IN 2017, a man with the first name Ramon was driving for Uber in the suburbs of Atlanta. He worked the late shift, ferrying intoxicated passengers around, often until early in the morning. Ramon’s cab likely reeked of alcohol from his myriad riders. During one of these late-night rides, a driver swerved into his lane. Ramon jerked the car to one side in an attempt to avoid a crash. The riders came to the conclusion that Ramon was driving drunk and reported it in their review. Ramon was immediately deactivated from Uber. He had effectively been fired on the spot by an algorithm.

Ramon is diabetic, which makes it very difficult (though not impossible) to drink to intoxication. Being fired for a debatable mistake mattered to Ramon because driving a cab was his livelihood. In fact, he drove so much, he wouldn’t be faulted for believing that he worked for Uber and that Uber was his employer.

As sociologist Alex Rosenblat explains in her new investigative book, Uberland: How Algorithms Are Rewriting the Rules of Work, many other factors would give Ramon this impression too:

The company determines the types of cars that are eligible on its platform, and it sometimes modifies the list of acceptable types at will; sets and changes the pay rates as it wishes; controls the dispatch; targets drivers unevenly with incentives; retains the full power to suspend or fire drivers without recourse; and mediates and resolves conflicts at its discretion.

However, Uber doesn’t consider Ramon to be an employee.

In fact, Uber considers its drivers to be everything but employees. They are simultaneously customers of Uber’s proprietary software and private contractors providing the company a service (they also provide a service to another group of Uber’s customers, the riders). If this seems confusing, uroboric, or like a contortionist’s exercise in semantics, it is.

So what’s really going on? Are Uber drivers employees, clients, or customers? This is an important question that Rosenblat’s new book asks. And for drivers like Ramon, finding an answer to that question is imperative. Do they deserve the fundamental protections that employees get in this country, like the right to organize, workers’ compensation, minimum wage, or any number of other rights?

Rosenblat’s book is a combination of sociological analysis, excerpts from Uber-driver online forums, communications with Uber executives and employees, and an avalanche of in-person interviews with drivers from all over the United States and Canada. Her analysis isn’t a polemic; it is balanced and measured. To answer the question of Uber drivers’ true place in the economy, she asks us to question the official story that Uber tells us about itself. Uber, at different times, has asserted it is a technology company, a logistics firm (whatever that means), and a broker between two parties. It seems to be everything but a taxi-cab company.

Questioning the official story that Uber tells is important because the stories frame our understanding and guide our expectations about companies. They can affect how those companies are regulated, bought, and sold. Snapchat once tried to tell the story that it was a camera company, to distract from the fact that it was truly a social media company, one who looked much more puny compared to Facebook and Twitter instead of Kodak and Nikon. The story that Uber tells is one of a technology company that only owns a proprietary software and algorithm. It has few employees and fewer assets.

Owning no assets and having few employees has become the holy grail of business in the last 50 years. It is what we call the “lean corporation.” Assets like factories and stockpiles of goods are a drag. They grant the employees who work these assets great power, because the employees are necessary to use them. Being a “technology” company theoretically avoids all that. Software requires close to zero space to store and zero cost to ship. Software’s great promise was the abolition of necessary employees.

And so it came to be in the 1980s that being a technology company, as opposed to a huge manufacturing corporation like General Motors, was the best way to sell yourself. Simultaneously, and really for the same reason, corporations began hiring temps instead of full-time employees, who are entitled to all kinds of protections dreamed up in the 1930s in response to the Great Depression, and by the 1950s came to expect pensions and vacation time. For lean corporations, the fewer full-time employees, the better.

Uber sells this vision to investors, for whom Uber often seems like a technology company because it owns so few assets and has so few employees. Uber is the proud owner of zero cabs. Investors think of Uber as simply a broker between two parties: riders and drivers. As co-founder Travis Kalanick once said, “We are not setting the price. The market is setting the price.”

