Self-driving cars may create more traffic congestion than they solve, expert says

Waymo says it has a couple hundred autonomous vehicles on the roads across metro Phoenix — that number is closer to 800 in San Francisco.

And while that pales in comparison to the number of human-driven vehicles in those cities, David Zipper says the advent and availability of the internal combustion engine could provide a lesson about what might be on the horizon with AVs.

Zipper is a senior fellow at the MIT Mobility Initiative and a contributing writer at Vox. In a piece called "A self-driving car traffic jam is coming for U.S. cities, he argues the increase in autonomous vehicles could bring "the most tumultuous shift in transportation since cars first rumbled their way into the scene."

Zipper joined The Show to talk more about this.

David Zipper

/

Handout

Full conversation

MARK BRODIE: I was really interested in how you compared the coming increase of AVs into cities to when the internal combustion engine was mass produced about a century ago.

How similar or dissimilar is what we're on the cusp of now to where we were about 100 years ago?

DAVID ZIPPER: I mean, I think that the level of excitement in some categories, particularly among some in the transportation industry or some who are thinking about the future of technology, I think it really is comparable. I mean, there was so much enthusiasm 110 years ago for how the automobile could really liberate cities from horses, which had a lot of problems, right.

And cities really raced to sort of accommodate themselves to facilitate car trips by, you know, narrowing sidewalks, by installing on street parking, by making walking across the street illegal by establishing the jaywalking as a crime. And you know, if you look back now, I think we think that a lot of those moves were misguided or they were mistakes.

A lot of cities are trying to undo them. And, and I think that self-driving cars could be equally transformative in changing how we travel in cities and also how we think about the public space that is our streets today.

BRODIE: Yeah, it sounds like you very much feel as though autonomous vehicles will be maybe as disruptive to cities and how cities function sort of on a day-to-day basis as the initial automobiles were.

ZIPPER: I think they can be at scale, Mark. And if you think about what a city full of that many robotaxis would look like, I do think there's some, some things that are to me quite concerning as somebody thinks a lot about the future of cities. And I think really at the top of the list I would put congestion because the sort of raison d'etre of self-driven cars is to make the experience of being in a car nicer. And I think it is nicer, Mark.

And I think humans react to things that are pleasant by doing a lot more of them. And if you imagine people taking a lot more car trips because of trips they would have otherwise taken, perhaps on transit or by biking, or maybe just trips they wouldn't have taken at all, also some trips that might be longer because they're happy to be stuck in congestion since a self-driving car is sort of pleasant to be inside.

You scale that out when you've got tens of thousands of cars and, and hundreds of thousands of people making decisions to be inside cars more often. And we're talking about gridlock the likes of which we've never seen before. And that's a real problem, I would say, for urban life and for urban economies.

BRODIE: Well, it also seems like it would kind of defeat the purpose of one of the main benefits that supporters of these robotaxis tout, which is there are fewer cars on the road. People aren't driving sort of in a human way where they might be distracted, they might be tired, and that with autonomous vehicles there are fewer cars on the road, so there's less congestion.

You're saying it might have the exact opposite impact?

ZIPPER: Oh, I think there's very good reasons to think that it would have the exact opposite impact with many more cars on the road, partly because of these sort of newly induced trips.

But there's another factor, Mark, and that's what's known as deadheading. For those who don't know the term, deadheading occurs when a taxi or an Uber or a self-driven car is en route for a pickup or is waiting for the next summons.

It's just sort of driving around empty. Deadheading is as much as 40% of the miles of ride hail like, like Uber and Lyft. Right now. It may not be 40% with robotaxis, but it's going to be something. And those empty cars are going to be taking up street space, highway space and slowing everybody else down.

BRODIE: Well, so you have advocated for maybe an expansion of what New York City has done in terms of congestion pricing and, you know, trying to reduce the number of cars on the street.

I wonder, though, given the, I'm not even sure what the right word is, but given the vitriol with which opponents derided it and fought it in New York City, is that a viable way forward for other places?

ZIPPER: Well, I would note, Mark, that you're right. It was very controversial at implementing congestion pricing in New York, but its popularity has risen sharply this year since it was implemented in January. And by the way, this is something we see in every city that has adopted congestion pricing.

But I think there's other ways of addressing the sort of added congestion from robotaxis that don't go all the way toward a Manhattan-style congestion pricing.

One possibility that I noted in the Vox story is to create attacks simply on deadheading to induce those companies to find ways to minimize the extent to which their vehicles are occupying street and road space without any passengers inside whatsoever.

BRODIE: One of the other things that you write about is this idea of cities needing to rethink their curb space and maybe rethink their parking spaces. And I'm curious about this because we've seen some cities start to rethink how loading and unloading zones, for example, in commercial areas areas, or passenger pickup and drop-off zones in places that have lots of ride hailing or even autonomous vehicle activity.

And I wonder how big of a deal do you think this might be in terms of cities trying to rethink curb space and maybe put more of a value on it than they have in the past?

ZIPPER: Oh, I think it could be a very big deal, particularly in the future when we have a whole bunch of robotaxis. And I would make one other point about this one, Mark, which is that when we're talking about mileage-based fees or deadheading fees just now, I think that there would be some tension between what the public interest is and what the ride-hail and robotaxi companies would be advocating for.

But when we start talking about curb-based reforms, there's really nice alignment between what the companies are looking for and what public officials and the public interest would call for.

And I'll give you a sort of simple example focusing on robotaxis. You know, a robotaxi doesn't really need to park the way that a human driven car does after arriving at its destination. So what's problematic is if there's no place to pause and pick somebody up and, or drop them off, what would be great is if you create specific spots on the curb where these pickups and drop offs can occur.

The problem now is that most cities really couldn't tell you where they want robotaxi companies to do these pickup and drop-offs because they don't really know how their curbs work. They don't have a great up-to-date map of all of their curb space and designated spaces that they might change or what the parking rules are. So step one is just measuring everything.

BRODIE: What kind of timeframe do you think we're looking at, and what do you think cities are looking at in terms of when this robotaxi future might be that some of these things might be really important for cities to be looking at and doing doing well?

ZIPPER: I have to be honest with you, Mark, I don't know. And more than that, I think anybody who tells you they can they can tell you exactly when robotaxis are going to be scaling and ubiquitous, you might want to look a little bit askance at them because these predictions have been famously wrong for a decade. Let's be candid.

But what I would add is that, you know, we've been talking about congestion pricing, we've been talking about understanding your curb. These are all steps that are advantageous to city life. I would argue regardless of when, or even if, robotaxis truly scale, there's really no reason to wait.