefit of reminding drivers that they should still be in control of the vehicle, and shielding themselves from the type of liability Tesla faces today.

 

Cite as: Daniel A. Crane, The Future of Law and Mobility, 2018 J. L. & Mob. 1.

Introduction

With the launch of the new Journal of Law and Mobility, the University of Michigan is recognizing the transformative impact of new transportation and mobility technologies, from cars, to trucks, to pedestrians, to drones. The coming transition towards intelligent, automated, and connected mobility systems will transform not only the way people and goods move about, but also the way human safety, privacy, and security are protected, cities are organized, machines and people are connected, and the public and private spheres are defined.

Law will be at the center of these transformations, as it always is. There has already been a good deal of thinking about the ways that law must adapt to make connected and automated mobility feasible in areas like tort liability, insurance, federal preemption, and data privacy. 201 201. See, e.g., Daniel A. Crane, Kyle D. Logue & Bryce Pilz, A Survey of Legal Issues Arising from the Deployment of Autonomous and Connected Vehicles, 23 Mich. Tel. & Tech. L. Rev. 191 (2017). × But it is also not too early to begin pondering the many implications for law and regulation arising from the technology’s spillover effects as it begins to permeate society. For better or worse, connected and automated mobility will disrupt legal practices and concepts in a variety of ways additional to the obvious “regulation of the car.” Policing practices and Fourth Amendment law, now so heavily centered on routine automobile stops, will of necessity require reconsideration. Notions of ownership of physical property (i.e., an automobile) and data (i.e., accident records) will be challenged by the automated sharing economy. And the economic and regulatory structure of the transportation network will have to be reconsidered as mobility transitions from a largely individualistic model of drivers in their own cars pursuing their own ends within the confines of general rules of the road to a model in which shared and interconnected vehicles make collective decisions to optimize the system’s performance. In these and many other ways, the coming mobility revolution will challenge existing legal concepts and practices with implications far beyond the “cool new gadget of driverless cars.”

Despite the great importance of the coming mobility revolution, the case for a field of study in “law and mobility” is not obvious. In this inaugural essay for the Journal of Law and Mobility, I shall endeavor briefly to make that case.

I. Driverless Cars and the Law of the Horse

A technological phenomenon can be tremendously important to society without necessarily meriting its own field of legal study because of what Judge Frank Easterbrook has described as “the law of the horse” problem. 202 202. Frank H.Easterbrook,Cyberspace and the Law of the Horse, 1996 U. Chi. Legal F. 207, 207-16. × Writing against the burgeoning field of “Internet law” in the early 1990s, Easterbrook argued against organizing legal analysis around particular technologies:

The best way to learn the law applicable to specialized endeavors is to study general rules. Lots of cases deal with sales of horses; others deal with people kicked by horses; still more deal with the licensing and racing of horses, or with the care veterinarians give to horses, or with prizes at horse shows. Any effort to collect these strands into a course on “The Law of the Horse” is doomed to be shallow and to miss unifying principles. 203 203. Id. ×

Prominent advocates of “Internet law” as a field rebutted Easterbrook’s concern, arguing that focusing on cyberlaw as a field could be productive to understanding aspects of this important human endeavor in ways that merely studying general principles might miss. 204 204. Lawrence Lessig, The Law of the Horse: What Cyberlaw Might Teach, 113 Harv. L. Rev. 501 (1999). × Despite Easterbrook’s protestation, a distinct field of cyberlaw has grown up in recent decades.

“The law of the horse” debate seems particularly apt to the question of law and mobility since the automobile is the lineal successor of the horse as society’s key transportation technology. Without attempting to offer a general solution to the “law of the horse” question, it is worth drawing a distinction between two different kinds of disruptive technologies—those in which the technological change produces social changes indirectly and without significant possibilities for legal intervention, and those in which law is central to the formation of the technology itself.

An example of the first species of technological change is air conditioning. The rise of air conditioning in the mid-twentieth century had tremendous effects on society, including dramatic increases in business productivity, changes in living patterns as people shifted indoors, and the extension of retail store hours and hence the growing commercialization of American culture. 205 205. Stan Cox, Losing Our Cool: Uncomfortable Truths About Our Air-Conditioned World (and Finding New Ways to Get Through the Summer) (2012). × The South’s share of U.S. population was in steady decline until the 1960s when, in lockstep with the growth of air conditioning and people’s willingness to settle in hot places, the trend abruptly reversed and the South’s share grew dramatically. 206 206. Paul Krugman, Air Conditioning and the Rise of the South, New York Times March 28, 2015. × The political consequences were enormous—from Richard Nixon through George W. Bush, every elected President hailed from warm climates.

