1 Introduction

Kant’s theory of peace is likely the most popular part of his critical philosophy and certainly the one that reached the most diverse audiences, with an impact extending beyond the circle of interpreters or academics, a significant influence on policy makers, and concrete repercussions in the building of major regional and global institutions.Footnote ¹ There are obvious reasons behind this popularity. One is the intrinsic importance of a philosophical project meant to overcome what Voltaire wittingly described as the self-inflicted human tragedy: war. Another is the attractiveness of a model based simply on three major reforms: (1) the overcoming of despotism through the republicanization of national regimes, (2) the building of a peace-promoting supranational institution, and (3) the recognition by all states of the cosmopolitan right of each human being to visit other states. Yet another reason is the style that Kant adopts, significantly more direct and enjoyable than the one reserved for the three critiques. When one takes a closer look, however, the model reveals a number of puzzles, problems, and difficulties: some long debated among generations of Kant scholars, others less analyzed but relevant to conceptualizing some important political issues of our world.

This Element has two major goals. On the one hand, it introduces the reader to Kant’s theory of peace and to the way in which it fits within the broader context of the critical philosophy (Sections 1–3). On the other hand, it digs into one aspect of the model related to the “federalism of free states” (ZeF 8:354), an aspect that has generated much debate among Kant interpreters, given its centrality to the entire project and Kant’s changing attitude toward the nature and the powers of the supranational institution that he advocates as a necessary ingredient of his model (Sections 4–8). The Element offers an interpretation of the Kantian federation that in many regards departs from the mainstream reading. Special emphasis will be placed on the problematic coexistence of two conflicting theoretical desiderata: on the one hand, the necessity of establishing an international institution, either in the form of a world republic or a federal state with coercive powers, for securing peace; on the other hand, the necessity of avoiding the risk of an excessive erosion of states’ sovereignty.

These two major goals will be pursued in seven steps. In the first and second sections, the Element illustrates Kant’s project of perpetual peace. Kant thought that if we want to abolish forever interstate wars we need to globally implement three major institutional reforms: At the domestic level, all states should overcome despotism and become republican (in a sense that will be explained in due course); at the international level, all states should enter a supranational institution meant at the very least to serve as a forum for handling controversies peacefully; and at the cosmopolitan level, all states should recognize the right to visit on the part of foreigners entering their territory. These three reforms are the object of the three “definitive articles,” the backbone of Kant’s entire project. During the process leading to a global implementation of these reforms, political leaders should follow six additional rules that Kant calls “preliminary articles,” mainly meant to decrease the likelihood of war and keep international relation above a certain degree of civility, so that that very transition does not prove too difficult. Kant complements his project with two supplements and an appendix in two parts. Since excellent commentaries on Toward Perpetual Peace already exist, I will mention the essentials of the project and quickly move to the discussion of its most problematic and fascinating aspects, particularly those still in need of interpretive clarification.Footnote ² Among these are: (a) the far-reaching theoretical consequences of the “division of work” between preliminary and definitive articles in terms of what we would today call non-ideal and ideal theory; (b) the ensuing survival in the model of some space for just war theory despite Kant’s harsh criticism of the “sorry comforters” (ZeF 8:355) Grotius, Pufendorf, and Vattel; (c) the probabilistic nature of the argument resting on citizens’ aversion to war (Reference CarantiCaranti 2017a, Reference Guyer and CarantiGuyer 2006); (d) the question of whether only republics or any kinds of states are allowed to become members (Reference CarantiCaranti 2017a, Reference CavallarCavallar 2015, Reference DoyleDoyle 1983a, b, Reference MacMillanMacMillan 1995, Reference RawlsRawls 1999); and (e), the question of whether cosmopolitan right (the right to visit) is supposed to contribute to the cause of peace by merely enabling international economic interdependence or, more broadly, by fostering peoples’ reciprocal knowledge and understanding.

The third section focuses on the place of Kant’s peace project in the architectonic of the critical philosophy. I will address the question of compatibility between some aspects of the model and Kant’s epistemological and moral commitments as presented in the three critiques. On the one hand, the focus will be on Kant’s idea that nature “guarantees” that humankind, willingly or unwillingly, will reach a global durable peace. This thesis will be defended – to the extent that this is possible – in the face of three major challenges – epistemological, moral, and anthropological – raised against it (Reference Caranti, Goldman, Patrone and FormosaCaranti 2014, Reference Eberl and NiesenEberl and Niesen 2011, Reference Guyer and CarantiGuyer 2006, Reference Ludwig and CarantiLudwig 2006). On the other hand, the focus will be on the relationship between knowledge, morality, and politics. It will be highlighted how Kant interprets politics as relatively independent of morality, prescribing that politicians (especially the moral ones) interpret their role as irreducible to a mechanical application of the principles of right to the complexity of human affairs. It will be shown that moral politicians need to avoid two opposite and yet equally dangerous epistemological excesses: on the one hand, the disregard for the repercussions of their pursuing justice in a specific context, fueled by an acritical endorsement of the “fiat justitia, pereat mundus [Let justice be done, though the world perish]” maxim; on the other hand, the ultimately paralyzing attempt to predict with any degree of precision the consequences of one’s political initiative, an ambition that merely mimics the responsible attitude of the moral politician while in reality it turns out to be closer to a technical understanding of politics that Kant rejects and attributes to what he calls “political moralists.”

After having illustrated Kant’s project of perpetual peace and discussed its place in the critical system, Sections 4–8 focus on international right, with the second definitive article’s prescription to build a “federalism of free states” (ZeF 8:354); that is, an international federation among states to overcome – at least to a certain extent – the lawless, hence war-prone condition of international relations. Section 3 reconstructs the evolution of Kant’s thought on international right. A textual analysis will show how Kant reshapes the notion of international right itself over time, with a major breakthrough in Toward Perpetual Peace. Moreover, it will be shown that Kant progressively abandons the idea of conceiving the exiting from the state of nature at the international level in strict analogy with the domestic case, and thus moves from a preference for the institution of a supranational institution with coercive powers, either in the form of a world state or in that of a state of states, to one based on normative, not merely empirical or pragmatic considerations, for a league of states without coercive powers.

Section 5 discusses the arguments Kant offers to justify his preference for the federation of nations (Völkerbund) over the state of nations (Völkerstaat). I will distinguish three lines of argument – empirical, logical, and moral – and will suggest that Kant’s strongest argument in favor of the federation is the moral one, which turns on the moral personality of states, in turn ultimately resting on the freedom of citizens.

Section 6 confronts some of the most interesting readings on the federation versus world republic issue offered in the first two decades of the twenty-first century. I will focus on four major critics of Kant’s preference for the federation over the ideal of the world republic: Thomas Pogge, Jürgen Habermas, Pauline Kleingeld, and Otfried Höffe. Independently of their differences, they all propose an “enhanced” league or minimal world republic as the kind of institution Kant either defended or should have defended, given the function the federation is supposed to play.

In replying to these critics in Section 7, I will suggest that: (a) Kant never abandoned the idea that the only true guarantor of peace is a world republic or a federation with coercive powers; (b) the reason why he ended up advocating for a voluntary league without coercive powers is not that he “did not/could not” think of the solution of breaking down sovereignty, with some competences transferred to a supranational institution and others kept for nation states; it is rather that he correctly saw that at least in the crucial area of using force toward external actors – the area that matters most for a theory of international peace – a state may not compromise on its exclusive competence, if it is to remain sovereign at all; c) Kant thus favored the solution of the voluntary league because he wanted to avoid the merging of all states into one, a goal shared also by those critics who reproach him for his lack of courage and/or consistency; (d) the league, while by itself not a guarantee, would still be a significant peace factor because it would offer a permanent channel of diplomatic relations where states have the chance to have their controversies discussed and assessed by something like a third, independent party – the partial overcoming of the international state of nature through a court recently highlighted by Ripstein (Reference RipsteinRipstein 2009: 228); and (e), despite his preference for the federation, Kant never ruled out the possibility – in fact he kept hoping – that states might voluntarily choose to dissolve in order to enter a larger global institution.Footnote ³ Yet I will suggest that it is not entirely clear from a Kantian perspective how the transition should unfold without violating the basic rights of citizens who may oppose it.

