How self-driving cars could determine the future of policing

The Waymo driverless car is displayed during a Google event. (Eric Risberg/Associated Press)

I was recently asked to predict the future of policing. My first reaction was curmudgeonly. It can't be done, I responded, at least in a useful way. My second reaction was to run with the question and to make a prediction, or at least to hazard the following guess: Self-driving cars will cause major shifts in policing strategies.

Here's my thinking. A lot of police enforcement of the law is focused on traffic enforcement. Part of that is based on a legitimate interest in ensuring safe driving. Cars can be very dangerous, so the state has a strong interest in ensuring that drivers are alert, trained, licensed and driving cars safely. The Supreme Court has recognized that interest by giving the police significant powers to enforce the traffic laws. Officers can stop a car when they see a traffic violation, and they can force the driver to stay by the side of the road while the officer completes the mission of the stop.

Although part of the police interest in traffic enforcement relates to traffic safety, a lot of it also relates to enforcing other laws. The Supreme Court has repeatedly held that the police can use traffic stops pretextually. What the officer is really trying to do doesn't matter, the court has ruled, as long as the officer has a valid basis for stopping a car for a traffic violation. The reason is that Fourth Amendment doctrine rejects subjective intent as a basis for determining whether a stop is lawful. So long as the officer had a valid basis for a stop (speeding 55 in a 50 zone), the fact that the officer is really acting for other reasons (determining if there are drugs in the car) is irrelevant.

This is hugely important because almost every driver routinely violates the traffic laws. It's actually hard to drive without breaking the traffic code, and doing so can lead other drivers to honk and yell at you as you creep slowly along in traffic. Allowing pretextual stops gives the police lots of power to stop people for non-traffic reasons. And as a lot of people know, especially in minority communities, that power can be abused.

Now consider self-driving cars, also known as "autonomous vehicles." We are just beginning to see some cars on the market that can drive themselves. And a lot of big companies are betting on self-driving cars in the future.

Right now the technology is expensive, but technologies that are expensive today often become very cheap in a few years. Consider smartphones. When they were introduced a decade ago, they were amazing new things only few people had. In 2016, 77 percent of adults and 92 percent of people ages 18 to 29 had smartphones.

So let's imagine, in 20 years or so, that the norm is for cars to be self-driving. Maybe you will have your own self-driving car that is parked on the street and comes to your door and drives you to your destination when you want. Or maybe you will just rely on car services, like Uber or Lyft, and you will pay to have a self-driving car come get you when you need to go somewhere. It won't be a special or rare thing. It will just be what everyone does, like carrying around a smartphone today.

Here's the interesting part of the picture, I think: Self-driving cars can be programmed to drive in perfect compliance with the traffic laws. If the speed limit is 55, your car will drive exactly 55. And if everyone is driving around in self-driving cars, every car will drive exactly 55. Maybe speed limits will change in response. If most or all cars are self-driving, maybe speed limits can be changed because they will drive safely at higher speeds. But the key idea is that a future of self-driving cars may have very few traffic violations. I suspect that will be great for society, on the whole. Most obviously, it will mean fewer accidents and fewer deaths caused by drunken driving.

But it may also cause an unintentional change to policing strategies. If traffic violations are mostly a thing of the past, the police won't be able to use pretextual traffic stops anymore. They can follow self-driving cars all day but won't be able to pull them over. Because the power to make pretextual stops requires a traffic violation, we may see a sea change of enforcement away from traffic enforcement.

This doesn't mean that the police will go away, of course. When it comes to the dynamics of police investigations, there is often an equilibrium at play. Less police power in some areas leads to changes that increase it in others—and vice versa.

My guess is that the shift away from traffic enforcement will come at the same time as a shift toward using car trips to reconstruct past events. As Uber and Lyft customers know, ride-sharing services keep detailed records of trips. When you get your receipt, for example, it tells you not only the start and stop point and time but also the exact roads you traveled. If everyone is getting around using self-driving cars, those cars will keep records of trips that the government can use.

It's hard to predict the future, and of course all of this may be hilariously wrong. But if I'm on the right track, the future of policing may be a future without the pretextual stops and traffic enforcement common today but with the addition of routine reconstruction of past car trips.

In the case of records kept by ride-sharing services, one obvious legal question is whether the records of past trips are protected under the Fourth Amendment or statutory privacy laws. The traditional Fourth Amendment answer would be that the service user does not have rights in the information. The records are just the company's business records of where their cars went, and that is the company's business record and not the customer's. Whether that remains the law is open right now, thanks to the recent grant in Carpenter v. United States, the cell-site case. It may be that Carpenter will make those records protected by a warrant requirement, so that reconstruction is permitted but requires the government to establish probable cause that the trip will reveal evidence of a crime. As always, stay tuned.

Finally, I was reminded after tweeting about this idea that, way back in 2007, Elizabeth E. Joh predicted a variation on the idea. Joh predicted that perhaps in the future roads will be automated and will report traffic violations automatically. Tickets will be issued by computers without human involvement and therefore without human discretion. A decade later, self-driving cars look to me more likely. And unlike automated roads, self-driving cars won't give the police the option to enforce in person if there is investigative advantage in doing so. But I wanted to flag Joh's article, especially given the technological trends in the last decade.