In 2015, Uber cut the rates it paid per mile and per minute. “LOWER PRICES = HIGHER EARNINGS,” the company announced to incredulous drivers, who already had to foot the bill for fuel, maintenance, car leases, and benefits like health care.
Uber insisted that the cost-per-ride cuts would cause passenger demand to rise, and that the increase would translate to higher earnings for its gig workers. But drivers recoiled at this “Uber math”—the less-than-affectionate term some started using to describe the company’s communications—recognizing that the equation really meant that they’d have to do more trips and work longer hours to see the alleged boost. This certainly wouldn’t be the last time that Uber would push drivers to up their workload either.
For ride-hail drivers, such pressures have become routine. The companies make regular changes to pay and conditions, often with the effect of squeezing more out of their drivers. Uber is known to intervene based on the app’s tracking of driver braking and acceleration, ride acceptance and cancellation rates, and, of course, passenger-sourced ratings. For years, ride-hail apps have also sent out frequent nudges to incentivize workers to get or keep on the road. For example, when a tired Uber driver working near Penn State University tried to go offline at 4 a.m. one morning in 2015, a playful in-app notification from Uber popped up to say, “YOU’RE $1 AWAY FROM EARNING $40. Don’t stop now, keep driving!” He posted the notification to a driver forum, commenting, “They better cough up $40 or there will be heck to pay!” He continued working until he received another ride request.
So when, two weeks ago, Uber announced a new policy limiting drivers to 12-hour shifts in the name of helping to “keep riders and drivers safe on the road while preserving the flexibility drivers tell us they love,” its stated motivation struck me as only part of the story. Since 2014, I’ve been studying Uber from the perspective of the drivers. I’ve made daily visits to online forums where drivers post updates about their work, traveled to interview local drivers in more than 25 cities in the U.S. and Canada, and kept up with its corporate policies and communications. While Uber may have taken a step to get drowsy drivers off the road in the name of public safety, it obscures the fact that it still pushes its drivers in other concerning ways.
Uber’s official new policy states that drivers who engage in a total of 12 hours of driving time will be shown a prompt requiring them to go offline for a six-hour break. As an Uber executive told the Washington Post, the app will use GPS and telematics to measure an individual’s cumulative driving time. The app won’t count long stops, such as waiting for a ride request in an airport parking lot (a policy that highlights the fact that Uber doesn’t pay workers for all of their time on the clock), but will count short traffic stops and multiple stints in a day—say, three four-hour shifts—taken without a six-hour rest in between. It will also give drivers periodic notifications when a driver is close to hitting the 12-hour max before automatically going offline. It’s similar to restricted hours Uber had already rolled out in New York City and to caps implemented by the NYC Taxi & Limousine Commission to prevent driver fatigue.
Drivers have had mixed responses to this change. For example, a driver I recently interviewed who works for Uber and Lyft in Boston said that she thought it was a good policy. She had started driving part time for the ride-hail apps last year to supplement her income working full time at a restaurant, but said she soon found the ride-hail work less exhausting and more flexible than her regular job. She suggested that the new rules might get individuals who start working before the morning rush at 5 or 6 a.m. and drive nonstop late into the night to take a break sooner. “It’s the human body, you get tired, you may doze off a little … and that could be dangerous for everyone involved,” she told me. She said she thought drivers could make up for the lost income by better timing the hours they picked to be on the road.
On an online driver forum, Ethan, a part-time driver who’s been working for Uber and Lyft for a little more than a year in and around Burlington, Vermont, posted, “From a safety standpoint this is important. My honest opinion is that people deep down aren’t really pissed about not being able to work this much. They’re pissed because they HAVE to work this much to make ends meet and can’t now.”
Considering that, as independent contractors, drivers aren’t entitled to a minimum wage, benefits, or reimbursements for costs, this squeeze can be very real. A recent study released by MIT of more than 1,100 Uber and Lyft drivers in 2017 seem to bolster this point: The authors determined that 74 percent of those surveyed earned less than the minimum wage in their state. What’s more, the researchers also determined that a startling 30 percent were actually losing money once vehicle expenses were included. The study’s methodology has since been disputed by Uber’s chief economist, Jonathan Hall, who wrote that the paper’s figures put earnings at “less than half of the hourly earnings reported in the very survey the paper derives its data from,” which estimated that drivers had hourly earnings of $15.68 before expenses and taxes. Hall said that Uber did not take issue with the MIT paper’s estimation of drivers’ costs, but said that the authors erred in their interpretation of survey questions in a way that created “deeply flawed” earnings numbers. The lead author of the study, Stephen Zoepf, told Reuters he will rerun his analysis. Regardless of the outcome, this isn’t the first time Uber’s been implicated in disputes over what drivers earn. In 2017, Uber agreed to pay a $20 million Federal Trade Commission settlement over charges that it recruited drivers with exaggerated earnings claims. [Update, March 12, 2018: Zoepf released a formal statement in which he addressed Hall’s criticism and reran the numbers accordingly. Based on two possible methodologies, Zoepf found either 54 percent or 41 percent of drivers surveyed made less than the minimum wage in their state ($8.55 or $10.00 per hour respectively), and that 8 percent or 4 percent of drivers lose money.]
