1. The Obligation to Obey the State

Kant is unequivocal that persons have a duty to obey the state they are in. The obligation is categorical: persons cannot evaluate laws according to their own moral sense and choose which ones to obey, nor can they judge rulers and decide whether they deserve allegiance. Public law determines what is permitted or prohibited, and these laws impose strict duties on citizens. Strict duties require specific performance, so that a person fulfils the duty only by acting in a particular way. To the legal duties are affixed penalties and punishments, and a transgressor will suffer consequences according to the severity of the crime. The most serious transgression is to attack the person of the monarch or to seek to kill him – this is high treason and punishable by death (MS 8: 320).

Kant is also unequivocal that the obligation to obey the law applies even if the government is unjust. A government is unjust when it fails to govern according to the rule of law and instead arbitrarily imposes its will on the population. This typically takes place when the division between legislative, executive and judiciary is not respected, or when the government seeks to interfere unduly in the private sphere by imposing on citizens a particular view of the good life or a religion. Kant typically characterizes that as despotism. Nonetheless, even when the government behaves unjustly, and it therefore would not in principle have been wrong to stop government from so behaving, private persons cannot refuse to obey the law. Persons who are wronged must instead seek to contest public authority through existing institutional means, in particular by registering protest in the public sphere.

The justification for the categorical duty to obey the law is that it is the only way equal freedom can be secured. Freedom in social interaction, which Kant sometimes refers to as external freedom in contrast with internal transcendental freedom, is to have a range of choices protected and to be independent from the arbitrary choices of others. This freedom must be equal since every human being is endowed with the same innate right to freedom (TP 8: 289; MS 8: 238, 256, 306; SF 7: 85). Kant defines justice, or ‘right’, in the universal principle of right (UPR), as the formal compatibility of choice among every member of a state according to a universal law:

Maxims of acts that asymmetrically limit the freedom of some more than others (Kant's favourite example is hereditary nobility) could not possibly be established in the form of universal law and so are incompatible with a juridical condition. The test of whether a maxim is apt to become a law in the doctrine of right is therefore, as in Kant's moral writings, the possibility of a maxim being made universal without contradiction.

I will refer to Kant's view of justice as a public recognition view because it relies on positive law as a necessary condition for validity. Kant himself rarely uses the term justice (Gerechtigkeit) and speaks usually of something being right (Recht) and of a condition being right (Rechtzustand). This terminology signals a close connection between justice and positive law. Any claim to have a right (ein Recht) must be capable of being established in a body of law (das Recht), and only becomes conclusively valid once this is the case. Natural laws (natürliche Gesetze), which are those laws that can be recognized as obligatory a priori by reason without positive law, are indeterminate and can only become a justification for action in the state once they have been specified and recognized in the form of positive law.

The reason rights only become valid through law is that entitlements that can be enforced against others must be secured by an ‘omnilateral’ will, that is, a will in which everyone could see themselves as included. The opposite is a merely unilateral will, expressing the private views of one person. If somebody unilaterally claimed to have a right, but refused to abide by shared institutions to define and enforce it, it would be the same as refusing to be in a civil condition, where conflicts among rights-claimants are solved by public determinations rather than private violence. Individual parties to a conflict can only impose a private will on others, and as such act despotically. The state is the only organization that can function impartially and provide an omnilateral will, because it is sovereign, purportedly neutral, and provides a system of laws and procedures for establishing and arbitrating claims to right. Therefore, its verdict is of a different order from that of parties to a conflict: it has authority, the right to bind others to its decisions by its mere choice (MS 8: 224).

Because justice requires positive law, it is a duty to be in a juridical condition (a state), even an imperfect one. Kant spends considerable efforts showing that it follows from this view of justice that resistance and revolution cannot be just. These arguments revolve around the paradox that any disobedient person finds him or herself in, that is at one and the same time wanting public legal authority but reserving the right to be the final judge of right and wrong. Whether an individual or a people contest the ruler, there will never be a neutral authority to decide in the dispute between the two, and so an individual or a people will in effect be judge in their own case.

