Marianna Porras said she started driving for Uber in 2017 because, as a single mom, she was attracted by the flexibility it offered in terms of allowing her to make her own schedule.
At the time, the company took 25% of fares and was transparent about what it was taking and what Porras was getting, she said. She could drive a reasonable amount of time and make a decent living.
Not anymore — in the wake of the pandemic, Uber has steadily cut driver fares, she said. To make as much as she used to in a day, she has to drive double the hours. And the translation between what her customers pay and what she earns is now anything but clear.
“This is financial abuse,” Porras said.
Porras is far from the only ride-hailing driver complaining about declining wages. On Wednesday, drivers in San Francisco and around the country gathered for protests and strikes against Uber and Lyft. Drivers blocked off part of Third Street in front of Uber’s headquarters and dozens chanted outside its offices, shouting “Shame on you!” and demanding to speak with Uber CEO Dara Khosrowshahi.
Academic experts who have studied driver pay think they know why Porras and other drivers have seen their pay rates fall. The decline is due to something Veena Dubal, a law professor at UC Irvine who studies gig work and workers, dubbed “algorithmic wage discrimination” in a research paper last year.
That’s a fancy way of saying that the companies are determining what they pay drivers for trips based on what they know about the drivers. The companies may offer two drivers who are right next to each other significantly different rates for the exact same trip — a phenomenon drivers have actually documented.
Among drivers, “there is a real sense of outrage that their labor is being priced differently based on what the companies know about them and their behavior,” Dubal said. “These are pretty unprecedented practices in terms of the history of capitalism.”
When they launched, Uber and Lyft paid drivers and charged riders in much the same way traditional taxi services did — on per-minute and per-mile bases. The companies took a fixed percentage — traditionally 25% or so — of customer fares and drivers got the rest.
The companies did charge more during peak demand, but that was fairly transparent and didn’t affect the essential nature of the system, which was that the longer the ride in duration or distance, the more customers paid and the more drivers earned.
But the companies gradually moved away from that model, first announcing upfront pricing for customers and then, in the fall of 2022, moving to upfront payments for drivers.
Instead of charging customers or paying drivers based on the actual duration of a ride or distance driven, the companies offered a flat rate that wasn’t subject to adjustment. Although both companies say they include estimated time and distance in their driver-payment calculations, they both acknowledge that they use other factors, too, including demand for rides and incentives that may be applicable to particular drivers.
Lyft spokesman CJ Macklin declined to comment on whether the company takes into account what it knows about individual driver behavior when setting rates for particular trips.
In an email, Uber spokesman Zahid Arab said it was “patently false” that the company sets rates for individual drivers based on their “personal characteristics or past behavior on the platform.” Likewise, Uber doesn’t group drivers together by certain characteristics or offer different rates based on those group characteristics, he said.
But such statements appear to be at odds with what CEO Khosrowshahi said earlier this month on a call with investors and analysts after Uber announced its first-ever annual profit since going public. Based on the data it’s collected since it launched upfront pricing, Uber has realized “drivers are quite idiosyncratic in terms of their desire” for the types of trips they want to accept, he said.
“What we can do better is actually targeting different trips to different drivers based on their preferences or based on behavioral patterns that they’re showing us,” Khosroshahi said. “That really is the focus going forward, offering the right trip at the right price to the right driver.”
Already, he noted, with upfront fares, “you’ve gone from just flat time and distance to now kind of point estimates for every single trip based on the driver ... We’re making these point estimates both in mobility and delivery. We’re doing it globally.”
Meanwhile, in a blog post last year describing Uber’s new upfront payment model, Miriam Chaum, its head of work, economic policy and advanced technologies, indicated the company does in fact use what it knows about drivers to group them into different buckets. It just doesn’t classify drivers according to protected characteristics such as race or sex.
