Bits of Books - Books by Title


How Algorithms Are Rewriting the Rules of Work

Alex Rosenblat

(NY Magazine)

More books on Work

As much as we all hate bosses, you have to admit they have gotten slightly better over the years. Frederick Winslow Taylor, the odd and efficiency-obsessed father of management theory, was fond of dispatching managers to stand over workers with stopwatches and direct their every movement as if they were trained animals. Reacting to the obviously soul-crushing nature of Taylorism, a wave of touch-feely business gurus in the 1960s aimed to inspire people into becoming more productive. This has led, over the course of the last few decades, to the insipid corporate culture of team-building exercises and black-bordered posters of windsurfers at sunset, but at least there's free coffee now.

One thing you get from reading Alex Rosenblat's Uberland: How Algorithms Are Rewriting the Rules of Work, is that there is nothing inevitable about management trending in a positive direction. Drawing on four years of ethnographic research among Uber drivers, Rosenblat has produced a thoroughly dystopian report that details how millions of drivers are now managed by a computerized system that combines the hard authoritarianism of Frederick Winslow Taylor with the cynical cheerleading of Michael Scott.

But wait: Isn't the whole point of Uber that you can be your own boss? After all, Uber talks of its drivers not as employees but 'partners.' In its propaganda, Uber portrays itself not as a taxi company at all but a technology platform that connects drivers directly to riders. 'FREEDOM PAYS WEEKLY,' reads one recruitment ad reproduced in Uberland. Next to it, there's a picture of a breezy millennial with shaggy hair and a five-o'clock shadow, a scarf draped rakishly around his neck. He looks so noncorporate that he might not be wearing any pants.

In order to put that idea to rest, Rosenblat must first untangle the myths that made it seem possible in the first place. If you think about it, it's bizarre that taxi drivers became a symbol of cutting-edge technological disruption. Cab drivers have typically occupied a benighted role in the public imagination: hustlers, criminals, or, at best, misanthropic folk philosophers. Rosenblat offers a valuable history of the ideological work that went into the 'gentrification' of the profession. Uber was born in 2010, a time when all sorts of lowly regarded labor was being seen in a new, exciting light thanks to the hype around Silicon Valley. This was the age of the 'sharing economy.' Sharing economy boosters promoted the notion that the world is full of underutilized resources, which, if shared through technology, could greatly increase general prosperity. The message resonated widely during the slow recovery after the recession. One way to look at it is that the sharing economy reframed a crisis of global capitalism as a matter of inefficient resource distribution that could be solved with an app. Uber, Airbnb, TaskRabbit, and their peers and imitators urged us all, in clean sans serif fonts, to become 'partners' and press our personal vehicles, extra bedrooms, and spare time into 'helping' other members of the online 'community.'

The sharing economy was itself built on a myth that work done through a technology platform was not labor but a sort of 'communion,' as Rosenblat evocatively puts it. Reviewing media coverage of Uber in the early days, I was indeed struck by writers' quasi-religious faith that technology could miraculously make driving a taxi into a comfortable and lucrative middle-class career. How else can you explain the credulousness of reporters who relayed a claim by Uber in 2014 that an average driver made $90,000 a year? The Washington Post gushed: "Uber's remarkable growth could end the era of poorly paid cab drivers."

Update: It didn't. A recent study by the ride-sharing trade publisher Ridester suggests half of all Uber drivers make less than $10 an hour, while BuzzFeed examined leaked Uber data and found that, after expenses, the average Uber driver in the U.S. takes home $10.87 an hour. It seems a full-time Uber driver can easily earn a poverty wage. And in Rosenblat's telling, the rates only go down as Uber becomes more established and more drivers flood the streets, trapping drivers who take out subprime car loans at usurious rates to drive for Uber in a state of near-indentured servitude. One driver that stuck out to me in Uberland was Raul, a New York City Uber driver who had to boost his shifts from 8 to 9 hours to 12 to 14 in the face of falling rates. Rosenblat, who maintains a sometimes-unnerving cool while narrating tales of outrageous exploitation, writes: "The autonomy to choose which fourteen of the twenty-four hours in a day to work doesn't create the sense of freedom implied by 'flexibility' rhetoric" of Uber.

Uber's image of freedom is borrowed from the swashbuckling Silicon Valley entrepreneurs who sit at the top of the sharing economy. Unsurprisingly, the reality of working for Uber is nothing like this. A real entrepreneur is free to chase wealth through skill and instinct. Uber drivers are given neither the information needed to act in their economic interest, nor the autonomy to do it. When they log on, they are pinged automatically with rides and must accept them without knowing the destination. So they might drive 15 minutes to give a passenger a measly half-mile ride, earning negative dollars. Even if drivers did have the information to choose their rides, Uber's policy requires them to accept 90 percent of all rides or risk deactivation. Uber drivers could be described as 'entrepreneurial' only if entrepreneurs were required by law to take 90 percent of the crackpot ideas offered to them by people they met at dinner parties and attempt to turn them into profitable companies.

