Some urbanists have become skeptical about the future of autonomous vehicles even as unstaffed, autonomous taxis are now serving customers in Phoenix and Japan. Others worry that AVs, if they are ever deployed widely, will make cities worse. Angie Schmitt posits that allowing AVs in cities without implementing deliberate pro-urban policies first will exacerbate the problems of cars in urban areas. However, cars themselves aren’t to blame for the problems they’ve caused in cities. Policymakers created rules that dedicated public space to cars and prioritized ease of driving over other important goals. Urbanists should be optimistic about the arrival of AVs because urbanist policy goals will be more politically tenable when humans are not behind the wheel.
To avoid repeating mistakes of the past, policymakers should create rules that neither subsidize AVs nor give them carte blanche over government-owned rights-of-way. Multiple writers have pointed out that city policymakers should actively be designing policy for the driverless future, but few have spelled out concrete plans for successful driverless policy in cities. Here are three policies that urban policymakers should begin experimenting with right away in anticipation of AVs.
Price Roadways
Perhaps the biggest concern AVs present for urbanists is that they may increase demand for sprawl. AVs may drastically reduce highway commute times over a given distance through platooning, and if people find their trips in AVs to be time well-spent, when they can work, relax, or sleep, they may be willing to accept even more time-consuming commutes than they do today. As the burden of commuting decreases, they reason, people will travel farther to work. However, the looming increase in sprawl would be due in large part to subsidized roads, not AVs themselves. If riders would have to fully internalize the cost of using road space, they would think twice before moving to far flung suburbs.
Now is the time for cities and states to implement congestion pricing policies to manage the demand for scarce road space. Congestion pricing programs in Virginia and London provide potential models. And Singapore provides a model of using congestion pricing not just for highways, but arterial roads as well. Broad-based pricing for road space would encourage a ridesharing model rather than individually-owned AVs, allowing riders to spread the cost of road use over multiple passengers.
In downtown areas, the arrival of AVs will mean a from curbside parking to curbside loading zones. And just as underpriced curbside parking contributes to congestion by causing drivers to cruise for parking, passengers getting in and out of cars will cause traffic if curb space is priced too low. City policymakers should begin exploring options for reallocating curbside parking to loading zones and pricing curb space for short stops. Washington, DC has already started a trial program.
Donald Shoup’s principles for managing curb parking apply to pick ups and drop offs as well; policymakers should set prices high enough so that there’s at least one available pick up/drop-off spot on each block at all times. Since taxis, rideshare vehicles, and delivery trucks are currently the primary users of short-term curb services, cities could begin enforcing prices just for these vehicles using a payment mechanism like EZ-Pass.
Adopt Shared Streets
The adoption of driverless technology presents an opportunity to reform policies designed to support car traffic in dense urban areas at the expense of other road users. Stephen Smith pointed out years ago that AVs will struggle to move in areas that are crowded with pedestrians because walkers will lose their fear of being hit if they step out into slow-moving traffic. Without drastic changes to pedestrian traffic rule enforcement, pedestrians may take over the streets in areas where sidewalks are crowded and in places where there’s a steady stream of people crossing streets. And that’s wonderful! It provides an opportunity to return busy city streets to multi-use spaces that are safe for all types of road users.
Absent policy intervention, driverless cars — or just widespread automatic braking — could turn streets with lots of pedestrians and cyclists into de facto woonerfs. A key promise from AV boosters is that time spent in AVs can be time spent working or doing something fun, so there should be less need to speed AVs through urban areas relative to cars today. AVs are not yet at woonerf-level navigation ability — they would probably come to a complete standstill in a crowded woonerf rather than moving at a walking pace. But testing in San Francisco and Tokyo shows that more difficult environments for navigation may not be far behind.
Cities should ramp up experimentation with shared streets and pedestrian-only streets now to begin determining how to adapt their bus systems to having some streets where traffic moves at a walking pace. Solutions could include grade-separated bus lanes within otherwise shared streets, or rerouting buses to major arterials that have lower pedestrian density.