Section 8 deals with the reading of Kant’s international right offered by Katrin Flikschuh. This reading is consonant with our suggestion that states cannot divide up their sovereignty as easily as Habermas and so many others assume. Yet, in overlooking how this extends to the strict impossibility on the part of states to compromise on their exclusive competence to go to war, I will argue that Flikschuh does not get to the heart of the paradox Kant is facing.

Both the overview of Kant’s theory of peace in Sections 2–3 and the relatively detailed analysis of international right offered in the remaining sections are carried out by staying close to Kant’s text. This should enable readers to grasp the intricate and interesting problems that Kant’s model reveals, despite its apparent simplicity.

2 Kant’s Project of Perpetual Peace

Zum Ewigen Frieden: Ein Philosophischer Entwurf, published in 1795, was probably the greatest literary success Kant had in his life. Shortly after its publication, translations in French and English appeared, and by around the end of the century the essay had gained centrality in the intellectual debate in France and Germany. By the time Kant died (nine years after the book’s appearance), his “philosophical project” had reached its tenth edition. After that, its popularity started to decline. Probably the cosmopolitan orientation of the essay, meant as an antidote to the excesses of nationalism, as well as its ahistorical, abstract, universal definition of the just state displeased the taste of an increasingly Romantic intellectual milieu. After various ups and downs of fortune, with “returns to Kant” especially in the aftermath of the two world wars – when both the League of Nations and the United Nations were inspired by Kant’s plea for a global federation – a number of factors, including the birth of democratic peace theory in the early 1980s; the fall of the Berlin Wall (with the subsequent third wave of democratization), and the occasion of the 200th anniversary of its publication gave the essay a new elan. Today, one can hardly dispute that Kant’s project (or better, the mainstream interpretation of it) is part of the background knowledge not only of philosophers but also of international relations scholars, historians, political scientists, and, most importantly, political leaders of contemporary liberal democracies.

Kant articulates his project according to the structure typical of the international treaties of his time. We thus have preliminary and definitive articles, supplements, and a two-part appendix. Undoubtedly the definitive articles are the center of the entire work. Here Kant provides his “philosophical project” in favor of peace. The other parts, albeit peripheral, are crucial for a proper understanding of the definitive articles themselves.

3 The Project of Perpetual Peace in the Context of the Critical Philosophy

The three definitive articles unequivocally constitute the backbone of Kant’s project. This does not mean, however, that they can be read in isolation – as they often have been. They are obviously embedded in a broader philosophical project constituted by Kant’s critical philosophy as a whole. To begin with, Kant’s model evidently rests on his teleological view of human affairs. One aspect of this view, or perhaps I should say the most extreme (hence clearest) part of that view, is the famous thesis according to which nature provides a “guarantee” that humanity will one day reach perpetual peace. It is fair to say that it is this thesis that has caused most problems of compatibility between Kant’s peace project and his overall philosophy to be noticed. Another point of tension is undoubtedly the relation (or apparent disagreement, as Kant says) between politics and morality. Without passing over the extent to which some tension remains, I am going to suggest that no real incompatibility occurs in either of the cases.

4 The Evolution of Kant’s Thought on International Right

Very few interpreters deny that Kant has some hesitations when it comes to designing the nature and powers of the supranational institution that should contribute to world peace, along with republican institutions within states and the right to visit, recognized by states, of every human being. Even sympathetic scholars, who believe that Kant never changed his mind and always consistently preferred the option of a federal state of states, recognize that he designed different models: the single world state, the federal state of states, and the league of nations or confederation (Reference Byrd and HruschkaByrd and Hruschka 2010: 196–205). They also acknowledge that at least in the Doctrine of Right Kant advocates a fourth model of a “permanent congress,” which they themselves construe as even weaker than the league of nations, because while the latter rests on some multilateral treaty signed by member states, the former remains in a precontractual stage (Reference Byrd and HruschkaByrd and Hruschka 2010: 204–5).

In fact, the very idea that Kant does not change his mind over the years on the point at issue is hard to defend. Kant first suggests the necessity of a supranational institution for the sake of peace in Idea for a Universal History with a Cosmopolitan Purpose (1784). As noticed by Mori, even if the alternative between a federation and a world state is not formulated, the reasons behind Kant’s oscillation are already present (Reference MoriMori 2008: 107–8). Kant first stresses the analogy between the domestic and the international case in the move from the state of nature to the civil condition. The same unsocial sociability that forced individuals to enter the civil condition leads peoples and states to abandon a lawless state and to enter “a federation of nations, where every state, even the smallest, could expect its security and rights not from its own might, or its own juridical judgment, but only from this great federation of nations (Foedus Amphictyonum), from a united might and from the decision in accordance with laws of its united will” (IaG 8:24). The logic of the analogy and the reference to a “united might,” and, even more forcefully, to a “united will,” suggest that Kant has something more in mind than a voluntary and contingent defense alliance. And yet the reference to the Foedus Amphictyonum goes in the opposite direction, because this was merely an alliance of neighboring Greek poleis to protect and maintain a religious site. Elsewhere in the essay Kant describes the federation as a “large state body” (IaG 8: 28) and as a “perfect state constitution” (IaG 8: 27). Both descriptions again suggest something similar to a world state, and yet the reference to “a law of equilibrium” (IaG 8: 26) to regulate the essentially healthy hostility which prevails among states, as well as the characterization of the federation as a “cosmopolitan condition of public state security” (IaG 8:26), which is “not wholly without dangers” (IaG 8:26), indicate that Kant has no intention of overcoming the state of nature at the international level in perfect analogy with the domestic case, but merely intends to establish a security system based on bilateral or multilateral treaties signed by members that remain fully sovereign. The fact that this system is characterized as “cosmopolitan” should not be taken as evidence of an endorsement of something more ambitious than a tentatively permanent defense alliance. In 1784 Kant has not yet distinguished international from cosmopolitan right with full clarity (Reference CarantiCaranti 2017b).Footnote ²⁸ Indeed, at that time Kant does not have a notion of cosmopolitan right as a distinct and independent branch of right that has nothing to do with the federation. Indeed, in the Feyerabend lectures (1784), there is no mention at all of cosmopolitan right. Moreover, there Kant says that “the jus gentium is merely the possibility of a Völkerbund” (V-NR/Feyerabend 27: 1393).Footnote ²⁹ Hence, characterizing the supranational institution as “cosmopolitan” does not necessarily mean that Kant is advocating something more ambitious than an international federation that encompasses all states. Obviously, the ambiguity regarding the very nature of the institution translates itself into an ambiguity in terms of the powers the institution is supposed to have. It is not clear whether Kant in 1784 is ready to concede at least a partial transfer of national sovereignty to a higher political authority for the sake of peace.

