Kant’s Doctrine of Right gives us a theory of property and argues that as a matter of public right the state has authority to tax to secure its preservation and to support “charitable or pious institutions” (RL 6:326), but Kant does not articulate or defend a theory of distributive justice.
The fact that Kant does not articulate a theory of distributive justice has not kept political philosophers from citing Kant as inspiration and support for whatever theory of distributive justice they favor2 – including those who argue that the notion of distributive justice is itself mistaken.3 This widespread reliance on Kant invites the question, “Does the Doctrine of Right imply a theory of distributive justice?”
To address this question, we discuss Paul Guyer’s argument that Kant’s Doctrine of Right implies, roughly, the principles of distributive justice as found in Rawls’s justice as fairness. Guyer’s argument is that Kant’s theory of property implies a contractualist theory of distributive justice; in turn, this implies that the distribution of property rights must be fair, and that fairness is secured only by something like Rawls’s second principle of justice.
We have doubts about each step in Guyer’s argument. That is, we question whether Kant meant for his theory of property to imply a Rawlsian form of contractualism, whether Kant’s contractualism implies fairness, and whether fairness implies the difference principle.
Regarding Kant’s relation to contractualism, we hold that, although Kant clearly maintains a conventionalist view of property rights, the normative basis of his theory of the Rechtstaat contains both contractualist and non-contractualist elements. We argue that Kant’s non-contractualist commitments in the Doctrine of Right pose problems for a reading of Kant’s state of nature as a hypothetical choice situation analogous to Rawls’s original position.
Next, we consider the steps of Guyer’s argument that move from contractualism to fairness, and from fairness to the difference principle. Guyer’s argument draws attention to the fact that the contractualist ideal can be read in two ways. First, the contractualist ideal typically states the requirement that the terms of society must be acceptable to every member of society and identifies the terms that satisfy this requirement. Second, and with greater originality, Guyer argues that on Kant’s contractualist view, there are moral constraints on the terms it is permissible to offer other members. Guyer’s insight is that one can approach a contractualist argument from two different directions: One can ask what terms are rationally acceptable, or one can ask what terms are morally offerable. We argue, however, that neither direction gets Guyer to the conclusion he draws. The first direction gets us a principle of mutual advantage, while the second direction gets us a principle of formal equality. Neither direction gets us to a criterion of fairness. Thus, the third step in the argument, from fairness to Rawls’s difference principle, also fails.
The outcome is that Kant’s Doctrine of Right is compatible with, but does not require, a number of principles of distributive justice – including those principles we find in Rawls’s justice as fairness. Kant’s Doctrine of Right, then, does not imply a single theory of distributive justice.4
In the next section, we present Guyer’s argument. In Sections 3–5, we offer reasons to doubt the main moves in that argument. In Section 6, we offer some concluding reflections about the relation between liberty and distributive justice that are suggested by our discussion of Guyer’s argument.
Guyer argues that Kant’s theory of property as developed in Private Right implies a principle of distributive justice along the lines of Rawls’s difference principle.
The universal principle of right states: “Any action is right if it can coexist with everyone’s freedom in accordance with a universal law, or if on its maxim the freedom of choice of each can coexist with everyone’s freedom in accordance with a universal law” (RL 6:230). A correlate of the universal principle of right is that every person has an innate right to “Freedom (independence from being constrained by another’s choice), insofar as it can coexist with the freedom of every other in accordance with a universal law” (RL 6:237). The innate right to freedom is the only innate right. All other rights are acquired, meaning that they are established by an act (e.g., a crime or a contract) (RL 6:237). Acquired rights are of three kinds (RL 6:247–8; 6:259–60): rights to land and objects; rights to the performance of acts by other persons; and rights of authority that allow one person to command another in some respect.
The essence of a property right is a claim to exclusive possession of an object, so that all other persons are required to defer to the claimant’s will in regard to said object.5 Kant’s account of the nature of property rights can be helpfully contrasted with that of Locke. Locke famously holds that one acquires a right to a thing by “mixing one’s labor” with it. On Locke’s view a property right is thus a direct relation between a person and an object.6 Even if there were no other persons on earth, Locke’s position implies that I could still acquire a right to own an object by mixing my labor with it. On Kant’s alternative to Locke’s account, a right to property is not a relation between the right-holder and the object, but a relation between the will of that person and the wills of all other persons with regard to that object (RL 6:260–1, 6:268). Property rights are thus a matter of social convention; that is, of agreement between rational persons about who is entitled to what.
