Class, capitalism and the tech industry

June 24, 2015

David Judd and Zakiya Khabir look past the myth-making and public relations propaganda to find out who does what and for who in the celebrated U.S. tech industry.

LAST FALL, a video of Dropbox employees arguing with neighborhood kids over a soccer field in the Mission District of San Francisco became emblematic of a conflict generated by the tech boom. Anyone who lives in the Bay Area is inundated with articles, conversations and occasionally protests about the negative impact of the sudden injection of tech wealth on residents.

There's no question that the expansion of the information technology industry has often come at the cost of the displacement of its neighbors (see here, here and here)--and while Silicon Valley might be its epicenter, tech-driven gentrification is by no means an issue confined to the Bay. But there are also fault lines within the tech industry itself, whose surface appearance often belies the depth of the conflicts, potential and actual.


Class, Income and Power

Conventional wisdom in the United States has that we are all--all of us who matter, anyway--middle class. If we define class by income percentiles, and bucket the middle half or three-fifths together, that will be true by tautology. But let's see how far we can get if we try a different approach, as advocated by Michael Zweig:

Inside Yahoo's corporate offices
Inside Yahoo's corporate offices

"Class" must be understood in terms of power rather than income, wealth, or lifestyle, although these do vary by class. Using power as the starting point allows us to see class as a dynamic relationship rather than as a static set of characteristics. Investigating class as a question of power also makes it possible to find the organic links among class, race, and gender...which are [themselves] best understood as power relationships...

The working class are those people with relatively little power at work--white-collar bank tellers, call-center workers, and cashiers; blue-collar machinists, construction workers, and assembly-line workers; pink-collar secretaries, nurses, and home-health-care workers--skilled and unskilled, men and women of all races, nationalities, and sexual preferences. The working class are those with little personal control over the pace or content of their work and without supervisory control over the work lives of others...The United States has a substantial working-class majority.

This working class includes many people in the tech industry: manufacturing workers, janitors, office workers and user support staff. As we will argue below, it also includes most programmers, despite their relatively privileged status in the current labor market. And it includes a newly prominent category of workers like Lyft and Uber drivers and TaskRabbit "Taskers," whose independent contractor status is a legal fiction covering precisely managed working time.

Tech workers, in this broad sense, make up a growing segment of the U.S. working class. Employment in technical occupations across industries grew by 77 percent from 1997 to 2012, as companies outside the tech industry hired hundreds of thousands more developers and IT staff.

Within the information technology industry itself, jobs have been moving sideways from manufacturing to services. In the decade leading up to 2012, a 27 percent drop in "computer and electronic product manufacturing" jobs and stagnating employment in "other information services" was just barely made up for by a 41 percent increase in jobs in "computer systems design and related services" and a 13 percent increase in "software publishing" jobs.

In general, though, true employment by tech companies is probably higher than what is counted in government statistics, given the continued trends toward outsourcing and obfuscated employment relations. Altogether, at least 4 million people, or 3 percent of the workforce--and probably twice that--either work for tech companies or in an IT role in another sector. Some 3.9 million people worked in technical occupations in 2012, and an overlapping but distinct 3.4 million worked in computer, software and Internet industrial categories.


Transforming the Working Class or Making It Obsolete?

But is this a misleading picture? Are "knowledge workers" actually doing something very different from traditional work--and, in fact, making the traditional working class obsolete? Jeremy Rifkin argues in his 1995 book, The End of Work:

The information and communications technologies and global market forces are fast polarizing the world's population into two irreconcilable and potentially warring forces--a new cosmopolitan elite of "symbolic analysts" who control the technologies and the forces of production, and the growing numbers of permanently displaced workers who have little hope and even fewer prospects for meaningful employment in the new high-tech global economy.

