2. The Constitutive Relation between Public Law and External Freedom

I begin from the fact that Kant regards peace as a matter of right, rather than of ethics: ‘This rational idea of a peaceful, even if not friendly, thoroughgoing community of all nations that can come into relations affecting one another is not a philanthropic (ethical) principle but a juridical principle (ein rechtliches Prinzip)’ (MM, 6: 352).Footnote ³ A juridical principle is a principle for the public and positive legal ordering of external relations between persons, rather than a principle which persons must make the motive of their actions (6: 213–14, 218–21). Insofar as peace relates to the community of all nations, to speak of it as a juridical principle or a principle of right indicates that it is a principle for the right ordering of a global public legal order. Moreover, to speak of peace in terms of right means that it is a condition that must be established by agreement and one that must provide assurance of compliance to the parties. Peace cannot merely be based on good will or sympathy; it requires the rules and assurance provided by a lawful condition (TPP, 8: 349).

Next, the idea of equal external freedom is evidently central to Kant’s philosophy of right. Indeed, Kant writes that right ‘proceeds entirely from the concept of freedom in the external relation of people to one another’ (TP, 8: 289). According to the constitutive interpretation, this notion of freedom cannot be seen as an independently conceived end for which a public legal order is a mere instrument. The distinctive feature of Kant’s notion of external freedom – freedom as independence – is its internal relation to a public legal order (rechtlicher Zustand). External freedom entails independence from the unilateral or private will of others but at the same time it requires ‘the dependence of all upon a common legislation’ (8: 349–50, emphasis in original; see also MM, 6: 316). Only by submitting to a common public legal order can a multitude of human beings whose choices affect one another avoid being dependent on the unilateral will of another. According to Kant, ‘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, is the only original right belonging to every man by virtue of his humanity’ (MM, 6: 237, emphasis changed). And this ‘principle of innate freedom already involves … innate equality, that is, independence from being bound by others to more than one can in turn bind them; hence a human being’s quality of being his own master (sui juris)’ (6: 237–8).

Note then that, in Kant, coercion is internal to the very concept of right. Indeed, ‘Right and authorization to use coercion … mean one and the same thing’ (MM, 6: 232). The external relation between right and coercion follows from Kant’s notion of freedom as a matter of not being constrained by the choice of another. The only way in which I can enjoy freedom as independence is if I am assured that other persons’ interference with my choices is obstructed. Because freedom as independence concerns external freedom and not internal freedom or autonomy, it can be regulated by coercion, and because freedom as independence requires assurance and security of freedom, ‘as a hindering of a hindrance to freedom’, it requires coercion (6: 231). The fundamental concern and normative core of Kant’s legal philosophy – freedom in the external relation of people to one another – cannot then be based on mere goodwill or a balance of powers. Freedom and individual rights require common legislation and coercion: ‘[H]owever well-disposed and law-abiding human beings might be, it lies in the rational idea of such a condition (one that is not rightful) that before a public lawful condition is established individual human beings, peoples and states can never be secure against violence from one another’ (6: 312).

Despite some family disputes, it is generally well understood among proponents of the constitutive interpretation that external freedom in Kant should be understood as relational, law-governed and dependent on a common, coercive will (Flikschuh Reference Flikschuh2017: 62, 71–87; Hodgson Reference Hodgson2010; Ripstein Reference Ripstein2009; Zylberman Reference Zylberman2016). However, a clear and elaborate argument for the connections between external freedom, common will and coercion is provided only for the case of domestic right. Thus, the (sovereign) state plays a prominent role in the constitutive interpretation of Kant’s political philosophy. More specifically, it is argued that a public legal order qua a state solves three problems in the state of nature, namely the problem of unilateral determination of rights, the problem of assurance and enforcement of rights and the problem of indeterminacy of rights (Ripstein Reference Ripstein2009: 145–81). The state provides procedures for establishing a common will which can lay down what is right, and which has the power to assure everyone subject to it of the security of their rights, as well as judge cases of indeterminacy. For Kant this is particularly important in relation to property rights, which have a conventional element and can be enforced in accordance with the coexistence of freedom only by a common will, which has no purposes of its own. If property rights were enforced by a unilateral will, some would be subject to the will or purposes of another: ‘So it is only a will putting everyone under obligation, hence only a collective general (common) and powerful will, that can provide this assurance’ – that is, the assurance of securing ‘freedom in accordance with universal laws’ (MM, 6: 256).

