The new paradigm for work as epitomized by Uber—low-paid piecework without job security, benefits, or even formal employer-employee relationships—extends beyond Silicon Valley’s gig economy. Just one example of another industry plagued by precarity is higher education. In the past few decades, universities have increasingly relied on contingent professors called “adjuncts” who, like Uber drivers, are low-paid, hired by the gig (i.e., a course), have little job security, and are deemed part-time despite the number of hours put in and the many who teach as their primary source of income. Their “flexibility” allows university administrations to hire or lay off adjuncts to accommodate fluctuating student populations and budgets, just as Uber provides work only when there is customer demand.
Poverty and precarity are not, however, the only working conditions that adjuncts and Uber drivers share. The more fundamental commonality is their atomization. Because an Uber driver’s workday is spent entirely in their own car, they are isolated from their coworkers, in the same way that adjuncts are separated into classrooms.
There are few social ties through which workers can commiserate and develop trust and solidarity. The lack of social connections are a major barrier to bringing people into campaigns and lasting organizations. Atomization at work is an obstacle to uniting with coworkers, building collective power, and fighting for better working conditions. It shifts the balance of power between bosses and workers toward the former and thereby perpetuates the symptoms of precarity and exploitation. If the point of production is to remain a key arena in leftist politics, then we need to understand the inner workings of atomization, especially given its accelerated spread by pandemic-induced experiments in remote work.
Marx’s analysis of capitalists’ tendency to bring workers together into large, complex systems of production, exemplified by the factories of the Nineteenth and Twentieth centuries, contextualizes the utility of atomization today. This analysis is supplemented by his dissection of Nineteenth-century production into a few categories, which shows how different forms of production require different amounts of social coordination and thereby provide or preclude foundations for resistance. Applying and extending this theoretical framework to production today—both the production of profits, e.g., in the gig economy, and the production of society, e.g., in higher education—can shed light on the ways in which atomization functions as one of capitalists’ best weapons in the class struggle.
The utility of atomization lies in what Marx and Engels called a “fundamental contradiction” in capitalist production: commodities are collectively produced by a company’s workforce but profits are privately appropriated by the capitalist.1Freidrich Engels, “Socialism: Utopian and Scientific,” The Marx-Engels Reader, edited by Robert C. Tucker, Norton, 1978, 176. Irony and injustice aside, the contradiction is a political one: to increase productivity and therefore profit, capitalists bring more and more workers together into production processes that increasingly rely on cooperation among workers. In so doing, however, capitalists provide the social and spatial foundations upon which workers unite and resist. Historically, this contradiction culminated in the factory, which became the hotbed of working-class resistance in the Twentieth century.
Although neoliberal governance crushed workers’ collective power and fostered a culture of individualism in its place, capitalists still face the problem of sidestepping that “fundamental contradiction” not in society but inside production itself. This is because production processes are often themselves social.2Note that throughout this article, I use the word “social” in the sense of socializing and social relationships, not in the sense of societal (as in social reproduction or social democracy) or collective (as in social production, i.e., production by multiple people) or socializing companies (nationalizing them).
Think how much verbal communication it requires to make a McDonald’s Happy Meal or a TV show. Socializing at work and for the purpose of work is the medium through which trust, solidarity, and unity is transmitted: it is the basis of social relationships between workers. This sociality at work is therefore the foundation of workers’ unity, which is in turn the foundation of their power. But capitalists can interrupt this chain, breaking the link between the sociality of production and the development of the unity necessary to build power. They do so by threading the needle between economically integrating the labor of workers without socially integrating the workers themselves.
As Mario Tronti wrote in his continuation of Marx’s analysis of the factory, “only from within labour”—i.e., through the concrete organization of production—“can capital disintegrate the collective worker and then integrate the isolated worker.”3Mario Tronti, Workers and Capital, Verso, 2019, 29. If capitalists can disintegrate the collective worker in the moment of production, they can disintegrate the collective worker in the moment of remuneration, achieving what Tronti called “the ideal for the most modern capitalism, … that of recuperating the primitive relation of simple purchase-and-sale contracted between the individual capitalist and the isolated worker.”4Ibid., 30. In other words, Tronti’s analysis of capital’s ideal points directly to the gig economy we’ve become increasingly subjected to. The atomization that workers experience in the gig economy and beyond is the consequence of asociality that capitalists built into the organization of production.
