They know what you’re doing. They know where you’re going. They know everything,” Maurice tells me. “Once the app is on, they know. They know.”
Maurice has driven for Uber for three years now. He reckons that, on average, he spends 30 hours a week behind the wheel, starting his orbit between either of Chicago’s airports and destinations downtown at around 4 a.m. most days. “It’s nice little money,” he says of the work. “I’d rather ride around and make money than, you know, ride around and don’t make money.”
It’s hard to argue with that logic. Maurice isn’t drilling into bedrock, clearing a virgin forest, or carrying bags of feed. He’s making a few bucks cruising around town. At times, I’m sure, it can almost feel like it isn’t even work at all. (“I’m just driving right now. Dropping you off, that’s it. Nobody to answer to. Just going with the flow.”) But it is work, of course, and if Maurice ever forgets that, he merely needs to look over at the phone that’s mounted to the center of his dash.
They know. They know. What no one knows is how many people there are like Maurice. I don’t mean Uber drivers specifically— of whom there were nearly 4 million by the end of 2018, according to the Securities and Exchange Commission report the company filed in advance of its May 2019 initial public offering—but all of the workers who qualify as members of the so-called gig economy. The uncertainty is partly explained by the fact that there is no broad agreement on what counts as gig work. Side hustles are as American as political acrimony and apple pie, so if you define gig work to include everything from babysitting to hawking heirlooms, you might find, as the Federal Reserve did for a recent survey, that nearly three out of 10 Americans are so engaged.
If, however, we narrow the focus to the revolution in economic life by which work is almost entirely mediated by handheld devices, we discover a smaller but rapidly expanding universe. Consider that when the JPMorgan Chase Institute began studying what it calls the “online platform economy” in the first quarter of 2013, it found that only 0.3 percent of all Chase checking accounts received regular payments from what it deemed online platform companies such as Uber or Airbnb. By the first quarter of 2018, that number had grown to 1.6 percent, with 4.5 percent of all Chase accounts receiving income from at least one such company in the past year.
Alex Rosenblat, author of the recently released Uberland, calls the dynamic between Uber high command and its legion of drivers an exercise in “algorithmic management,” a fine term, at once futuristic and sterile. It highlights the fact that Uber and other companies like it are exploring a brave new world of workplace superintendence done almost entirely by machines, while also reminding us that the algorithmically tailored interaction between labor and management is never more than an innovation on an age-old relationship. With apologies to The Who, it may not exactly be a case of “Meet the new boss. Same as the old boss.” But for drivers like Maurice, when the Uber app is on, it’s still a boss nonetheless.
Bosses have always been a tricky matter in the ideological debates over capitalism. Champions of the system are keen to underscore the individual liberty assumed by free markets and, therein, the power of individuals to determine their destiny apart from the visible hand of government. At the same time, however, no less than Adam Smith emphasized the discipline assumed by the division of labor. No child nurtures the ambition of spending her adulthood grinding the edge of a pin, but such stultifying attention to detail is what is assumed by specialization, and a pin manufacturer is well advised to have a foreman looking over her shoulder to make sure the young woman is busy filing.
It is one of the great infelicities in the history of political economy that Smith, the man whose work did so much to augur and accelerate the Industrial Revolution, did not live long enough to see its ferocious bloom. He died in 1790, as another revolution, that in France, burst into view. Wrestling with the moral and practical implications of a commercial world typified by heavy industry and a growing horde of wage workers was left to others, most notably Karl Marx. “Owing to the extensive use of machinery and to division of labor, the work of the proletarians has lost all individual character, and consequently, all charm for the workman,” Marx and Friedrich Engels wrote in the Manifesto of the Communist Party in 1848. “He becomes an appendage of the machine, and it is only the most simple, most monotonous, and most easily acquired knack, that is required of him.”
If a company pushed any employee to the very edge of his capabilities, it would cause him to soon break down, the company’s investment in him squandered.
While Marx took such conditions as a gross affront to our fundamental humanity, others regarded the transformation of men into pistons, spindles, and pulleys as the principal challenge for managers in an industrial setting. Frederick Winslow Taylor’s proposed solution was presented in The Principles of Scientific Management, a pioneering work of workplace rationalization that was first published in 1911 and is now regarded as the foundational text for management consulting. “Under scientific management,” Taylor boasted, “the ‘initiative’ of the workmen (that is, their hard work, their good-will, and their ingenuity) is obtained with absolute uniformity.”
