Service Workers Aren't Servants: A Response to Jodi Dean

Ben Burgis

May 26, 2025

One of the great themes of Marx’s work is that capitalism isn’t natural or eternal. Socialists have always hoped that, when it finally shuffles off the historical stage, it will be replaced with something better. But they’ve also worried that it will be replaced with something worse. Marx and Engels wrote in 1848 about class struggles sometimes ending not with the triumph of a class that reorganizes society in its own image but with the “common ruin of the contending classes.” Karl Kautsky and (later) Rosa Luxemburg expressed a similar thought by saying that humanity faced a choice of “socialism or barbarism.” The phrase became a staple of twentieth-century radical rhetoric, in part because “barbarism” is usefully vague. While Kautsky seems to have intended “barbarism” to mean the apocalyptic collapse of industrial civilization, the word can mean whatever horrors seem to be on the horizon at any given time. In the 1930s, for example, many socialists suggested that the only alternative to a global proletarian revolution would be a global descent into fascism.

A far less well-remembered version of Marxist “something worse” predictions came from a far less well-remembered Marxist, the early American socialist leader Daniel De Leon. In 1912, he warned that if the workers’ movement doesn’t move forward and achieve socialism, society could “[roll] backward toward feudalism generally—a feudalism adapted to improved material possibilities of production—a sort of social mongrel, Feudal Capitalism or Capitalist Feudalism.”

To the best of my knowledge, DeLeon never elaborated exactly what he meant by that formulation. In 2025, however, we have a book by Jodi Dean (Capital’s Grave) in which she asserts that we are in fact right now in the process of rolling backwards in just such a direction, and that society as it currently exists is becoming more and more a capitalism/neofeudalism hybrid. After all, she says, a bigger and bigger portion of the global workforce is transitioning from being commodity-producing proletarians to largely unproductive “servants,” big capitalists are increasingly turning toward accumulation strategies revolving around political influence, rent-seeking and predation rather than improving the efficiency of the production process, and the “parcellated sovereignty” on display in everything from nations surrendering part of their national sovereignty in international treaties to the rising importance of private security contractors is interestingly analogous to the chains of vassalage characteristic of medieval feudalism.

In a chapter called “Forward Can be Backward: On Transition and Temporality,” Dean says that many Marxists might resist the suggestion that the mixture of modes of production in our society is becoming more feudal because we’re attached to a “linear” and “stagist” conception in which “history only moves forward.” She opposes this assumption in the name of a conception of historical transition (which she associates with Lenin, as well as a more nuanced understanding of Marx himself) that reveals the process to be “open, complex, and honeycombed with contradictory forms and tendencies.” On this conception, it’s to be expected that “[o]stensibly obsolete economic forms” will “sometimes reappear.”

In a review essay in UnHerd pairing her book with Bruno Leipold’s Citizen Marx, I parsed this as Dean saying that “the dialectic of history” was “going into reverse.” I clarified that while she doesn’t go so far as to claim that we’re already living under “neofeudalism,” she does “think that’s the direction in which we’re transitioning, and that the capitalist or semi-capitalist order around us already has many neo-feudal elements.”

I made several objections to Dean’s argument:

One problem with all of this is that, in seeking to update Marx’s analysis, Dean shows that she misunderstood it. Marx emphatically doesn’t exclude “service workers” in toto from the commodity-producing working class. In explaining the distinction between productive proletarians and unproductive servants, for example, Marx uses the examples of a cook in a hotel and a cook in a private home. The former is helping to generate the revenues out of which his capitalist is paying him. It’s hard to imagine that Marx could have picked an example closer to a Starbucks barista (the archetypal “service worker” in our time) without access to a time machine.

And a deeper problem is that so many of the trends Dean is talking about are reversions not to the pre-capitalist Middle Ages, but to the very form of capitalism Marx was writing about in the 1860s. What is Uber if not a high-tech application to the transportation sector of exactly the kind of casual and insecure labor practices previously tamed by organized labor and the regulatory state?

Nor did the monopolistic robber barons of the 19th century restrict themselves to accumulation strategies that would meet the approval of libertarian economics professors. . .

