If Driverless Cars Really Are the Future, the Government Might Have to Pay for Them

The crystal ball predictions about the future of driverless cars keep on rolling in. The latest is from a highly publicized study last week from IHS Automotive, which forecast that self-driving vehicles will hit the streets in six years and explode from there, until nearly all cars are driver-free by 2050. That's quite a bold estimate, especially considering only a few countries and US states have even started to grapple with the legal implications of the new technology—the biggest hurdle to clear before cars start driving themselves around the world's roads.

So what needs to happen to achieve this foretold era of ubiquitous autonomous vehicles? According to the policy think tank RAND Corporation, policymakers need to study the pros and cons of the futurist vehicles and draft legislation to maximize their benefit to society. On that note, the group released a report today describing what it found to be the public benefits and disadvantages of self-driving cars. It found that, overall, the upsides outweighed the downsides—enough so that it might be worthwhile for the government to subsidize the emerging automobile.

The report, funded by "philanthropic donations," viewed the issue through some pretty rose-colored glasses. Like plenty of others have, it lauded driverless cars’ potential to reduce crashes, cut back on energy consumption and pollution, and provide "last mile" mobility to disabled, elderly, or poor people who aren't able to drive. It also suggested autonomous vehicles would minimize the "cost" of congestion by making the time spent stuck in traffic productive—especially if lawmakers ease up on distracted driving laws and make it legal to use your smartphone or laptop while en route.

Interestingly, the study pointed out that, down the road, eliminating the driver from vehicles could change the way land space is developed. One, because people may be more willing to commute further to work, pushing out suburban sprawl; two, because the considerable space currently tied up for parking in cities could be freed up and used for development if cars could drop off their passenger and go park themselves further away.

Of course, there's a flip side to this techno-utopian outlook. While the report echoed the conventional belief that autonomous vehicles are the key to a collision-free transportation system—a commonly cited statistic is that Google's driverless cars have driven more than 500,000 miles without a crash so far—I’m more skeptical.

Sure, it's logical: Eliminating the human could cut down on drunk driving and careless mistakes, by replacing them with autonomous controls triggered by sensors that automatically brake if there's a quick stop in front of the car, and so on. But what about hacking connected cars? Or a cyberattack on the digital infrastructure powering the future autonomous highways? What about automation failure or faulty information? Plus, as the report mentioned, smart cars have to be able to make sense of the data coming in about their environment. Is that bulge in the middle of the road a deer or a tire? Can the technology “think” quick enough to act accordingly, without human insight?

The environmental benefits are also ambiguous. The report suggests AVs could save energy through more efficient use, say by driving closer together at lower speeds to make better travel time, much like a train. Plus, if crashes did go down, companies could build lighter cars to lower their carbon footprint. On the other hand, if driverless cars became super popular, that could mean more people are on the road in general because it's easier and more fun to travel, which would make congestion worse and make energy emissions—and fuel prices—shoot up.

Which leads us to the money. As Motherboard's Derek Mead recently wrote, even though people are excited about the idea of driverless cars, they're not likely to actually buy them unless they're cheaper than the "dumb" alternative—which they won't be, at least at first, even though optimistic predictions expect insurance prices to drop if the cars prove to be safer. A recent report found self-driving models would cost $5,000 to $20,000 more than conventional vehicles, plus higher maintenance costs.

To know if autonomous vehicles really are safer, or advantageous to society, or if it’s even feasible to pull this idea off on a large scale, we have to get the robot cars on the road to study their impact, which means people have to start buying and using them. Some form of government incentive, like with electric vehicles, are a possible solution to this Catch-22.

As the report points out, the state could step in and absorb some of the liability risk from private companies, or support nervous insurance companies to encourage them to lower rates. Another option would be to redirect funds now going toward public transportation to a kind of futurist car-train hybrid—though it’s easy to argue that wouldn’t benefit the public more than, say, high-speed rail or a more efficient and affordable public transit system.

For its part, the report's outlook is optimistic. The technology "has considerable promise for improving social welfare but will require careful policymaking at the state and federal level to maximize its promise," it concluded. "Such research would also be helpful in determining the optimal mixture of subsidies and taxes to help align the private and public costs and benefits of this technology." Your move, lawmakers.

@meghanneal