Last week, Bloomberg published an article on Google’s “shadow workforce”: its workers who are not direct employees and thus do not receive Google’s celebrated benefits. According to Bloomberg, these contractors now outnumber Google’s full-time employees, and they are increasingly discontent with being shut out from Silicon Valley’s meteoric success.

Now, the revelations in the article are not exactly news to anyone who has been paying attention to labour issues in the tech industry. Tech Workers Coalition, an organisation trying to build worker power in the tech industry, was originally formed with an eye to building solidarity between full-time and contract employees, and a panel co-hosted by Tech Workers Coalition and Notes From Below in June included a discussion of the working conditions faced by contractors. On the other hand, such discussions have been largely confined to the more critical fringes of the tech industry. The Bloomberg article brings the conversation out into the open, raising awareness in a way that is to be welcomed.

Still, there are limits to what can be said in a mainstream publication like Bloomberg. This is an attempt to fill some of the gaps in that article, drawing on my own conversations with people I know as well as the wider economic context around the tech industry, in an attempt to go beyond what had been published in the article. My own vantage point is unashamedly partisan: although I have worked in tech, I am no longer a tech worker, and my relationship with the industry for the past year has been as someone who writes about it from an increasingly critical perspective.

Ultimately, my goal is to shift the conversation around contractors into a more active register: to make the case for building worker power in the tech industry. In the last few months, we have seen unprecedented displays of tech worker resistance, prompted by a political awakening responding to tech companies’ collaboration with various arms of the US government. Contractor segregation is just the latest front in this ongoing battle, and the Bloomberg article could spark a fresh wave of attention towards working conditions in the industry in a way that politicises tech workers further.

The company’s side of the story

Let us begin with the story the companies like to tell. Bloomberg reports that Google has ~90,000 direct employees and around the same number of contractors. Speaking anecdotally, it seems that other large tech companies have similar ratios. These contractors comprise a number of different roles across the industry, including content moderation, serving meals, cleaning offices, providing security, and writing code.

Why is this a problem? Well, for the company, there is no problem. Hiring contractors is merely an efficient and cost-effective way of meeting staffing needs within the existing system. At most large companies, the typical recruitment process for full-time employees is quite lengthy, so hiring temporary staff is a way for individual managers to bypass that structure, while also gaining the flexibility to let them go with shorter notice. It is also usually more cost-effective to hire contractors, especially for less technical roles. The total costs of a contractor are lower since the company does not have to pay benefits, and in fact are legally restricted from offering certain benefits like free transportation, invitations to company parties, or stock. Finally, as contractors work for intermediary companies, they cannot sue cash-rich tech companies over issues like wrongful termination.

In the case of Google, part of the rationale behind cutting costs is to appease a frugally-inclined CFO who has issued directives to limit hiring. Such directives have a deeper cause, though, and it is one that affects other public companies as well: investors have come to expect tech companies to post impressive quarterly profits while still “maintaining skinnier workforces than older corporate titans”, as Bloomberg reports. Given this dynamic, the company itself is merely doing its best to manage shareholder expectations. In any case, most contractors are non-technical and thus outside the company’s direct area of expertise, so it makes sense to subcontract out to a staffing agency that can find workers more readily.

For their part, contractors are often grateful for the opportunity to work for these companies. For non-technical roles, working at a large tech company usually brings better benefits and feels more prestigious than comparable roles in less lucrative industries. For technical roles, contractors are more likely to have non-traditional backgrounds or less experience, and may find it difficult to get full-time job offers based on their credentials. The temporary job might be seen as a stepping stone to a more permanent job at the company, which they will secure if they prove themselves worthy. In any case, contractors are choosing to work at these companies. If their conditions were that unpleasant, so the argument goes, they would choose to work elsewhere.

The other side of the story

Such is the company’s perspective. There is another perspective, of course, a perspective where “staff” are not merely costs to be cut but instead people: people who are struggling to get by. Such a perspective starts off by recognising that people who take these jobs do not always have much choice in the first place. They may be able to decide where to sell their labour power, but rarely do they have the choice to not sell their labour power at all. If the “choice” is between a temporary contract job at a company like Google, and a much more exploitative job at a different company, then their options are already limited.

The company, on the other hand, enjoys much more favourable labour market conditions for all but the most prestigious contractor roles. Faced with a large supply of un- and underemployed potential workers who are desperate to make ends meet in an era of general precarity, and with a name that would look good on a resume, the company has the most bargaining power in the majority of employment situations. Plus, given the general profitability of the tech sector - and this is especially true for the tech giants - it can afford to pay better wages than other companies for the same role. As a result, it can extend enticing offers, which makes it easier for contractors to tolerate the fact that their full-time colleagues are getting a better deal for similar work.

