Barman and Waitress (1920–1925) | Unknown / Courtauld Gallery, London (Samuel Courtauld Trust) / Image courtesy of the Courtauld
Eighty percent of the US workforce now does service work, yet the protagonist that still anchors most histories of capitalism is the unionized, hourly, goods-producing manufacturing worker. In Beneath the Wage: Tips, Tasks, and Gigs in the Age of Service Work (Zone Books/Princeton University Press, 2026), Annie McClanahan argues that this framing is not just incomplete but distorts our understanding of the conditions of waged labor. McClanahan traces an alternative history that runs from eighteenth-century domestic servants through Pullman porters to today’s rideshare drivers and Amazon Mechanical Turk clickworkers.
The book is organized around three paradigmatic forms of contemporary service labor: superexploited in-person tipwork, deskilled clerical microwork, and informalized circulation gigwork. In each case, McClanahan attends to the material conditions of service labor—how workers are paid, how they are controlled, and how technology mediates their work—but also to the aesthetic forms through which workers represent and make sense of those conditions. In doing so, she sets historical working-class genres like the workers inquiry alongside the novels and reality TV that gig workers have produced about their own twenty-first-century experience. The result is a study that moves nimbly from Emoji Dick to Raven Leilani and Priya Guns to Marx, Braverman, and Claudia Jones.
McClanahan is Associate Professor of English at the University of California, Irvine, Co-Director of the UC Materialist Institute for Research, and the author of Dead Pledges: Debt, Crisis, and 21st Century Culture (Stanford University Press, 2016). Recently, she sat down with Lina Moe to discuss Beneath the Wage and what happens when we center service work within the story of labor under capitalism.
You’re an English professor writing economic history and labor process theory. What does that position let you do that a labor economist or a historian can’t?
I think of the book as a mix of historical work and cultural studies, and honestly, historians can do a lot of what I do. But the cultural piece often gets left out. In the context of this book, that means attending to documentation of workers’ actual experiences—novels written by former gig workers, workers’ inquiries written by current ones trying to describe the conditions of their labor. I’m looking at those not just as empirical documents, though they do give us a pretty profound sense of the labor process, but also as cultural objects. The aesthetic modes in which workers describe their labor tell us something about their experience of the work that might fall out of a strictly labor historian’s account.
One of the book’s central arguments is that service work is the elephant in the room in the history of capitalism. Since you are bringing the attention to form and genre of a literary scholar, how does changing the main character also change the whole story of capitalism? What happens to the plot, the structure, or the arc of capitalism when service work moves to the center?
Oh, I love that framing. A few things shift. First, you go back further. If the core protagonist is the industrial worker, you mostly begin in the nineteenth century. But if you’re centering service work, especially domestic servitude in the pre- or proto-capitalist eighteenth century, you start looking for the moment when people first begin depending on wages for subsistence, and that takes you to domestic service first. Second, the gender of your protagonist changes. You’re no longer looking at a predominantly male portion of wage labor, but at a part of the economy that has always included women and that becomes increasingly feminized over the twentieth century.
And then there’s the larger structural point. So many of what we take to be the defining features of wage labor in the developed world—paid hourly, highly regulated, scientifically managed, formalized under law—all of those assumptions have to go out the window when you’re talking about service work, which has historically been paid non-hourly wages, managed indirectly or by the method of payment itself, unregulated, highly informalized. Once you let go of those assumptions, you get a much better picture of labor in the underdeveloped world today, where economies are basically skipping from agriculture straight to services without an intervening industrial phase. And it’s the industrial phase that almost inevitably brings certain kinds of regulation with it—it’s the assembly line that makes the hourly wage possible, and the hourly wage that makes possible the entire regulatory apparatus we tend to take for granted. Without that phase, you never build that system. Which is why you see, in the underdeveloped world, what Jan Breman evocatively calls “wage hunting and gathering”—piecework, gig work, informal work of all kinds.
You tell the (to me) surprising history of what was called “master-servant law.” This was a legal framework built to regulate domestic workers that ended up being used to exclude them from labor protections. Can you walk us through that, and connect it to the irony you identify: that workers have to prove they’re controlled in order to receive protections?
