The prevalent approach suggests that constitutions are silent about secession and may even implicitly allow it. But an examination of world constitutions reveals that the overwhelming majority of countries vigorously protect territorial integrity. This is true even of countries classified as consociational or consensus democracies. Scholars further point to the existence of secessionist political parties as proof that secession may align with constitutionalism. This Article, however, explains how democracies engage in a delicate game to chase and eliminate secessionist political mobilization. Democracies have been able to conceal their fight against secessionists by creating a large gap between “the law on the books” and “the law as practiced.” Some constitutions also use doublespeak to declare territorial integrity inviolable while nevertheless setting procedures for territorial change. In fact, democracies employ the most unconventional constitutional weapons to fight against secession. These include a ban on secessionist political parties from participating at elections and a constitutional eternity clause that makes territorial integrity an eternal value, not subject to constitutional amendment. These tools raise democratic paradoxes so extreme that democracies appear to be using the tools of authoritarian regimes. Militant democracy must be re-characterized to capture not just bans on political parties, but also eternity clauses. It serves to protect existential needs of states, not just their democratic regime. Even when democracies allow for secession, they set such hurdles that secession becomes all but impossible to achieve. Countries’ total prohibition on secession may be explained on strategic as well as principled constitutional law considerations. This Article argues that constitutions’ treatment of secession reveals that “We the People” is a territorial concept. The Canadian landmark Reference re Secession of Quebec decision’s precedential value for world constitutionalism must thus be qualified.
INTRODUCTION
In 1945, the United Nations General Assembly began with fifty-one member states; today, it has almost quadrupled its membership to 193 states.1 The literature suggests that almost three-quarters of new states created in the twentieth century owe their births to secession.2 This number includes decolonization of states that some today may consider as outside of the secessionist phenomenon because the respective territories should never have been considered part of the territory of the colonizing state.3 However, one should bear in mind that it took two World Wars (WWI and WWII) to truly recognize and implement a people’s right to self-determination, free of colonial domination. Even then, decolonization often required blood and war.4
This Article defines secession to include situations in which both citizens and territory depart from an existing state that enjoys sovereign power (ultimate political power) over them.5 It argues that, to minimize world unrest, international law is strategically vague regarding the right to secede, outside colonial or alien occupation.6 The Article further argues that, at the same time, constitutions by and large ban secession using their most unconventional weapons, but may conceal their fights against secession through doublespeak and camouflage to accommodate both internal and external pressures.7 I argue that this strategy, as well as the Canadian landmark constitutional case of Reference re Secession of Quebec,8 succeeded in convincing scholars that constitutions by and large implicitly allow secession.9 The prevalent absolute constitutional ban on secession may be explained on strategic as well as principled constitutional law grounds.10 I argue that when there is a departure of both citizens and territory from an existing state, this combined challenge requires a new act of self-constituting on the parts of both the seceding and remaining populations.11 The Article suggests that “We the People” is a territorial concept.
There are secessionist movements in all parts of the world, encompassing both democratic and non-democratic countries—Scotland in the United Kingdom; Catalonia and Basque Country in Spain; Flanders in Belgium; Kurdistan in Iraq; Vermont, Texas, and Alaska in the United States; Kashmir in India; Corsica in France; Crimea, Donetsk, and Luhansk in Ukraine; the Islamic State in Syria; and Tibet in China.12 Wherever one places a finger on the globe, one is likely to find a secessionist movement.
