2.1 Kant on Property

Kant’s political philosophy is an attempt to deduce the idea of the liberal state from a single, pre-institutional, ‘innate’ right to freedom as independence (Unabhängigkeit) from the constraining choices of others (DR 6: 237).Footnote ² A liberal state, for Kant, minimally includes the panoply of coercively enforceable individual rights, from the basic liberties (freedom of speech, conscience), through property rights in external things, to a welfare state (providing equality of opportunity and poverty relief). Kant’s deduction of the liberal state has two movements. The first movement, Private Right, proceeds from a premiss about the innate rightFootnote ³ of independence, to the justification of a provisional set of property rights (DR 6: 261 f.). This is a move from innate right to private property as its condition of possibility. But Kant also thinks that in the absence of a universally legislating public power, these private rights are indeterminate, nonbinding, and unenforceable.Footnote ⁴ The second movement, Public Right, is designed to correct these defects in the state of nature: enter omnilateral rule by a ‘powerful and united will’ that creates, interprets, and enforces law (DR 6:312ff). This establishes a ‘rightful condition’ under universal laws of right.Footnote ⁵ I will discuss the first movement in this section and the second in the next.

The first movement of DR is a movement from the innate right of humanity to private right.Footnote ⁶ In DR, Kant defines private property as follows:

This section disentangles these three moments of Kant’s concept of property: possession, capacity, and will, respectively. Each moment justifies a particular relation between a person and a thing. Kant establishes the first moment through his ‘Universal Principle of Right’, the second through a ‘postulate of practical reason’, and the third through the idea of ‘provisional right’. Taken together, these three moments make up his theory of property. This theory not only rejects Locke’s idea of a natural right to private property but also offers a structurally egalitarian, broadly Rousseauvian, account of property.

I begin with what Kant calls ‘possession … (according to the law of external freedom)’. Possession, in general, is the ‘subjective condition of the possibility of the use of an object’ (DR 6: 245). My possession of a thing is rightful if anyone’s interfering with it without my consent would wrong me. That wrong, for Kant, is tantamount to a violation of my freedom as independence from the constraining choices of others (DR 6: 237).Footnote ⁷ This innate right to independence, Kant thinks, follows analytically from the

Right, for Kant, is connected with the ‘authorization to use coercion’ (6: 231). So to have rightful possession of some external object is to be able to coercively exclude others, by right, from using it. Note that Kant thinks that the UPR only establishes innate, not acquired, right: the possibility of rightful occupation of spaces, of rightful performance of basic actions, and of the holding of things. Consider what this means for possession. Suppose that, by the UPR, I can rightfully raise my arm. This means that I can also rightfully raise an apple by holding it. This set of authorizations, Kant thinks, is all contained in my innate right to independence. The problem is that innate right does not explain how I can rightfully perform the extended action of using the apple without holding it. Extending authorization from held to unheld things is the task of Kant’s theory of property.

To clarify this crucial distinction, Kant draws from his Critical discussion of the analytic and the synthetic a priori. Consider your taking an apple from my hand without my consent. This wrongs me – it constitutes battery. The wrong, Kant thinks, follows analytically from my innate right. Contrast your taking, without my consent, an apple which I own but I am not currently holding. This also wrongs me – it constitutes theft. But this violation does not follow analytically from my innate right, says Kant. Once it lies beyond my body, the representation of the apple as mine becomes a representation distinct from the representation of the apple as belonging to my body. Kant thinks that this makes the former representation into an analytically distinct idea. So if these two distinct ideas (what he calls ‘intelligible’ and ‘physical’ possession) are to be connected a priori, they must be connected synthetically. This, Kant says, requires a ‘deduction’ (DR 6: 250). The deduction must show that I can rightfully use the apple without holding it, or, equivalently, that there are freedom-consistent, coercively exclusionary apple-uses. So far, most Kant interpreters agree. Things get considerably more complicated once we turn to the details of Kant’s deduction.Footnote ⁸

The second moment in Kant’s theory of property attempts to deduce a rightful capacity to use external things. Kant thinks this is established by the

That the Postulate requires a proof, or at least an elaboration, transpires from the structure of Kant’s argument (DR 6:250). I will assume, following recent Kant scholarship,Footnote ⁹ that the Postulate seeks to establish dominion, that is, a set of complete and enforceable property rights, including use, benefit, and alienation of external objects of choice (e.g. DR 6: 270). Even so, Kant’s proof of the Postulate is obscureFootnote ¹⁰ and seems question-begging.Footnote ¹¹ Here is one non-question-begging interpretation. Suppose I set the end of making a fruit salad. I am rightfully holding the apple in one hand and the pear in the other. Suppose that, in laying down the apple to chop the pear, I thereby lose title to the apple. This makes the rightful pursuit of the end of a fruit salad impossible, placing it into what Kant calls res nullius. But this leads to an inconsistency. Recall that independence requires that I be able to set and pursue ends independently of others. This includes fruit-salad-type ends. So I must be able to rightfully lay down the apple. Rightful ‘intelligible’ possession of the apple is therefore possible.

If this reconstruction is correct, then Kant’s Postulate establishes that there exist rightful uses of external things. But we are not much the wiser, for this says nothing about who can own what. And this is what makes Kant’s next move indispensable to his defence of private property. Kant writes that the Postulate

On the present reading, the permissive law turns the Postulate into an agential authorization: an authorization to those individuals with the relevant capacities to enlist external things in their pursuit of their own ends.Footnote ¹³ This makes certain things ownable. How do we go from ownability to ownership?

The final and crucial step in Kant’s argument for private property is the possibility of a united will, or ‘what I will should be mine (in accordance with the idea of a possibly united will)’. This is a Rousseauvian general-will-type authorization applied to external possession. Throughout DR, Kant insists on a distinction between

1. (a) ‘common ownership’ or the ‘original community of land’ (Gemeinbesitz) and

2. (b) ‘private community’ of private owners joining their properties together to form a ‘collective possession (Gesammtbesitz)’ (DR 6: 251; 267).

Kant criticizes the latter notion as presupposing contract and therefore private property. The former notion, by contrast, is an idea that ‘has objective (rightfully practical) reality’ and is connected a priori with the idea of the civil condition – the just state. Kant needs this connection with the civil condition because he explicitly denies that there are innate or natural rights to property. He also mocks any Lockean appeal to labour-mixing as the ground of property acquisition (DR 6: 260 f). If Kant is right, then nothing can be nobody’s in the state of nature. That is, if there are no natural rights to property, then nothing can belong to nobody by natural right, just as nothing can belong to somebody. But Kant does think that we can make sense of the idea of external things belonging jointly to all – things that can be used only with everyone else’s permission. He reasons that to make something mine is to withdraw it from the community by the hypothetical permission of all.Footnote ¹⁴

Baynes sums up this argument as follows:

There is some consensus among Kant scholars that this ‘obligation’ is a natural duty to support just institutions. On this interpretation, the structure of Kantian property rights must anticipate the structure of the just state. Rafeeq Hasan, for example, argues that in the ‘state of nature property claims do not wrong others, but anticipate a condition in which the authority to make such claims can no longer be unilaterally determined’ (Hasan Reference Hasan2017, 850; see also Guyer Reference Guyer1996). It follows that the authority to make private property claims in the state of nature is dependent on the structure of the rightful condition which the state of nature ought to anticipate.

Now return to this passage:

I have broken this down as follows. For any external object of choice, x, the ‘law of external freedom’ says that x is ownable. This moment of ‘possession’ says nothing about who owns what, not even provisionally. This indeterminacy is resolved by the Postulate, read as a permissive law. The Postulate says that there are agents who can own x. These agents receive an authorization, or normative power they would not otherwise have, which puts others under an obligation not to interfere with their exclusive uses of x. But this is not to say who owns what, even provisionally. To establish provisional ownership, I must ‘will’ that x be mine, subject to that willing anticipating a rightful condition. In anticipating, I am to ask whether everyone could hypothetically permit me to withdraw x from communal use. Then, and only then, is x mine. So Kant not only rejects natural rights to private property but also derives property from an idea of original communal possession. Kant de-Lockeanizes property theory by Rousseauvifying it.

So much for summarizing Kant’s theory of property. I now discuss the second movement of Kant’s argument for the liberal state, from Private to Public Right. Its central feature is a theory of citizenship.

2.2 Kant on Citizenship

Citizenship plays a crucial role in Kant’s second justificatory step for a liberal state (see TP 8: 290–297 and DR 6: 314–316). In DR §46, Kant again opens with a Rousseauvian argument about the unity of sovereign power. He argues that the state’s legislative authority belongs to ‘the concurring and united will of all,’ where ‘each decides the same thing for all and all for each’. He adds that, since one ‘can never do wrong in what he decides upon with regard to himself’, it follows that the legislation of such a united will must be consistent with the demands of right. But who are Kant’s co-legislators?

Kant’s co-legislating citizen has three normative attributes (rechtliche … Attribute): heFootnote ¹⁵ enjoys freedom (Freiheit), ‘the attribute of obeying no other law than that to which he has given his consent’; equality (Gleichheit), the attribute of ‘not recognizing among the people any superior with the moral capacity to bind him as a matter of right in a way that he could not in turn bind the other’; and civil self-sufficiency (Selbständigkeit),Footnote ¹⁶ the attribute of ‘owing his existence and preservation to his own rights and powers as a member of the commonwealth, not to the choice of another among the people’ (DR 6:314). Freedom, equality, and civil self-sufficiency, then, are the ideal public expression of Kantian independence. What remains to be explained is what the third attribute adds to the first two.

