When software workers ask why the industry has organized so slowly, the usual answers are cultural: engineers are individualists, compensation was too high, or people could simply leave a bad job. Each explanation catches part of the history. None explains the whole workforce, and none turns low union density into evidence that collective bargaining cannot fit software work.
A more useful account looks at the institutions surrounding the worker. Professional identity, career mobility, venture-backed growth, subcontracting, distributed teams, and sharp status differences all made collective identity harder to build. Those conditions shaped behavior for years. They did not remove the underlying conflict over who controls staffing, pay, surveillance, workload, product decisions, and job security.
Begin with the claim being tested
The claim is not merely that software workers have had low union membership. The stronger claim—often implied in workplace conversations—is that something about technical work makes durable collective organization unnatural.
That stronger claim needs evidence. Low organizing rates can have several causes:
- workers may be satisfied with existing conditions
- workers may leave rather than fight to change a workplace
- the workforce may be divided across employers, vendors, and classifications
- professional norms may discourage a worker identity
- organizing may occur around ethics or discrimination without initially using labor language
- employers may have enough resources and flexibility to contain collective pressure
Those explanations can operate at the same time. The important question is which conditions changed workers’ incentives and relationships—not whether software workers possess a fixed cultural trait.
Professional identity offered a different story about work
JS Tan, Natalia Luka, and Emily Mazo begin their ILR Review research from the observation that tech workers have not traditionally been associated with labor activism. Their article, Unlikely Organizers: The Rise of Tech Worker Labor Activism, examines how professional workers who appeared distant from labor politics became participants in collective workplace action.
Software workers have often been invited to understand themselves as builders, experts, creators, or future founders. That identity can be materially attractive. It promises autonomy, influence, and a relationship to the product that feels different from ordinary employment.
It can also obscure a simpler fact: the employer still decides whether a team exists, what is measured, which work is funded, who receives equity, when a product is cancelled, and whether a worker keeps access tomorrow. Professional discretion may be real while those decisions remain concentrated elsewhere.
The effect is not that technical workers forget they are employees. It is that workplace problems arrive in professional language. Understaffing becomes a planning challenge. Unpaid emotional or after-hours labor becomes commitment. Arbitrary promotion becomes a need for greater “impact.” A conflict over control can therefore remain individualized longer than it would in a workplace with a stronger labor tradition.
The entrepreneurial ideal made exit feel more natural than voice
Robert Dorschel’s research on tech-worker subjectivity describes the “entrepreneurial self”: a worker oriented toward networks, projects, autonomy, and personal adaptability. His article, A new middle-class fraction with a distinct subjectivity, connects that orientation to a wider transformation in professional work.
In practical terms, many workers were encouraged to manage workplace risk as a personal portfolio problem:
- learn a more valuable technical stack
- negotiate an individual exception
- transfer to a better team
- collect equity and leave
- move to the next company before conditions deteriorate
For a worker with strong mobility, that strategy can be rational. It can also leave the underlying workplace unchanged and make collective action harder to imagine. The person who exits is no longer present to build trust with the people who remain.
Exit was never equally available. QA, support, operations, content moderation, outsourced development, staffing-firm workers, visa-dependent workers, and people outside major hiring centers often faced different labor markets and different exposure. Treating the most mobile engineer as the representative software worker hid those divisions.
“The tech workforce” was never one employment relationship
A software product can be built by people who share meetings and repositories while receiving paychecks from different companies. A directly employed engineer, a staffing-firm tester, an outsourced support worker, and an independent contractor may experience the same release pressure without sharing the same legal employer, benefits, job security, or bargaining rights.
Distributed work adds another layer. Workers may be separated by office, time zone, country, product group, acquisition history, or remote-work status. Communication tools make coordination technically easy while leaving relationships thin. A large channel is not the same thing as durable trust, and a visible reaction is not the same thing as willingness to act.
