According to one IT industry expert, the search engine is twenty-first-century infrastructure, and “building and maintaining a search engine is so expensive and labor-intensive that it requires the same kind of planning and upkeep that, say, the Golden Gate Bridge does.”1 This statement defies the popular perception of the digital economy as knowledge-based, immaterial, and weightless,2 and, unlike the industrial capitalist economy, needs very little in the way of human labor and low-wage workers, if the industry's mythology is to be believed.
In reality, digital activities facilitated over network technology as a social process embody human labor, and the search engine is no exception; every link on the web and each keystroke on a computer, tablet, or smartphone contains human labor. Search engine technology is so seamlessly embedded in our daily lives, however, that it masks a cut-throat competition to maximize profit that manifests itself in a whole series of complex labor processes and divisions of labor that enable and animate it. This chapter uncovers the occupational structure of the search engine industry to clarify both the social relations between capital and labor and the search industry's profitable expansion.
What are the distinctive modes and forms of labor processes and organization that assist in the search engine industry's profitable accumulation? Who is actually laboring to deliver information instantaneously and in a seemingly highly automated way in response to humans’ never-ending search queries?
To answer these questions, this chapter draws on Harry Braverman's theory of labor, in which the labor process is increasingly mechanized and Taylorized as labor is broken up into ever-smaller, simpler, and more discrete tasks to speed workers up and disassociate them from the entire labor process in order to increase productivity and reduce labor costs.3 Though new industries and new technologies require new kinds of occupations and skills, under capitalism, that does not negate the impact of the division of labor, mechanization, and rationalization that constitute the degradation of working conditions and deskilling of labor affecting all levels of workers. Braverman's concept of the degradation of labor refers not only to precarious working conditions with routinized and repetitive tasks but also to alienation from the workplace and social structure.4 This process also involves structural displacement of labor as labor-saving technologies replace workers and create structural unemployment, which he describes as systematic surplus labor, or the “reserve army of labor” (RAL). Drawing from Karl Marx, Braverman writes that RAL is a necessity of working for the capitalist mode of production; it controls labor by increasing or absorbing the entire workforce depending on the economic cycles of expansion and contraction.5
The political economy and technology have changed since Braverman originally wrote in the 1970s, but this chapter demonstrates that the central features of work within capitalism that Braverman observed persist in the digital age. The chapter argues that with increasing competition and the introduction of digital technologies, labor degradation and deskilling have intensified as capital mechanizes, rationalizes, and innovates in the division of labor to increase profit. Thus, his critique of the capitalist mode of production is still pertinent because capital has continually deployed networked information technologies across economic sectors by attacking labor processes as they exist, with the aim of altering and even reconstituting everything from the sequencing of specific tasks to the technical division of labor within companies and industries to the location of production processes. By applying Braverman's work, this chapter underscores the broad occupational division of labor and the nature of work that enables capital accumulation and expansion in the search engine industry, centering on Google as a representative of the sector and the wider Internet sector in general.
The chapter begins with the changing technology sector within the broader political economy within which search is located. It shows that there is a pool of RAL in the tech industry that is the result of ongoing restructuring, technical innovation, and periodic capitalist crises and their impact on paid labor. Second, it discusses three broad categories of work in Google that not only represent the search engine industry but reflect wider trends in labor structures today. The chapter discusses Google's hierarchical occupational structure—specifically, skilled labor, low-wage processing workers, and unpaid labor. These categories seem overly simplistic, but the purpose is not to identify a comprehensive and detailed division of labor; rather, it is to draw attention to the broader characteristics of today's labor organization in the Internet industry. These divisions illustrate that, despite the industry's creating a new segment of highly skilled occupations, it also generates an extremely large segment of low-wage processing and appropriates unpaid user/consumer labor. Rather than looking at this feature as a distinctive phenomenon, this chapter locates it within the longer development of capitalism. By demonstrating how the highly automated and science- and technology-based search industry generates and relies on masses of precarious processing workers and unpaid labor for its capital accumulation, the chapter offers an unmatched window into the labor organization of the most dynamic economic sector in contemporary capitalism.
Braverman explains that the RAL, or body of surplus labor, consists of various forms—people are unemployed, underemployed, and precariously employed with extremely low wages. The reserve army of labor is a structure of capitalism that shouldn't be separated from employment given that the RAL dynamically affects the precariousness of the working class.6 As Fred Magdoff and Harry Magdoff put it, “One of the central features of capitalism is the oversupply of labor, a large mass of people that enter and leave the labor force according to the needs of capital.”7 Capital's continuing introduction of labor-saving technologies with crises-prone capitalism expand the RAL, which has become the instrument for capital to control workers, keep wages down, and maintain profits. Looking at the pool of surplus labor is helpful as a starting point in discussing labor in the search engine industry within the broader political economy.
Five years after the great recession, which wiped out 8.4 million jobs between 2008 and 2009 (6.1 percent of all payroll employment),8 former US president Barack Obama celebrated the nation's US economic progress and declared, “We have got back off [sic] our feet, we have dusted ourselves off…. Construction is up. Manufacturing is back. Our energy, our technology, our auto industries, they're all booming.”9 He reminded the public that the unemployment rate was at its lowest point since September 2008 and that the economy was improving. In fact, in 2019 the Department of Labor reported the unemployment rate as 3.6 percent, down from its official recession peak of almost 10 percent.10 Carrying on Obama's economic optimism, the Trump administration boasted that this was the lowest unemployment rate since the 1960s and that the economy was booming.
Although the official unemployment rate is typically used as an indicator of the health of the economy, it doesn't offer a full picture. The official statistics conceal real counts of unemployment. The number does not account for missing workers who are neither employed nor actively looking for jobs because they are discouraged by job prospects and have given up searching for a job. If missing workers were added to the official count, the unemployment rate in June 2019 would have been 7.2 percent.11 This was at a time when unemployment was at a record low. Moreover, there is a significant number of workers who are chronically underemployed, which is the case when workers want to work full-time but are forced to work part-time (involuntary part-time) due to economic conditions or work at jobs for which they are overqualified. These members of the RAL are often muted from the US government's official unemployment rate and the media. Thus, even before the current economic crisis was compounded by the coronavirus, the “booming economy” had never quite resonated to the millions of poor working people who were struggling to survive and being left behind.
The RAL is a constant threat to currently employed workers. Their need to compete with the unemployed keeps wages down12 in addition to serving an ideological function for the Internet industry to serve its interests. The industry promises over and over that it will create a large quantity of good, high-paying jobs as such tech firms as Google, Facebook, Amazon, Apple, and Twitter are increasingly occupying cities and suburbs around the world with their mega data centers and shiny new campuses complete with associated amenities. Political elites and their allies offer massive government subsidies to tech firms to bring these Internet giants to their towns and cities with a promise to their citizens of economic prosperity. Meanwhile, the existing tech industry is being restructured to renew its profits as the industry faces new competition.