If we think that only a state can have a common, powerful will and if we add that, for Kant, in order for the state to be a state, the ruler must be sovereign and have no legal duties to his subjects ‘that he can be coerced to fulfil’ (MM, 6: 319), then it is tempting to conclude that Kant provides a strong argument for state sovereignty and exclusive territorial rights (Mikalsen Reference Mikalsen2017; Stilz Reference Stilz2011). However, it is a mistake to tie the idea of public law exclusively to the sovereign state, particularly to any old state. Right, in Kant, is not merely a description of the legal order of any particular state (6: 229–30; Höffe Reference Höffe and Newton2006: 82; Wood Reference Wood and Timmons2002: 6). First, Recht is a normative idea and not an empirical description. Right has normative elements; there are content constraints on what the sovereign can decide in order for its commands to be proper laws and constitute right. This is the reason why Kant posits a Universal Principle of Right (UPR) and an innate right of humanity (6: 230, 237). The UPR is the basic principle of Kant’s Rechtslehre. It 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’ (6: 230). As I explain below, the UPR is not addressed to each individual but concerns the proper ordering of a juridical or rightful state (rechtlicher Zustand). Such a state or condition must secure the coexistence of external freedom of individuals.

Second, ‘we have to take into consideration not only the relation of one state to another as a whole, but also the relation of individual persons of one state toward the individuals of another, as well as toward another state as a whole’ (MM, 6: 343–4). Right concerns not merely the internal legal order of a state (domestic right) but also the relationship between states (international right) and the relationship between states and foreign individuals (cosmopolitan right). In Kant, there is one ‘general concept of right’ with three levels, and ‘if the principle of outer freedom limited by law is lacking in any of these three possible forms of rightful condition, the framework of all the others is unavoidably undermined and must finally collapse’ (6: 311).

The innovative idea of Kant’s philosophy of right is that public law is not justified merely as the best means to a predefined end, but rather as creating a specific relationship of freedom between individual human beings. There is a risk that the constitutive interpretation fails to incorporate this insight when it comes to international and cosmopolitan right. If the state fully and conclusively constitutes individual freedom, the two other levels of right can at best have an instrumental relationship to individual freedom. Therefore, I propose that we regard the constitutive relationship between public law and freedom as a matter not only of the relationship between the state and its citizens, but between right at all three levels and ‘freedom in the external relation’ of all human beings.

3. The Duty to Continual Approximation to Peace

Arthur Ripstein’s Force and Freedom (Reference Ripstein2009) is paradigmatic for the development of an interpretation of Kant’s legal and political philosophy that concentrates on the constitutive relationship between public law and a relational view of freedom. I say this despite misgivings about an alleged ambivalence in Ripstein’s position (Flikschuh Reference Flikschuh2017: 78–81; Zylberman Reference Zylberman2016). It is a core idea in Ripstein’s interpretation of Kant that public law is not a mere means to an independently defined end but, rather, constitutes a system of equal freedom among mutually dependent human beings. For Kant, ‘[public legal] institutions and the authorization to coerce are not merely causal conditions likely to bring about the realization of the right to freedom … Instead, the consistent exercise of the right to freedom by a plurality of persons cannot be conceived apart from a public legal order’ (Ripstein Reference Ripstein2009: 9, cf. 33–5). For present purposes, we can leave possible departures from this idea aside. Ripstein’s Kant interpretation has clearly influenced Anna Stilz’s (Reference Stilz2011) argument for the territorial rights of legitimate states and other arguments for the importance of state sovereignty in Kant (e.g. Mikalsen Reference Mikalsen2017). Ripstein himself gives little space in his book to the international and cosmopolitan levels of right (2009: 225–30, 295–8), but focuses his interpretation on the relationship between domestic right and freedom as independence (Reference Ripstein2009: 10).

Ripstein (Reference Ripstein2009: 225–30) argues not merely that Kant has some historical and empirical hesitations regarding the need for legislative and executive (coercive) powers beyond the nation state but that there is in principle no need for such powers at the international and cosmopolitan levels of right. This is the case, he argues, because only one of the three defects of the state of nature among individual human beings is present in the relations between sovereign states. According to Ripstein (Reference Ripstein2009: 227), in his argument about international right Kant ‘offers an analogue neither of the assurance argument nor of the argument from unilateral choice’. Internationally, there is only the problem of indeterminacy and hence only need for an international court. In a short discussion of cosmopolitan right, Ripstein (Reference Ripstein2009: 295–9) accepts that it entails a right to asylum, even if a state has no duty to accept outsiders as settlers. For Ripstein, neither international nor cosmopolitan right create any problems or dilemmas regarding the state’s supreme authority over its territory, that is, its sovereignty.Footnote ⁴

