A recent study by the National Bureau of Economic Research (NBER), a nonprofit research group focused on public policy and business matters, found that passengers with "African-American sounding names" were more than twice as likely than those with "white sounding names" to have their ride canceled by Uber drivers. Those same riders also had to wait 35 percent longer than passengers with "white sounding names" for rides from Lyft. (The difference between "white sounding" and "African-American sounding names" for the purposes of this study was taken from a 2004 paper by other researchers, which cited "names such as DeShawn, Tyrone, Reginald, Shanice, Precious, Kiara, and Deja" as examples of black names.)
When it came to gender, the NBER report—which was co-authored by researchers at the Massachusetts Institute of Technology, Stanford, and the University of Washington—found that women were often taken on longer, more expensive routes, than necessary when riding with Uber.
It should be noted that it's not just app-based ridesharing services that have these problems: previous research and anecdotal evidence has shown discrimination by taxicab drivers against riders based on race and sex.
For this new study, researchers chronicled 1,500 Uber and Lyft cab rides over a two-year-period in cities on both coasts: Seattle, Washington and Boston, Massachusetts.
In Seattle, they found that African-American passengers were forced to wait 35 percent longer than white passengers when hailing a Lyft or Uber cab. In Boston, riders hailing Ubers with African-American sounding names were more than twice as likely to have their trip canceled than if they requested a ride with a white sounding name, despite the fact that Uber penalizes drivers for canceling trips too often.
Worse still, these passengers were more than three times likelier than white riders to have their trip nixed if they were trying to get picked up in a low-density area—just the type of place that suffers from poor public transit and where an Uber or Lyft is extremely helpful to get around.
It should be noted that Uber cabbies don't see the name of the passenger until they've accepted the fare, which is why they have to cancel the ride if they choose not to pick the person up. Lyft drivers, on the other hand, do see the name and sometimes even a profile photo of the person hailing a ride, before they accept the fare. So while Lyft drivers canceled rides far less often than Uber drivers, it's possible they were screening out African-American passengers before they even accepted fares.
Women also endured discrimination at the hands of cab drivers, albeit in a different way. In Boston, female passengers picked up by Uber drivers were often taken on much longer drives than needed. One passenger reported being carted onto the freeway for multiple exits, when her stop was a mere mile away. Not only did these women have to shell out more cash for what should've been a short trip, but they were often forced to listen to chatterbox drivers—uncomfortably so. The researchers described this behavior in the paper as a "combination of profiteering and flirting to a captive audience."
Uber, which has a less than stellar reputation on matters of discrimination against the disabled, treatment of its employees, and rider safety, defended itself with the following statement: "Discrimination has no place in society and no place on Uber. We believe Uber is helping reduce transportation inequities across the board."
The company did, however, suggest it could improve, also stating "studies like this one are helpful in thinking about how we can do even more."
The researchers behind the study noted that fixes to systemic problems of discrimination won't be easy, not even with the "move fast" ethos espoused by Silicon Valley companies. Taking out passengers' names from trip bookings, however, could remedy the immediate problem. The trick with this solution, however, is that it could end up confusing the driver and messing with pickups.
Overall, the study counters one of the big selling points of app-based ride-hailing services—namely, that people can get rides on demand no matter the color of their skin or their appearance.
Get six of our favorite Motherboard stories every day by signing up for our newsletter.