Digital Labor The Internet as Playground and Factory

Increasingly thrown on the defensive, Hufington insisted that, in her new position as AOL’s head of content, she was pushing for the hiring of hundreds of professional journalists to staf the bureaus the company had opened as part of its Patch.com local news operation. That was a valid argument. But closer examination suggested that these new recruits would be servicing operations that are dificult to distinguish from what is known as a content farm—a site with shallow, non-original stories written specifically to trigger popular search queries and to game Google algorithms into placing the site on the first page of search results. Leading content farms such as Demand Media and Associated Media churn out low-quality articles and video in the field of online advice, paying a measly piece rate to their free agent creatives. As Dan Roth reported in his original 2009 Wired article on the topic, “pieces are not dreamed up by trained editors nor commissioned based on submitted questions. Instead they are assigned by an algorithm, which mines nearly a terabyte of search data, Internet trafic patterns, and keyword rates to determine what users want to know and how much advertisers will pay to appear next to the answers.” As a gauge to the fast growth of this spam-like sector, Roth estimated that Demand Media alone would soon be publishing “the equivalent of four English-language Wikipedias a year.”²

Just as these sites are ushering in a fast food revolution in content, they are engaged in a race to the bottom when it comes to remunerating employees. The filmmaker featured in Roth’s 2009 article was paid \$20 per clip for each how-to video he shot on location, edited at home on Final Cut Pro, and then uploaded to Demand Media. Given the growth rate of this sector, that \$20 piece rate has undoubtedly come down in the intervening years. AOL’s own business model for its big push into online content proved to be one of the factors driving the wage depreciation. The AOL Way, the company’s expansion plan that was leaked in February 2011, revealed how it would pay a pittance to in-house writers who were expected to pen up to 10 blog articles per day, each prepped for search engine friendliness and for maximum ad exposure.³

Creatives who have been knocked to the ground by the recent Great Recession feel pressured to sign up with this kind of word factory when, increasingly, it is the only game in town that pays. After all, the alternative to churning out junk product for a content farm is to play the reputation game by posting for free, like the Hufington Post bloggers. The former option involves the kind of routine toil that is anathema to aspiring creative professionals. The latter option promises the kind of unalienated expression of thought that is closer to their ideal. Yet only one of these will guarantee food on the table.

On the face of it, this does not appear to be a new dilemma. Creatives have been facing this kind of choice since the eighteenth century, when the onset of commercial culture markets ofered them the choice of eking out a living with the scribblers on Pope’s Grub Street or of building a name-recognition relationship with the fickle public. Literary agents, unions, and other professional organizations sprang up or evolved in order to protect their livelihoods from the rough justice of the marketplace, and while the explosive growth of new media has outpaced and outsmarted the traditional agents of bargaining and regulation (such as the press unions), ever-fresher versions are likely to emerge. The Freelancers Union, for example, was founded in 2001 specifically to respond to the needs of the selfemployed, and it has been the fastest-growing union in the United States in recent years. Its members are learning how to acquire an ever-larger share of social insurance and political clout while surfing each new wave that washes over the ever-mutating creative/digital landscape. Beginning in the 1990s, WashTech pioneered the business of labor protection for permatemps in the tech industry, and other Communications Workers of America (CWA) locals are following suit in their eforts to recruit independent contractors.

But it would be wrong to conclude that in the realm of digital labor there is nothing new under the sun. On the contrary, each rollout of online tools has ofered ever more ingenious ways of extracting cheaper, discount work from users and participants. The transition from web 1.0 to social web was a quantum leap in this regard. The youthful zeal that went into the first generation of web designs was bought with cappuccinos and beaming admiration from clueless elders. Building the pioneer environment of the web was like a massive barn raising, largely dependent on uncoordinated volunteer efort. Its successor also trades on the openness of youth, but the sophisticated operations of its hidden labor economy bear as much resemblance to the block-building of web 1.0 as the exotic derivatives of today’s Wall Street do to the origins of pork belly trading on the Chicago Mercantile Exchange. The social platforms, web crawlers, personalized algorithms, and other data mining techniques of recent years are engineered to suck valuable, or monetizable, information out of almost every one of our online activities.

