What some environmentalists miss about Uber and Lyft
Uber and Lyft make dense urban living more fun and convenient.
David Roberts—a former Vox colleague who now runs the excellent Volts newsletter (Subscribe!)—isn’t a fan of ride-hailing services like Uber and Lyft.
In a Monday post, he pointed to a new study from a trio of Carnegie Mellon researchers finding that ride-hailing trips create more problems—including higher greenhouse gas emissions, congestion, crashes, and noise—than if the customer had driven a personal vehicle. They calculated that these extra harms were worth around 35 cents per trip.
The reason is simple. When you drive yourself, your car is only moving when you’re in it. In contrast, when you hail an Uber, its driver first has to come to you. These extra miles of travel—known as deadheading—means more of all the negative consequences that come with traveling by car.
Roberts draws some pretty sweeping conclusions from this. “The only solution to the problem of cars is fewer cars,” Roberts writes. “For those who care about the public good, Uber and Lyft are a distraction.”
It’s an interesting study, but unlike Roberts I don’t think it comes close to telling us whether ride-hailing apps serve the public good, on net. Indeed, the authors of the study readily acknowledge this.
“Our paper focuses on the relatively near term, assuming that individual trips are shifting from personal to [ride-hailing] vehicles without other major changes to the transportation system,” they write. “In the long term, large scale [ride-hailing] travel could lead to fundamental changes in land use, development, transit, and lifestyle (as opposed to per-trip) patterns.”
I suspect that if you take account of these more complex impacts, the environmental impact of ride-hailing becomes more positive. In particular, ride-hailing apps make dense urban living more fun and convenient. That, in turn, encourages the development of denser cities. And as Roberts himself acknowledges, denser cities have substantial environmental benefits.
How Uber and Lyft promote density
When a city has a lot of people per square mile, people live close to stores, restaurants, schools, and other amenities. Errands are shorter, and a higher fraction of them can be made on foot or by bicycle. Higher densities support better transit systems and higher transit utilization. Apartments and townhouses also take less energy to heat and cool than single-family homes. All of this means fewer greenhouse gas emissions.
But one of the biggest obstacles to high-density development is demand for parking. When I first moved to the DC metro area in 2003, I lived in a large apartment building in the Virginia suburbs. This was a high-density development of a sort, but a lot of the benefits of high-density living were negated by the vast parking lot surrounding the building. It wasn’t very convenient or pleasant to walk anywhere, and there wasn't a subway stop anywhere nearby. So most people in the building—including me—owned cars and drove for most trips.
My experience illustrates the chicken-and-egg problem any suburban area faces if it tries to increase density. There’s a middle range where a neighborhood is dense enough for parking shortages to become a concern, but not yet dense enough that it’s convenient to get around in other ways. You frequently see this when a developer proposes putting an apartment building in a suburban neighborhood: existing residents come out in droves to fret about parking shortages.
Ride-hailing apps help here. By making it easier to get by without a car, they encourage some people to live without one. And this matters most for new developments.
If an apartment building already has a big parking lot around it, then parking will likely be free to tenants and a lot of them will go ahead and get a car. And once you have a car, it’s still going to be easier and cheaper to drive most of the time—especially if big parking lots make it hard to walk anywhere.
But if you’re designing a brand new apartment building, you’ll want to allocate fewer parking spaces per unit—they take up valuable land, and many of your future tenants might be happy without one. When a bunch of developers do this over a period of years, the result will be a more walkable neighborhood that benefits everyone.
Over the last decade, dense urban living has become increasingly fashionable relative to single-family homes in the suburbs. Uber and Lyft obviously aren’t the only factor driving this shift, but it seems likely that their existence contributed to this trend by making it easier for people—especially young people who like to go out drinking on the weekends—to get around without a car.
Self-driving technology could make ride-hailing more environmentally friendly
The main reason ride-hailing is bad for the environment is deadheading: the extra driving a vehicle does between dropping off one passenger and picking up the next one. Unfortunately, Uber and Lyft customers often find themselves waiting for long deadheads as a driver comes from a far-flung neighborhood to pick them up.
Uber and Lyft could fix this problem by increasing the density of drivers: having more cars in more places, ready to pick people up at a moment’s notice. This would be convenient for customers and good for the environment. But it would be economically inefficient, because drivers’ time is valuable.
Self-driving technology could transform the economics of the ride-hailing business and make this practical.
Human drivers aren’t willing to sit around for hours waiting for their next hail. Self-driving cars are entirely willing to do that. Obviously, the owners of these vehicles would like to achieve high utilization to maximize the return on their investment. But the cost of a self-driving car sitting idle for an hour is far lower than the cost of a human driver doing so.
This technology is not so far off. Google’s sister company Waymo is already testing a self-driving taxi service in the Phoenix suburbs and is preparing to expand to San Francisco.
In the long run, we can expect companies running self-driving taxi services to over-supply their networks with vehicles, relative to the current industry norms. As a result, the average vehicle will be substantially closer at the time you hail it. We can also expect self-driving vehicles to be substantially cheaper (since they won’t have to pay a driver), which should increase utilization. That, too, should increase the density of rideshare vehicles and hence reduce average deadheading distances.
What this means in practical terms is that when you pull out your smartphone in 2041, there will likely already be a self-driving car waiting a block or two away from you. If you want to, you’ll be able to hail it five minutes ahead of time and have it wait until you’re ready to go.
One place where this will make a particularly big difference is in neighborhoods located a mile or two from a subway station. Uber and Lyft have fairly high minimum fares, which is largely a reflection of the deadheading costs drivers face before picking up a passenger. These high fares (and uncertain wait times) make it impractical for someone to take a 1-mile Uber ride to the local subway station as part of their daily commute.
Not only are self-driving taxi rides likely to be cheaper in general, they’re likely to especially be cheaper for short fares. So more people should be able to take a cheap, short taxi ride to a subway station, then take the subway the rest of the way to work.
At least one startup is already experimenting with services like this. A couple of years ago, I visited a company called Optimus Ride that was testing prototype self-driving vehicles in a new mixed-use development near a subway stop in the Virginia suburbs of DC. The developer of the project was going to pay Optimus Ride to offer a free self-driving shuttle service that takes residents to the subway as well as to a nearby grocery store, restaurants, and other locations. Not only would this make the development more attractive to residents, the developer hoped it would save them money by reducing the number of parking spots they need to supply in all of these locations.
I don’t know if Optimus Ride will succeed in achieving fully self-driving capabilities. Last time I checked in with them they still had safety drivers behind the wheel on most trips. But sooner or later someone is going to get this technology working at scale. And when it does it will have a substantial impact on the way we organize cities. And I expect many of those changes to be positive from an environmental perspective.
How do Uber and Lyft impact your travel patterns? Would your personal carbon footprint be larger or smaller in a world without these services? Please share your thoughts in the comments.