Actually, this is an old argument. In 1930, the economist John Maynard Keynes predicted a utopian 15-hour workweek within a few generations. Less comfortingly, the revolutionary activist and author James Boggs wrote in 1963 that "automation and cybernation are shrinking rather than expanding the workforce...[and] work is becoming socially unnecessary." He concluded:

Today the creative work of production is being done by the research engineers, the program planners, the scientists, the electronic experts...What they are creating is a mode of production which, as long as the present system continues, excludes more and more people from playing any productive role in society.

Therefore, Boggs concluded, there were already living "millions [who] have never been and never can be absorbed into this society at all."

Boggs' book was in many ways prophetic, but not in its catastrophism. Total nonfarm employment in the U.S. was 56 million at the beginning of 1963. It was 121 million in January 1997, when Rifkin wrote. And it started out this year at 141 million.

Employment as a percentage of the population doesn't show quite the same trend--it started 1963 at 55 percent, grew to 63 percent by the height of the 1990s boom in 1997, and started this year at 59 percent--higher than in 1963, but still a long ways from recovering from the dramatic drop of the Great Recession.

Nevertheless, technological advancement and growth have so far produced more jobs, not fewer. Even as automation has slashed employment in the U.S. auto industry and other mid-century heavyweights of the American economy, capitalism has drawn whole new segments of the population into the workforce: women in the U.S., hundreds of millions of peasants in China. This is out of the need for profit, not charity.

Under capitalism, the advance of technology has always thrown workers out of work. Even in the 19th century, Karl Marx could write:

Modern industry never looks upon and treats the existing form of a process as final. The technical basis of that industry is therefore revolutionary, while all earlier modes of production were essentially conservative...[It] revolutionizes the division of labor within the society, and incessantly launches masses of capital and of workpeople from one branch of production to another...[This results in] incessant human sacrifices from among the working class, in the most reckless squandering of labor-power and in the devastation caused by a social anarchy which turns every economic progress into a social calamity.

But what Marx was describing was a continuing series of calamities, without an inevitable end point. At the same time, capital also relies on living labor for its reproduction. Machines, even the most sophisticated, don't add surplus value to the production process. Without exploiting human labor, capital cannot be turned into profit. And so profit-seeking capital finds new areas of investment, new markets and new industries, even as old ones are transformed by mechanization.

Even in the newest and most technologically advanced sectors of the economy, the working class, as we have defined it, remains not only a majority, but a majority with interests that conflict with those of its employers--and a majority with the power that comes from being necessary to the creation and realization of profits.


Who Is a Tech Worker?

In the U.S., the stereotype of the "techie" is probably a young white or Asian man who earns a relatively high salary as a programmer in Silicon Valley or some large coastal city. But while this image has a basis in the exclusionary demographics of programming, it shouldn't be taken as representative of workers in the tech industry as a whole.

This remains the case even if we define the industry in relatively narrow terms--as made up of information technology businesses, ranging from titans like Google, Microsoft and Apple to the myriad smaller software companies and Internet startups that share commonalities of interest, strategy, competition and dependence.

It might be defensible to exclude telecommunications and aerospace businesses that tend to sell hardware to very different markets, are more likely to be unionized and have "old economy" business models, although they employ thousands of engineers and other technically skilled workers.

But it would seem false to exclude companies like Amazon and Uber, which share software companies' origins, culture, venture capital-backed funding model and focus on technical innovation and "scalability"--even though they sell traditional goods and services like books, groceries and cab rides. And it would be truly strange to exclude computer hardware businesses when these, in the form of companies like HP, Intel and Apple, not only have often grown or pivoted to compete in software, but also established Silicon Valley in the first place.

There are still a million people employed in computer and electronics manufacturing in the U.S., many of them performing classic assembly-line work. Production workers in Silicon Valley's factories have historically skewed immigrant, female, and Latina or Southeast Asian. The 55 mainly Latina women who struck a Versatronex plant in 1992 to protest low wages and lack of benefits represent a side of Silicon Valley that is very real, despite being very unlike what we see on HBO's series of the same name.

