Capital demands that we dissolve the difference between thing and person as a [mere] prejudice. – Hermann Cohen (609)Footnote 1
Law hews to all the forms and deformations created by the obligation of labour. Capital is the great inventor; law must build the catacombs in which it buries itself. – Hermann Cohen (610)
1. Introduction
Many authors see Kant as uncritical of market relations and even capitalism. His formal account of the just legal order emphasises individual rightsFootnote 2 to property and contract; it shows little concern for material inequalities. This article considers the response of a leading ‘neo-Kantian’, Hermann Cohen. In his commanding Ethics of Pure Will (1904), Cohen invokes Kant’s ethical principles for socialist ends. Capitalism, he argues, fails to respect workers as ends in themselves, using them merely as means.
My moral sympathies are with Cohen: I agree that Kant’s principles call for a critical attitude to systematic inequalities in wealth – above all, if these inequalities involve powers to use people as mere means. While Cohen’s account is suggestive, however, I argue that it suffers serious flaws; we can remedy these by drawing on some aspects of Kant’s approach, as well as a central aspect of Cohen’s.
More specifically, Cohen’s reservations concerning Kant’s account of right lead him to downplay its fundamental materialism: that right is essentially coercive. I suggest that the Doctrine of Right offers critics of capitalism more than Cohen allows, because Kant insists on this most brutal material reality: coercion. When they are rightful, legal relations bring coercion within the rational structure of human interaction. Equally, this gives us compelling reasons to critique legal structures: are they fully rightful, or do they give some people more coercive authority than others?
Following the same logic, I also criticise Cohen for invoking legal ideas without appreciating their material implications. Rather than Kant’s ‘fact of reason’, Cohen invokes the following ‘fact’ of jurisprudence. As a legal person, a cooperative (Genossenschaft) can incorporate individuals as free and equal members. For Cohen, this legal idea provides ethics with a worldly point of reference – unlike Kant’s individual ‘fact of reason’ or ideal ‘kingdom of ends’. From a Kantian perspective, however, Cohen’s figure is also unworldly. It downplays the exclusions and constraints involved in any actual legal person, be it a cooperative or trade union, a business corporation or a state.Footnote 3 At the same time, Cohen’s ‘fact’ invites an important question: what role do legal persons play in capitalism?
On the interpretation offered here, capitalist legal structures lend authority not only to individual property owners. They also grant legal power to capital. Beyond individual property, capitalism rests on money, employment, and a specific sort of legal person: the business corporation. To see the resulting form is to see why capitalism is unjust. It betrays Kant’s animating principle of equal freedom; it thereby commits the wrongs Cohen alleges in my first epigraph: it invests things with legal agency; it allows many people’s use as mere means.
2. Scepticism about Kant’s formalism
On one view, Kant’s emphasis on the form of legal relations can never be adequate to the realities of capitalism – and especially not its most exploitative forms. I have mentioned Kant’s unconcern with material inequality. Most starkly, he writes:
this thoroughgoing [formal – GW] equality of individuals within a state, as its subjects, is quite consistent with the greatest [material – GW] inequality in terms of the quantity and degree of their possessions, whether in physical or mental superiority over others or in external goods… the welfare of one is very much dependent upon the will of another (that of the poor on the rich)… thus one serves (a day labourer) and the other pays him, and so forth. (8:291f)Footnote 4
Worse, Kant heaps formal insult on material injury. On his view, many people should be merely ‘passive’ citizens. Formally, everyone is a legal subject, hence neither slave nor serf. But this does not mandate equal citizenship. A ‘shop clerk, a day labourer, or even a barber’ (8:295n), not to mention women and servants: all are dependent on others. So they should not enjoy the rights of ‘active’ citizens: legal action and political participation (6:314f). We may doubt whether Kant would have been more generous to nineteenth century factory workers.
Hence, Kant’s formalism seems not merely blind to material inequality. It sees it in order to disqualify many people from full citizenship. Twice over, material factors undercut Kant’s promise of equal external freedom. As Cohen puts it, ‘Kant did not exercise that free, unbiased, sovereign criticism of positive law that gives his transcendental criticism its true life and its powerful fruitfulness’ (Reference Cohen1910: 399, as cited by Widmer Reference Widmer2024a: 76).
In the face of these betrayals, the obvious response is to insist, first, on the importance of equal civic rights and, second, on the material realities. The first battle was still on-going when Cohen wrote: women were passive citizens; German men’s civic status differed with their taxation class. Cohen argued for equal, universal voting rights in a short essay published in the same year as Ethik des reinen Willens (Reference Cohen, Görland and Cassirer1928 [1904]; see also 518f).