Three Categories of Production
To better understand the varying degrees of sociality in different industries, we can use Marx’s categorization of production processes according to the extent of the division of labor. The three categories he developed—which he called simple cooperation, manufacture, and large-scale industry—were successive stages of development only in some industries. In general, each category simultaneously coexisted in the economy alongside the others and still does today. There is no strict historical or logical requirement that one form be more advanced than another, or that all industries converge on a most “advanced” form. Each category encapsulates a qualitatively different material organization of production, i.e., a different form of cooperation among coworkers.
The qualitatively different forms of cooperation permit different quantities of socializing among coworkers, with consequences for workers’ collective power. Furthermore, each form permits different degrees of spatial concentration, which is a precondition for socializing in production. In reviewing these categories, we should think creatively about how they can be applied and extended to production today, and how they can explain obstacles or reveal openings to organizing our workplaces.
A production process in which workers do the same type of work with little division of labor can be thought of as “simple” cooperation.5Karl Marx, Capital, vol. 1, Penguin, 1990, 445, 454, 465. Cooperation without the division of labor can improve efficiency, like evoking competitiveness among coworkers doing the same task or saving money by sharing the means of production, e.g., weavers doing the same work side by side in a single workshop. When workers just work side by side on separate but identical tasks, rather than coordinate production of a given commodity or assist one another, production requires minimal social coordination among workers, which undercuts their unity and thus power.6Ibid., 442.
Simple cooperation can be completely asocial if workers are not in proximity. For example, in the original cottage industry, peasants made textile products at home when they could find time away from the fields. Although multiple peasants cooperated to produce cloth for the same capitalist, their spatial dispersal across multiple cottages was an obstacle to their collective power.7Ibid., 591. Their work in the fields, planting or harvesting crops in parallel, was just as asocial and disempowering; “the dispersal of the rural workers over large areas breaks their power of resistance.”8Ibid., 638. Without spatial and social integration, the peasantry of Nineteenth-century Europe lacked association—“much as potatoes in a sack form a sack of potatoes,” nothing more than the sum of its parts—and therefore lacked the foundation for resistance.9Karl Marx, “The Eighteenth Brumaire of Louis Bonaparte,” The Marx-Engels Reader, edited by Robert C. Tucker, Norton, 1978, 608.
The second form of cooperation, which Marx called manufacture (not to be confused with how we use the word today), is based on the division of labor.10Marx, Capital, vol. 1, 485. In manufacture, the division of labor is limited by the handicraft skills of various artisans, skills that science has not been able to incorporate into machines.11Ibid., 458. What was innovative about manufacture was that it integrated artisans into a single production process—often concomitant with spatially concentrating them in a single workshop—so that their handicrafts can be specialized for the final product, e.g., wheelwrights specializing in making wheels for carriages.12Ibid., 455-456.
The specialization of each worker in a part of the production process can give rise to a mutual interdependence among them and therefore a necessary amount of socialization, especially when they all coordinate on producing each commodity. For instance, in wire production, one worker draws the wire, another straightens it, a third cuts it, etc.13Ibid., 464. But manufacture is less social if the final product is just an amalgam of various parts, each made by a specialized worker in isolation. In this case, cooperation is so asocial that the handicrafts could be dispersed into separate workshops, as Marx observed of watchmaking in Switzerland.14Ibid., 461-462. This precludes any socializing in production, interrupting the chain that could otherwise be used to leverage sociality into unity and from such unity, workers’ power.