The Internet sector has spoken ad nauseum of its ability to create more good jobs and boost economies; meanwhile, traditional tech firms such as IBM, Cisco, Hewlett-Packard (HP), Intel, AT&T, and Microsoft, faced with competition from the emerging Internet services sector and a rapidly changing IT market, have been showing a lot of volatility in their labor forces and laying off record numbers of workers as they constantly restructure and automate their businesses. Cisco slashed 5,000 workers in 2012 and has shown only modest workforce gains since 2018 despite the rapid implementation of 5G networks.13 Since 2011 HP, whose garage in Palo Alto is the mythical birthplace of Silicon Valley, has eliminated 300,000 jobs, almost equivalent to the populations of two Mountain Views and two Palo Altos combined.14 Hewlett-Packard's workforce contracted from over 340,000 workers in 2011 to 51,000 in 2021.15 Between 2013 and 2022 IBM cut its workforce from 466,000 to 345,000.16
As it reoriented its business toward cloud computing, Microsoft cut 14 percent of its workforce in 2014—as many as 18,000 people—mostly from Nokia, the mobile business it had acquired that year. According to the company, this was the largest layoff in its history.17 In response, Chinese workers at Microsoft's Nokia factory at Yizhuang Industrial Park took to the streets to protest. In his public memo to Microsoft employees, CEO Satya Nadella said that this drastic change was necessary for the company to become “more agile and move faster.”18 After 2016, Microsoft gradually expanded its workforce because it had succeeded in shifting its business to the cloud.19 While the Internet companies earned record-breaking profits during the COVID-19 pandemic, AT&T cut almost 9,000 jobs between June and September 2020.20 In fact, AT&T, the world's largest telecommunication company, currently has 78,000 fewer workers than it had in 2015.21 This continuing trend across the tech sector was succinctly described by Kay Roger: “Layoff is a permanent feature in the tech sector.”22
On the surface, these frequent layoffs seemed contradictory, given that tech is the most dynamic economic sector. Yet this was indicative of an IT sector that was demonstrably undergoing a wide-ranging reorganization with the rise of new Internet businesses but was also highly uneven and volatile in terms of workforce restructuring. Also, it exhibited disparate and even contradictory trends that were deeply marked across the length and breadth of the information workforce. This permanent layoff/hire cycle constantly replenished the RAL while weakening labor power and depressing wages.
Labor as incarnated in the search engine industry needs to be situated and understood within this vortex. I now turn to an examination of some characteristic trends of the high-pay, high-status segment of the occupational structure that receives the most media adulation in the information industry in general and the search industry in particular.
As the Internet sector expands, it restructures existing tech sectors and generates new kinds of occupations, labor demands, and workplace structures. One of the characteristics of the labor structure within Internet industries such as search is a concentration of disproportionately well-paid and highly skilled workers at the top. As the search industry extends its profit territory to include mobile, the cloud, app development, autonomous vehicles, and artificial intelligence, workers categorized as high-skilled by the industry are computer and data scientists, software engineers, computer programmers, product managers, quality assurance engineers, machine learning engineers, and those doing similar jobs.23 These IT workers are often said to be akin to well-paid young “Googlers” who have four-year or advanced degrees in computer science, engineering, and business, and the nature of their work has to do with computer systems design, scientific research and development, or management and business. Initially, many of them came from elite institutions such as Stanford University, University of California, Berkeley, MIT, Carnegie Mellon, and UCLA, the historic hubs of the Academic-Military-Industrial Complex. This was not surprising given that the origins of the search engine industry are within this complex, where engineers have historically had access to both technical training as well as government and private capital. Engineers, working hand in hand with venture capitalists and guided by capital expansion, have transformed search into one of the most powerful sectors of the information industry.
Google described itself as an engineering company. The long-time Silicon Valley journalist Ken Auletta has stated that Google is run by engineers: “Google's leaders are not cold businessmen; they are cold engineers.”24 These engineers, under the guidance of corporate goals, design and build search technologies to be productive and profitable. Tech firms typically do not reveal the number of engineers they have, but Google reported having 27,169 employees in “Research and Development” in 2016.25 Although not all workers in R&D are engineers, this number indicates the scale of the company's research unit. In 2018, when Google acquired the design unit of the Taiwanese smartphone company HTC, the major part of the deal was the assimilation of more than 2,000 HTC engineers into Google—thus the portmanteau “acqhiring,” a combination of “acquisition” and “hiring.” To give some perspective, the total number of tenured and tenure-track engineering faculty members from across twenty-three engineering disciplines in the United States was 28,521 in 2017.26
This engineer-driven Internet sector has taken in highly skilled IT workers for quite a while. After the 2008 economic downturn, the worst since the Great Depression, when other established IT companies such as IBM, Cisco, and HP were shedding workers in a rapidly restructuring market, Google announced that it added more than 4,500 workers in 2010 in engineering and sales and recruited more than 6,200 workers in 2011.27 In that year, when other tech firms were still trying to recover from the recession, Google's workforce increased by 33 percent, or more than 8,000 employees.28 Facebook, Apple, and other Silicon Valley companies also were on a hiring spree, and by 2016, the number of tech jobs in the San Francisco bay area surpassed the peak of the dotcom era.29 Clearly, there was a surge in technology jobs in Silicon Valley and beyond driven by Internet firms.
This demand for highly skilled labor by the Internet companies resulted in pushing up wages for this class of young IT workers with an average age of less than thirty-five—the average age of workers at Google was thirty, at Facebook twenty-eight, and at Apple thirty-one in 2017.30 Google's median pay package was $246,804 in 2018, followed by Facebook with a median pay of $228,651.31 The median salary at Google, Facebook, Netflix, and Twitter surpassed that of Exxon, Chevron, Goldman Sachs, and Verizon.32 By comparison, in the United States the average starting salary of a public school teacher was $38,617 and the average teacher's salary was $58,950 for 2016–2017.33 Worse, when adjusted for inflation, the average teacher's salary was down by 5 percent, and in some states down by as much as 15 percent, since the Great Recession of 2008.34
This tells us that new Internet companies indeed create well-paid jobs; however, what is concealed is that they tend to nourish and enrich only a small cadre of very highly skilled workers by generating a relatively narrow array of jobs that require considerable education and expertise. For instance, according to the 2019 Science and Engineering Indicators, the STEM workforce consisted of approximately thirty-six million persons, representing 23 percent of the total US workforce.35 This number includes technical jobs that do not require a bachelor's degree. The majority of highly skilled STEM jobs require at least a bachelor's degree, but less than half of the STEM workforce has a college degree. Even if all STEM workers with a college degree worked in highly paid occupations, they would make up only 10 percent of the overall workforce.