The 1793 essay “On the Common Saying” introduces for the first time a clear distinction between “cosmopolitan constitution” under a common sovereign and “a rightful condition of federation under a commonly agreed upon right of nations.” Kant says that the latter is to be preferred, if the former proves itself even more dangerous to freedom than the status quo “for it may lead to the most fearful despotism” (TP 8: 310–11). The first hermeneutical problem here is to understand whether, as it is at times argued (e.g., Reference MoriMori 2008: 109), Kant intends the cosmopolitan constitution as the solution dictated by theory and the federation as the one suggested by practice. If this is the case, Kant would be grossly incoherent because the general goal of the essay is precisely to show that what is true in theory (read: dictated by practical reason) can always be realized and should never be abandoned on the basis of “practical” considerations. However, there is no clear indication that the cosmopolitan constitution is the only option suggested by theory. Kant appears to be presenting the alternative – that is, the federation under international right – as more palatable to a single world state and yet fully still consistent with what theory demands. This suggests that what he has in mind is not merely a voluntary confederation, but a federation with coercive powers. Moreover, the fact that Kant describes the supranational institution dictated by theory as subject to “enforceable public laws to which each state must submit” is too thin to serve as evidence to conclude that he is talking about a world state. A federation with coercive powers established by multilateral treaties among states would still satisfy that definition. Not accidentally, in a Reflexion (8065) from the eighties Kant explicitly describes the alternative to the world state; that is, the federation as that institutional setting that guarantees “the freedom of each state under a universal law” (Refl 19: 600 my emphasis).Footnote ³⁰

To be sure, in 1793, on the crucial point of what shape and powers the supranational institution required by theory is supposed to have, Kant does not seem to hold firm beliefs. The example of a solution that is true in theory is what “the theory of Abbé St Pierre or Rousseau” (TP 8: 313) proposed. Now, while Rousseau may have endorsed a European federal state, that is a federation with coercive powers, St Pierre, whose position Kant knew well, never dreamed of proposing either a world state or a federation with coercive powers. His “Grand Alliance” was nothing but a defensive league and his “European Assembly” nothing but a permanent seat for diplomatic intercourse, that is, a (con)federation without coercive powers.

But if the federation (with or without coercive powers) is not the solution merely dictated by practice, what does practice suggest then? The “practical” surrogate to the solution true in theory is not the federation under international right, but the status quo or the highly unstable “balance of powers in Europe” (TP 8:313). Far from betraying the general orientation of the essay, this position squares perfectly with it.

If up to “On the Common Saying” Kant remains noncommittal as to the exact form the supranational institution is to take, from Toward Perpetual Peace (1795) on, his preference for the federation over the world federal state is stated clearly. Moreover, the federation itself becomes unequivocally “light.” In the 1795 essay, right after having introduced the second definitive article, Kant clarifies that what must be established is a “league of peoples” which need not be a “state of nations” (ZeF 8:354). And the reason for this preference ‒ roughly, the fact that the very notion of one international state is inconsistent ‒ comes close to a repudiation of his previous opening up to the possibility of a world state. In the Metaphysics of Morals (1797), Kant will end up denying any coercive power to it, while affirming the freedom of member states to leave the federation at will.

Things are not so simple, however. Within the same essay Kant keeps recognizing the world-state/state-of-nations option as the rational one, or, as he puts it, the one true in thesi: “In accordance with reason there is only one way that states in relation with one another can leave the lawless condition, which involves nothing but war; it is that, like individual human beings, they give up their savage (lawless) freedom, accommodate themselves to public coercive laws, and so form an (always growing) state of nations (civitas gentium) that would finally encompass all the nations of the earth” (ZeF 8:357) In other words, the rational way to prevent war is to endorse that international state (Völkerstaat) which Kant had described a few lines before as intrinsically contradictory. Moreover, the reasons that lead Kant to abandon what is true in thesi to endorse what is possible in hypothesi are fascinating. Since states, according to their conception of international right, place limitations on what is right in thesi – that is, they veto any solution that includes their disappearance as independent and sovereign entities – “in place of the positive idea of a world republic only the negative surrogate of a league that averts war, endures, and always expands can hold back the stream of hostile inclination that shies away from right, though with constant danger of its breaking out” (ZeF 8:357). The surrogate of the league is the solution possible in hypothesi because states are what is “posited below” (literally, in Greek, hypo-tithemi means “to put below”): They are already given and with their legitimate interest in remaining in existence, constitute a formidable obstacle to what is right in thesi.

While in the 1793 essay Kant conceded the existence of difficulties but remained confident that what is right in theory must also be possible in practice, now the fact that something is right in thesi is no longer a sufficient guarantee regarding the possibility of its implementation and can be put aside because of the obstacles states “posit” on the way toward the world state. So this is a double shift: on the one hand, from a noncommitment as to the best institutional form to an open endorsement of the federation as the only possible solution. On the other hand, it is a shift from trust that what reason commands is certainly feasible, to just the opposite: Something may be true in theory but does not apply in practice. Finally, if in 1793 Kant still thought that the supranational institution, no matter whether in the form of a single world state or a federated state of nations, had to be endowed with coercive powers to enforce the common laws binding all members, he is now careful to clarify that if the federation has to have any power, this is not analogous to the power the civil state has over citizens: “This league does not look to acquiring any power of a state but only to preserving and securing the freedom of a state itself and of other states in league with it, but without there being any need for them to subject themselves to public laws and coercion under them (as people in a state of nature must do)” (ZeF 8:356). Finally, and this point is crucial, unlike the world state, the federation may be feasible, but there is a price to pay. While the former guarantees that there will be no war, the latter merely makes war less likely (ZeF 8:357).

As anticipated, Kant’s evolution toward a clear preference for the federation over the world state culminates in 1797’s Metaphysics of Morals. Here Kant repeats that only a world state is capable of guaranteeing peace, by transforming international relations from a condition in which by necessity there can be only provisional right to a condition in which right becomes peremptory, in perfect analogy with the exit from the state of nature at the domestic level. And yet not only is the non-oercive nature of the federation reaffirmed (“the alliance must … involve no sovereign authority (as in a civil constitution)” (MS 6:344) but also its permanence in time is put at risk. While it remains true that, unlike a contingent gathering of states, the association is to be conceived as a “permanent congress of states,” the federation “can be renounced at any time and so must be renewed from time to time.” And to reinforce the point, Kant offers two further considerations. On the one hand, he indicates the 1719 Assembly of the States General at the Hague – a gathering of the major European powers ultimately leading to the 1720 Treaty of The Hague – as a good example of the federation as it should be understood.Footnote ³¹ On the other hand, he warns against conflating the federation with the United States of America, “which is based on a constitution and can therefore not be dissolved” (MS 6: 351).

If this is the evolution of Kant’s position, it may be of interest to identify the models that Kant seems to have entertained and evaluated in this long and difficult journey. Byrd and Hruschka think there are three models, but I think that one can reasonably distinguish at least five, listed below in descending order of institutional ambition:

1. 1. Single world state (TP 8:311), world republic (ZeF 8:357), or universal monarchy (ZeF 8:367). Transfer of states’ sovereignty.

2. 2. Federal state of states/nations (Völkerstaat or allgemeiner Staatenverein) TP 8: 312, 313; ZeF 8: 354, 357; MS 6: 311, 350. Transfer of states’ sovereignty.

3. 3. Federation with coercive powers (Föderation), different from the Völkerstaat because it is not a unified political body, does not have legislative authority on all areas, and yet there is a power (likely of judicial nature) issuing binding resolutions of controversies among members on the basis of the existing treaties, and possibly some form of executive power to execute sentences. (TP 8:311, Refl 19:600). Transfer of states’ sovereignty on specific areas.

4. 4. Federation without coercive powers or league (Völkerbund or equivalent expressions used by Kant such as Friedensbund, Föderalität, freier Föderalism, föderative Vereinigung), likely generated through international treaties with, so to speak, no expiration date, hence tentatively permanent. This is roughly equivalent with what we today call a mere confederation (ZeF 8:354, 356, 357). No transfer of states’ sovereignty.

5. 5. “Permanent” congress of states (Staatencongreß), still in a precontractual stage, to be continuously renewed from time to time and dissoluble because members are allowed to pull out anytime (MS 6:350–1). No transfer of states’ sovereignty.

