In the urbanist sphere, much debate over AVs (self-driving cars) boils down to whether they will encourage further sprawl - and the negative externalities sprawl creates when built to excess - or denser construction and more urbanized living. There are compelling arguments in both directions. By making driving safer, and ostensibly decreasing congestion by decreasing the required minimum distance between vehicles, it’s plausible that an autonomous future could result in more dispersion. By reducing the need for parking near destinations, on the other hand, autonomous vehicles might encourage urban commerce.
Overall, the trend seems to be somewhat in the latter direction. Pilots of self-driving shuttle services, combined with the popularity of ride-hailing apps and their easy translation to driver-less vehicles, indicate that the self-driving future will heavily involve shared vehicles. Autonomous buses are also being developed, with much interest from the transit industry. In any event, I would suggest that people who view a new technology as the deciding factor behind land use changes are missing something. Whether new travel methods change where people live depends as much on policy decisions as on ability to travel.
Take traffic. On one hand, if the optimistic projections are correct, then AVs will deal with less total congestion, improving road throughput. However, a mass of small vehicles merging onto one road will inevitably cause congestion. Los Angeles and Houston won’t become meaningfully faster to drive through with autonomous vehicles alone. Yet autonomous vehicles combined with charging for use of the road and, where appropriate, mass transit, may just be the right combination that makes commutes far faster than they are now.
Or, consider the impact of land use regulations. Many would scoff at the question of whether driverless vehicles would hurt urban areas, because of the perceived preference in America for suburbs and against urban living. However, much of this preference is because it's difficult to build enough housing for urban areas – and remember, “urbanized” does not need to equal “major city” – to become affordable. Increase the housing supply of existing cities, and allow suburbs to fill in, and the future becomes more walkable and accessible – with autonomous vehicles filling in the gaps that walking and transit can’t quite effectively meet.
Curb space is another consideration. Many complaints over ride-hailing apps involve cars taking up scarce space to pick up and drop off passengers (and deliveries). Yet the decreased need for street parking, combined with a regulatory scheme that charges vehicles for use of curbs, would mitigate these nuisances. The greater throughput of vehicles could allow for many “stroads” to be decreased in width, and for high-occupancy vehicles to get exclusive access to lanes – decreasing total car traffic.
Not to suggest that such a rosy scenario is guaranteed. Yet by inveighing against self-driving cars themselves, many well-meaning advocates not only ignore the potential benefits of this technology; they also lose sight of the bigger picture. The automobile would have become popular with or without the subsidies bestowed upon it by government at all levels in the postwar era. Yet its sheer dominance and negative impact on cities is a result of policy decisions which gave utter deference to private cars, and cushioned road users against paying the true cost of the expensive infrastructure they used. To prevent a future where AVs have a similar impact, the focus needs to be on eliminating subsidies. Even if the driver-less revolution never comes to pass, this subsidy paradigm still needs to change.