In reality, then, new Internet companies need only a relatively small number of highly skilled, well-paid workers. Compared to traditional IT companies, Google, Facebook, and Twitter didn't require many workers to design their algorithms or service-based products until they expanded beyond their core business domains.36 In 2012, seven years after its founding, the social news site Reddit had only eleven employees servicing a site handling over four hundred million unique visitors;37 the photo-sharing social network company Instagram, with a $1 billion valuation before it was acquired by Facebook, had a team of only sixteen to support thirty million users.38 Thus, the perception that the tech industry creates a large number of high-level jobs does not comport with reality. In actuality, the search engine industry operates with only a small cadre of very high-skilled workers at the top. This labor structure is reflected across Silicon Valley, where Google, Facebook, and many of the Internet companies reside.
According to the 2020 Silicon Valley Index, which tracks the economic trends in the region, since 2010, although there is a slight uptick in highly skilled, high-wage occupations in the Valley, the greatest job growth was in the low-wage service and infrastructure sector, with a shrinking number of middle-income jobs.39 The region has turned into one of the most economically unequal areas in the world. In 2021 the average annual income in the Valley was $170,000 but the average income for service workers was $31,000.40 This disparity is not limited to the Valley; rather, the growth of the Internet industry is polarizing the entire US workforce.41
Moreover, even this small class of highly skilled workers is no longer insulated from capital's competition-driven efforts to cut labor costs and increase profits, with a vast global reserve army of labor resulting in a new division of labor deployed by multinational corporations and enabled by network technologies. Today, higher-skilled tech jobs such as engineering and computer programming are being automated or are outsourced to places such as India, where there is an increasingly abundant supply of lower-cost, highly educated IT workers. Tech companies are rallying together to further open and reach the global skilled labor market.
Companies in Silicon Valley have long been the most outspoken protestors against US caps on visas for highly skilled foreign workers—though they were silent while the Obama administration deported more than two million working-class immigrants.42 Google, Microsoft, Apple, Facebook, and other major IT companies are at the forefront of lobbying Congress to reform immigration laws to increase the number of H-1B visa holders so that they can bring more highly skilled workers from other countries, if not completely remove the barriers to bringing in an infinite number of foreign IT workers.
Since the 1990s the US Department of Labor has issued 85,000 H-1B visas per year to allow foreign-born workers with specialized skills to work in the United States on a temporary basis.43 More than half of these H-1B visas have been issued for technology-related positions. Software and systems engineers, financial analysts, computer systems analysts, and marketing specialists—which make up a large segment of H-1B applicants—are the most commonly sought after by Google, Amazon, Apple, Facebook, Microsoft, IBM, and other tech companies.44
In 2019, the top H-1B visa holders were multinational information technology consulting and outsourcing firms such as Deloitte Consulting LLP, Tata Consultant, Cognizant, and Infosys, as well as Internet companies such Google, Amazon, Microsoft, Apple, Qualcomm, Salesforce, and Uber.45 Google was the fifth-largest H-1B employer in 2019, holding 9,085 H-1B positions certified by the Department of Labor. Apple and other Silicon Valley firms desperately want more access to the global pool of highly skilled workers because those companies seek cheaper and more easily disposed-of, docile workers. Tech companies have long lobbied the US government to enlarge this segment of the workforce.
In 2007 the vice president of Google's People Operations, Laszlo Bock, testified before the US House Judiciary Subcommittee on Immigration and urged Congress “to significantly increase the annual cap of 65,000 H-1B visas, to a figure more reflective of the growth rate of our technology-driven economy.”46 In 2012 Brad Smith, Microsoft's general counsel and executive vice president, said that tech companies were facing a workforce crisis because of the lack of qualified job applicants.47 He noted the shortage of labor, pointing out that Microsoft had thirty-four hundred open positions for researchers, developers, and, engineers—an increase of 34 percent from 2011—and that the skills gap was one of the biggest problems for the company.48 Microsoft founder Bill Gates testified before the US House Committee on Science and Technology, stating that US companies were facing a severe shortage of scientists and engineers with the skills necessary to develop future innovative information technologies. Gates asked that Congress reform US immigration policy to increase the numbers of highly skilled foreign workers to work for US companies.49
In 2013, publicly warning of a skills gap, Facebook CEO Mark Zuckerberg, supported by Bill Gates and Google CEO Eric Schmidt, formed the political lobbying group FWD.us to focus specifically on immigration reform and rally around bills that would serve the group's business interests. The IT sector also banded together to back immigration bills such as the Science, Technology, Engineering, and Mathematics (STEM) Jobs Act, the Startup Act 2.0, the Brains Act, and the Immigration Innovation Act. Put succinctly, the industry continues to demand that the US government intervene to enlarge its skilled labor force because the barrier to transnational mobility of IT workers can only be resolved by regulation on the part of nation-states.
But the Trump administration's draconian immigration policy complicated the US tech sector's efforts to secure and enlarge its flexible pool of labor. It suspended access to new H-1B visas and H-4 visas awarded to the spouses of H-1B holders through 2020. In response, more than fifty tech companies including Google, Facebook, Amazon, and Microsoft filed a brief backing a lawsuit challenging the ban on entry of foreign workers. During the Trump administration, there was a rise in the rejection rate for H-1B visa applicants. The multinational outsourcing firms that received the majority of new H-1B visas suffered from the declining approval rate.50 This was one of the reasons why Silicon Valley tech companies joined the rising social movement against Trump's immigration policy.51 While Internet firms were willing to fight against immigration policy as it related to H-1B visas, they saw no contradiction in working with the US Immigration and Customs Enforcement Agency to supply technology to facilitate the arrest and deportation of immigrants from Central America. Shortly after the Biden administration took over in January 2021, the Trump-era rule on the restriction of H-1B visa applications was rescinded. According to Axios, which obtained a confidential document,52 a group of tech industry CEOs led by Eric Schmidt and Jared Cohen (CEO of Jigsaw and former adviser to Condoleezza Rice and Hillary Clinton) made several policy recommendations to the Biden administration under the premise of competing against China. They included an immigration policy that encouraged the United States to retain and expand the number of highly skilled workers in the fields of science and engineering.53 As of this writing, the Biden administration was seeking to allow more foreign skilled workers in the United States.
