Federalism and Secession

by Jorge Cagiao y Conde (Volume editor) Alain-G. Gagnon (Volume editor)
Edited Collection 198 Pages
Series: Diversitas

Table Of Content

←12 | 13→


How can we explain our decision to offer readers a book on federalism that deals mainly with secession? In fact, there is no shortage of reasons. For example, many of the contributors to this book make the observation that the question of secession has been largely ignored in the specialized literature on federalism. This silence alone is enough to attract interest from researchers, who would be just as surprised to see a legal expert study the complex topic of marriage, or a lawmaker legislate on the same matter, only to drop everything when faced with the question of divorce. Could we imagine, today, a study of marriage or a civil code – in a liberal democracy – failing to cover the subject of divorce? This, however, is the situation in the field of federalism (our equivalent of the “political union”) and secession (the “political divorce”). All of which leads to a simple question: “Why?”

Federalism has often been presented as the best mechanism for accommodating diversity in unity in a complex society. It appears to be a system that, by opting for a vertical (territorial) separation of power, is able to keep territories and cultures endowed with enough autonomy (self-rule)1 to make their own political choices united under a joint government (shared rule). However, federalism is by no means immune from the problems it is expected to solve. This is because, first, the theoretical virtues of the federal idea, an ideal balance between a centre and a periphery, are sometimes difficult to put into practice, for example when the values of federalism have to give way to a centralizing vision. Carl Schmitt referred to this centripetal dynamic within a federation in a particularly apt way: “a federal state without a federal foundation”.2 Secondly, federalism is a little like the worm in the apple: it carries the germ (the freedom without which peoples would not even ←13 | 14→consider uniting) of disaggregation.3 From this point of view we can only agree with the position of Will Kymlicka, that the greater the territorial autonomy and ethnocultural or national diversity present in a federative system, the more the system will tend to be challenged by its constituent units. European federalists will probably agree with this as they face the fallout from Brexit. When the centrifugal effect of the federative relationship dominates, the values of federalism and the reasons that encourage states or peoples to unite will fade, and unity will be, and will remain, precarious.

The authors whose work is presented here recognize the potential of federalism as a way to organize relations between several different states, peoples, nations or territories under the same government, in a relatively balanced and harmonious way.4 However, they are not naïve or idealist about the ability of the federal idea to succeed in the complex situations in which it is applied. In some cases success seems assured (the United States, Switzerland, Germany, etc.), and the merits of federalism can be showcased. But there are also failures (the former Yugoslavia, or more recently Brexit) and semi-failures (sometimes only perceived as such) that have generated turbulence in recent years in devolutive systems (Scotland in the United Kingdom, Catalonia in Spain) or federative systems (Québec in Canada).

The question here is not to decide whether federalism can be an efficacious remedy against secessionism. This has already been subjected to extensive analysis, the results of which mainly indicate a positive ←14 | 15→answer, with some qualifications. These qualifications are necessary because there are, in fact, two different cases: one in which federalism is applied to a people or nation that sees itself as the only people or nation in the future federation; and the case in which federalism is applied to several different peoples or nations that see themselves as distinct, and generally wish to remain so. The problem of secession concerns mainly the second case.

Territorial federalism and multinational federalism

Secession poses an existential problem, as Carl Schmitt has pointed out.5 For this reason, federation-nations (territorial federalism) and multinational federations (pluralist federalism) are not affected in the same way. In a federation in which the constituent entities see themselves as forming part of a single nation (like the United States or Germany), the risk of secession is low.6 The same cannot be said of federations based to a greater or lesser degree on national or ethnocultural pluralism, like Belgium, Canada, the United Kingdom or Spain (two decentralized states), or the European Union.

In these contexts, national or ethnocultural groups in a specific territory that is part of the federation or state are more likely to raise the question of secession. In Canada, Québec, rather than the other provinces or territories (majoritarily English-speaking), was the province that launched a referendum process in order to secede. In the United Kingdom, it was Scotland (and, in earlier times, Ireland); it is hard to imagine that England would follow this route to separate from Wales, Northern Ireland and Scotland. In Spain, which has faced secessionist challenges from Catalonia and the Basque Country, the same claims would be unlikely to come from other regions such as Andalusia, Galicia ←15 | 16→or Valencia. Likewise in Belgium, where secession appears attractive only to the Flemish nation, rather than the national and linguistic group that was formerly dominant, the French-speaking Walloons.

In short, secession is an issue – and a concern – for multinational federalism, rather than for territorial or national federalism. When the nation-building process has had the expected effect in a federation that has also consolidated its functional democracy, the risk of seeing a strong secessionist movement emerge is so low that it is easy to understand why, in these federations, the hypothesis of secession is greeted with a mixture of distaste and incomprehension.

Multinational federations, which are more directly concerned by the problem of secession, must also deal with the greater legitimacy granted to secessionist demands within their borders. The same logic applies: just as it is difficult to imagine that a French région (except in a colonial context), a German Land or a US State would ever ask its nation-state for permission to separate and form an independent state (by reason of the degree to which the populations concerned see themselves as forming an integral part of the nation and the state), so it is easy to understand and to accept, in the state- and nation-based logic of our modern political world, that a territorialized human group, aware of being a distinct nation or people, should want to control its own state. In other words, a demand from the second group is generally seen as having more legitimacy than a demand from the first group. The question of legitimacy is of capital importance here. If, like the United Kingdom (under article 50 of the EU treaties), Québec, or Scotland, the political entity claiming a right to withdraw is recognized as a people or nation, secession seems to follow an easier path and to find a place in the legal order; on the other hand, if the political entity claiming a right of secession is not perceived by the central state or organization to be a separate people (as in the case of Catalonia, for example), then the legal order will remain inflexible.

Structure of the book

It is difficult to separate legality from legitimacy in such a sensitive and complex area. This is probably one of the key lessons we can take from the court decision that has attracted the most attention: the advisory opinion given in 1998 by the Supreme Court of Canada concerning Québec’s ←16 | 17→unilateral secession.7 This is the leading case8 systematically referred to today in the all discussions of secessions in liberal democracies. The authors presented here are no exception to this rule, and the advisory opinion of 1998 is discussed extensively in this book.

In the first chapter, Christophe Parent reviews the juridical nature of secession. Going beyond the question of secession as a fact (when it succeeds despite violating the legal order to which the secessionist territory formerly belonged), he examines the normative dimension: “in a federal framework, does there exist, or can there exist, a right of secession?” Parent offers a vast and rich array of experiences drawn from doctrine and positive law to help readers understand the contrast between federal theory and practice over the course of history, and the hostility of federal positive law towards secession. Parent presents the thesis, in a sense following the path traced by the Supreme Court of Canada that the juridical value of secession in constitutional law depends on a mechanism for constitutional amendment to accredit the legal avenue for the outcome targeted.

Next, extending his theory of federation, Olivier Beaud defends the idea that secession from a federation should be treated as an autonomous concept. In his Théorie de la Fédération,9 Beaud presents the federation as an intermediate stage between the two dominant federative models, the federal state and the confederation. Since a federation is not a federal state (in the author’s view), and since the relation between federalism and secession has generally been examined in the fields of constitutional and international law in connection with the federal state (because confederations are associations governed by international public law), Beaud appears justified in asking if there is a difference between secession from a unitary state and secession from a federation. He details his approach and outlines the possible consequences.

Chapter 3, by Jorge Cagiao y Conde, tests the dominant thesis in studies of federal systems, which posits that there is a logical ←17 | 18→incompatibility between federalism and secession, to the point that the positive law enacted by federations has no choice but to exclude the right of withdrawal. The small number of cases in which federations have constitutionalized a right of secession should, under this thesis, be classified in the “confederation” category (like the EU, if it can be so defined), or as “non-federal” exceptions to the general rule (because they are contrary to the principle of federalism). Cagiao y Conde discusses the thesis in relation to known federal experiences drawn from both doctrine and practice, on the basis of what he calls the “legal logic” that applies when the question of federalism and secession in raised. He arrives at a conclusion that introduces considerable nuances to the dominant doctrinal thesis.

In Chapter 4, Dave Guénette and Alain-G. Gagnon review Québec’s secessionist experiences in Canada. With a constitution that remains silent on both the unity of the Canadian federation and the organization of a referendum concerning a province’s independence, Canada offers an example that clearly illustrates the ability of a federative legal system – and a federal political culture – to mobilize its constitutional resources in order to channel a secessionist conflict using peaceful legal means. It comes as no surprise that the Canadian experience has become, since the famous reference of 1998, the leading case from which politicians and researchers around the world draw both arguments and inspiration, despite a small number of unanswered questions and uncertainties (a clear majority, the constitutional amendment procedure) that remain in the Canada-Québec debate, as Guénette and Gagnon point out.

The last chapter presents what could be considered the perfect counterexample to the lessons drawn from the Canadian experience. Lucía Payero analyses the conflict in recent years between Spain and the independence movement in Catalonia, supported by a political majority. Unlike Canada, Spain is not a federation, and unlike Canada, it has not chosen to explore some of the federative resources that certain authors have glimpsed in its constitutional order. Payero’s thesis is as follows: given the (historical) hostility in Spain with respect to the federative arrangement demanded by the Catalan nationalists and also by left-leaning federalist parties in Spain, which form a tiny minority, based on a revision of the constitution, and given also the failure of infra-constitutional or informal attempts to reform the system (revision of the Statute of Autonomy of Catalonia in 2006, fiscal pact, etc.), support for ←18 | 19→the Catalan independence movement, which was marginal at the start of the century, has actually increased. In the case of Spain, the demands for Catalan secession can be seen as a direct consequence of Spanish hostility to federal solutions.