In other words, these companies treat their contractors the way they do because they can: because they have more bargaining power than the majority of individual contractors, especially those applying for roles conventionally thought of as “unskilled”. As public companies must follow the whims of bottom-line-driven shareholders, they are structurally incentivised to save on labour costs wherever they can. Overspending on technical talent is still acceptable because of the founding myths at many of these companies, leading to a glorification of the “rockstar” or “10x” engineer; paying $1 million per engineer when acquiring small startups is not uncommon, nor is paying astronomical salaries for those with expertise in AI. However, when it comes to workers who are less obviously crucial to production, or whose skills are considered more interchangeable, the company has no reason to consistently pay more than necessary unless there is a countervailing force - either resistance from below, or advocacy from above.

Once you recognise the imbalance inherent in the company-contractor relationship, the story these companies tell to justify their use of contractors starts to fall apart. The most blatant bit of management-speak is their explanation for bringing in outside agencies for roles outside their area of expertise, agencies that are often umbrella organisations who themselves bring on subcontractors for the different roles they have fill. What “expertise” could such agencies really provide? The middlemen in this transaction - those who own, manage, and profit from these outsourcing companies - are not in that business out of charity. They certainly do not have any magically innovative technology for coordinating workers dramatically more efficiently - if they did, Google would have bought them already. What they offer, then, is an arms-length structure designed to allow Google to dodge accountability while paying less for an equivalent quality of work.

This is the same outsourcing story that we have seen in other sectors in recent decades. The whole point of this setup is to shift the burden of responsibility away from Google - to avoid legal entitlement to benefits or liability of any kind - in order to pay less for labour overall. Even if that means some of the money goes to agency fees instead of directly to workers, that is acceptable, as their wellbeing is not Google’s responsibility. The drive to save money on labour is agnostic to where these cost-savings actually come from - whether it is sick pay, or health insurance, or wages. Performance reviews are also off the table, despite frequent requests by contractors for formalised feedback, as performance reviews are tied to raises, which employers would like to avoid. Even the claiming of overtime pay - technically a legal requirement for many of these roles - is, as is common in many industries, often discouraged by management, implicitly or explicitly.

At the same time, a toxic environment is created whereby contractors are implicitly put in competition with each other for the chance to convert into full-time. This is neoliberalism in its most visceral form: the individual worker becomes an atomised unit, forced to display enough perseverance and “merit” to convince their supervisor of their economic value. With such an individualised subjectivity, the possibility of collective consciousness or solidarity between workers is undermined, along with the perception of labour as a class with common interests. For individual contractors in this environment, where the burden of responsibility for one’s survival is placed solely on the individual, it feels like the only option is to go along with the game. The upshot is that labour becomes more fragmented, disempowered, and made to accept fewer material gains despite deepening exploitation.

On a macro scale, this amounts to a shift in class power towards capital and away from labour, both in terms of where money flows and in terms of decision-making ability. An optimist might say that a smaller share of income going to workers means that these companies have more cash to productively re-invest in the company, which would improve production in the long run. But that is not what happens in practice: the tech giants are all hoarding cash right now and, of course, barely paying taxes in the process. This contributes to an ever-growing concentration of capital, exacerbating already unsustainable levels of inequality at a time when many workers are increasingly turning to credit to make ends meet.

Fighting back

Challenging this state of affairs will not be easy, but it is also not impossible. It starts from the ground up: building power from below, one conversation at a time. The example of Bay Area tech company SurveyMonkey is heartening. In response to concerns from full-time employees over the employment conditions of contractors at the company, raised via an internal employee survey, SurveyMonkey worked with its subcontractor agencies to improve the benefits packages on offer. This suggests that solidarity from full-time employees can go a long way towards instituting change.

On the other hand, SurveyMonkey is a private company with only a few hundred employees. Not only are they free from the pressures facing publicly-traded companies, but they are small enough that their internal culture actually matters to current and potential employees. The larger companies, however, have enough inertia that they are essentially machines, devoid of the concerns over culture they may have once had. Microsoft is often spoken of in this lens; Google is not far behind, and Facebook only a little way after that. Forcing policy changes at the larger companies will be more difficult and will likely require a prolonged campaign.

Developing collective subjectivity

Before that, of course, lies the challenge of getting contractors and full-time employees to develop some sort of collective subjectivity in the first place. This is especially difficult at the companies that have a reputation for being innovative and fun to work at - these companies have a way of developing an obstinate loyalty among their workers. Newcomers may be dazzled by the perks, seeing them as something to be grateful for rather than the very attempt to buy their gratitude at a discount. Such a reaction is understandable, especially among those new to the company or industry, as their current roles may indeed be preferable to previous ones. But the scales will fall from their eyes eventually - they certainly did for me.

That point - the point when disillusionment starts to set in - is also the point at which the possibility of change emerges. As a presentation from Tech Workers Coalition explains, disillusionment can be triggered in various ways: standard workplace grievances like bad management or unfair processes; issues around social composition like sexism and racism; and ethical concerns with how their products are being used. The traditional reaction to such disillusionment in our present atomised atmosphere is to individualise and internalise: minimising expectations; rationalising away uneasiness; contorting the self to fit the needs of the corporation.

As inevitable as that sounds, that is not the only option. The alternative is to find new ways that workers can have agency. But this is not something that can be done on an individual basis; for it to be effective, and for it to succeed in making lasting changes, such agency must be collective. Building collective power, from below, is the only way to challenge processes institutionalised from above. Exercising agency on the individual level requires building solidarity on a collective level: recognising that you have more power together. That means bridging the artificial divide between contractors and full-timers.