Whether all this activity can or should be classified as labor according to any traditional criteria of political economy is a case in point, and one of the themes of this book. To address the question more fully, as I will do in the pages that follow, involves delving far below the visible surface of the digital landscape on which the Hufington afair was exposed.

From the outset, however, let us bear in mind that new media are not determining agents. Like any other technology, they are facilitators, not causes, of changing social forces. So, too, as Marx and many others have noted, technologies are not simply weapons of class war, designed to control and deskill workers, they also harbor the potential to eliminate wage labor, socialize production, and free up our time. Whether they are deployed for the latter purpose depends not so much on their technical development as on what Marx called the “relations of production”—that is, the state of our socioeconomic relationship to capital, property, and governance. Reverse engineering begins with technology, but unless it is also taken up as a social challenge, the chances are that the outcome will only benefit tech-savvy elites.

Formerly Known as Employment

In the heyday of the labor movement, it was commonly observed that the bosses needed workers but that workers didn’t need bosses. Yet in the third and fourth quarters of 2010, corporate America posted record profits at a time when the U.S. Bureau of Labor Statistics reported the real unemployment rate at 17%. Does this yawning disjunct between profits and joblessness mean that the bosses have learned how to get by without workers? Not exactly, no, but the statistics, which can be dissected a hundred ways, might suggest that a sea change is occurring in the world of work.

Two of the reasons for the high earnings seem to be beyond dispute. Corporations are moving more and more of their operations ofshore, especially jobs in high-skilled sectors, where the largest savings in labor costs can be gotten. So they still need workers, but not expensive ones in the North. A second explanation rests on what business economists call increased productivity. Roughly translated, this means that employees have been pressed, by the stif threat of redundancies, either to work harder and longer for the same paycheck or to take a cut. In any downturn, employers will push their advantage in this way, but in a soul-sucking recession like this one, there is no quarter; the assault comes from all sides, whether in the form of pay freezes, concessions, furloughs, layofs, or further casualization. A third reason—and this is the unfamiliar quantum—is the growing reliance on new kinds of free labor to boost the balance sheet of companies that are canny enough to harvest it. Hard evidence for this footprint is not so easy to muster, but the strong anecdotal record suggests it is large enough to be statistically significant.

Free, or token-wage, labor is increasingly available though a variety of channels: crowdsourcing; data mining or other sophisticated digital techniques for extracting rents from users/ participants; expanded prison labor programs; the explosion of unpaid, near-obligatory internships in every white-collar sector; and the whole gamut of contestant volunteering that has transformed so much of our commerce in culture into an amateur talent show, with jackpot stakes for a few winners and hardluck swag for everyone else. The web-enabled developments have attracted the most media attention, not least because the tidal surge of free online content directly threatens the livelihoods of professional writers and artists. After all, the widespread shuttering of newspapers, magazines, and overseas news bureaus has seen a generation of union jobs scattered to the winds. Professional pay scales are reduced to dust as the online content aggregators sweep all before them, and resistance was few and far between until the Hufington afair came along. In most corners of the information landscape, working for nothing has become normative, and largely because it is not experienced as exploitation.

From the early days of the Great Recession, business press pundits have wondered how far firms could go in taking advantage of new sources of free labor in order to stay afloat and improve their market positions. How can we take advantage of all the free time (or “cognitive surplus,” as net evangelist Clay Shirky puts it) that people have, especially the newly unemployed? Since many of the latter will be spending their newly free time online, how can we exploit their willingness to explore any avenue in search of the possibility of employment? Can we take advantage of their inclination to take on tasks that feel like fun? Or, more ominously, how can we harness their habitual need to participate in something that feels like work, in the absence of paid work and just to keep their hand in? Advocates for this line of thinking have seen it as a viable business strategy. They have also made overblown bonanza-scale claims for the potential windfall, inspired no doubt by the high valuations of social media firms. Inevitably, it has been suggested that social networking is the oil of the twenty-first century; yet, so far at least, it looks as if oil is still the oil of the twenty-first century.