Marx’s third category, large-scale industry, covers production processes based on complex machinery, which enabled capitalists to maximize the division of labor. By dissecting the production process into minute parts that can be implemented by machines, handicraft skill is transferred from workers to machines, and the former are left to mind or assist the latter.15Ibid., 545. In contrast with manufacture, workers are now “deskilled” and replaceable, but their work is even more social. They are engaged in a sophisticated choreography synchronized with the machines and each other, requiring face-to-face, verbal, social coordination to keep commodities flowing through the machinery. Sociality is inherent in large-scale industry: “the co-operative character of the labour process is in this case a technical necessity dictated by the very nature of the instrument of labour.”16Ibid., 508. Spatially concentrating workers is necessary, too, because the means of production must be located in a single place so that machinery can be jointly operated and production can flow seamlessly from one step to the next. Large-scale industry inevitably brings workers together.
Marx and Engels were optimistic that the material organization of production within advanced industrial factories would be the social foundation upon which workers would unite, fight back, and ultimately revolt:
The advance of industry, whose involuntary but willing promoter is the bourgeoisie, replaces the isolation of the workers, due to competition, with their revolutionary combination, due to association. The development of large-scale industry, therefore, cuts from under its feet the very foundation on which the bourgeoisie produces and appropriates products for itself. What the bourgeoisie, therefore, produces, above all, are its own gravediggers.17Karl Marx and Friedrich Engels, “Manifesto of the Communist Party,” The Marx-Engels Reader, edited by Robert C. Tucker, Norton, 1978. 483.
Although those graves were not dug in the West, the Congress of Industrial Organizations (CIO) took advantage of the social associations among factory workers intrinsic to industrial production to develop and successfully apply a sophisticated organizing methodology.
Promoted today by Jane McAlevey, this methodology prioritizes social relationships among workers and leverages informal social groups in a workplace by identifying “organic leaders” who have clout in those groups, persuading them to join the struggle, and winning the active support of their followers.18Jane McAlevey, No Shortcuts: Organizing for Power in the New Gilded Age, Oxford, 2016, 30–34. The methodology was successful in organizing factories in the 1930s and it continues to be successful today in workplaces that can be categorized as large-scale industry but are not factories, namely, logistics warehouses.
Amazon’s distribution centers require socially and spatially concentrated cooperation to efficiently take in, pack, and ship out products. This has given organizers the social foundations for unionization campaigns in the US and abroad, including what is now the most celebrated labor victory in decades: the unionization of the JFK8 fulfillment center in Staten Island by the independent Amazon Labor Union. Despite Amazon’s best attempts to reduce the time in which workers can chat with each other by speeding up and monitoring work and by spatially separating workers under the pretext of Covid-19 social distancing, labor in large-scale industry is an inherently social process. After worker-organizers in Staten Island learned the methodology by reading communist CIO organizer William Z. Foster, they successfully leveraged the social relationships forged by production itself as well as sites of socialization outside the workplace, such as the sidewalk, bus stops, and bus rides.19Justine Medina, “How We Did It,” Labor Notes, April 1, 2022; Luis Feliz Leon, “Amazon Workers on Staten Island Clinch a Historic Victory,” Labor Notes, April 1, 2022, https://labornotes.org/2022/04/amazon-workers-staten-island-clinch-historic-victory.
While comparative analysis between the successful unionization of the JFK8 warehouse and the more difficult campaign being undertaken by the Retail, Wholesale, and Department Store Union at an Amazon warehouse in Bessemer, Alabama is complex—and, surely, forthcoming—one can’t avoid observing the latent power of social relationships between workers themselves, in contrast with relationships between workers and union staff organizers. Organizing workplaces where social relationships between workers are sparse or weak is, therefore, a major challenge.
Analyzing Higher Education and the Gig Economy
Applying and extending Marx’s three categories of production today can help us understand the variation in workers’ collective power across different industries and investigate where it’s lacking. We can then formulate concrete responses suited to the kind of labor and setting we’re organizing in. We should not expect that all facets of the organizing methodology developed for Twentieth-century factory workers will necessarily apply to other forms of cooperation today. The categorization also sheds light on where capitalists have had the most success atomizing workers, namely, industries where production is organized by simple cooperation or manufacture, where workers were already weakened by the low degree of sociality and therefore vulnerable to further desocialization, disempowerment, and exploitation.