It's true that the hardware industry in the U.S. has shrunk in recent years, although it's unclear whether the trend will continue, and if it does, at what pace of decline. The semiconductor fabrication engaged in by Versatronex has largely moved overseas in the last two decades, cutting short attempts by workers to organize. But even if the offshoring of manufacturing continues, that will not mean the Internet has moved the U.S. economy beyond material goods.

Instead, Internet companies are increasingly integrated with a growing logistics industry that relies on work which is just as regimented. Journalist Mac McClelland made this clear in a Mother Jones article describing her time in one of the warehouses that companies like Amazon and Apple rely on to store and ship their products:

The place is immense. Cold, cavernous. Silent, despite thousands of people quietly doing their picking, or standing along the conveyors quietly packing or box-taping, nothing noisy but the occasional whir of a passing forklift. My scanner tells me in what exact section--there are nine merchandise sections, so sprawling that there's a map attached to my ID badge--of vast shelving systems the item I'm supposed to find resides. It also tells me how many seconds it thinks I should take to get there.

These companies, like those which sell the services of drivers, cleaners or delivery people, can rely on the pressure of millions of unemployed people to keep workers stuck in jobs with low pay and unsafe, humiliating conditions.

But for good measure, they typically rely on "temporary" staffing agencies and "independent" contractors to maintain some distance from their worst-treated workers--which helps reduce the potential of either consumer backlash or joint organizing with those segments of their workforce that have more immediate leverage. In addition, these giant employers have placed most of their larger facilities in isolated areas with few potential allies or alternative employment opportunities for workers.

This doesn't mean that logistics workers are powerless--far from it. In principle, in the era of same-day delivery and just-in-time production, the potential to disrupt chokepoints like warehouses offers real power to an organized workforce.

In practice, whether the leverage of those who work at these critical points in the supply chain correlates with the necessity of their labor or with the number of substitutes available on the market depends on whether these workers are negotiating collectively or as individuals. As labor historian Kim Moody points out, the longshore workers who today have some of the strongest and most militant unions were once among the most precarious of employees.

Still, with the exception of some strikes in Walmart's supply chain, this potential is mostly unrealized, at least in the newest corporate giants. As a result, these workers are left with a miniscule slice of the tech industry's revenues.

The era of "big data"--of companies built on collecting, centralizing and monetizing information about human activity in new ways--still requires humans to sift and act on that information at key points. We are talking about as many as 100,000 workers--often based overseas in low-wage economies like the Philippines, but also employed near corporate headquarters in the U.S. Facebook might have most of us working on our own time to produce its raw content, but that content still needs to be filtered and curated by paid labor into usable forms.

Even those tech companies which sell only software rely on a physical infrastructure of office buildings, fiber-optic cables and datacenters. They also have the same need for administrative work, receptionists and office managers, and HR and support staff as any similar type of organization.

Despite pay that tends to be modest at best, the people who fill these positions often have better conditions than workers performing sweated labor in a warehouse or small factory, and in some circumstances, they have been able to organize.

Silicon Valley janitors have been unionized for two decades--since a bitter 1992 campaign during which employers used workers' immigration status to fire them in large numbers. Facebook's bus drivers voted to unionize last fall. These drivers are now paid $18-20 an hour: above the minimum wage, but not above a living wage in a region where the cost of living for a family of four is $75,000 per year and up.

Many of the white-collar non-technical workers in the industry are isolated at smaller companies. They are disproportionately women, who are particularly likely to face gender discrimination working with better-paid engineers and managers who are predominantly male. It's rare that a startup is willing to admit in a job ad that it's seeking a "Girl Friday"--but it's common for the industry to expect a woman who might be hired as user support to serve, in the words of Melissa Gira Grant, as "a kind of domestic worker, a nanny, housemaid and hostess, performing emotional labor that is at once essential and invisible."

Google employs about as many men as women in non-technical roles, but more than four-fifths of its technical employees are male--and only 2 percent and 1 percent are Latina/o or Black, respectively. The numbers at other big tech companies that have released them, such as Apple and Amazon, are similar.