This was not a novel argument; Locke and Rousseau had already explored this paradox and used it to rule out there being a contractual relation between people and sovereign because there would be no neutral authority to adjudicate the contract. They nonetheless defend a people's right to revolution because in their view a people can spontaneously act in a unified manner and pursue justice. This option is not open to Kant, however, because of his public recognition view of justice. An act is not justified merely by the outcomes it has (such as the deposition of a despot and the creation of a new regime); its justice depends on whether its maxim can be made universal without contradiction. But the maxim to rebel (even under despotic government) cannot be made universal because it would mean that anyone unilaterally could judge the ruler to be despotic and set themselves over the law. This would render the law only conditionally binding and, in Kant's view, that is incompatible with a juridical condition. Since a juridical condition is necessary for the sake of equal liberty, resistance and revolution are excluded, ultimately, because they are incompatible with freedom.

2. Defining the State

Kant's argument has led to a good deal of dissatisfaction because, paradoxically, it seems to require persons to obey despotic governments for the sake of their own freedom. But recently commentators have started to explore the possibility that Kant may have rejected a right to resist only to constitutional states, and not to ‘any thug who happens to wear a crown’ (Waldron Reference Waldron2006). Interpretations wary widely in terms of what that might mean. Byrd and Hruschka hold that any polity that is not a republic (government that protects equal liberty and the rule of law and institutionally separates the legislature from the executive) fails to deserve obligation. This is a difficult view to maintain, for two reasons. First, because it conflicts with Kant's many statements that one must obey even very imperfect rulers (TP 8: 290, 299, MS 8: 320), and second, because it saddles Kant with an implausible view. If nothing short of a republic must be obeyed it would be difficult for a polity ever to develop towards that stage. Ripstein, more cautiously, sets a lower limit and argues that regimes that enslave or kill parts of the population cannot be counted as true states (his chief example is Nazi Germany) (2009: 338). Because such regimes deny persons the right to life and liberty they must in Kant's terminology be classified as ‘barbaric’ and therefore not true states at all.

Ripstein's view is less controversial, but it remains unclear exactly how persons should judge the matter and why it does not contradict Kant's theory of authority. Byrd and Hruschka describe the dissolution of the state as a right of revolution (Reference Byrd and Hruschka2010: 91), but that seems to conflict with Kant's many statements that revolution cannot be right. Moreover, it makes no sense to speak of revolution when in fact the state has dissolved, since a revolution is when force is used to overturn the existing constitution in a state. To understand this more precisely, the first step is to elucidate the Kantian notion of the state.

There are two necessary and sufficient elements of the state: power and the rule of law. Kant refers to the ideal of the state as the Platonic ideal of the respublica noumenon: ‘the idea of a constitution in harmony with the natural right of human beings’ (SF 7: 90). The first aspect of the respublica noumenon is that it has superior coercive power: ‘he who does not have enough power to protect each one among the people against the others does not have the right to command the people either’ (ZEF 8: 383). A condition of anarchy cannot be a state; there must be a univocal determination of right and wrong (Waldron Reference Waldron1996). The sovereign institution in a state is the legislature. Its task is to create laws and ensure that sanctions are attached so that transgressors are apprehended and punished. To this end, the legislature has the ultimate control, through the executive branch of government, over all public institutions of order, including the police and the military.

The second aspect of the ideal state is the rule of law. On the public recognition view, justice must be defined through legislation. This is why Kant calls a state a ‘juridical condition’ (Rechtszustand) and defines it as ‘a union of a multitude of human beings under laws of right’ (Rechtsgesetzen) (MS 8: 313). The basis for legally established justice is the procedure of the UPR, which requires laws to be such that they secure equal liberty. The state must govern through established law (that which is rechtens), in support of that which is rightful (rechtlich). To ensure this, there are three powers in the juridical condition: the legislature, executive and judiciary – and they are separated. The legislature is the sovereign, creating positive laws that are impartial and held to be the general will of the people. Positive laws are backed by force, providing uniform rules of right and wrong. These laws are employed in specific cases by the judiciary, and enforced by the executive. Not all positive law is according to justice. Like modern legal positivists, Kant distinguishes between law and its merits, between positive law, ‘what is laid down as right’ or rechtens, and laws that are right (iustum) according to ‘a moral concept’ or Recht (MS 8: 229, 267, 297, 306). If the state completely lacks one of its two defining aspects it would not be a state. In reality every state is more or less deficient in one of the two dimensions. Deficiencies in power and justice can intersect in four possible ways, creating four different regimes: just and powerful, just and weak, unjust and weak, and unjust and powerful. The particular difficulty is not caused by very just and efficient regimes or their opposites, but those regimes that are either fairly unjust or inefficient. For a deficiency to cast doubt on whether the entity is a true state, the transgression must be systematic and severe. Discrete violations of justice (e.g. unjust sentencing in a criminal trial) occur in every legal system and the normal remedy is to appeal through the existing procedures. If deficiencies are minor, the state will in general be capable of supporting equal freedom and as such qualify as a state.