A preliminary explanation follows Kant’s account of the franchise. After suggesting that ‘the only qualification for being a citizen’ is ‘being fit to vote’, Kant adds that this

Kant’s passive citizensFootnote ¹⁷ enjoy freedom and equality if they possess legal standing as parts (≠ members) of the commonwealth and if they can work their ‘way up from this passive condition to an active one’ (DR 6: 315). Both conditions are satisfied, Kant implies, in the case of apprentices, domestic servants, minors, and women, albeit passively. Minors, for example, enjoy a right ‘to the care of their parents until they are able to look after themselves’ (DR 6: 280), a right entailed by their innate right to independence. It follows that enjoyment of innate right is compatible with de facto lack of control over its active exercise. On the interpretation I will present, Kant purports to justify this lack of control by the dependent nature of the right-holder’s labour. Civil self-sufficiency, on this view, establishes a presumption for independent use of your rightful powers, including your productive powers. If I am right, then self-legislating citizens in the Kantian state must enjoy interdependent independence, that is, the ability to make their own (productive) choices independently of the private choices of others. I now explain the relevance of this interpretation.

My interpretation of civil self-sufficiency is key to understanding Kant’s disenfranchisement not only of apprentices, domestic servants, women, and minors – all of whom are subject to fiduciary ‘status’ relationships – but also of wage labourers and landless farmers – who are not. In the DR passage immediately following the definition of civil self-sufficiency (see the quotation above), Kant writes:

The Indian travelling blacksmithFootnote ¹⁸ contrasts with those subject to status relations, in that she is merely contractually bound to ‘let and hire’ her productive powers to her employer(s) (DR 6: 285).Footnote ¹⁹ But the blacksmith’s position is similar to those subject to status relationships in that she, too, lacks civil self-sufficiency. According to Kant, the Indian blacksmith enjoys freedom and equality insofar as she could come to own enough iron and thereby come to bind her employer to as much as the employer actually binds her. This is the significance of Kant’s equal opportunity proviso, that ‘anyone can work his way up from this passive condition to an active one’ (DR 6: 315). But the blacksmith presently lacks iron ownership, which means she lacks independent use of her productive powers. In order to exercise these powers, she must get permission from the iron owner(s) to use the iron they own, which means she must put her powers at their disposal. She therefore lacks civil self-sufficiency. Like the minor, the Indian blacksmith remains subject ‘to the choice of another among the people’ (DR 6: 314),Footnote ²⁰ quite independently of a supervening status relationship.

The rest of this section argues for three claims. First, Selbständigkeit is the idea that free and equal people can co-legislate if and only if they have independent use of their rightful powers as members of the commonwealth. Absent such independence, dependent producers cannot legislate on behalf of their own agency. Second, the relationship between citizenship and property, for Kant, does not have the educational, psychological, or anti-corruption significance it has for Sieyès, for the French constitution of 1791, or for contemporary neo-Roman interpretations of citizenship.Footnote ²¹ Rather, property is relevant only insofar as it gives citizens unsubjected formal discretion over the exercise of their rightful powers. By enjoying such discretion, Kant’s interdependent citizens independently facilitate the conditions of their mutual independence. Third, Kant takes the empirical form of independence in the modern state to be independent commodity production.Footnote ²² That is, Kant’s citizen uses her productive powers to produce external means for others, without having to make her powers into their means. So, in depending only on the content of exchange relationships – what she produces with her own powers – each member of the commonwealth depends only on her own ‘ability, industry, and good fortune’ (TP 8:296). After introducing the interdependent independence interpretation, I will defend each of these claims in what follows.

2.3 Citizenship and Selbständigkeit

Here’s a reconstruction of DR §46, Kant’s main discussion of citizenship. This reconstruction must explain why, for Kant, the franchise follows economic independence and not vice versa. More precisely, economic dependence suffices to disenfranchise dependents in fiduciary status relationships – e.g. the minor and the domestic servant – and non-fiduciary contractual relationships – e.g. the Indian blacksmith. The structure of this interpretation is as follows (Kant on the left column, my reconstruction on the right):

In a nutshell, Kant could not have thought that dependents should be disenfranshised, unless he thought that, in legislating, they would be making ‘arrangements about another’. Economically dependent citizens, for example, would be legislating not for themselves as independent citizens, but only for their propertied masters as citizens. But then, Kant reasons, those dependents cannot possibly legislate rightfully. The interpretative challenge consists in offering an account of Selbständigkeit broad enough to explain why the dependent cannot legislate (claim (5)) and why that suffices for their disenfranchisement (claim (7)).

The only interpretation that makes these inferences exegetically palatable, I now argue, is the interdependent independence interpretation of civil self-sufficiency (hereafter: interdependent independence).Footnote ²³ This interpretation establishes a presumption in favour of enjoying unsubjected exercise of your rights and powers, including your productive powers,Footnote ²⁴ as a member of the commonwealth. Interdependent independence sees legitimate public power as regulating the compossible choice sets of interdependent choosers. These choosers (re)produce their own lives by exercising their commonwealth-conferred rights and (productive) powers independently of the private choices of others. Here is some further textual evidence for this reconstruction.

DR §46 calls the active citizen a ‘member’ (Glied). This contrasts with the passive citizen, a mere ‘part’ (Teil) of the state. Kant suggests that the choices of active citizens form a community (Gemeinschaft) and that the concomitant status of passive citizens is ‘as it were, mere inherence’ (Inhärenz). These terms are explicitly connected in Kant’s Critical philosophy. In the Critique of Pure Reason Kant contrasts the reciprocal determination of ‘coordinated relations’ (as in members of a body) with the ‘subordinated’ relation of effect to cause. The former relation, unlike the latter, constitutes a community (Gemeinschaft) between ‘agent and patient’ (CPR A80/B106). Inherence, on the other hand, is contrasted with subsistence, in that only subsistent entities – substances – have causal powers. It follows that only substances can partake of community: they are independent as substances but interdependent as reciprocally dependent on other substances.Footnote ²⁵

Kant applies these categories explicitly to politics in the Critique of the Power of Judgment. In a footnote discussing the French Revolution, he applies the idea of organization to the ‘body politic’ and suggests that ‘in such a whole, each member (Glied) should certainly be not merely a means but at the same time an end … whose position and function should be determined by the idea of the whole’ (CPJ 5: 375). On the present interpretation, political community is a form of reciprocal interdependence through the social division of labour. The ‘body politic’ is the united will of independent ‘substances’, that is, of free and equal people in independent possession and use of their rightful powers.

Now, the criteria for voter-fitness, on this view, are entailed by three a priori premisses (1, 2, 4). The first two (1, 2) reflect Kant’s views of the appropriate object of legislation, namely the self-legislating will herself. Kant then identifies that will with free and equal people in independent possession and exercise of their rightful powers (4). It follows that the domestic servant, the private tutor, and the Indian blacksmith cannot legislate. For in legislating, they would not be self-legislating: their agency, as ‘inhering’ in the agency of others, cannot be exercised in their own name. As I explain presently, none of these inferences makes sense unless Kant thinks of the rightful condition as a system of cooperative production, in which the subjection of labour capacity to a private will undermines one’s formal standing to legislate universally.

I now offer a unifying explanation for these assertions, based on the idea of interdependent independence.

2.4 The Unifying Explanation: Interdependent Independence

Kant’s disenfranchising exclusions are not, I now argue, about how citizens are disposed to vote. Rather, they are about the nature of the object of universal legislation in the Kantian state. The idea is that only interdependent independents can, in legislating, self-legislate, that is, make arrangements about their own agency.

Suppose the European and the Indian blacksmith share a qualitatively identical productive power of working iron into hammers. Both produce, say, an equal output per hour out of equal inputs; both serve the commonwealth by making it richer by some hammers. Suppose, further, that both enjoy the same set of commonwealth-conferred legal rights to exercise these productive powers, in addition to the right to own iron. They are therefore free and equal before the law. Kant implies that the Indian blacksmith owns no iron, such that, if she is to exercise her productive powers by working iron, she must ask permission from the iron owner(s).Footnote ²⁶ The Indian blacksmith’s ironlessness thereby gives the iron owner(s) discretion over the exercise of the blacksmith’s productive powers – her ability to work iron into hammers. According to Kant, it follows that she lacks Selbständigkeit. That is, her labour process, and therefore the condition of exercise of her commonwealth-conferred legal rights and powers, is de facto unilaterally controlled by others.Footnote ²⁷ This means that her legislative choices do not just affect what belongs to her: the Indian blacksmith is interdependent but not independent. Kant infers that she cannot legislate rightfully. In elaboration of this inference, consider an analogy with musical production.

Suppose you own the musical instrument I need in order to perform as a member of the orchestra. Then you can, by legal right, deprive me of it at will. It follows that I cannot exercise my music-making powers as an orchestra member by playing that instrument, except through your unilateral permission. This way my music-making, even if realized, serves two masters: you and the orchestra. I might, of course, be unaware of that. But that’s irrelevant. Kant’s objection is not that my dependence on you might make me fearful, obsequious or servile in my music-making, inclined to pursue your ends as opposed to mine, and so on. These empirically contingent effects on my psychology are not at the centre of Kant’s democratic theory (which is why they are never mentioned).Footnote ²⁸ Rather, the ground of my exclusion from orchestra membership just is the dependence of the realization of my music-making ends on your unilateral will. That is, even if I were to perform, my performance would still be conditional on your permission. And that would be true whether I know it or not. Such dependence, Kant thinks, normatively disqualifies me from making arrangements about my own music-making. This disqualifies me from occupying the office of music-maker altogether.

Contrast the case where I or the orchestra itself control(s) my instrument. Perhaps the orchestra has public rules meant to serve the goal of music-making: who can occupy the office of violinist, pianist, and cellist, who can use the instruments under what conditions, and so on. Then, in setting the instrument as a means to the realization of my music-making powers, the exercise of these powers is subject to nothing but the end of music-making. I therefore no longer serve two masters; my agency is no longer alienated to you.Footnote ²⁹ Now note that this independence does not preclude interdependence—in fact, the former presupposes the latter. As an orchestra member, I am dependent on you, just as you are on me, for nothing but the mutual exercise of our orchestra-conferred musical powers. That is, we are mutually dependent for doing out our respective parts in the division of musical labour. So the orchestra’s powers are now only the musicians’ interdependent musical powers, independently exercised.