Ripstein provides two reasons why, internationally, there is no issue of either unilaterality or assurance, both of which I will challenge. First, he argues that there is no problem of unilateral acquisition among sovereign states because a state’s territory is like a person’s body, which it does not acquire but simply has. As a result, the state’s territory is not in need of omnilateral authorization (Ripstein Reference Ripstein2009: 227–8; cf. Mikalsen Reference Mikalsen2017: 16). While it is true that Kant speaks of the state as ‘a trunk [with] its own roots’, and ‘as a moral person’ (TPP, 8: 344), I do not think this is sufficient to show that Kant regards a state’s territory as innately possessed. Indeed, in Toward Perpetual Peace, Kant distinguishes between the way of acquiring a territory (Erwerbungsart) and the status of possession of a territory (Besitzstand), and he writes that we must accept current status of possession, not because it has ‘what is required in order to be called a right’ but rather because their way of acquisition was ‘taken to be legitimate according to the public opinion of every state at the time’ (8: 347). Kant can only speak of a state’s territorial possession in this manner, because he regards it as acquired and not innately possessed (Ypi Reference Ypi2012: 304). Moreover, the fact that Kant in this connection proposes laws regarding territorial acquisition indicates that he does think there can rise problems of unilateral acquisition.

Second, Ripstein (Reference Ripstein2009: 228–9) argues that no problem of unilateral acquisition will arise among states, because states as rightful conditions, in contradistinction to individual persons, have no private purposes. However, it is only from the perspective of its own members that the state can be seen as an omnilateral will which has no other purpose than establishing a rightful condition. Seen from the international and cosmopolitan perspective, the state is still a unilateral will when it enforces its right to territory (and its inhabitants’ property rights) vis-à-vis other states and their citizens (Reglitz Reference Reglitz2019). Kant seems to accept this when he writes that, while a state in relation to its own subjects is rightful, in relation to other peoples or states it is ‘simply a power’ (MM, 6: 311). Ripstein (Reference Ripstein2009: 228) argues that a state as a rightful condition ‘could never have grounds for’ an aggressive war. But since, as we have seen, Kant’s general argument for the need for positive, coercive law does not depend on how well disposed and law-abiding human beings are (6: 312), it is unclear why that should be different in the international realm.

Regarding cosmopolitan right, Ripstein (Reference Ripstein2009: 299) writes that Kant ‘provides an account of the right to refuge as a right of world citizenship. If you cannot go to your home state without being met with violence, any place of safety becomes your home, because, as Robert Frost put it, “they have to take you in”’. What is unclear is what kind of obligation on the part of the state we are talking about here. If the right to asylum indeed is a right, the state’s obligation cannot be a philanthropic or ethical obligation but must be a juridical obligation. The right to live in a rightful condition, the right of the asylum seeker, must be a juridical right in order to be a right and cannot depend on the goodwill of the state where she seeks asylum. In order for a refugee not to be dependent on a foreign state as simply a power, all states must be publicly and juridically obliged to take in refugees. Without a common, coercive will for the refugee and the states to which she could take refuge, there will be no security of hindering a hindrance to freedom for the refugee.

A different argument for regarding state sovereignty as essential is that the public legal institutions of a state are a great achievement, which must be respected (Stilz Reference Stilz2011: 592–5). There are three aspects to this argument: (1) state institutions protect rights and provide public goods; (2) such institutions are difficult to establish; and (3) they are a product of a people’s collective effort, which must be respected. These are important considerations and can be supported by Kantian principles and the constitutive interpretation. However, this type of argument proceeds as if we can solve the issue of public law and freedom one state at a time (Huber Reference Huber2016: 691). Thus, Stilz assumes that she can defend states’ territorial rights in a first step and deal with control of borders, international freedom of movement and resource privileges separately and ‘in a second step’ (Stilz Reference Stilz2011: 574). But this is to ignore the all-encompassing and systemic perspective of Kant’s tripartite division of right.Footnote ⁵ My point is that we cannot turn to global issues only after having settled the matter of legitimate authority domestically. Rather, we must think all relations among human beings and states into one system. Thus, while it is true that state institutions are difficult to establish and that they protect important rights, the questions that remain are (1) whether they can protect all human rights and (2) whether the very sovereignty of states threatens some human rights.

