A Google Inc. driverless car, circa 2010, which navigates using video cameras, radar sensors, a laser range finder and detailed maps. Photograph: Google Inc., via Bloomberg
A Google Inc. driverless car, circa 2010, which navigates using video cameras, radar sensors, a laser range finder and detailed maps. Photograph: Google Inc., via Bloomberg

Here’s the dream: You want to go out to meet some friends for a drink, 30 miles from your house. You get into your autocar, set the destination, and then lie back for a quick disco nap so that you’ll be fresh enough for dancing afterward. The car drops you off at the restaurant and obediently takes itself off to the parking lot. It will come back to pick you up when you send an electronic signal that you’re ready. You can even have it drop off a friend who lives five miles beyond your house -- after it’s dropped you off. It will be back in the driveway in time for your morning trip to the farmer’s market.

Or maybe you don’t own a car, but want to go to the crowded ballgame on the outskirts of town. You pull out your smartphone, which informs you that an autocab is five minutes away -- seven minutes if you want to cut your costs by sharing with a nice couple who are already on their way to the game. Because you’re meeting friends at the park, you’d love to save money by sharing. You spend five minutes getting your wallet and keys together, and checking out tonight’s lineup, before your autocab pulls up to the door with a nice couple already in it … wearing the jerseys and caps of the opposite team.

Ah, well, nothing’s perfect, even in the world of driverless cars.

But to hear tech futurists talk, it will be close. Driverless cars will be safer and quieter. They will fetch you to your destination, and then trundle off to park themselves. They will all but eliminate the auto accidents that kill tens of thousands every year. I believe all this. But I’m still worried about the future of driverless cars, not because they’re technically impossible, but because the liability possibilities are enormous.

To see why, a brief primer on manufacturer liability. Currently, liability is limited by two things. On the manufacturer side, user negligence: If you run your car into a school bus, it is you, not the manufacturer, who is liable. Even if better brakes might have helped, the fact that you were going 50 in a school zone is probably going to make it hard for the plaintiffs to recover from a manufacturer.

On the user side, liability is limited by the fact that most people are judgment proof, or close to it. Plaintiffs will usually sue up to the limits of your liability policy, but not beyond, because it’s not worth the hassle of trying to collect. Even most middle-class people have a bit of a retirement account and a smidgen of home equity, both of which are usually protected in bankruptcy. They have a few thousand dollars in the bank, which can be converted into cash and “spent on ordinary living expenses” long before the jury delivers a verdict. The lawyer can get a judgment against you for $10 billion, if he wants, but the expected value of that judgment is … the limits on your liability policy.

(Note: This does not apply to people with large amounts of liquid assets, who should have lots and lots of liability insurance. But hopefully, they already know that.)

On the other hand, the expected value of a $10 million judgment against Ford is … $10 million. That gives lawyers more incentive to file claims, and increases the expense to the company of the claims that are filed.

And even if the overall number of accidents drops, the number of accidents where the automaker is perceived to be at fault will approach 100 percent. After all, they’re the ones who designed or installed the software that made the decision. And while in theory, a jury should be able to say, “Well, this was a hard design problem, you can’t make everyone happy, and this is an unfortunate tragedy," in practice, this is unlikely. If the machine built by a corporation made a decision that killed or seriously injured a person, the jury is going to give the person money at the expense of the corporation.

These issues make me very worried for the future of driverless cars. Understand that I’d love to be wrong -- I, too, want a car that will let me nap while it does the hard work. But I think this is a big hurdle for the nascent industry to jump. They may “jump” it by specifying that drivers are expected to be alert and at the wheel at all times. That would still be good from a safety standpoint -- auto fatalities would fall a lot. But it would be far from The Dream.

Here’s one way that we could fix this: Do a radical overhaul of our liability system. In fact, we’d get rid of the liability system, and replace it with a no-fault accident insurance system like New Zealand’s. It's mandatory, but it’s “true insurance” -- which is to say, it’s paid for entirely by user fees. If you have an accident, the system doesn’t try to assign fault; it just pays you off, according to a schedule of damages.

This will make trial lawyers, and some liberals, somewhere between uncomfortable and hopping mad. (The scheduled damages in New Zealand are relatively modest; that’s what keeps the system affordable.) Deny people their right to sue, and potentially win a huge verdict from a company?

But the liability system is not a very good way to handle these sorts of problems. For one thing, if you have a bad accident, why should you be able to recover only if (1) you can prove someone was negligent; and (2) that someone is well-insured and/or rich? Nor is the liability system a good way to punish negligence. Studies have repeatedly found that the system errs in both directions: It punishes people who have done nothing wrong, and it doesn’t punish people who have actually been grievously negligent. The New Zealand system takes care of people who have had bad accidents, and leaves punishment up to professional boards, who don’t do an obviously worse job than the malpractice system here. Certainly, it lowers the use of “defensive medicine” (tests and monitoring, and even procedures, done largely to avert potential lawsuits). That not only lowers the cost of health care; it lowers the number of people scarred or even killed by unnecessary c-sections, biopsies, and other invasive procedures. That’s a win for everyone.

Conservatives won’t like it because it’s mandatory. But as government programs go, this is pretty minimally intrusive. It consists of a cash grant, not service provision, so it’s not particularly distortionary. It doesn’t involve much redistribution, except for people at the lowest income level (to whom we’re already redistributing via various welfare programs). And it replaces a system that’s even worse: the tort lottery.

If we want to have an innovative society, we need to have a legal structure that will support that society. This would be a good first step, not just for driverless cars, but for all the inventions we haven’t imagined yet.