Rebooting the Tech Worker Movement
Workplace organizing in the tech sector has undergone a steep decline. What went wrong?
When we speak to new workers entering tech, few seem to know about the ferocious activism that spread through the industry a little less than a decade ago. Back then, these battles seemed to be happening everywhere, grabbing headlines in major outlets, week after week, for years on end, from Google and Microsoft employees’ fight against military contracts to Amazon workers’ open letter against the online retailer’s complicity in the climate crisis. According to a recent research paper—co-authored by one of us, JS Tan—such campaigns went from just four cases in 2017 to nearly fifty in 2019, coalescing into something unmistakable: a tech worker movement.
Nearly a decade on, however, that movement has faded. Since the COVID-19 pandemic, tech worker protests have steadily waned, and aside from a few cases, we now see far fewer stories of the industry’s workers uniting to hold their employers accountable. What explains this decline?
One critical factor was a shift in the financial environment. When the movement emerged in the late 2010s, tech workers enjoyed a tight labor market with substantial individual bargaining power. But organizing conditions changed dramatically in 2022. In response to rapidly rising inflation caused by global supply chain disruptions during the pandemic, the Federal Reserve raised interest rates, wiping out hundreds of billions of dollars in the tech industry’s market capitalization and instigating, for the first time since the dot-com bubble burst, a wave of hiring freezes and layoffs across the industry.
For the industry’s activists, this was catastrophic. In the past, tech workers participated in workplace activism from a position of power, knowing that their employers were unlikely to cut employees in such a tight labor market. But now, job security was no longer guaranteed. These layoffs, however, were not only a response to the new financial environment; they were also a clear attempt to take back control over a workforce that employers believe had gotten too unruly. At the heart of this was an all-out attack on what they believed to be the progressive foundations that had turned their workforce against them. Employers took advantage of the new climate to explicitly target “woke” havens like AI ethics teams, DEI initiatives, and content moderation roles. They also took it upon themselves to squash workplace activism—firing organizing workers, cutting off internal communications, and disallowing employees from speaking to the media. In the 2010s, most tech workers wouldn’t have thought twice when signing an open letter or participating in a walkout; now, most have become far more cautious, fully aware of the risks of showing any sign of dissent.
There is no question that the industry’s new hostility toward its own workforce was a factor in the movement’s retreat. Yet deciphering its decline requires looking not only at the external forces but the internal logic of the movement itself. As we see it, the decline was also caused in no small part by its own participants and strategies. While any effort would have faced strong headwinds in recent years, what were the limits to the tactics and organizing strategies that the tech worker movement has employed?
To understand where the movement went wrong, we need briefly to touch upon how it began. In its initial phase, from roughly 2017 to 2020, tech organizing mostly relied on public pressure campaigns that mobilized the press. In an era when tech companies still maintained relatively untarnished reputations, open letters, petitions, walkouts, and op-eds could push employers into changing course. Initially, these strategies—combined with varying degrees of internal organizing—produced some notable victories, resulting in Google canceling a military contract and Amazon taking its climate impact seriously. However, as employers realized that bad press was fleeting and did little to affect their operations, the public pressure strategies used by the worker movement were increasingly deflected or simply ignored. By the end of President Donald Trump’s first term, the tech industry had abandoned its progressive façade, neutralizing the shame-and-blame strategy on which this public pressure approach relied.
It was in this context that a small group of workers at Kickstarter, then still relatively disconnected from the rest of the tech worker movement, did something that many others had not yet considered. While the larger movement focused on external pressure, Kickstarter employees turned to internal pressure, winning the industry’s first wall-to-wall union.
Such a victory was long thought to be impossible in the tech sector. Labor unions, some posited, were antithetical to innovation. Even in the early days of Silicon Valley, when it was actually known for producing silicon hardware, the industry was highly averse to unions. Robert Noyce, cofounder of Intel and a towering figure in the Valley during this time, explicitly said of the industry: “Remaining non-union is essential for survival.” For decades, unions in tech were virtually nonexistent.
This was what made the successful union drive at Kickstarter a watershed moment for a movement in search of new strategies. Unlike the previous era of tech activism, Kickstarter’s workers had achieved lasting structural power that they could codify into a contract. For the first time in the industry, workers would have a seat at the table. And this victory was as significant for the members of Kickstarter’s new union as it was for the wider movement. Upon seeing the campaigns’ success, seeking union recognition under the National Labor Relations Act (NLRA) became the tech worker movement’s north star.
Some participants in the tech worker movement had organized unions elsewhere, but most had never heard of the NLRA, let alone studied the process by which a union gets certified. And even fewer were practiced in the specific work of building a union from scratch.