According to Kant, land and objects are originally held in common (RL 6:262). This raises the question of how one person can acquire rights of exclusive possession in these in a manner that is consistent with the equal freedom of others.
Guyer finds Kant’s ultimate solution to this puzzle of rightful acquisition in the self-same ontology of property that gave rise to the problem.7 If a claim to exclusive possession could meet with the rational agreement of all those limited by it, then the implied limitation on their freedom is consistent with their will, and so consistent with their freedom. Thus, a unilateral provisional claim to property can be rightful so long as it is such that others could rationally agree to it. Kant holds that such an original acquisition is “provisionally rightful” (RL 6:257). An acquisition is conclusively rightful only in the civil condition (RL 6:257, 6:264).
Thus, according to Guyer, a property right factually consists in the agreement of others to defer to the authority of the owner. As Guyer writes, “a property right consists in a relation among wills, regarding an object, namely the consent of all those persons who could control and use an object that one among them can” (Guyer 2005b:263). Moreover, a property right normatively depends on the possible agreement of other persons. For Guyer, just as the original acquisition of a provisional property right is valid only if others could rationally agree to it, so the origination of property rights in a civil state requires the possible consent of all to those rights (or to the system of rules that defines and maintains those rights). In a slogan, the conclusion that Guyer draws from Kant’s account of property in Private Right is that “the ontology of property [that property consists in a relation between wills of persons with regard to an object] demands its morality [that property rights could be agreed to by all affected]” (Guyer 2000e:281).
Since Kant holds that we must have property rights (by the postulate of practical reason with regard to rights, RL 6:250), and since (both normatively and factually) we can have these conclusively only in a rightful civil condition, we have a moral obligation to create and maintain a civil condition (a state).8 The state both keeps track of the distribution of property rights (the state as recorder of deeds), and offers assurance that all have sufficient incentives to respect the property rights of others (the state as sheriff).9
So far, Guyer’s reading has taken us from the moral law to the conclusion that property rights require the possibility of agreement by all those affected by them, both for their normative validity and for their factual efficacy. And Guyer has shown how this conclusion, together with certain facts about human nature and the human condition, entails that we must form a civil condition with the institutions sufficient to secure the innate and acquired (property) rights of all.
At this point, Guyer has not identified the principle of distributive justice that ought to guide the design and activities of the state with respect to property and related economic institutions. The next three steps in Guyer’s argument are meant to identify the sort of principle of distributive justice that Kant’s doctrine of right implies.
According to Guyer, property rights normatively require that all affected by them could agree to them or to the system of rules and institutions that sustain them. Agreement can be either forced or free (Guyer 2000b:238). Since “unprovoked threat or use of force … would violate the innate right to freedom … of anyone on whom it was exercised” (Guyer 2000b:249), the right to property can “be grounded only in the freely and rationally given consent of those whose deference is needed.”10 While it is permissible to use force to create the needed assurance that property rights will be respected, the agreement required by the normative dimension of property rights cannot be forced; Guyer concludes that “for any system of property rights to be morally acceptable, all affected by that system must be able to freely consent to it” (Guyer 2006b:270).
Since we have a duty of right to create a system of public right that institutes, tracks, and secures property rights, it follows that we have a duty to create a system of public right that could be freely agreed to by all those affected by it.
Thus, Kant’s doctrine of right and theory of property imply three claims: 1) we are morally required to create and maintain a state that defines and secures for all a system of property rights; 2) this system of property rights must be constituted so that all affected could freely agree to it; thus, 3) we are morally required to create and maintain a state where the distribution of property rights could be freely agreed to by all.
It follows that Kant’s theory of right and property together imply a contractualist approach to distributive justice: “on Kant’s view, property cannot exist at all except by means of agreement, so a social contract or proviso of distributive justice cannot be merely added to already existing property rights; it is inherent in the very idea of rightful or morally acceptable property” (Guyer 2006b:271).
Guyer’s next move is to assert that a system of property rights can be freely agreed to by all only if it is fair to all: “a reasonable person will consent to a system of property rights only if he sees it as sufficiently fair… So, if a rightful system of property must be consistent with the universality of external freedom, and reasonable persons would only freely consent to a system of property that meets some minimal standard of fairness, then a rightful system of property must meet some such standard” (Guyer 2005b:264).