Of course, the complexity increases if we distinguish between republican or despotic forms of the models that presuppose centralized coercive powers (models 1–3). There is also the crucial question of whether states can be coerced to participate in any of the models indicated above. On this very important point, Byrd and Hruschka are positive that for Kant states may be coerced (Reference Byrd and HruschkaByrd and Hruschka 2010: 194). They cite as evidence an indirect, which they call “formal” argument, and a more direct, explicit one. The first is that the postulate of public right, which introduces the general obligation to leave the state of nature and enter a civil condition, applies to all articulations of right. Indeed, in the Doctrine of Right Kant presents the postulate before he moves on to distinguish between the three juridical dimensions, and this suggests that it applies to all three of them. Interestingly, a similar point was made by a very early critic of Kant, Johann Gottlieb Fichte, who pointed to a footnote in Toward Perpetual Peace (ZeF 8:349) in which Kant has another version of the same postulate. Fichte, however, makes the point in order to reach the opposite conclusion; namely, that Kant is not faithful to his postulate because, when it comes to international relations, he is ready to make an exception. Fichte realizes that Kant wants (more or less consistently) the obligation for states to leave the state of nature to coexist with their freedom not to enter any supranational institution, especially if this has coercive powers and therefore promises to erode their sovereignty. And that this is the case is quite evident from what Kant says as early as 1795, when he describes the enlargement of the federation as starting with a powerful and enlightened people that establishes republican institutions and serves as a “focal point of federative union of other states, to attach themselves to it and so to secure a condition of freedom of states conformably with the idea of the right of nations; and by further alliances of this kind, it would extend further and further” (ZeF 8:356). Kant’s language suggests a gradual and spontaneous process as opposed to one in which all states are simultaneously forced into the federation.

The second argument is that in paragraph 54 of the Doctrine of Right Kant says that states living in a state of nature are obliged to leave it (MS 6:344). But, of course, this does not seem to be enough evidence to conclude that Kant wants states to be coerced into respecting this obligation.Footnote ³² In fact, the opposite seems to be true. A few lines later, Kant indicates that the supranational institution which states are to enter is the league of nations that “can be renounced at any time and so must be renewed from time to time” (MS 6:344). Thus, on Byrd’s and Hruschka’s reading, Kant in this passage is saying that states can be coerced into entering an institution which they can leave at any time. This would be quite bizarre.

To sum up: Kant changed his mind over the years about the nature and powers of the supranational institution that he indicates as the second pillar of his model for peace. He leaned more and more toward a thin and loose league, which states have only a generic obligation to enter without being coerced to do so, and that they may dissolve at any time. However, Kant never changed his mind about the fact that the only guarantee of peace would be a world republic or a federation with coercive powers. This is obviously compatible with the idea that he ended up preferring, an institution that at best decreases the chances of war and that consequently needs the concurrence of the other two dimensions of right to deliver something approaching a guarantee of peace.

5 Three Lines of Argument in Favor of the Federation

Having reconstructed the evolution of Kant’s position on the powers and nature of the supranational institution, the most fascinating and yet most difficult point still remains to be addressed. Why did he come to prefer the federation over the world state? Do his reasons stand up to scrutiny, or is the dissatisfaction expressed by many readers from very early on (Fichte) justified? Kant has profoundly different arguments to ground his preference, from empirical-pragmatic, to normative, to logical ones. Indeed one can distinguish three main lines of thought: (a) empirical-pragmatic arguments (a belief in the administrative inefficiency of a state covering too large a territory combined with its alleged tendency to move toward a soulless despotism and possibly a final stage of anarchy); (b) a logical argument (in the very idea of a state of states lies a contradiction, because a world state removes the very notion of international right); (c) a moral/juridical argument resting on the limits of the domestic analogy (states may not be forced to abandon the state of nature and enter a juridical condition because, unlike individuals in the state of nature, they already have a moral personality).

The first line of thought starts from an empirical claim concerning the past experience of an inverse relation between administrative efficiency and extension of the territory over which a government rules: “as the range of government expands laws progressively lose their vigor” (ZeF 8:367). This is combined with another claim, no less empirical, regarding the necessary degeneration of such states. A “universal monarchy,” but likewise also a state of states, would first degenerate into a “soulless despotism,” probably in the desperate attempt to implement its rule over so large a territory. Alternatively, such a system would degenerate into anarchy, with the emergence of factions which the central authority had no sufficient power to keep in check. This is obviously not acceptable because anarchy is the state of nature which the supranational institution was meant to avoid. As Kant puts it, “if such a state made of nations were to extend too far over vast regions, governing it and so too protecting each of its members would finally have to become impossible” (MS 6:350). At times Kant does not make the last step in the argument from despotism to anarchy, because he takes as self-evident that even if a universal despotic government were capable of implementing its laws, they would still be the “graveyard of freedom” (ZeF 8:367), thereby guaranteeing peace at the cost of denying the supreme value (freedom) that all articulations of right (including the international one) are supposed to secure. In addition, and this may count as a further twist in this pragmatic line of thought, a world state would remove that healthy hostility among states which is necessary to prevent the powers of humanity “fall[ing] asleep” (IaG 8:26), and thereby hindering the cultural and moral improvement of the species.

As rightly noticed by Mori (Reference MoriMori 2008: 117), no matter how sound these empirical claims are, they hardly harmonize with the general orientation of Kant’s thought.Footnote ³³ If a world state were “right in theory” one would not expect Kant to say that it is to be abandoned as a political goal only because experience suggests that it is difficult to realize in practice. The fact that large empires in the past have followed the path indicated by Kant could never be a sufficient reason for making what is right nonobligatory. As Kant himself puts it: “experience cannot teach what right is” (TP 8:306). And even more forcefully, what reason prescribes as right does not follow “from perceiving the ways of the world” and it remains obligatory “even though no example of this could be found” (MS 6:216). So any empirical argument meant to show either that the world state is too ambitious a political goal or inefficient/prone to degenerating should not be taken, at face value, as Kant’s last word of the matter.

What about the second line of thought? The contradiction argument, presented in Toward Perpetual Peace, starts from the lawless condition of international relations and the obligation each state has to overcome it, by entering a league of nations. Kant is keen to clarify that this league

Kant’s reasoning is now that a world state, in establishing – like any other state – the relation between a superior and an inferior would end up denying the independence and sovereignty of states, thereby merging all of them into one. This would contradict the assumption that we are looking for an international institution, that is, something that preserves the plurality of juridical subjects signing the contract.

Three things are worth noticing. To begin with, notice Kant’s language: “need not be a state of nations [kein Völkerstaat sein müßte]”. If there is a contradiction in the very idea of a state of nations, then one would expect Kant to say something stronger to rule it out, like “may/should not be a state of nations.” The fact is that, as I have said repeatedly, Kant never abandons the idea that the only true solution to the problem of war is precisely either the single world state or the federal state of states. And this emerges even when he is about to present an argument that supposedly spots a contradiction in that very idea.

Secondly, one could say that Kant does not consider the possibility of a federal state of nations, which allegedly would preserve the plurality of states while establishing a federal power above them. Obviously Kant does consider (and rule out) this possibility in the Doctrine of Right, when he clarifies that the league he wants should not be conflated with the United States of America. But by the light of the contradiction argument only, the impression is that Kant overlooks the possibility of a federal state that does not merge all states into one. This criticism is well grounded and yet – I believe – does nothing but reveal the ultimate reservation Kant seems to have toward anything more than a loose federation. This is the idea that if the federal state advocates competence for itself over the use of force against external actors, even if it leaves competence to states in other areas, then member states’ sovereignty is completely lost. I will come back to this point because it constitutes the essence of my reading of Kant’s preference for a league of states. For the moment, it is sufficient to notice that it is unlikely that Kant simply did not think of the federal option. He rather already considered it – albeit implicitly – incompatible with the sovereignty of the federated states.