Second, if formal equality masks injustice, we should look to the material realities it hides.Footnote 5 Even granted equal civic rights, labourer and factory owner enjoy quite different rights and duties: to live, the labourer must submit to the owner’s terms. Law functions as a tool for the wealthy; the worker becomes a mere means for the factory owner (320f, 611; also Reference Cohen1889: 139; Reference Cohen1910: 280ff, 326f). In Cohen’s potent formulation: when the labourer’s actions are dictated by the factory owner, not by the labourer himself, this violates the worker’s status as a unified ethical personality and end in himself. Material inequality turns formal equality into an ideological mask; worse, it tears the labourer in two (604f).
Like Cohen and many other readers, I find Kant’s unconcern with material inequality deeply troubling; still worse, that he cements it with the formal category of passive citizenship. I admire Cohen’s determination to use Kant’s thought creatively, to resist realities that Kant wrongly endorsed (the day labourer’s servitude, the poor person’s dependency, many people’s second-class citizenship) or never witnessed (the vast, grinding brutality of nineteenth century capitalism). However, I will argue that Cohen’s ethical critique is too gestural, and that we can take more from Kant’s formal account of right than Cohen allows. Kant’s account can deepen Cohen’s ethical critique by revealing the legal form of capitalism – not least, if we follow an important clue at the centre of Cohen’s own theorising: the idea of legal personhood. The resulting legal form betrays both Cohen’s ethical emphasis on humanity as an end in itself and Kant’s formal principle of equal freedom.
Kant summarises his emphasis on legal form as follows: ‘the pure doctrine of right… concerns only what is formal about a faculty of choice that is to be limited in its external relations according to laws of freedom, regardless of any end (as its matter)’ (6:375). The last clause brackets even the most basic ‘ends’ of living beings – not least, our need for subsistence. The focus on the ‘form’ of private property brackets all inequalities in holdings – that is, the ‘matter’ of these property rights.
Still, Kant’s account has a material basis: ‘limit[s] in… external relations’ – or in other words, the power to coerce. Reason itself says that coercive limits are fundamental to right:
the universal law of right: so act externally that the free use of your choice can coexist with the freedom of everyone in accordance with a universal law… reason says only that freedom is limited to those conditions in conformity with the idea of it and that it may also be actively limited by others [i.e. through coercively enforced laws – GW]… (6:231, bold is my emphasis)
With these words, Kant brings might within right. ‘Might makes right’ on the condition that might upholds the principle of right: that no one may exercise greater power to bind others, regardless of ‘physical or mental superiority over others or [superiority] in external goods’ (8:291). This cannot be the whole story: as we have seen, Kant grants greater civic rights to just a few people; ‘superiority in external goods’ (money, factories, the bare means of subsistence) has decisive effects on people’s lives and relationships. Nonetheless, Kant’s account is not merely formal: the authorisation to coerce belongs to the form; positive law institutionalises this.
3. Cohen’s ideal ‘legal person’
Cohen takes a different approach. Coercion, he says, ‘has not grown on the ground of transcendental freedom… Coercion forms, both logically and ethically, the end of reason’ (Reference Cohen1910: 395ff). Cohen knows that actual states use force, of course; even the ideal state must punish (373ff). Unlike Kant’s ideal of the rightful state, however, Cohen’s two central ideals have no conceptual relation to force.
First, Cohen emphasises the importance of social relations that uphold ‘humanity’, understood as people’s capacity for moral agency (Reference Cohen1910: 274; Widmer Reference Widmer2024b: 615f). Cohen’s Ethics of Pure Will argues that the will is pure by virtue of its fellowship with others, its ethical orientation to every person’s status as end in themselves (185ff). As against the formalism of Kant’s moral theory – that is, Kant’s insistence on the formal quality of the moral law, and not only of legal rights – Cohen emphasises the substantive end of each person’s humanity (320ff). Cohen’s ‘ethical socialism’ condemns capitalism for treating ‘the difference between thing [that which is merely a means] and person [as end in themselves] as [mere] prejudice’ (609). On this point, I think every Kantian must accept Cohen’s claim. Notwithstanding its formal aspects, Kant’s moral theory is anchored in humanity; to confuse persons and things is to betray morality at its root. If social and economic relations systematically treat many people merely as means, we are duty-bound to challenge and change them.
