The city of Seattle had been a source of irritation for Uber and Lyft for years. Last fall, that annoyance reached a peak. After passing a tax on all rides, Mayor Jenny Durkan announced a set of measures to help Uber and Lyft drivers.
Among them was likely a particular sore spot for the ride-hailing companies: a minimum wage for drivers.
To set the minimum wage, the largest city in the state of Washington first needed to know how much drivers actually make. It wasn't as simple as applying Seattle's current minimum wage of $16.39 because driving for Uber and Lyft is different from most other jobs. Drivers cover all their own expenses, including gas, insurance and vehicle maintenance, and those costs needed to be factored in. So, Seattle commissioned two well-known economists from the New School and the University of California, Berkeley to crunch the data.
Unbeknownst to the city and the economists, however, Uber and Lyft had quietly commissioned researchers from Cornell University in the hopes of telling another data-driven story. The companies said they paid the university a total of $120,000 for that work.
The two research teams came up with very different conclusions. The New School and UC Berkeley economists calculated that after paying expenses, Seattle ride-hail drivers made an average of about $9.73 per hour, according to their 80-page analysis released on July 6. The Cornell researchers put the number at $23.25, according to their 130-page report, also released on July 6.
"I was just amazed, outraged really," said Michael Reich, the UC Berkeley economist who worked on the study for Seattle. The Cornell report "just made my head spin. I couldn't believe it."
The divergent studies have set off an academic tussle involving PR blitzes, data-shaming and a sharply worded rebuttal. The Cornell report also prompted a critical internal memo from a professor at the university that raised questions about the school's involvement in research commissioned by private corporations.
Uber and Lyft may be young companies, but they had borrowed a page from a well-used corporate playbook. Companies have long paid academics to conduct studies that might influence policy makers and public opinion. Tobacco, pharmaceutical and chemical companies are notorious for it. Coca-Cola reportedly used the tactic to shift blame away from sugar as a contributor to obesity. Google allegedly spent millions of dollars to fund research papers that challenged antitrust regulations. If successful, these kinds of studies can lead to relaxed laws and a lot of saved money for the companies.
In the case of Uber and Lyft in Seattle, the right research could help the companies save millions of dollars by influencing the city's policy to avoid paying its current minimum wage. Likewise, for drivers, a minimum hourly pay that accurately takes into account expenses could be a lifeline at a time when they're risking their health because of COVID-19 to shuttle people around town. Seattle, which has about 30,000 ride-hail drivers, is expected to announce a decision on the driver-specific minimum wage in coming weeks.
Reich and his colleague, James Parrott, calculated that the minimum wage for Seattle drivers should be $28.19 per hour to account for expenses. The Cornell researchers, led by professor Louis Hyman, didn't come up with a number based on an equivalent analysis but instead concluded that 92% of drivers in the city already make more than the $16 minimum wage, even after deducting expenses.
Although Seattle hadn't requested the alternate analysis from Uber and Lyft, the companies said they anticipated regulators might still take it into account.
"We hope policymakers will take a fact-based approach as they consider new policy proposals by using the insights from Cornell's in-depth analysis of detailed data," an Uber spokesman said.
Cornell University didn't return a request for comment.
War of words
Over the past several months, Parrott and Reich worked on surveying nearly 7,400 drivers and making their computations. The Cornell researchers were quietly doing their own analysis using data on 14,000 drivers provided by Uber and Lyft. Less than a week before Parrott and Reich were to release their findings on July 6, Cornell's researchers sent their report to the city.
Chelsea Kellogg, a spokeswoman for the mayor's office, said they were all taken by surprise. "We only got a little bit of a heads-up," she said.
The opposing academic teams agreed to have a Zoom call to discuss the findings. But that didn't smooth things out. And, as the two reports were publicly released last Monday, both sides put up their defenses.
Parrott and Reich penned a five-point rebuttal to the Cornell study, spelling out the differences between the two reports. They said the Cornell researchers didn't add in all of the drivers' expenses, including vehicle cleaning, smartphone bills and car insurance. They also noted the Cornell study included tips in its earnings calculations, which isn't allowed under state and local laws. In addition, they said the Cornell research for the most part only took into account the time drivers were on a ride. Time spent waiting for rides, which can be as much as 50%, wasn't always factored in.
"If drivers are really being paid that well," Reich said, "there should be a lot more people wanting to be drivers."
Meanwhile, Uber and Lyft spent the last week pushing their narrative. Lyft emailed reporters a link to the Cornell study on July 7 saying it shows "drivers in Seattle make $23/hour." The company didn't mention the city's commissioned report. When CNET asked Uber and Lyft about the Parrott-Reich study, both companies said they had issues with the research.