The principles outlined in his book promise this eventuality. Key among them is the decoupling of effort from understanding. “[I]n almost all of the mechanic arts the science which underlies each act of each workman is so great and amounts to so much that the workman who is best suited to actually doing the work is incapable of fully understanding the science,” Taylor wrote. And thus, he believed, there was no reason to bother explaining it to him. Indeed, doing so was not only a waste of time; it was counterproductive because it invited dissension and second-guessing. The point of a science of management was to furnish any workplace with “many rules, laws, and formula which replace the judgment of the individual workman.” A factory was most efficient when managers did the thinking, and workers did what they were told. As a factory hand described one such workplace to the journalist Edmund Wilson in the early 1930s, more than a decade after Taylor’s death, “a man checks ’is brains and ’is freedom at the door when he goes to work at Ford’s.”
More than any of his industrial peers, Henry Ford embraced the logic and aim of scientific management, though he came by his particular practices through trial and error in the laboratory of his factories rather than through the writing of Taylor. “We expect men to do what they are told,” Ford wrote in his memoir, My Life and Work. “The organization is so highly specialized and one part is so dependent upon another that we could not for a moment consider allowing men to have their own way.” Workers were vetted by the Ford Motor Company for jobs that best fit their mental and physical aptitudes and then “scientifically arranged” along the assembly line, “not only in the sequence of operations, but to give every man and every machine every square inch that he requires and, if possible, not a square inch, and certainly not a square foot, more.”
Such exactitude in the production process was not merely a matter of determining which worker did what best—and, for that matter, how a particular task might be done most efficiently. It was also concerned with conserving the most expensive investment any company makes: its workforce. Much like the careless driver who races his Ferrari in first gear, if a company pushed any employee to the very edge of his capabilities, it would cause him to soon break down, the company’s investment in him squandered.
In certain respects, Marx failed to price in this possibility when he conceived the fate of the proletariat. Rather than the gears of some great machine, their destiny seems more akin to a lump of coal dropped into the belly of an insatiable furnace, the vitality soon consumed in pursuit of the clinical promise of more units per minute. For Taylor, Ford, and other oracles of industrial management, this was a category mistake. Workers’ purpose was not fuel but function, and insofar as considerable investments had to be made to harness their full potential, you wanted their services for a lifetime, not a fortnight.
The cardinal concern, therefore, was not to break the individual laborer, or at least only to break him the way you would break a beast of burden rather than a bowstring. Then again, as Taylor understood, a superior example of workplace management didn’t involve the breaking of a worker’s spirit any more than it did her back. It required “another type of scientific investigation,” he said, one that went well beyond research into particular task techniques and the skeletomuscular system to include “the accurate study of the motives which influence men.”
Economists will sometimes conflate the discussion of motives with that of incentives, and one incentive in particular: money. No doubt, a longing for money is a common motive, and a powerful one at that, but in addition to being expensive to appeal to, it hardly encompasses what Taylor had in mind. If individuals only do what you want them to do because you will pay them for their efforts, they will hardly be enthusiastic employees, and you will be left to constantly reach for either a carrot or a stick to keep them honest and dependable.
Accordingly, especially as the focus of modern management has shifted from the conveyer belt to the cubicle, from jobs where one risked his body to those where he feared selling his soul, the science of management shifted from a concern for precise physical arrangements to a study of the subtle art of persuasion. “The formal aim, implemented by the latest psychological equipment, is to have men internalize what the managerial cadres would have them do, without their knowing their own motives, but nevertheless having them,” the late sociologist C. Wright Mills wrote in White Collar, his mid-century treatment of the American middle class. Under such a regime, he continued, “[m]any whips are inside men, who do not know how they got there, or indeed that they are there.”
For capitalism to work, we had to be as free to leave the conveyor belt, or turn over the keys, as to start a venture or strike a deal.
Notwithstanding its ominous tone, Mills’s vision of the science of middle management is still a matter of “easier said than done.” Indeed, valiant efforts to embed these whips have long made for scabrous satire precisely because they are typically impotent and lame. Take Office Space, a film that depends for so much of its humor on abortive attempts by bosses to elicit even a scintilla of esprit de corps from the spreadsheet servants encaged in their cubicles. Whether through the banal entreaties of office banners (“Is This Good for the Company?”), patronizing invitations to self-expression (“Friday is Hawaiian Shirt Day”), or intra-office social events with all the charm of a root canal (“Happy birthday, Mister Lumbergh”), a movie such as Office Space provides a useful reminder that it is far easier to cajole a worker’s body than control his mind.
That’s a good thing, too, for the great promise of algorithmic management, when combined with the power of big data, is the ability to harvest untold information from gig workers that may be used to make sense of, and best appeal to, their individual motives. And given the centrality of the app to everything a driver does, the opportunity is one uniquely afforded to Uber.