Second, the pure will is not bound to others by communal belonging – not by affection or family ties, ethnic belonging or cultural affiliation, nor by attachments that expand as they encounter outsiders and strangers.Footnote 6 Rather, Cohen’s ‘transcendental method’Footnote 7 turns to an idealised jurisprudence, whereby people are incorporated into a single ‘legal person’: a Genossenschaft (fellowship or cooperative) or rightful state (e.g. 224f). This upholds their capacity to pursue their own ends. But it also transforms them by clarifying the task of moral agency (237, 338f): each must recognise the claims and moral agency of every other member, regardless of similarities or differences; just as law demands that we recognise every person as a bearer of rights, regardless of specific bonds or affiliations. For this reason, ‘the term “moral person” is used for the “legal person”’ (236; see also 249). Cohen calls his method ‘transcendental’ because it concerns what makes possible this ‘Faktum’ (227f) of pure jurisprudence: a unified moral world or ‘Allheit’ (231 and passim). For Cohen, this approach overcomes Kant’s individualistic focus on maxims and appeals to conscience (thus the notorious ‘Faktum’ of reason: 5:47, 55; cf. Reference Cohen1910: 255f). It also leaves behind the Christian otherworldliness and paternalism that Cohen finds in Kant’s kingdom of ends (4:433f) – a ‘transcendent shepherd-world’, as Cohen cruelly dubs it (28).Footnote 8
Cohen’s moral turn to legal personhood is intriguing and provocative. I will dwell on both aspects below. But let me note one provocative omission straightaway. Republicanism is the political theory of how citizens might constitute a single legal person: the body politic or ‘public thing’ that constitutes the freedom of each. Cohen makes just one glancing reference, and then only to Roman antiquity (324). This omission is remarkable since it bears on all Cohen’s key points of reference: the unifying moral or legal person, Kant, and socialism. Cohen knew well that Kant sees the state as a ‘moral person’ (6:274, 6:343) and that this ‘person’ must be republican (Reference Cohen1910: 448f, 536f). Cohen should also have been aware of intense contemporary debates among socialists about republican ideas (see e.g. Leipold Reference Leipold2024). That is to say: Cohen sets aside a key political resource for understanding how people might be unified as free and equal.
Instead, Cohen refers to the ‘pure form’ of the Genossenschaft, as developed in the German legal tradition. He especially relies on Heusler’s Institutionen des deutschen Privatrechts (Reference Heusler1885) – that is, the institutions of private law, rather than those of public and political life.Footnote 9 Cohen pictures a fellowship or cooperative (Genossenschaft) whose members have equal rights to knowledge and participation; where property serves communal ends rather than selfish ones (229ff). This is not meant as a blueprint for reform, nor as a ‘“solution” to the problem of how to achieve unified ethical subjectivity’ (Hollander Reference Hollander2021: 46). Rather, it represents an inspiration for political struggles: only actual political engagement can decide how or in what form cooperatives or other associations might challenge current injustice. Still, just by its nature, the Genossenschaft has two features that trouble Cohen’s notion of ethical unity or Allheit. Its membership is bounded: that is, it excludes. It must involve government: even a self-governing fellowship imposes duties and must reckon with disputes.
As regards the public realm, Cohen also offers this legal person as an inspiration for the rightful state: the Rechtsstaat incorporates all citizens, uniting the free choice of one person with that of others (269). Again, this is not meant as a political blueprint. Cohen’s aim, instead, is to help us gain insight into ‘immoral prescriptive concepts manifested in legal laws’ (Widmer Reference Widmer2024a: 78): the injustice of unequal voting rights (518f; Reference Cohen, Görland and Cassirer1928 [1904]), for example, or capitalist economic relations. Cohen’s concern remains the never-to-be-completed political task of giving form to ethical ideals, as we seek to overcome the injustices of actual historical constellations. Still, the state as legal person shares those awkward features of the Genossenschaft. However republican or democratic or socialist, every state must impose policies over many disagreements. Its membership and jurisdiction are bounded; it excludes.
Cohen is not unaware of these limits. In particular, Cohen interprets ‘love of neighbour’ so as to overcome the problem of inclusion/exclusion. He draws on Leviticus 19:34: ‘The stranger shall be to you as the native-born’ (214f; Hollander Reference Hollander2021: 199ff). Which is to say: love of neighbour is love of human beings; the pivotal figure is not the neighbour, nor someone who is already a fellow in any given Genossenschaft. Instead, it is the ‘stranger’, who must be treated as a member and not as a stranger. Hollander puts the point beautifully: ‘The “stranger” is no exception for Cohen, no interruption of the native-born, territorial order; rather, it bespeaks the possibility of political and ethical action as such’ (Reference Hollander2021: 260). Seen in this way, the Cohen-ite legal person represents ‘an open-ended task’ (ibid: 54). Nonetheless, the task is inherently demanding, and demanding on both sides. Members are not just rights-holders: they must also fulfil duties. To the extent that the non-member has a choice in the matter, it is between exclusion or the scheme of government which brings members together.