A preview of the future

There is another factor to consider, and it is one that adds extra urgency. What we are seeing in terms of rising precarity and instability among contractors is likely a preview of things to come for full-time employees too - even the technical ones. It will most likely manifest as a gradual creep upward, affecting certain roles and demographics more than others. With the rise of coding bootcamps and other avenues for attaining technical skills and accreditation outside the standard 4-year degree, conjoined with the increasing availability of tools to abstract away more software development tasks, the supply of potential workers to fill basic technical roles increases. In the long run, in the absence of a strong organised labour movement in the industry, an enlarged supply of potential workers will only lead to depressed wages for those roles. The outcome will be widening polarisation as the ever-shrinking pool of technical workers who are still in-demand - who have expertise in whatever capital considers valuable at any given time - fetch increasingly higher salaries, while the rest are made to compete for the remaining full-time jobs.

Beyond the bottom line

So where do we go from here?

Those working in tech who still believe the story it tells to the world about itself - that it is making the world a better place - should take a cold, hard look at the conditions within their own backyard. Fulfilling the industry’s long-buried promise to be a force for good will require recognising where power currently lies, and challenging it when it is skewed too heavily in one direction. At the moment, power is concentrated at the top and is directed downward, undergirded by a particular economic rationality that prioritises growth in order to fulfil fiduciary responsibilities to keep Wall Street happy. We need to recognise that the current balance of power is not sustainable: not only has it led to drastic inequality among its own employees, but it has had rippling effects on the economy at large, contributing to widening inequality and deepening exploitation on a global scale. Challenging this model will take more than an executive pledge to behave more ethically; it will require structural ballast, which can only come by building power from below.

In some ways, a more empowered workforce could actually be good for these companies. For one, workers might be more productive if they have the ability to challenge inefficient processes or reverse management decisions they believe will harm the company. It could also help their image: these companies love to pay lip service to improving “diversity”, and revisiting pay structures with the goal of dismantling occupational segregation and the contractor/non-contractor distinction could narrow gender and racial pay gaps. Plus, sharing wealth more equally among employees, or even among the wider community, could go a long way toward alleviating negative perceptions among locals who are experiencing tech-enabled gentrification.

Still, it should be emphasised that the point is not to help the company as such. Although building worker power may occasionally lead to an improved bottom line, the latter should only ever be thought of as an incidental outcome and not the ultimate goal. The interests of workers, as a class, are different to those of the company or the executives who currently run them, as we have recently seen in the case of Amazon continuing its relationship with law enforcement despite employee protests. Building worker power is more important than a corporation’s short-term (or even long-term) profits, partly because such a perspective neglects the important externalities that money as a metric is incapable of capturing, but mostly because the interests of an entity as artificial and lifeless as a corporation should always be secondary to the collective interests of the real human beings who work there.

Looking ahead

Building collective power is about so much more than achieving slightly better wages or benefits for contractors. As laudable and worthy as that is in itself, the organising efforts necessary to achieve such gains will also indelibly change the power dynamics within these companies. At the moment, decisions are largely made at the top, with very little real autonomy on the part of workers, despite the fact that they do the actual work that keeps the company running. This is not the only way to run things, but it is what most of us are used to, and so it is difficult to contest without a critical mass of support. Building power from below is a way to reclaim agency by investing decision-making power much more democratically, as a way for workers to realise the control they should have over the products of their own labour.

In the long run, the prospect of building worker power in tech has implications that go far beyond the industry itself. Changing the balance of class forces in an industry as strategically key - and so often celebrated in the media - as tech will have ramifications outside the industry as well, by exposing workers in other industries to ideas of collective organisation. Any successes in organising within the tech industry will further the broader process of rebuilding consciousness around workers as a class, pushing back against the political project to suppress the power of labour over the last few decades.

At this stage, it is not entirely clear what the next steps are. With the exception of some recent organising successes in games development studios, the tech sector is what is known as a ‘green field’ industry without much history of successful unionising, so there are a lot of open questions. On the other hand, tech has always been famed for its ability to innovate and disrupt. Seen in that light, our current conjuncture could be an historical opportunity to invent new models of organising, in order to avoid the tendencies toward ossification that many associate with the organised labour movement. Such an endeavour would be a much more worthy use of tech workers’ skills and efforts than building another app to monetise users’ eyeballs while soothing their conscience via philanthropy.

This is just the beginning, though. As the history of labour organising has shown, the process of organising is itself transformative, and will open up new avenues of possibility at each step. What starts off as everyday conversations between colleagues about workplace grievances can become a powerful movement, grounded in collective solidarity and engaged in a strategic struggle to advance the interests of workers as a class. The path between here and there is unclear, but maybe that is okay. One step at a time.

Further reading


Wendy Liu (@dellsystem)

Wendy Liu is the author of Abolish Silicon Valley (Repeater, 2020). She built the website you are looking at right now.