One could say, without exaggeration, that the Willis Carrier’s frigid contraption exerted a greater effect on American business, culture, and politics than almost any other invention in the twentieth century. And, yet, it would seem silly to launch a field of study in “law and air conditioning.” Air conditioning’s social, economic, and political effects were largely indirect—the result of human decisions in response to the new circumstances created by the new technology rather than an immediate consequence of the technology itself. Even if regulators had foreseen the dramatic demographic effects of air conditioning’s spread, there is little they could have done (short of killing or limiting the technology) to mediate the process of change by regulating the technology.

Contrast the Internet. Like air conditioning, the Internet has had tremendous implications for culture, business, and politics, but unlike air conditioning, many of these effects were artifacts of design decisions regarding the legal architecture of cyberspace. From questions of taxation of online commercial transactions, 207 207. See, e.g., John E. Sununu, The Taxation of Internet Commerce, 39 Harv. J. Leg. 325 (2002). × to circumvention of digital rights management technologies, 208 208. See, e.g., David Nimmer, A Rif on Fair Use in the Digital Millenium Copyright Act, 148 U. Pa. L. Rev. 673 (2000). × to personal jurisdiction over geographically remote online interlocutors, 209 209. Note, No Bad Puns: A Different Approach to the Problem of Personal Jurisdiction and the Internet, 116 Harv. L. Rev. 1821 (2003). × and in countless other ways, a complex of legal and regulatory decisions created the modern Internet. From the beginning, law was hovering over the face of cyberspace. Al Gore may not have created the Internet, but lawyers had as much to do with it as did engineers.

The Internet’s legal architecture was not established at a single point in time, by a single set of actors, or with a single set of ideological commitments or policy considerations. Copyright structures were born of the contestation among one set of stakeholders, which was distinct from the sets of stakeholders contesting over tax policy, net neutrality, or revenge porn. And yet, the decisions made in separate regulatory spheres often interact in underappreciated ways to lend the Internet its social and economic character. Tax policy made Amazon dominant in retail, copyright policy made Google dominant in search, and data protection law (or its absence) made Facebook dominant in social media—with the result that all three have become antitrust problems.

Whether or not law students should be encouraged to study “Internet law” in a discrete course, it seems evident with the benefit of thirty years of hindsight that the role of law in mediating cyberspace cannot be adequately comprehended without a systemic inquiry. Mobility, I would argue, will be much the same. While the individual components of the coming shift toward connectivity and automation—i.e., insurance, tort liability, indemnification, intellectual property, federal preemption, municipal traffic law, etc.—will have analogues in known circumstances and hence will benefit from consideration as general questions of insurance, torts, and so forth, the interaction of the many moving parts will produce a novel, complex ecosystem. Given the potential of that ecosystem to transform human life in many significant ways, it is well worth investing some effort in studying “law and mobility” as a comprehensive field.

II. An Illustration from Three Connected Topics

It would be foolish to attempt a description of mobility’s future legal architecture at this early stage in the mobility revolution. However, in an effort to provide some further motivation for the field of “law and mobility,” let me offer an illustration from three areas in which legal practices and doctrines may be affected in complex ways by the shift toward connected and automated vehicles. Although these three topics entail consideration of separate fields of law, the technological and legal decisions made with respect to them could well have system-wide implications, which shows the value of keeping the entire system in perspective as discrete problems are addressed.

A. Policing and Public Security

For better or for worse, the advent of automated vehicles will redefine the way that policing and law enforcement are conducted. Routine traffic stops are fraught, but potentially strategically significant, moments for police-citizen interactions. Half of all citizen-police interactions, 210 210. Samuel Walker, Science and Politics in Police Research: Reflections on their Tangled Relationship, 593 Annals Am. Acad. Pol. & Soc. Sci. 137, 142 (2004); ATTHEW R. DUROSE ET. AL., U.S. DEP’T OF JUSTICE, OFFICE OF JUSTICE PROGRAMS, BUREAU OF JUSTICE STATISTICS, CONTACTS BETWEEN POLICE AND THE PUBLIC, 2005, 1 (2007). × more than forty percent of all drug arrests, 211 211. David A. Sklansky,Traffic Stops, Minority Motorists, and the Future of the Fourth Amendment, 1997SUP. CT. REV. 271, 299. × and over 30% of police shootings 212 212. Adams v. Williams, 407 U.S. 143, 148 n.3 (1972). × occur in the context of traffic stops. Much of the social tension over racial profiling and enforcement inequality has arisen in the context of police practices with respect to minority motorists. 213 213. Ronnie A. Dunn, Racial Profiling: A Persistent Civil Rights Challenge Even in the Twenty-First Century, 66 Case W. Res. L. Rev. 957, 979 (2016) (reporting statistics on disproportionate effects on racial minorities of routine traffic stops). × The traffic stop is central to modern policing, including both its successes and pathologies.