Even so, the financial profile of these companies is remarkable. In 2011, Face-book took in an estimated \$4.3 billion in revenue, and almost \$1 billion of that was net profit. Leaving aside its pre-IPO valuation at more than \$100 billion, these numbers are big enough, especially if you consider that the firm only had not many more than 2,000 employees on payroll. This ratio of employees to revenue is unusual by any historical standards, but it is typical of firms that dominate the upper stratosphere of information services. In 2011, Google, for example, had around 30,000 employees, but it pulled in an estimated \$35 billion in revenue for a \$13 billion profit. The other fast-growing social media companies—Twitter, Groupon, Zynga, LinkedIn, and Tumblr—are in the same boat.

For the rapidly shrinking population that are not Facebook users, Aaron Sorkin’s film The Social Network must have presented a conundrum. On the one hand, the story of creative conception that it presents is reasonably familiar. Take a hothouse Ivy League environment where collegiate values are easily trumped by the predatory marketplace ethos already incubating on the campus and add a cast of recognizable characters: a socially challenged white male engineer; a brainy white girlfriend who challenges him even further through humiliation; a socially desirable male entrepreneur; assorted and primarily Asian female groupies who are irresistibly attracted to Jewish men; and a neobohemian start-up crew working 24/7 to make a market breakthrough. These are all updated components of the standard Hollywood template for myth making in the field of technical invention. This is how national champions labor to bring dazzling innovations and lustrous wealth into existence.

On the other hand, there is no actual social networking depicted in the film, and so uninformed viewers could plausibly conclude that the firm’s huge financial success rests on the subscription base of Facebook’s half a billion users. Yet these users are not consumers in any traditional sense of paying customers. Rather, the variety of activities they perform (technically known as click signals) is the source of valuable data that is sold to the true customers—advertisers or behavior market vendors such as Bluekai, TargusInfo, and Acxiom. According to Eli Pariser, “Acxiom alone has accumulated an average of 1500 pieces of data on each person in its database, which includes 96% of Americans.”⁴ Some of these customers pay to advertise on the site, though most use the information to follow users around the web with personalized spot advertising. The trade-of for users, of course, is free access to the platform and the software, but, from the company’s perspective, the cost of hosting and maintenance is dwarfed by the tradeable value of the information it can extract from the daily churn on its site. By far the majority of social network users are unaware of how the platform owners profit from the volunteer content of their communications, or indeed how they themselves are generating monetizable product for the owners. But as Andrew Lewis has succinctly put it, “If you’re not paying for something, you’re not the customer, you’re the product being sold.”⁵

The 1960s futurist Alvin Tofler coined the term “prosumers” to describe the class of consumers who had evolved beyond a passive acceptance of marketplace choice. But he could hardly have imagined how the term would come to be used, as it is today, to denote the mass of Internet users whose devoted eforts to build relationships and polish their online social identity are the raw material for tidy profits enjoyed by others. Nor could any of the other sunny 1970s prognosticators of postindustrial society such as Daniel Bell have imagined the new order being driven by an attention economy, or that it would be sustained not by the gainful labor of cognitive workers but by the selfpromotion of ordinary, unpaid individuals. Moreover, the rewards that underpin this economy are, in some respects, redolent of the kinds of assets that secured social standing in an era that preceded industrialization, when the careful and laborious nurturing of relationships with wealthy and powerful names were sources of considerable worth.

Today, we can see the resurgence of such a culture based on the cultivation of social capital, whether for those in search of breakthrough or blockbuster attention in the reputation stakes (in Twitter trending and top viral links) or in the more low-hanging circuits of Internet self-exposure. In some quarters, this afective currency has replaced the wages of industrialization, especially for professionals who used to earn a structured living from paid content and who now disseminate their bylines far and wide in hopes of securing a niche livelihood from name recognition. But by far the most substantial rewards are allocated, on an industrial basis, to those who build and maintain the technologies of extraction, who hold the system’s intellectual property, and who can trade the aggregate output of personal expression as if it were some bulk commodity like grain or beets.⁶ The real spoils, in other words, do not go to the aspiring stars, ranked and rated by the battery of metrics that measure Internet sentiment and opinion, but to behind-the-scenes content hosts and data miners, who utilize these and other metrics to guarantee their profits. The outcome, for this latter group, is a virtually wage-free proposition. When all is said and done, the informal contract that underpins this kind of economy is a profoundly asymmetrical deal.