These arrangements give employers several ways to keep a problem local. A workload issue can be described as one team’s process failure. Unequal pay can be attributed to geography or employment category. A vendor’s conditions can be declared outside the client company’s responsibility. Organizing has to reconstruct the relationships that the employment model keeps administratively separate.
Relative advantage delayed a shared diagnosis
Parts of the software workforce received high salaries, equity, flexible schedules, and considerable bargaining power during periods of rapid growth. Those gains were meaningful. They also supported the belief that workplace problems could be handled one person at a time.
The weakness in that belief becomes visible when the remedy depends on a decision no individual worker controls. One person can negotiate a raise; one person cannot create a transparent leveling system. One person can leave an understaffed on-call rotation; leaving does not establish a staffing floor for the people who inherit it. A worker can decline one invasive tool by finding another job; that does not determine whether the tool becomes normal across the industry.
Collective organization becomes more legible when workers see the limit of individual exit. That recognition can happen during layoffs, return-to-office mandates, surveillance rollouts, discrimination disputes, or conflicts over what a product is used to do. The triggering issue varies. The shared discovery is that professional status does not guarantee decision-making power.
Tech labor activism often developed through adjacent conflicts
Tan, Luka, and Mazo argue that workplace social activism helped create a path toward labor activism. Workers organized around product ethics, discrimination, sexual harassment, military or immigration contracts, and other questions that were not always framed initially as wages-and-hours disputes.
Those campaigns gave workers experience that ordinary professional life often did not provide. People learned who would take a risk, how management responded to collective pressure, which communication networks could be trusted, and where formal values ended when they conflicted with executive authority.
This does not mean every ethical campaign becomes a union campaign. It shows how collective identity can be produced through action. Workers may arrive at labor politics after a concrete conflict reveals that the company’s mission, internal channels, or individual influence cannot resolve the dispute.
Competing explanations still matter
The research does not justify one universal story about software workers.
Some workplaces may have low organizing interest because conditions are genuinely good. Some workers may value mobility more than changing one employer. Some campaigns may fail because the issue is narrow, relationships are weak, or the proposed organization does not fit the workforce. National law, local labor markets, employer size, and union strategy also affect what becomes possible.
The available research is selective. Visible activism at prominent firms can receive more attention than quiet organization in smaller companies, vendors, public institutions, game studios, or non-U.S. workplaces. Professional workers who speak publicly are not necessarily representative of everyone who produces and maintains software.
Those limits are reasons to investigate one workplace carefully. They are not reasons to return to a cultural stereotype.
What this evidence cannot tell you about your workplace
Industry history cannot determine whether a current group is ready to organize. Readiness depends on present relationships and behavior:
- Do coworkers compare facts with one another?
- Are the issues widely and deeply felt?
- Do people follow through on small commitments?
- Does the group cross teams, locations, and employment categories?
- Can workers disagree without the network collapsing?
- Have people identified the decisions they want to change?
A company can sit inside an industry with low union density and still contain a well-connected group prepared to act. A highly publicized wave of tech activism can also coexist with a workplace that has not built those relationships.
Use Workplace Mapping and Organizing Conversations to examine the actual shop rather than treating industry identity as a substitute for evidence.
The practical conclusion
Software workers have been slow to unionize because several institutions made individual adaptation easier to imagine and collective organization harder to sustain. Professional identity, mobility, employment fragmentation, and employer control reinforced one another. None of them eliminated common workplace interests.
The useful lesson is conditional: do not assume technical workers will organize because the industry has become harsher, and do not assume they cannot organize because earlier generations rarely did. Look for the relationships, shared issues, and small acts of follow-through that turn a workforce category into an organized group.
Sources
- JS Tan, Natalia Luka, and Emily Mazo, Unlikely Organizers: The Rise of Tech Worker Labor Activism, first published online October 9, 2025, ILR Review: article page
- Robert Dorschel, A new middle-class fraction with a distinct subjectivity: Tech workers and the transformation of the entrepreneurial self, first published online July 4, 2022, The Sociological Review: article page