The industry has warned of catastrophic labor shortages in an attempt to increase the mobility of highly skilled labor and justify the reform of immigration laws in order to have access to foreign labor. Although firms are required to pay foreign workers comparably to American workers, in 2008 there was a significant difference in the salaries of foreign workers holding H-1B work visas and their American counterparts in computer-related occupations.54 A 2020 study by Costa and Hira of the Economic Policy Institute revealed that all major US Internet companies, including Amazon, Microsoft, Google, Apple, and Facebook, exploit the H-1B visa program and pay their foreign workers a salary that is below the local median wage.55
With the increase in demand for tech workers in the growth segments of the sector, the tech industry continues to seek foreign workers to enlarge its labor supply for its accumulation projects because this global labor pool gives capital greater power to hold down salaries and also the flexibility to hire and fire employees depending on market dynamics. In addition, to expand their labor supply, corporations also exploit government subsidies as they use institutions of higher education as their training grounds for both foreign and domestic workers. Workers constantly have to adjust their skills in response to market needs rather than pursuing meaningful work that fulfills their aspirations.
As the competition between and among global Internet companies has intensified, the industry has made a concerted effort to further open global markets for highly skilled labor and reorganize not only low-wage workers but, increasingly, highly skilled workers to create a large and more casualized labor pool for its rapidly changing sector. As Biao Xiang points out, since the high tech industry was financialized in the 1990s, its value has been tightly linked to fluctuations in the stock market, making way for large-scale hirings and firings.56 The logic of capital here is to have not only a sufficient supply of skilled labor but also a mobile workforce that can quickly “respond to market fluctuations with minimum time lag.”57 This signifies extreme volatility in the tech labor market, and there is no job security even if one is highly skilled in the information-based economy.
Does this mean that the search engine industry depends only on a small number of highly skilled workers? The answer is no.
I turn now to show some characteristic features of the information industry at the other end of the pay scale. As I will demonstrate, Google is able to depend on a small number of skilled workers not only because it is a highly automated industry but also because it has appropriated a massive pool of invisible low-wage and unpaid workers to prop it up.
Braverman captured the essence of labor processes under capitalism, explaining:
Every step in the labor process is divorced, so far as possible, from special knowledge and training, and reduced to simple labor. Meanwhile, the relatively few persons for whom special knowledge and training are reserved are freed so far as possible from the obligation of simple labor. In this way, a structure is given to all labor processes that at its extreme polarizes those whose time is infinitely valuable and those whose time is worth almost nothing.58
Braverman's description sheds light on the highly polarized and mechanized labor process of the 1970s, which resembles the broader organization of labor in today's Internet sector.
Until Google workers exposed the issue to the media, the company barely mentioned an entirely different category of workers toiling at the Googleplex: temporary workers, vendors, and contract workers, known colloquially as TVCs. Google hires a wide range of TVCs, of whom there were 121,000 as of 2018, outnumbering the company's full-time employees.59 The TVCs are hired by third-party staffing firms such Adecco, Cognizant, and Zenith Talent. The common perception of TVCs is that they are part-time workers, but many actually work full-time. Using TVCs, Google is able to save labor costs such as employee health and pension plans while hiring more workers in more in-demand fields such as AI, cloud computing, and data mining. More important, given that the industry is so fast-changing and competitive, it also wants to have a flexible workforce so that it can easily hire and dispose of workers as needed. Because the pandemic and the chronic economic crises reloaded the reserve army of labor as described above, TVCs have long been disposable workers living in fear that their jobs would no longer be needed soon. In the midst of the global pandemic, for instance, Google revoked job offers to more than two thousand temporary and contract workers globally.60
Google's contract workers are far from a homogenous group. They perform a wide range of tasks, including working as software engineers, data analysts, user experience researchers, project managers, linguists, recruiters, lawyers, and content moderators, not to mention nontechnical workers such as bus drivers and cafeteria workers. They are dispersed around the world. Their hourly pay varies from minimum wage to $125 per hour.61 This illustrates that not only low-wage labor is outsourced. Among this contingent and “flexible” workforce, however, is a large swath of entry-level workers who are paid minimum wage, often isolated from their co-workers or working exclusively online. This groups’ much-needed work in the development of the industry is largely obfuscated by seemingly magical technology.
For years, Google has attributed the supremacy of its search results to its automatically configured algorithm. But Yahoo!, one of the leading early search engines, employed human indexers—trained librarians—to collect and organize the information on the web. Human labor was the foundation of its search business, and this distinguished Yahoo! from other early search companies. There are no longer human indexers and librarians per se because automated search engine technology has become the norm, but their work has not been entirely eliminated; rather, this automation has led to the emergence of a new class of low-wage workers who are practically interchangeable and invisible. As David Harvey puts it, “What is on capital's agenda is not the eradication of skills per se but the abolition of monopolizable skills.”62
Some of these low-wage process workers are called human evaluators, often referred to as “quality raters” or “search engine evaluators,” whose task it is to determine the relevance of search engine results before the company releases an alteration to its algorithm. The head of web spam at Google, Matt Cutt, once explained the role of quality raters in responding to the search engine optimization community, which had expressed concern that raters were affecting search results. In 2012 Cutt defended the objectivity of Google's search results and said that human raters were working under Google's Search Quality Evaluation Team only for initial testing phases for proposed search algorithm changes.63
Google started to advertise for quality rater positions in late 2004 and at first hired them directly, but today it uses outsourcing companies specializing in supplying a global labor pool to large multinational corporations.
With titles such as Multimedia Judge, Internet Search Administrator, Web Content Assessor, Query Understanding Judge, Ad Assessor, Internet Crowd Worker, Web Content Assessor, Internet Assessor, and Social Media Internet Assessor, these positions are advertised as flexible, telecommuting, temporary work from ten to thirty hours per week. Since the positions are often advertised as work-at-home jobs, they may target stay-at-home mothers on websites like workathomemom.com, telecommutingmommies.com, and baycenter.com. These workers make up Google's search quality evaluation team, but their salaries and working conditions are far from the idyllic conditions on the Google campus that the media touts. They are required to pay their own expenses for high-speed Internet connections and to have a smartphone and tablet and ever-changing computer technologies for their tasks. The average salary for these permanent temporary quality raters is between $12 and $16 per hour with neither benefits nor job security. The positions often require employees to have a bachelor's degree or four years relevant work experience; however, quality raters do not enjoy the prospect of moving up to a full time career at Google or other Internet companies. While their work is tightly connected to the engineers who design Google's algorithm, quality raters do not have any direct interaction with engineers or other workers at the company because they are managed remotely.
The major tasks of quality raters are to evaluate search and/or advertising “relevancy,” label spam, and flag problem pages as engineers constantly tweak the algorithm. According to Search Engine Watch, whereas the nature of the job for quality raters is presented as flexible and self-directed, it is routinized, mechanized, and tightly managed. The raters perform their tasks according to a manual of specific strict guidelines more than 170 pages long provided by the company.64 Google points out that the work of quality raters is far from simple because they deal with many complex cases.65
According to a Google rater in her 2012 interview with Search Engine Land, raters had to meet their productivity goals to stay on the job. She described the nature of the work and said that there was a certain number of tasks that they had to complete every minute.66 If they fell behind in terms of productivity, they could be put on probation and could not work during that period.67 The quality of raters’ work was tracked based on staying within the time period for rating tasks and the number of tasks that had to be returned.68 If raters’ quality was not up to the company's standard, they were terminated. She offered that “it's a very controlled work environment.” Given a large pool of RAL who are always waiting to be pulled into the workforce, Google can easily dispose of them as needed.