1 Elazar, D. J., Exploring Federalism, Tuscaloosa, University of Alabama Press, 1987.

2 Schmitt, C., Théorie de la Constitution, transl. by F. Deroche, Paris, Presses universitaires de France, 1993, p. 537.

3 Kymlicka, W., “Is federalism a viable alternative to secession?”, in Lehning, P. B. (ed.), Theories of Secession, London, Routledge, 1998, p. 111–150.

4 The question of secession does not arise in a system in which everything is running smoothly for both the federated parties and the federation. Secession is always connected to another problem, a symptom for a serious dysfunction within the federation. Rainer Baubock describes the situation as follows: “None of the current Anglo-American theories of secession gives proper consideration to the most common grievance voiced by national minorities in multinational states: that the terms of federation are either unfair or have been violated by the majority. If this charge is indeed a plausible and necessary justification for threatening with secession, then it would also follow that a national minority is morally bound to maintain the unity of the existing sate as long as fair terms of federation are respected.” (“Why Stay Together? A Pluralist Approach to Secession and Federation”, in Kymlicka, W., Norman, W. (eds.), Citizenship in Diverse Societies, Oxford, Oxford University Press, 2000, p. 367).

5 Schmitt, C., op. cit., p. 518 ff.

6 It is worth noting that the Civil War (a war of secession) in the United States occurred at a point in its history when the system was undergoing nationalization and the states (at least those in the South) saw themselves as sovereign and in charge of the constitution (compact federalism). The nationalization of the system (and of the population), successfully concluded and consolidated since the 19th century, makes it highly unlikely that today, in Texas or California, a large part of the population would support secession.

7 Reference re Secession of Québec, [1998] 2 S.C.R. 217.

8 Delledonne, G., Martinico, G. (eds.), The Canadian Contribution to a Comparative Law of Secession: Legacies of the Quebec Secession Reference, London, Palgrave Macmillan, 2019.

9 Beaud, O., Théorie de la Fédération, Paris, Presses universitaires de France, 2007.

←20 | 21→


Federalism(s) and secession: from
constitutional theory to practice

Christophe Parent


The word “secession” (from the Latin secedere, “to withdraw”) dates back hundreds, and even thousands, of years. The first traces can be found in the masterful history of Rome written by Titus Livius, known as Livy, in the 1st century BC.1 He describes an episode familiar to all historians of Antiquity, referred to by Livy as the “Secession of the Plebs” (per secessionem plebis), which led to the creation of the well-known Tribune of the Plebs in the 5th century BC. From the same revolt we gained the expression “to withdraw to the Aventine”, an allusion to the fact that the Plebs deserted Rome in their conflict with the aristocratic patrician class and withdrew to the Sacred (Aventine) Mountain until they obtained political equality.2 The original scope of the term “secession” therefore extended well beyond federalism. As used by Livy, a close associate of the Emperor Augustus, it referred to the separation of a social class rather than a separation between two political entities both aspiring to sovereignty. At the outset, then, secession was connected to a historical and semantic reality, and cannot be reduced simply to the ←21 | 22→separatist dynamic within a federal system. In fact, it is interesting to note that the notion of secession is practically absent from the constitutional semantics of the main federal states. None of the constitutions of the United States, Germany, Switzerland, Canada, Australia, Austria and Brazil contains the word “secession”, either to authorize it or prohibit it. This gives rise to a paradox: secession is a notion that is largely ignored in the constitutional law of federations, even though it initially emerged in domestic law.3 We will come back later to the theoretical reasons for this situation.

However, a more in-depth historical examination reveals that the possibility of secession within a [con]federal structure was known to Greek historians and legislators as early as the 5th century BC, thanks to the (numerous) defections from the Delian League. Readers can immerse themselves in the story in the History of the Peloponnesian War, in which Thucydides uses the idea of defection several times to describe cities that wanted to break their ties with Athens.4 Since history can be used to illuminate the future, it is interesting to note that already, during this period, democratic Athens attempted to oppose the secession of various cities using military force.

Naturally no lesson can be drawn from this example, however illustrious. The conflict arose more from an imperialist shift in Athenian thinking than from the normal operation of a defensive league of cities that, today, would be considered more as an alliance or confederation.5 However, law (as a discipline) remains uncomfortable with the idea of secession, as though legal science suffered from the paradox of Buridan’s ass: it faces an impossible choice between the right of a ←22 | 23→people to self-determination on the one hand, and respect for a state’s territorial integrity on the other. To understand the reserves felt by international law (except in the specific case of colonization), matched by the almost deafening silence of constitutional law, we need to go back to the fundamentals. An ancient theory that can be traced back to the 19th century and Georg Jellinek, but which is still current in legal circles, underlies the anxiety felt by legal experts. Secession, and more generally the birth of a state, is considered to be a matter of “pure fact”, one that by its very nature lies outside the purview of legal science.6 Kelsen states, for example, that the birth and death of a state are metajuridical facts.7 Lying outside the control of the law, secession depends on its own failure or success. This explains the more arcane aspects of the positions taken in international law, often apparently contradictory, on the question of secession,8 dependent on a power relationship that is itself contingent. The state exists in law only because it exists in fact – this is stated in black and white in international law handbooks.9 A new state will only be recognized if a government has effective control over a given territory. International law, as an obedient pupil of the Heidelberg master, has enacted a principle of effectivity that consists of endorsing the fait accompli,10 meaning that the secessionist combat is legalized a posteriori by its success. It appears that, here, law is written by the victor!

←23 | 24→

Can a jurist really be happy with an approach that reduces secession to a question of Realpolitik? Positivism refutes the idea that a normative statement can be inferred from a fact,11 and the role of a judge is not to take the side of the stronger party which is able to impose its point of view. The role of the law is to state a priori, rather than simply a posteriori – as influenced by Realpolitik – which party is in the right. This question is one of the key issues dealt with in this paper: to go beyond the false pretences raised by pure and practical concepts in order to rehabilitate an objective approach to secession. In the Kelsenian sense, this refers to the urgent need to develop a legal argument without subjecting secession to a value judgement. Legal science cannot become an advocate or defender of a cause,12 whether to defend the right to secession or the right of an existing state to territorial integrity. This confrontation based on values would, on the contrary, lead to a “war of the Gods”, to borrow an expression from Weber, escaping the control of scientific rationality and feeding an endless debate.13 The question must therefore be examined with this requirement in mind, avoiding any syncretism between law, morals and justice, between sein (the “is”) and sollen (the “ought”). Secession, for a jurist, can be neither good nor bad, fair nor unfair. It must be free from all ethical dimensions: it is either legal or not legal, from the standpoint of a higher norm.

Now that these guidelines have been laid down, we can return to the question of federalism, since secession – in our contemporary world – is indissociably linked in the public discourse to the federal model. Naturally, the US Civil War is a key focus, but in fact unitary states must face the question of secession too, as reflected in the exponential growth in the number of UN member nations, which has been multiplied by four over a period of seventy years. From fifty members in 1945, the UN grew to almost 150 member states in 1984, a trend that can be explained by decolonization. But since 1990 the UN has expanded to include another thirty-eight states to reach a total of 193 member states, mainly ←24 | 25→as the result of secessions, and not only from federal states. Nevertheless, few constitutional texts include specific provisions on secession. Only a few states mention a clear right to separate. To explain the inaction of the federal drafters, who remain practically mute on this essential subject, it is important to note that federalism is, in its essence, a model that respects and promotes diversity. A naïve – or optimistic – stance is to suppose that the federated entities will have no reason to leave a federal union and its countless economic, political and military advantages,14 but it is clear that federal unions are not immune from separatist temptation.15 On this topic, Will Kymlicka has expressed reservations about the ability of federalism to avoid secession, and others have even suggested that federalism may even accentuate the secessionist tendencies of ethnic groups.16 One immediate example is the dissolution of the federal states of Czechoslovakia, Yugoslavia and the Soviet Union. Both the Yugoslav and Soviet constitutions – in countries where separation sometimes led to violence – expressly recognized the right to secession. Today, liberal thinkers do not necessarily consider secession, within a “perfectly just state” that upholds the principles of justice, to be justified or even desirable.17 Here, constitutional democracy plays the role of a template ←25 | 26→to limit the moral legitimacy of this possibility. It is true that a right to secession could lead to strategic, and even egotistical, behaviour on the part of political sub-units18 such as rich regions that could attempt, by threatening to withdraw, to avoid funding a social system based on their ability to pay, preventing the federal state from achieving its mission of fair redistribution. However, we can only reiterate that the legality of the right to secession is not the same thing as its legitimacy, whether moral or political. This once again emphasizes the relevance of our question, which in fine lies at the normative level: in a federal framework, does there exist, or can there exist, a right of secession? Federalism can take one of three distinct institutional shapes, all of which revolve directly or indirectly around the state, whether in the form of a union of states (a confederation) or of a single state (a federal state) or, on the contrary, in a form defined in opposition to the state and its constituting principles. We are referring here to the federation, a model that remains purely theoretical, but which probably offers the purest form of federalism. Theoretically, and this word is important, the institutional federal model chosen will have a considerable influence on whether or not a right to secede exists.