Economy-wide, according to the Bureau of Labor Statistics, Black and Latina/o employment among software developers and programmers rises to one in 20 each, still disproportionately low--while women remain at one in five. Women leave technical roles at more than twice the rate of men, while Blacks and Latinos are structurally excluded from entering these jobs in the first place.

One study of the tech industry found significant racial disparities around compensation, with white men paid best and Latinos making around $15,000 less, on average, than their non-Latino colleagues after controlling for education, occupation, age, geography, gender, citizenship status, marital status and children in the home.

A great deal of attention is being paid the industry's lackluster numbers around gender and race. Less discussed is how sexual orientation, gender identity, family status and age can also count against tech workers in companies where the phase "culture fit" is a cover for hiring only those that fit into the "bro-grammer" stereotype.


Are Developers Part of the Working Class?

Software developers are much better compensated than the average worker in the tech industry. Last year in the U.S., the median worker earned a paltry $35,540 compared to $91,320 for software developers and programmers. And this doesn't include the comprehensive benefits packages that are the industry norm. Even the lowest 10th percentile of programmers makes $50,920 per year on average.

Programmers typically have more control over their working lives than workers on an assembly line or at a checkout counter. Nobody has yet figured out how to Taylorize software development. The task of writing a working program has yet to be broken down into subtasks that can be performed without specialized knowledge and some grasp of the whole system. Individual output is hard to measure effectively, especially in the short term, which makes it difficult for employers to control the pace of work or use carrots and sticks to make up for a lack of internal motivation on the part of employees.

For these reasons, programmers are often thought of as "professionals," like doctors or lawyers, and not part of the "working class." But this is misleading. Most programmers are neither self-employed nor employed in an organization where they have a meaningful say, and most can't expect any kind of tenure or partnership status, no matter how long they work. If anything, older programmers instead start to experience age discrimination.

Programmers are typically hired at-will, and are not generally accepted to have any professional obligations that could supersede management's authority, nor the ability to do research or pro bono work on company time. Management styles range from the likes of Dilbert's pointy-haired boss to a hierarchy so flat that it's almost invisible--until it's time for somebody to get fired. Still, for most programmers, management authority is a constant.

Of course, this is increasingly true of people who work in hospitals and law offices, too. The larger trend doesn't bode well for the relative privileges that programmers have acquired.

Since the Industrial Revolution, capitalism has shown a long-term tendency to deskill labor and separate conception from execution in a way that increases the power of management. As industries mature, they tend to try to make workers interchangeable units. That hasn't happened to developers yet--but it's worth taking noting when perks that used to grant some workers unusual freedom from oversight--like Google's "20 percent time" or remote work at Yahoo and Reddit, are stripped away. (This despite the disproportionate impact of a forced relocation on working parents, and thus on the industry's vaunted diversity initiatives.)

In part because the measurement of programmer productivity is an unsolved problem, software developers are often expected to work long hours. In Marxist terms, management settles for extracting more absolute surplus value because it can't extract more relative surplus value.

At crunch times, particularly in startups and the gaming industry, 60- to 80-hour workweeks are routine. Perks like catered meals--increasingly common, at least in areas like Silicon Valley and San Francisco with highly competitive labor markets--also serve the purpose of keeping employees in the office longer.

And programmers can rarely expect to leave work behind when leaving the workplace. Even if they are not on-call to fix a crashed server, they typically need to read, program and attend meetups and conferences on their own time in order to keep up with a rapidly changing field.

The imperative to extract every drop of productivity from the workforce is not unique to the tech industry. It appears anywhere that profit comes from the difference between the value of what workers produce and the wages they are paid--that is, throughout capitalism.

It may sound strange to suggest that someone coding, for example, a smartphone app whose single purpose is to let you send your friends the message "Yo" is not only producing value, but a surplus on top of an above-average salary. Yet in a society in which cash value trumps human need, it's true--at least assuming the Yo app, a real example, ever makes money for its investors.

