“Could driverless cars really kill congestion?” asked Mónica Guzmán in the Seattle Times, a thought she admits gained special urgency while she was parked in SR 99 gridlock. On the one hand, the technological benefits are easy to see. But what about the technological impacts we don’t see? In what ways is a driverless car not really a car, but a network?
Guzmán lists off the usual near-term benefits we can see coming:
Packed with advanced sensors, location-aware technology and links to real-time traffic data, the cars would see better, respond faster and know more. Speeds go way up, accidents go way down, and backups are history.
But she gets immediate push-back from Seattle Subway’s Ben Schiendelman, who points out that you wouldn’t see much improvement in traffic congestion until most cars were driverless. (It takes only one distracted driver to cause an accident.) And WSDOT’s Mark Bandy wonders if driverless cars might bring even more congestion, if they can be sent off on errands with no one in them. Maybe you’d just send your car home from work, rather than pay for parking downtown.
Others fear “driverless sprawl”; what’s a three-hour, one-way commute if you can work on your laptop the entire time? Sprawl can be a problem but it’s not clear that the driverless car adds that much fuel to its fire. If someone is effectively doing six hours of work in a mobile office, after all, it raises the question of why they’re driving into the office at all. If what’s required is office facetime, will that many people be interested in tacking a six-hour commute onto an eight-hour workday?
But on Sightline, Ben Schonberger and Steve Gutmann have a different take on a self-driving-car future, asking whether this disruptive technology might be most disruptive to the concept of the privately-owned car, at least in urban areas. They envision fleets of on-call cars making an even greater dent in the inefficiency of private-car ownership. (The Atlantic‘s Alexis Madrigal has written on the topic of a driverless transportation system, too.)
Their “Car2Google” service would work much like car2go already does, except that you wouldn’t have to hoof it to the nearest car; it would come to you, a driverless taxi service. If the company can legitimately promise its customers a car when they need it, ownership becomes a moot point. Schonberger and Gutmann imagine that you’d be billed incrementally, but a subscription service doesn’t seem out of the realm of the possible, either.
However it works, fleets of driverless cars offer that systemic shift that Schiendelman brings up, while still providing point-to-point transportation. (This doesn’t obviate the need for mass transit: point-to-point service strengthens a light rail or subway system’s utility, especially if there’s no parking involved.)
What’s most disruptive about the technology of driverless cars might be the economics of scale they offer, making their use more of a mandate than any legal requirement. Part of car2go’s challenge in Seattle has simply been scaling up to meet latent demand, but here and elsewhere they’re serving membership of around 20,000 with 500 cars. In contrast, across the U.S., there are 439 privately-owned cars for every 1,000 people.
It seems rational to spend $30,000 on a car, until it doesn’t. People in more rural areas, where distance is a prohibitive factor in a driverless car just zipping over, would still probably purchase their vehicle. And while commuters park their cars, farmers live in their pickups.
But what if driverless cars effectively offered you a $25,000 housing subsidy, so long as you live within an area dense enough to generate that economy of scale? (This is just car ownership, not even considering externalities like on-street parking and garages.) People who want a home in the country and will put up with a two-hour commute are not, I’d argue, the real engine of sprawl. That’s caused by the drive to qualify (for a mortgage). The network effects that dense urban (even suburban) areas reap the benefits of could help rewrite that equation.
Driverless cars, like horseless carriages, may represent a larger leap than we can at this moment suspect. And though they’re still predicted to be “eight years out,” their widespread adoption as part of transit systems (public or private) could come much more swiftly than we imagine.
