Kill-Switch Rule Under Consideration

Section 24220 of the 2021 infrastructure law directs the National Highway Traffic Safety Administration (NHTSA) to require that automakers include a device in new cars that will passively detect whether the driver is drunk and keep them from operating the car. NHTSA has until November of this year to write the rule and then new cars made beginning two years after that must comply. The law specifically states that if NHTSA can’t find a device that will accomplish this, then it doesn’t have to require one.

U.S. Navy photo by Photographer’s Mate 2nd Class Nathanael T. Miller.

This law has led to fears that the government will require automobiles to have “kill switches” that the government can remotely control, thus shutting down people’s freedom of mobility at any time. Defenders of the policy argue that these fears are overblown. Last month, NHTSA issued its advance notice of proposed rule making on the subject and it’s worth looking at it before comments are due next Tuesday, March 5.

First of all, the law doesn’t require or even mention the idea of a remote kill switch. NHTSA’s notice of rulemaking also doesn’t mention the possibility. While this doesn’t mean it won’t be proposed, it seems unlikely.

Second, the law doesn’t require or even mention the idea of recording or tracking someone’s movements. It does suggest that a vehicle could measure “impaired driving” and not just someone’s blood alcohol level. But NHTSA’s rulemaking notice specifically notes that “privacy considerations are critical” to any system and suggests that it might prohibit recording devices that could be a part of determining if someone is driving impaired. However, such devices are still a possibility.

A third concern that critics have is that a tiny percentage of false positives in any device designed to determine if someone shouldn’t be driving could be highly disruptive. Americans make hundreds of millions of auto trips per day, so 0.01 percent false positives could impact tens of thousands of people each day.

Another concern is cost. In the past, NHTSA has required such things as seat belts, air bags, antilock brakes, and vehicle stability control. All of these have added to the cost of new automobiles yet they are partly responsible for a major decline in auto fatalities since 1970. How much would an impaired driver device add to the cost of an automobile and what that cost be worth the benefit?

NHTSA’s notice points to the increasing fatalities since 2011 and notes that a large percentage of drivers in fatal accidents were drunk. But the actual data (shown in a chart on page 10) shows that drunk driving fatalities have declined much faster than total fatalities. In 1982, almost half of all fatalities involved a drunk driver, but that’s down to around 30 percent today. While total fatalities have grown since 2011, drunk-driving fatalities in 2019 were almost exactly the same as in 2011. They grew in 2020 but that’s largely due to changes in behavior from the pandemic.

Still, 30 percent is about 10,000 deaths per year. While the cost issue is worth addressing, if costs can be kept down anything that can greatly reduce that number sounds like a good idea. However, people who don’t drink might not agree to having to pay for something they don’t use.

There is also the possibility that NHTSA might decide to propose a system that doesn’t measure someone’s alcohol consumption but just senses whether they are driving erratically. “Impaired driving,” the notice says, could include distracted driving or drowsy driving. Considering recent research has concluded that smart phone distractions are responsible for much of the increase in fatalities since 2011, devices that try to assess whether someone is distracted could be worthwhile.

Here I need to point out that NHTSA’s own data say that distracted drivers are reported to be involved in only a small percentage of fatal accidents. Of course, few people will admit they were distracted by their smart phone when the accident took place. But I have some problems with the Cambridge Analytics report claiming U.S. drivers are more distracted. First, how do they know that people using smart phones in a particular car are the driver and not passengers? Second, since we didn’t have smart phones before about 2008, how do they know that drivers weren’t equally distracted by something else?

Assuming that the correlation between the increase in smart phone use and the increased fatalities since 2011 is more than a coincidence, something that would sense a driver’s distraction and force them to return their attention to the road would be a good thing.

Highway safety shouldn’t be a partisan issue. I hear the right saying that Biden is trying to take away our freedom to drive, but I don’t see that in this proposed rule. I hear the left making knee-jerk reactions that drunk driving is bad, but that problem seems to be declining anyway.

Instead of making a political issue about this, we should make it a technical issue and insist that NHTSA use a data-driven process to identify the most cost-effective measures we can take to reduce fatalities and injuries. If that includes an in-auto detection system, we should be glad we can develop such a system that can potentially save 10,000 or more lives per year.