Kantian Selbständigkeit supports this analogy between orchestra and modern state.Footnote ³⁰ Take the European blacksmith, who produces means that others need in order to set and pursue their ends as members of that state. By selling the hammers she makes in return for wigs, cloth, or money, she exercises her productive power without having to ‘alienate’ it to others (TP 8: 295). That way, she produces means that facilitate the unsubjected purposiveness of the wigmaker or the tailor, without making her own purposiveness into their means. The European blacksmith thereby serves ‘no one other than the commonwealth’ (TP 8: 295), the condition of her own independence. The Indian blacksmith, by contrast, serves two masters: the commonwealth and her employer(s). For to alienate your productive powers is to serve a (series of) private master(s), rather than the conditions of your own freedom. So, even if the Indian blacksmith were deemed fit to vote by dint of her independent-mindedness, she would still be bound to making her productive powers into the means of others.Footnote ³¹ By interdependent independence, she would thereby lack the normative power to legislate on behalf of her own agency. Crucially, the nature of this disempowerment is independent of the Indian blacksmith’s psychological dispositions or the contingent content of her material ends. Her disempowerment is only about the formal exercise of her own powers as subject to the private choices of others.Footnote ³²

The orchestra analogy can explain why Kant’s democratic theory is progressive for his time (Maliks Reference Maliks2014; Weinrib Reference Weinrib2008). Suppose that orchestra membership can be expanded in either of two ways: the orchestra itself provides more instruments to more musicians or more musicians bring their own instruments. Both arrangements allow for the free, unsubjected exercise of this extended membership’s music-making powers.Footnote ³³ In a similar vein, Kant proposes to extend independence – and therefore the franchise – by extending public poverty relief to ‘those who are unable to maintain themselves’ (DR 6: 326);Footnote ³⁴ by enforcing the opportunity of anyone to ‘work his way up from his active condition into an active one’ (DR 6: 315);Footnote ³⁵ and by giving every active citizen an equal vote. This egalitarianism of voting shares is most explicit in Kant’s discussion of the estates. Having raised the question ‘how it came about that many human beings who could otherwise have acquired a lasting status of possession were thereby reduced merely to serving [the landowner] in order to be able to live’ (TP 8: 296, my emphasis), Kant explains that enfranchisement should depend only on the ‘status of possession, not … the size of … possessions.’ So possession matters because and insofar as it evinces independent use of one’s powers, including her ability to feed and protect herself by right.Footnote ³⁶ I will return to this in Section 2.5.

To sum up the argument so far: according to interdependent independence, an active citizen is someone whose exercise of her state-conferred rights and powers is not subject to another (set of) non-state agent(s), just as an independent musician is someone whose exercise of her orchestra-conferred musical powers is not subject to another (set of) non-orchestra agent(s). Kant’s schoolteacher and European blacksmith, for example, enjoy both (a) a set of commonwealth-conferred political powers and (b) unsubjected formal discretion over their conditions of exercise.Footnote ³⁷ Kant’s private tutor and Indian blacksmith, by contrast, enjoy (a) but not (b). Both must get permission to use their powers, which trivially entails that they lack discretion over their exercise. So the private tutor and the Indian blacksmith lack civil self-sufficiency. This, Kant thinks, suffices to disenfranchise them: only the independent producers can, in legislating, legislate on behalf of their own rightful powers.

Contemporary liberalism solves this problem by severing the link between franchise and independence, thereby enfranchising wage labour. But this move, which proceeds by denying the presupposition in premiss (4), misunderstands the spirit of Kant’s argument. Kant wants to make independence – unsubjected exercise of one’s interdependent rightful powers – central to political life in the modern state. The rightful condition must therefore reflect more than free consent by juridical equals. This is what Selbständigkeit adds to the Freiheit-Gleichheit diptych. And this is also why the Indian blacksmith, lacking Selbständigkeit, cannot self-legislate. The way Kant sees it, her productive dependence means she would not be legislating on behalf of her own freedom, even if she had the right to vote.Footnote ³⁸

I now argue that interdependent independence is the only consistent exegesis of Kantian citizenship. If I am right, then Kant’s state does not just represent a republic of property owners united by state-enforced relations of property, contract, and status.Footnote ³⁹ After all, this bucolic picture is compatible with a republic of non-interdependent subsistence farmers – something resembling a Jeffersonian agrarian republic. Kant’s picture is different. For him, an omnilateral will represents a republic of free citizens enjoying all-round interdependence under conditions of independent use of their rightful powers. It is more port city than agrarian republic.

2.5 Against the Neo-Roman Interpretation

I have, so far, argued that Kant’s active citizens enjoy, in addition to freedom and equality, interdependent independence. This third attribute, Selbständigkeit, establishes a presumption against alien unilateral control over the exercise of their rightful powers, including their productive powers. This section rebuts a broadly neo-Roman interpretation of active citizenship.

Some philosophers take Kant’s disenfranchising exclusions to depend on an empirical claim from cognitive, volitional, or motivational corruption. Luke Davies, for example, argues that, for Kant,

This corruption reading of Kant’s exclusions resonates with Claus Dierksmeier’s idea that ‘having sovereignty over your own business and household would educate you to participate adequately in the affairs of political government and sovereignty’ (Dierksmeier Reference Dierksmeier2002, 50). In a similar vein, Jacob Weinrib cautions against the ‘potentially particularizing’ nature of the will of dependent persons, who cannot, by dint of that dependence, ‘achieve the requisite level of impartiality to contribute to the general will’ (Weinrib Reference Weinrib2008, 11). Sarah Holtman underlines the possibility of ‘fearful decisions of subordinates’ in response to power over them, which she contrasts to ‘the well-informed and skillful reasoning of those who comprehend and are committed to justice’ (Holtman Reference Holtman2004, 100). And Rafeeq Hasan argues that ‘Kant’s ostensible point about the passive citizen is that … he is likely to feel pressured into voting the interests of his boss …’ (Hasan Reference Hasan2017, 922, emphasis added).

These broadly neo-Roman concerns are not Kant’s concerns: his account of civil self-sufficiency is not premissed on a corruption idea. For one, Kant never mentions the psychological dispositions presupposed by such an idea. Furthermore, there is no reason to think that a disposition towards servility (Davies), partiality (Weinrib), or fearfulness (Holtman) will necessarily accompany the dependent.Footnote ⁴⁰ Indeed, the independent might manifest these dispositions and to a greater extent. Finally, the corruption reading cannot explain why dependence makes rightful self-legislation impossible. Consider the following inference from the argument of Section 2.3:

There is no textual evidence in Kant that (8) refers to the empirical dispositions of voters. Kant offers no evidence that extending the franchise would make new voters fearful, obsequious, or motivationally less likely to legislate rightfully.Footnote ⁴¹ If anything, one would expect the opposite. These corruption readings therefore misunderstand the thrust of Kant’s position, which follows validly from certain a priori claims about the proper object of political legislation and the nature of political community (claims 1, 2, 4). As DR§46 makes clear from the outset, Kant is interested in the purely normative features that define independence, as well as the accompanying non-psychological features of agents ‘as members of the commonwealth’.Footnote ⁴²

Consider a further illuminating contrast. Kate Moran has recently traced the neo-Roman ideas underlying the corruption reading back to Sieyès and his discussion of the ‘lackeys of feudalism’ (Moran Reference Moran2021, 9). In that discussion, Sieyès offers an argument for why ‘women, children, and foreigners’ should only be accounted as passive citizens, namely that they ‘contribute nothing to the maintenance of the public establishment’.Footnote ⁴³ Davies draws a suggestive parallel between this argument and an important passage from Kant’s drafts:

Pace Davies,Footnote ⁴⁴ this passage refutes the Sieyès reading. For consider: Kant’s surplus-producing landowners can certainly contribute to taxation or to ‘the maintenance of the public establishment’. So, by Sieyès’ reading of ‘contribution’, these landowners should be enfranchised. Yet Kant disenfranchises them (‘… and yet they are citizens only to the extent …’). Why? Because ‘contribution’, for Kant, means something more specific than it does for Sieyès, namely independent labour contribution: participation in a division of labour among juridical equals with independent use of their (productive) powers. How else are we to explain Kant’s argument that only those landowners who produce a surplus consumed by nonlandowners (those who ‘do not depend on the land’) can be citizens? Since Kant’s landowners do not partake of that interdependence, they are, just for that reason, disenfranchised.

Unlike the neo-Roman reading, interdependent independence illuminates both sets of contrasts implied by the LDPP passage. First, civil self-sufficiency contrasts with the non-civil self-sufficiency of the subsistence farmer and the surplus-producing landowner who only sells to other landowners.Footnote ⁴⁵ Second, it contrasts with the civil non-self-sufficiency of servants and wage labourers. The first group enjoy independence, but are not interdependent; the second group enjoy interdependence but are not independent. The European blacksmith, by contrast, enjoys both. She enjoys interdependence by selling the product of the exercise of her powers in the market. And she enjoys independence by not having to make her productive powers into the means of others. She thereby acts from her ‘own choice in community with others’ (DR 6: 314). In the terms of the Critical philosophy, she is a ‘substance’ in ‘community’ with all other ‘substances’ of the ‘body politic’. This has implications for political obligation. On the present interpretation, political obligation – the obligation to obey the law – is at least partly an obligation of juridical equals to contribute to the maintenance of the state, in proportion to their ability, under conditions of labour independence. Kant does not deny that the dependent producers have political obligations, but he does think that productive independence triggers special obligations.

Now, Kant’s early characterization of citizenship in the loose sheets (see the LDPP passage above) is even more restrictive of the demos than the published account in TP. But it is in keeping with Kant’s vision of the modern state as a system of interdependent independence, featuring cooperative production under a division of social labour. That vision brings the contradictory status of the dependent producer – at once formally free and equal but substantively dependent – into view. This contradiction will only become fully transparent in Marx’s later discovery of the necessary unity between the egalitarianism of exchange and the inegalitarianism of production under capitalism.

I now explain why Selbständigkeit, read as interdependent independence, is exegetically superior to an alternative, property-based interpretation of Kantian citizenship.