Quality raters for searches and ads are not exclusive to Google or to search engine companies; rather, they are a standard workforce in many Internet companies, including Facebook, which rely on search as a basic function in terms of access and revenue generation.69 It is no longer a hidden fact that there are tens of thousands of quality raters around the world working for the Internet industry.70 It is important to note that search quality raters are merely a small segment of the substrate of increasingly low-wage processing workers in the industry. Other low-wage workers include content moderators, image viewers and taggers, data labelers for AI, and speech transcribers for voice search. Mary Gray and Siddharth Suri call online process staff “ghost workers,” and these are the ones who empower search engines, mobile apps, social media, and e-commerce.71 Tarleton Gillespie calls them “custodians of the Internet” who labor behind platforms.72 Ursula Huws intervenes and notes that this kind of work is often characterized as a typical and new kind of labor practice opposed to full-time permanent positions with benefits, but she rightly points out that their work can't be fully explained without looking at the broader changes in the labor market over the decades and separating them from the rest of the workforce.73
In the late 1970s, with the dawn of the neoliberal era—which Harvey posited was a project of the corporate capitalist class seeking to curb the power of labor74—deregulation of industries, privatization of the public sector, attacks on trade unions, and weakened labor laws began to restructure the existing working class.75 Facing a crisis of falling profits and heightened international competitive pressure, companies embraced new electronic technologies to increase productivity and reorganize labor to ensure continuing accumulation. Companies adopted so-called flexible organization along with “lean” production to quickly respond to rapidly changing market conditions and international competition. The flexible organization combined with new technologies enabled companies to automate, standardize, and mechanize by breaking down and compartmentalizing much of the labor process. The lean or “just-in-time” production model, with extended supply chains supported by network technologies, segmented various labor forces to increase efficiency and control.76 These factors have resulted in the growth of the flexible and contingent labor sector—temporary staff, on-call workers, independent contractors, and the like—in the production of goods and services. Contingent work is labor's version of the just-in-time inventory system for lean production: a just-in-time workforce.77 With weakened labor power and constant restructuring of work processes with new labor-saving technologies, the postwar employment model of full-time permanent jobs with benefits has precipitously diminished; the majority of jobs being generated today are low-wage positions with few fringe benefits. This trend was reaffirmed during the 2008 Great Recession and onward.
In the “recovery” from the 2008 recession, the main employment gains were concentrated in lower-wage occupations.78 According to a 2021 Brookings Institute report, 44 percent of all US workers aged eighteen to sixty-four have low hourly wages of $10.22 and annual earnings of $17,950 on average, nearly one-third live below 150 percent of the federal poverty line, and females make up 54 percent of low-wage workers, representing a higher percentage than the 48 percent of females in the overall workforce.79
Advanced networked technologies are no longer simply applied to routine tasks; rather, they are being extended across occupations such as service, media, transportation, manufacturing, retail, healthcare, and education. Networked digital tools have been deployed to further atomize tasks, aggregate individuals who connect through the network, and control them remotely across national boundaries. Huws describes them as the “automatized workforce” in which workers are made easily interchangeable and exchangeable by stripping the collective power of labor.80 Today's Internet platforms, built by tech companies such as Uber and TaskRabbit to organize the labor market, are extensions of this trend as they intensify work and give more power to capital to control and monitor workers. The deskilling and degradation of work that Braverman noted decades ago persist today throughout the production of goods and services. A new wave of digitization, automation, and standardization is once again repositioning and restructuring labor.
By locating Google's low-wage processing work within this broader changing labor market, one can see a commonality in labor organization. Despite the seemingly different nature of their businesses, tech companies such as Amazon and Uber have a similar labor structure. Like Google's, Amazon's engineers are paid between $125,000 and $270,000, with equity ranging from $25,000 to $150,000 or more;81 however, the companies rely on a massive number of low-wage workers whose work depends heavily on and is controlled by intricate supply chains enabled by network technologies. Amazon, trailing only Walmart, is the second-largest private employer in the United States with more than one and a half million workers, among them warehouse staff and delivery drivers. In 2021 Amazon hired 33 percent of all warehouse workers in the United States.82 It breaks down its individual pieces of the global supply chain to speed up the delivery process, and the workers are pressured to adapt to this supply chain. The difference between Google and companies such as Amazon is that Google's low-wage processing workers are less visible because their work is dispersed over networks and their tasks are separate from the production process. Meanwhile, workers in Amazon's highly automated fulfillment centers and Uber drivers have been more visible recently because of their efforts to organize in response to exposure to extremely grueling working conditions.83 The commonality between them illustrates the deteriorating working conditions of the working class.
This shows that the search industry's occupation structure is far from an exception or limited to the tech sector. Rather, the model is prevalent across corporate America and even public institutions, reflecting the broader labor structure in today's capitalism. The majority of tech firms have used contract workers for 40 percent to 50 percent or more of their workforce.84 Similarly, McDonalds, Nike, and the federal government all rely on a form of precarious labor. In the federal government, the largest employer in the nation, more than 40 percent of the workforce consists of contract workers.85 Today more than one-half of faculty positions in higher education are for part-time adjunct instructors,86 and between 20 percent and 22 percent of workers in the National Public Radio newsroom are classified as temps.87 As Jamil Jonna and John Bellamy Foster posit, precariousness is not new; rather, it is “a defining element in working-class existence and struggle” and at “the fulcrum of the general law of capitalist accumulation.”88
Why, then, was Google's reliance on a large number of contract workers so surprising at first to the public and the media in the United States? This was the success of the ideological work done by capital, the capitalist state, and post-industrialists to privilege the information sector as a site of capital accumulation. They have persistently crafted a message that the new economy, built on information, is a path to prosperity for all and different from industrial capitalism, which historically profited from exploitative labor practices. Yet Google exemplifies a new information industry with the same old industrial capitalist practices.
If the “automatized workforce” represented by Google's quality raters is at the bottom of the paid labor supporting the Internet industry, there is another category at the bottom of the bottom: a great mass of unpaid user and consumer labor. Braverman didn't explicitly address the relation between capital and consumers; however, his conceptualization of the deskilling process inscribes consumers’ unpaid labor, which is enabled and expanded by automation and standardization. Anyone, including the platform's users, can perform tasks.
