Driverless cars and the future of law enforcement

The Texas Tribune has a story out on a new report (pdf) from UT-Austin's Center for Transportation Research that opens thusly: "Self-driving cars once seemed the stuff of science fiction. But with such projects from companies like Google already being showcased in Texas, lawmakers and transportation officials see an opportunity to use new technology to relieve congestion and improve safety in the rapidly growing state."

Volvo has suggested they'll have cars on the market by 2020 that are "impossible to crash"; the Tribune estimates driverless cars will be marketed to the public by 2025. And of course, it'll take some time to transition to the new tech and/or retrofit older vehicles. But a couple of decades from now, when the technology has been well-established, one can foresee a day when insurance companies may charge more or even refuse to cover people driving without such high-tech crash prevention technology.

From a traffic safety perspective, self-driving cars can't come soon enough as far as I'm concerned. When it finally happens, though, it's going to change law enforcement forever. For starters, no more drunk drivers, roadside breathalyzers, DWI cases in court, or for that matter $1,000 per year surcharges . Ethan Couch could be as drunk as he wanted in Google's driverless car and get home safely every time. But the even bigger change could be redirecting law enforcement away from the traffic stop as their primary enforcement strategy, since presumably vehicles would be programmed to obey the traffic laws. In that case, what would become of law enforcement's much-beloved pretext stop? As pointed out in this Grits post from January:

In 2008, according to the Bureau of Justice Statistics (pdf), 59.2% of all citizen contacts with police were traffic related. That year, police conducted searches at about 5% of traffic stops nationally, discovering contraband in about one out of every10 searches. While most drivers (84.5%) thought they were pulled over for good cause, an overwhelming majority of drivers searched, said BJS, said the search was "perceived as not legitimate." Police arrested 2.6% of drivers they stopped that year.

When police no longer have traffic enforcement as a pretext for getting around the Fourth Amendment, law enforcement in this country will lose what's become a primary crime fighting strategy. In that sense, Grits views the focus on traffic enforcement and pretext stops as an example of short-term, numbers-driven myopia. Just as fishermen fish where the fish are, law enforcement focuses on traffic enforcement not just to prevent crashes but also because that's the most cost-effective way to maximize arrests for other offenses, allowing police to find contraband when they otherwise would have no cause to stop, much less search average citizens. For now that makes sense if their goal is to maximize arrests. But police seeking to maximize arrests in the future may have to rely on different tactics that today they tend to downplay, like investigating and solving reported crime.

Overall, I'd view such a shift in law enforcement tactics and culture as a good thing.

On the negative side, the privacy implications are fairly profound because GPS tracking is pivotal to the functioning of driverless cars. There would need to be some way to ensure anonymity before driverless cars are unleashed en masse. That's not an insurmountable problem, but if driverless cars rolled out in large numbers without some sort of anonymizing feature, it would be a privacy concern on the scale of the NSA mining cell-phone location data.

There will be pros and cons to change but make no mistake: In the words of Sam Cooke, "Change is gonna come."

MORE: Here's a piece from Forbes on the massive amounts of data accumulated by new cars and potential privacy implications. The author also articulates some of the limits of anonymizing vehicle data.