1. Confederation: a “free union” that hides its true nature?

Praesumptio sumitur de eo quod plerumque fit (“A presumption arises from that which usually occurs”). This old legal maxim, although based on common sense, is no help when looking at secession as part of a theoretical approach to federalism. The history of confederations reveals only total indifference with respect to the legal principles set out in textbooks concerning the right to withdraw. Concepts, positive law and empiricism largely contradict each other, calling into question the categories and ←26 | 27→standpoints set out in the academic literature. If, like Elizabeth Zoller, we consider that “a theory must be useful in understanding the world; it must help explain what occurs and anticipate what is likely to occur”,19 then we are forced to recognize that confederative theory fails to explain the right to separation.

1.1 From a union of sovereign states …

Fortunately, the pure theory of law remains unaffected by factual reality.20 This is fortunate because, as we shall see, history has no regard for theoretical constructions.

1.1.1 Confederation as viewed in Kantian political philosophy

If one looks at theoretical constructions, and this is particularly true in the work of Emmanuel Kant to edify a cosmopolitical constitution, confederation is never considered otherwise than as a free union. Kant, who dreamed of a confederal union between states to ensure the peaceful coexistence of peoples and eradicate war,21 compared his model to a “permanent congress of states”.

←27 | 28→

The formula is not without ambiguity, but – properly understood – confirms the idea that a confederal pact offers each member the right to withdraw freely. Kant defined his congress as “a species of voluntary union of the several States, which should be at all times revocable and not, like that of the States of America, a union founded on a public constitution and consequently indissoluble. It is in this way only that the idea can be realized of a public law of nations, which may terminate the differences between peoples by a civil process, like the judicial proceedings among individuals, and not according to the barbarous manner of savages, that is to say, by war”.22 This, as we can all agree, was an insight of great import for the future. Clearly, the states in a Kantian confederation were given a right of withdrawal. In his Perpetual Peace: A Philosophical Sketch of 1795, Kant described a “federation of free states” (Foederalismus Freier Staaten) bound in an alliance of peace by a joint, revocable pact in which each member state retained its sovereignty, since he was resolutely opposed to the creation of a super-state. This philosophy matches, point for point, the legal framework of a confederation as illustrated today in public law textbooks.

1.1.2 The concept of confederation in public law

Confederation is a well-known model among legal experts,23 given that it is the oldest form of federalism. Some people even trace it back to the Greek amphictyonic leagues. However, to provide a contemporary definition, we can say that a confederation is an association of independent, sovereign states that entrust, by way of an international treaty, the management of certain matters (diplomacy, defence) to a joint organization.

Legal experts find it difficult, today, to find any actual examples of a confederation, or at least any that are unanimously recognized as such. However, legal theory24 generally considers that a confederation has five characteristics that distinguish it from other forms of association or political organization.

←28 | 29→

a. A confederation is first and foremost an association of states, and is not a state itself. It is therefore based on an international treaty and not on a constitution, which distinguishes it from a federal state.

b. A confederation has a restrictively listed set of attributed powers, generally limited to economic, monetary, customs-related or military matters.

c. A confederation recognizes each member’s right of veto on any change to the founding treaty, and even – for less developed confederations – the need for unanimous agreement for all decisions.

d. A confederation has no sovereignty. As a result, there is no “confederal citizenship” and the citizens of each member state have no vote for electing the confederation’s political authorities. Only a member state can have direct relations with its citizens (confederal mediacy versus confederal immediacy).

e. Last, and this is the key point for my purposes, a confederation of sovereign states recognizes each member’s right to withdraw from the association. This is one of the criteria traditionally used to distinguish a confederation from a federal state, which prohibits any form of secession. This binary opposition can be compared to the nature of the founding act: a treaty/pact for the confederation, and a constitution/statute for the federal state.

From this point of view the right to withdraw from the European Union highlighted by Brexit, arising from article 50 of the Treaty on European Union, matches the template (although the right was only formalized in 2009). Because the EU is not a federal state but a confederation, even sui generis, it offers its members a right to withdraw.25

1.2 … to a “perpetual confederation”

However, looking behind the scenes, we can trace a completely different history of confederation which, as we will see, is a long way from the principles set out in textbooks with respect to member states’ right to withdraw.

←29 | 30→

1.2.1 Confederal laws against secession

In fact, the permanency of the confederal union has always been a central objective.

The Delian League, or Athens’ imperialist temptation

From the time of the Delian League, secession was prohibited de facto, and even de jure. We learn this from the Decree of 446–445 BC voted by the Ecclesia of Athens with respect to Chalkis. Athens required its Chalkidian allies to swear an oath, on which they could not go back, to have the same friends and the same foes, which, within a military alliance, was equivalent to being unable to leave.26 The Decree was adopted after the Peace of Callias had been signed, in other words after the original goal of the alliance, to defend the cities from the Persian threat during the Greco-Persian Wars, had been attained. The alliance had become perpetual, despite the disappearance of its original objective. In reality, the cities that rebelled were forced back into line, or simply razed.27

Other military alliances, including some in the 20th century such as the Warsaw Pact, followed the same trajectory. It is hard to forget the tanks entering Budapest in 1956 when Hungary was considering leaving the pact signed a few months earlier; or the invasion of Czechoslovakia in 1968, which put an end to any reforming zeal in the country.

The Iroquois Confederacy

Thomas Jefferson could rely on three sources of inspiration for drafting the Declaration of Independence. The first was theoretical, the work of John Locke. The second was historical, the secession of the United Provinces of the Netherlands. The last was Indigenous, in the form of the Iroquois League.28 This confederacy, the most powerful in ←30 | 31→North America for almost two centuries before the arrival of Christopher Columbus, was based on a “constitution” that was transmitted orally and then set down in writing in 1720. It was known as Gayanashagowa, which can be translated as “Great Law of Peace”. However, the council of the five Iroquois nations condemned secession, which was dealt with in the same article as treason.29

The American Confederation of 1777

The “Articles of Confederation” signed by the thirteen original states in 1777 follow the same path. Although they specify in article 2 that each state retains its sovereignty, they also exclude any right of secession. The thirteenth and last article clearly states that “the union shall be perpetual”. A state had no right to unilateral withdrawal, and this prohibition of secession can be seen as a constant feature of the tradition in the New World.

The Swiss Confederation and Sonderbund War

After the fall of the First French Empire in 1815, a federal pact was concluded between the Swiss cantons, replacing the Act of Mediation imposed by First Consul Bonaparte in 1803. However, after recovering their sovereignty, the cantons quickly experienced tension between the rural, conservative and Catholic cantons, on the one hand, and the more industrialized, liberal and Protestant cantons, on the other. Anti-Catholic measures were adopted by the liberal cantons, leading seven conservative cantons to form a defensive alliance in 1845, quickly referred to as a Sonderbund (separatist alliance) by its detractors. In 1847, after a tight ←31 | 32→majority vote, Parliament ordered the dissolution of the Sonderbund. It is important to note that the pact of 1815 prohibited the cantons from “forming between them any link detrimental to the federal pact” (article 6). After the separatists refused to disarm, a short war was fought and quickly won by the Confederation. There is nothing surprising about this war, since even the pact of 1291, between the three communities of Uri, Schwyz and Unterwald, considered to be the founding pact of Swiss federalism, was intended to be perpetual. And fifteen years before the Sonderbund War, the Confederation had opposed the withdrawal of Neuchâtel.

In practice and in historical fact, then, the distinction between a confederation and a federal state, based on the licit or illicit nature of secession respectively, cannot be demonstrated.

1.2.2 The philosophical turning-point in the 16th century: the Dutch influence

The 16th witnessed a fundamental turning-point in the way in which secession was addressed. At the start of the Eighty Years’ War began, for the first time ever, a written document supported the legitimacy of secession. The Dutch, in revolt against Spain, claimed their independence and religious freedom, and rejected the centralism of the Hapsburg Empire. The “Act of Secession”, adopted on July 22, 1581,30 was followed four days later by the Act of Abjuration signed in The Hague, proclaiming de facto the independence of the United Provinces.31 Both the time and the place are linked, obviously, to the publication about twenty years later of Politica by Althusius, the father of modern federalism. As a municipal syndic defending the freedom of his city, Emden, from the Count of Frisia, he logically drew inspiration from the Dutch experience to design a plural political order implicitly based on the idea of revocable consent.

1.2.3 Universalization of the right of secession: the US Declaration of Independence

We know that events in Holland were a source of inspiration for Thomas Jefferson when he drafted the Declaration of Independence. ←32 | 33→However, the result was more than just one more chapter added to the history of secession, because the Americans universalized the right to independence. In 1581, it had been seen simply as the right of the Dutch to free themselves of the Hapsburgs, but the American Revolution made the right of secession a right held by all peoples in the world to recover their freedom. The Declaration of Independence opens by pleading for what we could, today, call “secession as a remedy”. The thirteen colonies considered it important to justify breaking away by listing the innumerable harms caused by the Crown’s exactions.32 The right of secession was not presented as an unconditional right, but as a response to injustices listed point by point, in order to convince the “opinions of mankind” in the words of Thomas Jefferson.