“Drivers are the subjects of constant experiments,” Uberland’s Rosenblat told me. “In much the same way that [a news publication] A/B tests what headline works better for a news story, drivers have constant A/B testing on their working conditions.”
It is inevitable that some of these tests concern what a driver takes home at the end of a shift. Rosenblat gives the example of the “destination filter,” a feature on the app that Maurice complained to me about. Uber had previously introduced a twice-a-day option drivers could select that would give them rides in a chosen direction, home typically, a perk of sorts whose luster dimmed when the company further introduced a 30 percent rate cut for any ride so offered. (As one driver described his impression of Uber’s decision on RideGuru, a popular online forum, “They are basically saying, ‘OK, we are going to rob drivers on at least two rides a day.’”)
The pay cut has since been suspended, and an Uber spokesperson told me that its intention was to redistribute the fees it withheld from drivers for using the feature to others in the same area who abstained, though the company did not make clear whether it actually succeeded in its efforts. Regardless, the course of events underscores the fact that, while drivers can protest changes foisted upon them and pop a trial balloon or two, they have little say in whether those changes are released in the first place.
Not driving—the most abrupt form of protest—is always an option, but hardly a viable one for the increasing number of workers whose gigs don’t complement their income so much as compose it. These drivers depend on Uber as much as anyone depends on a primary employer, so they are especially vulnerable to a steady diet of nudges, newfound constraints, and systematic surveillance by means of the all-seeing app mounted to the middle of their dashboards.
In fairness to Uber, on its face, most of this activity seems benign, with much of it ostensibly designed for the worthy end of passenger protection. For example, consider the way in which Uber compels “best practices” when it comes to behavior behind the wheel: “If I start going over 55 miles an hour, that goes red, and it starts listing my speed per hour,” one driver explained to me, pointing to her phone, as we hurtled down the highway. “If I went, like, over 80 mph, it would start dinging.” Or there is the way in which Uber prevents the kind of profiling that cab drivers have long been notorious for. Rather than identifying the destination and passenger the instant the ride is requested, in the US at least, Uber keeps passenger information hidden until the ride is accepted and destination information hidden until the driver arrives for the pickup, effectively preventing drivers from vetting their fares.
Whatever one makes of these policies, let’s not confuse the moral nature of these management techniques for the laudable aims to which they aspire. Badgering someone to slow down or withholding essential details of a ride she has agreed to until she cannot easily back out of it are instances, however trivial, of coercion and manipulation. They compel a driver to do what Uber wants her to do regardless of the driver’s opinions.
There are other practices, of course, that are far more morally ambiguous. Riders are constantly wielding carrots and sticks in the form of feedback, the consequences of which can be the difference between drivers making more money or losing their jobs entirely. At the same time, the company offers an array of incentive opportunities for completing so many trips in certain periods or for driving at certain specific times and places (respectively, a quest and a boost in Uberspeak). Such features help the company to ensure that, notwithstanding the endless encomiums to work flexibility, Uber has drivers driving whenever, and wherever, it needs them.
To be sure, humans have used similar techniques for millennia to command, govern, and control one another. Scientific management merely codified such practices within a commercial realm that assumed individual liberty for its efficient functioning. For capitalism to work, we had to be as free to leave the conveyor belt, or turn over the keys, as to start a venture or strike a deal. And yet it is precisely because of this freedom, and the unruliness and unpredictability it allows for, that working men and women must be strictly controlled.
The gig economy, and the decentralized workforce it assumes, therefore presents an asymmetrical set of hazards and hopes. For companies such as Uber, it affords the opportunity to harvest the time and talent of countless individuals without ever having to house or hover over them, or even honor many of the commitments, legal and customary, that have long structured the work of full-time employees. Indeed, with cutting-edge behavioral-science insights and the steady stream of information available to them through the personal devices that mediate such work, these companies have the opportunity to harness the power of algorithmic management with the kind of efficiency that Taylor could only have dreamed of.
For gig workers, on the other hand, beyond the possibility of moral restraint on the part of employers or worker protections put in place by an evolving regulatory scheme, there is the hope that maybe the science fails, that the cussedness of the human spirit is such that one will never be able to create a legion of flesh-and-bone automatons cheerfully dispatching tasks—or else, if the science does succeed, that it only does so in a fashion that ensures the experiences of individuals such as Maurice are consistent not only with a superficial sense of freedom but with the latitude of greater opportunity and deeper fulfillment as well.
Forget a few more rides per hour. That is a brave new world worth working toward.
John Paul Rollert is adjunct assistant professor of behavioral science at Chicago Booth.