Internet user and consumer labor as a form of unpaid work has been the subject of numerous debates among scholars.89 The notion of user activities on the Internet as a form of free labor reignited the interest in the theory of labor value. Christian Fuchs draws from Dallas Smythe's concept of audience commodity and argues that users as prosumers are being exploited in order to generate surplus value for the Internet and social media platforms.90 Challenging Fuchs, Arvidsson and Colleoni posit that capital accumulation and realization of value in social media are not due to the exploitation of user labor, rather, they come from rents in a reputation-based financial market.91 The debate revolves around the kinds of commodities and values that ad-sponsored Internet companies are producing. Despite disagreement among scholars, however, there is little question that unpaid labor rightly offered critiques of and intervention in the initial celebratory techno-utopian optimism of the Internet industry, and the debate draws attention to the role of unpaid labor in Internet companies’ profit making. There is a long history of appropriation of unpaid labor by capital. Thus, it is helpful to examine unpaid labor within the historical development of capitalism and question the role of unpaid labor generally and in relation to paid labor.
Ursula Huws points out that there are various forms of unpaid labor in capitalism including domestic labor, consumption labor, unpaid artistic work for self-expression, trade, and volunteer labor.92 It is a challenge to discern specifically how Internet users’ labor is deployed in the Internet sector considering its complexity and scale. Yet Huw's concept of consumption work or consumer labor is instructive as a starting point, for it describes how formerly paid work is replaced by unpaid labor that takes place both online and offline today and directly contributes to capital accumulation.93
One of the most prominent examples of capital's appropriation of unwaged consumer labor is taken from the history of self-service, an increasingly common business model today. It illustrates how portions of the work processes in diverse industries have been transferred to unpaid consumers to increase profit and productivity. Widespread today, this business model goes back to the early twentieth century in the retail industry and the late nineteenth century in the fast-food restaurant sector.94 With the expansion of mass production of food manufacturing, this business model can be seen in the early self-service chain grocery store Piggly Wiggly, established in 1916 by Clarence Saunders in Memphis, Tennessee.
Saunders at first modeled his grocery store on that of Albert Gerrard, who owned a small grocery store in California; but Saunders popularized self-service, adopting a cafeteria-style restaurant and grocery model where consumers served themselves. Saunders franchised Piggly Wiggly across the United States using his patented interior store design with its highly standardized floor plan and fixtures.95 Tracy Deutsch's study revealed that the first principle of Piggly Wiggly was uniformity. One chain manager noted, “Every store must do everything in exactly the same manner…this is one of the greatest advantages of our system. Clerks, goods, fixtures are interchangeable.”96 This self-service mode was increasingly adopted in other retail stores during World War I with its labor shortages and rising labor costs, and it accelerated with intense competition between chain stores and independent grocers in the late 1930s.97
This adoption of a self-service model was not exclusive to the retail food industry. The Bell Telephone Company, for example, started to automate its local phone service after World War I. Michael Palm's study demonstrates that by 1930, after a long labor struggle, nearly one-third of all telephones in the United States were rotary phones that automatically connected local calls, replacing many telephone operators,98 though the full brunt of the evil combination of automation and self-service in US telephony was only felt beginning in the 1950s and 1960s.99 Today, the self-service principle has been diffused and normalized throughout many other retail and service industries such as home improvement, do-it-yourself furniture, banking and online banking, pharmacies, airports, online ticketing, automated phone systems, print-on-demand publishing, and post offices.
The logic of self-service in capitalism is to transfer tasks from paid workers to the unpaid consumer, allowing firms to cut labor costs. Thus, many service industries in which consumers and firms interact at the stage of production continue to make an all-out effort to incorporate unpaid consumer labor as a means of improving productivity and lowering labor costs.100 In the 1970s service sectors in the economy were on the rise and service businesses were increasingly rationalized and standardized, mirroring the model of the mechanized manufacturing industry. The concept of the “production line approach to service” or “industrialization of service,” introduced by the Harvard economist Theodore Levitt (no relation to William Levitt, whose assembly line produced the Levittown suburban housing developments) applied manufacturing logic to service industries in which processes were simplified, standardized, and routinized.101 The adoption of this concept created the conditions to further incorporate consumer labor into the production of goods and services.
The question for capital had been how much value could be extracted from unpaid consumer labor, taking into consideration the cost of incorporating and managing that labor. Although the methods and levels of utilization of consumers’ unpaid time in capital's profit-making ventures today vary depending on the business, from simply contributing comments about a service or a product to improve quality and quantity for production to completing tasks on behalf of and replacing employees, the use of unpaid consumer labor is radically increasing with the introduction of new information technologies that offer newly direct interactive links between consumers and producers. Capital continues to innovate mechanisms and territories where consumers can be plugged into the labor process. It is constantly reinventing and reconfiguring labor processes via new technologies, management skills, and business models.
What is distinctive about today's unpaid consumer labor, which is integrated into everyday activities, is that it involves the appropriation of various forms of involuntary as well as freely donated labor: usability tests, software bug reports and fixes, comment and content generation and moderation, spam monitoring, map building, translation, service rating, and providing all manner of feedback. While these tasks are not productive labor in the sense of paid labor, they contribute directly to capital accumulation by the performance of various concrete tasks alongside paid labor or in displacement of it. The search engine industry came to rely on unpaid labor as a business strategy.
As the search engine industry became automated and industrialized, the initial need for and use of human labor did not disappear; rather, it intensified and changed form, as large-scale commercial search engines predicated their business models on relying on unwaged labor as well. In the process of its expansion, the search industry has not only created low-wage labor as described above but also externalized many of its labor processes by incorporating a global pool of paid labor. Although user labor is not paid, it is visible and highly valued; in this respect, it contrasts with unpaid domestic labor, which has long been invisible despite its value to capitalists in allowing them to set below-subsistence wages.102 Given the significant role of unpaid labor in capital accumulation, it is insufficient to understand the labor process of the search engine industry only through the analysis of paid labor. The analysis needs to be extended by relating the search engine industry's incorporation of unwaged labor and its relationship to waged labor within the longer historical process by which capital has continually sought to reorganize labor and lower its costs, among other ways by making greater use of unpaid consumer labor.
Debate about the role of unpaid labor in corporate capital's profit making has been reignited as Internet technologies have provided a ubiquitous platform on which to easily aggregate and manage the entire process for a globally dispersed unpaid user workforce. From its early stages the search engine industry viewed unpaid labor as a valuable resource and competitive advantage. As the industry developed, a range of work including content creation, usability testing, translation, map making, and data correction was performed by unpaid people that could be done by paid workers. It's not easy, however, to extract the precise part of work that is performed by volunteer unpaid labor and that performed in exchange for services given the complexity of the kinds of work involved in the search industry. Thus, the focus here is simply to highlight some of the work that is performed by unpaid labor.