Despite the most celebrated tech companies' aversion to explicit hierarchy, their widespread use of stock options as compensation, and other corporate techniques to convince workers that every member of "the team" is on the same side, it's clear that the people at the top are know that wages come at the expense of profits.

Thus, Apple, Google, Intel, Adobe, Intuit and Pixar have not yet disentangled themselves from antitrust and class action lawsuits resulting from an agreement--initiated by Google and Apple, and eventually involving dozens more companies, with more than a million employees in total--to cap wages by refusing to compete for each others' employees.

This cartel resulted in several billion dollars in lost wages, which went straight to corporate profits. But this transfer, which the law might recognize as theft, is in fact only the tip of the iceberg. It represents a deviation from the ideal of a competitive market, which itself offers no guarantee to workers that they'll receive value equivalent to what they provide.

Across the whole U.S. economy, hourly productivity grew by 80 percent in the last 40 years, while hourly income for the median worker grew only 10 percent. In the software industry, measured productivity has grown 12 percent a year for the last 25 years--meaning that it doubles about every six years. Wages have increased in tech, but not that fast.

Software developers often buy into the idea of advancement by individual merit--either through a liberal lens in which meritocracy is an ideal we still need to work toward, or a libertarian one in which everyone is already where they deserve to be, top or bottom.

This is partly a trickle-down illusion, based on an aspiration to have more in common with Larry Page, Sergey Brin and Mark Zuckerberg than a bank teller or barista. But it also has a basis in the following reality: Most programmers personally know a lot more colleagues who became unhappy at some workplace and quit for a better job than who achieved something through any kind of collective organizing. In this context, it's easy for those who believe they are being mistreated to feel isolated and even personally inadequate, rather than seeking solidarity from co-workers.

The cultural barriers aren't insurmountable. It's hard to explain this year's union drive at Gawker as the product of some sclerotic old-economy work structure, for example. And collective organizing could very obviously bring benefits that are hard to negotiate individually: transparency around salary and promotions, with equal pay for equal work; reasonable scheduling and accommodations like child care for people with families; the right to contribute to open source projects or veto the release of insecure, privacy-violating or otherwise unprofessional code.

Most important, however, might be the idea of solidarity. One of the only workers' organizations among programmers, WashTech, was launched at Microsoft in 1998 by contractors who were fed up with being treated as second-tier employees. Immigrants on H1B visas frequently face similar problems. Issues of discrimination based on gender and race constantly simmer in tech and occasionally boil over.

Yes, it's hard to imagine software developers on a picket line, especially in the short term. But it's also hard to imagine real progress against sexism and racism within tech--or a world where white-collar tech workers relate to the cities they live in as something other than foot soldiers of gentrification and displacement--without a revival of the old radical idea that an injury to one is an injury to all.


Tech Capital

Even in a world where the rich like to be worshipped as "job creators," the people at the top of the tech industry stand out. It isn't every CEO who gets multiple hagiographic biographies, but it seems that more tech companies than not have a garage-to-riches version of their origin story, in which individual genius triumphs over adversity--even if the adversity is, say, getting booted as CEO of Apple and having to settle for running Pixar for a few years instead.

There are plenty of less glowing stories to tell about tech's top executives: narratives of privilege, excess and incestuous ties to the rest of the power structure. The tale of Forrest Hayes, who died of a heroin overdose on his private yacht, triggering dubious manslaughter charges against a sex worker he'd brought aboard, sounds like something that belongs in a movie about the decadence of Wall Street. But in fact, Hayes worked for Google, the geek vanguard.

Tech bosses are just as fond of less-colorful rich-people-things--like buying politicians and busting unions--as their old-economy counterparts. For example, venture capitalist Tim Draper actually tried to get California split into six states by voters' referendum, though his hubristic initiative failed to qualify for the 2014 ballot.