In Kant's view, states often start as violent power grabs and proceed to work themselves clean through history. State sovereignty is therefore a precondition for reform, hence power must exist before a civil condition can be established. To explain why unrightful states are provisionally justified he devised the category of ‘permissive laws’ (leges latae, lex permissiva), which mean that injustices (wrongdoings) are temporarily tolerated if that is necessary for the transition to a just regime.Footnote ³

Ripstein as well as Byrd and Hruschka argue that a system that treats its population or parts of it as serfs or slaves cannot be said to constitute a juridical condition. As such they use a similar approach to that of Julius Ebbinghaus who distinguishes between injustices and acts of inhumanity (Reference Ebbinghaus1953: 21). Injustices are arbitrary decisions by the government to restrict liberty more than is required for the sake of justice, or unfair implementation of the law. Acts of inhumanity, however, reject the right of individuals to have legally protected rights – to property, freedom of speech and even life. Excluding persons from the community of rights is the most severe form of transgression and governments that do so go beyond their competence to such an extent that they cannot be considered governments, but rather private individuals with no authority.

But while it may be comparatively easy to conclude that regimes that completely refuse to secure the rights of its population or parts of it do not satisfy the conditions for a juridical condition, one may wonder about harder cases where a regime to some extent provides rights but withholds some rights from everyone and distributes other rights unequally. One may also wonder about the case of a regime that does not satisfy perfectly the conditions of sovereignty because it lacks superior power. Where exactly should the line be drawn between a highly imperfect regime that is still entitled to obedience, and a regime that has crossed the line and is no longer to count as a juridical condition?

This is not a question of merely academic interest, since persons are sometimes faced with situations where multiple parties claim the right to rule. If the situation slides into one of civil war, there is no juridical condition, but before it comes to that there will normally be one party that claims and perhaps deserves the right to rule. In those extreme circumstances persons will need principles to decide by and the question is whether such principles can be elicited from Kant's principle of the juridical condition. The task of finding this point can be compared to the Lockean question of judging exactly when a ruler forfeits his authority to rule. Locke is quite clear that one or a few transgressions by the ruler are insufficient; there must be a ‘train of abuses’ (1988: §§210, 225). But he makes no attempt at explaining how long that train must be, and we may wonder whether it can be determined at all.

In the case of Kant, we must expect that the distinction will revolve around the question of whether reform of the present regime is possible. After all, he is prepared to admit that an unjust regime can be provisionally justified in so far as the rulers take seriously their obligation to reform. We must assume that he meant not just regimes that unduly restrict freedom, but also regimes that to some extent fail to provide individuals with rights, as long as they are intent on reform (ZEF 8: 373–5). But this condition may be vexing for two reasons. First, it is difficult to know whether rulers are serious when promising reform, second, by waiting for reform to happen persons may put themselves in a weak position where they can easily be targeted by the present regime.

To distinguish between imperfect regimes – between those credibly developing towards justice and those that are not going to reform – is a matter of identifying the intention of the ruler. Since one cannot literally get inside the mind of another person, intentions must be identified in two ways. First, statements can be analysed in order to establish whether the ruler is at all committed to the principles of equal liberty. If the ruler's expressed will shows a maxim that is incompatible with a civil condition there is little reason to expect reform. Second, if the ruler claims to be committed to the principles of equal liberty, behaviour must be observed to reveal whether there is reason to trust those intentions. Violation of contracts is one indication; another is if the ruler refuses to permit freedom of expression, since this is how the people can communicate their wish for reform (as emphasized by Ripstein Reference Ripstein2009: 342). The rule-breaking behaviour must be persistent if it is an indication of lacking an intention to reform.