Guyer suggests that the minimal standard of fairness is the one articulated by Rawls’s difference principle: “Rawls’s difference principle is a plausible interpretation of the conditions under which it would be morally permissible to demand the agreement of others to a system of property and rational for them to consent to it” (Guyer 2000e:281). Depending on the specifics of the contract situation (like relative bargaining power, knowledge, and interests of the consenting parties), it might be rational to consent to principles other than the difference principle, but Guyer suggests that even if it might be rational to consent to other principles, other principles do not satisfy the moral requirements on what persons can demand that others consent to (and so cannot be permissibly offered as terms of distributive justice): “even if the difference principle does not follow from pragmatic considerations alone, it could be argued to follow from the moral constraints on demanding the consent of others to one’s property claims” (Guyer 2000e:281).
Guyer thus concludes that “Kant is committed to the conclusion that there can be external or public legislation enforcing the right to property, but only under conditions of equality like those defined by Rawls’s second principle of justice” (Guyer 2000e:281). Since we must have a system of public legislation enforcing the right to property, it follows that Kant’s theory of justice not only permits but requires that we implement Rawls’s second principle of justice (or something close to it) (Guyer 2000e:266).
In this section, we argue that Kant’s Doctrine of Right involves significant non-contractualist normative commitments, and that these non-contractualist elements of Kant’s view pose problems for Guyer’s Rawlsian contractualist reading of the Doctrine of Right.
As we have seen, Guyer argues that property rights require the agreement of others both normatively and factually: normatively, because a property right limits the freedom of others and so is impermissible by the universal principle of right unless it can meet with their agreement; factually, because to have a property right implies that other persons actually tend to defer to one’s will with respect to the object as a matter of social convention. If most others do not uphold the convention, then one does not actually have a right. So Guyer’s interpretation combines a conventionalist factual account of property with a contractualist normative account.
It is tempting to think that in the case of property rights, conventionalism implies contractualism. After all, if property rights are a matter of tacit or explicit agreement among persons (that is, if property rights are conventions), it is quite natural to suppose that the moral justification for the convention of property must be explained in terms of the fairness or rationality of the persons’ agreement. That is, it is natural to suppose that the moral status of a convention must be analyzed in contractualist terms. Guyer himself seems to invite such a view when he writes that “the ontology of property demands its morality” (Guyer 2000e:281).
However, there is no necessary connection between conventionalism as a factual or ontological account of property rights and contractualism as a normative account of property rights. Consider Hume’s theory of property. Like Kant, Hume holds that there are no natural rights to property in the Lockean sense. Instead, property rights are simply a matter of human convention or custom.11 And Hume holds that our property conventions are morally justified. But Hume’s account of the normative justification of the convention of property is decidedly non-contractualist. Hume argues that the convention of property is justified because of its beneficial consequences for human societies, not because of any agreement, whether tacit, actual, or rationally possible.12 Thus, Hume’s account of property illustrates that conventionalism about property does not imply contractualism about property.
Of course, Guyer does not think that conventionalism by itself implies contractualism. Guyer’s contractualism instead arises from the conjunction of conventionalism with Guyer’s contractualist analysis of Kant’s theory of external freedom. On Guyer’s argument, consent to the institution of property is either forced or free, and since the use of force is permissible only when it hinders a hindrance to freedom, force cannot permissibly be used to secure consent to a system of property rights. Thus, property rights require the free and rational consent of those whose freedom is limited by them.
This argument is too quick. The second premise of Guyer’s argument assumes that forcing someone to give her consent to a system of property rights cannot itself be a hindrance to a hindrance to freedom. But in the Doctrine of Right, Kant argues that we can force others to enter into a civil condition with us, even against their will, because to remain in the state of nature is wrong (RL 6:256, 6:257, 6:307–8). Remaining in the state of nature is wrong because it hinders the secure and universally consistent use of external freedom. Therefore, we have a duty of right to give our consent to the institution of a civil condition, which establishes a system of property rights.13 Thus, forcing someone to consent to a scheme of property rights can be justified as a hindrance to a hindrance to freedom, for withholding consent to such a scheme can itself be a hindrance to the exercise of external freedom.14
This outcome is important, for three reasons. First, it contradicts a key premise of Guyer’s argument: that we cannot rightfully force others to give their consent to the establishment of a property regime. Second, it points to an important non-contractualist source of normativity in Kant’s theory of private right: the duty to exit the state of nature, which Kant derives from non-contractualist premises.15
Third, it points to a fundamental disanalogy between Rawls’s original position and Kant’s state of nature. For Rawls, there is no duty to reach an agreement in the original position, and no authorization to coerce others to agree. Kant’s state of nature, by contrast, exhibits both of these features: There is a duty of right to form a civil state with others, and there is an authorization to coerce others to join with one in forming such a state. Guyer’s move from conventionalism to (Rawlsian) contractualism, in other words, is complicated by Kant’s endorsement of a prior, coercively enforceable, non-contractualist duty to exit the state of nature and form a civil state. Guyer’s contractualist reading does not sufficiently take account of the hybrid character of Kant’s view of the normative foundations of the Rechtstaat, which involves both contractualist and non-contractualist elements.
