Google's newly announced self-driving cars have the potential to completely transform cities.
The biggest shift that may result from self-driving cars is an end to widespread auto ownership, with taxis taking their place. The idea is that a driver's wages are what make taxis expensive. Without drivers, then, taxis could become cheap and ubiquitous. There will be no need to own a car because it'll be more cost effective to just summon one whenever you need it.
We're already headed down this road, with Uber revealing this week that someday its entire taxi fleet will be self-driving.
The average person will likely benefit from the reduced need to own a car.
Right now in the United States, the average cost of owning a car is $8,876 per year. American drivers pay 59.2 cents per mile. And that's just the cost of the car itself; if you buy a home with a garage or rent an apartment with a parking space, you're also paying more for housing so you can own a car.
The promise of self-driving cars is that these costs would disappear, or at least shrink.
With fewer privately owned cars, we won't need as many parking lots.
Donald Shoup, a professor of Urban Planning at UCLA and probably the world's foremost expert on parking, has pointed out that cars remain parked 95% of the time. Think about it; when you're at work, asleep, eating, showering, hanging out with friends, or doing whatever, your car just sits there, taking up space (and money). And because cars are parked so much of the time, there are actually far more parking spaces in the U.S. than there are vehicles.
Self-driving cars may eliminate this kind of waste. After dropping someone off, they could go pick up another person, and later they could be stored remotely and compactly. And with fewer parking spaces, more people would likely choose to walk and cities could be built with more density.
Self-driving cars may also make streets safer and less congested.
Google is already figuring out how to make its cars safer than human-controlled vehicles. And if the cars become widespread, entire transportation systems could be "hyperoptimized" so all the vehicles on the road interact efficiently. Imagine a world with no erratic drivers, no pointless lane changing, and no accidents due to human error. Suddenly the evening commute would become a lot less hellish.
That said, these benefits could be decades away because they would require vast numbers of self-driving cars on the road.
But if there's one reason to be skeptical, it's that cars themselves are simply less space efficient than other forms of transportation. Here's a GIF that makes that point obvious:
Google's self-driving cars are smaller than the ones in the image above, but they're still a lot bigger than a couple of seats on a trolley, bus, or subway car. That means even if the world embraces self-driving cars we'll still have to devote more space to transportation than would be necessary if public transit systems simply improved.
Self-driving cars also will have to use infrastructure that has historically damaged cities.
Engineers spent much of the last century cutting through large swaths of existing cities to put in freeways, with devastating results often affecting minority communities. These freeways are extremely costly and continue take a toll on their respective cities.
Neither more public transit nor self-driving cars would automatically fix these problems, but in the case of the latter they actually exacerbate them. Some believe, for example, that self-driving cars could actually lead to more overall driving. The cars also could result in increased urban sprawl, which tends to drive demand for even more cars and more car-only infrastructure like freeways. And in places where most people already use public transit, self-driving cars could produce more congestion.
All of which means that the rise of self-driving cars will maintain demand for things like freeways — where cars are welcome but that are alienating to anyone using a different way to get around.