Under such limitations, working for Uber becomes sustainable only if you chase its ever-changing array of incentives and bonuses, most notably surge pricing, which offers higher rates during prime time and in certain areas. Rosenblat refers to the constant notifications of surges and bonuses as 'gameified' work. This might suggest that winning could require some entrepreneurial ingenuity and risk-taking. But it's really more like the fixation of a compulsive gambler trying to exploit a loophole in a slot machine. Drivers scramble from surge area to surge area, which dangle the possibility of '2.1x rates.' Drivers who make the trek often find they are offered non-surge rides, which has led some to the conclusion that the whole thing is a bait and switch. There is a saying among veterans who populate the numerous ride-sharing message boards: "Don't chase the surge." The message boards and blogs on which drivers congregate to try to make sense of ever-changing terms of services and rates are home to the only true entrepreneurs in Uberland: the ride-sharing gurus who claim to have cracked the code. For just $47 you, too, can learn the Five Potent Strategies You Can Use to Find More Pickups No Matter What City You're In, exclusively from

Being managed by an algorithm sucks in so many other ways I can only briefly go into them here. Probably the creepiest is the pervasive surveillance. Uber tracks nearly everything drivers do on the app, including how much their phone shakes to make sure they are driving safely. Passenger ratings are incredibly important, and drivers must maintain a high rating despite the infuriating fickleness of users, who are now able to inflict their prejudices directly on them. If drivers have any issues with customers or payments, the only recourse is an outsourced 'customer service' operation, which leads to women drivers reporting sexual harassment to some call-center worker in the Philippines who responds with a form letter. A long and tedious section of Uberland explains how various design features and policies systematically cheat Uber drivers out of wages, which just reinforces how effectively the apparent objectivity of a technology platform can hide built-in biases that screw over its users.

Not everything is so bleak in Uberland. Many of the drivers that Rosenblat speaks to seem more or less satisfied with the job. They do value the flexibility of being able to choose when to drive, which allows them to work around a second job, see family members, or pursue higher education. For many, it's better than working at McDonald’\'s. One driver who works 12- to 14-hour shifts for Uber and Lyft when he has time off as a fast-food manager gushes to Rosenblat, "I'm a people person. I love meeting people." Uberland conveys a variety of driver experiences that you rarely get from exposes about Uber's work conditions, which focus on extreme cases. While there have been many angry protests over rate cuts, and lawsuits challenging exploitative labor practices, the stories Rosenblat collected mainly convey weary resignation. "For some drivers, this job is a millstone, and for others, it's a life preserver," Rosenblat writes. At times, I felt churlish for my growing conviction that Uber should be wiped from the face of the earth. Who am I to deny someone their lifeline? At the same time, reading about the many indignities inflicted every day on drivers under algorithmic management made me yearn for a world in which such an arrangement could not plausibly seem like a refuge.

So, work sucks. What's new? Everything, according to Rosenblat. She concludes that Uber has brought about a fundamental cultural shift in what it means to be employed. To Rosenblat, Uber's algorithmic management techniques and its rhetoric of driver empowerment means that its drivers are consumers of Uber's technology as much as they are workers. This new consumer-worker category allows Uber to squeeze as much value out of drivers as possible without running afoul of the legal and cultural norms that have traditionally reined in unchecked exploitation. Here, Rosenblat seems to fall into the technological exceptionalism she blames for the hype around the sharing economy. After efficiently exposing how Uber's rhetoric about 'driver-partners' is essentially a rhetorical smokescreen, it was disorienting to see Rosenblat basically come around to their main point, that work done on the platform might not really be work, even if her emphasis is more critical.

I don't know if it is that deep. Rosenblat's argument would go down easier if it was rooted in the specific history of the taxi industry, instead of an abstract notion of 'work.' But a quick glimpse at that history suggests Uber's algorithmic management is less a revolution than a refinement. Taxi drivers have long been defined by an ambiguous mix of freedom and dependency. In New York City, many taxi drivers lease cars and the medallions that allow them to ply their trade from powerful brokerages. These drivers are called independent contractors, with all of the instability and lack of benefits that implies, but their livelihoods rely on patronizing shady middlemen who constantly innovate new ways to screw them over. Leasing largely replaced the older commission-based business model in the 1960s, as medallion owners realized that collecting rent from drivers was less risky and more profitable than assembling a fleet of their own. Biju Mathew argues in Taxi! Cabs and Capitalism in New York City that this was an early example of outsourcing - the systemized moving of risk from the center to the margins. If you look at Uber as a giant broker, you can see that algorithmic management has made this outsourcing more efficient, while hiding it behind layers of technology and ideology. Computerized management is new and unsettling, but it's only driving us faster toward the same bad place.

Books by Title

Books by Author

Books by Topic

Bits of Books To Impress