In the 1990s Yahoo! was the leading search engine in the market with its web directory indexed by paid workers. NewHoo was launched to compete with it. NewHoo's founders, four Sun Microsystems engineers, noticed that Yahoo! was struggling to keep up with the speed of web growth and maintain fresh content to draw user traffic.103 They recognized that building a directory was extremely labor- and capital-intensive and decided to apply the open source idea to building a web directory by recruiting volunteer workers as their main source of labor to speed things up.104 As capitalists provided tools for their waged workers, NewHoo provided tools to their unpaid workers, who, in exchange, selected, described, and organized websites and added to the directory. By the time NewHoo was acquired by Netscape in 1998, the search engine had compiled one hundred thousand websites with the effort of more than forty-seven hundred volunteer editors, compared to Yahoo!'s seventy paid editors.105 NewHoo sold its directory for $1 million to Netscape, which promptly renamed it the Open Directory Project (ODP). At that time, ODP already had 1.6 million entries, surpassing Yahoo! to become the largest human-edited directory on the web.106 The major search engine companies, including Lycos, HotBot, Ask Jeeves, and Google, incorporated ODP to augment their search databases and to get fresh content, which allowed them to level the content playing field. When Netscape was acquired by AOL in 1999, the ODP was one of the key assets included in the acquisition.
Ironically, the ODP was often held up as one of the best examples of the new mode of production occurring outside the capitalist market. Yet the ODP was initially built by appropriating unpaid labor specifically to contribute to a profit-making investment rather than to challenge capitalist social relations. The ODP was firmly rooted in a capitalist market in which unpaid labor was integrated into and subordinated to a capitalist accumulation project. NewHoo was not the first or only Internet company to deploy volunteer labor to supplement paid workers. In 1994 a start-up called GeoCities was built with a business model based on “community.” GeoCities founders David Bohnett and John Rezner needed a large amount of content and traffic but had very limited capital with which to build their site. Instead of hiring paid workers, Bohnett and Rezner utilized this community-based business model as a way to extract labor power from its users. GeoCities provided free web hosting, suites of utilities, and other “Geotools” to its members, who were called “homesteaders.” In exchange, unpaid users created content by building websites focusing on their interests and organizing collections of member web pages by theme and subject for GeoCities, which then used members’ websites for advertising to generate revenue. Contributing to the firm's success was its cost-efficient editorial structure: the most labor-intensive part of the work was performed by volunteers.107 Bohnett touted the fact that GeoCities had seventy-five employees and nine hundred thousand editors, which meant seventy-five paid workers along with nine hundred thousand unpaid editors.108 The volunteers didn't work forty hours per week, but their effort was nonetheless significant because it was absorbed into GeoCities's profit making. The company leveraged the content created by unpaid users to generate more market share and more traffic. In 1999 GeoCities was purchased by Yahoo! for $3.57 billion.109
The extensive use of unpaid labor by Internet companies went unchallenged until a group of volunteers at AOL's “Community Leader Program” reported AOL to the Department of Labor, asking the department to investigate whether the company had violated the federal Fair Labor Standards Act. They pointed out that they were treated like any other paid employees, filing time cards, working specific shifts, and so on, and therefore should be compensated.110 America Online, the largest ISP at the time and a major market actor across the length and breadth of cyberspace, had recruited a large number of volunteers as “community leaders” to perform routine tasks such as answering subscribers’ questions, maintaining chat rooms, and offering technical support in exchange for waived or heavily discounted monthly AOL connection fees. At its peak, AOL had up to sixteen thousand volunteers, some as young as twelve years old,111 compared to twelve thousand paid employees.112 One former executive estimated that the value of work performed by volunteers was as much as 30 percent of the company's annual revenue,113 and Forbes reported that AOL saved almost $1 billion in expenses from 1992 to 2000.114 In 2001 the Department of Labor declined to take any action against AOL, reasoning that the agency had limited resources and that it was “inappropriate” for the government to intervene in the dispute between AOL and its volunteers. Later, the group of volunteers filed a class-action lawsuit against AOL, which settled the case for $15 million.
This brief moment when the public questioned the use of unpaid work within social relationships between labor and capital quickly disappeared as the Internet sector increasingly became a site of economic growth. The deployment of consumer labor in the new economy has been described in the business literature with such neologisms as “co-creation,” “co-innovation,” and “democratization of innovation.”115 The definition of co-creation or coinnovation here is the creation of value jointly with consumers at the behest of capital. Leading business scholars have started to point out that the future of competition depends on this approach to value creation, one based on a supposed individual-centric co-creation of value between consumers and companies rather than company-centric value creation.116 This approach goes beyond earlier forms of self-service such as pumping one's own gas or making a withdrawal from an ATM because unpaid consumers are now actively drawn into capital's profit-making pursuits and participate directly and indirectly in the creation of value.
The concept of co-creation was once a business mantra—“users must be treated as co-developers”—at the very heart of many early Web 2.0 projects.117 The ubiquitous appropriation of unpaid user labor by Internet companies usually is no longer visible today; while its status as labor is effaced, user labor is hailed as typifying a supposed Internet-based culture of participation, democracy, and “open innovation.” In the Internet sector today, the incorporation of unpaid labor in profit making has become increasingly standardized in improving companies’ competitive advantage.
In the case of the search engine industry, companies depend on unpaid user labor for the most capital- and labor-intensive part of their work—providing feedback on algorithms, creating content, constant testing of new products, and correcting the data in digitized content. One of the most apparent of these tasks is providing feedback on search engine algorithms. Users are unwitting co-developers with engineers, assisting in the refinement of a search engine company's core algorithm technology as they go about their everyday search activities.
On average, Google changes its search ranking algorithm five hundred to six hundred times annually.118 These changes are based on more than two hundred different ranking factors, including user interaction/user intent, which signals Google to tune the search algorithm.119 In 2018 Google had devoted about one-third of its workforce—around 30,000 employees out of 98,771—to research and development. This segment of the workforce has responsibility for working on the design of information systems, including its algorithm and new products and services. Along with them, Google might have approximately ten thousand paid quality raters, and an unknown number of paid usability testers, but the most significant group of workers who provide the constant feedback used to improve the algorithm are unpaid users who perform searches every day.