In its release about the 12-hour driving cap, Uber suggested that the new rule won’t affect the “nearly 60 percent of U.S. drivers [who] use Uber less than 10 hours a week.” But the statement fails to acknowledge that the change will hit Uber’s most invested drivers the hardest. While it’s true that most drivers work part time, a minority of them do work full time or longer—and do so without overtime pay. They’re the individuals most likely to rely on the ride-hailing apps as their primary income, and the most likely to make investments, like buying a new car, to do this work. They’re also the most affected by compensation cuts, hence pushing themselves to work dangerously long hours to make up for the losses. Some part-timers who devote long hours to intense weekend shifts after working at their primary jobs during the week will be hit too. Plus, migratory workers who commute from their homes into a city with more work—some of whom sleep over in supermarket and airport parking lots to make ends meet—will certainly feel the impact when they can’t log in to work but can’t easily go home either.
There’s also the question of whether such policies will even be effective. Some drivers running up against Uber’s 12-hour limit quickly found an easy way around the system: going to competitors. Many Uber drivers had already been working for rival platforms like Lyft, toggling between the apps depending on demand and rates (a fact that also suggests any company’s claim that the majority of their gig workers only drive part time deserves greater scrutiny—many are essentially cobbling together full-time hours from multiple part-time ride-hail jobs). Uber may limit its drivers to 12-hour Uber shifts. But if drivers can simply switch apps in a patchwork job market, the policy may do little to curb drowsy driving—especially if they still feel financial pressures to stay on the clock.
The Boston area driver I interviewed who praised the change, for example, told me she hasn’t used Uber for 12 straight hours in some time. But she said she has definitely worked to the limit for Lyft, which has a policy that requires drivers to take a six-hour break after 14 hours in driver mode. David Aguirre, an Uber driver in Houston who started driving as a side gig but switched to full time after the company he worked for went under, said he also already felt the pull of putting in just a few more hours. He told me that the first time he reached the 12-hour limit under Uber’s new rules, he thought that it was a good policy to help keep drivers and passengers safe. But he admitted, “That day I just wanted to keep going a little longer to get to $500 for the day.” He said one of the first things he did after the automatic timeout was to apply to drive for Lyft.
Karen Levy, an assistant professor in the Department of Information Science at Cornell University (who’s also affiliated with Data & Society, where I work as a researcher) found a similar conundrum in her research with truckers. For years, they have technically been restricted from working too long by law—and, in some cases, are also electronically monitored for enforcement—but found ways to skirt it in practice. “Truckers don’t work without sleep for dangerously long stretches (as many acknowledge having done) because it’s fun,” she wrote in an op-ed for the Los Angeles Times, “They do it because they have to earn a living. The market demands a pace of work that many drivers say is impossible to meet if they’re ‘driving legal.’ ”
I certainly agree with Uber’s stated goal of making the roads safer for drivers, passengers, and others. But, short of implementing troublingly invasive monitoring of off-app drivers (which, considering Uber’s shady history of secretly tracking drivers working for competitors, may not be a far-fetched scenario) or giving drivers fairer compensation, Uber and Lyft won’t be doing enough to significantly address the safety issues that come with overworked drivers. Instead, the new policies respond to public pressure about the problem by shifting a shared liability for these dangers onto its drivers. The companies now bear less of the responsibility if a driver goes beyond the shift restrictions—the app timed out! The blame falls entirely on the worker, who managed their time poorly, who irresponsibly circumvented the safety feature.
These moves from Uber and Lyft seem to align with their gig-economy model of employment, which structures work as an individual pursuit and individual liability. But even this sell is misleading. While, for many drivers, the idea of being independent at work is very appealing, their ability to make entrepreneurial decisions is consistently constrained by the ride-hail apps’ nudges and other algorithmic management, rules, external costs, and wage cuts.
Perhaps the shortened amount of hours they can get out of a single driver might spur Uber and Lyft, who often compete for workers as well as passengers, to offer more attractive compensation. Or perhaps the poor pay and new limits might tip the scales for some superdrivers to trade the flexibility the ride-hail apps offer for other types of work. There’s only so long they can take the “Uber math.”