Ridesharing apps have made it easier to get home after a night out without driving, and that appears to be saving lives. Using Uber’s own trip data, two economists estimate that ridesharing has reduced traffic fatalities in the United States by 5.2% in the areas where it operates.
The study is by Michael L. Anderson of the Department of Agricultural and Resource Economics at the University of California, Berkeley, and Lucas W. Davis of Berkeley’s Haas School of Business. It was published online by The Review of Economics and Statistics on October 27, 2023, under the title “Uber and Traffic Fatalities.”
Why earlier studies disagreed
Previous research on ridesharing and road deaths reached conflicting conclusions. Most of it relied on the dates when Uber launched in each city. In an earlier working paper version, the authors showed that a simple “Uber is available” indicator explains less than 3% of the variation in actual rideshare activity across census tracts, because use varies so much within and between cities. When they repeated that approach, they too got inconsistent and mostly insignificant results.
So they used something better: proprietary Uber data on monthly trip counts for every census tract in the country, excluding Seattle and New York City, covering July 2012 to January 2017. Airport trips were left out. The data also showed what share of a tract’s trips happened between Friday at 8 PM and Sunday at 4 AM, when drunk driving is most common.
They merged these trip counts with the National Highway Traffic Safety Administration’s Fatality Analysis Reporting System, which records the location and date of every fatal crash, and compared tracts where ridesharing grew quickly with tracts where it grew slowly.
Fewer deaths at night and on weekends
With the detailed data, the effect was consistently negative. Ridesharing reduced traffic fatalities, and the impact was concentrated during nights and weekends, the hours when people are most likely to be driving home from bars, restaurants and parties. That timing is what the authors expected if rideshares are replacing trips that would otherwise end with someone driving impaired.
Checks for pre-existing trends found no sign that tracts with fast future ridesharing growth already had different fatality patterns before Uber arrived.
Overall, the authors conclude that ridesharing has decreased traffic deaths by 5.2% where it operates. Valued in the standard way economists put a price on lives saved, the annual benefit comes to $6.8 billion.
A benefit as large as the business itself
The authors compare that figure with the value Uber creates in the market. Back-of-the-envelope calculations, they write, suggest the life-saving benefits are of similar magnitude to the surplus captured by Uber shareholders or by Uber riders.
There is one notable gap. The data cover Uber only, not competitors such as Lyft, which the working paper puts at roughly 6% of Uber’s US bookings in 2015 and 14% in 2016. The lives saved by all ridesharing would therefore be somewhat higher than the Uber-only estimate.
Study Details:
- Title: Uber and Traffic Fatalities
- Authors: Michael L. Anderson, Lucas W. Davis
- Journal: The Review of Economics and Statistics
- Publication Date: October 27, 2023
- DOI: 10.1162/rest_a_01385