"We had questions about the procurement process they used to choose the labor-affiliated researchers," said Julie Wood, a Lyft spokeswoman. Parrott and Reich are economists and aren't affiliated with any union. The pair was also tapped by New York City in 2018 to do a similar minimum wage study for ride-hail drivers.
Uber and Lyft pointed to the fact that Parrott and Reich were using data the economists collected from driver surveys, whereas the Cornell researchers were working with ride data given to them directly by the companies.
"Parrott and Reich's study is based on limited data and flawed assumptions about drivers' experiences that are unsupported by facts, evidence or reality," an Uber spokesman said. "Parrott and Reich throw the kitchen sink into their per-mile cost estimates to artificially depress driver net earnings estimates and wildly inflate the pay standard needed to ensure that drivers earn minimum wage."
Lyft's Wood said, "The fatally flawed combination of an anecdotal data set and a series of incorrect assumptions means this report's conclusions are divorced from reality."
Parrott and Reich said they surveyed drivers to gather their data because Uber and Lyft, citing confidentiality issues, declined to provide ride data when Seattle requested it. As of today, the economists say the companies still won't hand over that information.
"If the authors of the Uber-financed study had bothered to talk to even a single Seattle driver," said Nurayne Fofana, who has been a full-time ride-hail driver for four years and was surveyed by Parrott and Reich, "they would know that the vast majority of this work is done by Black and brown immigrant drivers, working long hours for low pay to support our families."
The lead author of the Cornell study, Hyman, said Uber and Lyft were "uncomfortable" providing the data directly to Seattle, so instead they gave it to his team. Hyman said that neither he nor the other researchers on the project received any of the $120,000. Instead it went to the university. He added that even though the study was commissioned by the companies, it didn't shape his team's research.
As far as the results of his report, Hyman said, "We expected to find full-time drivers making less than minimum wage. We were very surprised to see so many drivers working part-time and earning average wages."
Hyman stressed, however, that the $23.25 average hourly pay number the companies are touting shouldn't be the focus of his research. He said his team calculated the median hourly pay for full-time drivers at around $18, which means half the drivers were making more than that and half were making less.
"I know there's that number the comms people at Cornell and Uber and Lyft are pushing," he said. "For us, we really wanted to emphasize the variation."
Reputation at stake
As the dispute played out last week, an internal memo written by an unaffiliated Cornell professor circulated among the university's researchers involved in the study.
Citing "serious worries," the author of the memo, which was seen by CNET, said there were concerns Uber and Lyft will use the findings to promote their own interests. The author indicated that the department should be involved in research that furthers public policy, rather than "advance the monetary gains and power" of specific private companies.
This isn't the first time Cornell has been challenged over a report for ride-hailing companies. A publication based out of the university's Industrial and Labor Relations school had published a study in 2018 by an economist who works for Uber, Jonathan Hall, and the late labor economist Alan Krueger. That report examined driver demographics and earnings and said 81% of drivers were satisfied working for Uber. Those findings were criticized by other economists as being favorable toward the company.
In the memo, the Cornell professor said that after the Hall-Krueger study was published, the reputation of the ILR Review was "very seriously damaged."
According to various media reports, Uber has used the Hall-Krueger report (and earlier iterations of it) as a tool to establish itself in new cities and to influence regulators, including defending itself for keeping drivers classified as independent contractors.
After speaking with CNET and other reporters last week, Hyman emailed CNET on Saturday saying he realized his team should've written an introduction to its report to put its findings in context. That seven-page intro, which was added to the top of the study, emphasizes the variation in Seattle drivers' earnings.
"Our role is not to serve as defenders or apologists for the [ride-hailing companies]," the report's intro now reads. "Furthermore, these findings cannot be generalized to other cities, domestically or internationally."
As for Seattle's regulators, Durkan's office said it doesn't plan to use the Cornell study in determining its minimum wage for ride-hail drivers. Instead, the city intends to use the numbers from the report it commissioned from Parrott and Reich. This means Uber and Lyft will likely be required to pay drivers in the city a minimum wage of $28.19 per hour within coming weeks.
Kellogg, the spokeswoman for the mayor, said that given the economic hardships facing workers in the US right now, it's important to make sure ride-hail drivers earn a living wage.
"This study comes at a time when protecting workers without a safety net is even more critical," Kellogg said. "Seattle faces a global pandemic, historic financial downturn and is in the middle of a movement to undo centuries of systemic racism, including disparate access to economic opportunity experienced by people of color and immigrants."