Proposition 103, approved by California Voters in 1988, created an elected insurance commissioner and mandated a rate setting process to protect consumers. It also required an immediate 20% reduction in auto insurance rates and set forth the criteria that can be used to set future auto insurance premiums as follows:
(a) Rates and premiums for an automobile insurance policy, as described in subdivision (a) of Section 660, shall be determined by application of the following factors in decreasing order of importance:
(1) The insured’s driving safety record.
(2) The number of miles he or she drives annually.
(3) The number of years of driving experience the insured has had.
(4) Those other factors that the commissioner may adopt by regulation and that have a substantial relationship to the risk of loss….(California Insurance Code section 1861.02)
The issue when applying this criteria to a self-driving or autonomous vehicle is that the safety record of an individual and the number of years of driving experience that individual has are irrelevant and immaterial when that person is not actually driving the car. While the number of miles driven may be of significance, what is more important is the real world performance of the hardware and software used in the autonomous vehicle.
The NHTSA Federal Automated Vehicles Policy, published in September of 2016, adopted the SAE International taxonomy for levels of autonomy as follows:
At SAE Level 0, the human driver does everything;
At SAE Level 1, an automated system on the vehicle can sometimes assist the human driver conduct some parts of the driving task;
At SAE Level 2, an automated system on the vehicle can actually conduct some parts of the driving task, while the human continues to monitor the driving environment and performs the rest of the driving task;
At SAE Level 3, an automated system can both actually conduct some parts of the driving task and monitor the driving environment in some instances, but the human driver must be ready to take back control when the automated system requests;
At SAE Level 4, an automated system can conduct the driving task and monitor the driving environment, and the human need not take back control, but the automated system can operate only in certain environments and under certain conditions; and
At SAE Level 5, the automated system can perform all driving tasks, under all conditions that a human driver could perform them.
There are no level 5 vehicles in operation today but there are definitely level 3 and level 4 vehicles, and the hardware is reportedly in place tor Tesla’s to become level 5 vehicles at any time. That means no human driver would be required at any time. If there is no human driver you can’t base insurance of years of experience and safety record.
At a hearing conducted by California State Senator Tony Mendoza Chair of the Senate Committee on Insurance on March 8, 2017, it was reported by Ian Adams in the Insurance Journal that although Senator Mendoza recognized that “strict adherence to Prop 103 does not fit” with self-driving vehicles, Deputy Insurance Commissioner Chris Schultz told the committee that the department believes Prop 103 works, for now, and that no immediate changes are needed.
This is a major disconnect with the California DMV. The DMV’s new proposed Driverless Testing and Deployment Regulations were released March 10, 2017 and these expressly allow for the testing of autonomous vehicles without the presence of a driver.
It is also surprising given California Insurance Commissioner Dave Jones’ attempt to get out ahead of these issues at a Public Hearing he held on September 15, 2014 at the Tech Museum in San Jose. At that hearing, the constraints of Prop 103 were discussed, as was the fact that Prop 103 only applies to insurance policies for personal vehicles; it does not apply to vehicles that are used as a public or livery conveyance for passengers, nor vehicles rented to others. See California Insurance Code Section 660(a)(1). This basically means that Prop 103 does not apply to taxis, limos, black cars, TNCs (like Uber or Lyft), or car rental companies.
As to why all this is important; the key to developing accurate insurance underwriting in this area will almost certainly require what is referred to as telematics. Telematics is computerized information that is conveyed from the vehicle to a third party. Traditionally this information included GPS which allows services like GM’s OnStar to provide roadside assistance, remote diagnostics or stolen vehicle location.
Currently a number of insurers have usage based insurance. One model relies on the number of miles driven by an Insured. This model does not require real time vehicle information other than an accurate odometer reading. An example of this is Metromile which is licensed in California. State Farm, AAA-Southern California and CSE were also early California adopters of this approach.
Another model of insurance does require vehicle telematics and looks at how someone is driving in addition to where and how many miles are driven. This uses real time information such as GPS location, acceleration, and other items to provide instant feedback on risky driving behavior and can result in cost savings. An example of this second model is Snaphot by Progressive which is not licensed in California. There are many benefits to the use of telematics outside of autonomous vehicles such as encouraging less driving, incentivizing safe behavior, and creating a proper economic allocation. Insurance provided in this manner is based on real risks and is not biased by the societal or socio economic factors that consumers have complained of in the past.