When Cohen calls Kant’s kingdom of ends a ‘transcendent shepherd-world’ (28), he faults it for Christian other-worldliness, for floating above actual economic and legal arrangements. As Kant himself says, it is ‘only an ideal’ (4:433). However, Cohen’s own figure is open to similar criticism. Cohen emphasises the moral power of legal personality: mutual recognition and fellowship orient our moral agency toward Allheit, transcending natural bonds such as kinship or ethnicity. But he sidelines the uncomfortable features of actual legal persons: by their nature, they involve limits and constraints. Kant’s kingdom of ends and Cohen’s ethical jurisprudence both promise to incorporate every person as an end in themselves. But neither corresponds to any worldly reality; neither shows how people can overcome the exclusions and constraints involved in even the most egalitarian association.
This is not to disagree with Cohen’s insistence on the social relevance of Kant’s ethical principles, nor Cohen’s reservations about Kant’s supposedly rightful state. If morality ‘really is of this world’ (394), then social relations and legal arrangements must recognise people as ends in themselves. At the same time, I have raised one reservation and hinted at another. Cohen offers his ideal legal person as a reference point for concrete historical struggles rather than an institutional blueprint. Still, any responsible ideal must also reckon with human finitude – manifest here as the limits, conflicts and constraints inherent in actual legal persons. Second, while Cohen is right to challenge Kant’s acceptance of material inequalities, Cohen is also too quick to focus on them. He gestures to other aspects of law, but never explains their role. Most notably, he omits the role of legal persons in capitalism, despite his own focus on legal personhood.
In response, I would like to return to Kant’s legal ‘formalism’,Footnote 10 and to highlight its materialism. Of course, ‘formal’ sounds abstract and immaterial. But Kant’s starting point is external freedom, whose contrary is the most brutal material factor of all: coercion. Kant could hardly foresee modern capitalism, neither its legal forms nor its material exploitations. If he had, perhaps he would have spoken for privilege rather than principle. Nonetheless, I suggest that Kant – or at least, the first principles of his legal thought – can help us make out the unjust form of capitalism, which structures the material inequalities that Cohen decries. Following Cohen’s own pointers, I will suggest that this form has three central elements: money, employment, and a specifically capitalist form of legal personhood. Together, these betray freedom and equality, legally and coercively.
4. The form of capitalism: money, workers, capital
It was a frequent refrain among the neo-Kantian socialists, that the labourer is reduced to a thing – a mere means, with use value only. As Cohen stresses, this contradicts the basic principle of Kant’s ethics: that each person is an end in themselves (e.g. 320).
Alongside this devaluation of the person, capital personifies property. As Cohen puts it:
With capital, property loses the impersonal character otherwise attached to it, which presents the contradiction between thing and person. Capital seems no longer to be merely a thing; it becomes a person, for it acts like a person. Just as the owner employs the worker to produce labour by providing him with his working existence, so capital is apparently capable of such production in and of itself. (609)
But Cohen does not explain how this happens. He stresses unequal rights to productive resources and unequal bargaining positions, as well as rights to bequeath and inherit property. He emphasises the role of contract – above all, contracts of employment (more below). He stresses the role of money (more in a moment). Further, he claims that law plays an essential role in all of this. Still, to go beyond moral denunciation, we should say more about the role that law plays.
4.1 Money
Cohen recognises that money is essential to capitalism. It is, he says, the perfect ‘idealisation’ of property (608). Cohen claims that this idealisation would be impossible without law (609). But he does not explain why. To justify the claim, we need to know how law enters the matter.
Kant’s own comments are also brief, indeed flawed.Footnote 11 But they contain an insight missed by Cohen. Kant acknowledges the everyday function of money – the ‘universal means by which human beings exchange their industriousness with one another’ (6:287; also 7:274). But he also points to money’s legal and political basis: the state’s power to tax (6:288).Footnote 12
In this, Kant prefigures modern state theories of money. Money has currency because it is units of ‘sovereign debt’.Footnote 13 Every British bank note illustrates: the sovereign ‘promise[s] to pay the bearer on demand the sum of…’. In other words, if you own a £10 note, the state has a £10 debt toward you. This claim is hard to make sense of.Footnote 14 It certainly sounds peculiar if you focus on the everyday features of money: it is scarce (608); it facilitates market exchanges by placing commensurable values on disparate forms of property and service (thus Cohen’s ‘idealisation’).
As the name suggests, a state theory of money draws attention to the legal-coercive structure underlying these familiar features. To describe money as ‘sovereign debt’ suggests that the state owes a debt to its citizens – or at least, to citizens who hold money. This may sound no more tangible than the state as social contract. However, it has a key practical implication.