Thirdly, even if this is correct, one may still ask Kant why we should preserve a plurality of states. What is it about nations that is so inviolable that it prevents us from imagining anything except a solution that overcomes them? Certainly, a world state would preclude the healthy competition among states that we saw in the pragmatic arguments, but this is too thin a basis. After all, much of the competition that stimulates human progress may happen between individuals, as we know from the domestic case. Even if the competition between peoples (or states) adds something to that between individuals, is the gain significant enough to give up the only institutional solution that would guarantee the permanent overcoming of war?

Alternatively, and more convincingly, one may try to boost Kant’s argument by establishing a link between the existence of a plurality of states and the freedom of individual citizens living in those states. As we saw, Kant says that in a state a relation has already been established between a superior and an inferior. One way to understand this is that once a state is created, a rightful condition (more or less perfect depending on how institutions approximate the republican ideal) is also established. This juridical condition is the guarantee (again, more or less perfect) of the freedom of the consociates. They are part of a contract through which they had at least some of their rights and freedoms secured. A melting of all states into one, or even the erection of a federation with a central power and ultimate authority on possible disputes between federated states, would only be compatible with individual freedom if citizens freely renounced the juridical condition they had established in their domestic state and decided to enter a new juridical condition, this time at the global level.

Eberl and Niesen makes a similar point when they argue that a world state would violate the individual freedom of citizens already belonging to established domestic states, especially if the existing states in question were republics (Reference Eberl and NiesenEberl and Niesen 2011: 240).Footnote ³⁴ This, however, does not explain how Kant can rule out through merely logical or normative arguments a world republic, where, at least in theory, individuals would have their sphere of freedom protected and secured just as well as they have it in domestic republics. I suggest that if one takes this question seriously then Kant’s best argument, by no means conclusive and yet quite powerful, which rules out anything more than a loose federation as a viable option, becomes evident. What is key is to focus on the normative importance of states as “hypotheses,” understood as what is already “posited below,” as I explained in Section 4. The fact that states already exist is not a mere empirical accident that can be overcome by following blindly what reason commands and what peace requires (the world state or the federal state of states). Since states as “hypotheses” (ZeF 8:357) have already established a limited juridical condition in which a group of individuals has signed a pact empowering that specific national commonwealth to protect their freedom and inborn right(s), and to do it in one peculiar manner among the many possible, the burden to protect their freedom can be transferred to a higher, global institution only if the parties agree to annul the original contract and enter a new one. For sure, states cannot be forced to enter, because what reason commands at the international level cannot ignore what it commanded at the domestic level when the domestic juridical condition was created.

This introduces us to the third line of thought, which paves the way for Kant’s most convincing case for the federation. The argument is based on the limits of the analogy between the domestic and the international case regarding the duty to abandon the state of nature:

Kant evidently thinks that the fact that states have established a lawful condition “inside” makes their case different than that of individuals in the state of nature. While in the latter case there is still no trace of legality (at most there are innate natural rights of individuals to be protected through the rule of law), at the international level each commonwealth is already a legal entity, with a right to exercise its autonomy. As noticed by Reference MoriMori (2008: 119–20), Kant expresses this conception of states’ autonomy in the important Reflexion 8065 (Refl 19: 600) from the 1780s that we already cited, where the condition of individuals in the state of nature is likened to that of accidents in need of a substance (the civil constitution), while that of states is considered already a condition of substance. A world state would downgrade substances to accidents. It follows that we have to discard the world-state option and adopt the federation, where each nation state keeps its sovereignty.

One important thing needs to be noticed. The legal autonomy already established internally does not remove the obligation to overcome a lawless condition in which each state is a constant threat to others. If there is an institutional setting that at the same time removes or mitigates that threat and maintains each state’s autonomy, the state’s obligation appears to be as strict as that of individuals. Even conceding that the analogy between the domestic and international case has limits, this at most removes the possibility for one state to coerce another to join a supranational institution under common laws, not the obligation for each state to do that spontaneously. Indeed, the autonomy of each state can never remove this obligation:

6 Cosmopolitan Concerns: A Neglected Middle Term Between Völkerstaat and Völkerbund

Since the first appearance of Toward Perpetual Peace, many readers of Kant, including sympathetic ones, have been dissatisfied with his ultimate preference for a federation rather than the world republic or the federal republic of republics as a viable political option.Footnote ³⁵ Particularly interesting is the position of moderate cosmopolitans who think that Kant could have endorsed a more ambitious solution without endangering the plurality and autonomy of states. Something along these lines animates four authoritative readings that have tried to show that, after 200 years of political experiments, we can aim higher than the mature Kant ultimately came to believe was possible. They probably constitute the best criticism of the federation available in the literature, and, more importantly, come close to the last-resort attempt to vindicate the possibility of a cosmopolitan solution to Kant’s peace project. While in our opinion ultimately unsuccessful, they deserve attention because they reveal a contrasto the strength of Kant’s arguments in favor of the federation.

Thomas Pogge (Reference Pogge and CarantiPogge 2006) argues that Kant was blind to the possibility of something more ambitious than a loose federation because he was still operating with a modern conception of sovereignty, a conception based on the idea that sovereignty can only be absolute: undivided and unlimited. From Aquinas via Dante, Marsilius, Bodin, Hobbes, and Rousseau all the way to John Austin in the past century, the idea has always been that if sovereignty is limited, we have no sovereignty at all. Indeed, if a sovereign is to operate only within a limited mandate, who is to decide whether a certain decision legitimately falls within the scope of that mandate? We have only two possibilities. If the decision rests with the sovereign itself, then sovereignty is unlimited because the sovereign is the ultimate judge about the limits of its competence. If it is an external agency (imagine a supranational court), then the sovereign has no power because any exercise of its authority would depend on the supranational agency’s acquiescence.

Pogge rebuts this classic manner of conceiving of sovereignty by noticing that in contemporary liberal democracies there is no institution within the system that exercises full sovereignty: Neither the parliament, nor the supreme court, nor any other institution has ultimate authority to decide possible conflicts of competences. This is even more evident – we might add – for the liberal democracies belonging to the EU, which not accidentally is considered by Pogge as a good example of the federation Kant should have endorsed. The sovereignty of member states is limited by the competences of the EU in many different areas and yet one could hardly deny that member states are in a substantial manner still sovereign or respected in their political autonomy. The EU has exclusive competences in the areas of the Customs Union; rules necessary for the functioning of the internal market; monetary policy for euro area countries; conservation of marine biological resources under the Common Fisheries Policy; and common commercial policy. But it has only a shared (with nation states) competence in the single market; employment and social affairs; economic, social, and territorial cohesion; agriculture; fisheries; the environment; consumer protection; transport; trans-European networks; energy; justice and fundamental rights; migration and home affairs; public health; research and space; development cooperation; and humanitarian aid. While all this certainly amounts to a diminution of sovereignty, there are still areas in which member states can exercise their political autonomy.