2.6 The Mutual Irrelevance between Citizenship and Property

Interdependent independence takes Kant’s concerns with property and propertylessness as wholly derivative from a more fundamental concern with the independent exercise of one’s rightful powers as a member of the commonwealth. It therefore contrasts with a proprietarian interpretation of civil self-sufficiency.Footnote ⁴⁶ According to that interpretation, you enjoy civil self-sufficiency if and only if your making a living depends exclusively on the exercise of your rightful property in external things. The proprietarian interpretation has numerous exegetical advantages. It explains why Kant takes domestic servants, minors, and women to lack self-sufficiency. It also explains Kant’s insistence that property is a precondition of citizenship (TP 8: 295), as well as his belief that propertylessness – poverty – undermines self-sufficiency. This is ostensibly why he argues for state-provided poverty relief (DR 6: 326 f).Footnote ⁴⁷ According to the proprietarian interpretation, the poor lack self-sufficiency because their making a living depends not on their own property but on the property rights and associated powers of the propertied, including their benevolence and goodwill.

Before proceeding, it might be worth explaining why the proprietarian interpretation is not tantamount to libertarianism and, more generally, why Kant is not a libertarian. Libertarianism is the idea that all rights flow from ownership, including self-ownership of one’s body and its powers. In the TP passage elaborating on the conditions of citizenship, Kant writes that ‘being one’s own master (sui juris)’ means ‘having some property (and any craft, fine art, or science can be counted as property) … that is, if he must acquire from others in order to live, he does so only by alienating what is his’ (TP 8: 295). This would seem congenial to a comprehensive account of the possible objects of property, according to which the skilled are propertied, regardless of their ownership of external things. Indeed, if skill is property, then why not also the ability to work – one’s labour power? This reflexive account of the possible objects of property entails libertarianism. Self-owners sell their property, including their own labour power, to other property owners, subject to a well-regulated system of contract and private property rights. Kant does not, however, affirm a reflexive account of property. He is therefore not a libertarian. Consider five reasons why Kant rejects libertarianism.

First, Kant explicitly denies that a person’s body and its powers are ownable. These powers are the objects of innate, not acquired, right. And since all property rights are acquired rights (DR 6: 237), one’s body and its powers are not ownable. Second, ownable objects have no rights (DR 6: 270), but persons do have rights, so persons are not ownable. Third, if mere skill and knowledge are property, then the skilled unemployed who lack rights to external things are not poor. But Kant does not exclude the skilled from poverty relief (DR 6: 326). Fourth, if possession of skill or mere labour power counts as property proper, then the skilled who lack external property – e.g. skilled domestic servants – cannot be counted as passive citizens. But Kant does so count them. Fifth, including skill or labour power as possible objects of property would make Kant’s attempt to justify private property viciously circular. So property, for Kant, is not reflexive, as libertarians think. The proprietarian interpretation is narrower: it concerns property rights to external things only, not reflexive property rights to the owner and her powers.

Despite its numerous exegetical advantages, the proprietarian interpretation is inconsistent with Kant’s texts. The inconsistency arises in empirical examples of two kinds: (i) examples involving property but not civil self-sufficiency and (ii) examples involving self-sufficiency but not property. In respect of (i), consider wage-labour. According to interdependent independence, wage-labour undermines self-sufficiency by making the exercise of the worker’s productive powers dependent on the rights and powers of private employers – the owners of productive assets. This is true both for finding employment and for the exercise of these productive powers once in employment. Yet one can imagine a wage labourer whose income and wealth exceed those of Kant’s civil servant and craftsman, both of whom are deemed to be juridically self-sufficient.Footnote ⁴⁸ So in cases like the wage labourer, the proprietarian interpretation generates false positives.Footnote ⁴⁹ In respect of (ii), consider cases of independent commodity production, in which each producer owns some, but not all, of the means of production she needs in order to set and pursue ends. Each can recover her purposiveness by selling to others the net product of the exercise of her unsubjected powers. In these kinds of cases, Kant allows for self-sufficiency without property in productive assets. So the proprietarian interpretation generates false negatives.

I now elaborate on these two kinds of cases, arguing for the exegetical superiority of interdependent independence over the proprietarian interpretation.

3.1 The Antilibertarian Gambit

Suppose that the British government decides to privatize the roads and sidewalks. And suppose that the putative justification for this policy is not efficiency, but ostensibly equal freedom. Perhaps the British government consists of tough-minded luck egalitarians, who think that everyone should own an equal share of road and sidewalk. What could one object to this policy, on grounds of equal freedom? A devastating objection is that it would make every commuter subject to the private choices of each individual road owner. That is, even if every citizen owned an equal road share, such ownership would make the purposiveness of all subject to the purposes of each. I am not free to move around if my movement depends on your permission, even if you’re inclined to grant it. So this government policy would violate every commuter’s right to equal freedom.

This kind of objection is central to the debate between G.A. Cohen (Reference Cohen1995) and Robert Nozick (Reference Nozick1974) on the nature of property rights. One of Cohen’s dialectical arguments against Nozick’s use of the term ‘liberty’ in Anarchy, State, and Utopia asks whether Lockean self-ownershipFootnote ⁶⁰ protects individual autonomy. To show that it does not, Cohen imagines a system of joint ownership, in which everyone must get permission from others to perform any external action, including the occupation of physical space. Cohen then asks whether such a system would be compatible with the form of self-ownership prized by Nozickian libertarians. Suppose, on the one hand, that self-ownership and joint ownership are compatible. Then self-ownership is no freedom at all, for it is consistent with having to ask permission from others to set and pursue any possible end.Footnote ⁶¹ Suppose, on the other hand, that self-ownership and joint ownership are not compatible. Then capitalist private property draws exactly the same objection as joint ownership – the value of being able to independently pursue ends requires a state more extensive than a minimal state. That is, effective self-ownership requires a state providing poverty relief and redistributing wealth so that nobody requires permission from others to pursue her ends.Footnote ⁶²

This argument has a distinctly antilibertarian flavour, because it starts from a premiss about the external freedom of the poor, not about their needs or welfare. It has been revived in the work of Ernest Weinrib (Reference Weinrib2003) and Arthur Ripstein (Reference Ripstein2009) – the ‘Toronto Kantians’. One of their contributions to the debate has been to add more flesh to Cohen’s barebones argument against libertarianism, taking their cue from Kant’s political philosophy. I begin by elaborating on this social-democratic – ‘liberal’ – strand in Kant exegesis.

3.2 Taking Toronto Roads Seriously

The Toronto Kantians, especially Ripstein, think there is something special about roads. The term ‘roads’ doubles up in meaning here: it signifies roads, but also any all-purpose means (e.g. law, knowledge, health, education). So here is Ripstein, discussing the problem of the landlocked commuter:

As his fellow Toronto Kantian Chris Essert puts it, lacking a road is unlike lacking a boat; roadlessness is unlike boatlessness in that having a boat serves a narrow set of ends (Essert Reference Essert2017). Roadlessness, by contrast, deprives you of access to any end involving proximity to others; it deprives you of end-setting capacity tout court. ‘The solution to this problem is obvious: roads, understood as a system of public rights of way, guaranteeing that there is a path from every piece of privately held land to every other’ (Ripstein Reference Ripstein2009, 248). The contrast between boatlessness and roadlessness can be extended further. Consider homelessness. Homelessness is not just lack of a home, because a homeless person can consistently have access to shelter. So her problem is not lack of housing but lack of an enforceable right to a home (Essert Reference Essert2017). In lacking such a right, she is subject to the permission of some landowner or other, even if they are (individually or collectively) inclined to grant it. So homelessness draws the same objection as roadlessness.

Now, one of the core ambitions of Toronto Kantianism is to show that private property requires, for its justification, a system of omnilateral rights guaranteeing each her independence. Ripstein argues that this presupposes economic redistribution:

It follows that the Kantian state must provide publicly funded poverty relief. Recall that Kant’s argument for the liberal state starts from a premiss about independence: you cannot give yourself laws that would make you into the servant of another. The Toronto interpretation adds that a system of laws which countenances poverty – propertylessness – would turn the poor into servants of property. So poverty is not in conformity with right. More succinctly: free and equal people cannot covenant their way into poverty. To do so would be to ‘throw away their freedom’ (DR 6: 329). It follows that the Kantian state must include mandatory coercive redistribution from the propertied to the propertyless. Kant explicitly supports this conclusion when he writes that

On this reading, Kant was a social democrat before it was cool.

Now note that the Toronto position is stronger than Cohen’s, because it asserts the necessity of redistribution coerced by public laws. It would not, in other words, suffice if some philanthrocapitalists got together and offered poor relief equal to what a Kantian state would offer. Oprah-funded poverty relief is necessarily unjust when Leviathan-funding is available. Cohen’s joint-ownership scenario, by contrast, allows private persons to accommodate the external actions of others. Again, the problem with this move is that Oprah’s permission is always revocable by Oprah’s private will.Footnote ⁶³ This is why the Toronto argument is more robustly egalitarian than Cohen’s.

I now explain how Kant’s treatment of work shows his views of economic independence to be even more demanding than Toronto liberalism allows. To anticipate the rest of the argument, suppose that public roads are of low quality compared to private roads, to such an extent that everyone prefers private to public roads. Then, pace Ripstein, Toronto’s public roads do not solve the problem of mutual dependence. What we need, instead, is a set of public roads that minimize or altogether eliminate the extent of private discretion over our walking lives. By the same token, a system of public right that only gives us some property but allows full private discretion over our working lives would be similarly unjustified. So Toronto leaves us unfree in both our walking and our working lives. I now make the case for this conclusion.

3.3 Kant beyond Toronto

In Section 2, I argued that if the blacksmith in India lacks iron ownership, then she lacks independent use of her productive powers. In order to exercise these powers, she must get permission from the iron owner(s) to use the iron they own. This means she must put her powers at their disposal. And this is why Kant disenfranchises her, on the grounds of lacking civil self-sufficiency. None of these inferences makes sense unless Kant thinks of the rightful condition as a system of cooperative production, in which the subjection of labour capacity to a private will undermines one’s formal standing to legislate universally.