Regarding “parcellated sovereignty,” I made two points. First, at least in advanced capitalist countries, there are at least some respects in which ‘internal’ sovereignty seems to have been a lot more parcellated in the 19th century than it is now. The “Pinkertons that robber barons like Carnegie would bring in to crush strikes” were a lot less hesitant than their contemporary equivalents about using live ammunition. Second, while the post-World War II period in particular involved significant losses of nation-states’ ‘external’ sovereignty to treaties and global institutions, if anything, the most striking thing about the world scene at the moment is how much that trend is being undermined. Great powers, at least, are dramatically asserting their national sovereignty in the traditional ways that great powers do—through protectionism, blatantly claimed spheres of influence, and even imperial land grabs all of which “littered the military and diplomatic histories of the 19th century.”

All in all, I argued, to the extent that we’re witnessing a process in which “forward can be backwards” (in however “open” and “complex” and “honeycombed” a manner), the place to which we’re heading back looks a lot more like a return of the era of capitalism Marx was originally writing about than a return to those “forms” ostensibly made “obsolete” by the rise of capitalism itself.

Dean isn’t persuaded by this critique. That’s to be expected. But her response in Sublation is extremely odd in several ways. For one thing, she objects so strenuously to my use of the phrase “going into reverse” to describe her claim that ostensibly obsolete feudal or feudal-like forms are reappearing as capitalism slowly dies that she speculates that I may have “skipped the chapter on transition and temporality (which could be challenging to readers accustomed to skimming for quick takeways)…” This accusation is, I have to say, a new one for me. My detractors have, in my experience, been a lot more likely to accuse me of being an obsessive nitpicker than of not doing the reading. For the record, I didn’t skip any chapter (or sentence) of her book, even though the experience of reading it involved a fair amount of frustration. (In all fairness, I did think the section reflecting on the phrase “no one cares” was very good, and I loved the critique of Silvia Federici, or at least I did right up until Dean fumbled it by herself dismissing “productivism” in a very Federici-like way.) In any case, even a reader who really was content with a quick skim would have surely noticed that the title of the chapter in question starts with the words Forward Can Be Backwards, which would strongly suggest that Dean herself isn’t entirely averse to the spatial metaphor of capitalism “going into reverse” to the degree that it neofeudalizes. Perhaps such a reader would have even noticed that she spends a lot the chapter talking about how “elements from the past” persisting or even “reappear[ing]” in the present is well within the sphere of what we should expect given an appropriately open and honeycombed reading of historical materialism, all of which would lead them, as it led me, to thinking that, even if it’s not a clean or simple reversion (which I never suggested it was), what she’s talking about can be reasonably described even by her own lights as involving some element of reversion.

(Slightly) more substantively, she faults me for my “quick dismissal of parcellated sovereignty with a gesture to strike-crushing Pinkertons, as if the privatization of public law was not a significant and well-documented dimension of our present.” This is the sum total of what she says about either of my two objections to her account of parcellated sovereignty. She also implies that what I say about this subject is a “misrepresentation” but she not only doesn’t tell her readers what specifically I’ve misrepresented or how I’ve done so, but she doesn’t so much as give them a hint. It’s true enough that what I said, I said “quickly” (necessarily so, given the demands of a review essay covering two books at a magazine with less generous word limits than Sublation), but I’m less clear on why the reader is supposed to conclude that I was making a bad or unpersuasive objection to her claim that the internal diminution of sovereignty by means of security privatization represents a reversion to the ostensibly obsolete form of feudalism rather than just a reversion to the classic era of capitalism. Privatization of various kinds of security and enforcement services is indeed real. The question is whether this is best understood as neofeudalization or a reversion to an earlier era of capitalism. On that issue her response just isn’t responsive. And the half of my critique about external diminution of sovereignty she ignores completely.