Higher education is one such industry and it’s an industry I can speak to since it’s where I work and organize as an adjunct at the City University of New York (CUNY). The form of production that professors are engaged in is manufacture: putting research aside, what we produce is a given student’s college education, which is the product of several instructors who divide this labor according to their academic specialties. Because a biology instructor never talks to a history instructor, never mind coordinates with them in charting a single student’s academic progress, our cooperation is primarily asocial, mediated not by communication but by the division of labor. This asociality is compounded by the spatial dispersal of instructors into separate classrooms, silos of production that represent an advancement beyond concentrated workshops in Nineteenth-century manufacture. What little social cooperation remains occurs in rare intra-departmental meetings.
When state governments began defunding higher education in the neoliberal turn of the late 1970s, university administrations coped by taking advantage of the low degree of sociality in production to rely more and more on low-paid, precarious adjuncts. Without social production, faculty lacked the unity and power to resist austerity-driven “adjunctification”—i.e., the recuperation of individual contracts, as Tronti put it—and even further desocialization of production.
For example, at CUNY, most departments don’t invite adjuncts to faculty meetings and exclude them from cooperating in the pedagogical work of setting curriculums. Furthermore, shared adjunct offices are crowded, ill-equipped, and have only a handful of desks for dozens of adjuncts. Consequently, most prefer to do any non-teaching work required for the job while at home, exacerbating the desocialization of their labor. Since they don’t stay on campus before or after class, they lose opportunities to socialize, cooperate, or bump into each other in the halls. Most adjuncts don’t even know the coworkers in their own departments.
The atomization of a faculty into isolated instructors proves to be a significant barrier when we try to organize our coworkers for contract campaigns or fights to fix common issues. The most basic obstacle is that, due to such desocialization, they are difficult to reach. Adjuncts don’t check their mailboxes, canvassing adjunct offices yields few conversations, and targeting specific instructors requires waiting outside their classes.
The CIO organizing methodology of identifying and persuading “organic leaders” can work well in spatially concentrated workplaces with social cooperation, like factories, distribution centers, and hospitals, but since adjuncts don’t socialize with each other or cooperate face-to-face in meetings, there is little to substantiate informal social networks from which organic leaders would emerge. This contrasts sharply with K–12 education, where the methodology has been successful. Because teachers cooperate socially within and between departments and are spatially concentrated for the school day, production in K–12 provides the social infrastructure for this style of organizing. The version of manufacture that K–12 teachers are engaged in has not been desocialized to the extent experienced by higher education workers.
Recently, advances in online teaching have further desocialized labor in higher education by dissolving the bare minimum of personal contact that was previously possible. This process, accelerated by the response to Covid-19, has had the effect of making higher education workers into guinea pigs for the atomization of K–12. Drooling over the efficiencies it could bring, university administrators used the pandemic to undertake a massive experiment in remote instruction, learning which courses could be moved online and how students and faculty would fare with minimal institutional support.
While health and disabilities should no doubt be humanely and carefully addressed by remote work, the aim of these experiments was instead to reduce costs. Perhaps most importantly, they learned that courses can still run if instructors live in different states or even different countries, opening up the possibility of offshoring higher education by globalizing the adjunct model.
Zoom did not, however, yield comparable gains in our ability to organize. Direct actions to confront the boss became impossible, and Zoom meetings were less effective at building relationships and campaigns than in-person conversations or even ordinary phone calls. Without socializing for the purposes of production, there were no pre-existing social networks to tap into, no way to reach coworkers who wouldn’t pick up the phone or attend a Zoom meeting, and no way to rope in the majority of coworkers who don’t self-select as organizers or activists. The issue with online teaching, then, is not only the intensification of labor or the disconnect with our students, but also the obstacles it brings to organizing our coworkers and building power.
