On November 14, 2022,  teaching assistants, student-researchers, tutors, and readers at the University of California launched the largest strike in the history of higher education in the United States (1). In the popular and labor press, the strike has been noted for many things: its size, its scale, its duration, its achievements and its shortcomings. Much less discussed, however, is that it marked one of the first-ever strikes of research labor. As we discovered at UC Santa Cruz and across the state, mounting and sustaining an effective research strike presents unique challenges rooted in the labor process of research work itself.  As grad unionization efforts and contract campaigns ratchet up at lab-heavy institutions like MIT, Cal-Tech, Dartmouth, and Pardee RAND Graduate School, it is critical that worker-organizers extract lessons from the UC experience. In this article, I hope to provide a preliminary report on what we have learned. 

In December 2021, the University of California voluntarily recognized Student Researchers United-United Autoworkers (SRU-UAW), a newly formed union of over sixteen thousand graduate student-employees. These workers, many of whom belonged to STEM fields and worked in laboratories, were previously ineligible for collective bargaining rights, because they were legally considered students rather than employees. However, this changed in 2018 with the passage of CA SB 201. Shortly thereafter, organizers with UAW launched a campaign to form SRU. By May 2021, a supermajority of researchers across 10 campuses had signed cards in favor of unionization, an incredible feat. When the University refused to recognize the entirety of the bargaining unit, researchers responded with a strike authorization vote wherein 97.5% of those who voted cast a ballot in favor. The vote seemingly prompted management to swiftly abandon their contestation and, in a few short months, researchers found themselves at the bargaining table alongside their teaching assistant counterparts. 

The core demands of researchers did not differ greatly from those of other academic student-employees. They wanted to be paid enough to live where they worked, to work in an environment free of harassment and bullying, and to end the exorbitant fees placed on international student-workers. The University was uninterested in meeting these demands and, after many months of failed negotiations, strike action was once again on the agenda. 

For those of us working as teaching assistants, the assignment was pretty straight forward. A strike meant ceasing all activity pertaining to the courses we were hired to teach. This meant no classes, no office hours, no emails, and absolutely no grading. However, for researchers, the question of what labor to strike was more difficult to discern. This is because, for one, workers’ own scholarly research is thoroughly entangled with the work that they perform for their lab’s primary investigator (PI), often a tenured faculty member. Because academic research is iterative and cumulative; setting down one’s instruments and walking out of the lab could mean significant setbacks in your personal academic career and catastrophic loss of your own research, a point that life scientists often illustrated by describing the critical labor they did to keep their mice, worms, or various other microorganisms alive. In contrast, for student-workers on the instruction side, a strike may provide an opportunity to make progress on your dissertation, a confession that was not uncommon to hear on the picket line. 

As workers across the state began discussing the very real possibility of a walkout, researchers in statewide meetings started to raise questions about the viability of a research strike. Such concerns were, for the most part, dismissed as official leaders and staff argued that a failure to strike all research would open the union up to legal action. This was because, they said, workers would be engaging in a “partial strike,” something they claimed was not protected under labor law. At UC Santa Cruz, we disagreed that this legal matter would have any bearing on the efficacy of our action, as union and management lawyers regularly suit one another during the course of a strike only to drop all claims as part of the tentative agreement (2). The terms of such an agreement, we argued, would ultimately be a function of power, not the law. Therefore, our task was to identify where that power was located and how we could fortify and leverage it. 

Our understanding of power differed from that of union officers and staff, who espoused something like an aggregate and uniform theory of power, evidenced by our contract campaign’s slogan “48,000 strong!” The logic underpinning this theory is that each individual worker represents an equivalent amount of power. Therefore, in order to increase the power of a strike action, you needed to increase the raw numbers of workers on strike or threatening to strike. The practical activity that flows from this, we observed, is essentially a voter turn-out operation, where atomized individuals are mobilized by self-identified organizers and staff to participate in an authorization vote or fill up space on a picket line. One problem with this formulation, we argued, is that it neglects the nature of our work and the ways in which leverage is unevenly distributed across our worksite. 