The shortcoming of arguments that focus on protection of state sovereignty and territorial rights and see no reason of right for common legislation and force beyond the sovereign nation state is that they fail to see the problems of right(s) that remain unresolved. Most worrisomely, they fail to recognize that these problems cannot be resolved if left to a separate and independent step. Furthermore, Kant clearly did see some issues of right as unresolved in a world without levels of proper right, which includes coercion, beyond the nation state. To be sure, Kant unequivocally rejects a unified world state (Universalmonarchie) (TPP, 8: 367), and he acknowledges that current states are not ready to accept a universal state of nations (Völkerstaat) (8: 357). Moreover, he seems to accept a league of nations (Völkerbund), which has no sovereign authority and which would work only as an arbiter in the disputes of states (MM, 6: 344, 350).Footnote ⁶ However, insofar as Kant settles on the latter, it is not because he thinks that there is in principle no need for more than this. Indeed, right and perpetual peace require more. In addition, while Kant writes that perpetual peace is unachievable, he does not stop there, as Ripstein does (Reference Ripstein2009: 230), but argues that states have a duty to ‘continual approximation to it’ (6: 350). This means that the non-coercive league of nations is not all that is required by reason or by Kant’s conception of right. The league of nations ‘is a right in subsidium of another original right, to avoid getting involved in a state of war among the other members’ (6: 344). Thus, states ought to go further than a league of nations and approximate ‘a universal association of states’ in which alone ‘rights come to hold conclusively’ (6: 350).

Kant is committed to the idea that the fact that something has not succeeded does not mean that it cannot succeed in the future. Nor does past lack of success mean that we do not have a duty to continual approximation to a rightful peace (TP, 8: 309–10; MM, 6: 350, 354; Kleingeld Reference Kleingeld2004: 315–18; Pogge Reference Pogge and Thomas2009: 201). We should not confuse what are the best means to protect rights today with our long-term duties in terms of a continual approximation to real peace and original right in all relations among human beings, including those rights that can never be realized by states that insist on protecting their own independence against the claims of non-members. Hence, I suggest that the constitutive interpretation of right must be combined with a developmental reading of Kant that emphasizes the transitional status of his argument for a non-coercive league of nations.

4. Moral Individualism and Conclusive Rights

The notion that international and cosmopolitan right do not depend on common will and coercion, as well as the conclusion that legitimate states have supreme authority over their territory, fails to appreciate what Katrin Flikschuh (Reference Flikschuh2010) calls ‘Kant’s sovereignty dilemma’. This is the dilemma that states cannot be subjected to coercion without losing their status ‘as that type of a moral agent whose will is juridically sovereign’ (Flikschuh Reference Flikschuh2010: 480; cf. MM, 6: 319), on the one hand, and that states must be subjected to a coercive international and global common will in order for there to be such a thing as international and cosmopolitan right. In other words, Ripstein’s argument simply ignores that the very concept of right in Kant entails common will and coercion (MM, 6: 232).

While I agree that one can detect a sovereignty dilemma in Kant, Flikschuh goes too far when she suggests that we find in Kant a ‘morality of sovereign states’, rather than a commitment to universal moral individualism. We might think that universal moral individualism belongs to a form of cosmopolitanism which regards legal cosmopolitanism as a mere means to realizing a moral ideal and thus fails to appreciate Kant’s distinctively constitutive, relational and juridical cosmopolitanism (Flikschuh Reference Flikschuh2010: 469–70; Reference Flikschuh2017: 2). There are two reasons to think that Kant is not committed to universal moral individualism that are relevant for our discussion. First, he speaks not only of the rights of individuals but also of the state ‘as a moral person’ that should not be made ‘into a thing’ (TPP, 8: 344). Second, Kant’s justification of right is not based on properties of individuals but on how persons morally relate to one another (Flikschuh Reference Flikschuh2017: 69–99). I argue that none of these reasons is sufficient to show that Kant is not a universal moral individualist. While it is true that Kant rejects one form of moral individualism, he does not reject all types of universal moral individualism. Indeed, we might learn from Kant to think of universal moral individualism in a different way than what has become standard. Thus, Flikschuh fails to appreciate another dilemma in Kant, which has to do with his commitment to both a morality of sovereign states and a form of universal moral individualism.

