Why are travel speeds in Manhattan dropping?

There’s a kind of ominous tidbit in New York City’s most recent mobility report.  Travel on streets in Manhattan is getting slower.

Here’s the data, computed from the report, showing the average travel speed on streets south of 60th Street in Manhattan:

The difference between 9.35 miles per hour and 8.21 miles per hour doesn’t seem like a lot, but it works out to noticeably longer trips in this urban environment. If we take a 1.5 mile trip (the median taxi ride in New York is between 1 and 2 miles), the average trip now takes about a minute a twenty seconds longer than it did just a few years ago: about 11 minutes rather than 9 minutes and 40 seconds.  That delay affects a lot of people: there are about 360,000 taxi trips daily in Manhattan, and about 1.8 million people ride buses on the city streets, and are also affected by delay.

There are lots of reasons why travel might be getting slower, among them:

The economy is rebounding. More jobs in New York means more commuters, more shoppers, more business and more traffic.

There are more “for hire vehicles.”  The number of taxi trips is down, but that’s been more than offset by the huge increase in for-hire vehicles, especially those provided via ride-hailing transportation network companies (TNCs), like Uber and Lyft. Bruce Schaller’s report earlier this year confirms that the growth of ride-hailing has contributed materially to the decline in traffic speeds.

Interestingly, despite the rebounding economy, fewer vehicles are entering Manhattan. With fewer vehicles, it’s puzzling as to why travel speeds should be getting worse. The evidence here is at best circumstantial, but here’s a hypothesis: While fewer vehicles are coming into Manhattan, those that are coming are more likely to be taxis or other for-hire-vehicles looking for customers. They may be lingering longer.

If so, there’s a kind of tragedy of the commons at work here. Additional cars and drivers are drawn into Manhattan due to the dense concentration of potential fares and customers, enhanced by the prospect, at least in the case of TNCs like Uber and Lyft, of earning even more when prices are multiplied during “surge” periods. And neither taxis nor TNCs have a strong incentive to leave Manhattan (fewer fares, and at least for trips to New Jersey, the prospect of bridge tolls). As more and more cars concentrate in Manhattan, streets become even more clogged and traffic moves more slowly. And it isn’t just cars that are affected, of course, its the city’s buses as well. Slower bus transit times mean longer trips for bus passengers and lower efficiency for the transit system (driver costs, which are largely time based, are higher per passenger carried and per mile traveled when the buses move more slowly. City wide bus speeds have dropped 2 percent according to the Mobility Report. And bus ridership in New York has been declining since 2009.

Of course, what’s been happening below ground, in the New York subway has been very much in the news in the past couple of weeks. Charles Komanoff has estimated the congestion and delays on the subway cost its riders $1.4 billion a year. And there’s a connection: the slower buses go, the more likely it is people will take the subway. And that appears to be exactly what’s happened. Since 2010, the number of bus trips has declined by 46 million annually, while the number of subway trips has increased by 159 million per year.  One way to reduce the pressure on the subway system is to make buses more attractive.

The Transit Center has put forward a series of ideas for improving bus service and travel speeds in New York:  better signal timing, optimizing routes, automating fare collection, allowing all-door boarding, increasing deployment of “Select Bus Service”, with exclusive lanes for buses on some routes.  All of these would help, but for buses operating in mixed traffic, arterial travel speed may be the most important constraint on productivity improvements.

Ultimately, the only way to improve travel times on the street may be to price the roadway. Charging a per mile fee, adjusted to time of day could reduce traffic on busy streets, and enable buses to move faster. Faster buses would be both more productive (more passengers carried per driver hour) and more attractive to riders, potentially relieving pressure on the subway system.

And this isn’t just a New York City problem. The Seattle Transit blog reports that the city’s buses spend 18 percent of their time stopped in traffic, which produces schedule unreliability, and lowers the efficiency and raises the cost of operations.

Ride-hailing services like Uber and Lyft aren’t solely to blame for traffic congestion. But what they do is accentuate the problem, particularly in dense urban environments. Places with a high concentration of customers, the lure of surge pricing, and free streets will attract lots of TNC vehicles. This is exactly what’s happened in San Francisco, where as much as 20 percent of downtown traffic consists of ride-hailed vehicles. And with more traffic comes slower speeds, which imposes high costs on all road users–and importantly, undermines the productivity and attractiveness of transit buses. This is a problem that is only likely to become worse as ride-hailing grows.