Microsoft's Bill Gates and Facebook's Mark Zuckerberg have both spent tens of millions of dollars more soberly and successfully, buying support for dramatic changes to public education meant to break the power of teachers' unions. Marissa Meyer, the former Google executive and now Yahoo CEO, spends her spare time serving on the board of Walmart, which has spent years fighting against any kind of organizing by its low-wage workforce.

But regardless of the personal qualities or connections of the people at the top of the tech pyramid, it should be clear that their interests--as members of the 0.01 Percent--will diverge from those of their workers on questions of public policy as well as the simple division of revenue between wages and profits.

A more ambiguous category is the small entrepreneur--the startup "founders" who are now processed in bulk through institutions like Y Combinator and other incubators, as well as hustling on their own. Many of these people have put their whole lives into their businesses, and very few strike it rich. Some are sincerely interested in changing the world for the better.

Nevertheless, at the end of the day, those who take venture capital funding are accountable to their investors, while those who remain independent need their employees to produce surplus value in order to keep the companies afloat. While the venture capital-funded founders, most of whose businesses will never make a profit, are arguably better seen as participants in a big R&D department for their investors than as classic entrepreneurs, their interests as stockholders are still much more clearly aligned with their bankers than with ordinary workers.

In any case, this isn't a stable group. Many of the "founders" go back and forth between working for themselves, contracting and working as straight-up employees. This fluidity both mitigates the financial risk of a business failure and contributes to the strain of pro-business libertarianism among tech workers with marketable skills.

Among those who do make it big, any rebel spirit more often manifests itself in a disdain for regulation and disregard for social costs than a challenge to the interests of the 1 Percent as a whole. For a prime example of the former, see Uber and Lyft sabotaging one another, arbitrarily slashing pay and plotting to smear journalists--stories like that are far more common than Valve's much-touted attempt to do away with bosses.

In the current economy, whether the tech sector is best described as a bubble or a genuine boom, it's undeniable that there are plenty of examples of making it big. That leaves space for some tech workers to get a slice of the pie--even as others, positioned closer to the millions still unemployed, are squeezed.

However, even genuine booms end, and we know all about bubbles deflating. When conditions get worse again for business and margins get tighter, how will tech capital respond? History gives us some evidence.

When they were at the cutting edge of an earlier iteration of Silicon Valley, IBM and HP had reputations as great places to work, much like Google and Facebook today. To be sure, the details differed: rather than over-the-top fringe benefits, the "HP Way" promised employment security, with an explicit no-layoffs policy. But the carefully cultivated sense that labor and management were on the same side was a striking parallel.

That didn't survive the recession of the early 1990s, and new competition during the boom that followed. IBM and HP laid off hundreds of thousands of workers--leading, among other things, to the formation of Alliance@IBM, a minority union, but still one of the only organizations of its type in tech.

The tech industry, especially in its Silicon Valley locus, has been consciously anti-union for a long time. Intel co-founder Robert Noyce proclaimed that "remaining non-union is essential for survival for most of our companies."

These days, that sentiment among tech bosses has, if anything, expanded its ambition. Venture capitalist Marc Andreessen spoke for many of his ilk when he said in 2012 that "there may have been a time and a place for unions," but that was once upon a time, it was never in tech, and that world is long gone now.

As for the motivations of the tech bosses themselves, they may stem from a belief that giving employees a say means losing the flexibility to react to new technology and a dynamic market, and not just a desire to cap salaries and benefits. But it's better taken as a warning that tech's bosses understand their interests don't coincide with those of their employees in the long run. Even that neutral-sounding need for flexibility can rapidly translate into a need to fire people and abandon past promises--as it did for the blue chip giants of the 1980s.

It may be the case that few tech workers currently treated as skilled will see themselves as part of a class with common interests that conflict with those of the bosses until they are confronted with mass layoffs or other attacks on things we've come to take for granted. If so, we won't be well positioned for that struggle.

But no matter who we are in the industry--cleaner, coder, designer or picker/packer--if we don't go to work, our bosses can't make a profit from our labor. In the end, to keep delivering things of value to people, we don't need them: They need us.

Further Reading

From the archives