The art of producing gratis media content by showcasing the vox populi has a long history; its origins could arguably be traced to the establishment of letters to the editor columns in print publications. Since these contributions were selected, edited, and, in many cases, fabricated in order to support the editorial line of newspapers and magazines, they ofer a good illustration of how supposedly unsolicited public opinion can be generated, shaped, or even ventriloquized. Websites that depend on user input, whether for the main action or in the form of comments posted in response to a featured item, are in direct linear descent from these first letters to the editor. More raw and unfiltered by far, they build on the popularity and cost-efectiveness of their print antecedents.

In recent years, as the open comments sections (“Comment Is Free”) have lengthened and proliferated, more and more online newspaper versions have turned to crowdsourcing appeals for readers to generate free columns, images, videos, designs, fact checking, and other information supplies.⁷ The principle underlying these appeals is that readers will be gratified to participate and that the results will be more authentic, especially if they are drawing on skills and knowledge unavailable to a commissioned reporter. Outside of the mainstream media, this principle also applies to the widespread uptake of crowdsourcing as a semi-industrial technique for extracting ideas, opinions, designs, and intelligence with little or no compensation for the provider other than name recognition. Informal evidence suggests that as long as a task can be advertised as fun or cool, there is a good chance you can get it done for free, or for a pittance from the seemingly ever-obliging crowd. Moreover, if some of the input seems to be very professional, that is because either the crowdsourcing call is specifically crafted to appeal to professionals on their downtime, or else because it quite probably comes from someone who used to be a professional employee and has been cast into the amateur demimonde of the volunteer content provider.

At the other end of the spectrum are more routine tasks, such as those put out for bid on programs like Amazon’s Mechanical Turk. The bids are accepted in return for a minimal fee set by the requester and are assigned to tasks that may take no more than a few minutes to perform.⁸ The registered taskers of the Mechanical Turk and other e-lance operations would not be thought of as remotely resembling temporary employees any more than the uncompensated creatives who respond to the more skill-intensive kinds of crowdsourcing. They leave no trace of their employment, and certainly nothing to implicate an employer in any legal or regulated network of obligations. What they do, however, is bring the definition of a job much closer to its etymological source—a discrete lump, or piece, of work that exists only for the duration of its fulfillment.

Distributed Workplaces

Distributed labor has been suggested as a way of describing the use of the Internet to mobilize the spare processing power of a widely dispersed crowd of discrete individuals. This should not be confused with an older use of this term to describe the business process outsourcing business model for coordinating geographically dispersed workplaces, whether from telecommuting or from distant nodes on a global production chain. That model was especially critical to the wave of whitecollar ofshore outsourcing in the first half of the last decade, and it depended on sophisticated work-flow platform technologies to slice up, allocate, and recombine work. So, too, the new model should be distinguished from the distributed work-place known today as the mobile ofice. Business strategists advocate on behalf of the benefits to employee morale of allowing high-wage corporate talent to work anytime, anywhere, and on any device. But it is the boost to eficiency and productivity that commands the most attention in their reports. Untold revenue can be extracted from the steady erosion of the boundary between work and leisure time—a long-held dream of employers—which results from putting employees on an unforgiving 24/7 leash.