Ex // Top Stories
Torrey will play the Make Out Room as part of an album release party for their second full-length record
When people come to The City, London Breed opines, visitors won’t residents hanging their heads — they’ll find San Franciscans building the future
Oakland native Andrew Wilson’s solo exhibition at Jonathan Carver Moore gallery showcases a body of artwork 10 years in the making
“In order to continuously improve the Driver app and earnings offerings, we ... run tests of new features and models that may result in different fares for drivers across different groups, which are never based on protected characteristics,” Chaum said in the post.
Regardless of whether Uber and Lyft are using individualized or aggregated data on drivers to set their pay rates, they appear to be benefiting from the move to upfront pricing.
Uber’s revenue from its taxi service was up 41% last year from 2022 and the unit’s adjusted operating profit rose 50% to nearly $5 billion. Lyft’s sales were only up 7% year-over-year, but its annual loss shrank 79%, falling from $1.6 billion to $340 million.
But while the companies are benefiting, drivers are losing out, researchers say. With upfront pricing, the companies have made it impossible for drivers to know their pay rate or to have any certainty about that rate, Dubal said. And there’s the potential that the companies will be able to use even more personalized data, such as how much debt drivers are in or how much rent they owe, to determine what they pay drivers.
“It’s kind of like a sci-fi dystopia,” she said.
And there’s evidence beyond individual complaints that the use of such models is already affecting driver pay.
The average amount of money Uber drivers earned per trip was down 12% in the third quarter of 2023 from the same period in 2022, according to Len Sherman, an executive in residence and adjunct professor at Columbia Business School who examined data from Gridwise, which collects driver pay information. Meanwhile, Sherman estimates Uber’s take rate — its cut of customer fares — has risen to 40%.
Dubal, who interviewed drivers for her research paper, and Katie Wells — a researcher at Georgetown who tracked a group of 40 Uber drivers in Washington, D.C. for five years, starting in 2016 — have also found declining pay rates among the drivers they surveyed and interviewed.
That drop they’ve documented has come even as other employers were boosting pay in the wake of pandemic-related labor shortages and as unions representing auto workers, UPS delivery drivers, airline pilots and actors won big gains in wages in their latest rounds of contract negotiations.
“To me, this is incredibly striking that while labor unions have scored some pretty major successes and the only question is how big a raise did they get, not did they get a raise ... [Uber and Lyft drivers], and they alone, are hit with a double-digit pay cut,” Sherman said.
Both Uber and Lyft say that driver pay has actually increased rather than fallen. At the median, Lyft drivers made $30.68 per “engaged” hour, if tips and bonuses are included, according to Macklin, the company spokesman. That was up 18% from the second half of 2018, he said.
Additionally, Lyft announced earlier this month it would guarantee that drivers earned 70% of what’s left of their customer fares each week after accounting for taxes, commercial-drivers expenses and other fees.
In the fourth quarter of last year, Uber drivers made about $33 per “utilized” hour, according to company spokesman Arab. Over the last six years, median driver earnings for making trips, including tips and other incentives, is up 30%, he said. Uber’s share of customer fares has stayed flat; excluding commercial insurance costs, it’s “well below” 20%, he said.
By “engaged” and “utilized” hours, Lyft and Uber mean the time between when drivers accept and complete a trip. It doesn’t include their downtime between trips. And their pay is gross pay; it doesn’t include their own insurance, gas or maintenance costs.
Regardless, numerous drivers at the Wednesday protest said what they’ve experienced is the opposite of what Uber and Lyft claim. Their pay has declined from just a few years ago. They’re seeing significantly smaller portions of what customers are paying and they’re having to work longer and longer hours to make the same amounts.
Mahamed Algaheim said he’s seen examples in which Uber charges $45 for a ride to San Francisco International Airport, but the driver only gets $16, or rides for which it charges the customer $60 and offers to pay drivers $17.
“It’s getting out of hand,” said Algaheim, who’s been driving for Uber for three years. “It’s not even fair.”