To fill this gap, tech workers turned to international unions. In the heyday of the public pressure approach, some tech employees consulted with union staffers—but such interactions were rare, and affiliation with an international was rarely on anyone’s mind. That all changed when the movement started to realize the limits of ad hoc media campaigns. In Kickstarter United, which was affiliated with the Office and Professional Employees International Union, many—including the two of us—saw a new model for organizing that offered a path for lasting power. The first step was to affiliate with an international.
Internationals transformed the tech worker movement. Gone were the reckless public pressure campaigns of the past. Workers now had access to expert guidance on labor law and the hard-won lessons that internationals had accumulated over decades of labor struggle. This institutional memory, we thought, would allow a sector entirely new to the tradition, like tech, to bootstrap organizing.
Internationals also gave the movement a clear goal: to win union recognition through a well-defined, legally protected process—what we call the Recognition Playbook. This approach was optimized for establishing supermajority support under the NLRA, which protects private-sector employees’ right to organize without fear of retaliation. Once recognized, management is then required by law to negotiate in good faith for a labor contract, allowing workers to codify the changes they wished to see.
Under the Recognition Playbook approach, the nature of the movement had also transformed. Whereas ethical and social issues—curbing emissions, canceling military contracts, and the like—animated the previous era of organizing, internationals instead focused on bread-and-butter issues: pay, healthcare, working conditions, and job security. Because winning a union required majority support, they had to prioritize demands that were deeply felt and widely held.
Before long, internationals were playing a central role in the tech worker movement, powering a new surge of labor activism across the industry. Around the same time as the Kickstarter campaign, Google contract workers at the outsourcing firm HCL won their union election, and software workers at Glitch followed soon after—the start of a rapid run of labor victories. Union drives also scaled up: In February 2021, workers at the online publishing platform Medium went public after signing up 70 percent of the workforce, and a few months later more than 650 tech workers for the New York Times won their campaign. Even at larger, harder-to-unionize shops, the Recognition Playbook became the go-to framework. Google workers, for instance, launched the Alphabet Workers Union. By the end of 2021—just a year after Kickstarter’s landmark win—a dozen new tech unions had won recognition.
Having both been among the Recognition Playbook’s most committed advocates—for a time even working for an international to push it across the tech sector—we believed this approach would help give the movement a stronger foundation. But even during the most militant, well-organized campaigns, we found instead that tech workers faced a number of difficulties adapting the playbook to their sector. And while not all of their challenges were unique to tech—and the tech sector is certainly not the only one in which workers have failed to utilize the playbook to win unions and contracts—it is important to understand how these challenges worked to undermine the most well-worn path to worker power.
A first challenge stems from the fact that the average tenure at a tech company is just two to three years, meaning that many workers will leave before organizers can build enough support for an election. Where the playbook has worked, it has been at smaller and more stable companies; organizing at larger tech firms would require workers to commit to a struggle whose rewards are unlikely to materialize within their own tenures.
Tech companies are also made up of highly varied roles across different product lines and time zones. Even within the same company, the concerns that animate one group rarely resonate with another. Engineers would often point to stressful on-call duties as their most deeply felt issue—something wholly unrelated to the grievances of product managers. Because of the nature of project-based work, workers are also highly atomized, with daily interactions confined to just a few colleagues. This fragmentation makes it nearly impossible to identify crosscutting issues that can unite the workforce at large. If the playbook depended on widely held and deeply felt issues, the tech industry’s structure made finding them much more difficult.
For our purposes, however, the critical problem with the playbook lies in its assumptions about the labor process itself—and, in particular, where tech workers have leverage. Striking has long been considered the strongest expression of the power and solidarity that workers can accrue in organizing. And throughout the history of labor struggle, the most effective strikes are ones with extremely high participation—so that when workers withhold their labor, it causes employers immediate economic pain. In factories, schools, and logistic hubs, production is directly linked to the hours that workers labor. When workers withhold that labor, production stops and revenue dries up: the assembly line stalls, classrooms sit empty, and trucks stop moving. But at tech companies, the primary work is automation itself. The whole job of a software engineer is to build systems designed to run without them. Because of this, these strikes don’t inflict the same economic pain on tech employers that they might in other sectors.
This is what we saw when the Tech Guild, one of the most militant unions in the industry, went on strike. While bargaining their contract, management was stalling and unwilling to budge on any of the union’s demands. It is normal for contract fights to take a long time, as research from Bloomberg Law shows that the average time-to-contract after winning a recognition was roughly 500 days. And so, for the Times Tech Guild, their strike was meant to force management back to the table and settle on a fair contract.