The new kind of distributed labor does not need to be performed by payroll employees in far-flung branch locations, or by notebook toters in wired cofee shops, the default workplace for a generation of contract freelancers who forsake the privacy of their homes (Tofler’s “electronic cottage”) to work in public view, braving, or feeding of, the gregarious hum of society. Rather, it is done either by users who do not perceive their interactive input as work at all, or else it is contracted out online—through a growing number of e-lance service sites—to a multitude of taskers who piece together lumps of income from motley sources. As in the ofshore outsourcing model, the dispersion of this labor is highly organized, but it is not dependent on physical relocation to cheap labor markets. Instead, the cost savings can be derived from either the latent talent of the crowd or the microdivision of labor into puzzles, stints, chores, and bits, which, if they amount to anything more than distractions, require only fitful bursts of concentration.

The devising and parceling out of these microtasks is arguably only the latest development in a lineage of work management that derives from Taylorism. Taskers are efectively deskilled, dispersed, and deprived of any knowledge about the nature of the product to which their labor contributes. The coordinating manager, by contrast, is in complete control of the labor process. As for the donor labor of the crowd, that has a longer historical lineage since it owes a lot to the traditions of creative work, where sacrifices in monetary compensation are commonly made in return for job gratification or for the opportunity to test and advertise one’s talent. This willingness to donate labor was referred to as self-exploitation when it first emerged as an industrial prototype in the formal employment ofered by the New Economy or dot-com firms of the late 1990s. In the course of my own ethnographic research on these new media workplaces, for No-Collar, I recall that one of my interviewees told me her job ofered “work you just couldn’t help doing”—a description that seemed to sum up the mentality of passionate, or sacrificial, labor.⁹ Subsequent ethnographic studies of knowledge and creative industry workplaces have shown that job gratification comes at a heavy cost—longer hours in pursuit of the satisfying finish, price discounts in return for aesthetic recognition, selfexploitation in response to the gift of autonomy, and dispensability in exchange for flexibility.¹⁰

One of the ways to contextualize the rise of the creative industries over the course of the last decade is to interpret it quite literally as an efort to industrialize creativity, aimed, of course, at the market prize of intellectual property.¹¹ Adapting the tempo of creative work to an industrial template is an acute managerial challenge, however, and, in a jackpot intellectual property economy, the costs of competing are considerable. The turn to crowdsourcing ofers a more impersonal solution that slices costs and delivers owners from any employer-type obligations. The crowd is not only smarter than trained employees, you don’t need to make social security contributions to take advantage of its wisdom or put up with the wayward personalities of the creatives on payroll.

Crowdsourcing and allied techniques are the progeny of strange bedfellows. On the one hand, there is a clear debt to the collaborative basis of the open-source movement, shareware, and the hacker ethic, which is profoundly proto-anarchistic in its embrace of the principle of the commons. The underlying spirit of mutuality, or what Trebor Scholz and Geert Lovink call “the art of free cooperation,” has been surprisingly tenacious in the face of concerted eforts on the part of would-be monopolists to enclose, privatize, and commercialize the digital domain.¹² After all, a handful of corporate giants—Google, AOL, Facebook, Yahoo, and Microsoft—now account for the overwhelming majority of daily web trafic. On the other hand, the corporate race to the bottom in pursuit of cut-price labor costs is also drawing heavily on the same collaborative spirit. It is no surprise that entrepreneurs scouting around for a fresh, dressed-to-impress business model have seized on crowdsourcing as a technique that unleashes the latent, or native, genius of Internet culture.

Many readers will no doubt conclude that this dual utilization is all part of some big-picture trade-of. After all, the social web has opened up a whole new universe of information-rich public goods—including the potential for anticapitalist organizing; really, really free markets; peer-to-peer common value creation; public access culture; cyberprotest; and alternative economies of all sorts (and, if you believe any of the cyberhype about the Twitter revolution, it is even the key to overthrowing authoritarian rulers in Middle Eastern and North African states). On balance, then, it could be said that the role social web platforms are playing in new modes of capital accumulation is simply the price one pays for maintaining nonproprietary networks whose scope of activity is large and heterogeneous enough to escape the orbit of government or corporate surveillance. Though the enclosers are pushing hard, the balance, for the time being, is still in favor of the commons. From this point of view, all of the free labor that gets skimmed of can be seen as a kind of tithe we pay to the Internet as a whole so that the expropriators stay away from the parts of it we really cherish.