Algorithms are central to search engine businesses, but they cannot be effective without a large quantity of fresh content to index and deliver to users and advertisers; thus, generating original content has long been a major task for search engine businesses. For this, companies again rely heavily on user labor in creating, uploading, sharing, filtering, and commenting, supplemented by acquired proprietary content and partnering with content providers such as NBC Universal, Sony Pictures, and Disney. Take Google's YouTube site for instance. YouTube, which handles more than 1.8 billion unique users each month, is the world's largest video sharing platform and the third-most-visited site on the web after Google search and Facebook. In 2020 more than five hundred hours of video were uploaded to YouTube every minute.120 Although Google doesn't disclose the exact number of YouTube staff, the unit has between one thousand and five thousand employees, according to the company's LinkedIn page,121 and upwards of ten thousand paid content moderators122 to deal with this huge amount of content and usage.123
Given the enormous scale of the site and the massive amount of work needed to improve its algorithm, functionality, and interface and to upload and filter new content, this is still a relatively small number of full-time workers. Google once admitted that it was overwhelmed by the sheer amount of uploaded YouTube content needing to be evaluated. In a rare moment of transparency, Google Public Policy Manager Verity Harding openly stated in 2015, “We do rely on our one billion-strong community to help us flag violations of our policies.”124 The filtering of media content is one of the most labor-intensive parts of YouTube's business. Since 2012 Google has used volunteers for two different tasks: Trusted Flaggers evaluate and report content that violates the site's guidelines, and YouTube Contributors look after the site's help forum and social media and respond to users. In 2018 YouTube generated more than $15 billion in ad sales. In the same year, Netflix earned $15.8 billion but spent $8 billion creating original content.125 If YouTube had to pay its one billion volunteers, could it generate $15 billion in revenue?
With its creative deployment and appropriation of unpaid labor, Google was one of the industry's early trendsetters in co-creation. This was manifest as the company employed from its inception the so-called beta business model, which mirrors the open-source community's bug-fixing approach, in which its products are released incomplete and during development. At Google, the motto has long been “Launch early and iterate,”126 meaning that the iteration process relies on work by users to assist Google in perfecting its products. A product would normally be tested by paid workers or internally by Googlers, but Google decided to release products that are perpetually in beta versions to large numbers of users. According to Jeff Jarvis, this was Google's way of saying, “There are sure to be mistakes here and so please help us find and fix them and improve the product,” as the company monitors user activities to see how new “free” products and services are used and which features are rejected and adopted.127 The logic behind releasing “perpetual beta” products is not to conduct a technical experiment; rather, it is a business strategy to transfer part of the work to unpaid labor.
All of Google's products and services, for instance, Google Maps, Google Translate, and Google Health tap into unpaid labor from around the world. Google has boasted that its translation service was powered by AI, but it was actually a combination of machine learning and volunteer translators. Since 2006 Google volunteers have run the webmaster forum, covering fifteen languages and dealing with fifty thousand threads per year. In 2018 Google revamped and renamed its volunteer program, now the Product Expert Program, with a goal of recruiting volunteers to run Google forums, test beta products, and clean spam on Google Maps by correcting fake listings, bogus reviews, and so on.128 This volunteer program is organized hierarchically, from community specialist to Silver Product Expert, Gold Product Expert, and Platinum Product Expert, assigning different tasks depending on volunteers’ levels of contribution. In exchange for free labor, Google rewards volunteers with “perks” such as testing Google's new products before release, direct feedback to Google employees, and having “special badges.”129
Imagine how many quality raters, usability testers, translators, and cartographers Google would need to hire to perform these myriad tasks. Instead of paid workers, Google is using the old but familiar community-based business model. This “community” of covert strategic labor has been a vital force in the industry's profit making and development and in the transformation of search engine technologies into one of the most dynamic information industries.130
The appropriation of unpaid labor in capitalism is not new; rather, various forms of unpaid labor—by women, prisoners, consumers, and interns—have always been part of the capitalist system. What is new in the search engine industry is the sheer scale and range of unpaid labor taken by capital in our daily lives. An important question needs to be raised: What's the relation between paid and unpaid labor considering that within the very definition of capitalism, workers must sell their labor to the market in order to survive?
As illustrated above, the search engine industry's accumulation process relies on both paid and unpaid labor, which are dynamically intertwined. With automation, mechanization, and routinization, capital has deskilled workers, resulting in the hiring of low-wage processing workers and increasing the RAL as well as the transferring of tasks previously performed by paid workers to users. With capitalists’ cut-throat competition and constant restructuring of labor for ongoing accumulation, the combination of a reserve army of labor, appropriation of unpaid labor, and new technologies is driving down wages, increasing exploitation, creating precarious working conditions, and greatly increasing corporate profits.
Several questions remain: Why are people willing to perform voluntary work for Google and the other Internet firms? Users are not merely performing work on behalf of Google's business. Their everyday information-seeking activities are also rewarding in and of themselves, providing entertainment, communication, work, and education. Given that users receive benefits from commercial Internet services, one could argue that users are willing to trade their labor in exchange for services. The question then needs to be shifted from “why?” to “at what price?” The price visitors pay to use those services is to live in a constant state of surveillance in which their moves are gathered, monitored, extracted, and analyzed for corporate gain. Is this a fair trade? The absence or lack of public information provision offers few choices for the public; however, as Chapter 1 illustrated, the choice wasn't absent, it was consciously removed via policy and capital.
Braverman argues that capitalism tends to deskill work so that tasks can be interchangeably performed by any group of workers. This feature of degraded work has been carried into today's Internet sector even though new kinds of skilled work are being generated. No longer are librarians cataloging websites behind the search engine; now thousands upon thousands of quality raters are looking through the web link by link. The common myth of the highly automated search engine industry is that it is generally reliant on highly skilled workers. Yet the need for and use of human labor have by no means been eliminated by automation; on the contrary, these needs have actually intensified. Google deploys an array of highly skilled and educated employees but also has legions of contingent and low-wage processing staff, unpaid users, and volunteer workers who substitute for paid workers. That is how Google and other Internet companies are able to support thousands of highly paid workers and generate billions of dollars of profit. At the same time, even highly skilled workers are not secure because the industry is constantly trying to enlarge its pool of labor through automation and tapping into foreign labor markets. Google's accumulation and growth are mainly attributable not to its algorithm but to its use of unequal, exploitative occupational structures and a large cohort of low-wage and unpaid labor.
Google is not an exception; rather, it reflects today's changing occupational structure, which is due to ongoing competition and the reorganization of labor through automation, standardization, mechanization, and declining working conditions and wages. The challenge today is not to define digital labor but to locate a new opportunity for solidarity among the fragmented working classes who are working on- or offline—search quality raters, domestic care workers, fast-food workers, assembly workers, immigrant farm workers, unpaid domestic labor, logistics workers, Amazon workers, Uber and Lyft drivers, and content moderators—within ever-changing organizations and complex social divisions of labor.
The next question to be examined is how the search engine industry controls its workers and how workers are resisting. Google's current hierarchical labor organization alone is not sufficient to generate over $76 billion in net income.131 Google is known for its “innovative” labor management, which supposedly is participatory and democratic; however, Chapter 4 will demonstrate that it is firmly ingrained in the historical tradition of capitalist labor management built on the twentieth century's welfare capitalism and Frederick Winslow Taylor's scientific management of the workforce.