An NHTSA study indicated that 99% of all accidents are caused or contributed to by driver error. It is without doubt that an always on, safety conscious, rule abiding and ever vigilant system with superior senses (LIDAR, sonar, radar, video, gps, mapping, multi-vehicle communication), processing speed and reflexes (many vehicles are already drive by wire) will significantly reduce if not eliminate most of the accidents caused by human error or reckless behavior. There is a reason autonomous vehicles can operate in a tight high speed peloton when humans cannot. The potential cost savings from reduced accidents are significant, (experts estimate a reduction in insurance cost of approximately 72%), but only if those savings are based on to consumers.
If telematics are not used as the ratings basis, then there is no transparency into the safety of an autonomous car, and no ability to reduce the cost of insurance based on a reduced risk. It is not merely a profit motive by insurance companies that is driving this, it is also consumer driven. As reported by the AAA in a recent survey, many Americans still feel unsafe sharing the road with autonomous cars. They are however also ready to embrace the technology as 59% want the technology in their next vehicle.
As we know, SAE level 5 vehicles are not on the road today; that means we must look to level 3 and 4 in the short term. The underlying problem with SAE Level 3 or Level 4 Vehicles is that they require the system to be turned on, and therefore require telematics. If a system is not “always on” then without data, an insurance company will likely assume it is never on, and never pass any savings along to its insureds.
This boon to the insurance industry will not result in early adoption of the technology. The lack of a cost savings incentive will delay and reduce the true societal benefits of reduced accidents, property damage and physical injuries that should arise out of use of autonomous driving technology. The problem in California is, absent some change to Proposition 103, telematics and the fact that an autonomous system is engaged cannot be one of the “top 3” ratings factors.
Other factors at work here are the opposition to any insurance changes by Harvey Rosenfield, Consumer Watchdog and other groups that benefit greatly as intervenors in the current rate setting process. If telematics are used then there is little need for this onerous process. The intervenors received $17.6 million since 2003 according to an article by Jim Miller at the Sacramento Bee. Mr. Rosenfield’s opposition to any changes to Prop 103 or any decrease in the opportunity to intervene can reasonably be expected, but it apparently came as a surprise to Senator Mendoza at the recent California Senate hearing. Undoubtedly these groups will use warnings about invasion of privacy as a red herring to mask the real economic incentives at work.
Additionally there will likely be opposition from taxis, limos, black cars, TNCs and car rental agencies as those companies stand to gain a potential significant cost advantage by using the technology in California, the largest car insurance market in the country. They alone will be able to use telematics to get reduced insurance rates. They therefore have a clear economic incentive not to open up discounted rates for autonomous vehicles to the general public. There is already the huge economic incentive to these companies of eliminating the labor cost of drivers, but if they can at the same time maintain a barrier to the entry and get an insurance discount, the economic incentive will only be that much greater.
The last strange bedfellows created in opposition to allowing the use of telematics in connection with individually owned autonomous cars are transportation planners, traffic regulators, environmentalists and employers.
Planners and regulators prefer a hub and spoke or trunk model which optimizes the use of mass transit. Public fleets of automated vehicles and how to manage them by Bern Grush and John Niles is an example of this thinking. It goes without saying that there are many environmental and traffic benefits to simply having less vehicles on the road. A recent study by Shirley Zhu and Alain Kornhauser, Professors at Princeton in New Jersey found that 5 out of 6 vehicles could be eliminated using a shared vehicle model.
Employers at technology companies have become early adopters of commuter shuttles based on the fact that active work time can be recovered through Wi-Fi enabled vehicles. When an employee can work for one hour or more rather than drive through traffic, their employer sees a significant increase in productivity. If all vehicles were capable of level 5 the resulting increase in productivity could be astronomical.
All that is certain at this point is that autonomous vehicles are a reality and there will be a great many societal benefits when they fully arrive. However, if California wants to maximize those benefits and make its implementation as broad based and consumer driven as possible, then significant changes will need to be made to Prop 103 as soon as possible. Otherwise there will be a number of winners and losers based on the existing regulatory scheme. California is undoubtedly the current technology leader in the autonomous vehicle field, but its position could be in peril if it doesn’t make the right adjustments soon.