Guyer might reply to our argument by granting that, on Kant’s view, others may be coerced into joining or respecting a system of property rights, but insisting that one can coerce others only if that system of property rights is rightful. And the rightfulness of a system of property rights, Guyer could argue, depends on whether the system is such that free and rational persons could consent to it. If and only if the system of property rights is rightful in this sense – if hypothetical free and rational persons could or would accept it – does the system of property rights uphold the external freedom of all. Only in such a system can actual persons legitimately be coerced into compliance.16
This response raises the question, to which we return in the next section, of the legitimacy conditions for a rightful Kantian system of property rights. We agree that, for Kant, not just any system of property rights counts as rightful, and that individuals in Kant’s state of nature may not rightfully force others to join them in establishing a non-rightful civil condition (which would not really be a civil condition at all). But we disagree that, for Kant, the relevant legitimacy condition is that a scheme of property rights must meet the high threshold of Rawls’s second principle of justice. Individuals in Kant’s state of nature may rightfully coerce others to enter into less egalitarian regimes of property rights, and those who refuse to consent to such regimes do wrong. We make the case for these claims in the following two sections.
After arguing for a contractualist reading of Kant’s theory of property, Guyer’s next step is to maintain that free consent can be expected only where the terms of agreement are fair to all parties and that Kant’s theory of justice, therefore, implies that the distribution of property rights must be fair to all. Although we have cast doubt on the idea that, for Kant, a legitimate state can be based only on unforced consent, we grant that assumption for the sake of argument in this section and the next. We then argue that even on this assumption, Kant’s theory of property does not imply a Rawlsian conception of distributive justice.
It would be somewhat surprising if Kant’s universal principle of right together with a few facts about human nature and the human condition implied a principle of distributive fairness. Guyer’s argument is that, on Kantian premises, liberty implies equality, so that one “cannot subsume the right to property under a principle of liberty akin to Rawls’s first principle of justice without also acknowledging something very much like Rawls’s second principle” (Guyer 2000e:282). This move is controversial, and not one that Rawls would endorse. With regard to the apparent conflict between liberty and equality, Rawls rejects that either can be derived from the other; rather, liberty (the basic liberties) and equality (of opportunity and the difference principle) together present the basic terms of cooperation appropriate for a democratic society.
Moreover, it is notable that Rawls did not argue from contractualism to fairness, but from fairness to contractualism. Rawls’s idea is that if we conceive of society as a system of social cooperation between free and equal persons, then we must find a way to fairly distribute the benefits and burdens among the cooperators. He then moves from the need for principles that secure distributive fairness to contractualism as the way to identify these principles.
Of course, Guyer’s aim is not to discuss Rawls, so Guyer could shrug off these worries as insignificant: The point remains that free consent can be expected only to fair terms. But is that true? It seems there are many examples of free (and informed) consent to unfair terms. When the indigent consent to harsh terms of employment, is that not free consent to unfair terms?
Guyer might reply that such consent is not free, since the circumstances force the choice. But then it seems that Guyer relies on concepts of coercion and freedom different from those found in Kant. For Kant, coercion is a relation between persons and the freedom that is the subject of the universal principle of right is freedom from those limitations persons impose on each other.17 So, if someone cannot meet her needs due to misfortune, and she therefore consents to unfair terms of cooperation with others, then we cannot by Kant’s concept of freedom say that her consent is not free or that she was forced to consent to these terms. No person forced her, and her consent is based on the benefits agreement brings and so is rational. She freely and rationally consented to terms that are unfair insofar as they exploit her situation. And that, of course, is partly the point of Rawls’s original position: Free consent is indicative of fairness only if the parties are situated as equals and ignorant of facts that they might use to offer terms that favor those they represent.
So, it is not the case that people cannot freely and rationally consent to unfair terms of agreement. Any free and rational agreement must in some sense benefit all the agreeing parties, but not all mutually advantageous agreements are fair.