Higher education exemplifies how bosses can take advantage of the low degree of sociality in manufacture to further desocialize production and increase their own power over workers. It is, however, far from the only such example. The gig economy provides a variety of systems of production following simple cooperation that are ripe for the same developments.
Uber, for one, represents an advancement upon older styles of simple cooperation. Traditional taxi drivers are engaged in simple cooperation, doing the same type of work simultaneously, in parallel, and for most of their shift, spatially dispersed in separate cabs. Socializing is therefore limited to chatting with coworkers when picking up and dropping off cabs at the depot and occasional cooperation among drivers, cleaners, mechanics, radio dispatchers, etc. The industry was open to further desocialization.
Uber bypassed the sociality intrinsic to the material organization of the taxi industry by reorganizing production so that workers drive and maintain their own cars. The depot as a site of social connection was thereby eliminated. Furthermore, the internet enabled Uber to use technology—servers that calculate routes and algorithmically dispatch them—to deskill workers without spatially concentrating them, as is necessary for the machinery of large-scale industry. This combination of deskilling and desocializing labor inhibited workers’ ability to unite and resist. These innovations therefore perpetuate Uber’s recuperation of individual contracts.
In contrast, one of the few sectors of the gig economy in which workers have been able to organize strikes—food delivery— happened to provide couriers with the material conditions to socialize and unite. Deliveroo couriers in London and Foodora couriers in Turin were expected to wait for orders at central meeting points. These spatial concentrations allowed workers to meet one another, chat, exchange WhatsApp info, and agitate.20Arianna Tassinari and Vincenzo Maccarrone, “Striking the Startups,” Jacobin Magazine, www.jacobinmag.com/2017/01/foodora-strike-turin-gig-economy-startups-uber/;
Facility Waters and Jamie Woodcock, “Far From Seamless: a Workers’ Inquiry at Deliveroo,” Viewpoint Magazine, viewpointmag.com/2017/09/20/far-seamless-workers-inquiry-deliveroo/.
Fighting Back When Production is Desocialized
If we care about organizing the working class at the workplace, we must confront the menagerie of different processes of production that exist today, including not only modern instances of large-scale industry—where the sophisticated organizing methodology used by the CIO can succeed—but also desocialized workplaces where that methodology is limited. So a preliminary step before organizing one’s workplace is to determine which form of cooperation is employed—simple cooperation, manufacture, large-scale industry, or something else. One also needs to investigate the concrete material organization of production to understand where social cooperation and socialization more generally happen, if at all. If production is social and if workers are spatially concentrated, then traditional forms of organizing might suffice. Otherwise, we need to experiment with new organizing methods and adaptations of traditional methods suited to desocialized and dispersed industries.
One option, which has successfully been taken by some workers’ centers that organize domestic laborers (spatially dispersed simple cooperation), is to use ethnic, immigrant, or religious communities as alternative social foundations for building unity and power. Another possible option—especially when there is no such external community among workers to tap into, as is the case in higher education—is to fight for the resocialization of production. The struggle would be for a material reorganization of production that brings back social cooperation. Tronti would agree: “Our task,” he wrote, “is to continually recompose the material figure of the collective worker against capital, which itself seeks to dismantle this figure.”21Tronti, Workers and Capital, 30. Since it is only by dictating the material organization of production that capitalists can desocialize cooperation, fighting on that terrain—the organization of production—can recover the sociality of our labor and thereby reestablish the foundation necessary for uniting with our coworkers and building power. Ironically, this is the one of the few terrains in which the ossified official labor movement still fights, i.e., the struggle to stipulate working conditions in a written contract.
How a struggle for resocialization would be concretely implemented depends on the industry. CUNY, however, may prove a fertile test case for approaches that can be adapted to other industries or at least serve as food for thought. For example, our last contract gave adjuncts a paid office hour for every class, which required us to spend more time in adjunct offices and consequently bump into each other more often. To build upon this incidental resocialization, we could demand social cooperation, too, such as co-teaching opportunities and the inclusion of adjuncts in department meetings with pay. Social, face-to-face, verbal cooperation can foster the bonds necessary to reverse our atomization and eventually build solidarity and power with our coworkers.