When the state imposes taxes, it places citizens in debt. How should citizens meet this newly coined debt? In just the same way that two IOUs cancel each other out: you return tokens of the debt that the state owes to you. If you owe £10 in tax, you must hand over a £10 note or its banked equivalent. By accepting this, the state fulfils its ‘promise to pay the bearer’. It makes good on the ‘public’ element in the famous statement on every dollar bill: ‘legal tender for all debts, public and private’.Footnote 15
To spell this out a bit more, recall Robert Nozick’s notorious claim: ‘Taxation of earnings from labour is on a par with forced labour’ (Reference Nozick1974: 169). It is easy to be distracted by the (im)moral hyperbole: how can income tax be ‘on a par’ with labour under whip or chain? I want to point out a deeper problem in the conceptual plumbing. To explain why money is valuable, we must appreciate its relation to taxation. Citizens value money, not just because it is useful for market exchanges, but because they must pay taxes. Nozick’s hyperbole about taxation mirrors his libertarian fantasy that money is a matter of voluntary coordinationFootnote 16 – a fantasy belied by the sovereign imprimatur on all currencies.
But then notice: those who command financial resources – and especially those who can save and speculate, or in other words, the wealthy who command ‘means in excess of [their] own needs’ (6:453) – have the legal power to settle ‘all debts, public and private’. Rather than ‘exchanging their [own] industriousness’ with others, they have legal rights to deploy other people’s industriousness: that is, to use them as means. It would be Nozickian hyperbole to say that they can force others to labour. Still, if we bring two further factors into sight, we can see how capitalism’s vaunted ‘free labour’ may come closer to ‘wage slavery’ – that is, to workers’ use merely as means. Not, pace Nozick, because governments tax wages, but because law invests wealth with governmental power.
This happens through two factors that Cohen draws attention to, but does not examine. The first is the legal structure of employment; the second is the legal personhood of the joint-stock company or, as we would now say, the business corporation.Footnote 17
4.2 Workers as mere means
Cohen claims that ‘the meaning of the employment contract is domination [Herrschaft]’ (605). But he says little to explain what makes this domination possible. He repeats a familiar Marxist theme: the parties’ unequal control of the means of production. This is certainly important. Again, however, it hardly justifies Cohen’s claims about the role of law. I think Kant can help us see further.
Having made the argument elsewhere (Williams Reference Williams2024, Reference Williams2025: §4.4), I will summarise briefly. Employment contracts lack the formal equality of contracts for goods and services. On top of anything specified in a written contract, law imposes two duties on workers (Deakin & Wilkinson Reference Deakin and Wilkinson2005: 108). The first is loyalty: employees should serve their employer’s purposes. The second is obedience: employees should follow their employer’s instructions.
As such, employment contracts resemble a key relation of Kantian domestic right.Footnote 18 The master-servant relation involves a contract, which stipulates ‘matter’ such as duration or rate of pay. But it rests on a distinct form of right. In Kant’s well-known terms, this form consists in rights to a person ‘like rights to a thing’ (auf dingliche Art: 6:276ff) Accordingly, Kant assumes the master’s voice: ‘I get a right to make arrangements about him [i.e. the servant]’ (6:259). The servant is meant to serve ‘the good of the household’ (6:360 – or in other words, loyalty); the master instructs the servant as to the activities required (obedience). The Hausherr is Herr – i.e. master – not just because he owns the house, but also because he, and not the servant, defines the good of the household; he, and not the servant, directs the servant’s activities to this end.
Employment in a factory or organisation differs from domestic service in many ways. But the legal form is analogous. Subordination reflects not just property relations, but also the legal authority built into employment law. This is no coincidence – as labour historians have detailed, this law descended from master-servant law.Footnote 19 Like heads of households, employers are legally authorised to dictate purposes and duties. Within a certain role, employers determine how workers use their time and active powers; who they must work with; what access to resources and information they shall have; and countless further specifics. The employer’s ongoing authority gives them valuable power to navigate an uncertain world; the worker’s formal subjugation is the employer’s formal gain.