How about the other direction of Guyer’s argument – the direction that appeals to what it is morally permissible to offer as terms of agreement, rather than what terms others can freely consent to? If we follow this direction, Guyer’s argument is: (i) the system of property must be structured in accordance with terms to which all could freely agree; (ii) it is morally impermissible to offer unfair terms of agreement; so (iii) the system of property must be structured in accordance with terms that are fair to all.
The problem with this argument is that when Kant discusses the moral restrictions on the terms that we can offer others, the restriction he identifies is not fairness, but consistency or reciprocity:
When I declare (by word or deed), I will that something external is to be mine, I thereby declare that everyone else is under obligation to refrain from using that object of my choice… This claim involves, however, acknowledging that I in turn am under obligation to every other to refrain from using what is externally his; for the obligation here arises from a universal rule having to do with external rightful relations… the universality, and with it the reciprocity, of obligation arises from a universal rule.
What Kant says here is that a claim to property rights must acknowledge the equal entitlement of others to claim property rights. This consistency requirement does not, however, entail anything about the content of the rights of others.
The requirement of consistency can be elaborated in terms of the equality that Kant frequently says is one of the three defining properties of republican citizenship (the other two are freedom and independence).18 This is “equality as a subject” and it means that, with the exception of the head of state, “Each member of a commonwealth has coercive rights against every other” (TP 8:291). Kant makes clear that this formal equality of citizenship “is quite consistent with the greatest inequality in terms of the quantity and degree of their possessions … and in rights generally (of which there can be many) relatively to others” (TP 8:292). The equality in question, in other words, is an equal right to have rights and to have one’s rights enforced. Citizens cannot claim a right to have rights that others could not have, and no person “can rightfully bind another to something without also being subject to a law by which he in turn can be bound in the same way by the other.”19
The equality that Kant affirms thus turns out to have little to do with Rawlsian equality or a principle of substantive fairness, but is instead a formal equality which means that (apart from the head of state) there can be no distinctions between citizens when it comes to the right to have rights and to have one’s rights protected by the state: “all are equal to one another as subjects; for, no one of them can coerce any other except through public law … no one can lose this authorization to coerce (and so to have a right against others) except by his own crime” (TP 8:292).
Thus, though Guyer is correct that Kant thinks that there are moral restrictions on the terms of agreement we could offer others, and that these restrictions imply a standard for the system of property we should institute in a civil condition, the requirement that Kant affirms is not one of fairness, but one of formal equality. This requirement rules out a scenario where some are formally excluded from ownership rights, but the requirement does not get us to an egalitarian (or prioritarian) principle of distributive justice.
Guyer’s distinction between the two directions that a contractualist argument can take is important. And, in fact, we have shown that this is a substantive distinction, producing different requirements of distributive justice based on which direction we follow. If we follow the direction of what terms persons can freely and rationally agree to, we get a principle of mutual advantage. If we follow the direction of what terms persons can permissibly offer as terms of free agreement, we get a principle of formal equality. Yet neither of the directions gets us to a principle of fairness.
The final step in Guyer’s argument aims to show that only the difference principle (or something like it) satisfies the requirement of fairness.
Guyer does not offer much of an argument that the difference principle is required by the principles of Kant’s theory of justice. He again suggests that the argument would proceed from the moral requirements on what terms of agreement one can offer others as terms of society, rather than relying on what those agreeing to the terms could freely consent to. But, as we have already indicated, it is not clear why it is impermissible to offer any other terms of agreement than the difference principle. Guyer acknowledges that his argument here is incomplete:
[E]ven if the difference principle does not follow from pragmatic considerations alone, it could be argued to follow from the moral constraints on demanding the consent of others to one’s property claims. I do not attempt to pursue this issue further. Let us just stipulate that Kant has established in a general way that one can neither reasonably expect nor reasonably demand to establish a property right of one’s own without the consent of others, and that they will give that consent only if they see the system of property rights thereby implied as in their own best interest as well. In other words, one has a right to property as an exercise of one’s own freedom, but only under the condition that one is willing to concede an analogous right to property to others.
This is an interesting passage. In addition to admitting that his argument is incomplete, Guyer here returns to the argumentative strategy of asking what terms persons could freely consent to (not what terms they could permissibly offer), and then states the two principles of distributive justice that we have suggested are implied by the sort of contractualism that Guyer finds in Kant, namely, the principle that the system of property rights must be in the interest of all, and, second, the principle of formal equality.20
Moreover, a closer look at Guyer’s interpretation as it is expounded across various writings reveals that he finds support for multiple principles of distributive justice that could be freely consented to or permissibly offered by all in Kant’s writings.
