Slamming the brakes on self-driving cars
What if your car could keep you out of an accident by knowing what was coming around the corner?
No, I’m not just telling you what you already know about self-driving cars. Most of the self-driving prototypes we are familiar with rely on sensors, which have limited range and can’t see past obstacles. That’s enough to prevent a lot of accidents —but what if trouble is lurking outside their range of “sight”? Enter V2V: vehicle-to-vehicle communications that could make the dream of self-driving cars come true. Or not, if regulators decide not to push us in that direction.
The idea is that our cars could use radio spectrum to chat with each other even when they can’t see each other. Your car might learn that a car far ahead of them has suddenly braked hard, enabling you to safely brake well ahead of the problem. Or that a car approaching on a crossroads has lost its brakes and is about to career through the intersection.
Last December, the Barack Obama administration pushed ahead with a rule that would require automakers to start deploying these systems in new cars within a few years. But not everyone was happy with the rule: It uses scarce spectrum that other industries would like for themselves, and there are questions over whether the technology’s usefulness outweighs its hefty cost. Under President Donald Trump’s transportation secretary, the rule has become “mired in regulatory review”.
This is the worst possible time for regulators to be stepping on the gas and then suddenly slamming on the brakes. Self-driving cars are coming: Maybe we can’t yet gauge their arrival time, but we certainly know they’re headed down the road at a pretty brisk clip. And while they are often hailed by techno-utopians as a triumph of the free market, their arrival is going to be partly determined by the government—especially by regulators who outline safety and communications standards and by judges who settle the lingering question of who has liability for any accidents that do occur.
If self-driving cars are going to include vehicle-to-vehicle communications, then developers should know that now, rather than wasting time developing workarounds for problems that could more easily be solved with V2V. Or conversely, basing their plans around expected V2V systems that cease to be required by regulators, forcing them to go back to the drawing board.
Such slowdowns to innovation will have a price—in dollars, in human lives. Last year, about 40,000 people were killed in auto accidents in the US. When a life-saving technology is in development, even a year’s delay means thousands, perhaps tens of thousands of needless deaths. It also means many more people struggling with serious injuries and chronic pain. Regulators need to settle this question. Which means we need to settle an even bigger question about regulation.
One of the first things Trump did as president was to sign an executive order requiring that two regulations be rolled back for every new one that was promulgated. As a longtime advocate of rolling back regulatory complexity, I found a lot to like in this rule. Unless strenuous effort is made to regularly prune them back, regulations have a tendency to blindly grow until they have wrapped the economy so tight that nothing can breathe, much less thrive. Trump’s executive order forces us to do some very necessary maintenance.
Unfortunately, it’s a rather crude instrument for the job. It’s possible that nothing more nuanced would actually work; nuance and flexibility, alas, give regulators quite a bit of discretion to put off an annoying chore. Just the same, we should recognize the dangers of cutting with such a dull knife.
In agencies that are dealing with a genuinely new and disruptive force—one for which innovators need regulatory clarity before they can bring a product to market—such a strict regulatory requirement faces them with an unpalatable dilemma. Every new regulation that is vitally necessary means finding two regulations that aren’t. But each of those dusty old rules was created for a reason. Some of those reasons were bad; in some cases the rule never worked as intended; and in others the rule later became obsolete. But that does not describe all of our regulations. And imagine the brutal publicity that will ensue if someone dies or gets hurt in an accident that could plausibly have been prevented by a rescinded rule.
An even larger problem is that the “two for one” rule doesn’t tell us about the relative quality of the regulations involved. An agency could satisfy that executive order by issuing a terrible, stupid, costly regulation that forced power companies to rip out their windmills, as long as it repealed two sensible and necessary regulations that (for now) prevent factories from dumping toxic waste into our watersheds.
Reducing regulatory complexity is an important goal, but it cannot be our only goal. We ought to strive for less regulation, yes, but also for good regulation, and for regulation that gives companies enough clarity to innovate and build their businesses. Unfortunately, while the Trump administration has proven enthusiastic about the first goal, we’ve seen much less talk about the others.
Republicans have too often failed to have that conversation—to formulate a serious and workable philosophy of what regulations our modern economy needs, as well as noting the many it could do without. And the voters have allowed it. We don’t want to waste our time thinking about complicated decisions and confusing details; we want to stake out our position, for or against “regulation,” and leave it at that. Bloomberg View
Megan McArdle is a Bloomberg View columnist.
Comments are welcome at firstname.lastname@example.org
- Opening bell: Asian markets trade mixed; Tata Steel, Infosys, Maruti in news
- The growing importance of the finance sector in the stock market
- Are housing companies such as Unitech too big to fail?
- Smaller firms did badly in September quarter, but it has nothing to do with GST
- The steadily shrinking share of the construction industry