When Uber and Lyft first launched, the vision was that they’d help get drivers out of their cars, which would ultimately reduce traffic on city streets. A new study using detailed data from San Francisco says that the opposite is true: Traffic congestion has gotten significantly worse in the city over the past few years, and ride-hailing companies are a large part of the reason.
“San Francisco has been at the center of a lot of the evolution and revolution in the transportation industry and technology,” says Joe Castiglione, deputy director for technology, data, and analysis at San Francisco County Transportation Authority, the organization that worked with academic researchers on the new study. “When [ride-hailing companies] showed up on the scene, back around 2011, I think there was a tremendous amount of enthusiasm around having this new, convenient, affordable mobility option. Flash forward five or six years later, and people were saying, ‘it really feels like there’s a lot more congestion in the city.'”
The agency, which has the responsibility of managing traffic congestion in San Francisco, wanted to find out if ride-hailing vehicles were to blame. The agency started by trying to get data on each ride, which companies didn’t want to fully divulge (the California Public Utilities Commission, which also has that data, wasn’t willing to share it either.) But computer scientists at Northeastern University were able to build a profile of each trip using the limited data that was available.
The research revealed that most trips are happening in the most congested parts of the city, at the most congested time of day. The new study looks at the next question: Vehicle “hours of delay,” or the time people spend stuck in traffic compared to traveling on free-flowing roads, increased 62% in the city from 2010 to 2016. But since the city and jobs have also been quickly growing, how much of the traffic is coming from those factors compared to ride-hailing? A lot of it, it turns out.
In the past, some studies had suggested that Uber and Lyft could reduce traffic by helping more people get to public transit stations, or decide not to own cars, or by encouraging people to share rides instead of driving alone. But those studies were theoretical. “There are good reasons that if those things were true, then they actually could reduce congestion,” says Gregory Erhardt, the lead author of the new study and an engineering professor at the University of Kentucky. “The problem is when we look at the data, we find that those aren’t happening enough to offset the ways in which [ride-hailing companies] increase congestion.” Car ownership hasn’t really shifted downward in San Francisco. Data from other cities suggests that only a minority of riders choose to share rides.
“The main thing that distinguishes this study from much of the past research is it does have this unique data set to base this analysis on that is much more detailed, in both space and in time, than anything that’s been available in the past,” says Erhardt.
The study compares the actual changes in traffic to what might have happened without Uber and Lyft by using a transportation planning model. “Essentially, it’s sort of a real-life version of Sim City,” he says. “It simulates where people live, where they go to work, and how they move throughout the city. And this particular one has been in use in San Francisco for about 15 years now.”
There’s a caveat: The model doesn’t fully account for the growth in delivery vehicles from companies like Amazon, something that Uber and Lyft were quick to point out when the SFCTA first issued a report based on the study last fall. Still, the study has now been validated through the peer review process, and Erhardt points out that it’s possible that delivery vehicles have also reduced some other local trips to the store. That data isn’t yet available. Traffic also increased most in the areas where ride-hailing trips are happening most often.
Local government can begin to use the data to find solutions. The Transportation Authority is taking another look at the idea of congestion pricing downtown–not specifically to target Uber and Lyft, but as one potential strategy to reduce traffic. The study found that picking up and dropping off passengers is one contributor to congestion, so the city could also look at new strategies to change what can happen at the curb. In November, San Francisco will also have the chance to vote on a new tax for ride-hailing companies. The study might also be useful for other cities that have similar characteristics, like a dense downtown and strong public transportation.
“There’s a whole set of kind of policy recommendations and policy implications for the research,” says Castiglione. “I think those recommendations will be on a much more solid foundation because they’re not based on people’s fears or their fantasies but rather based on actual facts.”