Establishing whether there is reason to expect reform from someone holding power will be a matter of judging a particular case. This cannot be an easy judgement. Consider Kant's own Prussia, which in the Allgemeines Landrecht of 1794 established a confused mix of proto-liberal principles and late feudal privileges. While it condoned serfdom, and as such violated the humanity of many of its subjects, the legislation went through a large but slow overhaul and it eventually abolished serfdom in 1807. But this may never have happened, or at least not have happened so soon, had it not been for Napoleon's victory over Prussia in 1806. This development would have been hard to predict for someone living in Prussia in the 1790s. It was nonetheless clear by then that the regime was intent on reform and that it was willing to listen to the public, since it favoured a fairly large sphere of free expression. We know that Kant himself judged his state to be on a path towards republicanism. Other regimes may be less obviously committed to reform. Among the autocratic regimes in the Arab world prior to 2011 some, like Egypt, seemed set on a path of reform, while the great bulk were not committed to principles of equal liberty nor were they opening a public sphere where the country's future could be discussed. As such, they may, in Kant's terms, have been doing wrong ‘in the highest degree’.

3. The Role of Judgement

What does it mean to make oneself the judge about the state, and how could anyone be entitled to do so? For Byrd, Hruschka, and Ripstein this question does not surface perhaps because they assume that no judgement needs to be made about whether a state exists. In cases of despotism and barbarism the state has simply disappeared and individuals are free to act. But that begs the question because in order to know that the present regime is despotic or barbaric a judgement has to be made, and this is a complex moral judgement, not merely a matter of surveying who has the greatest coercive force. How can it be made and who is entitled to make it?

Judging consists of two parts: forming a verdict, and acting on it. To simply form a verdict without acting on it does not interfere with the freedom of anyone else and could therefore not be prohibited according to the UPR, which limits justice to the compatibility of the freedom of choice of each according to a universal law (MS 8: 230). Persons are entitled to freedom of conscience and speech. But matters are different when it comes to acting on the judgement. As we have seen, Kant argues strenuously that a right to act on any private judgement cannot be supported by positive law, because it would lead to anarchy. Who, then, could be entitled to act on a judgement that the state has failed?

First it is necessary to establish exactly how it can be possible to act on a judgement about the state given that Kant's argument against revolution rejects the view that anyone may make themselves judge over the sovereign. After all, the state is instituted in order to end the state of nature, where everyone acts on private judgements of right and wrong. In the state, judging about law is solely the office of the sovereign, and Kant approvingly quotes Frederick II's motto: ‘Argue as much as you will and about whatever you will, but obey!’ (WA 8: 37). Acting according to existing right does not require judgement, because it is simply to do what positive law requires, prohibits or permits. These are strict duties, because they do not allow for latitude in how to implement them, in contrast with ethical duties, which, as wide duties, allow individuals leeway in terms of how ends are to be realized (MS 6: 411 and 390). As we have seen, the reason persons have no right to act on their personal judgements about positive law when they are in the state is that this is incompatible with public authority. But if persons may not judge about the law, how could they have a right to judge about the existence of the state?

To judge about the state is actually different from judging about laws. A judgement about the state is a verdict about whether the social configuration with its laws and public institutions counts as a civil condition, based on the application of the concept of the respublica noumenon in a given context. This might seem similar to judging about laws, since judging about the state too will involve a verdict on laws (it will simply be a verdict on all the laws and public institutions in a society). But judging about the state is different because in this case there is no authority that could decide, since, as Ripstein rightly emphasizes (2009: 342), the verdict of the professed sovereign only has authority in case he or she in fact is the sovereign (and at stake is that very question).

A judgement nonetheless has to be made, because given Kant's view of the state it is not obvious when it ceases to be one and the state of nature sets in. Who, then, has the responsibility to judge? One might have thought the judgement should be left to the people as a whole – as Locke famously expressed it ‘the people shall be judge’ (1988: §240). To Locke (as well as to Rousseau) collective judgement by the people acting as one is possible because the people is united through a referendum prior to entrusting a government.Footnote ⁴ Kant too was committed to the idea of popular sovereignty, but collective decisions by the people are possible only through procedures binding on everyone, and that presupposes positive law and sovereignty. In the absence of law the people is merely an uncoordinated multitude (MS 8: 318, TP 8: 302). Hence, if the people were in a position to judge, the question would already have been decided. A referendum on revolution could not be binding because there can be no procedurally correct way to hold it. In order to set up a referendum about whether the polity is a state one needs to settle on certain rules (about who should be permitted to vote, how the question should be put, what should count as sufficient majority and so on), but that decision presupposes an entity with the authority to define the procedures. The only legitimate entity to do so would be a sovereign entity employing law and superior coercion, and (in the absence of international or cosmopolitan organizations) that can only be the state. Again, the method for settling the question presupposes what needs to be decided. There can exist no established domestic procedures for binding judgements of whether a state is present because at question is exactly whether such procedures exist (Flikschuh Reference Flikschuh2008: 380ff.).