The first principle is the minimal principle of formal equality of access to rights. This criterion is grounded in a claim of reciprocity: Those claiming property rights must be willing to grant the same rights to others who satisfy the relevant and non-discriminatory legal criteria.21 This principle is consistent with any number of distributions of benefits and burdens, even distributions where some are worse off than they would be without the system of property rights or distributions where they have less than they need to sustain a decent existence.
The second principle is that the system of property rights must be to the benefit of all: “it will be in our interest to agree freely to a property claim when the system of property rights within which such a claim is being made promises all of the participants in it some reasonable level of access to its benefits.”22 The term “benefit” is relative, but at least every person must be better off with the system of property rights than they would be without it. The most straightforward interpretation of this requirement is that the system of property must make everyone better off than they would be in the state of nature.
Guyer suggests a third principle of distributive justice: the sufficientarian principle that all persons must have at least enough resources to lead a decent life: “the right to an opportunity to property sufficient to maintain existence or an equivalent that can produce the same result provides a minimum standard for the rational acceptability of any system of property rights, where the rational acceptability of such a system is in turn a necessary condition of its morality.”23
A fourth principle that Guyer suggests is that the system of property rights works better than any realistic alternative in securing the ends of those affected by it: “any system of property rights … must … [be seen as] at least working better than any realistic alternative” (Guyer 2006b:270). It is not immediately clear how we should interpret this principle (best by what standard?). But a reasonable interpretation would be that no person’s interests can be sacrificed to better satisfy the interests of others.24 An alternative interpretation yields a tempered utilitarianism (maximizing aggregate preference satisfaction combined with a guaranteed minimum sufficient for decent living), though Guyer does not himself suggest this interpretation.
Finally, we reach the principle that Guyer argues is required by Kant’s doctrine of right: the principle that the system should be designed to maximize the position of the least well off. It seems plausible that this principle could be freely agreed to by all, and that it could be permissibly offered as terms of agreement. This shows that the principle is compatible with the contractualism that Guyer finds in Kant, but not that it is required by it.
Are any of these principles required by Guyer’s Kantian contractualism? It depends on the direction of the question: Are we asking what people could freely consent to or what terms people could reasonably (morally) ask others to consent to?
If we ask the first question, what people could freely and rationally consent to, it seems that people could freely consent to any principle that makes them better off. Thus, the second of the principles we listed, the principle that all must be better off with the system of property than they would be in a state of nature, is required.
If we ask the second question, what terms people can morally offer as terms of agreement, it seems clear that on Kant’s view reciprocity demands formal equality, so that the system must provide an equal right to have property rights (but not rights to equal property).
None of the other principles are required – no matter the direction of the argument. The other principles are all principles that persons could freely consent to if the alternative is a state of nature (no stable system of property). Conversely, we do not see any Kantian reasons why offering the other principles as terms of agreement is morally prohibited.
One consequence of our argument is that the Doctrine of Right does not single out any principle or set of principles as required for distributive justice.
In Section 3, we cast doubt on Guyer’s Rawlsian interpretation of Kant’s state of nature. But even granting that part of Guyer’s argument, we argued in Sections 4–5 that Guyer’s interpretation does not establish that Rawls’s difference principle is required by Kant’s Doctrine of Right. Leaving aside the question of whether Guyer’s interpretation is faithful to Kant’s text, we showed that it does establish two minimal requirements that any system of property rights must satisfy. First, if, as Guyer argues, a system of property rights must be one to which all could freely and rationally consent, then everyone must be better off in that system than they would be without any system (i.e., in a state of nature). Second, if the system of property rights must be one that all could permissibly offer others as terms of agreement, then the system must respect formal equality, so that all affected by it have the same rights of access to property rights.
