The Slow Approval of Self-Driving Cars Is Costing Lives
Laws requiring a "driver" in driverless cars make as much sense as requiring a horse to be yoked to the front of an automobile, just in case.
A recent news story gave the anti-self-driving car pearl-clutchers some ammunition, although it was more comical than tragic. In West Hollywood last month, a goofy sidewalk-delivery robot collided with a robotaxi at around 4 miles per hour. There obviously were no injuries and neither vehicle was damaged, although writers had a field day depicting this as a preview of the coming dystopian war between robots.
Critics of autonomous vehicles, or AVs, can fixate on nonsense if they choose, but widespread legislative efforts to limit the development of these vehicles actually endanger us given that robots are much better drivers than human beings. They're not chatting on cellphones, fussing with the radio, or distracted by crying children. They are more attuned to surrounding traffic—and can "see" much farther ahead than the average driver. But fear of the unknown often drives policy.
Granted, I used to share the "isn't that cool, but I'm not getting into one of those" view as I've watched Waymo robotaxis covered with futuristic-looking cameras and lidar sensors zip around San Francisco. Then I grabbed a ride in one in Phoenix and was shocked by the non-shocking nature of the experience. It was like taking any taxi—except the car was a nice Jaguar, I could play my own music and I didn't have to listen to the driver's political theories.
The taxi ride from my hotel to tour Waymo's operations center south of downtown actually was a bigger ordeal as the human driver got lost and took me miles out of the way. The robotaxi drove carefully. I'm not saying things can't go wrong, but they can go wrong when riding in any vehicle. Accidents in traditional taxis and other human-piloted vehicles aren't exactly an unusual occurrence.
News reports also made a big deal out of a minor robotaxi accident in San Francisco, even though there were scores of human-related car accidents the same day. The latest peer-reviewed research from Swiss Reinsurance Co. found that robotaxis "significantly outperformed…the overall driving population" with an 88-percent drop in property-damage claims and a 92-percent drop in bodily-injury claims. Insurance research tends to be extremely reliable as insurers need to know the risks as they set rates.
Nevertheless, lawmakers and regulators continually try to slow the adoption of self-driving vehicles. One key reason is unions, which fear the loss of jobs—especially when it comes to the trucking industry, taxi businesses, and transit systems. Beyond raw protectionism, the key stated concern is safety. In that case, the data doesn't support the call for onerous regulation. Foes of these vehicles are trying to gin up unrealistic safety concerns to limit a life-saving advancement.
For instance, many states are pushing "driver-in" laws that require a human driver in the driver's seat, which defeats the purpose of the technology. It's as if early 20th-century legislators required a horse to be yoked to the front of emerging automobiles just in case. Others ban AVs for interstate commerce. This restricts the industry's expansion, forcing companies to spend time lobbying rather than creating a new industry.
In the new legislative session, California lawmakers are expected to revive a bill (Senate Bill 915) that would let larger municipalities impose their own limits on AVs. These include permitting rules, caps on the numbers of such vehicles, and mandatory inspections. The state Department of Motor Vehicles and California Public Utilities Commission already regulate AVs, so don't buy the nonsense that they are somehow unregulated. This measure is pernicious given that vehicles would have to comply with vastly different laws as they crossed city lines.
This regulatory situation reminds me of debates about the federal Food and Drug Administration, which is responsible for approving medicine and medical devices. Obviously, the nation needs some regulation to protect against dangerous products, but the FDA is known for its excruciatingly slow approval process. If the agency approves a drug too quickly and it leads to ill results, that means bad headlines and congressional inquiries.
Therefore, in bureaucracies, the incentive structure is to go slowly and impose myriad impediments to new developments. With the FDA, that has meant years of delay of life-enhancing drugs. As others have noted, the victims of this delay do not know why they can't access these treatments. These downsides are unseen, but they cost thousands of lives.
Back to vehicles, more than 42,000 Americans die in collisions every year. Based on the above-mentioned research, AVs have dramatically lower bodily-injury rates. If governments slow approval of self-driving cars—or give local governments the ability to stop their use based on anecdotes and irrational fears—then we'll likely have more deaths and injuries. And no one will know why.
So, sure, it might be funny to think about dueling robots on L.A. streets, but let's not let minor foibles cause us to lose sight of the bigger picture.
This was first published in The Orange County Register.