But Rosenblat shows us that the everyday experience of a driver would make you question whether Uber is an impartial broker in a free market. When drivers get a notification they have received a ride, Uber gives them 15 seconds to respond. If they don’t respond (because they are say, in traffic and legally can’t use their phones) they are considered to have declined the ride. Drivers must accept over 85 percent of the rides offered or face deactivation.

A driver named Tim tells Rosenblat about how Uber will give him a bonus if he accepts over 90 percent of the rides in a shift. He loves the flexibility of his schedule but every time he wants to quit for the night or go pick up his daughter, often another ride pops up. If he declines the ride, he will blow his bonus and might not even cover his expenses of driving for that night. While Uber once advertised that drivers might make up $90,000 a year, Rosenblat cites studies that indicate most drivers are making way less, closer to $10 an hour.

Another problem is that Uber doesn’t let him see the details of the ride before he accepts it. He might have to drive 45 minutes to pick up a rider (which drivers refer to as “deadhead” rides because they don’t get paid for the time they spend getting to a client), and then drive them to an area that has relatively few riders, which forces him to make another uncompensated drive back to a busier area. Uber drivers end up footing the cost of the extra gas, because Uber wants to ensure riders can get a fare from anywhere. Uber does this because it knows that if it left the decision up to drivers, the market might deny some riders service. Kalanick is breathtakingly wrong: to serve one of its customers (the riders) Uber needs to regulate and totally control its market.

But is it possible that Uber drivers are independent contractors? In New York, for example, a worker must have certain freedoms and responsibilities to be considered self-employed. They must be able to set their own schedule, to negotiate their own rate, and pay for and choose their own tools. Uber drivers don’t set the rates for riders or get to choose what vehicles they use. So is their flexible schedule, their theoretical ability to log in and out of work, enough to qualify them as self-employed?

Many workplaces allow their workers flexibility in where and when they work, while still managing to pay them as employees. In a paper called “Does Uber Redefine the Firm?” law scholar Julia Tomassetti argues that companies like Uber aren’t really that new or different. They are founded on what she calls “regulatory arbitrage.” Uber is regulated like a tech company, while getting the profits of transit firms. This has opened up a fissure in labor law, which Uber, for the time being, has exploited. But algorithmic management is still management. Uber’s story obscures what it really is: an international taxi company with thousands of employees.

Another part of Uber’s story about being a technology company is its algorithm, which manages a great number of people and imposes strict rules on their conduct during work. In Ramon’s case, being terminated by an algorithm was equally consequential as being fired by a human. So algorithmic management doesn’t fundamentally change the employer/employee relationship. If you are managed by an algorithm, there is nothing to suggest you are less of an employee.

Uber isn’t really the problem. It provides workers with what Rosenblat calls “good bad jobs”: part-time work for those whose full-time income is not enough. However, this influx of workers needing a second job because their first job is unstable destabilizes the field of cab driving for those who intend to drive full-time. It’s a vicious cycle. Labor historian Louis Hyman has referred to Uber as the “waste product of the service economy,” which is pretty much right on, but it is also a waste producer. Uber is a trough that feeds those who can’t get enough to eat at home, but leaves those who rely on it exclusively starving.

Rosenblat’s book is rife with many testimonials from Uber drivers who truly enjoy driving for the company. It shows that people need Uber, and if you don’t mind sitting down all day and talking with strangers, it’s great. The problem is the misclassification of drivers as customers and independent contractors. It’s important to remember that Uber’s misclassification of workers is contingent on the public, investors, and the courts thinking that Uber is so new and different that it is exempt from the laws that govern this country.

Uber’s necessity springs from 50 years of stagnant wages — a failure of our political and economic system. It grew fat by feasting on the precarity that wage stagnation wrought. Rosenblat’s book urges us to wipe the techno-enthusiasm from our eyes and see Uber isn’t much different than any other major corporation with obligations to scores of employees.

¤

Robin Kaiser-Schatzlein is a journalist living in Brooklyn, New York.