Will there continue to be routine police stops in a world of automated vehicles? Surely traffic stops will not disappear altogether, since driverless cars may still have broken taillights or lapsed registrations. 214 214. See John Frank Weaver, Robot, Do You Know Why I Stopped You?. × But with the advent of cars programmed to follow the rules of the road, the number of occasions for the police to stop cars will decline significantly. As a general matter, the police need probable cause to stop a vehicle on a roadway. 215 215. Whren v. U.S., 517 U.S. 806 (1996). × A world of predominantly automated vehicles will mean many fewer traffic violations and hence many fewer police stops and many fewer police-citizen interactions and arrests for evidence of crime discovered during those stops.

On the positive side, that could mean a significant reduction in some of the abuses and racial tensions around policing. But it could also deprive the police of a crime detection dragnet, with the consequence either that the crime rate will increase due to the lower detection rate or that the police will deploy new crime detection strategies that could create new problems of their own.

Addressing these potentially sweeping changes to the practices of policing brought about by automated vehicle technologies requires considering both the structure of the relevant technology and the law itself. On the technological side, connected and automated vehicles could be designed for easy monitoring and controlling by the police. That could entail a decline in privacy for vehicle occupants, but also potentially reduce the need for physical stops by the police (cars that can be remotely monitored can be remotely ticketed) and hence some of the police-citizen roadside friction that has dominated recent troubles.

On the legal side, the advent of connected and automated vehicles will require rethinking the structure of Fourth Amendment law as required to automobiles. At present, individual rights as against searches and seizures often rely on distinctions between drivers and passengers, or owners and occupants. For example, a passenger in a car may challenge the legality of the police stop of a car, 216 216. Brendlin v. California, 551 U.S. 249 (2007). × but have diminished expectations of privacy in the search of the vehicle’s interior if they are not the vehicle’s owners or bailees. 217 217. U.S. v. Jones, 565 U.S. 400 (2012). × In a mobility fleet without drivers and (as discussed momentarily) perhaps without many individual owners, these conceptions of the relationship of people to cars will require reconsideration.

B. Ownership, Sharing, and the Public/Private Divide

In American culture, the individually owned automobile has historically been far more than a transportation device—it has been an icon of freedom, mobility, and personal identity. As Ted McAllister has written concerning the growth of automobile culture in the early twentieth century:

The automobile squared perfectly with a distinctive American ideal of freedom—freedom of mobility. Always a restless nation, with complex migratory patterns throughout the 17th, 18th, and 19thcenturies, the car came just as a certain kind of mobility had reached an end with the closing of the frontier. But the restlessness had not ended, and the car allowed control of space like no other form of transportation. 218 218. Ted v. McAllister, Cars, Individualism, and the Paradox of Freedom in a Mass Society. ×

Individual car ownership has long been central to conceptions of property and economic status. The average American adult currently spends about ten percent of his or her income on an automobile, 219 219. Máté Petrány, This Is How Much Americans Spend on their Cars. × making it by far his or her most expensive item of personal property. The social costs of individual automobile ownership are far higher. 220 220. Edward Humes, The Absurd Primacy of the Automobile in American Life; Robert Moor, What Happens to the American Myth When You Take the Driver Out of It?. ×

The automobile’s run as an icon of social status through ownership may be ending. Futurists expect that the availability of on-demand automated vehicle service will complete the transition from mobility as personal property to mobility as a service, as more and more households stop buying cars and rely instead on ride sharing services. 221 221. Smart Cities and the Vehicle Ownership Shift. × Ride sharing companies like Uber and Lyft have long been on this case, and now automobile manufacturers are scrambling to market their vehicles as shared services. 222 222. Ryan Felton, GM Aims to Get Ahead of Everyone with Autonomous Ride-Sharing Service in Multiple Cities by 2019. × With the decline of individual ownership, what will happen to conceptions of property in the physical space of the automobile, in the contractual right to use a particular car or fleet of automobiles, and in the data generated about occupants and vehicles?

