From scientific management to data-driven labor organizing
As data is collected about workers, how does it impact them and how do they respond to it?
Contemporary workplaces have increasing capabilities to log, collate, and analyze data about workers to make critical decisions about labor. This is often to the disadvantage of workers, who are monitored and surveilled in great detail. At the same time, union movements in the US have made some critical wins in the past year. For example, successful unionization efforts at major corporations like Amazon, Starbucks, and Google point to a resurgence in the labor movement. The looming role of data alongside the renewed labor movement raises the question: as data is collected about workers, how does it impact them and how do they respond to it? Understanding this is key to imagining what the future of work could (or should) become.
In this interview, we discuss research on the role of technology and data expertise in supporting worker advocacy movements conducted by Dr. Vera Khovanskaya, an NSF/CRA/CCC Computing Innovation Postdoctoral Fellow at the University of California, San Diego.
The following is an edited transcript of our conversation.
Sohyeon: To start, what were the questions drawing you into your work on data and labor unions?
Vera: While in the middle of another project on labor, I got really interested in questions about unions’ engagement with technology. How has the labor movement responded to technological change in the workplace?
Sohyeon: A lot of your work is very historically grounded — when did you make the connection there?
Vera: My research led me to a book called A Trade Union Analysis of Time Study, by William Gomberg, who, at the time, was working as the head of the Management Engineering Department of the International Ladies Garment Workers Union (ILGWU). And it was just shocking to me that a labor union spearheaded a systematic takedown of the “scientific objectivity” claims underlying the foundations of scientific management. This led me to the records of the ILGWU Management Engineering Department at the Kheel Center for Labor-Management Documentation & Archives at Cornell. These archives document the union’s strategy of hiring in-house industrial engineers to do time studies where the union thought a union time study would be useful to counter the management’s time study, as well as their efforts to assert the role of organized labor in academic and professional discourses around scientific management. My dissertation was trying to understand, how did this analysis and data work happen? And what did it do and at what cost?
Sohyeon: When you say at what cost — what is it about the relationship between data, technology, and labor that you find compelling?
Vera: Well, I think data collection is a key part of automation. People who don’t study automation sometimes think that automation is robots or mechanization. But the first step of automation is to collect data about how people do work. First you collect data about each part of the job, how long it takes, the order in which the parts are done, and what dependencies exist between tasks that are being done in sequence. This paves the way for the reorganization of work so that everybody does a smaller part, thereby rendering the worker more replaceable, and the work more amenable to mechanization.
There’s a strong centralizing tendency in the data collection and usually, it’s not good for the worker. On the other hand, you can also use aggregated data to push back against work conditions. So, the data collection is both the fundamental source of labor degradation and a potential wellspring of opportunity for organizing.
Sohyeon: Could you say more about what “labor degradation” means?
Vera: Labor degradation occurs when the task of planning how the work ought to be done is reassigned from the workers doing the work to another part of the organization, such as management. Data collection allows management to take over the planning function.
Sohyeon: One thing I really appreciate about your work is how the same data processes can be flipped to empower users, much like how the ILGWU conducted their own data work to resist scientific management. How does that flip work?
Vera: Yes, okay: we have to be really careful about the stakes of using worker data collection as part of worker empowerment. Let’s say I am part of a worker advocacy organization and we start collecting workers’ data and our goal is to dispute wage mechanisms. Now let’s say the employer says “we’d love to see your data and see if we can work with you” — now your advocacy organization is in an uncomfortable position being invited to manage the workers they represent. That could have tough consequences, especially if workers weren’t expecting their data to be used in that way. To avoid this, the data work must be in concert with a broader organizing strategy, and it must be done strategically.
Sohyeon: But the cases you examine in your work do use data-driven approaches, right?
Vera: Yes. I think they [the workers in those cases] were operating in a moment where labor’s engagement with those methods was strategic [for them]. Further unpacking the costs and benefits of data-driven approaches connects to the other work I’m doing, which is asking how are data tools currently being used within the labor movement? One role for data tools is in the data-driven pushback against workplace automation, but another important role is for tracking and consolidating membership organizing data within unions.
Unions are incredibly complex, data-driven organizations, with membership information moving between the local, regional, and international levels. There’s also power and expertise at work there. And it all plays out in the arena of data tool adoption, maintenance, and enforcement. Bringing these two strands of research together, my work on the role of data collection in the union shops gives me a unique perspective into the effects that data tools have on the organizational relationships within the unions themselves.
Sohyeon: If unions are interesting examples of complex organizations, what are some broader implications of your research you’re thinking about?
Vera: Studying the role of data tools in unions can help us understand the role of data tools in other membership organizations such as nonprofits and political organizations, as well as other advocacy and organizing groups. But the most exciting thing about unions is that because labor consciousness is high, data experts in the unions are often sensitive to the centralizing tendencies of the data tools and look for ways to stave off labor degradation in organizing work. For example, they do this by giving organizers explicit autonomy in their data collection and by offering data trainings so that local organizers can have an agential role in determining data strategy. In short, the ways that data tools already work in the labor movement can inform how we might support more data-driven advocacy strategies in the future.