This is the difference from contracts for goods or services. In these contracts, the parties decide at the start what each owes the other. Of course, many implications are hard to anticipate. But this problem is formally symmetrical; it inheres not in the form but in the matter, since every specific commitment is bound to raise problems of foresight. In a relationship of servitude or employment, by contrast, the employer’s authority creates a formal asymmetry, which in turn creates a formal barrier to the worker’s ‘informed consent’. As a point of logic, I cannot prospectively consent to something that another person subsequently decides.Footnote 20
In Section 2, I noted that Cohen speaks not just of domination and instrumentalisation. He also suggests that the worker’s ethical personality is torn in two:
[T]he controller of property becomes the controller of the person bound by this property. Hence the [worker’s] act is isolated. Isolation does not damage a thing; but it inexorably tears apart [reisst] the whole person, the unity of the person. (605)
The legal structure of employment helps show how this can be. It is not just the property ownership that Cohen stresses, or related factors familiar from Marxist accounts – ‘private ownership of the means of production’, the small number of capitalists as against potential workers, and the resulting differences in bargaining power and access to productive assets. The legal form of contracts for goods and services allows the parties to specify particular duties; the legal form of employment imposes an open-ended command of obedience and loyalty. Hence the employer gains far more than individual actions (607); they gain the worker’s active powers and purposiveness. Cohen’s claim that employment ‘tears apart’ the worker is sharpened if we appreciate this formal inequality.
Kant says little to explain why outright instrumentalisation will not follow. Servants are not slaves: the legal form is not property in things. But slavery is not the only form of instrumentalisation. Kant is explicit that auf dingliche Art means rights to use as means (6:359). Worse, as noted, he even adds formal disempowerments – not least, servants’ passive citizenship.
Of course, by contrast with slavery, it was not merely facetious to speak of ‘free labour’, as was standard among the economists of Cohen’s day. As a matter of legal form, even passive citizens have legal rightsFootnote 21 ; workers agreed to jobs and could leave them. At the same time, it was not mere hyperbole to speak of ‘wage slavery’. Neither legal nor economic factors mitigated employers’ legal and economic power to dominate and exploit.
To overcome this, one sort of response looks to material factors, such as the relative bargaining position of the parties.Footnote 22 For employees with special skills or social privileges, there may be nothing dominating about employment contracts. Workers with fewer social resources may also be able to resist attempts to make them into mere means, if they have other good options for work or if there are decent social security provisions.
Still, once we notice the formal inequality, a Kantian view must demand formal remedies. Formally speaking, this is the only way to find reciprocity in a formally unequal relationship. These must be both individual and collective.
In terms of the individual employment relation: if the worker is bound to loyalty and obedience, then the employer must be bound to stringent duties, too. Under nineteenth century capitalism, just as eighteenth century master-servant law, employers were burdened with few duties.Footnote 23 Cohen was familiar with social-democratic and socialist campaigning for some of these protections, if not the eventual successes: duties regarding health and safety, rules on working time, sick and holiday pay, minimum rates of pay, and much more.Footnote 24 Where these provisions are absent, the realities for many workers remain as brutal as in nineteenth century factories – in sweatshops across the world, in Chinese factory towns, or even in conditions for lower-ranking employees in much of the US.Footnote 25
At the collective level, trade unions or professional associations draw on the power of legal personhood to enable collective action and representation, and to press for the sort of worker rights just mentioned.Footnote 26 This remedy gains in importance, and even becomes compelling, once we consider the meaning of ‘capital’ and its legal form. Almost always, employers are themselves legal persons – that is, corporations of some form.
4.3 Capital as an end in itself, or at least, as a legal person
I started this article with Cohen’s words: ‘Capital demands that we dissolve the difference between thing and person as a [mere] prejudice’ (609). As with money and employment, Cohen sometimes says that law is involved in this grotesque dissolution. Thus my second epigraph: ‘law must build the catacombs in which [capital] buries itself’ (610). But Cohen does not elucidate his subterranean (in)version of Marx’s superstructure metaphor. He does not explain how law and capital become capable of action. He raises, but does not answer, the million-dollar question: ‘What distinguishes… property from capital?’ (611).
I have noted the central weaknesses of Kant’s approach: his tendency to downplay the domination and dependency enabled by material inequality, and then to heap legal inequality on top. Here, in his references to capital, I think we find the most serious weaknesses of Cohen’s approach: vague gestures to actual law and ideal jurisprudence, alongside careless borrowings from the MarxismFootnote 27 his ethical socialism was meant to improve on.
Cohen relies on a broadly Marxist view of capital, as the means of production as owned by a small number of people (605, 609, 613). Cohen’s framing of this ownership relation is alternately individual and collective. He talks about capitalists as individual people who own capital (612f). He also refers simply to the capitalist class (615f). Both ways of talking are problematic. Individualism is misleading, since we are dealing with a social and economic structure (capitalism). Assuming a class structure begs important questions as to how people maintain such a structure – especially when we consider the competitive aspects of capitalism, whereby individual capitalists (if such there be) or businesses compete in a highly dynamic process. To understand capital and capitalism, we need to see how the individual and collective come together, to form both structure and agency.