Nor could that judgement be left to intermediary associations like the military, churches, or political parties engaging in elite negotiation. These associations can exist in the state of nature, but they can have no authority over individuals, since authority must come from the state. In principle it is possible to set up an international or cosmopolitan certifying board to judge on the recognition of entities as states, perhaps similar to the existing system of recognition in the United Nations. Contrary to the UN system it would apply comprehensive moral standards to the judgement and not be satisfied with a judgement of effective power. If a federation had this authority, states would in principle no longer be sovereign because they would recognize a superior authority. For Kant this is not an alternative because he, for a variety of reasons, holds state sovereignty to be of supreme importance (ZEF 8: 355–7). Moreover, an international certifying board addresses the problem only on one level and the difficulty reappears on a higher level because one would need a certifying board to certify the certifying board, which leads to an infinite regress.

In the absence of established authorities, individuals are left with the responsibility to judge. This follows from the postulate of public right, which requires that persons exit the state of nature and join a civil condition.

The duty to enter a state would make no sense unless persons first judge whether they may already be in a civil condition. The first (and necessary) step for someone who wants to be in a state is necessarily to judge whether he or she might already be in one. Moreover, persons who are in the state of nature are entitled to enforce the judgement. To enter a rightful condition is an external duty, one that requires specific performance (not that one adopts a particular end), which means that it can be enforced from the outside through external sanctions. The corollary of the duty to join the state is a right to force others to join if they refuse to do so voluntarily:

Kant describes this entitlement to use force variously as a permission (Erlaubnis), a provisional right (ein Recht), and an authorization (Befugnis). He writes that one may (dürfen) force others and that one is not bound (verbunden) to respect their desires to remain in a state of nature (MS 6: 256–7, 307–8, 312–13). This is not a right privately to secure one's possessions in the state of nature, or a right to punish transgressors (as it is for Locke).Footnote ⁵ In Kant's state of nature rights are provisional because of the lack of public law that would guarantee them, and there can be no entitlement unilaterally to enforce them because respect for others presupposes enforcement through public legal authority. The condition of rightful reciprocal obligations is public law, and law only comes into being with the state. The authorization persons have is therefore not to secure their property but to force others to join a rightful condition where a sovereign can determine through law, and with the help of courts, what justice requires.Footnote ⁶

That forcing recalcitrant persons to join the state is consistent with justice is evident from Kant's justification of coercion:

Refusing to enter the state is wrong (‘in the highest degree’) because it is incompatible with equal liberty and therefore in conflict with the innate right to freedom. For that reason it must be right to coercively hinder people from remaining in the state of nature. But to speak of having a right in the state of nature is odd from a Kantian point of view, since positive law is necessary for establishing rights. This is what was earlier referred to as the public recognition view of justice. Indeed, in the quote above Kant described duties of right as those ‘for which external lawgiving is possible’. But in the state of nature there exists neither public law nor coercive institutions, so it is impossible to have external lawgiving for a right to force others to join in creating a state.

The authorization to use force must therefore be thought of as a provisional right, justifiable with reference to what it seeks to achieve. This provisional right is not retained within the state and can only justify the use of force before it comes into being or when it has been dissolved. The basis for it cannot be a constitutional procedure but must be the requirement of practical reason to enter a lawful condition. The act to force others into the state is omnilateral only in the sense that it is performed for the sake of creating public institutions; it will not be omnilateral in the sense of being sanctioned by institutions that represent a universal will. Often those who claim sovereignty by force are mere strongmen who are not motivated by the duty of right to enter the state, and who do not aim to establish constitutional government. But it is at least possible to imagine persons using force to establish the state in pursuit of a duty of right, and therefore it can be a duty to create a state (since ought implies can) and to do so by force if necessary.