On the instructional side, for example, the disruption of core or prerequisite classes creates greater administrative pressure than electives. This is because internal processes in the University require completion of core or prerequisite courses in order to determine graduation or advancement in major. When students lack a grade for these courses, it creates a crisis that the administration must contend with. Additionally, enrollments in core classes tend to be considerably higher than others, which means that striking them is necessarily more impactful. So, in the summer, our organizing committee set out to map out every course offering in every department, paying special attention to classes that, when struck, would inflict the most pain on our employer. We then worked with department stewards to identify who was TAing for each course and organized on the basis of these collectivities. On the research side, the uneven distribution of leverage was more difficult to chart. Instead of courses, we mapped labs, focusing on those that seemed to generate the greatest amount of money (the UC collects overhead, or indirect costs, on all grants) or prestige for the University.  Where we could, we noted which labs had contracts with industry, as these tended to be on shorter funding cycles and required deliverables on a quarterly or yearly rather than multi-year basis. We then centered our organizing efforts on these strategic sites. 

Given the concerns that researchers had about sabotaging their own research, unionists working in labs convened organizing meetings to identify what parts of the labor process, when struck, would have the biggest impact on our employer, not researchers themselves. For example, researchers might decide that data could be collected, but not entered or submitted. The reasoning behind this, as UCSC’s “Strike Guide for Researchers” stated, is that “data and analysis are the deliverables of much of our research, so data reporting and analysis is one of the most important aspects of a GSR work stoppage.”  The guide continued, “this stoppage will apply pressure to PIs who require reports of collected data for their funding sources.” The value of this approach was not only to avoid self-harm, but to hone in on the harm that researchers could do to the boss. Indeed, it was the latter that assured workers that they could strike to win. 

In a life science laboratory at UC Santa Cruz workers openly debated the question of what labor to strike. Then, workers collectively agreed upon what work to refuse. As one worker put it in a report she drafted for researchers on other campuses looking to replicate our model,  “One of our key goals was to minimize the extent to which people are left to agonize over decisions individually.” Indeed, we observed that this process seemed to produce a sense of shared commitment and collectivity among labmates who had previously worked in relative isolation from one another. 

Through power mapping and department organizing, we learned a lot about research labor. We discovered that power is not uniformly distributed within labs, as jobs are not standardized and different workers are responsible for different parts of the labor process, at different times. This means that during a strike, certain workers’ will possess more leverage than others, exposing them to greater intimidation and pressure from above. This is unlike instructional labor, where teaching assistants and graduate student-instructors all cross the grading deadline together. Work is not synchronized across or within labs. Labs do not share the same deadlines and research timelines usually stretch over months to years rather than days to weeks. Leverage is dispersed and its impacts are not immediately visible. As a result, researchers’ sense of collective power proved more difficult, though not impossible, to establish. 

On our campus, mapping out the leverage points within and across laboratories helped to produce confidence in researchers who might otherwise doubt the efficacy of withholding their labor. Moreover, encouraging researchers to collectively discuss and agree upon what labor to strike, based on their assessment of the labor process, produced feelings of solidarity, collectivity, and power that were able to combat feelings of isolation and powerlessness rooted in the asynchronicity of the labor process of lab work. 

As is true in any workplace struggle, when people feel alone, they are more vulnerable to intimidation. And, in the lab, intimidation was rampant. This too, we think, has to do with the structure of lab work, where researchers directly report to their faculty PIs who act as fundraisers, HR managers, and supervisors. As one researcher from UC Merced said, “These people not only control your salary and are single-handedly able to decide whether the University will credential you or not, but they are also your primary connection and reference to any future in-field employment.” Moreover, he added, “their own careers depend on the work you are doing; walking off the job is a direct threat to them.” The structure of instructional labor is quite different, as the work that you perform for the University is unlikely to be directly supervised by, for example, your dissertation chair. Whether or not instructional labor is struck almost certainly has no bearing on their career. 

We are only beginning to understand the complexities of striking research labor. However, what is clear is that an adequate understanding of the labor process is essential to mounting a strategic labor action and building the kind of collectivity that is necessary to carry it out.

(1) Post-doctoral scholars and Academic Researchers also struck, though they returned to work after 4 weeks.

(2) This is to  say nothing of how impossible it would be to target "partial strikers" with any accuracy. 

Sarah Mason is a PhD candidate in the Sociology department at UC Santa Cruz where she works as a teaching assistant and researcher. She is a head steward of UAW 2865.