If these are the only two requirements of distributive justice that follow from Guyer’s interpretation of Kant (no matter the direction of the contractualist argument), then Guyer’s reading of the Doctrine of Right is compatible with a number of principles of distributive justice. Many principles satisfy these two requirements, and many principles that do not satisfy them could be conjoined to them. Thus, sufficientarianism, tempered utilitarianism, varieties of egalitarianism where all are better off than in the state of nature, and the difference principle are all compatible with Guyer’s Kant.25
That Guyer’s view is compatible with a number of principles of distributive justice does not mean that it is toothless. The two distributive requirements that follow from Guyer’s reading rule out varieties of pure consequentialism, since, first, pure consequentialism might leave some worse off than they would be in a state of nature if necessary to maximize the good, and, second, pure consequentialism allows only a conditional commitment to formal equality. Guyer’s Kantian contractualism also seems to rule out varieties of Nozickean libertarianism, meaning theories of distributive justice that (i) reject the very notion of a system of public property rights apart from the natural rights that persons have independently of the state; and, therefore, (ii) permit any distribution of property rights that actually results from free transactions (or if all past wrongdoings have been corrected). Nozickean schemes therefore (iii) deny that a system of property rights can be measured by whether it benefits those affected by it. So, though Nozickean libertarianism accepts formal equality, Kant would reject it because (i) it relies on a mistaken theory of property, (ii) it has a mistaken view of the state (rejecting the idea that the state must be governed by a general will rather than the sum of unilateral and bilateral wills), and (iii) it rejects that the justice of a system can be measured by whether those affected by it benefit or not, and so does not require that all are better off than they would be in the state of nature.
All of these remarks have referred to the consequences of Kant’s view for distributive justice on Guyer’s reading. But since we disagree with one of the premises of Guyer’s interpretation – that on Kant’s view consent to a property scheme cannot be forced – there is still a question about the implications of Kant’s view for distributive justice on our reading. As discussed in Section 4, we accept that a notion of reciprocity is at the core of Kant’s idea of a civil condition, and therefore accept that Kant’s view entails formal equality – each person must have an equal right to have property rights. However, things are murkier regarding the first consequence of Guyer’s interpretation – that a system of property rights must, for Kant, be to the advantage of all. The case for this requirement that we discussed in Section 4 depends on Guyer’s assumption that people must freely accept the terms of social cooperation. But this is precisely the assumption that we called into question in Section 3. Can our reading provide a separate Kantian route to this seemingly quite plausible conclusion?
We have suggested that, for Kant, consent to a scheme of property rights need not be free for the reasons we offered in Section 3, but it still must be rational. And the requirement of mutual advantage can be regarded as following from the nature of a rational agreement – a rational agreement must be in the interest of each party. This suggestion finds some support in Kant’s texts. The basis for the duty to exit the state of nature is the postulate of practical reason with regard to rights, which holds that objects must be useable in the interest of human freedom. And it is plausible that all will be better off within a regime of property rights that allows for the stable use of objects than they would be within the state of nature, for a regime of enforceable property rights and contracts makes possible new modes of economic production and cooperation.26 Thus the rational agreement to exit the state of nature can be expected to be mutually advantageous.
On Guyer’s interpretation and on our own, then, two requirements governing distributive justice follow from Kant’s Doctrine of Right: the requirement of formal equality, and the requirement of mutual advantage. Neither requirement individually, nor the conjunction of the two requirements, entails a specific conception of distributive justice.
Finally, we draw attention to two ways that the conclusion that Kant’s theory of justice is compatible with a number of principles of distributive justice. This is an interesting, non-trivial conclusion.
First, since many have argued that Kant supports their particular theory of distributive justice – Rawls, Nozick, and others mention Kant’s moral philosophy as a source of their theories (and anti-theories) of distributive justice – it is remarkable that Kant’s theory of justice actually leaves distributive justice underdetermined. Since we agree with Guyer that Kant meant his theory of justice to be implied by (and to be the only one implied by) his moral philosophy, we have reason to doubt those who argue that Kant’s moral or political philosophy implies their favored theory of distributive justice.
Second, Kant’s political philosophy works from a single and widely accepted principle of equal liberty: the principle that, from the perspective of justice, every person is and ought to be at liberty to decide what to do for themselves, as long as what they do is consistent with the equal liberty of others. Kant’s theory of justice unfolds from this basic principle, the correlative innate right to freedom, and a few facts about human nature and the human condition. Many political philosophies accept Kant’s basic principle and the facts he works with. If our argument is correct, attempts to derive a theory of distributive justice simply from the principle of equal liberty and the facts that Kant works with fail and will continue to fail, since these Kantian materials do not decide the question.
1 Earlier drafts of this chapter were presented and discussed at the conference “Nature and Freedom in Kant,” Brown University, October 2013, and at a meeting of the DC/Baltimore Kant Workshop in July 2014. We are grateful for the feedback and criticism we received from audience members and workshop participants on these occasions. For her commentary at the Brown conference, we especially thank Marcy Latta. For editorial assistance, we thank Lauren McGillicuddy and Kate Moran. Lastly, we wish to express our gratitude to Paul Guyer for his generosity as a scholar, mentor, and friend.