One may wonder whether the authorization to use force to create the state after all is identical to Locke's concept of natural law, which entitles persons to use force against tyrannical rulers. In a basic sense the concepts of justice of the two thinkers are clearly different: Locke's notion of right is orientated towards securing an end (such as the individual or common good), whereas Kant's notion of right is formal and results from the procedure of universalization without contradiction inherent in the UPR. But Locke and Kant do share the idea that reason itself, in restricted circumstances, can justify private individuals in using violence. For Locke this takes place within the state, specifically against a government whose authority has been dissolved. For Kant, however, natural right can never justify private action against the government for the various reasons we have seen that he gives against a right of resistance and revolution. The private use of violence can only be justified once the state itself (not just the authority of the ruler) has been dissolved. In that case, reason does indeed authorize the creation of a state through private use of violence.

4. Contradicting Authority?

The argument so far has been that persons always have a responsibility to judge whether the polity truly is a state. But this might seem like just a different way of justifying revolution, and therefore contrary to Kant's basic principles. After all, any revolutionary might claim the state no longer exists and proceed unilaterally to seek to overthrow the rule of law. In his anti-revolutionary argument, Kant made it clear that the state's authority cannot depend on the more or less well informed opinions of one or several persons but must be above the shifting factions in society. Its authority comes only from governing according to principles of right. But if one faction in society can be justified in claiming that the state has vanished it would seem that the state's authority nonetheless depends on the consent of one or several groups.

Yet, this is not a consequence of the theory. First it is important to recall that judging consists of two elements: forming a verdict and acting on it. Persons are always entitled to form a verdict, but they are not always permitted to act on it. No one can legitimately be prevented from thinking what they like about the present government, and they should be free to voice that opinion as long as they do not conspire to overthrow the state. But they are free to act on the judgement only if in fact they are in the state of nature. If they are in the state of nature there is of course no state to overthrow so the claim is not that the state's authority is limited by or dependent on private judgements.

This might sound unsatisfactory because the consequence may often be similar to a revolution. After all, it is impossible to establish authoritatively whether the condition in fact is the state of nature or a highly imperfect state. There is no neutral party to decide who is right between the person claiming to rule a state and the person denying the claim. Hence, it would be easy for a would-be revolutionary simply to claim that the state no longer exists and the result might be a constant stream of challenges to the state. Yet, while this may be worrisome, it is a contingent empirical consequence, whereas Kant's argument is moral and juridical. Kant is concerned to deny that anyone can claim moral legitimacy against law, and that a constitution coherently can allow persons to break the law when they judge it to be necessary. But the claim that persons have a responsibility to judge is not that there should be a positive law permitting persons to make the judgement and to act on it (it would be a law with limited use, since it would only be legitimately appealed to when the state has ceased to exist). Nor is the claim that morality can overrule legality, because it would only be legitimate to act on the judgement when the state has ceased to exist. Thus, Kant's arguments against a moral or legal right to revolution do not exclude that persons have a responsibility to judge the state.

Persons who inappropriately deny that a juridical condition exists challenge the stability of the system, not its authority. Authority derives from providing the rule of law and having the power to secure it. If persons dishonestly challenge this system such authority does not disappear. The appropriate public answer is to secure the stability of the legal system by striking down the threat and treating the disobedient subjects as rebels. Such rebels are no different from anyone else who breaks the law. Their motives may be noble, but noble motives cannot justify anyone in unilaterally resisting the rule of law.

Because the idea of judging the state does not justify action against actual states, it is not a justification for revolution. A revolution is when force is used to overturn the existing constitution in a state, and since there is no constitution in the state of nature there can also be no revolution. For this reason, other arguments Kant provided against a right of revolution do not exclude a responsibility to judge whether the state exists. He ruled out revolution because it returns society to the state of nature, but in this case the state of nature already obtains. Furthermore, he argued that the idea of a revolution is incoherent because a people cannot act collectively and with authority in the absence of state institutions and public procedures of coordination (MS 8: 318, 340). It might be thought that acting on a judgement about the state likewise presupposes an untenable view of non-institutional collective action. But, as was shown earlier, it presupposes no collective action because the agents are individual persons, and if they band together they do not act in the capacity of a people but as moral agents. If someone attempted to coordinate their actions, that person would not act with authority and could not require obedience.