Most potential woonerfs are in large cities or vacation destinations, and they’re disproportionately in Manhattan. New York policymakers in particular should continue their woonerf and car-free pilots and should plan to adapt public transit accordingly. Places that should begin experimenting with woonerfs outside New York include Georgetown and Chinatown in DC, the French Quarter and Marigny in New Orleans, and State Street in Chicago.
The vast majority of American streets do not have crowded sidewalks or even a steady stream of pedestrians. Without drastic changes to land use, they won’t be reasonable candidates for woonerfs. In these places where pedestrians are sparse, today’s traffic laws may continue working fine even with widespread adoption of driverless cars. Without high pedestrian density, AVs will generally be able to proceed when they have the green light.
Eliminate Parking Requirements and Auction Public Parking
Parking is one of the biggest obstacles to walkabiltity in American cities. With AVs, it will be possible to dramatically reduce car storage in urban areas. Rather than parking when not in use, autonomous ridesharing cars can continuously drop off and pick up passengers. Individuals who own AVs can send them home while they’re at work or to a far flung parking lot that doesn’t take up space in an urban core. Simultaneously eliminating the dead space in parking lots and parking garages and adding more urban residents and destinations would dramatically increase walkability.
Parking requirements — ill-advised at any time — are particularly damaging in a time when it’s foreseeable that parking cars in center cities will continue becoming less important. Now is the time for municipalities to eliminate parking requirements and to sell off city-owned parking for potential redevelopment. Requiring new buildings, with lifespans of several decades, to include space for car storage in places where real estate is valuable is mandating an enormous waste of space and resources as demand for parking decreases.
The private sector is already developing podium parking that is designed to be converted to indoor space once their buildings require fewer parking spaces. Developers are aware that their customers in center cities will increasingly use transportation options other than driving their own cars, and they are building space with the hope of being able to take advantage of reduced parking requirements in the future.
Eliminating parking requirements and selling off government-owned parking lots and garages is the simplest change cities can make right now to for adaptation to a world with less parking and much less center city parking. The introduction of AVs will give policymakers another shot to get this right when they’ll face less constituent pressure for convenient parking.
Driverless Politics
There are a few reasons to believe that the switch to driverless will move politics in a pro-urban direction. The legal system will likely take deaths, injuries, and property damage caused by autonomous vehicles much more seriously than it takes those caused by human drivers. Courts have failed consistently to hold drivers responsible for killing other road users through negligence or reckless driving. Because most judges and jurors drive cars, they can easily imagine themselves in the position of having injured or killed a pedestrian or cyclist. As a consequence, drivers rarely face criminal charges or even traffic tickets for their actions, and victims and their families rarely receive the type of compensation they could expect if their injuries came from a negligent corporation.
While autonomous vehicles are forecast to be much safer than human drivers, some rate of collisions will remain inevitable. But judges, juries and policymakers will be unlikely to show software or car companies anything like the leniency they’ve shown human drivers. After an Uber test car in autonomous mode hit a pedestrian in Phoenix, Arizona state policymakers banned the company from further testing. If a human had been at fault, they likely would have faced no consequences.
Similar politics may help deprioritize the speed of AV traffic in densely populated areas. When drivers are no longer behind the wheel, or even in their own car, politicians and citizens will likely be more open to ideas to level the playing field between cars and other forms of transportation. Cliff Winston and Quentin Karpilow point out that during the period of technological upheaval, when many people will be transitioning from paying for their own car to paying for ridesharing, is a politically opportune time to introduce congestion pricing with the least opposition.
Regardless of the AV industry’s progression, there’s little to now downside risk in pricing roads, trying out woonerfs, and eliminating parking requirements. With these policies in place, AVs present an opportunity to move toward urbanist goals and more walkable cities.