I suggest that the most distinctive feature of Cohen’s approach provides the clue: the Genossenschaft (fellowship or cooperative) discussed in Section 3. For Cohen, this legal person is both a source of unity and a form of agency. As noted, it can offer an inspiring ideal. But the same phenomenon can also operate to exclude, divide, and dominate. At the level of the state, Cohen is amply aware of this. He contrasts the actually existing ‘power state’ (Machtstaat) to the ‘rightful state’ (Rechtsstaat). The former reflects the interests of ‘the estates and the [ruling] classes’, not the pure idea of the state (615). This suggests a parallel line of criticism, one that can explain the structure and the agency of capital.
Capitalism has always involved legal arrangements to unite the people who Cohen calls ‘capitalists’. In Cohen’s day, these largely took the form of the Aktiengesellschaft or joint-stock company. With minor variations, this is the shareholder-based business corporation that dominates modern economies.Footnote 28 These arrangements constitute businesses as legal persons: law grants legal personhood to the means of production. It thus creates ‘capital’ in the sense that defines capitalism: not simply monetary or productive assets, however unequally owned. Instead: assets under the dual control of directors and shareholders, hence not individual property. This does not make capital into ‘a worker’ (609, quoted above). It does make capital as a legal person – albeit with a very different form to the Genossenschaft or Rechtsstaat.
Legal personhood means that the company or corporation is a legal actor in its own right, distinct from any named individual (Williams Reference Williams2025: §2.2, §4.2.3). Like the Genossenschaft, it has its own purposes and powers of action, its own property and legal obligations. Unlike the Genossenschaft, it has directors and shareholders, as well as workers – granted the legal structure of employment already discussed. Alongside the ‘joint stock’ – that is, the means of production – workers are the means by which the company produces goods and services and generates financial returns.
Like all socialist authors, Cohen stresses the gap between workers’ wages and the revenue generated (e.g. 610). But note that this value accrues, not to individual property owners (‘capitalists’) but to the legal person of the joint-stock company or business corporation. This legal fact raises distributive questions. Will the company invest, or pay more to workers, or give bonuses to senior managers, or dividends to shareholders? In law, directors have wide discretion – they may use their ‘business judgment’, as the courts have consistently put it (Johnson Reference Johnson2013).
Nonetheless, there is some legal accountability. Not to workers,Footnote 29 of course, but to shareholders. Details vary with time and jurisdiction, but the basic relation is governmental. Shareholders have limited powers to elect company directors; hence directors have good reason to attend to their wishes. And what shareholders mostly want is quite simple: money. There is no other reason to buy shares, unless to gain power over a company – that is, a money-making enterprise. Indeed, if we dig deeper into the legal structure, we see that shareholders have little power except to press their own financial interests, as opposed to other ends that the individuals who own shares might care about – be it the well-being of workers or communities or environments.Footnote 30
In the words of Cohen’s fellow neo-Kantian, Franz Staudinger, the result is an Ausbeutungsgemeinschaft (Reference Staudinger1907: 26): an ‘exploitation community’ of directors and shareholders. Both groups have an obvious interest in increasing the company’s size and market share, and in reaping some of its returns.Footnote 31 More than this, both groups occupy legal roles that belong to the constitution of these legal persons. The directors’ purpose is to keep the company going; this requires profits which make money for shareholders. Shareholders’ purpose is to elect directors who will ensure that money flows. The resulting ‘community’ relies on the use of workers. But it is set against their interests in anything more than bare life (perhaps not even this, if replacing workers is easy). It makes capital into a legal actor determined to grow in size and generate on-going profits; it makes workers into a means to this end.
In other words, Cohen’s claim that capital would ‘dissolve the difference between thing and person as a prejudice’ (609) lacks an explanatory basis. But his own emphasis on legal personhood provides the key. Law dissolves this distinction by creating capital (Pistor Reference Pistor2019). Only law can create legal persons who are not people; only law can grant a business corporation its existence and rights to act; only law can personify the profit motive.
Kant’s formal division of rights makes the point vivid. ‘Innate right’ does not apply. The company is not an actual human being: an ethical personality with, for example, rights against assault or confinement, or rights to civic participation. But the company commands all the ‘acquired rights’ of Kant’s typology.
First, corporations own property. This permits a formal account of capital.Footnote 32 Capital is a company’s ‘joint stock’ or assets, dedicated to producing goods and services for sale, so as to maintain and augment those assets, and to generate an income stream that benefits directors and shareholders.
Second, corporations make contracts – above all, to buy and sell goods and services. Hence the threefold process that Cohen mentions: mere things become goods for exchange so that they can take the abstract form of money (608). Beyond all material activities, the lifeblood of a company is money. The inability to deliver on contractual and other obligations is bankruptcy – the company’s equivalent of death.