We can also hijack capital’s impetus to make production more efficient and turn it to our own advantage. If universities merge several small classes into a single large lecture, we can demand more teaching assistants and regular, paid, collaborative meetings of all the staff involved in the class. If classes are too big to evaluate each students’ progress through traditional means, we can respond by creating a single merged discussion board for every section of the same course as a virtual form of spatial concentration and a medium for our cooperation.
In other industries where production has been desocialized, organizers could consider concrete ways to make work more collaborative—what tasks can be done in pairs? how can meetings incorporate more collaboration? how can safety procedures or quality reviews be done by multiple workers together? Where work is very atomized, perhaps socialization outside of work time could also be leveraged or even increased. For example, break time and break rooms can be utilized or, if absent, demanded. Shift changes and the subsequent convergence of workers at points of public transit can also be exploited, as Amazon Labor Union did. Admittedly, demanding mere technical adjustments to work itself can be uninspiring unless they are incorporated into bigger campaigns, but without recovering the sociality of production, we cannot reconstitute the social foundation for solidarity and power.
Simultaneously, we need to resist further dispersal. For example, working from home has disempowered teachers and many other white-collar workers. We must think creatively about how to maintain the safety and accommodations for disabilities afforded by working from home without letting remote work become a trojan horse of desocialization in the long-term.
To fight both against dispersal and for resocialization, especially in this moment of historic weakness for the left, we must rely on the inalienable strength of our position in the economy: that we alone produce commodities, profits, and society. Our necessity in production and our ability to stop it is the starting point for rebuilding power. From there we can restore the sociality of our labor and hopefully reestablish the foundation for unity and power in the workplace. It is a small but necessary step toward bigger fights and bigger victories.
 Friedrich Engels, “Socialism: Utopian and Scientific,” The Marx-Engels Reader, edited by Robert C. Tucker, Norton, 1978, 176.
 Note that throughout this article, I use the word “social” in the sense of socializing and social relationships, not in the sense of societal (as in social reproduction or social democracy) or collective (as in social production, i.e., production by multiple people) or socializing companies (nationalizing them).
 Mario Tronti, Workers and Capital, Verso, 2019, 29.
 Ibid., 30.
 Karl Marx, Capital, vol. 1, Penguin, 1990, 445, 454, 465.
 Ibid., 442.
 Ibid., 591.
 Ibid., 638.
 Karl Marx, “The Eighteenth Brumaire of Louis Bonaparte,” The Marx-Engels Reader, edited by Robert C. Tucker, Norton, 1978, 608.
 Marx, Capital, vol. 1, 485.
 Ibid., 458.
 Ibid., 455-456.
 Ibid., 464.
 Ibid., 461-462.
 Ibid., 545.
 Ibid., 508.
 Karl Marx and Friedrich Engels, “Manifesto of the Communist Party,” The Marx-Engels Reader, edited by Robert C. Tucker, Norton, 1978. 483.
 Jane McAlevey, No Shortcuts: Organizing for Power in the New Gilded Age, Oxford, 2016, 30–34.
 Justine Medina, “How We Did It,” Labor Notes, April 1, 2022; Luis Feliz Leon, “Amazon Workers on Staten Island Clinch a Historic Victory,” Labor Notes, April 1, 2022, https://labornotes.org/2022/04/amazon-workers-staten-island-clinch-historic-victory.
 Arianna Tassinari and Vincenzo Maccarrone, “Striking the Startups,” Jacobin Magazine, www.jacobinmag.com/2017/01/foodora-strike-turin-gig-economy-startups-uber/;
Facility Waters and Jamie Woodcock, “Far From Seamless: a Workers’ Inquiry at Deliveroo,” Viewpoint Magazine, viewpointmag.com/2017/09/20/far-seamless-workers-inquiry-deliveroo/.
 Tronti, Workers and Capital, 30.