There is indeed a difference between Kant and contemporary liberal cosmopolitanism, as shown by Flikschuh, but I think it is misleading to say that it lies in Kant rejecting universal moral individualism in favour of a morality of sovereign states. It is of course true that Kant writes that the state should be respected as a moral person. However, this well-known paragraph from Toward Perpetual Peace ends with the admonition that treating a state as a mere thing that can be acquired by another state through inheritance, exchange, purchase or donation has the result that ‘the subjects are thereby used and used up as things to be managed at one’s discretion’ (TPP, 8: 344, emphasis added). The position that Kant stakes out here is important, but it would be a mistake to read it as an indication of a non-individualist view. The fact that Kant ends the paragraph in which he speaks of the state as a moral person with how the subjects are treated demonstrates that he agrees with the moral individualist view, according to which ‘all moral wrongs are ultimately wrongs to individuals’ (Altman and Wellman Reference Altman and Wellman2008: 285). The reason why a state should not be made into a thing is that this would make its members into things. Thus, in Kant, the state has no intrinsic value in the sense of ultimate value. Only individual human beings have ultimate value.

It might be thought that treating a collective entity such as a state as constitutive of freedom and understanding freedom in relational terms is non-individualist (Flikschuh Reference Flikschuh2017: 73–98). Thus, moral individualism is sometimes rejected because it is thought to entail that collective goods can have only instrumental value. Joseph Raz (Reference Raz1986: 198–203), for example, thinks moral individualism entails this view of collective goods and rejects it, because he thinks collective goods can also have intrinsic value. According to the constitutive interpretation, Kant rejects the idea that a public legal order has value only as an external causal means to individual ends. We have said that the relationship between the public legal order and individual freedom is constitutive rather than instrumental, but does that mean that a public legal order or state is intrinsically valuable? I argue that the public legal order or state is intrinsically good only in one of the three ways that things can be intrinsically good according to Raz. That is, the state can be a constituent good, because it creates relations among individuals the value of which depends on a public legal order. However, this does not mean that the state is valuable in itself ‘irrespective of what else exists’, or that the state has ultimate value in the sense that it can explain ‘the value of non-ultimate goods’ (Raz Reference Raz1986: 200). In Kant, the value of the state cannot be understood irrespective of the existence of individuals and its value cannot be understood as bestowing value on other, non-ultimate goods. Thus, even if the value of the state is not merely instrumental in Kant, its value is also not intrinsic, if we regard ‘intrinsic goodness’ to include unconditionality and ultimateness.Footnote ⁷

If we see right in Kant not as a description of an actual legal order, domestic or global, but as a normative idea, we must begin from what Kant writes about right in the ‘Introduction to The Doctrine of Right’. The Introduction provides the definition of right and can be seen as the moral ground of the entire doctrine of right, including international and cosmopolitan right. In the Introduction, right is defined as a relation between individual choices under universal law (MM, 6: 230), and the Universal Principle of Right mentions only individual persons. Moreover, there is only one innate right, and it belongs to individual human beings (6: 237). The UPR and the innate right to freedom fit badly with an idea of some collective entity being good in itself and having ultimate value. This is true even if we accept the constitutive interpretation, according to which freedom should not be understood as an individual possession, but rather as a rightful relation among persons. The only way in which we can give the state equal value to the individual human being is by positing a different concept of right for the international than for the domestic realm. However, for Kant there is only one concept of right, and it concerns the universal coexistence of individual freedom. Right concerns ‘that relation of human beings among one another that contains the conditions under which alone everyone is able to enjoy his rights’ (MM, 6: 305–6, emphasis changed).

Now, it may be accepted that the state’s moral personality derives from its role in protecting the rights of individuals and that Kant is a moral individualist in this sense, but that this individualism is a domestic rather than a cosmopolitan matter. Thus, the argument could be that Kant is not a universal moral individualist in the sense of taking a universal or cosmopolitan view of relations of freedom. Ripstein (Reference Ripstein2009: 296) seems to indicate something like this, when he writes that, ‘as a general matter’, citizens of one country have no obligations in terms of rights to outsiders. However, this conclusion is difficult to uphold for two reasons.