There were really two stages of discovery in researching that chapter. The first was realizing that our entire image of what constitutes an employee, and what constitutes management, basically derives from master-servant common law—essentially the first version of labor law invented under nascent capitalism. That’s where we get the word “servant” used as a legal equivalent for “employee.” The master’s total control over the servant is the origin of our understanding of management itself. But then, very quickly, with industrialization, those original categories of workers get excluded from the framework, even though all the terms derive from them. And that has a lot to do with gender and, in the United States, with race—this work is increasingly done by women of color, in the context of battles between Northern and Southern legislators over how to regulate labor under two very different economic systems.
What you then see is certain types of work getting carved out, particularly work where the worker is in some form of motion—which is why I focus on what I call “circulation gig work,” whose regulatory roots go back to things like truck driving. In all these cases, control gets outsourced to technology. The independent contractor is “independent” because no one is visibly watching them work—you don’t have a master surveilling the household. Instead, you have an app. That gives the company an alibi for claiming independent contractor status even while exerting a massive level of technological control—more supervision, actually, than many workers who are classified as employees.
One of the real divides in gig worker organizing is whether to fight misclassification or pursue gains within independent contractor status. In your reading of Priya Guns’ Your Driver Is Waiting, one gig worker character says: “We don’t just want fair wages, they’ll put us on contracts with a low hourly rate and still make money off us.” Is the novel articulating a critique that can help us understand current policy fights?
It’s really complicated. My sense is that there’s a divide right now between East and West Coast labor organizing. On the West Coast, unions really wanted to use the California ballot initiative process to fight misclassification, but there wasn’t adequate gig worker buy-in for that as the main goal. Some of that lesson was learned on the East Coast, where organizers are saying maybe we need to be fighting on different terrain, because workers themselves don’t necessarily see employee classification as what they want. And I think if you have a workforce that includes a lot of migrant workers and people without full citizenship or work permits, the classification issue becomes very tricky. So I tend to think you have to follow the lead of what workers themselves want, and it seems to me that collective bargaining rights is the direction to go.
You suggest that when we tip a server or rate a driver on an app, we’re not just consumers, we’re taking up some of the work of management. Does that make us all complicit in the exploitation of service workers?
It’s a descriptive rather than a normative claim, in the first instance. Regardless of where any of us sits in relation to gig work, we are all essentially serving as employers of service workers all the time. And there’s something about our implicatedness in those relationships that is distinct from our relationship to industrial workers, precisely because of how service wages work. When we tip, or rate a driver on an app, we’re doing the work of management and that puts us in a much more vexing, intimate relationship with service workers than the distant, impersonal relationship we might have to the people who make our shirts or process our food.
But I don’t think the answer is simply to say we’re all complicit and leave it there. The visibility of those relationships is also the basis for possible forms of solidarity across the sector. A group like One Fair Wage is doing exactly this—organizing around a method of wage payment rather than a job category, which brings together valet parkers and sommeliers at high-end restaurants, the whole gamut. Thinking about service workers as a class that might have solidarity precisely because of the intimacy involved in the work—that seems like a really interesting and underexplored possibility.
And it connects to something I think about in my own sector. In higher ed, the absolute urgency is wall-to-wall organizing. Tenure-track faculty and the people doing landscaping on campus need to be in forms of institutional solidarity that can be realized at the level of collective bargaining, ideally in one union, minimally in unions working closely together. There’s no other way to do it. I’m speculating that’s true across the service sector.
Your chapter on contemporary conceptual poetry and its intersections with service work was extraordinary. I couldn’t believe projects like Emoji Dick or Of the Subcontract had made money off other people’s labor, and that tenured academics wrote about the work in ways that separated the aesthetic object from the labor practice.
It’s baffling to me. Take Emoji Dick—it’s a translation of Moby Dick into emoji, and it became this celebrated conceptual art object. But it was produced by parceling the entire novel out to Amazon Mechanical Turk workers, who were paid fractions of a cent to translate each sentence. The art world talked about it as a provocation about language and meaning; the labor that made it was basically invisible. I just can’t imagine treating those as separable questions.