2. The federal state: an “indissoluble union”?

The history of the 19th and 20th centuries includes several examples of peaceful secession: the secession of Hungary from Austria in 1867,33 of Norway from Sweden in 1905,34 and of Iceland from Denmark in 1944.35 However, if we look in more detail at each secession, its consensual character appears to depend far more on the time chosen, and on the political opportunity created by the weakening of the central state, than on a truly consensual arrangement. Another example is the secession of Singapore from Malaysia in 1965. Is it reasonable, though, in this case to posit a peaceful succession, given that Singapore was in reality ←33 | 34→thrown out of the federation less than two years after joining it because of economic and racial conflict?36

In fact, it is hard to identify examples of secessions in recent history that have not led to war or strong regional tension – one immediately thinks of the federation of Eritrea and Ethiopia, organized by the UN in 1952, which degenerated into a thirty-year war. Other examples of this type include Kosovo, the Tamil situation in Sri Lanka, Bangladesh, and Chechnya, among many others.37 States that expressly recognize a right of secession are extremely rare. However, the sometimes ephemeral nature of some federations, if not simple prudence, calls for a legal way to settle the question.38

2.1 Federal positive law

2.1.1 Federal constitutions expressly allowing a right of secession

Only two purely federal states have recognized a right of secession, which is less than the number of unitary states that have done so (for example, Denmark, Liechtenstein and Uzbekistan).39

←34 | 35→

a) Constitutions that recognized a right of secession in the past

Soviet Union

The Soviet Union was the first federal state to include a right of secession, in black and white, in its constitution.40 The right was present from the outset, since the Constitution of 1924 provided that “each federated republic is guaranteed a right to withdraw freely from the Union” (article 4). A similar provision was found in the Constitutions of 1936 (article 17) and 1977 (article 72), which state that “each republic is free to secede from the USSR”.41 However, we know how this fine-sounding principle was applied, since for many years Moscow repressed any movement considered to reflect “exaggerated nationalism”.42

Lenin believed in the right of nations to self-determination, but in reality the defence of the proletarian revolution was more important than the rights of the federated nations.43 His support for self-determination was really only a step in the process leading up to and/or necessary for constituting Marxist unity.44 Lenin did not introduce the right to self-determination to prepare for the dismemberment of the Soviet Union, but simply wanted to appease national fears while shoring up proletarian unity. This allowed Lenin to say that “recognition of the right to secession reduces the danger of the ‘disintegration of the state’ ”.45

←35 | 36→

The Burman constitution of 1947

The constitution of the Union of Burma in 1947 included, in Chapter 10, express recognition of the fact that “each state is entitled to separate from the Union”46. However, Myanmar went on to experience a long succession of military coups that made the principles of the federation inoperable. The right disappeared in 1974 when the Constitution of the Republic of the Union of Myanmar was adopted, authorizing nothing more than local autonomy until the constitution of 2008.

Former Yugoslavia

The constitution of Yugoslavia, amended in 1974, gave the republics a right of secession,47 but in the end Yugoslavia followed the tradition of socialist states that made the right of secession an example of petitio principii. When Croatia and Slovenia unilaterally declared their independence on June 25, 1991 the army of the Socialist Federal Republic of Yugoslavia (mainly composed of Serbs and Montenegrins) invaded Croatia and Slovenia to prevent them from seceding.

State Union of Serbia and Montenegro, 2003

Following the collapse of Yugoslavia, the State Union was formed in 2003 under the high patronage of the European Union, which persuaded Montenegro not to opt for independent statehood but, instead, to form a new, looser federation with Serbia. Cooperation was limited to a few powers. The founding document, signed in 2002, stated in its preamble that “after a period of three years, Serbia and Montenegro will have the right to initiate a procedure to re-examine their national status, in ←36 | 37→other words to leave the state union”.48 To nobody’s surprise (given the scepticism on both sides) a referendum on independence was held in Montenegro as soon as the probationary period ended. The independence side won with just over 55 % of the votes cast, leading to the dissolution of the federation, this time with no violence.

Transitional constitution in Sudan, 2005

The most-recently created member of the international community, South Sudan, results from a constitutional right of secession. The 2005 peace agreement led to the adoption of a transitional constitution which, in articles 118 and 222, enshrined the right to self-determination and the holding of a referendum to authorize the secession of the south of the country.49 The referendum was held in January 2011, following the transitional period provided for in the constitution, leading to a clear victory for independence.

However, it was the prevailing violence (the second Sudanese Civil War lasted from 1983 to 2002) that justified the United Nations Mission to South Sudan and forced the government in Khartoum to consent to the agreement.

b) Constitutions currently recognizing a right of secession

The federal or quasi-federal states that currently recognize a right of secession can be counted on the fingers of one hand:


First, Ethiopia, which – on paper at least – offers a broad right of secession, not only to the nine federated states but also to any nation, ←37 | 38→nationality or people in Ethiopia.50 The constitution also sets out the precise procedure and the conditions that must be met – first, a two-thirds majority vote of the country’s legislative council, followed by a referendum of the local population organized by the federal government. If the referendum favours secession, discussions are then held to define it in more detail. However, the actual importance of this right to self-determination needs is questionable. Sports fans probably remember an Ethiopian marathon runner who won the silver medal at the Rio Olympics and who crossed the finish line with his raised arms crossed, to draw attention to the arrest and repression of members of the Oromos, one of the country’s ethnic groups. The Oromos were attempting to exercise their right of self-determination, in particular via the OLF (Oromo Liberation Front), which was considered to be a terrorist organization by the government.

Saint Kitts and Nevis

Next, the two Caribbean islands of Saint Kitts and Nevis, which gained their independence from Great Britain in 1983. Nevis was authorized to separate from the island of Saint Christopher (Kitts) and this right was almost implemented in 1997 after the election victory of a secessionist party. The federation only survived because of the rigid procedure that required a double majority before Nevis could gain its independence: a two-thirds majority vote of the legislative assembly of Nevis plus a two-thirds majority vote in a referendum of the inhabitations of Nevis. Although the assembly voted unanimously in October 1997 in favour of secession, the population was less enthusiastic, since “only” 61.7 % voted for secession in August 1998,51 making it a close-run race.

This case highlights the question of the majority required to enable secession. We know that this question has not been settled in Canada, whether by the Supreme Court or in the Clarity Act. Is a simple majority sufficient? Should a larger majority be required, at the risk of undermining ←38 | 39→the political potential for secession? Should a two-thirds majority be required, as in Saint Kitts and Nevis? Montenegro has already been mentioned here; in this case the European Union imposed a threshold of 55 % of votes in favour of separation from Serbia.52


As a provocation, France can be added to this list. Its inclusion is provocative in the sense that France is – as we all know – a paragon of the unitary state. However, one section of the constitution is headed “Transitional provisions concerning New Caledonia” and deserves our attention here. The legal relationship it defines contrasts strongly with the ties that generally bind a community, even decentralized, with the central state. This results from the Nouméa Accord, which led to the inclusion of the section on New Caledonia and which included the establishment of “shared sovereignty” between France and New Caledonia, an approach applied by neither Spain nor Italy.53 The federal nature of the relationship, which reflects a form of federalism by disassociation, is clear.54 In fact, article 77 of the French constitution [which states that “the interested populations of New Caledonia will be asked to decide on accession to full sovereignty”] indicated that a referendum on self-determination would be held by 2018. The referendum has taken place, and French loyalists won it with over 56 % of the votes cast.

However, in France the right of secession is not limited to New Caledonia, and the question must be examined against the background of the right to independence of former colonies. The constitution of 1958 gave them – temporarily – a right of secession. Article 76 (in the 1958 constitution) provided for a period of four months during which overseas territories could choose either to remain in the Republic or to become an independent state (and to join the “Community”).55 Even ←39 | 40→after that period, the right of secession did not expire. Based on some audacious readings of the jurisprudence, France todays still maintains a right of secession. This may appear surprising since it is not clear from the constitution itself, even though the second paragraph of the preamble recognizes the right to self-determination of peoples overseas. In fact, a constitutional judge took it upon himself to construe a “right of secession” from another existing provision, article 53 of the constitution, which deals with “transfers, exchanges or adjunctions of territory”56. The first and third paragraphs of this constitutional provision set two conditions: first, a parliamentary votes in favour of a law authorizing secession and, second, the consent of the populations concerned.57 This was the process implemented by the French authorities in 1975 to introduce self-determination for the Comoros. This construction of article 53 was validated by the constitutional council which even gave it general scope.58 As a result, the government was able to pass a law authorizing the secession of this Indian Ocean territory following a referendum, and the Comoros form, today, an independent state. In ←40 | 41→other words, France today still has a procedure for the secession of a territory from the Republic.59

2.1.2 Federal constitutions excluding all forms of secession

The unitary Chinese state prohibits all forms of secession (article 52). The anti-secession law of 2005, which specifically targets Taiwan, reaffirms this position when it states that “the state will in no case authorize the secessionist forces supporting the independence of Taiwan to separate the island from China, under any name or by any means whatsoever”. China gives itself the power, if necessary, to use “non-peaceful means” (article 8). This is clearly a provision to be borne in mind following the breakthrough made by independence supporters at the local elections in Hong Kong in September 2016. However, few federal or quasi-federal constitutions follow China’s example in condemning all secessionist options so firmly. The approach is often more subtle even if the end result is identical: the prohibition of secession.

Union of the Comoros

The 2001 federal constitution of the Comoros is clearly drafted. Article 7-1 states that “Any secession or attempted secession by one or more autonomous islands is prohibited.”60 The provision highlights ←41 | 42→one of the main practical difficulties raised by any secession: “trapped minorities”. The exclusion in fact targets a specific case: the island of Mayotte. The Comoros, a small archipelago in the Indian Ocean north of the Mozambique channel, obtained independence from France in 1975, but relations between the two countries quickly descended into conflict because of the way in which the results of a referendum were interpreted. Certainly, 95 % of the archipelago’s population [spread over four islands] had voted for independence. However, one island – Mayotte – had voted by a large majority to remain part of the French Republic. This gave rise to a conflict that remains unresolved: which takes precedence, the territorial integrity of the Comoros or the choice made by the population of Mayotte and its own right to self-determination61? The French parliament eventually decided to treat the results of the referendum on an island-by-island basis, Mayotte remained French, and France has received at least twenty condemnations from the UN. This example – however circumscribed – deserves our attention because of the numerous inherent difficulties involved in secession. However tenuously, a parallel can be drawn with Kosovo, where 10 % of the population is Serb. Similarly, Brussels, although it has a French-speaking majority, is located in the Flemish part of Belgium.

