I am so honoured and grateful for the opportunity to think with my critics about Kant’s Theory of Labour. Every article or book is something of an experiment, and this one was no different: an experiment to see what arose when I put labour at the centre of an analysis of Kant’s political philosophy, as well as a methodological experiment in reading Kant from an intersectional feminist standpoint. As Helga Varden points out, the difficulty is that, as an Element, this is a very short book, and it would be the height of hubris to imagine – or intend – to develop a complete account of the Kantian state in only 55 pages. I intended no such thing, and I acknowledge with my critics that there are many gaps here and questions (I hope generatively) unanswered. My goal was to take up a very specific thread of Kant’s arguments and to use it to shed light on the systematic injustices in Kant’s vision of the state, in ways that might put Kant scholarship in conversation with new interlocutors from Marxist, materialist and social reproductive feminisms, and Black radical traditions.
This book is experimental in a second way: in writing it, I am pushing on my training as a Kant scholar, rather than prescribing a way of reading to anyone else. I am interested in how questions of freedom and liberation intersect with questions of accountability because I am not interested in simply sanitizing Kant in order to take ‘the good bits’ and leave the rest. The particular practice I name in the book is the tendency to treat Kant’s universalism as systematic, and his exclusions as erratic, and to prioritize the former accordingly. So I locate my work in an emerging tradition that aims to engage with Kant’s exclusions, and his commitment to injustice as systematic and intentional.Footnote 1 I am less interested in why Kant was committed to these systematic injustices and more with the impact of these commitments that followed from his role as a scholar, teacher, and public intellectual.Footnote 2 I agree with Varden that we need an account of how we, as philosophers and Kantians, move forward and theorize in better ways; my claim is that as we bring Kant forward with us, we must do so in light of, and not in spite of, the ways in which he embeds systematic injustices in his philosophical system.
This means, too, grappling with the problem of terminology that Varden names – which is of course not new to Kant studies (Kleingeld Reference Kleingeld1993). I emphasize that gender is always raced for Kant so that the ‘women’ he describes throughout his anthropological, historical, and political work are explicitly ‘white’, often middle-to-upper-class women in what he deems the ‘civilized’ parts of Europe. These are what I call ‘proper women’ for Kant, in that they fulfil nature’s two purposes for women, both biological reproduction and the reproduction and development of civilization and culture (Anth, 7: 305).Footnote 3 Varden troubles my use of the term ‘white women’ for these women, and I agree with her that it is not so simple: Kant’s hierarchies are not limited to racial ones but include national, ethnicized, and classed ones. The criteria one must meet to be a ‘proper woman’ in this sense are extensive – one thinks of Mr. Darcy’s similar description of the many attributes possessed by accomplished young ladies.Footnote 4 As Varden points out, there are many ‘white’ and European women who would not meet this standard. My use of the word ‘white’ as a placeholder here is thus perhaps too broad. I use ‘white’ in part because Kant is in many ways the originator of ‘whiteness’ as accomplishment (Larrimore Reference Larrimore and Judaken2013) and in part in the sense developed by Alcoff (Reference Alcoff2023) and Mills (Reference Mills, Gross and McGoey2022), to name the shifting and relative criteria of racial and cultural superiority, which are adaptable and sensitive to cultural and historical context. It may not be the right word. I look forward to discussions within Kant scholarship about how we should approach this difficulty once we recognize that we need better and more rigorous language to systematically name the exclusions built into Kantian categories like ‘woman’, rather than assuming that Kant’s term ‘woman’ is inclusive of all who now take ourselves to be women.
1. Conditions of dependency and the right to work one’s way up
Speaking of women, Varden argues that she and I disagree on Kant’s account of women’s right to work their way up from a passive state to an active one – but I think our views are closer than she credits. My claim is not that there are any legal barriers to women (or servants) ‘working their way up’; certainly, I do not think this is illegal in a Kantian state. My point is collective and structural: that if any woman, wage worker, or servant can work their way up, it does not follow that all can work their way up. I develop this argument by examining Kant’s nod to practices of outsourcing: that as any given woman, for example, works her way up, she outsources the labour of the household in order to do so. Someone else takes her place as a dependent labourer – and, given the raced and ethnicized division of service and caregiving labour, then and now, this perpetuates patterns of oppression and dependency. I do not dispute that Kant has offered a pragmatic and meritocratic liberal vision of the right of each individual to advance in society, but I want to make visible the patterns of structural servitude that this leaves untouched. The barriers with which I am concerned are economic and social, working at the collective level, rather than legal, working at the individual level. My priority is to draw our attention to the ways that this patterned economic dependency produces political disenfranchisement in ways that should trouble a Kantian account of ‘civil equality’.