First, the right to live in a legitimate state, accepted by Ripstein (Reference Ripstein2009: 295–9) and Stilz (Reference Stilz2011), is not a purely domestic matter but entails a distinctively cosmopolitan perspective. It is a universal or cosmopolitan right pertaining to all human beings in virtue of their humanity. But not only that, when we accept a right to asylum, as Ripstein does, the obligation to ensure the right to live in a legitimate state falls not only on the state where one is or was a subject but on other states that ‘have to take you in’. Thus, excluding outsiders in some circumstances does interfere with their rights (pace Ripstein Reference Ripstein2009: 296). It is often pointed out that Kant limits cosmopolitan right to ‘conditions of universal hospitality’ (TPP, 8: 357) and that this entails that a foreigner cannot force himself on another state (Flikschuh Reference Flikschuh2017: 60; Meckstroth Reference Meckstroth2018; Ripstein Reference Ripstein2009: 296). But the right to hospitality also means that a state cannot unilaterally exclude visitors. The implication is that the relation ought to be one of public law rather than violence on the side of both state and visitor. Thus, in the Metaphysics of Morals, Kant describes cosmopolitan right as preceding from ‘a thoroughgoing relation of each to all the others’, and as a right to attempt to engage in interaction across borders without being treated as an enemy: ‘This right, since it has to do with the possible union of all nations with a view to certain universal laws for possible commerce, can be called cosmopolitan right’ (6: 352).

Second, it is unsatisfactory to regard domestic right as conclusively and fully settled independently of international and cosmopolitan right. It is unsatisfactory, first, because in an interdependent world, we cannot neatly separate the relations a state has to its own subjects from the relations it has to other states, and second, because a state not only has relations to other states as states but also to their subjects. Kant, of course, acknowledges these points. They are the reason for his cosmopolitanism. Thus, he writes, ‘it has now come so far with the (narrower or wider) community of nations of the earth that a violation of right on one place of the earth is felt in all’, and thus cosmopolitan right is ‘a supplement to the unwritten code of the right of a state and the right of nations necessary for the sake of any public human rights (zum öffentlichen Menschenrechte überhaupt) and so for perpetual peace’ (TPP, 8: 360).Footnote ⁸ If cosmopolitan right is necessary for human rights überhaupt, it makes no sense to regard domestic public right as a fait accompli that already conclusively secures and determines the rights of individual human beings.

We should consider here what exactly Kant takes ‘conclusive possession’ to be. More precisely, we must ask if an individual state on its own, with no regard to other states and their subjects, can determine conclusively the rights of its own citizens, particularly their property rights. Kant makes a distinction between ‘provisional rights’ and ‘conclusive rights’. In the state of nature, individuals may have provisional rights to property ‘in anticipation and preparation for the civil condition’, but these rights are only made conclusive by the common and coercive will of the state (MM, 6: 257). Thus, ‘possession found in an actual civil condition would be conclusive possession’ (MM, 6: 257). First, it is important to understand that Kant is not making a purely empirical argument reminiscent of Hobbes, which says that only under the sword of the sovereign is the property of the subjects secure. For Kant, the civil condition solves a moral, not merely an empirical problem (Ripstein Reference Ripstein2009: 164). That is, it solves the problem not only of establishing and monopolizing sufficient power to secure property and prevent conflict, but also the moral problem of realizing the principle that no one should be subject to the unilateral will of another. Thus, if the notion of conclusive possession is to cohere with Kant’s normative conception of right, it must satisfy the normative criterion presented in the UPR, as well as realize the innate right to freedom as independence. We see this also from the fact that when Kant in this context speaks of a ‘civil condition’ (im Zustande einer bürgerlichen Verfassung), he is not speaking of any old state but a state that does not ‘infringe upon … a priori principles for a civil condition’ (6: 256).

Second, if an individual state on its own could secure the rights of its subjects conclusively, it is unclear why Kant regards the two other levels of right, international and cosmopolitan right, as part of the same general concept of right. Alternatively, we have the possibility that these further levels of right are needed for merely instrumental reasons, for example, that peace would be a mere causal means to protect the individual states’ capacity to uphold their internal rightful condition. But this would make naught of Kant’s claim that ‘universal and lasting peace constitutes not merely a part of the doctrine of right but rather the entire final end of the doctrine of right within the limits of mere reason’ (MM, 6: 355). In the same section, Kant clearly sees peace in normative and categorical terms, and as a condition of the conclusiveness of property rights. Thus, he continues:

We see here both the individualist basis of Kant’s conception of right – its basis in relations of human beings – and that it would be a mistake to understand the international and cosmopolitan levels of right as independent of this basis, or as mere instrumental or secondary additions to the right of a state. The international and cosmopolitan dimensions of Kant’s philosophy of right must be included from the beginning as a systematic part of the concept of right and cannot be brought in only in a second step after domestic right and rightful relations between members have been conclusively established.