Bolivarian Republic of Venezuela

The preamble to the constitution of Venezuela, which promotes the right to the self-determination of peoples, should not be taken at face value. The constitution contains, in fact, a long litany of provisions designed to protect the territorial integrity of Simon Bolivar’s birthplace. For example, article 126 carefully denies the right of the indigenous peoples to self-determination,62 while article 159 prohibits the states from doing anything to harm the country’s territorial integrity.63

←42 | 43→


The Federative Republic of Brazil was formed at the end of the 19th century. The constitution of 1891, promulgated after the proclamation of the Republic, repudiated the Empire and centralism and opted for federalism. This mirrored the republican slogan of “Centralization, Secession; Decentralization, Unity.”64 The same unity is, today, imposed. The first article of the 1988 constitution states that “The Federative Republic of Brazil (is) founded on the indissoluble union of the states, the municipalities and the federal district […]”.65


Australia has already faced secessionist movements, for example in Western Australia where a referendum was held in 1933. Although supported by almost two thirds of the electorate, the referendum result still needed to be endorsed by the British parliament, which refused to do so on the basis that the preamble to the 1900 constitution of Australia stated that “the people of New South Wales, Victoria, South Australia, Queensland, and Tasmania, […], have agreed to unite in one indissoluble Federal Commonwealth under the Crown of the United Kingdom of Great Britain and Ireland […]”.


The constitution of the Swiss Confederation contains no specific provisions on secession, but must be read in light of the spirit of the text. The absence of any “Schwexit” clause in the constitution of a country that promotes its own concept of willensnation is not accidental, but stems from a conscious decision by the founding fathers of 1848. In the wake of the Sonderbund War, it was feared that a right of withdrawal would imperil the newfound Swiss cohesion. The silence of the constitution of ←43 | 44→1848, which has been reconfirmed since, must be understood as a way to withdraw the right of secession from the cantons, and was intended as a new framework that did away with the principle of an alliance between sovereign cantons, re-established in 1815. The sovereignty of each canton is now constitutionally subservient to that of the Confederation.66


Like many other similar documents, the Mexican constitution does not mention “secession”, but leaves the reader in little doubt as to its unconstitutionality. Although article 2 of the 1917 constitution enshrines “the right of the indigenous peoples to self-determination”, the right is placed within a strict constitutional framework and cannot undermine the “preservation of national unity”. The constitution also specifies, at an early point, that “the unity of the Mexican nation entails its indivisibility”. Self-determination is reduced to its internal aspect: autonomy within the Mexican state.

And this should come as no surprise. The first federal document in Mexico, the constitution of 1824, written at a time when national disintegration was apprehended (Guatemala had seceded the previous year), aimed – through federalism – to defuse secessionist leanings and safeguard the union between the country’s various regions.

2.1.3 Constitutions that remain silent on the question of secession

It is true to say that most federal constitutions remain silent on the question of secession. This is certainly the case in the Federal Republic of Germany, where nothing in the Basic Law (or in any other law) regulates secession. However, without wishing to create any false controversy, the question is superfluous given that secessionist problems are not part of German political reality, other than in an anecdotal way, even in Bavaria. In any case, in other countries the silence of the constitutional texts gives their Supreme Courts more interpretational scope and leaves room for Realpolitik.

←44 | 45→

a) Interpretation by Supreme Court justices: a centripetal constitutional force

United States

The background to the US Civil War is well known. Following the election of Republican politician Lincoln67 and his project to abolish slavery, the southern states – sure of their legal position – seceded and formed the Confederate States of America. It is important to note that states’ rights had strong support in public opinion, not only in the South but also in the political class in the North. It is enough to look at a few episodes from American history, or a few statements selected from those made in the years leading up to the war, to understand the constitutional basis for the dispute.

This was not the first time that the United States had faced the possibility of a secession. During the War of 1812 between the United States and the United Kingdom, the New England states – opposed to the war launched by the federal government – had threatened to withdraw from the Union unless a compromise was found. Although nothing came of this in the end, it provided a foretaste of the Nullification Crisis that arose in 1830 and set the scene for the first secessionist attempt in the South, when South Carolina threatened to cancel the federal customs duties known as the “abominable tariff”. The duties were applied to imports from Europe with the objective of protecting nascent industries in the north-eastern states. The conflict worsened when constitutional lawyer Calhoun, Vice-President of the United States, set out some theoretical legal considerations while President Andrew Jackson threatened South Carolina with military action.68 The southern state retaliated with its own threat, to secede, and in fact secession was commonly used as a threat in the first decades of the American federal union.69

←45 | 46→

The right of secession was not necessarily contested in the North or by leading lawmakers. Even Thomas Jefferson defended a not dissimilar position in favour of freedom of choice by individual states. In a private letter, he wrote in 1816 that “if any state in the union will declare that it prefers separation […] to a continuance in union […], I have no hesitation in saying ‘let us separate’.”70 The sixth US President, John Quincy Adams (son of the second president, John Adams) stated in 1839 – a few years after leaving office – that “If the day should ever come […] when the affections of the people of these states shall be alienated from each other […] far better will it be for the people of the disunited states, to part in friendship from each other, than to be held together by constraint.”71 And in 1860, a few months before the Civil War, President Buchanan, even though he believed secession to be illegal, still condemned the use of force.72 Secession was clearly something that could be defended under the terms set in Philadelphia; at any rate, it did not justify federal execution, in other words the use of military force.

However, over the space of a few months secession became a casus belli, when the federal government declared the actions of the confederate states illegal and launched military action, a clear sign that nationalist ideology had replaced the idea of a pact. Of course, the context was specific: a fight between states either against, or for, slavery, but we know that in almost all federations the federal state tends to take action against the federated states in the name of equality and the promotion of individual rights.73 Lincoln stated this unequivocally and his statement ←46 | 47→goes beyond the context of a battle for individual rights: the states enjoy no right to leave the union, which therefore is perpetual in nature.74 This was confirmed at the legal level by the Supreme Court ruling in the famous case of Texas vs White (1868).


The Canadian constitution contains no provisions dealing directly with secession. It is only possible to note the absence of a provision similar to the one deliberately included in the preamble to the Australian constitution. The Canadian confederation does not claim to be an indissoluble union.

However, this did not prevent the Supreme Court from ruling that Québec possessed no unilateral right of secession, either under international law or under the Canadian constitution. Although it did not exclude secession on principle, it made it subject to three conditions: a referendum clearly manifesting Québec’s desire to leave Canada; negotiations between the federal and Québec governments based on four fundamental principles (federalism, democracy, constitutionalism and the rule of law, and the protection of minorities); and an amendment to the constitution to ratify the secession. The referendum itself would therefore not amount to secession, but a vote by a clear majority of Quebecers in favour of secession would create a constitutional obligation for the other players in the federation to negotiate.75

←47 | 48→


Bosnia-Herzegovina, which in early 2016 submitted a request to join the European Union, has been divided since the Dayton Agreement of 1995 into three federated entities: the Federation of Bosnia and Herzegovina, the Republika Srpska and the Brčko District. The constitution of Bosnia-Herzegovina, like those in Canada and the United States, does not mention secession explicitly, either to authorize or prohibit it. However, given the strongly separatist context (especially among the Serbs of Bosnia), and in particular since the independence of Kosovo was recognized,76 the constitutional court was asked, as early as 1998, to specify the status and rights of each of the country’s entities. This resulted in a ruling that denies the sovereignty of the entities, denies their standing as states, and above all rejects any right of secession.77

Russian Federation

At first glance the Russian constitution appears receptive to a right of secession. Faithful to Soviet tradition, the preamble enshrines the right of peoples to self-determination. However, a reading of article 5, § 3, of the constitution quickly suggests another interpretation, since it states that “The federal structure of the Russian Federation shall be based on its State integrity, the unity of the system of State power […] and self-determination of peoples in the Russian Federation.” Understood in this ←48 | 49→way, the Russian constitution therefore enshrines only a right to internal autonomy,78 a situation confirmed by the constitutional jurisprudence.

In 1995, Russia’s constitutional court was asked to rule on four presidential decrees connected with the dispatch of troops to Chechnya, giving it an opportunity to review whether or not a right of secession existed. The decision is interesting in several aspects.79 The court considered that the constitution of the Russian Federation, like the previous constitution of 1977, gave no unilateral right to the federation’s entities to change their status or, a fortiori, secede.80 The judgement continued by emphasizing that “State integrity is one of the foundations of the constitutional system of the Russian Federation”. The jurisprudence was further strengthened when the court was asked to rule on the attempted secession of the Republic of Tatarstan. It considered that the right to self-determination provided no legal foundation for a unilateral secession, but had to be exercised in accordance with the principle of territorial integrity.