Jürgen Habermas offers, in two important pieces of writing, a similar reading of Kant’s model. He argues that Kant and many of his interpreters start from a false alternative. Between the world republic, that Kant defends in the 1793 “Theory and Practice” essay, and the weak league of nations without coercive powers that he comes to support from 1795 on, there is a neglected middle term that would accommodate Kant’s persistent cosmopolitan ambition without any risk of degeneration toward a global autarchy (the famous “soulless despotism”). Suggested by the experience we have “at two hundred years’ historical remove” (Reference HabermasHabermas 1998) from Kant, this is “a politically constituted global society that reserves institutions and procedures of global governance for states at both the supra- and transnational levels” (Reference HabermasHabermas 2006: 135). Major international organizations (the UN, the World Trade Organization [WTO], the EU) exhibit a tendency, still in fieri, toward taming state power not through the monopoly of force of a supranational entity such as the world republic or even the “republic of republics,” but through the progressive transfer of competencies, previously reserved to the state, to higher institutional bodies. Even if member states remain free to leave the organization (and Brexit proves this is more than a logical possibility) the stability of the supranational organization is strengthened by states’ interest in remaining within the organization itself. In other words, as Habermas puts it, the

The question of what kind of supranational institution Kant’s theory of peace needs, quite independently of the form Kant ultimately endorses, is at the center of another influential reading worth considering for its qualified defense of the “world republic” option. In an essay devoted to our theme (Reference Höffe and HöffeHöffe 2004) and in the third part of Kant’s Cosmopolitan Theory of Law and Peace (Reference HöffeHöffe 2006), Otfried Höffe argues that the federation without coercive powers is helplessly insufficient to realize a juridical condition capable of securing an innate right to freedom for all human beings while at the same time delivering lasting peace. Kant’s endorsement of the federation is useful only as a reminder that Kant certainly rules out what Höffe calls “an exclusive cosmopolitanism.” Only a cosmopolitan order that “does not supplant national civil law, but supplements it” (Reference HöffeHöffe 2006: 140) can be considered as a solution. The necessity of a word republic, however, remains. A global institution with some of the powers reserved to states, hence with a minimal state form, is the sole true guarantee of individual spheres of freedom at the global level and of lasting peace. The global institution will have to guarantee the minimal legal equality among individuals sufficient for nondomination. Kant must then accept some analogy between a cosmopolitan legal order and the legal order of a state, but only to a point. Like Habermas, Höffe is keen to emphasize that the analogy between individuals forming a state and states forming a world republic is the wrong way to extend at the international level the duty to create a legal order. In contrast to “primary states,” the world republic “is assigned only a narrow range of powers; it is then a minimal world state” organized along republican lines. Such a world republic “demands ceding state sovereignty, but only to a minimal extent” (Reference HöffeHöffe 2006: 203). Thus, “the public safeguarding of rights in the legal form” culminates in the task of creating a subsidiary and federal world republic (Reference HöffeHöffe 2006: 195).Footnote ³⁶

The debate on the exact kind of institution Kant ultimately wants or should want for reasons of logical consistency and normative credibility (world republic, republic of federated republics, politically constituted global society, league of states with or without coercive powers) is also at the center of Pauline Kleingeld’s reading. In a series of publications culminating with her Kant and Cosmopolitanism (Reference KleingeldKleingeld 2011), Kleingeld rejects the common idea that Kant from 1795 on abandons the international federation with coercive powers as an ideal to be pursued. She holds that while Kant has good reasons to reject the world republic option, he consistently defends the solution of a plurality of federated states. The voluntary league is nothing but a first step in a process toward an international federation with powers and competences far superior to those commonly associated with a loose league (Reference Kleingeld and CarantiKleingeld 2006). While Pogge and Habermas assume that the modern dogma of sovereignty blinds Kant to a solution in which states alienate some of their sovereignty (mainly the unilateral power to decide when to wage war), but maintain full autonomy in all other political areas, Kleingeld thinks that Kant was not blind at all to this solution: The loose league of state is nothing but a step in an evolution toward a federation gradually and spontaneously taking upon itself more and more competences that were previously reserved to nation states, or better, receiving them as a result of peoples/states deciding freely to transfer them.

By suggesting a provisional role for the federation, this reading harmonizes the Völkerbund-Völkerstaat alternative, making the first a preparation for the second. First proposed by Fichte, who considered the federation nothing but an intermediate step (Mittelzustand) through which humanity should be able to reach the real target, that is, the world republic (Reference Fichte and FichteFichte 1845–6: 433), this reading still has a considerable hearing among contemporary readers. For example, Howard Williams argues that Kant “is both advocating an international state as the ultimate goal, but not advocating it as something to be realized in the immediate or near future. It is an objective to put to the back of our minds, but it is an objective we ought always to have in mind” (Reference WilliamsWilliams 1986: 256).Footnote ³⁷ However, crucial to this line of thought is the fact that nation states do not dissolve in the federation with coercive powers. States retain their sovereignty, which at least for Kleingeld includes their right to exit. In the next section I am going to argue that unfortunately the idea of a federation with coercive powers is incompatible with states keeping their sovereignty if states transfer competence as regards the military; and that if states do not do so, then the federation does not contribute to peace significantly more than a loose league does.

7 Tertium Non Datur: The Problem with Moderate Cosmopolitanism

Independently of their differences, Pogge, Habermas, Kleingeld, and Höffe all indicate the “enhanced” league or minimal world republic as the solution that Kant could and should have defended. This reading is attractive in that it promises to relax the conceptual dilemma Kant faces between guaranteeing peace through a dissolving of nation states into a world republic (or republic of republics) and preserving a plurality of states through a loose league by renouncing the ideal of a permanent and stable peace. As we saw, Kant never abandoned the idea that the only true guarantor of peace is a world republic or a federation with coercive powers. There is no reason to believe, however, that he ended up advocating for a voluntary league without coercive powers because he “did not/could not” think of the solution of breaking down sovereignty with some competences transferred to a supranational institution and others kept for nation states. Before charging Kant with the inability to see this elegant and convenient solution, one should wonder whether the proposal is truly viable. I will argue that the moderate cosmopolitan solution ignores the depths of the problem Kant is facing. And I further suggest that, in addition to the argument on the moral character of states, the nonviability of this solution is what led Kant to prefer the federation over anything institutionally more ambitious. The crucial point is that when it comes to who decides about using force against external actors – unfortunately the area that matters most for a theory of international peace – a state may not compromise on its exclusive competence, if it has to remain sovereign at all.

The point can be proven theoretically and illustrated through one very significant case study. Theoretically, it is sufficient to notice that any transfer of competences by a state to a higher body remains provisional as long as the state retains full sovereignty over its own defense policy, or, which is the same thing, over the use of force toward external actors. Military force is the last thing a state may appeal to in case it wishes to exercise its will against that of others, be they lower, higher, or same-level institutional bodies, partners within an organization, or nonstate actors, and so on. Think of a state as part of an international organization that is either unsatisfied with a certain interpretation of the rules of that organization or simply no longer wants to remain a member. Further imagine that it declares its intentions, but remains unheard by the organization at large and/or by its partners. How could it exercise its will if it does not control an army that could at least threaten the use of violence against opposing actors? Without full control of the army, there is no sovereignty, not even partially.

A major empirical confirmation of the theoretical point above comes from the history of EU’s integration process. The EU is particularly significant as a case study because it is the closest thing we have to the enhanced federation favored by moderate cosmopolitans. As Hueglin and Fenna argue, its “institutional framework has evolved gradually but steadily from intergovernmentalism to supranationalism” (Hueglin and Fenna 2015: 5), thus making the EU closer and closer to a federal, as opposed to confederal, body, in fact, as a case of the success of federalism. They continue:

In order to appreciate how sovereignty and exclusive competence on defense policy are indissolubly connected, it is instructive to look at the EU’s continued failure to establish, beyond proclamations and good intentions, a common defense policy. Despite its “gradual but steady” progress toward federalism, and despite repeated efforts and some progress made at an incredibly slow pace if compared with economic and monetary integration, the Union still lacks a common voice in foreign relations and, most importantly, in matters related to the use of force.Footnote ³⁸ The Treaty on European Union does give the EU competence to define and implement a common foreign and security policy (CSDP), including the progressive framing of a common defense policy. And yet it is the European Council and the Council of the European Union (Article 42 TEU) that take the decisions relating to the CSDP. These two collegiate bodies, respectively the forum of member states’ prime ministers and the forum of competent officials (ministers, ambassadors) from the executive of each member state, make decisions normally by unanimity, which obviously means that each state normally retains full sovereignty.