The preceding arguments entail the need for levels of right beyond the state that on the one hand are based on the principle of freedom in the external relation of individuals and on the other hand require a common and coercive will. Thus, we have not solved Kant’s sovereignty dilemma. Flikschuh (Reference Flikschuh2010: 488) tries to solve the dilemma by arguing that ‘a state which claims immunity from international juridical coercion on the grounds of its juridical sovereignty domestically is for that reason juridically obliged to enter into rightful relations with other states: its very claim to sovereignty domestically obliges it internationally’. However, insofar as there are no coercive powers and no common will internationally, it is unclear what Flikschuh means by ‘rightful relations with other states’. For rightful relations in Kant are not just relations that substantially follow some abstract moral principles but rather coercively bind everyone to common legislation. There is no solution here to the problems of common interpretation and assurance to settle disputes over territory, immigration, asylum, international trade and the like. To say that, because it is internally rightful, a state is obliged to follow moral principle externally or internationally is like saying that a moral person is obliged to act rightfully toward other persons. But in the state of nature among individuals Kant rejects this idea, because in such a situation ‘each has a right to do what seems right and good to it and not to be dependent upon another’s opinion about this’ (MM, 6: 312). As suggested earlier, it is difficult to see why we must not draw the same conclusion regarding international and cosmopolitan relations.

Additionally, in the ‘Introduction to the Doctrine of Right’, Kant is very clear that the principle of right is not addressed in the first person, as the categorical imperative is, but is a principle for a coercive public order. Kant writes that it cannot be required that I make the principle of right

Flikschuh’s proposed solution to the sovereignty dilemma turns right into something the individual states have an obligation to realize by limiting themselves, rather than it being a matter of establishing a common public legal order which limits the states from without, as required by Kant’s notion of right.

Without real rightful relations internationally and globally, that is, without common international and cosmopolitan public law, the rights of states and individuals will depend on the goodwill of others and on what seems right and good to others. Travellers, asylum seekers and traders will depend on the changeable will of the states where they go. In addition, states will have no assurance against the aggressions of one another except that a legitimate state as a rightful condition ‘could never have grounds for going to war except to defend itself’ (Ripstein Reference Ripstein2009: 228). However, as in the state of nature among individuals, the issue is not whether one can have reasons for violating others’ rights but that each has the right to follow her own opinion about what her rights are. Thus, applying Kant’s view on the state of nature among individuals to the international and cosmopolitan levels: however rightful states might be, before a condition of right (rechtlicher Zustand) is established globally, individual human being and states ‘can never be secure against violence from one another, since each has its own right to do what seems right and good to it and not to be dependent upon another’s opinion about this’ (MM, 6: 312).

5. Global Coercion and Legitimate States

The upshot of the preceding argument is that without international and cosmopolitan law backed by coercive power, states and individuals will be dependent on one another’s choices. Louis-Philippe Hodgson (Reference Hodgson and Ellis2012: 109) has similarly argued that a coercive (federal) world state can be defended on the premise ‘that states are entitled not to depend on one another’s choices’. Hodgson derives his argument for the right to the independence of states from Kant’s notion of freedom as independence for human beings. Thus, he writes,

Hodgson’s (Reference Hodgson and Ellis2012: 124) ‘Kantian argument for a world state’ as a precondition for ‘realizing external freedom’ assumes ‘that the world state’s authority to impose its rule through force would stop at the border of a legitimate domestic state’. And he assumes that each state on its own can ‘make rights conclusive for all agents dwelling on its territory’ (Reference Hodgson and Ellis2012: 108–9). This argument depends on a sharp separation between a state’s internal actions – how it relates to its own inhabitants – and the state’s relations to other states and their inhabitants. The federal world state, according to Hodgson, would have no authority regarding the first type of relations but only regarding the latter. A world state should only ‘have the authority to regulate matters concerning relations among states – territorial disputes, trade relations, immigration, and so on’ (Reference Hodgson and Ellis2012: 123). But can we really uphold this separation between internal and external relations of a state?

One difficulty is that issues like trade relations and immigration are not simply matters concerning relations between states. We must not forget Kant’s caution: ‘we have to take into consideration not only the relation of one state to another as a whole, but also the relation of individual persons of one state toward the individuals of another, as well as toward another state as a whole’ (MM, 6: 343–4). Moreover, if the world state has coercive authority to regulate trade relations, immigration and asylum, then clearly the individual states are no longer entirely independent to order their own internal affairs, and they can no longer conclusively determine the rights of their own inhabitants. Not only the states’ external but also their internal sovereignty will be affected by such coercive regulations.Footnote ⁹

In our contemporary world, many states insist on their sovereignty exactly in order to avoid immigration and global economic competition and to protect their own citizens’ independence from foreign influx and competition. On Hodgson’s independence argument, which includes a state’s right to determine its own citizens’ property rights independently and conclusively, it is difficult to argue against these sovereignty claims. Moreover, he cannot both uphold the state as conclusive authority on individual rights and allow a world state that coercively regulates trade and immigration. Global force can only stop at the border of the state if everyone stays in her own state, but that was hardly what Kant envisioned. The whole point of Kant’s cosmopolitanism is to theorize what is required in terms of right in a world where states and persons affect one another across state borders. In such a world, even coercive protection of state sovereignty is insufficient for securing that individuals are truly independent from the choices of others.