2 See, e.g., Rawls 1999:221–7, esp. 226–7; Lomasky 1987:99–100; Barry 1995:31 Tomasi 2012:96–8; Nagel 1991:chapter 5.
3 Both Hayek and Nozick cite Kant as support even while rejecting that distributive justice is a legitimate political concern. For Hayek’s statement of the view that distributive justice is mistaken, see Hayek 1960; for the tie to Kant, see Hayek 1960:160–77. For Nozick’s claim that distributive justice is a mistake, see Nozick 1974:chapter 7; for references to Kant, see Nozick 1974:32 and 228.
4 Thus, we agree with Mary Gregor’s claim that “students of Kant will presumably be interested not only in interpreting the text but in trying to develop on the basis of it a theory of distributive justice in the sense of fairness in the distribution of goods and burdens. Kant himself is not concerned with this problem… His concern … is not with the problem of what distribution of goods would be fair but with the more basic problem of how someone could acquire ‘goods’ of a certain kind…” (Gregor 1988:762).
8 Cf. RL, sections 42 and 44.
10 Guyer 2005b:249. In “Life, Liberty, and Property,” Guyer specifies that coercion to ensure agreement on a claim to property right could not “be justified in the only way that coercion can be justified, as a hindrance to a hindrance to freedom” (Guyer 2000e:281).
12 For the positive thesis that the rules of justice are morally justified by their consequences, see Enquiry pp. 83–9. For the negative thesis, that justice does not depend on a social contract, see Hume 1985:465–87.
13 The civil condition, for Kant, is equivalent to a society governed by the united general will. Thus, we could also say that we have a duty of right to consent to the authority of the united general will.
14 Kant’s theory of right thus does not rely solely on hypothetically rational consent for its normative foundation. It also relies on normative consent: consent that we ought to give. Cf. David Estlund’s notion of normative consent, Estlund 2008:10 and chapter VII. If Kant’s theory of right relies on normative consent, then some of the normative work is accomplished, not by the social contract itself, but by the prior duty to give one’s consent.
16 Cf. Guyer 2006b: “… we have the responsibility even prior to or in the absence of a well-functioning state to make only property claims that could be fairly enforced against others and to coerce them only into a state that would maintain a fair system of property rights.”
17 As Kant puts it, external freedom involves being sui iuris, being “one’s own master” (RL 6:238). Furthermore, in his statement of the innate right to freedom, Kant glosses freedom as “independence from being constrained by another’s choice,” which makes clear that external freedom is about relations between persons (RL 6:237). Our reading of Kant’s notion of being sui iuris focuses on an ideal of freedom as non-interference, not on a more demanding notion of freedom as non-domination. It could be argued that being sui iuris requires a background of relative economic equality to eliminate arbitrary power relations among private citizens, but that argument would go beyond anything suggested in Kant’s text. Kant seems to hold that formal equality, and perhaps certain basic rights to one’s “most necessary natural needs” (RL 6:326), are sufficient to establish one’s status as sui iuris.
18 Compare ZeF 8:349–50; TP 290–4; RL 6:314.
19 ZeF 8:350n, likewise in RL: no person has a “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” (RL 6:314).
20 As we discuss, there is an ambiguity in Guyer’s claim that one must extend to others an “analogous right” to property. “Analogous” could be construed materially or formally. We have argued that Kant is concerned with formally equal rights, not with materially equal rights.
22 Guyer 2000b:239. Likewise: “the system of property rights must work for the benefit of all” (2006b:271).
23 Guyer 2000b:254. Kant provides support for a sufficientarian reading in his discussion of redistributive taxation, RL 6:325–6; though it should be noted that 1) this passage is from Public Right, not Private Right, and 2) Kant does not allude to the property theory he developed in Private Right to support his argument for redistribution.
24 Perhaps that is what Guyer means when he writes that “a system of property rights can be freely agreed to by rational beings only if it is equitable to some suitable degree.”
25 Furthermore, although Kant himself develops his property theory as a theory of specifically private property, nothing in Kant’s view seems to us to rule out the possibility of social ownership (socialism). Since modern socialism did not develop until the nineteenth century, it is no surprise that Kant does not address this possibility.
26 Kant also suggests at RL 6:326 some rudiments of a welfare state that would ensure that all citizens have their “most necessary natural needs” met, which again suggests that all will be better off within the civil state than they would be without it.