Laura Bliss is CityLab’s west coast bureau chief, covering transportation and technology. She also authors MapLab, a biweekly newsletter about maps (subscribe here). Her work has appeared in the New York Times, The Atlantic, Los Angeles magazine, and beyond.
Since ride-hailing is putting more cars on the roads, one study finds that mobility services are also contributing to the rising number of fatal crashes.
Every convenience has its external costs, and economists love to count them.
With ride-hailing, the positive benefits to users, and society at large, are several. On-demand services like Uber and Lyft provide transportation in neighborhoods underserved by transit or taxis; safe rides home for late-night workers and partiers; and increased ease of access for people with disabilities, research has shown. There’s also some evidence that they reduce drunk driving. Besides providing a service to millions of riders, these companies are increasing access to safe mobility for groups that haven’t always had it.
But the growing demand for press-a-button transportation is coming from everybody, not just underserved riders. In cities, that is translating into more cars on the road. While the environmental and congestion impacts of this surge in vehicles have been much discussed, the potential uptick in traffic fatalities associated with Uber and Lyft usage has been one of ride-hailing’s lesser-studied negative externalities. In a new working paper, a team from the University of Chicago’s Booth School of Business is now attempting to pin this down. The authors estimate that 2 to 3 percent of the number of crashes in a given area can be attributed to the introduction of ride-hailing.
Here’s how they structured their research. Using the rollout date of Uber and Lyft’s various services in every city with a population of at least 10,000 in the U.S., they measured scale of adoption in those areas (based on Google search volumes for the companies’ names). They also took accident data from the National Highway Traffic Safety Administration and looked at changes in the number of collisions around the advent of ride-hailing. They found that fatal traffic crashes sharply tick up around the time that app-based rides were introduced into a given area, consistent with declines in gas prices, more vehicle-miles traveled (VMT), and increased traffic delays.
Their results stayed strong when the researchers controlled for time and place; rural areas saw a scant Uber/Lyft effect on their traffic death tolls, while denser cities with higher app adoption rates felt it more.
To be clear, the paper is not suggesting that ride-hailing drivers are getting into fatal crashes more than any other people on the road. Although the authors briefly mull how the average app driver might compare to the general population in terms of safety (and at least one report has suggested ride-hailing drivers are actually less likely to speed, drive aggressively, or fumble with their phones than civilians), that’s not what this new study measures. The finding is more straightforward: To the extent that these apps are likely putting more cars on the road, they’re probably also implicated in the number of deaths by cars.
“The fact is that ride-sharing allows for more VMT, and excess fuel consumption,” said John Barrios, a finance professor at the Booth School and one of the lead authors. “The technology makes it easier to be in a car and use it more than you would have before. The key thing is that there are externalities to that.”
And those externalities are getting worse. U.S. road fatalities rose to 35,000 in 2015, a 7.2 percent increase over 2014, representing the largest jump in 50 years. In 2016, the death toll increased another 5.6 percent. Preliminary data suggests that number leveled out in 2017, but that’s still about as high as it’s been in a decade.
There is some debate over the biggest driver of these gruesome numbers, especially about the extent to which digital devices are pulling our eyes off the road. But the most obvious suspect is cheap gas, which is encouraging folks to do more driving, and in some places, to buy more cars where they might have used transit. Economic growth is also a recent contributor to the rise in VMT. In short, the more Americans drive, the more opportunities they have to kill someone. That trend is mostly about private vehicles; apps are a small part of it, the new paper is saying.
The study is not perfect. For example, Uber and Lyft are famously loathe to release ridership data, so the study’s measurements use Google searches as proxies for levels of app adoption across the country. That’s not ideal. Also, a more robust version of this paper could include a narrow look at whether ride-hailing is linked to more deaths at the times and places when the service is especially popular, including late at night and around airports, as City Observatory’s Joe Cortright points out.
Both companies dispute the findings. “Uber has contributed to safety in many ways and we take our responsibility to help keep people safe seriously,” an Uber spokesperson told Business Insider. On Streetsblog, a representative for Lyft called the study “deeply flawed,” and said that “numerous studies have shown that rideshare has reduced DUIs, provided safe transportation in areas underserved by other options, and dramatically improved mobility in cities.”
That may be, but those benefits are all linked to two simple facts: Ride-hailing is growing, and it happens in cars. No matter who’s driving them, until there are fewer vehicles on the road—thanks to market forces, or regulations like the new “Uber cap” in New York City—society probably can’t expect to stop losing lives. Now there are some cold economics.