Remember that Hodgson (Reference Hodgson and Ellis2012: 124, emphasis added) argues that the coercive power of a global federal state ‘would stop at the border of a legitimate domestic state’. The same respect for the sovereignty of legitimate states we also find in Stilz’s (Reference Stilz2011) argument for states’ territorial rights. By ‘legitimate state’, they seem to mean, roughly, a state that respects the freedom or fundamental rights of their members. However, the question is whether this understanding of a legitimate state is adequate from the perspective of Kant’s tripartite understanding of public right. If right concerns not only a state’s relation to its own members but also the relations between states and the relations of states to non-member individuals, then clearly what defines a legitimate state must include all these relations. A legitimate state is not only one that fulfils its obligations to its own members; it must in addition fulfil its obligations to other states and foreign persons. Insofar as we are speaking about right, these are juridical and not ethical obligations. Juridical obligations cannot be defined by one’s own – or the individual state’s own – insight into the moral law but requires a public authority with a common, legislative will. In other words, the idea of a legitimate state depends on a common global public authority. Without such a global legal order, every person and state will depend for their rights on what foreign states take a legitimate state to be.

I have made two points regarding the idea of a legitimate state. First, we cannot define a legitimate state independently of how it relates to foreign states and foreign individuals. Second, the legitimacy of a state cannot merely be an ethical issue of goodwill but must be a juridical issue determined by a common will. Hodgson’s argument for a coercive federal world state accepts the need for a common will to determine when the relations between states are rightful and legitimate. But his argument fails when it comes to cosmopolitan relations that go beyond relations between states and concern relations of states and foreign individuals. This is clearest regarding the right of asylum, which I also discussed in connection with Ripstein’s position. If a state has an obligation to take in a legitimate asylum seeker, then this obligation must be a juridical obligation on the part of the state in order for the asylum seeker not to be dependent on the goodwill of the state. Thus, I suggest that the definition of a legitimate state must include its submission to the juridical obligation to accept legitimate asylum seekers. However, insofar as accepting asylum seekers affects how a state can provide for its own citizens’ rights and insofar as rights of asylum are laid down by a global common, coercive will, then the state cannot conclusively determine the rights of its own citizens without regard for the rights of outsiders. This is why cosmopolitan right is ‘a supplement to the unwritten code of the right of a state and the right of nations necessary for the sake of any public human rights’ (TPP, 8: 360).

Some might reply that cosmopolitanism without a strong principle of state sovereignty ‘risks becoming an ideology of powerful states in need of an excuse for going to war’ (Mikalsen Reference Mikalsen2017: 5). This is indeed an important concern, and it is true that Kant is wary of principles that could excuse aggressive warfare and colonialism (Meckstroth Reference Meckstroth2018). However, we must inquire further into the status of this argument. We should recall that for Kant peace is not an ethical principle that relies on the parties’ goodwill, it is a juridical principle that relies on establishing legal institutions that can provide assurance of compliance (MM, 6: 352; TPP, 8: 349). Thus, cosmopolitanism in Kant is not a set of principles that any state can use to fit its purposes. If a state uses ‘cosmopolitan principles’ as an excuse for going to war, it has not in fact acted on Kantian principles. It has acted unilaterally and without right. Just as institutions of domestic right are necessary to avoid unilateral and violent acquisition and protection of property according to one’s own ideas of right, institutions of international and cosmopolitan right can be justified because they establish juridical relations at the global level in order that no state can unilaterally determine when right requires the use of violence. It is exactly the point of establishing common global institutions of right to avoid that any state can have an excuse for going to war. To be sure, Kant also advances a second-best argument to the effect that, under present conditions, pushing for cosmopolitan institutions of right will be counterproductive and increase rather than diminish conflict and war. However, this pragmatic argument does not invalidate cosmopolitan principles and our duty to continual approximation to them.