Internet Explorer 11 is not supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

How Will Driverless Cars Really Change Cities? Who Knows.

There are plenty of theories about how they will reshape urban areas. But it’s anybody’s guess.

Driverless Ford sedan on the street
It’s normal to have long and passionate debates about the future that technological breakthroughs will bring us. Sometimes the boldest predictions turn out to be right. Other times they miss the mark by quite a bit.

I’m just old enough to remember the futurists who were convinced that television would destroy movies and radio. That didn’t happen, of course. Movies continue to be a shaper of our culture, and radio has evolved and survived.

A decade later came the speculation about jet airplanes. Their most enthusiastic boosters foresaw a magical era of speed, safety, comfort and convenience. That was about half right. Jets can take us most places quickly, and their safety record is admirable. But they’re less comfortable than ever, and they aren’t very convenient unless you live in a big city with a hub airport.

It’s fun to debate the impact of technology, and we do it all the time. But I’ve never seen a technology debate remotely like the one now going on about autonomous vehicles (AVs) and the impact they will have on our urban and suburban lives.

If you believe some urban futurists, AVs will usher in an exciting age for central cities, one in which parking lots and garages will all but disappear, parks and playgrounds will replace them, and the traffic congestion of the present will be a distant memory. If you believe others, the lure of riding in a car without having to drive will give us a whole new generation of exurban sprawl.

I have no idea how soon these vehicles will become ubiquitous. I suspect it will take longer than we think; that’s been true with most other technological innovations. But let’s drop that question and assume, for the sake of argument, that a day is coming when so-called Level 5 AVs, in which the occupant does nothing whatsoever, will dominate the nation’s traffic. With that in mind, I began looking into what might happen to cities and metropolitan life.

A few things hit me right at the start. The first one is that the most important issue isn’t technology -- it’s ownership. The most optimistic scenarios for the driverless car era implicitly assume that the net result will be a substantial reduction in the number of cars on the road. AVs will be more of a service than a product. People will get used to sharing them on their daily trips, and there won’t be much need to own a personal vehicle. Traffic will thin out, the air will be cleaner, and all the unused parking spaces will be available for urban amenities and even housing.

It sounds like a civic paradise. But let’s try a couple of thought experiments. What if you’re an 82-year-old middle-class retiree, living in a middle-distance suburb, who has just given up driving. You have the option of summoning an AV to your relatively remote location and waiting for it to come, or you can buy a little driverless runabout, keep it at home and just tell it where you want to go. Which would you choose? I think I know. Or suppose you’re a newly empty-nested family out in the suburbs that’s trying to get along with one car and finding it a little stressful. Wouldn’t a second one with no driving responsibilities be kind of tempting? I’m willing to concede that the technology will cut down on ownership by people in or near central cities. But that won’t be nearly enough to create the kind of revolution that some urban enthusiasts are talking about.

If we switch to AVs without reducing the number of cars, one of two things is likely to happen, neither of them very good. Either the sheer volume of automobiles will require vast amounts of parking, just as it does now, or the parking will go away but the vehicles will wander aimlessly all day through traffic-choked city streets -- zombie cars -- waiting for the next set of instructions from their master.

The point is that driverless technology, on its own, isn’t going to transform cities or end traffic jams. There will have to be other forces generating the change, new incentives that discourage car ownership and promote shared mobility.

One of these forces needs to be transit, and most important, microtransit. As AVs become cheaper, it might become practical for local governments to purchase and operate large fleets of them, not only in central cities but also in suburbs where residents could dial in a request and have them arrive in a few minutes. When they’re not dropping off or picking up passengers, these vehicles could be parked in relatively open suburban territory where congestion wouldn’t be a problem. Maybe our octogenarian retiree would be willing to rely on them and refrain from purchasing one for himself. Maybe. But for the process to work, he still has to be persuaded to share a relatively small space with one or more strangers. If he insists on traveling alone, that wouldn’t be much of a solution.

And no matter how many people are using these AVs, we wouldn’t want them shuttling back and forth between city and suburb and cluttering downtown streets. Ideally, they would transport their passengers a relatively short distance to a transit station, where people would transfer to a conventional bus or light rail, leaving the driverless car out in the suburbs where it would do the least damage.

Right now, however, our metropolitan transit systems fall woefully short of meeting that test. Once you get a few miles beyond city limits, as we all know, the transit network, even the bus service, thins out very quickly. To create a holistic transportation system that links microtransit to larger vehicles, we would need to spend a great deal more than we are spending now on expanding and improving service. That’s possible, especially if driverless technology brings the cost down, but it certainly isn’t on the horizon at the moment. Without a much greater public commitment to transit in general, it seems safe to predict that our mythical suburbanite would ride his AV straight into the center of the city, adding one more piece to the downtown traffic puzzle.

Unless. Unless we create another significant disincentive for bringing automobiles into city centers, whether they have drivers or not. This brings us to congestion pricing, the strategy that has been promoted tirelessly for decades by Samuel Schwartz, the former New York City traffic commissioner. Schwartz is the author, most recently, of No One at the Wheel, the fairest and most comprehensive treatment of the driverless car issue that I have come across. In the book, he takes up the whole history of congestion charges, from their relatively successful implementation in European cities to New York City Mayor Michael Bloomberg’s failure to get them approved a decade ago.

Schwartz offers several scenarios for how a congestion driving disincentive might work. It could mean charging a flat fee for entering the crowded center of a big city, as London and several other European cities do. It could involve a fee based on the number of vehicle miles traveled, an idea Oregon has tested. Or it could base a charge on the number of hours driven. Any of these might conceivably help. But in the end, as Schwartz concedes, they all require an effective and efficient transit system. People will restrain themselves from driving downtown only if they have another convenient, frequent and reliable way to get there.

It’s only fair, after all this urbanist futurism, to pay a little attention to the opposite argument: that people won’t mind 50-mile commutes if they can sit in the back seat and relax. This, too, involves some rather heroic assumptions. For one thing, it depends on how fast AVs are allowed to go. If they are traveling at 100 miles an hour, then yes, long-distance commutes might become more attractive. But if they are observing current speed limits, as they would probably have to do in the beginning, then a 50-mile commute could take longer than it does now. I’m not sure how many commuters would want this. The main beneficiary might be the liquor industry -- AVs would likely end up as the equivalent of the bar car on the 5:15 train from New York to Westport in “Mad Men” days. It doesn’t strike me as a big step forward.

But let’s concede a point and posit that long commutes into and out of the city will one day become pleasant and efficient. This could well mean, as some critics insist, a multiplication of new and distant exurban communities. But it wouldn’t mean a revival of 1950s split-level and cul-de-sac suburban enclaves. The millennial generation has already demonstrated its demand for dense and walkable places to live. That’s almost certainly going to define the exurbs of the driverless future. Central cities might or might not flourish in a driverless world; urban values would. 

Alan Ehrenhalt is a contributing editor for Governing. He served for 19 years as executive editor of Governing Magazine. He can be reached at
Special Projects