Now, Kant’s discussion of the two blacksmiths asserts a discriminating position: the European blacksmith is independent, whereas the Indian blacksmith is not. What explains this asymmetry? The Toronto answer is that only the latter is poor. But one can imagine the Indian blacksmith possessing property equal to that of Kant’s civil servant, craftsman, and European blacksmith. Yet only the latter are dubbed independent.Footnote ⁶⁴ So Kantian economic dependence must be broader than propertylessness. More precisely: propertylessness is sufficient but unnecessary for unjust economic dependence. How, then, does Kant propose to spell out the necessary condition? The answer, I think, must have something to do with the independent possession and use of one’s powers. To insulate the rest of my exegesis from Toronto liberalism, I will hereafter assume that wage labourers are not poorer than the median public servant – perhaps because there is high demand for labour, or there is public poverty relief, or an unconditional basic income. Being empirical, these assumptions do not beg the question against Toronto. I now explain how economic dependence, for Kant, extends beyond poverty.

A central step towards understanding Kant’s affirmation of the discriminating position must look to the role he assigns to the market. On the present interpretation, the European blacksmith produces for a competitive market in a self-standing way, that is, not for a boss or a series thereof (DR 6: 315). This is what I called independent commodity production. Independent commodity production, for Kant, guarantees the productive independence of interdependent people. But now notice that we are already outside Toronto territory, since independent commodity production requires more than property ownership. After all, the Indian and European blacksmith might own the same amount of property. Yet Kant only dubs the latter as ‘self-standing’.

An obvious rejoinder is that there is no discrimination to be had here. If nobody wants to buy the European blacksmith’s hammers, then she will starve. By the same token, if nobody wants to hire the Indian blacksmith’s labour, then he will starve. Therefore there is no sustainable discriminating position between the two cases – dependence on a series of consumers through the product market and dependence on a series of bosses through the labour market. Kant’s response to this objection is remarkably consistent across TP and DR. It consists in his distinction between opus and operamFootnote ⁶⁵ – that is, between

1. (1) private discretion over the use of your powers

   vs.

2. (2) private discretion over the product of the use of your powers

The distinction between (1) and (2) is basically a distinction between markets for labour, on the one hand, and markets for goods and services, on the other.Footnote ⁶⁶ Kant’s implicit assumption must be that competitive markets for goods and services draw a justification that labour (and capital) markets do not. For only the former are indispensable to the coordination of production and consumption. In Waheed Hussain’s (Reference Hussain, Ripstein and Vrousalis2023) apt expression, competitive markets for goods and services constitute a public ‘pattern of practical reasoning’ that coordinates the behaviour of private actors in a way that would otherwise be impossible to achieve. Capital and labour markets, on the other hand, are unnecessary for that allocative role, so they do not draw the same justification. Moreover, as I explained above, (2) alone is in conformity with independence, because (2) alone is consistent with the producer legislating on behalf of her own productive agency, independently of the productive choices of others. If that argument is sound, then Kant has a case for discriminating between his two sets of examples.

Of course, Kant accepts that independent commodity production implies market dependence. The European blacksmith depends on the will of the consumers as a whole. But this is just shorthand for dependence on the material preconditions of the just state. It is therefore dependence on the conditions for the freedom of all: ‘Thus a nation’s wealth, insofar as it is acquired by means of money, is really only the sum of the industry with which human beings pay one another and which is represented in the money in circulation within it’ (DR 6: 287, emphasis added).Footnote ⁶⁷ Such dependence on the mere ‘sum of industry’ contrasts with the Indian blacksmith’s dependence on the private will of a (series of) private employers. So the European blacksmith, but not the Indian blacksmith, can self-legislate, in the sense that she can make arrangements for herself with what is hers. The exercise of her powers is not permissionally conditional on someone else’s – only the sale of her product is.Footnote ⁶⁸ No other reading makes sense of Kant’s consistent assertion of the discriminating position and its associated contrasts.

This kind of argument would be roundly criticized by Marx some sixty years later. In the Grundrisse, Marx argues that independent commodity production – independent, small-scale production for a competitive market – is compatible with a set of merchants, usurers, or protocapitalists coming to control the net product. This means that they possess an ownership-conferred power over the labour capacities of others, even if they are not anyone’s boss and even if markets are perfectly competitive.Footnote ⁶⁹ This is possible because some market actors (in competition with each other) can command more labour than they give in its production. This possibility, broached by Smith in The Wealth of Nations, was hardly applicable to the port economy of eighteenth-century Königsberg, Kant’s city.

But the road to Manchester may yet begin in Königsberg. Kant’s insistence on independent commodity production, although unrealizable under mechanized industry, suggests that he is not just interested in poverty as propertylessness, as the Toronto interpretation suggests. I now present another example that builds on the distinction between (1) and (2). Suppose that omelettes are the only means of consumption and that, if I am to nourish myself, I must produce an omelette. As long as you and others own the eggs, I can produce the omelette only by your permission, which makes my ability to set and pursue the end of omelette production – my productive purposiveness – subordinate to your unilateral will. Omelette redistribution will not solve this problem, insofar as it leaves the mode of omelette production untouched, no matter how much it ameliorates the mode of omelette distribution in my favour. Egalitarian egg predistribution does better, but still subordinates my omelette production to your will, insofar as it does not preclude your ownership and control over the omelette-producing cookshop. Independent commodity production, Kant thinks, features neither form of subjection: in producing for the market with their own inputs, independent commodity producers are subject neither to the power of the lords nor to the power of the rich.Footnote ⁷⁰ Whether or not Kant is right to think this, it follows that Kant’s political economy of citizenship is not fundamentally about property and propertylessness. It is, rather, about forms of power over the productive capacities of the direct producers.

To sum up the argument so far: Kant says that those who cannot make arrangements about themselves, including the poor, cannot legislate on behalf of their own agency. They are therefore unfit to vote (DR 6: 315). But later in the same text he broaches the possibility that they need not be disenfranchised if given enough property to legislate on behalf of their own agency (DR 6: 326). I have argued that Kant’s examples evince the same ambivalence about wage labour: should it be abolished or should its agents be disenfranchised? The ambivalence arises because if nobody can covenant into property as her master, then nobody can covenant into capitalist property as her master. But then eliminating poverty (as general propertylessness) is only a subset of economic independence. The Toronto interpretation’s emphasis on propertylessness cannot account for this conclusion. Interdependent independence can.

I now explain what this interpretation has to say about a related but neglected issue: the gendered and racialized division of labour.

3.4 Race and Gender

According to Kant, Private Right is subdivided into three categories: property, contract, and status.Footnote ⁷¹ The third category involves ‘a right to a person akin to a right to a thing’, which makes certain uses of other humans ‘provisionally rightful’ (e.g. of a child until majority or of a marriage partner) in ways that would be impossible without law (DR 6: 276). More specifically, the subsumption of external things (property), of the ‘causality’ of others (contract), and of domestic society (status) under relations of Private Right makes labour flows within these relations provisionally rightful. Now suppose, through no great stretch of the imagination, that care work is overwhelmingly performed by women and migrants. It follows that Kantian Private Right establishes provisionally rightful labour flows that are gendered and racialized. Jordan Pascoe (Reference Pascoe2022) has recently argued, on this basis, that Kantian citizenship entails a set of unjust claims, on the part of a white male ruling class, to the surplus labour of gendered and racialized humans. More precisely, Pascoe argues that the domestic labour of migrants and women under Kantian Right is structurally dependent – as is standard wage-labour – but is also gendered and racialized – unlike standard wage-labour.

For the sake of concreteness, consider the following schematic representation of labour flows under Private Right. There are three hierarchically related groups: the As are active citizens, engaged in ‘self-standing’ market transactions with each other. The Bs are passive citizens engaged in wage-labour for the As. The Cs are also passive citizens, but this time in ‘status’ relationships with the As and Bs as ‘heads of households’ (DR 6: 282). In Kant’s characterization of domestic society, the Category includes wives, domestic servants, and children. There are two forms of interdependence here: horizontal (between the As) and vertical (between the As, on the one hand, and the Bs and Cs, on the other). Crucially, in all vertical relationships labour flows unilaterally from the subordinates (the Bs and Cs) to the superordinates (the As). With this schema in mind, I now return to Pascoe’s treatment of Kant.

I agree with Pascoe that Kant’s account of citizenship implies de facto claims, by a white male ruling class, to the alien labour capacities of other humans. This much follows from interdependent independence.Footnote ⁷² But there are some contrasts between Pascoe’s account and mine. First, Pascoe claims that the reigning injustice here is ‘coerced cooperation’ (Pascoe Reference Pascoe2022, 50). This is misleading. For Kant, coercion is a feature of every just society and a necessary complement to independent cooperation. This is why, for example, an able-bodied rich person can be rightfully coerced to contribute to general taxation in the interest of the freedom of a disabled poor person. Such coercion, according to Kant, would only be tantamount to a ‘hindering of a hindrance to freedom’ (DR 6:231). This is significant, because some coercible entitlements to one’s product are not generically objectionable from the point of view of rightful freedom. The As, for example, may have coercible entitlements to each other’s labour contribution or its fruits, consistently with everyone’s independence.

But perhaps Pascoe is objecting to power over the total stock of labour capacities, not the flow that constitutes their exercise. That is, the problem seems to be that the Bs and Cs are subordinate to the As, because the latter dominate the former. This domination means that the As, as a whole, are entitled not to alien labour but rather to the use of the stock of labour capacities of the Bs and Cs. The exercise of those capacities constitutes unjust labour flow.Footnote ⁷³ And if Pascoe is right about the ‘enclosed dependence’ of migrants and women under domestic right, then the Cs are excluded from the benefits of social cooperation. They are, therefore, in addition to being exploited, extorted. This is consistent with my interpretation of Kant.

