With Uber and self-driving cars, the time is right for congestion pricing

Why Uber and autonomous vehicles should be great news for fans of congestion pricing.

Hiding behind the safety debates around self-driving cars is what I consider the more important question: will our driverless car future bring convenient, safe, and efficient mobility? Or will we just get more congestion? Contrary to optimistic forecasts, there’s no guarantee driverless cars won’t be just as stuck in traffic as drivers are now, a point Robin Chase has made repeatedly. 

Autonomous vehicles might make congestion better—or they might make it worse

The risk is that fully autonomous vehicles will make taking a car so easy and cheap that everyone will do it a lot more. Uber may already be having this effect. The added efficiencies of AVs might not outweigh the congestion externality. 

But the worst can be avoided.

Fortunately, congestion is one of the rare real-world problems that actually has a simple, elegant solution. Economists and transportation experts agree congestion can be prevented by internalizing the additional cost through a spatially and temporary variable fee. When congestion is priced at the optimal level, enough potential drivers will decide to not drive, or to travel at a different time, or use a different route, and traffic will flow freely. Moreover, the resulting revenue can be used to improve public transit, or any other worthy cause.

If congestion pricing is such an easy solution, why don’t we use it? The reason is that it’s historically been extremely politically difficult.  Planners and economists have proposed it again and again, in place after place, only to see their proposals fall to political barriers. (Singapore, London, and Stockholm are the few exceptions).

This time the political equation is different

However, now with Uber, and later with AVs, congestion pricing is suddenly much more feasible, for two reasons.

1. Technologically it’s more feasible than ever

Ideally, the congestion price varies by time and location, so you need a system that knows a vehicle’s location at a given time and charges fees accordingly. Until recently this was technologically very difficult. But smartphones and GPS have changed that. Uber’s surge pricing already uses the necessary technology. It would not be technically difficult to add a varying congestion fee. 

2. The political equation finally works—at least temporarily.

Thus far congestion pricing has been a political non-starter. People don’t like being charged for something—road space—they’re used to getting for free. Politicians balk at imposing a fee perceived as a burden on ordinary working people, even though there are proven ways to mitigate the impact. Privacy concerns also loom large—people feel uncomfortable having their movements tracked.

Uber changes this political equation. If cities required a congestion fee for ride-hailing trips, the target would no longer be ordinary people, but a large corporation whose users are relatively well-off. Privacy concerns lessen, because Uber’s passengers are already accustomed to having their movements tracked. Plus, surge pricing has already shown users will (however grudgingly) accept variable prices. In the future, when driverless cars are provided by large mobility companies, a similar political equation will remain.

The main political opponent of such a fee would likely be Uber itself. In this regard cities have leverage because they can decide who gets to operate on their streets. And, Uber’s users and drivers don’t want to be stuck in congestion either. Such a fee might even be good for Uber’s business, if it allows users to take more trips. (Someone should model this.) Any company is unlikely to adopt such a charge unilaterally though, since would only work if also imposed on competing companies.

Cities can and should adopt congestion fees for ride-hailing services now.

Governments could levy a variable per-mile fee on all ride-hailing trips as a condition for operating on city streets. This is important to do now, while the political barriers are low. The policy should remain in place as we transition to autonomous vehicles. This will ensure we can enjoy the effortlessness the technology will bring, without getting stuck in traffic.  

The future of transportation, viewed from Washington

I spent the last four days in DC, wearing a suit and locked in rooms with some of the nation’s most influential transportation people, who were also all wearing suits. As a fellow for the Eno Center’s Future Leaders Development Conference, I had the opportunity to join in discussion of the future direction of transportation. We met with top people of the DOT, congressional committees, big companies, local agencies, even some startups, and many former secretaries of transportation. 

I came away feeling both depressed about the ability of the federal government to solve problems, and optimistic about the commitment of the people involved and the potential of technology. Here are my takeaways. 

1. Funding for public infrastructure is a huge problem

Federal funding for infrastructure investment is constrained and is going to continue to be for the foreseeable future. The federal highway trust fund is insolvent and by 2026 is projected to have a $107 billion deficit. One of the most memorable conversations was congressional staffers recalling their greatest recent achievement: working 24 hours a day on the FAST Act, which essentially kept federal funding for transportation exactly where it has been.[1]  You know something is wrong when Congress’ “success” is little but upholding an inadequate status quo. 

We need to keep pushing for more and better funding though. Without a major change in how transportation is funded, we will barely be able to afford to maintain our current roads, bridges, and tracks, let alone build more. 