Finally, Dean characterizes my objection to her misunderstanding of Marx’s view on productive and unproductive labor as a “flat out” misrepresentation on the grounds that she throws in a caveat somewhere that not all service work is unproductive. But let’s take a closer look at what she actually says about this subject in her book. At the bottom of p. 7 and the top of p. 8 of Capital’s Grave, Dean writes:

Although often overlooked by tech writers, the neofeudalizing tendencies of communicative capitalism show up most dramatically in the servant economy. I’m not referring here to the replacement of jobs by automation but to its opposite: the limits of automation…. As productivity increases, requiring fewer workers, those in need of a wage to survive are thrown into sectors less amenable to automation—that is, services… As Marx writes in Capital: “The extraordinary increase in the productivity of large-scale industry, accompanied as it is by a more extensive and more intensive exploitation of labor power in all other spheres of production, permits a larger and larger part of the working class to be employed unproductively. Hence it is possible to reproduce the ancient domestic slaves, on a constantly expanding scale, under the name of a servant class.” Marx’s argument resists simplistic progressivism. Capitalist industry itself reproduces social property relations characteristic of earlier economic forms. Capital’s social reproduction is not limited to the reproduction of capitalist social relations. It can also reproduce non-capitalist social relations of servitude.

A mighty service sector has arisen over the last decades. The majority of workers in advanced economies have moved from being producers to being servants. Employed in services and retail, they assist and help.

Dean’s contention here seems transparent, even when I reread it in the face of her claim to have been misrepresented. She’s saying that the service-sector workforce is (at least in general) a “servant class” and that the members of this class, like the servants Marx was writing about in Capital, aren’t “producers.” But, I’m sorry, this just is a misunderstanding on her part of Marx’s point. Being “productive” in his sense is about adding to the capitalists’ store of value by producing commodities that they can sell to consumers. Whether the commodity in question is a “good” or a “service” has absolutely nothing to do with Marx’s distinction. The servants he was comparing to ancient domestic slaves were literal servants, like maids and butlers, who weren’t adding to what Marx calls “the variable capital fund” (i.e. the part of a firm’s revenues out of which wages are paid). The economic interaction structuring the production of the services, rather than the nature of the services itself, is what made them unproductive. Marx is keen to emphasize that distinction. Hence, for example, in Theories of Surplus Value, he writes:

The cook in the hotel produces a commodity for the person who as a capitalist has bought her labour—the hotel proprietor; the consumer of the mutton chops has to pay for her labour, and this labour replaces for the hotel proprietor (apart from profit) the fund out of which he continues to pay the cook. On the other hand if I buy the labour of a cook for her to cook meat, etc., for me, not to make use of it as labour in general but to enjoy it, to use it as that particular concrete kind of labour, then her labour is unproductive, in spite of the fact that this labour fixes itself in a material product and could just as well (in its result) be a vendible commodity, as it in fact is for the hotel proprietor. The great difference (the conceptual difference) however remains: the cook does not replace for me (the private person) the fund from which I pay her, because I buy her labour not as a value-creating element but purely for the sake of its use-value. Her labour as little replaces for me the fund with which I pay for it, that is, her wages, as, for example, the dinner I eat in the hotel in itself enables me to buy and eat the same dinner again a second time.

More succinctly, in Capital itself, Marx gives the example of a private-school teacher who, Marx says, “is a productive worker when, in addition to belabouring the heads of his pupils, he works himself into the ground to enrich the owner of the school.” That this owner “has laid out his capital in a teaching factory, instead of a sausage factory, makes no difference to the relation.” I really don’t know how Marx could be any clearer about this. It’s fine, of course, for Dean to disagree with Marx (who was by no means right about everything), but it’s simply a fact that, in quoting Marx in an attempt to bolster her claim that the workforce of the retail sector, for example, is part of a class of “servants” rather than “producers,” she’s shown that she misunderstood Marx’s point.

While we’re on the service industry, a further ambiguity shows up in her response to me (as it does at several points in the book). In the Sublation article, she attributes the decline of class politics and the rise of various forms of contemporary identity politics in part to the fact that “waged laborers no longer see their work as driving social production and building a collective future.” They’re correct not to see themselves that way, Dean writes. They “aren’t producers.” Instead, “They are servants attending to the consumption requirements of the ruling class.”

Now, it’s a bit of a mystery to me how any of this is actually supposed to explain the decline of class politics (especially given the importance of e.g. hotel workers’ unions and transportation workers’ unions in the heyday of the American labor movement). But we can put all that aside. The bigger problem with this passage is that the great majority of contemporary service work does not in fact service the consumption requirements of the ruling class. It services the consumption requirements of other wage-laborers (and other non-capitalists). The world contains a great many more Denny’s waitresses and WalMart cashiers than it does sommeliers working at restaurants where any significant portion of the clientele can sanely be classified as part of the ruling class.