Another contrast between Pascoe’s interpretation and mine pertains to Pascoe’s reading of Public Right, which implies that interdependence precludes Kantian independence. But surely the latter presupposes the former. More precisely, the As are active citizens, who are reciprocally horizontally interdependent and also vertically interdependent with the Bs and Cs. Kant does see the problem with vertical interdependence – dependent labour – but dismisses it by allowing that any of those below can ‘work his way up from [passive citizenship] to an active one’ (DR 6:316). Pascoe is right to highlight the consistency of upward mobility with structural unfreedom. She does not, however, provide an explanation of how the disenfranchisement of the Bs and Cs is supposed to be justified, on Kant’s account. Interdependent independence fills this gap.

Crucially, according to interdependent independence, Kant’s disenfranchisement of passive citizens is only about dependent labour. So it can have nothing to do, in the first instance, the ascription of gender and race.Footnote ⁷⁴ A good example is Kant’s judgement of Kant himself. In DR 6:315 Kant contrasts the private tutor with the schoolteacher, claiming that only the latter performs independent (selbständig) labour. So only the latter is an active citizen – a citizen with a right to vote. Yet, as Pascoe points out, no relevant empirical feature of Kant’s changes in his transition from being a private tutor (in the 1740’s) to a university professor (in the 1770’s). This claim is generalizable. That is, no cognitive or motivational feature, or its ascription to some putative race or gender, plays any role in Kant’s account of independent labour. This is the only condition necessary and sufficient for the transition from passive to active citizenship. As I explained in Section 2.5, this contradicts a mainstream neo-Roman reading of independence.

This brings me to a final point. Most of Kant’s political writings are produced in the 1790s. They contain scarcely any association of Right with the putative cognitive and motivational dispositions of racialized and gendered humans. Kant does think of unilaterally determined purposiveness as merely sensuous and so not as an intellectual part of the republic. But this is not because dependent labour – and therefore racialized or gendered labour – cannot subsume sensibility to intellect. Rather, it is because that determinative power, when rightfully exercised, is de facto subject to the alien will of some ruling class. By definition, ruling classes rule through control over labour, whether or not that labour is gendered or racialized. So Kant’s disenfranchising exclusions, although unjustifiable, evince a fundamental concern about the independent determination of one’s sensuous powers, including her labour power, to the constitution of public power.

There is an interesting parallel here between Kant and Toussaint L’Ouverture, the Haitian revolutionary and founder of modern Haiti. Pascoe discusses a famous passage, in which Kant argues that a slavery contract is impossible. Kant mentions Haiti by name (‘the Sugar Islands’) and broaches an alternative to wage-labour: the possibility of ‘someone’ – presumably an emancipated slave – contracting to work ‘on his master’s land in exchange for the use of it instead of receiving wages’ (DR 6:330, my emphasis).

It is not far-fetched to suppose that Kant was here applying his third category of Private Right to the institutions of a possible post-slavery Haiti. L’Ouverture, for his part, had to invent a much more authoritarian version of the same category. In the 1801 constitution of Haiti he instituted a system of forced, plantation-based sharecropping, in which the contractual obligations of planters were to be overseen and enforced by the governor (see Hazareesingh Reference Hazareesingh2020, 401 n.88). So L’Ouverture was forced to institute, in practice, the same juridical category of right, lying between slavery and wage-labour, that Kant had invented in theory. In both cases, the putative justification for that category seems to have arisen neither from paternalistic opposition to idleness nor from any racialized assumptions about the dispositions of ‘lazy freedmen’, as Pascoe interprets Kant (Pascoe Reference Pascoe2022, 39 f). Rather, it arose from a commitment to mandatory cooperation as the material basis for the freedom and equality of all, under scarcely propitious circumstances. I now explain how these ideas extend to one specific form of racialized labour, namely migrant labour.

Kant’s account of Public Right comprises, in addition to rightful intrastate relations, rightful interstate relations (the ‘Right of Nations’), as well as rightful relations between states and individuals or non-state entities (‘Cosmopolitan Right’). Sylvie Loriaux has argued that rich states, according to Kant, may have redistributive duties to poor states. The ground of these duties is securing the ‘preconditions of self-determination’ for poor states as part of a global civil condition (Loriaux Reference Loriaux2020, 47). These duties are not coercible, but only subject to interstate agreement. Since this Element is about intrastate relations, I will not discuss interstate and cosmopolitan right.Footnote ⁷⁵ Yet there is, Loriaux argues, another class of duties to migrants from poor to rich states that are strictly intrastate:

On this view, European states have duties of justice to, say, Indian migrants, insofar as the former are implicated in the plight of the latter. Kant mentions colonialism, war, and fraudulent trade practices as forms of that plight. Loriaux is surely right that these are Kantian grounds for an open migration policy. But interdependent independence seems more demanding than Loriaux allows. For there is little by way of justification that Kant’s Europeans can offer to Indian migrants for denying them not just admission but citizenship. That is, Kant’s disenfranchising exclusions clearly apply to all dependent labour, whether performed by indigenous or by migrant workers. To return to Kant’s own example, the European blacksmith owns means of production. She therefore enjoys civil self-sufficiency and so is enfranchised. The Indian blacksmith, by contrast, owns no means of production. She therefore does not enjoy civil self-sufficiency and so is not enfranchised. Interdependent independence turns this QED into a reductio. Even if it were admitted that the rights and obligations of citizenship extend only to intrastate relations, they must cover not just the indigenous population but also in-state immigrants, such as blacksmiths from India.Footnote ⁷⁶ At least this much seems to follow from Kant’s own ambivalent assertion of the priority of independence over property. This concludes my discussion of racialized labour.

I now explain how variants of the idea of interdependent independence point the way towards a more radical critique of socioeconomic inequality, through Hegel and Marx, up to Rawls.

3.5 From Kant to Rawls, through Hegel and Marx

Interdependent independence establishes a presumption against power over the productive capacities of others. It is a ‘presumption’ because there are all kinds of special justifications for such power.Footnote ⁷⁷ One such justification is the existence of public offices. To produce under the direct authority of a public office, assuming it is geared towards the freedom and equality of all, is to serve the conditions of everyone’s freedom, including one’s own. It is therefore to serve only the conditions of one’s rightful freedom – to ‘make arrangements about yourself’ as free and equal. Kant’s contrasts between domestic and civil servant, private tutor and school teacher, copyholder and freeholder, all seem to presuppose this exculpating condition on power. How else could the same labour content (delivering letters, teaching students, cultivating the land) have different freedom valence across Kant’s contrasts?

But, as I explained in Section 3.3, Kant’s DR contrast between the two blacksmiths and his TP contrast between barber and wigmaker suggests a different special justification for power over productive capacity. That justification is self-standing production for a competitive market. This is a market in which producers are not necessarily poor or subject to the unilateral authority of bosses. Consumers in this market have power over producers, but that is unlike the power employers have over employees. Crucially, this is the only way to make sense of Kant’s consistent assertion of the discriminating position in all of his writings on work. Marx’s later criticism of the discriminating position, and his concomitant distinction between labour and labour power, shows that Kant’s argument can only be sustained under very restrictive conditions, namely under conditions where the capitalist mode of production does not predominate. But both Kant and Marx may be right: the former was wrong only about the restricted applicability of an otherwise sound principle. So the reason Toronto liberalism is inadequate is that Kant takes production, social reproduction, and their material conditions as themselves subject to the innate right to freedom. In a nutshell, Toronto neglects the supply-side of Right.

A corollary of Kant’s treatment of the market is a ‘dialectic’ of labour, or something resembling a materialist theory of history.Footnote ⁷⁸ Although Kant does not say so explicitly, DR adverts to the form of interdependent independence that comes with independent commodity production. That is, Kant’s conception of the independent free and equal citizen starts from the assumption of equal legal subjects. He proceeds by adding qualifications, including their market relationships, that justify ascribing citizenship status to them. Marx’s mode of operation in Capital is similar. He begins from the equal legal status of reciprocally-situated commodity owners. He proceeds by adding qualifications that would justify ascribing capitalist status to them. In both cases, the ascription (being a citizen, being a capitalist) is existentially dependent on, and derived from, the form of labour extracted or performed. In this sense, the road to Manchester must necessarily begin in Königsberg.

This ‘dialectic of labour’ is substantiated in Kant’s discussion of the foundations of the state in the Conjectural Beginning of Human History. Here Kant surmises that ‘the human being passed over from the period of comfort and peace into that of labour and discord, as the prelude to the unification in society’ (CB 8: 118, emphases in original). Since the division of labour and the concomitant ‘discord’ are indispensable to progress, the only way to humanize them is through the institutionalization of independence. Echoing Rousseau, Kant adds that this requires an ‘external lawfulness, … civil right’, which obtains because ‘the human being was to labour himself out of the crudity of his natural predispositions by himself, and yet was to take care not to offend against them even as he elevates himself above them’ (CB 8: 117, emphasis mine). My interpretation explains how this enfranchisement of specifically human production might work as part of Kant’s broader theory of justice.

Now, Kant could not have been a socialist, any more than Plato could have been a Christian. Like his contemporary Sieyès, Kant’s main concern was to eradicate the juridical vestiges of feudalism. But Kant went further than Sieyès, by thinking of the modern state as a system of cooperative production, in which our reciprocal entitlements to one another’s labour carry a justificatory burden. Moreover, the application of the idea of Selbständigkeit to the modern capitalist state unearths a tension between citizenship and poverty. Its full import culminates in Marx’s Critique of Political Economy, half a century after Kant:

According to interdependent independence, this ‘contradiction’ follows from Kant’s understanding of the proper object of universal legislation in the modern state. Most theories of citizenship since Kant have had to grapple, implicitly or explicitly, with this problem.

It may be useful to contrast Kant and Hegel here. In the Philosophy of Right, Hegel redoubles on Kant’s praise for commerce and the division of labour, which he calls ‘the system of needs’. Hegel goes as far as to argue that ‘civil society’ – the market system – is the distinguishing feature of modernity. He claims that the development of modern civil society expresses a ‘subjective particularity … allowed its rights’. This ‘becomes the animating principle of the entire civil society, of the development alike of mental activity, merit, and dignity’ (Hegel Reference Hegel and Knox2015, 133). For Hegel, labour under the liberal institutions of modernity completes the human being as such. So Hegel, like Kant, thinks of the ‘self-standing’ citizen as the pinnacle of modernity.