The coming transition from individual ownership to shared service will also raise important questions about the line between the public and private domains. At present, the “public sphere” is defined by mass transit whereas the individually owned automobile constitutes the “private sphere.” The public sphere operates according to ancient common carrier rules of universal access and non-discrimination, whereas a car is not quite “a man’s castle on wheels” for constitutional purposes, 223 223. See Illinois v. Lidster, 540 U.S. 419, 424 (2004) (“The Fourth Amendment does not treat a motorist’scaras hiscastle.”). × but still a non-public space dominated by individual rights as against the state rather than public obligations. 224 224. E.g., Byrne v. Rutledge, 623 F.3d 46 (2d Cir. 2010) (holding the motor vehicle license plates were nonpublic fora and that state’s ban on vanity plates referencing religious topic violated First Amendment). × As more and more vehicles are held and used in shared fleets rather than individual hands, the traditional line between publicly minded “mass transit” and individually minded vehicle ownership will come under pressure, with significant consequences for both efficiency and equality.

C. Platform Mobility, Competition, and Regulation

The coming transition toward ride sharing fleets rather than individual vehicle ownership described in the previous section will have additional important implications for the economic structure of mobility—which of course will raise important regulatory questions as well. At present, the private transportation system is highly atomistic. In the United States alone, there are 264 million individually owned motor vehicles in operation. 225 225. U.S. Dep’t of Energy, Transportation Energy Data Book, Chapter 8, Household Vehicles and Characteristics, Table 8.1, Population and Vehicle Profile, https://cta.ornl.gov/data/chapter8.shtml (last visited May 29, 2018). × For the reasons previously identified, expect many of these vehicles to shift toward corporate-owned fleets in coming years. The question then will be how many such fleets will operate—whether we will see robust fleet-to-fleet competition or instead the convergence toward a few dominant providers as we are seeing in other important areas of the “platform economy.”

There is every reason to believe that, before too long, mobility will tend in the direction of other monopoly or oligopoly platforms because it will share their economic structure. The key economic facts behind the rise of dominant platforms like Amazon, Twitter, Google, Facebook, Microsoft, and Apple are the presence of scale economies and network effects—system attributes that make the system more desirable for others users as new users join. 226 226. See generally DavidS.Evans& Richard Schmalensee, A Guide to the Antitrust Economics of Networks, Antitrust, Spring 1996, at 36; Michael L. Katz & Carl Shapiro, Systems Competition andNetworkEffects, 8 J. Econ. Persp. 93 (1994). × In the case of the mobility revolution, a number of features are suggestive of future scale economies and network effects. The more cars in a fleet, the more likely it is that one will be available when summoned by a user. The more cars connected to other cars in a fleet, the higher the quality of the information (on such topics as road and weather conditions and vehicle performance) available within the fleet and the steeper the machine learning curve.

As is true with other platforms, the mere presence of scale economic and network effects does not have to lead inexorably to market concentration or monopoly. Law and regulation may intervene to mitigate these effects, for example by requiring information sharing or interconnection among rival platforms. But such mandatory information sharing or interconnection obligations are not always advisable, as they can diminish a platform’s incentives to invest in its own infrastructure or otherwise impair incentives to compete.

Circling back to the “law of the horse” point raised at the outset, these issues are not, of course, unique to law and mobility. But this brief examination of these three topics—policing, ownership, and competition—shows the value of considering law and mobility as a distinct topic. Technological, legal, and regulatory decisions we make with respect to one particular set of problems will have implications for distinct problems perhaps not under consideration at that moment. For example, law and technology will operate conjunctively to define the bounds of privacy expectations in connected and automated vehicles, with implications for search and seizure law, property and data privacy norms, and sharing obligations to promote competition. Pulling a “privacy lever” in one context—say to safeguard against excessive police searches—could have spillover effects in another context, for example by bolstering a dominant mobility platform’s arguments against mandatory data sharing. Although the interactions between the different technological decisions and related legal norms are surely impossible to predict or manage with exactitude, consideration of law and mobility as a system will permit a holistic view of this complex, evolving ecosystem.

Conclusion

Law and regulation will be at the center of the coming mobility revolution. Many of the patterns we will observe at the intersection of law and the new technologies will be familiar—at least if we spend the time to study past technological revolutions—and general principles will be sufficient to answer many of the rising questions. At the same time, there is a benefit to considering the field of law and mobility comprehensively with an eye to understanding the often subtle interactions between discrete technological and legal decisions. The Journal of Law and Mobility aims to play an important role in this fast-moving space.


Frederick Paul Furth, Sr. Professor of Law, University of Michigan. I am grateful for helpful comments from Ellen Partridge and Bryant Walker Smith. All errors are my own.