Republic of South Africa

A similar logic prevails in South Africa, where the “rainbow nation” has a hybrid status strongly influenced by federalism. As in Russia, the constitution recognizes the right to self-determination, raising the question of whether this constitutional principle entails a right to separation. Fortunately the constitutional court in South Africa had an opportunity to rule on this question when it certified the constitution in 1996,81 and its answer is not surprising: the right to self-determination, ←49 | 50→while respecting the sovereignty of the state, is to be understood only in its internal sense.82

b) Federal realpolitik

Some states, whose constitution remains silent on the question of secession, suggest that the political practice, or history, of the country concerned should guide the debate.


This applies, for example, to Belgium, where the constitution does not give the regions or communities a right of secession. Some eminent Belgian constitutional experts argue that a unilateral secession could only occur outside any legal framework.83 In reality, however, the debate is essentially political.84 Legal arguments are not predominant in the public debate to counter the discourse of the Flemish nationalists. It is important to note that the case of Belgium is unusual compared to the general run of nations invoking the right of secession, since the Flemish community is not a minority and in fact represents 60 % of the country’s population. Belgium itself emerged from a (peaceful) secession from the United Provinces. The dissociative federalism that characterizes the Belgian state is seen by many as a transitional state prior to separation. Secession is therefore an integral part of Belgium’s history, and nobody would seriously consider refusing the secession of Flanders if it set itself firmly upon that path.

←50 | 51→


It has been said about India that it is “unitary in spirit, but federal in form”, or even “a unitary state with subsidiary federal principles”,85 and it is true that the constitution never uses the term “federation” despite this being an objective of the constituting parties in 1949, one of which stated that “what is important is that the use of the word ‘Union’ is deliberate […] The drafting committee wanted to make it clear that though India was to be a Federation, the Federation was not the result of an agreement by the States to join in Federation, and that the Federation not being the result of an agreement, no state has the right to secede from it […].”86

In any case, it is hard to imagine that Kashmir – or at least the part under India’s administration, Jammu and Kashmir – where independence is invoked by certain movements and where tension with Pakistan is high,87 could one day be left to secede by the federal government.88


Although the Nigerian constitution does not expressly forbid secession, it appears to be prohibited in practice. The Igbos ethnic group, which represents just under 20 % of the country’s population and is concentrated in the southeast, would like to see a right to secession incorporated into the constitution. The Igbo homeland, the former Republic of Biafra, whose attempt to secede in 1967 led, in less than three years, to the death of almost one million people, filed a request for ←51 | 52→a constitutional amendment in 2014,89 but to no avail. On the contrary, the Biafran independentist Nnamdi Kanu, director of Radio Biafra (based in London) and a leader of the prohibited movement “Indigenous People of Biafra” was arrested in October 2015 and tried in March 2016 by the High Court of Abuja on the charge of “propagating a secession agenda”.

Federated States of Micronesia

Micronesia is a federal state in the Pacific with four federated states. With barely 100,000 inhabitants on 700 km2 of land emerging from the ocean, it is not immune to separatist temptations – a referendum on secession was scheduled to be held on March 3, 2015 in the state of Chuuk. Although no legal provision explicitly prevented this, the President of Micronesia, Manny Mori, campaigned against independence on the basis of the unconstitutionality of the proposed secession and the need to first amend the founding legislation (requiring the assent of 75 % of voters and three quarters of the states).90 The proponents of secession invoked international law and the example of Kosovo, unsuccessfully since President Mori was careful to specify that a “yes” vote in the March 3 referendum would not necessarily make the State of Chuuk an independent nation. In the end the referendum was never held, being postponed sine die by the governor of the state, who considered that the public needed to be made more aware of the issues.


Malaysia has no right of secession. As is well known, Singapore left the federation in 1965, but as the result of exclusion rather than secession. Recent events have shown that the federal government tends to consider secession as an act of sedition. This can be seen in the authorities’ reaction to the emergence in recent years, in particular via social media, of a separatist movement in the eastern states of Sabah and Sarawak (Sabah Sarawak Keluar Malaysia). Taking advantage of Malaysia Day, the Prime Minister has since pointed out that these states on the island ←52 | 53→of Borneo are an integral part of Malaysia and that the very question of separation is inconceivable. The Attorney General, Tan Sri Abdul Gani Patail in turn specified that secession was against the “spirit of the constitution”. It is important to note that the old Sedition Act, which dated from colonial times, has reappeared at a time of political repression against opponents of the regime, and plans have been proposed to make secession a criminal offence.


Iraq acquired a federal constitution in 2005 in well-known circumstances, mainly to deal with the Kurdish question, since it was a condition set by the Kurdish people before joining the “new Iraq”. The land is strategically placed, since Kurdistan has substantial oil reserves in the North,91 and the difficult period Iraq is experiencing has once again highlighted this issue. Already in 2006, the main leader in Iraqi Kurdistan had threatened secession if Prime Minister Nouri Al-Maliki confirmed his choice of the flag formerly used by the Saddam Hussein regime as a national emblem. At the time, Iraq President Jalal Tarabani had attempted to offer reassurance by refuting “any idea of a Kurdistan separated from Iraq”. The separatist issue, however, refuses to disappear, and a few months ago Massoud Barzani, leader of Iraqi Kurdistan, called for a referendum on the creation of a Kurdish state. The Constitution remains silent on this issue. Like any other constitutional text, of course, it mentions the unity of the country, without indicating any clear conclusions, but Bagdad is not using legalistic arguments. A few months ago Prime Minister Haider al-Abadi, on a trip to Berlin, shared his hope that Kurdistan would continue to “be part of the country”, pointing out that the area is “part of Iraq and will, I hope, remain so”.

2.2 Using constitutional theory to cut the Gordian knot of secession

The question of secession pivots on the definition of the state. There is a historical reason for this: the federal state is a hybrid model containing two irreconcilable paradigms.

←53 | 54→

2.2.1 The trap set by the syncretism of the federal state

The federal state is a hybrid. Its father is the state. As conceptualized by Jean Bodin (Les Six Livres de la République, 1576), the state is founded on indivisible (and perpetual) sovereignty, with the ultimate goal of providing security, as shown by Thomas Hobbes (Leviathan, 1651). But the federal state also has a mother, federation, inspired by Althusius (Politica, 1603) and, later, Kant, and is based on pluralism. From its father, the federal state has kept the imprint of its origin: a social contract between individuals who abdicate some of their powers to a state, which then becomes the sole holder of sovereignty. As a result, there is no right of secession, which would be synonymous with anarchy since it would give right to the – not inconsiderable – number of around 6,000 ethnic groups identified on the planet to set up their own state. On the other hand, through its mother, the federal state has retained a focus on particularities, which must be respected or – if it is not respected – can create a right to separate from a union which would have become form of tutorship. In short, the federal state is a child of the state,92 with its Hobbesian conception of sovereign political authority, and the federation, derived from a multiplicity of holders of political authority, and a symbol of autonomy and freedom.

However, federal unions have, almost systematically, despite being initially the result of a pact, made federalism subservient to the construction of the nation-state.93 One example is the United States, which has gradually become a single nation, with the Civil War marking the starting-point for the transition. In English, the term “United States” only began to be used in the singular following the victory of the Union in 1865.94 The rejection of secession created a lastingly unitary reading ←54 | 55→of the Philadelphia Convention. In any case, the question of secession remains insoluble if it is seen as a binary choice between the supporters of state unity and territorial integrity, on the one hand, and the partisans of States’ Rights, on the other. This is why we must be beyond this apparent contradiction.

2.2.2 Redefining the constitutional basis for secession

In constitutional terms, it is not appropriate to consider secession from the point of view of a people’s right to self-determination. This principle comes from international law, and it has been remarked on numerous occasions that it is more akin to a political principle, in its effective form, rather than a rule of normative law. As a result, secession must be given a suitable legal definition in the field of constitutional law. The question that must be answered is this: what does any secession ultimately consist of?

Regardless of the procedure used, whether a unilateral declaration of independence95 or a referendum on secession96, secession has no constitutional and/or institutional consequence for the residual state until its fundamental law is revised to take note of the departure and suppress obsolete provisions. The legal value of a local referendum on secession is questionable and even, in some cases, null,97 because the thorny issue of ←55 | 56→legality can only be dealt with through revision. This approach offers a way to dispel the uncertainty about whether a federation is based on a pact (foedus) or a constitution (fundamental law).

a) Partial versus total revision

In the field of constitutional law, two types of revision are generally considered: partial revision, and total revision. Partial revision uses constitutional amendments to rewrite/add/strike out one or more provisions. Total revision is of another nature altogether. It may involve the amendment of a substantial portion of the constitution or the substitution of a new text; or else the revision of a fundamental principle of the existing constitutional order (for example, when a republic becomes a monarchy). In both cases, a total revision in fact masks an abrogation of the existing constitution. However, many eminent legal experts believe that an abrogation of the constitution contravenes constitutional legality.98 Raymond Carré de Malberg, repeating the position of Jellinek ←56 | 57→on the self-limitation of the state, considered that “however absolute the power of the state, and even if it was legally possible for it to do everything, it cannot abolish the legal order and found anarchy, because it would be destroying itself”.99 In his Constitutional Theory, Carl Schmitt refused to admit that constitutional laws could abrogate the constitution.100 The state, a “mortal God” in the apt description of Thomas Hobbes, is based on a constitution that has a “claim to eternity”.101

At a theoretical level, a total revision – an euphemism for the disappearance of the sovereign’s work – can only be unconstitutional. At a formal level, a constituted power cannot dissolve the work of the original constituting power. Taking a material approach to the law, it is necessary to state that the state cannot itself abrogate its constitution.102 “Political suicide is not a legal category.”103 On this basis, Georg Jellinek, who contrasted the right to leave, characteristic of a confederation, with the idea of the state, could logically write that “a union under public law such as the state […] can never be dissolved, legally, through the will of its members”.104 Only the people, by a revolutionary act (in the legal sense, meaning an upheaval of the established constitutional order), ←57 | 58→can adopt a new constitution; but it is not possible – legally – to amend an existing constitution completely or fundamentally.105 As a result, if the text says nothing about a right of secession, one of two scenarios is possible:

Scenario 1: If the secession of a province requires a total revision of the federal constitution, it is unconstitutional.106

Scenario 2: If secession requires only a partial revision of the constitution, the revision itself is constitutional, and the federal authorities are then responsible for noting the choice expressed by a province and launching a revision process in accordance with the constitution.

b) Actual cases

As we have seen, most constitutions fail to mention secession, and it is not possible to define a general rule to interpret their silence. Each constitution must be interpreted in situ to ascertain if the departure of a member state will require a partial or total revision.