But Hegel also highlights the dangers of economic instability and poverty inherent in the market system. And although he does not yet recognize the possibility of a specifically capitalist mode of production, he does identify poverty as a recurring, inescapable, and destructive feature of civil society. Poverty is inescapable, Hegel thinks, because it results from the mechanization of production and the concomitant – he thinks – drop in wages and employment. It is destructive because it tends to dissolve the estatesFootnote ⁷⁹ into opposing classes. This threatens the very existence of the modern state. So, although Kant and Hegel offer different solutions to the problem of poverty,Footnote ⁸⁰ both disenfranchise the poor for similar reasons: Kant, because the poor cannot freely self-legislate; Hegel, because they jeopardize the stability of free institutions. Crucially, both Kant and Hegel take the connection between labour and poverty to be central to the constitution of modern citizenship. The connection remains central to contemporary political philosophy.

I conclude this section by contrasting the Kantian picture of cooperative production with Rawls’. Rawls (Reference Rawls2001) famously criticizes welfare-state capitalism for being too lenient on inequality and therefore on the economic conditions of equal citizenship. Rawls’ view is, in that respect, more demanding than Toronto liberalism. It is therefore more in line with Kant’s, as I have presented him. In his later writings, Rawls explicitly endorses ‘property-owning democracy’ (egalitarian predistribution) and ‘liberal socialism’ (state-owned means of production plus worker cooperatives) as the only two eligible institutional arrangements for ‘justice as fairness’. This is how Rawls proposes to enfranchise ‘powers and prerogatives of offices and authority’ into the subject of justice. Now suppose that these powers fall under the purview of his ‘difference principle’,Footnote ⁸¹ and that work involves such powers. There is, then, no reason why a Rawlsian well-ordered society should not regulate work and the workplace to treat everyone as free and equal.Footnote ⁸²

Now, according to interdependent independence, the normative disempowerment of the poor – their inability to legislate universally because an alien will de facto controls the conditions of their agency – also afflicts the worker subject to a capitalist boss. It does so, moreover, whether or not the worker is poor. One Kantian institutional remedy to this disempowerment might be to endow workers with the requisite public powers over their conditions of work, quite independently of rights to property. These may include a system of pro-labour labour law, but might also include more substantive public control over production. On this reading, Rawlsian property-owning democracy would be insufficient for Kantian independence. And now we find ourselves, again, further to the left of Toronto liberalism. In the next section, I will sketch a Kantian argument for public control over investment that supersedes both Kant’s model of independent commodity production and Rawls’ model of property-owning democracy. That argument tracks the influence of Kant’s philosophy on the ‘ethical socialism’ of the late 19^th and early 20^th centuries.

I now summarize the argument of this section. According to interdependent independence, Kant’s political economy of citizenship is not fundamentally about property and propertylessness, as the Toronto liberals argue. It is, rather, about control over labour capacity. I argued that such control is, in general, inconsistent with Public Right, unless it draws a special justification. One such justification, for Kant, is self-standing (selbstständig) production for a competitive market – that is, independent commodity production. Although independent commodity production is not possible under mechanized industry, Kant’s normative theory, I argued, provides fruitful engagement with the topic of work that takes us beyond Toronto liberalism. In the next section, I explain how interdependent independence helps us make progress in contemporary political philosophy, by supporting a strictly egalitarian account of cooperative production.

4.1 Rawls’ Ambiguity

According to John Rawls, the just society is a system of cooperation for mutual interest, in which citizens cooperate in reciprocally justifiable terms as free and equal. In this, Rawls follows a long tradition from Rousseau to Marx. Rawls discusses the kinds of justifications that can be offered on behalf of social and economic equality on the basis of his two principles of justice. These principles are supposed to reflect the three values of the French Revolution: freedom – reflected in Rawls’ ‘liberty principle’ – equality – reflected in his ‘fair equality of opportunity principle’ – and fraternity – reflected in his ‘difference principle’ (Rawls Reference Rawls1971, 60 f).Footnote ⁸³ Towards the end of A Theory of Justice, Rawls offers the following defence of the division of labour:

Kant agrees with Rawls (and Marx) that the just society would preserve some form of the division of social labour.Footnote ⁸⁴ I want to dwell on the positive argument of this passage, which focusses on meaningful work. ‘Meaningful work’ here stands for two things: first, work that is democratically self-directed in some strong sense and, second, work that involves sharing out arduous tasks or toil to such an extent that all workers have a reasonably equal chance of realizing themselves at work. Rawls’ theory of justice proposes to buttress meaningful work through economic institutions reflecting the ‘meaning of fraternity’:

Can this idea buttress Rawls’ emphasis on ‘meaningful work in free association with others’? The answer is not obvious, because fraternity, as Rawls elaborates it, rests on an important ambiguity. This is highlighted by the following example, due to G.A. Cohen:

By Rawls’ own lights, just fraternity cannot be mere trickle-down from the privileged to the unprivileged.Footnote ⁸⁵ To see why, suppose that being a banker or singer offers a higher salary (or higher social esteem or is inherently more meaningful) than being a janitor. Schematically, the banker uses the market to leverage her savings, whereas the singer uses it to leverage her talents. Rawls asks: what kind of justification, if any, can the banker or singer offer to the janitor to elicit her productive contribution to their joint product, given that the janitor receives much less than the other two?

Whatever Rawls’ actual response, a Chicago-type justification is unavailable to him. This is because Chicago, compared to Boston, only gives breadcrumbs to the janitor. It is therefore ineligible as an expression of Rawlsian fraternity. A Kantian reformulation of this thought could go something like this: if the private will of the privileged, as a whole, determines access to goods and services that the unprivileged need to be able to set and pursue their ends, then there is pro tanto an injustice.Footnote ⁸⁶ I will elaborate on this Kantian twist in Section 4.3. Before I do, I must explain Cohen’s critique of Rawls at greater length.

4.2 Cohen’s Cut

Cohen (2008) attacks Rawlsian liberalism for its tolerance of inequality. Cohen insists that, since the owners of scarce productive assets in Rawlsian society will get more than non-owners, Rawlsian liberals are unjustifiably committed to ‘giving to those who have’ (Cohen 2008, 86).Footnote ⁸⁷ He asks: what justifies the fact that the rich or talented will, in a Rawlsian society, fail to contribute the same amount of work at higher rates of taxation (for the benefit of the less well-off)? According to Cohen, such behaviour is normally unjust. The normality qualification is important, because a refusal to preserve output at a higher tax rate can be caused by an unwillingness to work as much or by the practical impossibility of doing so without impugning one’s life projects. Sometimes the rich and talented can’t maintain production at a higher tax rate – and that may be a perfectly good excuse – and sometimes they just won’t maintain production.

To this distinction between ‘can’t’ and ‘won’t’, there correspond, roughly, two readings of the difference principle. The strict reading of the difference principle ‘counts inequalities as necessary only when they are, strictly, necessary, that is, apart from people’s chosen intentions’:

The lax difference principle entails ‘giving to those who have’, says Cohen. So we should favour the strict difference principle on justice grounds.Footnote ⁸⁸ Put in terms of the Two Brothers case, suppose that A and B can rightfully coerce each other to go from New York to Boston or Chicago – as Rawls’ theory of justice implies. Then they can rightfully coerce each other to go from New York to Boston, period. Chicago should not be on the cards. The argument for this conclusion goes as follows. According to the lax reading of the difference principle, the following argument is sound:

Cohen calls this the incentives argument for inequality. The incentives argument is an expression of the lax difference principle. According to Cohen, this argument has the same structure as the conditional threat proffered by a kidnapper to the kidnapped child’s parents:

The two arguments are relevantly analogous, says Cohen, because both the rich and the kidnappers wrongfully demand rewards to which they are not independently entitled.Footnote ⁸⁹ In a nutshell, Cohen’s case against the lax difference principle is that it countenances incentive demands that are not reciprocally justifiable to free and equal people. The related Kantian thought, which never figures explicitly in Cohen’s reasoning, is that incentive demands can never constitute an omnilateral justification for certain inequalities. Such demands are instances of might-makes-right, or attempts to return our social relations to the state of nature.Footnote ⁹⁰ I now discuss this Kantian argument for the strict difference principle more closely.

4.3 A Kantian Argument for Strict Maximin

I began this section by asking what justification the banker can offer to the janitor either for earning more or for having a more meaningful job. According to Cohen, she can offer very little. And the little she can offer must be about what she can’t do. Kant’s political philosophy has a powerful argument for supporting this Cohenite conclusion. The argument is that ‘won’t’-type justifications on the part of the privileged are always invalid, because they are attempts to make the direct producers into servants of the will and won’t of privilege. The idea is that, when the terms of social cooperation depend on the will and won’t of privilege, none of us are free. So when the rich and talented say that they ‘won’t’ keep up production when taxes are raised, they are not offering a genuine justification for not doing so. All they are offering is a conditional threat based on their collective might as owners of different kinds of productive assets.

Although Cohen never quite puts it in these terms, his analogy with the kidnapper suggests that he thinks of those who leverage their productive assets in the market in terms of such unilateral lawgiving. This does not mean that the privileged can, by right, be forced to work or to preserve their productivity. But it does mean that, if they decide to work, then they cannot use their market leverage to lessen their relative tax burden. So privilege is not wronged if it receives no incentives payments or if these payments are reduced or confiscated through taxation.