Third, corporations employ people. The company gains rights over workers like rights over things. If the company is left to its own devices, we should expect ‘auf dingliche Art’ to mean what it says, rather than what Kant hoped. Workers are not things for purchase and sale – that is, slaves. But as Cohen argues, workers are liable to be used in the same way as the physical means of production – that is, merely as means.
On the one hand, the company has few legal obligations to do anything more – at least if we stay with the nineteenth century context, pending social-democratic developments in employment law and trade unionism. It must only respect workers’ own legal right to walk away.
On the other, only shareholders and directors have governmental rights within the company. They have no interest in the company doing anything by workers except to use them as efficiently as possible. Of course, some commentators hope for the best – the old canards of ‘enlightened self-interest’ and ‘win-win’ solutions. But the underlying form is harsher. The company’s legal personhood entwines the interests of directors and shareholders. Those interests are directly opposed to workers’ interests – in better pay and conditions, or a meaningful stake in the company’s governance.
Here is capitalism’s central legal actor. It commands rights to property and contract and employment. It deals in the money that workers need, that directors are duty-bound to generate, and that shareholders have governmental power to extract.
5. Conclusion
Legal personhood is Janus-faced. One face promises unity and fellowship. Kant argues that the ‘moral person’ of the republican state can unite people as equals (6:343; cf. 236). Cohen holds that the Genossenschaft and the democratic state can unite people and form them as free and equal fellows. As such, the legal person offers a model for ethical thinking and a point of reference for political struggles. It may even suggest a path from capitalism to socialism.
However, two difficulties belong to artificial legal personhood. Necessarily, every actual legal person excludes many people; every actual legal person involves a scheme of government. And as Kant insists, ‘legal’ means: the right to compel. As a result, legal personhood can betray moral and political hopes, in at least three ways.
Internally, a legal person can be hierarchical and dominating. Like many others, Cohen finds fault with Kant. In Kant’s republic, some people are merely ‘passive citizens’ (6:314). When he consigns women and many workers to the status of mere ‘parts’ rather than active ‘members’, Kant chooses hierarchy over equality (Vrousalis Reference Vrousalis2022). Cohen also finds fault with existing states: thus his reference to the ‘power state’, as opposed to the rightful state (615).
Externally, two further problems arise, in opposing directions. Every actual legal person excludes many people. So it may harm or dominate outsiders, even if it serves its members well. Alternatively, a legal person may be a tool for outsiders. The opposite of republican self-rule, for example, is a colonised state ruled by a metropole. Again, Cohen’s ‘power state’ has been involved in both.
All these betrayals also appear in the legal person that Cohen ignores: the joint-stock company or business corporation.Footnote 33
Shareholders buy governmental rights, usually a mark of insider status. But their stake is only financial; they can sell at any point. For them, the company is merely a means to financial gains. Lacking personal involvement, we can hardly call them ‘Genossen’ – fellows or members or insiders.
Are workers insiders? They often spend large parts of their lives within the company; only their agency and activity enable the legal person to use and develop its means of production. But just like the property relations involved, the legal structure of employment subjugates workers; they are replaceable and disposable. The legal structure of the business corporation grants rights to use workers without combining with them, without granting them participatory or membership or governance rights. Despite their personal involvement, we can hardly call them Genossen, either. In Cohen’s powerful image, workers may be torn in two (605, as quoted above): their ethical personality betrayed, their active powers degraded to mere means.
Alongside legal personhood and employment, money structures these disconcerting inside/outside relations. Recall Kant’s words: money is the ‘universal means by which human beings exchange their industriousness with one another’ (6:287). Workers certainly give their industriousness: that is, they act as means. Mostly, they have no other way to gain enough money to live (or at least, to live decently). By contrast, every shareholder has more wealth than they need to get by. Beyond this material inequality, however, I have followed Kant’s formalism to point out the legal structure of capital, and the material, coercive aspect of this form. Shareholders enjoy a legal privilege by which their money works for them. As Cohen says in the context of my first epigraph, ‘capital itself becomes a worker’, generating interest (609) – that is, more money. It takes a lot of work – and law, and coercion – to sustain this preposterous privilege. Capital only ‘becomes a worker’ because it is formed as a legal person that grants rights to wealth and denies rights to workers. To the extent that workers lack legal protections or social advantages, they become ‘mere means’.
The name for the resulting structure is capitalism. Giving legal personality to property creates capital. These legal persons invest monetary wealth with legal agency and fail to incorporate workers as ends in themselves. They betray the ethical principles that Kant and Cohen share, just as they betray Kant’s principle of equal external freedom.Footnote 34
Open access