If one looks at the history of past attempts to achieve better integration in the matter of defense, the picture is rather clear. Europeans have tried many times to go beyond mere coordination in military affairs and parallel the economic, monetary and to a certain extent political integration they were building in the post–World War II era. And yet, something has always gone wrong, and never by accident. We start with France’s refusal to ratify the European Defence Community (EDC) in 1954. As noticed by Rosato, “Sovereignty concerns … meant that the defense community was never particularly popular in France. … The preexisting NATO system was far preferable on this score: although it was tightly organized for operational purposes, the United States, Britain, and France retained sovereignty over their own forces” (Reference RosatoRosato 2011: 63). While economic and monetary integration progressed in the years leading to the collapse of the Soviet Union, with the consolidation of an internal market over and beyond a custom-free zone, military integration languished, aggravated by the evaporation of the major geopolitical factor (the threat posed by the USSR) that could have provided an impetus to create an integrated defense. Indeed, with the USSR gone, Europeans felt confident enough to establish a common defense policy for the first time independently of their American ally, which in practical terms meant the creation of a structure that was an alternative to NATO. But again, its results were limited.

At the 1999 Helsinki summit, European leaders appeared to have taken decisive steps in this direction, contingently motivated by the way in which the air strike in Kosovo conducted by NATO forces, notably the United States, painfully exposed the inability of the EU to back up its diplomatic and negotiating efforts with adequate military power. This led to the creation of the European Security and Defence Policy (ESDP), which was renamed the Common Security and Defence Policy (CSDP) in 2009 by the Treaty of Lisbon. The CSDP was further developed in 2016 with a global strategy for the foreign and security policy of the EU, designed by EU leaders in the aftermath of the Russian annexation of Crimea, Brexit, and Trump’s promise to diminish the US’s military commitment within NATO.

What were the results of this effort at a common defense policy? It would be unfair to claim that the EU is wholly absent in the international arena. Since 1999 the EU has launched thirty-four missions and operations in fields ranging from conflict prevention through peacemaking, peacekeeping, crisis management, joint disarmament operations, and military advice and assistance to post-conflict stabilization. Currently, the EU is present on three continents through the deployment of ten civilian and six military missions. Yet even the most ardent fan of the EU knows that (a) the intergovernmental nature of the EU defense has never been overcome; (b) even today the EU is unable to show a united military power capable of working as leverage in what are currently the hottest contexts, even those clearly falling within its sphere of influence (just think of Turkey and Russia gaining control and influence in the Middle East and North Africa); (c) after seventy years of good intentions and fanfare, and despite the creation of structures such as the Permanent Structured Cooperation (PESCO), the European Defence Fund (EDF), and the High Representative of the Union for Foreign Affairs and Security Policy, we are not even close to anything resembling a Europe with a common voice in foreign affairs and a common army to back it up.

This less-than-exciting outcome cannot be accounted for in terms of chance or geopolitical considerations working against the result of a common European defense. If the fall of the Soviet Union removed one pressing reason to unite against a common threat, we have come far enough to have reasons to be threatened by new autocratic powers (mainly Russia with its quasi-satellite states such as Belarus, and China), and yet this is not sufficient to move beyond enhanced cooperation to true integration. One major reason, if not the reason for this is that member states can renounce full control of the use of force only by forfeiting the last and deepest level of national sovereignty. Such a move would end the EU as we know it and turn it, to the joy of many past and present continental federalists, into something like the United States of Europe. As long as there is an alternative way of guaranteeing security – and NATO provides such an alternative – European states will never move beyond fanfare about military integration, because they know that this is the only way to avoid their end as independent political entities.Footnote ³⁹

The lesson one should learn at this point is quite evident. Kant consistently ended up favoring the solution of the voluntary league, given that he wanted, like moderate cosmopolitans, to avoid the merging of all states into one. The league does not provide any guarantee of peace and yet it is to be preferred because it does not erode the deepest layer of sovereignty of states. Moreover, while unable to provide a guarantee of peace, the league is still a significant peace factor because it constitutes a permanent channel of diplomatic relations where states can have their controversies discussed and assessed by something like a third independent party, even if this judicial authority may have no ultimate power to enforce its decisions (a reluctant state may simply leave the federation if it refuses to comply). The contemporary experience of many intergovernmental organizations (IGOs), mixed or fully democratic, from the Arab League to the Assocation of Southeast Asian Nations (ASEAN), the Organization of American States (OAS), and the African Union (AU), proves precisely how institutions of this kind can do something to mediate controversies between members, even if they are in no position to prevent war in all circumstances. This limited contribution to peace would of course increase its own impact to the extent to which the other two factors of Kant’s model make progress, that is, members of the voluntary league become republican and economic and cultural interdependencies among them grow, thereby making war very costly, like in the case of the EU. While Kant never abandoned the idea that a guarantee of perpetual peace requires the full overcoming of the international state of nature in perfect analogy with the domestic case, he came to realize that states are not merely self-interested entities stubbornly refusing to comply with what practical reason demands. They are not mere “accidents,” as the Reflexion cited above nicely illustrates. They have a moral personality ultimately resting on the juridical condition that was created the moment (no matter how imaginary and hypothetical) in which a formerly scattered group of individuals followed the first-order command of practical reason, abandoned their specific (domestic) state of nature, and created a commonwealth. Their choice and the specific manner in which their freedoms and rights were (more or less) secured within that specific state deserve moral and political respect. Normative/moral, not merely pragmatic reasons set limits to the analogy between the domestic and the international case.

This is obviously compatible with the possibility that states freely decide to dissolve themselves to enter either a world republic or a federal state of states. This is the limited amount of overlapping that my reading has with the above-discussed provisional reading sponsored by Fichte, Kleingeld, Williams, and others. The point we share is that nothing in Kant prevents the possibility of a free and uncoerced overcoming of the international system in favor of some form of the world state. In fact, much in his philosophy encourages this development. The difference between my reading and the others is that I do not think that the problem is merely one of insufficient moral education on the parts of political leaders and/or citizens. In my reading, citizens of a state who refuse to dissolve their nation into some higher authority are not morally defective, because they can make the argument that what they are defending was generated by compliance with the first-level command of practical reason. Hence, it is entirely within the legitimate scope of their freedom to reject the dissolution of their nation.

This difference has consequences for the different way in which I read the possibility of a spontaneous and free overcoming of nation states in favor of the world republic or other forms of cosmopolitan authority. While champions of the provisional reading are rather elusive on the conditions that must be met to make such a move legitimate, I argue that such conditions must be spelled out carefully and are rather demanding. Here a comparison with a recent discussion in the theory of human rights may prove instructive. Many experts today argue that human rights should make room for a people’s choice of a nondemocratic form of government. Thomas Christiano, for example, thinks that instrumental and intrinsic considerations suggest that there is a human right to democracy, understood as a right to perfect civil and political equality (Reference Christiano, Cruft, Liao and RenzoChristiano 2015). At the same time, however, a state may choose to adopt a nondemocratic form of government without violating that individual right, but with a very demanding condition: All citizens (not a majority) need to waive their right to democracy. I wonder whether something similar holds for the decision to dissolve a state. The original decision of a people to form a state may be compatible with a later decision to dissolve it. But what kind of majority would be necessary for that? A simple one with at least 50 percent voter turnout? A qualified one? A vote that is unanimous, at least in principle, in the sense that there are reasons to believe that those opposing the dissolution are somehow wrong? Certainly, from a Kantian perspective, there should be clear reasons to believe that the decision springs from the general will, not from the interests of one faction, no matter how large.

At this point moderate cosmopolitans may argue that the general will cannot fail to endorse a decision that would guarantee the security of all citizens of the state. And yet, this would problematically blur any distinction between universal practical reason and the general will of each domestic context. The latter, taking into considerations its Rousseauian pedigree, is after the common good of a people, not of humanity.Footnote ⁴⁰ And as long as the will of a people favors a condition of local political autonomy over one of certainty about world peace, it would be hard to construe this preference necessarily as the preference of a faction or as the unanimous yet inauthentic decision inspired by the volonté des tous.