Secession of a Belgian community

Let us look at the case of Belgium. The secession of Flanders would lead, a minima, to changes to seventy-five out of just under two hundred articles in the constitution, raising the question of whether the Kingdom ←58 | 59→would maintain its legal personality if its constitution were to be so substantially amended. Beyond the quantitative aspect, it is above all the nature of the regime that would be irremediably affected. Belgium is built on its linguistic and cultural polarity, which is at the core of the federal state. As pointed out by Belgian constitutional expert Marc Verdussen, “if you remove one pole, the very foundation of the state will collapse”.107 If the Flemish community left, the Walloon community would be left on its own (ignoring the small German-speaking community).108 The residual Belgian state would become a Walloon state, and federalism would be only an empty shell. At this point, it would be appropriate to ask if Wallonia could claim status as a “successor state” under international law.109 The comments of Alexis Vahlas, a specialist in the question of state succession, are relevant here: “if […] secession occurs, but involves most of the population and territory of a state, or the seat of its government authorities, it will probably lead to the dissolution of the state. In this case, the remaining portion may be seen as being so different from the ←59 | 60→parent state as to no longer be considered as the state that succeeds to its legal personality.”110 This is exactly the case that would apply in Belgium.

To return to domestic law, it is not just amendments to the Belgian constitution that would be required – the change would result purely and simply in the abolition of the federal regime, to be replaced by a unitary regime. This type of constitutional upheaval, in the form of a regime change, resembles a total revision which, as mentioned above, is considered to infringe constitutional legality.111 It should also be noted that section 195 of the Belgian constitution, which governs constitutional amendments, appears to prohibit such a substantial change.112

Secession of a German or Austrian Land

Limits on the power of constitutional amendment are a common feature of European constitutions. In Germany, section 79, § 3 of the ←60 | 61→Basic Law and its explanatory note prohibit a total revision, as well as any revision intended to abolish federalism.113 In Austria, total revisions are authorized, but require a referendum of the whole federal population, or in other words majority support for secession outside the separating entity. In reality, however, the question cannot truly be asked in this way in either country.

This is because, in Germany as in Austria, the secession of a Land would not require a total revision. The German constitutional regime would not be deeply affected in its legal shape by the departure of Bavaria. As a result, the secession of a Land, given the lack of a contrary constitutional provision, could legally by ratified by a revision of the constitution.114 The same would apply in Austria as regards the secession of one of the nine Bundesländer (such as Tyrol).

Secession of a Canadian province

Canada is made up of ten provinces, including Québec whose cultural singularity is reflected in the constitution. The word “Quebec” occurs almost seventy times in the Constitution Act, 1867, meaning that many sections would have to be tidied up in the event of secession. However, the nature of the Canadian federal regime would not be deeply affected, and federalism would continue. Obviously, the departure of the province with the most singular nature would make the federation even more homogeneous, but this is a different problem.

This view of the situation is confirmed by the Supreme Court of Canada since, if the Reference Re Secession of Quebec is re-read through the prism of total versus partial revision, the Court has already made up its mind. Québec’s secession would – the Court states – require only a simple amendment of the constitution. The Court is even careful to signal its disagreement with certain authors who see a more complicated situation.115 If Québec secession requires only a partial revision, and this ←61 | 62→is the approach taken by the Supreme Court, then secession would be constitutionally valid.

To sum up, the legality of a secession is, in fine, conditional on the degree of the constitutional interdependency established between the separating entity and rest of the federation. The stronger the interdependency, the more it affects the very nature of the regime (and this is the case in a binational state in which the very reason for the existence of the federation disappears if one half leaves), and the harder it is to claim that the process is legal. The close relationship between federalism and the organic integrity of the union cannot be undone. Only a revolution, in the legal sense, could further the secessionist project. On the other hand, if the existing constitutional regime can survive the departure with a few simple amendments to the constitution, the secession can be considered as legal, since it requires only a partial revision.

1 Tite-Live, Ab Urbe condita libri, Liv. II, par. XXXIII.

2 This time during the second Secession of the Plebs in 449 BC, to denounce political imbalance between the plebeian and patrician classes. It led to the adoption of the Twelve Tables, the first written corpus of Roman law, which had been transmitted orally up to that time.

3 On the other hand, secession has received a lot of attention in international public law, in connection with the right to self-determination. One example is the advisory opinion on Kosovo’s unilateral declaration of independence issued by the of the International Court of Justice which was asked to, but avoided, giving its opinion on recognition in international law of the theory of secession as a remedy: ICJ , Advisory Opinion on the Declaration of Independence of Kosovo, July 22nd, 2010.

4 Jean Voilquin, in his French translation of the History of the Peloponnesian War, used the term “secession” four times; see T. 1, Paris, Librairie Garnier Frères, see Book I, XCIX; Book III, XIII, Book IV, CXXIII & CXXX. However, other translations tend to use the term “defection”.

5 In a contrary example, the United States could have legitimately declared war on France in 1966 when De Gaulle decided to withdraw from the joint NATO command!

6 Jellinek, G., L’État moderne et son droit, t. 1, “Théorie générale de l’État”, Paris, 1900 (2004), p. 269.

7 Kelsen, H., “Théorie générale du droit international public”, RCADI, 1932, t. 42, p. 261; see also Jellinek, G., op. cit., p. 269.

8 The Badinter Commission on Yugoslavia considered in 1991 that “the existence or disappearance of the state is a question of fact”, Advisory Opinion n°1, November 29th, 1991, RGDIP, 1992, p. 264.

9 Combacau, J., Sur, S., Droit International Public, 5th ed., Paris, Montchrestien, 2001, p. 279.

10 It is possible to find some exceptions to this principle, including Bosnia-Herzegovina. The former federated republic of Yugoslavia controlled, between 1992 and 1995, only 20 % of the territory, and yet Bosnia-Herzegovina was presented as a state as early as March 1992. Here, recognition created effective statehood, rather than the reverse. On the other hand, some secessionist movements have been able to establish their authority over a territory without obtaining recognition from the international community: the Bosnian Serves of the Srpska republic; Chechnya between 1991 and 1994, and especially between 1996 and 1999 (following a defeat of the Russian army); and Transnistria, Abkhazia, South Ossetia and the Turkish Republic of Northern Cyprus.

11 Kelsen, H., Théorie pure du droit, 2nd ed., transl. by C. Eisenmann, Paris, Dalloz, 1962. Ideally, Sein and Sollen would be the same; but, as Kelsen points out, there is no causal link between them, meaning that just because a secession passes the factual test, in cannot necessarily be included in a Sollen.

12 Troper, M., “Hans Kelsen”, in Huisman, D., Dictionnaire des philosophes, Paris, Presses universitaires de France, 2009, p. 1413.

13 The expression, used many times since, is borrowed from Max Weber and conforms to his principle of “axiological neutrality”.

14 This is one possible interpretation of the goals of the new and controversial President of the Philippines, Rodriguo Duterte, who announced in June 2016 that he wanted to revise the constitution to federalize the country. One of his stated objectives was to end any separatist temptation for the Muslim minority.

15 And sometimes more quickly than one might think. The goal of the ephemeral Transcaucasian Democratic Federative Republic – one of the shortest-live examples of a multi-national federation – was to unite Georgians, Azerbaijanis and Armenians. The federation was unable to reconcile the divergent national interests and, under an Ottoman threat, was dissolved after only three months in existence (February-May 1918) after the successive secessions of the Georgians, Azerbaijanis and Armenians.

16 Kymlicka, W., “Is Federalism a Viable alternative to Secession?”, in Lehning, P. (ed.), Theories of Secession, London, Routledge, p. 111–150; Nguyen, E., Les nationalismes en Europe. Quête d’identité ou tentation de repli?, Le Monde, 1998, p. 193; Snyder, J., “La gestión de la etnopolítica en Europa Oriental: una valoración de los enfoques institucionales”, in Ferrero, R. (ed.), Nacionalismo y minorías en Europa Central y Oriental, Barcelona, Institut de Ciències Polítiques i Socials, 2004, p. 56–57.

17 Apart from any hypothetical human rights violations or discrimination against part of the population, the principle of a constitutional democracy’s territorial integrity takes precedence over the principle of self-determination, because integrity is a fundamental factor in its ability to act as a state that guarantees justice, see Buchanan, A., “Theories of Secession”, Philosophy and Public Affairs, vol. 26, n°1, 1997, p. 31–61. This position offers a clear contrast with a number of libertarian philosophers, for whom the right of secession is open-ended, see von Mises, L., Liberalism in the Classical Tradition, transl. by R. Raico, 3rd ed., Irvington-on-Hudson, Cobden Press, 1985, p. 109–110.