Kant distinguishes between the equality of innate right, or our ‘equality as human beings’ (MM, 6: 315), and what he calls ‘civil equality’, which is tied to one’s standing to vote. In the drafts for the Doctrine of Right, he describes it as the ‘potential’ for being a legislator (VMM, 23: 293). The distinction is illustrated through reference to the ways that the wigmaker is an active citizen, while the barber is not. The problem Kant identifies is a pragmatic one: those who control their own means of production, and thus create a product that can be sold, have bürgerlichen Selbstständigkeit, often translated as civil independence or self-sufficiency.Footnote 5 Those who must directly work for others, and who do not therefore produce anything they themselves can sell, are dependent on ‘arrangements made by another’ (MM, 6: 314).
This picture is complicated by the fact that persons are not really ‘independent’ in the ways that it assumes, and Kant’s own emphasis on the dependency of workers belies this picture.Footnote 6 This is one way that I think we get a different vantage point on the Rechtslehre when we begin by asking questions about work. Because Kant links political and material dependency – heaping the former on top of the latter – addressing political dependency means addressing material dependency, namely, the dependency of workers upon their employers (and husbands).Footnote 7 Of course, the main problem here is that there is little evidence that Kant himself thought that this dependency needed to be addressed at all. As Axel Honneth recently noted, Kant is right that employment relations in a market economy create conditions of dependency, but from there he arrives at the wrong solution: that we resolve the problem by distinguishing the independent from the dependent by developing the category of ‘passive citizens’ and simply accepting that universal suffrage is neither possible nor desirable (Celikates et al. Reference Celikates2023). This leaves us with an account of ‘civil equality’ which is simply untenable – particularly for a philosophical system which prides itself on universality. The problem is not just that material and political inequality is here rendered ‘rightful’, but that the inequality of dependency will always be patterned. By analysing the divisions of labour in Kant’s account of the state, we can see that it is women and people of colour (and other classed, colonized, and ethnicized others) who will be systematically relegated to conditions of material and political dependency, generating the patterns of second slavery that Du Bois identified (Reference Kant and Gregor2017). My account shows how these patterns are politically enforced through what we might think of as a Kantian ‘passive citizenship’ strategy. That Kant, himself, did not see these patterns of exploitation and exclusion as a problem for his defence of a liberal meritocracy in which ‘each’ can ‘work his way up’ reflects the ways in which this pattern was likely a feature, rather than a bug, of his broader philosophical system, rendering the varied hierarchies explored in his geographical and anthropological works as politically ‘rightful’.
Contemporary Kantians are less complacent about giving up on universal suffrage through passive citizenship strategies, and so the key question – that posed by each of my critics – is what Kantian resources we find for working to address and remedy these patterns of exclusion. I acknowledge that these prescriptive moves (as Elvira Basevich rightly calls them) were less the target of Kant’s Theory of Labour than the diagnostic ones. There are, I think, two main pathways suggested by Basevich’s and Martin Sticker’s arguments, which are often taken as mutually exclusive: the state enforcement or socialization of employment contracts, and direct state support, in the form of welfare or a Universal Basic Income (UBI).
2. Kantian policy possibilities and the refusal of dependency frameworks
I want to reject the idea that these pathways are mutually exclusive, particularly when we examine them by focusing on patterns of gendered, racialized, and ethnicized labour. After all, as Sticker has argued elsewhere (Reference Sticker2023), many arguments for a UBI are not truly universal, meaning that they are not global or allow exceptions within the territory of a nation state. The economic logic of a UBI then relies upon access and entitlement to the cheap and exploited labour of (invisible) others. I want to note two things about this argument. The first is just how similar it is to Kant’s claim that all private rights are in some sense provisional until a rightful order is made global (MM, 6: 350); the rightfulness of a UBI likewise seems to hinge upon its universality. The second points to a problem central to Kant’s Theory of Labour: a UBI aims, as Sticker points out, to bypass the problem of labour in creating conditions of universal independence. Like Kant, it proposes to solve the problem of dependence with independence, without grappling with the ways in which independence is always already an illusion: all independence is premised upon access to someone else’s labour. All opting out of work and claiming leisure as the ‘highest good’ is premised upon entitlement to someone else’s labour – often, the kind of labour people would be unlikely to freely ‘opt’ into.