The strike should have easily achieved this. It was scheduled to take place on the day of the presidential election, perhaps the single most important day for the New York Times every four years. The picket line was well publicized, and with 95 percent support from its membership, the strike checked all the boxes of the Recognition Playbook: high participation, unity, and visibility. From the outside, the strategy indeed looked bulletproof. The plan was to shut down the country’s most influential paper at the very moment when millions of readers would be glued to its coverage of the presidential election. Most unions could only dream of this kind of leverage.
But as these tech workers struck, the paper’s core digital products continued to run seamlessly, and its primary revenue streams were untouched. Unsure if extending the strike would break the platform and force management back to the bargaining table, the Times Tech Guild decided to end the strike. When it was all over, management and the union seemed as far apart as when the strike began. The strike also ended up costing participating members 2.3 percent of their annual wages—more than $3,500 per worker. It wasn’t until more than a month later—over a thousand days after winning their union and double the average time-to-contract—that the Times Tech Guild ratified its agreement. The contract included wage increases of up to 8.25 percent, just-cause protections against arbitrary firings, additional compensation for on-call work, and more flexible hybrid schedules. These were not insignificant wins, but whether they came from the leverage built from striking remains unclear.
The problem is not that the industry is immune to strikes, but rather that the way tech workers can strike effectively has yet to be discovered. Had the strike lasted longer, more of the paper’s digital products might’ve been impacted, and the accumulation of unaddressed software incidents might have grown into an injury that management could not have ignored. Had it been done without warning—with management left unprepared—the site might have been less equipped for a labor stoppage. Or, had it started well in advance of the presidential election, workers’ refusal to participate in preparation for the massive influx of user traffic might have posed a more direct threat to the website’s stability. With such strategies, it’s possible workers could have had a better shot at true economic disruption.
But by the time of the strike, the Times Tech Guild had never tried exercising its leverage, and thus had little data about how effective the action would be. This is in fact the key problem with the Recognition Playbook. Its underlying strategy is to treat organizing much like building a bonfire. Each moment of unrest, each demand, each swell of solidarity is another piece of tinder, piled higher and higher toward the promise of a powerful union. Only when the heap is tall enough to win a majority union are workers encouraged to light it. The problem is that starting such a blaze is much harder in tech, and workers have never learned to strike a match.
Rather than only directing toward the slow and steady work of building majority support for a union, we think that setting these smaller fires is essential. They offer organizers the chance to experiment, test their collective strength, and discover where their leverage lies, so that by the time a majority is won, the union already knows not just how to build the fire, but how to set it ablaze.
The early years of the tech worker movement brimmed with visions of reshaping the industry itself: ambitious climate goals, divestment from the military-industrial complex, and a democratic say over the technologies being built. But the movement’s initial strategy focused too much on building external pressure—at a time when companies began to care less about their public image. The Recognition Playbook reoriented the movement toward building within the shop floor itself, but it also narrowed the movement’s ambitions to more bread-and-butter demands, leaving little room for the bigger ideas and concerns that had animated the movement in the first place.
This is not to say that contracts have no place. They remain powerful tools—and, in the right circumstances, indispensable ones—for protecting workers, codifying gains, and sustaining organization in the long term. But when it comes to the broader ambitions of the movement, they are simply tools. They should be picked up when they strengthen workers’ leverage and put away when they constrain it.
Despite these difficulties that the tech worker movement has faced, the need for it is more urgent than ever. Trump is back in office with an even more aggressive agenda. The most powerful tech tycoons have gone from mere market actors into oligarchs—wielding not just economic but overtly political power. Much of the traditional regulatory guidance around tech has been abandoned. At the same time, AI is being deployed in ways that could displace and discipline engineers, designers, and other knowledge workers, further concentrating power in the hands of this tech oligarchy.
Tech workers thus represent an ever more crucial lever of power. And given tech’s growing role as a political force shaping the conditions of democracy itself, it is not enough to organize solely within a traditional union framework that can only win shop-level demands. The tech worker movement must transform into a project that aspires to build an effective counterpower, and challenge an industry that now stands on the same side as an increasingly authoritarian state.
JS Tan is a co-author of Against Tech Oligarchy, a book about the rise of tech worker activism and the labor politics of Silicon Valley. He is getting his PhD at MIT and previously worked in tech. His work has been featured in the New York Times and MIT Technology Review, among other outlets, and his writing has appeared in the Drift, Jacobin, Foreign Policy, the Baffler, and the Guardian.
Clarissa Redwine is a co-author of Against Tech Oligarchy. She is a tech worker and labor activist. She helped organize the industry’s first wall-to-wall union at Kickstarter in 2019. Her work in the labor movement has been featured on the BBC and in the New York Times, the Guardian, the Verge, TechCrunch, and many other news outlets.