Nor is this a trivial point. It’s central. I don’t at all disagree with Dean’s premise that different modes of production often co-exist in a given society. But the way in which Marxists have traditionally (and, in my view, correctly) differentiated what the dominant mode of production was in any given society was by means of what the dominant “relation of production” was tying the “immediate producers” (whether slaves, serfs, or proletarians) to whoever controlled the production process. Capitalism, on this understanding, is a class society where the predominant relation of production is wage-labor. And that could hardly be the dominant relation in a society where capitalists only sold commodities (be they “goods” or “services”) to one another.

I have to say it seems like the capitalism of the 2020s includes an awful lot of capitalists pursuing accumulation strategies that rely on increasing productivity. If that weren’t the case, the rapid automation and AI-ization that Dean herself sees behind the shift to service work wouldn’t be taking place. Putting that point aside, though, there’s a more basic problem. However much predation and rent-seeking is going on in a given society, that which is being stolen or extracted in rent has to come from somewhere. As Marx famously points out in Ch. 5 of Capital, the source of the enormous wealth being so rapidly generated by the capitalist classes of his era couldn’t simply be some capitalists being better at business than others and thus doing a better job of buying low and selling high, since, “The capitalist class of a given country, taken as a whole, cannot defraud itself.” A similar point would seem to apply to the idea that the capitalist class, taken as a whole, is predominantly using the wage-labor-dependent majority of the population to sell services to itself. Whether the commodity being sold is an entire car or a twenty-minute Uber ride, in any thriving capitalist economy the consumers of the commodities capitalists employ workers to produce are mostly going to be other members of the working class. In mischaracterizing this reality, Dean’s story about service work obscures at least as much as it reveals.

Toward the conclusion of her response to me, Dean usefully highlights a point of agreement. “I agree with Burgis,” she writes, “that the future will be bleak without a powerful labor movement.” This isn’t a small point of agreement. It’s immense. But I’m confused by the claim with which she follows this up, that the “neofeudal hypothesis tells us where we can find [such a labor movement]: service workers.”

Now, I don’t think Dean is wrong to see an important frontier of organizing there. Reindutrialization has been a stated priority of both the current administration (in a strange, fumbling way with its incoherent tariff policy) and the previous one (which gave us the CHIPS Act) and if that trend continues in more effective ways in the future, manufacturing might increase a bit as a share of employment. If so, that would be very good for the labor movement given the greater strategic power of that sector of the class. Even so, Dean is right that, to the degree that automation (everywhere) and outsourcing (in more advanced capitalist nations) is shifting more of the working class of the United States and other countries toward service work, more labor organizing will have to take place in that sector. No argument there. But how Dean’s neofeudalism hypothesis, which analogizes service workers to domestic servants, is supposed to create the expectation that this layer of society will be the “vanguard” of future struggle is beyond me.

Under literal feudalism, there were centuries of periodic revolts by agricultural serfs that sometimes secured some concessions but were utterly incapable of overthrowing the overall mode of production. Only the rise of the modern capitalist class, whose ascent was solidified by the bourgeois-democratic revolutions that inaugurated the modern world, could do that job. And I’ve never heard anyone claim that medieval servants were typically even as insubordinate as serfs. I could be mistaken here, but my suspicion would be that some equivalent to Malcolm X’s memorable observation that domestic slaves were less likely than other kinds of slaves to rebel against their masters held true under feudalism as well, as it did in ancient slave societies. I understand that Dean isn’t being entirely literal in comparing service workers to the Victorian domestic servants Marx was comparing to ancient domestic slaves, but to whatever extent we’re supposed to take this chain of analogies seriously, it seems to me that this makes the expectation of a service-worker “vanguard” more confusing rather than less.

No form of class society prior to capitalism was overthrown in a process mostly led by that society’s labor force, acting on its own behalf (rather than being used as ground troops by some other class of the population). The Marxist expectation that the emancipation of capitalism’s commodity-producing working class will be the act of that class arises from a particular analysis of the role of that class in this particular mode of production, and it can’t simply be assumed that something similar will continue to be true if we’re truly exiting capitalism and entering something worse.

In other words, Dean’s good and healthy political instinct that the only way forward lies through a revitalized workers’ movement may be in conflict with her current theoretical analysis. If so, I hope she chooses the former.