This is not to say that I am arguing against the Kantian case for a UBI. On the contrary, I think it is a strong case, and it gets several things right about Kant’s conception of work, citizenship, and the role of the state. It takes the problem of passive citizenship to be not dependency per se, but dependency upon other private citizens, and it recognizes, as Sticker points out, that it is not one’s labour that makes one an active citizen, but one’s independence: the material stability and leisure that one’s labour makes possible. Kant was, after all, writing at a time when most of those who would count as ‘active citizens’ would do so on the back of inherited wealth and hereditary nobility; the independent bourgeois – the professor and wigmaker – are newly admitted into these ranks. Accordingly, participation in public reason, Kant reminds us in What Is Enlightenment?, must be separated from one’s work or one’s ‘office’: we participate as a citizen, rather than as a worker (WIE, 8: 37). This model of participation frees up wage labourers to participate, despite their dependency, but it also assumes that they have the leisure to do so.Footnote 8
This refusal of dependency and the associated claim to leisure have been important in feminist arguments about rethinking the domestic sphere, from the wages for household movement to Angela Davis’s vision of the ‘obsolescence of housework’ to contemporary calls for the abolition of the family.Footnote 9 But these arguments have consistently pointed out that a basic income alone – whether universal or targeted to domestic labour – will not suffice, as it leaves existing divisions of labour intact. It targets the material conditions of dependency but not their structure: a UBI may be individually unconditional, but it is collectively conditional on some working to provide the goods necessary for the survival and flourishing of all. This is not unlike Kant’s ‘each can work their way up’ proposal, which is similarly individually universal but collectively and structurally dependent on some being relegated to positions of dependency. In both cases, the problem is that of necessary, or ‘essential’, work. Without transforming our collective reliance on this labour – and our collective awareness of our reliance on this labour – we do not solve the structural problem. We merely pass it down the chain.
Thus, I think Sticker is right to say that work is not the only route to independence, but I would trouble this by pointing out that all independence relies on labour. This is to say that while a UBI is a route to independence, this public dependence must include, as Basevich argues, the state’s duty to regulate and enforce minimally just conditions of labour. That is, as long as labour is required to keep our sewers functioning, hospitals staffed, and food supply stable, we have a collective obligation to ensure dignified wages for such labour.Footnote 10
Let us begin with wage labour. As Basevich points out, contracts are a kind of ideal for Kant, but to operate as such, participants would need not only to align their ends but also to have the bargaining capacity to do so in a way that is non-exploitative. The structure of the contract is meant to ensure the former: Kant posits that contracts are structurally symmetrical so that although my employer is ‘enriched’ by my labour, I am ‘enriched’ by my pay (MM, 6: 274). As Basevich notes, there is little symmetry in the meaning of ‘enrichment’ here. And yet, Kant is not insensitive to the ways in which labour contracts must protect the rights of labourers. This is borne out by the perspective Kant assumes in the Dogmatic Division, where he refers to ‘granting another the use of my powers for a specified price’ and ‘carrying on another’s affairs in his place’ (6: 285) – descriptions of employment contracts that explicitly take the perspective of the employee, rather than the employer.Footnote 11 This suggests that Kant is concerned with the conditions that make labour contracts rightful from the employee’s perspective, since it is the selling of one’s labour power, rather than the buying of it, that most directly threatens innate right.
Nevertheless, the formally symmetrical structure of contract right is insufficient to ensure that the bargaining power of workers protects against exploitative contracts. And yet, as a gradualist and an institutionalist (rather than a revolutionary, like Marx), Kant must be able to account for workers’ rights to assert justice claims. I have suggested that wage workers’ right to participate in public reason (even if they could not vote) offers one such basis for labour organizing and unionizing (Pascoe Reference Pascoe2022: 12). However, as Kant’s own account of dependence suggests, minimal conditions of labour justice must be material as much as they are political. Ensuring just bargaining positions for workers could then be considered a critical task for the state’s right of distributive justice. As Sticker argues, a UBI could be one way to do this.
On the other hand, there is a logical order to Kant’s trichotomy: property right is prior to contract and domestic right. The assumption is that because we need to own things in order to trade or sell them, property precedes contract right, so that contracts merely move about the wealth that already exists, ‘enriching’ people in the process. The state, then, has a duty to enforce property rights, with a few exceptions (Messina Reference Messina2021), and the state’s capacity to enforce contract rights would seem to be limited by this prior commitment. Thus, Kant understands contract as an essentially private agreement between two independent persons (Ripstein Reference Ripstein2009: 109, 120) and thus views the state’s capacity to enforce it as limited by Private Right, rather than as a project shaped by the distributive justice that he attributes to Public Right (MM, 6: 306). The literature on Kantian approaches to distributive justice and welfare rights is vast,Footnote 12 but the question of how this informs either the state’s right to ensure that labour contracts protect innate right, or to distribute necessary labour, has been underexplored.