This Kantian argument from the independence of the worse-off has implications for how to think about economic power under capitalism. One of the major discoveries of Marx’s critique of political economy was that capitalist control over productive assets confers on the capitalists control over the total stock of the labour capacities of workers. This control is consistent with capital and labour being equal juridical subjects as commodity owners (of non-labour and labour inputs, respectively).Footnote ⁹¹ A corollary is the distinction between market and economic power. If markets are competitive, capitalists (individually and collectively) have no market power. But, at the same time, capitalists collectively do have a property-conferred economic power over the alien labour capacities of workers. In other words, capitalists collectively can but won’t maintain their contribution to public institutions in response to a tax hike. More generally, the citizens’ dependence for employment, social provision, production, and social reproduction, on the optimizing wills of the capitalists subjects the former to the constraining choices of the latter. This is why capitalism draws the same objection as poverty: in making the non-owners of productive assets into servants of capital, capitalism is inconsistent with the freedom of those non-owners. Cohen’s argument is thus kantianized.

What are the relevant Kantian alternatives to capitalism? A relevant contrast is with an economic constitution in which the producers can use the stock of their labour capacities independently of privilege. Under independence, their use of these capacities is not subject to the will and won’t of the owners of productive assets. On this reading, Kant’s emphasis on Selbstständigkeit – the independent possession and use of one’s rightful powers – contains in crux the democratic idea that the will and won’t of privilege cannot figure in the justification of social institutions and their reproduction.Footnote ⁹²

I now return to Rawls. Although he does not address ‘can’t’ and ‘won’t’ directly, he does say that social and economic inequalities are justified only if

This is an inchoate statement of the ‘can’t’ and ‘won’t’ distinction, indeed a concession to the strict reading of the difference principle. But the Kantian argument cannot stop here. It further requires that public power pick up the slack between ‘can’t’ and ‘won’t’. That is, just as poverty alleviation must be provided by the state, so must the direction of overall investment. The reason is that only public power can attenuate the kind of discretion that capitalist private power has over the stock of our labour capacities. This remains true even under conditions of distributive justice and even if there are reasonable exit options to working for capitalists.Footnote ⁹⁴ After all, the bare existence of these options does not preclude the domination of capital, unless these options are up to us in some strong sense.Footnote ⁹⁵ What is ‘up to us’?

Making the economy up to us could involve a combination of public control over alienable productive assets, together with high taxation over the use of inalienable assets. What form ‘public control’ should take cannot be settled here. But it would minimally require public ownership of major social utilities and credit institutions, as well as some form of planning of aggregate investment. In the interest of completeness, and for the sake of highlighting a broader point about structural domination under Kantian Right, I now sketch a well-known argument for this conclusion. My purpose is twofold. First, I want to show that influential socialist ideas have their roots in Kant’s political philosophy. Second, I want to show how these ideas also illuminate the limits of his philosophy.

4.4 From Immanuel Kant to Eduard Bernstein

Outlining an argument dating back to the early days of social democracy, Przeworski (Reference Przeworski1985) explains how democratic governments under capitalism are structurally dependent on capital. The capitalists have power over governments in part because they can credibly threaten to disinvest, fire, relocate, or a combination thereof. Carrying out such threats will precipitate job losses and wage cuts, undermining any government’s re-election prospects. Supposing that these threats are credible and that governments are interested in re-election, capitalists have structural power over governments.

But capital is not the only productive asset that can be leveraged in the market. As Cohen and others have shown, talent (and functionally equivalent inalienable assets) can be similarly leveraged. Suppose, pace Przeworski, that there is an economy of cooperatives, owned and controlled by their own talented workers.Footnote ⁹⁶ Cooperatives, in general, are more conducive to equal freedom than capitalist firms, because they are unlikely to relocate in response to democratically authorized tax hikes, or to lay off workers in response to minimum wage legislation, etc. These are the kinds of concerns animating democratic egalitarians, in general, but also Kantians, in particular.

Now note that, under a regime of private property in productive assets, these cooperatives must control their own investment, hiring, and pricing policy. There are, in general, two ways firms can affect voters and thereby leverage governments. They can cut worker incomes and they can increase unemployment. There are, in turn, two ways to increase unemployment: firms can lay workers off, or – assuming a growing labour force – they can fail to hire. Suppose that cooperatives are disinclined to fire their own members. They can still undermine democratic decisions by failing to hire new workers, or, less directly, by raising prices faster than nominal wages. Now suppose there is a democratically authorized decision to raise the minimum wage. Cooperatives can defeat this decision by threatening not to hire new workers. This ‘won’t’ threat is, in turn, made credible by their failure to invest. For under a regime of private control over investment, ‘the profits of enterprises today are the investments of tomorrow, and the investments of tomorrow are the employment of the day after’ (Helmut Schmidt, cited in Przeworski Reference Przeworski1985, 43). Private employment, in other words, presupposes private investment. It follows that the only way capital’s threat can be made non-credible consists in wresting control over investment decisions from its private owners.

To sum up: private ownership of capital, whether by cooperatives or by capitalist firms, can only be reproduced by maintaining the owner’s rate of profit. Maintaining profit sometimes requires higher unemployment and/or lower real incomes. Supposing that governments depend, for their re-election, on the overall level of unemployment, or of worker real incomes, and that private owners of capital control their own investment, hiring, and pricing policies, such owners will sometimes be disposed to wield power to the detriment of democratically authorized decisions. It follows that some form of public planning of investment may, for this reason, be indispensable to the enforcement of democratic authority.

How might we arrive at investment planning? Marx uses a familiar Kantian analogy to argue that capitalist private property begets its own gravediggers. The gravediggers are the capitalists themselves, who become obsolete once the division of labour develops to the point of separation between ownership and control. Marx writes: ‘a musical conductor need in no way be the owner of the instruments in his orchestra, nor does it form part of his function as conductor that he should have any part in paying the ‘wages’ of the other musicians’ (Marx Reference Marx1991, 511). Once the joint stock company comes into the historical scene, the owner is no longer the manager. Therefore the functions of the capitalist can be separated from her ownership of productive instruments, just as the functions of the conductor can be separated from her ownership of musical instruments. It follows, says Marx, that socialism – the socialization of the means of production under a democratic state – is compatible with economic efficiency and a complex division of labour. Eduard Bernstein, one of Marx’s literary executors, turned this idea into an argument for gradual approximation to socialism through parliament-led democratic reform.

Whether Kant would have endorsed such arguments is anyone’s guess. But Bernstein’s reformism did have a long Kantian pedigree. For the rest of this subsection, I explore two variants of Kantian socialism: ethical and juridical. Bernstein, for his part, took his reformism from the 19^th century Neo-Kantians, especially Hermann Cohen.Footnote ⁹⁷ Cohen offered a defence of ‘ethical socialism’ – a socialism founded not on Kant’s juridical idea of external freedom but on his ethical idea of the Formula of Humanity.Footnote ⁹⁸ Variants of that idea pervade the writings of Karl Vorländer, Lucien Goldmann, and the Austromarxists.Footnote ⁹⁹ A boon for their theory seems to be that it unites ethics and economics, effectively subordinating the latter to the former.

Unlike the ethical interpretation, the juridical interpretation I presented above emphasizes cooperative production under a division of social labour. Kant’s alertness to the possibility of the subjection of labour to property despite their formal equality as juridical subjects, his suspicion of large landownership and the concentration of property, and his commitment to state-led democratic solutions to these problems are all methodologically and substantively based on his juridical, non-ethical account of independence.

Now, a corollary of my interpretation is that these seminal ideas are unavailable if we focus solely on Kant’s ethics. Therefore juridical socialism has considerable advantages over ethical socialism. I will mention four. First, Kant’s juridical conception explicitly connects labour with free and equal citizenship, a move unavailable to his ethical theory as such. Second, his emphasis on property, contract, and status leads to a political economy of citizenship that foreshadows Marx’s critique of it.Footnote ¹⁰⁰ Juridical socialism therefore resonates with the critique of political economy (and related social science) in ways that ethical socialism does not. Third, juridical socialism does not presuppose any specific ethical or moral theory. It is therefore compatible with neutrality on controversial questions about the good life. Fourth, the focus on coercive obligations enables us to ask what free and equal people can rightfully coerce each other to do, quite independently of their ethical dispositions. By contrast, any socialist theory that begins by assuming away deep and pervasive ethical disagreement is assuming away politics. For all of these reasons, Kant’s juridical theory of citizenship seems like a better starting point for rethinking the basics of the socialist tradition than his ethical theory. And whatever the starting point, Kant’s influence on socialist thought is beyond dispute.

I now conclude by using this section’s structural examples to highlight the structural limits of Kantian Right.

4.5 The Limits of Kantian Right

Kant’s political philosophy offers a parsimonious and systematic framework for understanding the externalization of rightful relations between persons. Starting from an innate right to external freedom, through private right, we arrive at the rich diversity of the public institutions of modernity. In most cases, Kant proceeds deontically, by asking who wrongs whom and how. This is the private law model of litigants who require an impartial judge to give each her due. Perhaps predictably, this framework does less well dealing with structural unfreedom, that is, cases where the agency of either of the litigants is absent, mitigated, or intractable. Consider a Przeworskian illustration.

A newly elected government has a democratic mandate to raise taxes in order to build public roads. This mandate can be thwarted in one of two ways: through corruption and through structural dependence. Corruption involves the privileged (capitalists, landowners, the talented) bribing government officials not to raise taxes. Structural dependence, on the other hand, does not involve corruption. Rather, it operates on the type of mechanism explained in Section 4.4. For example, the privileged announce that, if taxes are raised, they will disinvest in the economy. Disinvestment means fewer jobs, unemployment, and eventually an outvoted government. So the government withdraws or waters down its democratically authorized plan.

Both cases involve a unilateral will thwarting democratic decisions. But no agent or group exceeds their public mandate in the second case. Indeed, in the second case the privileged are legally trying to maximize their rate of return, the voters are legally choosing the government that will create the most jobs, and the politicians are legally enforcing the will of the governed. Yet something goes awry; the political procedure is rigged in favour of the rich and powerful. Kant’s political philosophy has intellectual resources for embracing this conclusion. In this section I have tried to mobilize some of them. But mobilizing them to the full means looking beyond the private law interpretation of Kant’s theory of justice, whether in its Toronto variants or otherwise.