Cosmopolitans could also legitimately protest that the case under consideration is not different than the one applying to individuals before the constitution of the state. If they have a duty to establish a domestic state to preserve freedom and peace, and if they can be forced to enter one even if they are unwilling to do so, are they not under the same duty to move one step up and complete the transition toward a cosmopolitan condition in which the state of nature is finally overcome, not only internally but also externally? If the decision is ultimately that of individuals and not of the states to which they belong, are we not back to the possibility of coercing even those individuals who refuse to take the last step necessary for the establishment of a true cosmopolitan condition? Possibly so. We touch here what has been aptly described by Katrin Flikschuh as the sovereignty dilemma of Kant’s notion of international right. To this last point of our analysis, I now turn.

8 The “Sovereignty Dilemma” Reconsidered

Flikschuh notices a “sovereignty dilemma” affecting Kant’s notion of international right. For Kant, right is analytically connected to coercion. Moreover, sovereign states are the exclusive and supreme agents in charge of enforcing right. Hence they cannot be legitimately coerced by other states or supra-state institutions to enter a federation and to abide by the rules of its membership with no option to abandon it. At the same time, in the domain of international relations practical reason prescribes that states establish an international juridical condition as a necessary component of justice. Therefore, concludes Flikschuh, “Kant’s theory of Right simultaneously requires and prohibits the juridical compulsion of states” (Reference FlikschuhFlikschuh 2010: 471). Flikschuh further suggests that the dilemma can only be solved if we move away from the logic of analogy/disanalogy between the domestic and the international case, and focus on the fact that in Kant, right is a system in which each part (domestic, international, or cosmopolitan) cannot stand without the others. If we assume this perspective, then the domestic component cannot do without the other international and cosmopolitan right. This means that states cannot welcome the unlimited sovereignty that right assigns to them while ignoring the demands the very same right puts upon them at the international and cosmopolitan level. If right is a system made of three interdependent components, states cannot, so to speak, pick and choose among them.

This is an interesting reading but, by the light of what we have already observed, the sovereignty dilemma does not seem to get to the heart of the problem that Kant is confronting. The fact that states can never be coerced from the outside is obviously compatible, as we saw, with the possibility that they freely decide to transfer some of their competences to the supranational institution. This is indeed what Flikschuh suggests they can and should do to avoid the “picking and choosing” described above. But Kant’s real and more profound dilemma is that if states renounce or diminish – even freely – their exclusive competence in one particular area – the use of force toward external actors, which is the competence that must be transferred if war is to be overcome – they simply are no longer sovereign.

Notice that this point is not dependent on a Weberian conception of sovereignty as a monopoly of force, because Weber himself thought that in certain areas the state may allow its exclusive authority to be limited without compromising its sovereignty. One must subscribe even less to a Schmittian conception, which rests on the ability to declare the state of exception (Ausnahmezustand). Rather, the point is a Hobbesian one, with an important qualification. For Hobbes, a soverign state is one that keeps ultimate authority over deciding whether the terms of whatever agreement it has entered have been respected. We share this conception and qualify it by noticing that this is equivalent to saying that the state is sovereign if it keeps ultimate control over the force that it may need to use to counter any action by an external agent perceived as either violating existing agreements or posing an existential threat. A state is sovereign if it keeps exclusive competence over the use of force toward external actors (be they other states or supranational institutions). If it loses that competence and yet still reacts to a perceived injustice, its reaction can only be conceptualized as a violent rebellion within a state. In other words, we would already be in a world state or in a federal state of states.

This notion of sovereignty, which, I am suggesting, Kant endorses, is in stark contrast with the necessity, to which Kant also subscribes, of a supranational institution with coercive powers (the point emphasized in particular by Höffe), which obviously implies that states may not have at their disposal a military force to counter the supranational institution. In this sense, far more radical than the one suggested by Flikschuh, Kant’s conception of right simultaneously requires and prohibits the juridical compulsion of states. Only if the depths of the paradox are fully appreciated can one understand fully the normative reasons that led the mature Kant (a) to abandon what he kept on believing his entire life was the sole true solution to the problem of war (on this point Byrd and Hruschka are right) and (b) to endorse the limited diplomatic contribution to peace offered by the “permanent congress of states.” To save the plurality of states Kant came to believe that one needs to be satisfied with the difference (small or big, this is debatable) made by the diplomatic channels offered by the loose league, while hoping that the league’s contribution be magnified and rendered sufficient by the combined effect of the other two peace-promoting factors: republicanism inside states and cosmopolitan right.

In a sense, Thomas Pogge is right that the EU is a good example of the Kantian federation. It is the best example, however, because its attempt to further the process of integration and to extend it to the creation of a truly common defense while remaining a confederation of sovereign states mirrors the paradox with which Kant struggled in all of his political writings. The fact that he did not solve it completely and that his praise for the “rational” solution of the world state ultimately coexists with his preference for the federation is not a sign of intellectual weakness, even less of political cowardice. It reflects something that one could describe as a sort of antinomy of practical-political reason.

9 Concluding Remarks

Kant’s model of perpetual peace is still one of the most influential paradigms at our disposal in terms of thinking of world affairs and imagining political solutions to the most pressing issues of our time. The model changes in the hands of Kant over time, and behind its apparent simplicity lurk tensions and dilemmas, some of which I tried to highlight and, to the extent possible, solve. Among those that are particularly fascinating is Kant’s changing attitude toward the nature, shape, and powers (or lack thereof) of the supranational institution that plays the role of second pillar of the entire model. I suggested that Kant saw with absolute clarity a dilemma that was growing within his system of right between the never-abandoned belief that the sole solution of war is the overcoming of states and the necessity of respecting the “posited” moral personality of states, established at the moment in which a scattered group of individuals obeyed practical reason’s command to move from a lawless condition to one in which the inborn right(s) of human beings would be secured. I suggested that this dilemma could only be overcome if citizens of the existing nations unanimously decided to dissolve their states to merge into a global compound (it does not matter whether in the form of a single universal republic or in that of an only apparently less stringent federal state of states). In Kantian terms this would mean that the general will of each domestic republic would merge into a sort of global general will that decides to win peace at the cost of local political autonomy. Despite the principled preference for the federation, whose profound ground I highlighted above, nothing in Kant’s system prevents this evolution.

The only problem with this – it seems to me – is the rather demanding condition on which the passage would remain respectful of the rights of citizens of each existing state. I suggested that this condition is their principled unanimous consensus: All citizens should be able to agree, not just a majority. One could be puzzled by this suggestion, which of course merely tries to guess what Kant would or should say, given his silence on the point: the recent referendum held in Scotland to leave the UK proves that even decisions about the survival of nations and the birth of new entities are taken by simple majorities. Similarly, very few would object if a new referendum held in Kosovo and Albania established, against the will of some Kosovars, that the two states merge.

While it is true that Kant allows for decisions made by political actors without the explicit support of the people – think of the single-handed decision of an enlightened ruling class to transform a despotic state into a republic or to change the form of sovereignty to bring government more ino line with the spirit of the original contract (MS 6:340) – by his standards the decision to merge the state into a higher body would seem to be an altogether different decision. It would call into question the pact that individuals subscribed to when they followed the first-order command of practical reason and overcame their domestic state of nature. When they did that, they generated a rightful condition that is hard to conceptualize as one that could be overcome by coercing some unwilling individuals. They may oppose it by appealing to the terms of the original pact to which they subscribed. To be sure, practical reason now commands that states be dissolved because this is the unique manner to guarantee peace. It could also be said that the principled unanimous consensus I so much insist on is less demanding than it sounds. All citizens should agree to it, were their moral insight sharp enough. Or they would agree to the merging, were they set in an ideal condition of choice. Possibly so, and yet some doubts remain.