2. Private and public roles are changing, yet still the same 

This week reiterated for me that, although the real advances in transportation are coming from the private sector, that doesn’t mean government’s role is obsolete. Obviously autonomous vehicles pose safety problems that need to be addressed by government. But in my view the potential congestion problems arising from Uber and ultimate autonomous vehicles deserve more attention. 

Congestion is the kind of externality that will most likely only be solved through smart public policy, like pricing. The politically opportune time to introduce a pricing policy is when autonomous vehicles are new, when the safety regs are still being hashed out. Something like a per-mile or per-congestion unit fee on road use would go a long way to make sure our streets and highways stay relatively unclogged. Plus it would raise revenue to reinvest in road infrastructure, which, as I’ve said, are severely underfunded. 

3. We used to have too little data; now we have too much 

Everyone is talking about the overwhelming volume of data, at public transit agencies in particular. There’s an increasing need for solutions that can organize that data and use it to improve the product. Transit agencies can learn a lot from the big data analytics and UI/UX design now becoming standard in the tech sector. That could largely be the job of consulting groups like Remix, but transit agencies have to provide the data. There’s a potential role for government, in that a key step is standardizing data across transit agencies. 

[1] There are a few minor tweaks like making funding available for a wider range of modes, but this does not help the fact that the overall level of funding is insufficient and could even worsen the funding problem, since the money could be spread more thinly.

Why jitneys were regulated out of existence, and why Uber/Lyft will not suffer the same fate

If you’re interested in macro trends, Mary Meeker’s recently released internet trends report has a lot of good stuff. One thing that jumped out to me is this statistic: in 1915 jitneys in L.A. gave 150,000 rides a day. Today in L.A., Uber does 157,000 rides/day. Uber is as big now (in percentage terms bigger) as the jitneys were during the “jitney craze.” 

Of course, only five years after they appeared on the scene, the jitneys were regulated out of existence. Uber itself likes to promote this story as a cautionary tale against regulation today. But although the regulations for so called TNCs are still being hammered out, the regulatory scene today is far different than it was in 1915. So far it’s pretty clear Uber, Lyft, and others will not be regulated out of existence this time. 

Why the different outcomes? Why did jitneys get outlawed so quickly, and why are Uber and Lyft escaping the same fate? 

First I should reiterate that ridesourcing and jitneys have a lot in common. They emerged suddenly, making use of a new technology (automobile; smartphone). Both responded to demand by using flexible routes and pricing, and both were made possible by a supply of cheap labor. Both were a hit among customers and made their way into popular culture. Neither initially fit into existing regulatory categories. 

1. Jitneys probably created more directly obvious problems

Despite the similarity, the two forms of transport differ in important ways that matter for regulations. Compared to Uber and Lyft, the negative impacts of jitneys were probably more immediately obvious to the public. Once jitneys became ubiquitous in major cities, people blamed them for traffic congestion, accidents, and general chaos. Because jitney drivers’ income depended directly on the number of passengers served, it was not uncommon for jitneys to race each other on the street to pick up the next passenger. You can imagine what that did for road safety in an era with many fewer traffic laws. Uber is also controversial, of course, but not in a way that so directly impacts urban life. 

2. Central vs. fragmented organization

The more important differentiating factors, though, are organizational and political. Whereas Uber and Lyft are central companies that recruit independent drivers, jitneys started out as independent drivers. Although some entrepreneurs owned and operated entire fleets of jitneys, the industry was still very fragmented. The industry mainly consisted of small operators with limited political influence. Many jitney operators worked only part-time or for temporary periods and did not have enough vested to fight political battles. There was no centralized corporation as there is in Uber’s case. 

The jitneys were thus ill-prepared for the regulatory pressure brought against them. In 1915 streetcar companies held the political weight. Streetcars then dominated urban transport, and city and state governments depended on them for urban expansion and hence economic growth. Local officials often had close relationships with streetcar company leaders. Streetcar companies thus had the political influence to impose harsh regulations on their new competitors.

In contrast, ridesourcing companies have developed a structure to offer flexibility through independent contractors while maintaining the organizational and political advantages of a central corporate decisionmaker. Their competitors and political opponents, the taxi industry, were fragmented and politically weak, especially in California where the battles first broke out. The political weight was on the side of Uber and Lyft, and they have been able to influence regulatory policy much more than the jitneys ever could. 

I agree it probably was a mistake for cities to prohibit shared vehicles and hence jitneys early in the 20th century. Today, I believe we’ve already turned the corner on regulations for ridesourcing. Uber and Lyft, or services like them, are here to stay and the task now is to figure out what role policy plays in making sure they serve the public interest.