One option is to make as many wage workers as possible into civil servants, which makes those employment contracts public, rather than private. This is suggested by Kant’s own distinction between the dependence of wage and domestic workers and the independence of civil servants as those who labour on behalf of the state. The distinction has nothing to do with the sort of work required or the sort of compensation offered; the sanitation worker employed by the state is not necessarily better off than the housekeeper employed by a large hotel. The difference, on Kant’s account, is that the sanitation worker is not dependent on the private will of another, but on the public will of all, of which she is a part. Therefore, there is a sense in which she works for herself – she is ‘self-directed’, as Du Bois would put it – and so she counts as an active citizen. Such a socialization of work involves a transformation of private right in the name of public justice which is not without challenges on a Kantian account, but it offers one pathway towards addressing the problem of political exclusion shaped by Kant’s account of dependence.
3. The political value of the domestic sphere
The resources for challenging the enclosed dependency of wives and domestic and service workers are more complicated – not least because it means, as Varden puts it, considering how to make domestic relations less vulnerable to abuse and ‘bad’ dependencies. My account emphasizes these ‘bad dependencies’, drawing out a conception of the household that links marriage and domestic servitude. It distinguishes the ‘enclosed dependence’ that characterizes the household, which is organized hierarchically by a duty to ‘share the ends’ determined by the head of household, from the dependency of wage work with its enforceable practices of aligning ends. Basevich links my account of ‘enclosed dependency’ to Du Bois’ conception of ‘second slavery’: both name conditions of private dependence with economic and political consequences distinct from wage labour, which are structurally organized to prevent certain social groups from realizing the ‘potential’ of civil equality. What, then, is the prescriptive proposal for this bleak picture? The Kantian account of domestic right is not without political promise. Varden worries that my analysis of the Kantian state cannot offer a critique of the right to marry, and she is right that this was beyond the scope of the arguments in this project, although I have explored the question elsewhere.Footnote 13 Kant presents the domestic sphere as an essential structural element of the rightful state – a ‘new star’ rather than merely a ‘stella mirabilis’ (MM 6:358-9). He recognizes the domestic sphere as containing a set of activities – like sex and intimate and caregiving labour – that constitute the use of another person, rather than the danger of treating them as a mere means. This necessitates theorizing a domain of life in which persons have rights to one another, such that they share and align ends in ways that transform their external freedom. Varden is right that, if this sphere is so important, it cannot exclude same-sex or polyamorous couples. Attending to Kant’s vision of the domestic sphere as inclusive of domestic servants can, I think, help make this case: it need not be limited to two people (nor to kinship-based familial relations). But with this inclusion in mind, the deeper prescriptive question is: what kinds of political and material conditions would allow this condition of sharing ends to be a productive transformation of one’s external freedom, rather than a condition that creates political dependency in the name of generating leisure and independence for some of its members? Certainly, this is critical to the work of making marriage and the domestic sphere less susceptible to abuse or violative dependencies.
One possibility is that we have largely solved this problem by granting women the right to vote and participate in the public and market spheres and granting same-sex couples the right to marry, ensuring that access to these domains is not limited in the ways Kant imagined.Footnote 14 But to make the domestic realm a rightful condition, I think we must consistently recognize it not only as the site of marriage but also as the site of a distinctive form of labour, and we must reimagine the role of the state in protecting those who do that labour. I think there are resources in Kant’s account for doing so, beginning with the recognition of this domain of labour, as well as Kant’s distinctive recognition of this labour as involving the use of persons, beyond their treatment as a mere means. Accordingly, Kant describes this domain as a condition of ‘shared ends’ – precisely as the public realm is a condition of shared ends. The fact that Kant thought these ends were to be dictated by the head of household is a problem (much as his idea – possibly abandoned in his mature work (Kleingeld Reference Kleingeld and Watkins2018) – that public laws required only the possible, rather than the actual, consent of the governed is a problem for democratic theorizing). But like his limits on women’s citizenship and same-sex marriage, we could, as Varden suggests, reject this claim in order to work closer to an ‘ideal’ conception of domestic right, particularly when we remember that the supposed ‘independence’ of the head of household is made possible by access and entitlement to the labour within the household (troubling the idea that anyone, in a household, is really independent).