You present a nuanced treatment of automation. I sense that the debate in humanities and social sciences about how to theorize automation is split between those saying AI will destroy jobs, this is an emergency, and the more debunking tendency to say, Well no, AI doesn’t exist per se without the human labor behind it. Where do you locate your argument?
The way you’ve framed it is exactly right. The ghost work argument—that AI is not artificial intelligence but, as some are now saying, African intelligence, because most of the workers doing the tagging and checking are located in the underdeveloped world—is totally true, and we have to be talking about that. But AI is also a real thing that has really happened, and it will increasingly need less and less human intervention. The demand for Mechanical Turk labor has already decreased significantly because of AI’s growing computational power.
And I think the debunking work is really valuable for making a broader point: The levels of productivity growth we saw during the massive phase of industrial automation in the twentieth century are over. That’s what explains incredibly slow GDP growth in the US over recent decades, and the long downturn in the rate of profit more broadly. But it doesn’t mean employers across every sector—including sectors that once seemed completely unautomatable—aren’t trying to use technology to eke out smaller, more marginal gains, and ultimately to decrease labor costs. Technology isn’t just being used to replace labor. It’s also being used to deskill it, outsource it, reduce wages, intensify it—in the case of gig work, to force workers to work as fast and hard as possible under constant surveillance.
That’s where the term “so-so automation” is really useful. And anybody who’s interacted with a customer service bot in the last couple of years knows exactly what it means—automation is replacing a position that was already low-waged, and though it’s doing a mediocre job, the bot is saving just enough labor to be worth it to the employer.
“Workers’ inquiries” is a genre of writing documenting the day-to-day conditions of labor from the inside. They were devised by radical thinkers like C. L. R. James, and you trace how the form has been revived by scholar-activists who embed with delivery workers or call center employees to help them tell their stories. What did you want to add to that tradition?
In the last chapter some of the scholar-organizer dimension really comes in, because one of the things I learned from writing the book and engaging with these materials is how important it is to understand your own labor process as a precondition for being able to organize. I’ve been doing a lot of faculty organizing in the UC system over the last two years, and really trying to learn about other people’s experience of their jobs. People in STEM fields, for instance, have a very different experience of the labor process, a very different vocational self-conception, a very different relationship to the institution than faculty in the humanities and social sciences. Workers’ inquiries have been a really valuable way to do that preliminary organizing work in my own sector, and that’s something I’ve learned from looking at how the genre is being used in other parts of the service sector, particularly in gig work.
The book ends with a personal form of workers’ inquiry into your own industry: a coda on EdTech and the automation of teaching. Is there anything in higher ed organizing that’s making you optimistic? And what feeling did you want to end the book with?
We’re in a period of immense generational change in higher ed labor. For the first time, a majority of faculty are coming to their tenure-track positions having already been unionized as grad students. These are people who have accepted that unions are useful and that it’s okay to understand yourself not as a priest in that classic sense of pure vocation, but as a worker in a potentially antagonistic relationship with your institution. That generational shift is huge. And what’s happening at the federal level with higher ed funding is making people in STEM fields, which have historically been immune to austerity, suddenly aware of the conditions that humanities and social sciences faculty have been facing for a long time.
The obstacles are real too. The “it can’t happen to me” attitude is everywhere, and it’s shocking to me. We are not immune to the forces of history, but tenure-line faculty have a really hard time understanding themselves as being in even a potentially antagonistic relationship with their institution. We’re so vocationally identified with it and at least historically have had a role in running these institutions, so sliding between roles that look like management and positions that look like classic employment creates confusion about who the boss actually is that makes organizing more challenging.
As for the feeling I wanted to end with—I want people to understand this as a crisis moment, a tipping point. Not in the sense of inevitability, but as in a last moment of opportunity. I think unionization, and ideally wall-to-wall unionization that creates intimate solidarity with other unions on campus, is the only thing standing between us and our obsolescence. And I think that’s truer now than it was even two years ago. The horizon is vanishingly short. If we care about these institutions, if we think it matters that we can teach students in an actual physical classroom rather than as a video on a screen, we have to do something now.
