Self-driving cars are already here.  Automakers and software companies are testing them like crazy, and people like Elon Musk predict that they could be the norm within a decade.

There’s a lot to like about the technology: it can potentially make car accidents a thing of the past, allow us to avoid the pain and stress of driving and instead do something productive with that time, and potentially free up a lot of space in our built environment that is presently dedicated to parking cars, particularly if we switch to on-demand type ownership models (the “Netflix of cars” approach).

But from an environmental perspective, there’s a huge potential downside.  If driving is made easy, the technology could encourage more driving and therefore more traffic.  Just imagine how convenient it might be for people to live far from their jobs if the commute can involve a robot driving for them while they nap or work on a laptop.

So what can policy makers do to head off this dystopia?  The solutions look a lot like what we need to do to address traffic and sprawl anyway, even without the new technology taking over:

• First, we should stop investing in highway expansion projects and instead use available dollars to maintain existing roads and highways.  We shouldn’t build more capacity to accommodate endless self-driving traffic, which will only encourage more sprawl and congestion.
• Second, we should build more alternatives to traffic, such as bus-only lanes, bike and pedestrian infrastructure, and rail transit.  When we do build highways, we should include congestion pricing features, with the revenue reinvested in mobility options for that corridor.  With viable and plentiful alternatives to driving, people won’t opt for self-driving cars as readily.
• Finally, and most importantly, we need to price transportation better to reflect the true costs.  If people truly want to live far from job centers and use self-driving to navigate long commutes, that’s fine.  But they should then pay for the costs to us all from the increased traffic and environmental degradation.  That means switching from a gas tax to a vehicle miles traveled, or fee-per-mile, tax.  As cars get more fuel efficient and transition to battery electric models, gas tax revenue will decline.  Self-driving car owners will then need to help pay for the infrastructure they’re using.

There are other policy options to consider, but these three would be a great start to heading off a potential nightmare scenario of self-driving traffic and sprawl misery.  And in the meantime, we can encourage the environmentally positive features of self-driving technology by encouraging more car- and ride-sharing business models and the development of “right sized” vehicles that fit the vehicle size to the trip.

In the end, it’s always better to be prepared for the future than overwhelmed by it.