Once we reframe the domestic sphere as a site of interdependence – rather than of dependence and independence (made possible by that dependence) – then I think we ought to attend to the fact that the conditions of end sharing in the domestic and public spheres take a similar form. This offers us an ideal of the domestic sphere as the place where we learn to work and think together, towards common ends, transforming our own preferences in the name of this collective project. We would develop few of these skills in the other domains of private right, where we acquire property and enter into contracts as supposedly independent people, each out to enrich ourselves. The domestic sphere is a site of critical political learning, on a Kantian account, and reveals how it is not only the ‘leisure’ allowed by access to the domestic sphere but also the practice of interdependence – sharing and pursuing ends, together – that supports political participation.
If I am right about the political value of the domestic sphere on a Kantian account, then this makes a different case for state interest in rightful conditions within the domestic sphere. This is particularly important for Kant, who thought the activities of the household involved forms of use that could not be remedied through contract: in other words, that outsourcing these activities as wage labour would only exacerbate the problem of exploitation (see his views on sex work) (Pascoe Reference Pascoe2023). If the activities within the domestic sphere must be organized by shared ends, then there are two possible remedies: ensure that the shared ends within the domestic sphere are not structured so as to produce enclosed dependency and thus disenfranchisement, or outsource these activities to the state, which is another domain of shared ends.
One could accomplish the latter goal by, as Angela Davis (Reference Davis1983) argued, industrializing and socializing housework and reshaping private life so that as much house and care work as possible could be shared and by providing state-funded industrialized house cleaners to offset the collective burden. While we might hesitate to grant the state access into our homes in these ways (in the USA, we may reasonably not want the government that overturned Roe and justified surveilling our period-tracking apps to also empty our pockets and do our laundry), this offers us a thorough and radical vision of what transforming dependent labour into public labour might look like.
The former remedy might support the case for a UBI, as well as the state enforcement of just contracts and the expansion of the public sphere. For example, a UBI in conjunction with state laws limiting the length of the working day/week would allow citizens more time in their preferred domestic constellation (I do not assume, with Kant, that the marital or kinship-based family is the only option), and it might allow greater creativity in determining this constellation (Weeks Reference Weeks2020)). State-supported care labour and household labour might, in turn, allow greater equity in the distribution of labour within it. And if the domestic sphere is a critical site of political learning and practice, as a space in which we learn to share ends and pursue common projects, then these supports are not only about maximizing individual choices but also about supporting our common political projects and developing our capacities as citizens.
Did Kant understand the domestic sphere in these terms? He certainly nodded to it as a site where ‘civilization’ and ‘culture’ were developed; the role of ‘proper’ women was to cultivate these sorts of practices. But he understood political practice as distinct from these sorts of things, and for all his emphasis on the capacity of active citizens to vote, his own commitment to republican, as opposed to democratic, practice has been the site of much debate (Davies Reference Davies2020). The primary purpose of citizens, on Kant’s account, is participation in the rational deliberation and public reason of the public sphere, which requires both leisure and material independence.Footnote 15
My suggestion is that once we refuse this story about material independence or self-sufficiency, other purposes of citizens become possible. If the domestic sphere is where we learn to share ends and pursue common projects, then caregiving labour becomes not a merely domestic project but a political one: an activity proper to citizens. This not only reorients our understanding of the citizen but also reshapes our conception of the state’s duties: supporting the participation of citizens in domestic contexts (not limited to kinship families) becomes a political project, supportable by a basic income and the enforcement of just employment contracts that limit the working day. And state support for public institutions of caregiving – including but not limited to public schools, childcare, and eldercare – is critical infrastructure for developing the capacities of citizens in ways that protect them from depending on the private will of another (as Du Bois and the early freedmen who fought for public schools in the American South understood (Basevich Reference Basevich2024)). On this account, the labour characteristic of the domestic sphere ceases to be something one wants to ‘work one’s way’ out of, but rather critical infrastructure for the project of citizenship.
This is, admittedly, only a partial solution; it will not solve all our problems of labour exploitation or the patterns of local and global outsourcing that render them invisible. But a valuable feature of Kant’s insistence on treating the domestic sphere as critical infrastructure for the state is that we can take this ‘new star’ as a starting point for reimagining its role in a vision of justice and consider the constellations of institutional policy that might support it.
Acknowledgements
My gratitude to Eoin O’Connell, Helga Varden, Howard Williams, and Laura Pascoe for the discussion of this piece, as well as to my additional critics Maria Mejia, Garrath Williams, Martin Brecher, and Ewa Wyrębska-Đermanović, and to the audiences at author-meets-critics sessions at the PPE Conference in New Orleans in October 2023, the ‘Using People Well, Treating People Badly’ workshop organized by Martin Sticker in September 2023, and the Kant Reading Group organized through Cardiff University.
Open access