18 Other liberal thinkers, although they partially agree with Sunstein about the potentially negative impact of a right of secession on democratic debate and political stability, support recognition for a constitutional right of secession. See Weinstock, D., “Vers une théorie normative du fédéralisme”, Revue internationale des sciences sociales, n°167, 2001, p. 79–87.

19 Zoller, E., “Aspects internationaux du droit constitutionnel. Contribution à la théorie de la fédération d’États”, RCADI, vol. 294, 2002, p. 41–166.

20 In this connection, Kelsen disagrees with the sociological schools of thought that assign knowledge of the facts of human behaviour to legal science for the purpose of defining standards.

21 As Kant says in Perpetual Peace, “States do not plead their cause before a tribunal; war alone is their way of bringing suit.” Kant continues, “[…] reason, from its throne of supreme moral legislating authority, absolutely condemns war as a legal recourse and makes a state of peace a direct duty, even though peace cannot be established or secured except by a compact among nations. For these reasons there must be a league of a particular kind, which can be called a league of peace (foedus pacificum), and which would be distinguished from a treaty of peace (pactum pacis) by the fact that the latter terminates only one war, while the former seeks to make an end of all wars forever. This league does not tend to any dominion over the power of the state but only to the maintenance and security of the freedom of the state itself and of other states in league with it, without there being any need for them to submit to civil laws and their compulsion, as men in a state of nature must submit. The practicability (objective reality) of this idea of federation, which should gradually spread to all states and thus lead to perpetual peace, can be proved.” (Kant, E., Perpetual Peace, A Philosophical Sketch, 1795).

22 Kant, E., Métaphysique des mœurs, 1st part, “Doctrine du droit, Du droit public, Droit des gens”, Paris, Vrin, 2011, par. LIII.

23 Le Fur, L., État fédéral et confédération d’États, Paris, Panthéon-Assas, 2000.

24 Aubert, J.-F., “Essai sur le fédéralisme”, Revue du droit public, n°3, 1963, p. 404–405.

25 Parent, C., “Le droit de retrait de l’Union européenne”, Revue du droit public, n°3, 2016, p. 935–956.

26 Cloche, P., “Périclès et la politique extérieure d’Athènes entre la paix de 446–445 et les préludes de la guerre du Péloponnèse”, L’Antiquité classique, vol. 14, n°1, 1945, p. 94.

27 For example, when Naxos wanted to withdraw in 472, the city was besieged and forced back into the league by Cimon, son of Miltiades. Athens acted in a similar way a few years later with Thasos, and then with Mytilene in 428.

28 It would be a mistake to underestimate the influence of the union of the Iroquois nations on Thomas Jefferson. For instance, the future 3rd president of the United States declared in 1787: “I am convinced that those societies (as the Indians) which live without government enjoy in their general mass an infinitely greater degree of happiness than those who live under European governments.” (Letter to Edward Carrington, Jan. 16, 1787, The Papers of Thomas Jefferson, vol. 11, 1 January–6 August, ed. Julian P. Boyd, Princeton, Princeton University Press, 1955, p. 48–50). It is curious to note that Jefferson substituted the pursuit of happiness, in the Declaration of Independence, for Locke’s right of property.

29 Treason or Secession of a Nation, Article 92: “If a nation, part of a nation, or more than one nation within the Five Nations should in any way endeavor to destroy the Great Peace by neglect or violating its laws and resolve to dissolve the Confederacy, such a nation or such nations shall be deemed guilty of treason and called enemies of the Confederacy and the Great Peace.” “It shall then be the duty of the Lords of the Confederacy who remain faithful to resolve to warn the offending people. They shall be warned once and if a second warning is necessary they shall be driven from the territory of the Confederacy by the War Chiefs and his men.”

30 Mentioned by Martinenko, A., “The Right of Secession as a Human Right”, Annual Survey of International & Comparative Law, vol. 3, n°1, 1996.

31 Recognized by the Peace of Münster included in the Westphalia treaties of 1648.

32 “When in the Course of human events, it becomes necessary for one people to dissolve the political bands which have connected them with another, […] a decent respect to the opinions of mankind requires that they should declare the causes which impel them to the separation.”

33 It is only fair to note that the weakness of Austro-Hungarian Empire following the defeat at Sadowa in 1866 forced Emperor Franz-Joseph to agree to negotiate Hungary’s independence. The compromise was ratified by the Austrian parliament in 1867, after which the Austrian Diet amended the constitution to match the new political arrangement.

34 The decision of the Norwegian parliament was followed by a referendum in which 99 % of voters supported secession. Sweden decided to negotiate rather than be isolated from the international community in the event of war, and the Act of Union was repealed by both parliaments.

35 Taking advantage of the fact that Denmark was still under Nazi occupation.

36 Its departure was imposed by Malaysia, which had its parliament pass a constitutional amendment to remove any mention of its name in the union.

37 One case that should be mentioned is that of the United Arab Republic, founded by Nasser in 1958 as a political union of Syria and Egypt. It came to an end in 1961 following a coup in Syria for the purpose of secession.

38 We can cite the ephemeral Transcaucasian Democratic Federative Republic – one of the shortest-lived examples of a multi-national federation – whose goal was to unite Georgians, Azerbaijanis and Armenians. The federation was unable to reconcile the divergent national interests and, under an Ottoman threat, was dissolved after only three months in existence (February–May 1918) after the successive secessions of the Georgians, Azerbaijanis and Armenians. Other examples are Indonesia (1949– 1950), Libya (1951–1963), Mali (1960), Cameroon (1961–1972), the West Indies Federation (1958–1962), and Serbia and Montenegro (2003–2006).

39 The Act on Greenland Self-Government of 21 June 2009 gave Greenland a right of self-determination that could lead to independence. Chap. VIII, art. 21, par. 1, reads as follows: “Decision regarding Greenland’s independence shall be taken by the people of Greenland.” The micro-state of Liechtenstein, consisting of eleven communities, recently gave them the “right to secede from the state” (Art. 4, par. 2). The Uzbek Republic has given a right of secession to the Republic of Karakalpakstan (Art. 74 C.).

40 Article 4 of the Soviet constitution of 1924 specifies that “Each one of the member Republics retains the right to freely withdraw from the Union”, while article 6 required consent of all the member republics before and modification of the territory or limitation of modification of article 4. The constitution of the federal state was amended under Stalin in 1936, and then in 1977, without affecting the right of secession.

41 Article 70 of the constitution defined the USSR as the result of the self-determination of its nations and the “voluntary association of the soviet socialist republics”.

42 At least until the democratization of the regime launched by Gorbachev and the 1989 election of nationalists (such as the Lithuanians) to the Supreme Soviet.

43 “The aim of socialism is not only to end the division of mankind into tiny states and the isolation of nations in any form, it is not only to bring the nations closer together but to integrate them”, Lenin, Œuvres complètes, Éditions sociales, t. 22, 1960, p. 159.

44 “In the same way as mankind can arrive at the abolition of classes only through a transition period of the dictatorship of the oppressed class, it can arrive at the inevitable integration of nations only through a transition period of the complete emancipation of all oppressed nations, i.e., their freedom to secede” (ibid.).

45 Lenin, Œuvres complètes, t. 20, p. 437 et 445.

46 However, the constitution prudently included a strict formal framework. The constitution was frozen for ten years after its adoption, and a vote by two-thirds of the members of the state council was required along with a referendum of the population of the secessionist state.

47 “The peoples of Yugoslavia, proceeding from the right of every people to self-determination, including the right to secession, on the basis of their will freely expressed in the common struggle of all nations and nationalities in the National Liberation War and Socialist Revolution, and in conformity with their historic aspirations, aware that further consolidation of their brotherhood and unity is in the common interest, together with the nationalities with whom they live, have united in a federal republic of free and equal nations and nationalities and created a socialist federative community of working people” (Preamble).

48 This three-year period began with the adoption of the new constitution, which occurred in 2003.

49 Article 222 stated that, in the six months preceding the end of the six-year transitional period, a referendum would be held in South Sudan (paragraph 1) offering two options: confirmation of Sudan’s unity, or secession (paragraph 2). Article 118, paragraph 1, stated that if the result of the referendum on self-determination confirmed unity, the national legislator had to fulfil its duties in accordance with the provisions of the Constitution; and that in the event of a vote in favour of secession by the South Sudanese people, the seats of the members and representatives of South Sudan in the National Legislative Assembly would be deemed vacant (paragraph 2).

50 Article 39 of the constitution of Ethiopia, 8 December 1994: “Every nation, nationality or people in Ethiopia shall have the unrestricted right to self-determination up to secession.”


ISBN (Book)
Publication date
2021 (June)
Bruxelles, Berlin, Bern, New York, Oxford, Warszawa, Wien, 2021. 198 pp.

Biographical notes

Jorge Cagiao y Conde (Volume editor) Alain-G. Gagnon (Volume editor)

Jorge Cagiao y Conde is a researcher and lecturer (with HDR qualification) in the Department of law and languages at the University of Tours. His book Micronacionalismos. ¿No seremos todos nacionalistas? ("Micronationalisms—are we not all nationalists?") was published in 2018. Alain-G. Gagnon is a professor in the Department of Political Science at Université du Québec à Montréal, where he holds the Canada Research Chair in Québec and Canadian Studies. He is also the director of the new Centre d’analyse politique – Constitution et Fédéralisme (CAPCF: https://capcf.uqam.ca).


Title: Federalism and Secession