Secession is typically thought of as a topic of international law, not least because secessionists need the recognition of the international community to accomplish their objectives. It is assumed that constitutional law has nothing interesting to say on the subject. But constitutional literature is increasingly dealing with secession, debating whether it is advisable for a democracy to legalize secession and even set procedures for it in the constitution. Scholars like Patrick Monahan and Michael Bryant, Robert Young, Wayne Norman, Daniel Weinstock, and Susanna Mancini argue that, if secession is permissible but regulated by a constitutional document, then there is a greater likelihood that it will occur through peaceful means, governed by rule of law principles, than through violence.13 But others, like Cass Sunstein and Donald Horowitz, warn that to regulate secession is to invite a self-fulfilling prophecy, or at least strategic exploitations of threats to secede, and that such tactics can tear apart the fabric of cooperation and compromises that hold democracies together.14 They further argue that the easier it is to exit the community, the less likely it is that the population will attempt to bring about consensual changes from within.15 Vicki Jackson argues in favor of the intermediate solution of constitutional “silence” as a way to minimize the threat of secession.16
When we move from the theoretical discussion to that of the actual practices of democratic countries, the common wisdom is that “[i]n most cases, the constitution is simply silent on the matter.”17 Since constitutional democracies are typically founded on the principle of the consent of the governed,18 their constitutional silence might be interpreted as tacit permission for secession. In fact, Monahan and Bryant with Coté advised Canada to treat its own constitutional silence as permission to secede,19 and it seems the Canadian Supreme Court followed this advice in its landmark decision regarding Quebec.20 Wayne Norman even suggests it is a “freak of history” and a “historical accident” that constitutions do not include an explicit secession clause on how exit may be conducted.21
Scholars further argue that only a minority of constitutional democracies explicitly ban secession in their constitutional documents.22 They also suggest that such a ban may be easily overcome by a constitutional amendment.23 In addition, scholars found that a few countries explicitly allow for secession in their constitutional document, and even set out the ways to achieve secession.24 That some democracies enable secession in their constitutional documents is used by scholars to show that secession does not necessarily stand in opposition to constitutionalism, and they argue that other democracies should follow these examples.25
In contrast to the prevailing approach, this Article’s interpretation of constitutional provisions suggests that democratic constitutions are not silent about secession at all. Rather, they regulate it rigorously in explicit but indirect ways. I demonstrate how prominent democracies that typically serve as role models for other countries include in their constitutional documents extremely potent tools intended to prevent secessionist movements from achieving their goals. Consociational or consensus democracies that are supposedly based on consensus and veto powers to minority groups are no different than majoritarian democracies in this respect.26 The two tools this Article highlights are banning secessionist political parties and treating the territorial integrity of the state as unamendable (in “eternity clauses”), making secession subject to an “unconstitutional constitutional amendment” doctrine. As scholars have focused on eternity clauses and have not recognized the link between the ban on political parties and eternity clauses, they have concluded that constitutions by and large are silent on secession.27
An examination of 192 constitutions of both democratic and non-democratic states that are members of the U.N. reveals that only 28 countries are silent and do not explicitly protect territorial integrity and national unity.28 Among these countries are Canada, Ireland, Japan, the Netherlands, New Zealand, and the United States.29 Three additional countries provide procedures for territorial change with no conflicting language that protects territorial integrity and thus do not require constitutional amendment to achieve secession.30 There are seven countries that supposedly allow for secession by embodying an explicit secession clause, and Bosnia and Herzegovina, which left the constitutionality of secession to the discretion of its Constitutional Court.31 Additionally, the United Kingdom has enabled a referendum on the secession of Scotland and provided for a process for the secession of Northern Ireland.32 Thus, roughly 79% of world constitutions explicitly ban secession. The number may be higher were we to include implicit bans on secession. Thus, for example, constitutional silence did not prevent U.S. courts from developing an implicit prohibition against (unilateral) secession.33 These findings contrast with Zachary Elkins’s 2016 study, which suggested that “a full thirty-eight countries have included the prohibition [against secession] across their constitutions dating as far back as the 1850s.”34 Moreover, focusing on democracies and semi-democracies, as defined by the Freedom House’s Country Scores and the Economist Intelligence Unit’s Democracy Index,35 reveals that 74% of democracies and semi-democracies explicitly ban secession.36
I argue that constitutional democracies have successfully created the impression among the democratic world that these unconventional weapons of the constitutional system—ban on political participation and unamendability—are used solely to protect democratic values. But, in fact, democracies also use these tools to prevent secession and for existential—and not so existential—needs, rather than democratic reasons alone.
The Article further argues that secession reveals how both tools—the ban on political participation and eternity clauses—complement each other and should be understood as mirror images of one another. Because there are important theoretical, historical, and methodological links between the two mechanisms, a country that has one of the mechanisms may legitimately infer the implicit existence of the other. This is not trivial, as countries often have only one of the mechanisms explicitly incorporated into their respective constitutional systems. Once the courts understand the duality of the two mechanisms, they may develop a more coherent jurisprudence for applying both mechanisms.37
Democracies often conceal their fights against secession through doublespeak and insincerity. They may both ban secession and set the procedures for territorial change. They may ban secessionist parties but offer other convincing explanations for the ban that have nothing to do with secession. They use camouflage because of the unique challenges secessionists pose to democracies—notably, democratic commitments to the rights to self-rule and self-preservation. But democracies might believe that each of these values point to different resolutions in the secessionist context. Democracies might believe that the right to self-rule supports secessionism while the mother country’s own right to self-preservation negates it.38 Furthermore, democracies conceal their fights against secession to fare better in both the court of international public opinion and existing international tribunals. Thus, for example, if the European Court of Human Rights prohibits banning peaceful secessionist political parties, as it in fact does, then democracies may ban those parties for non-secessionist reasons.39
When we turn our attention from countries that ban secession to countries that purportedly allow for secession in their constitutions, it turns out that these latter countries, too, try to prevent secession. They say that they allow for secession but set such hurdles that secession becomes all but impossible to achieve in democratic ways. Scholars, who use these countries to illustrate why secession should be allowed, do not acknowledge how the enabling tools are undermined by the country providing them.
Yet, despite the prevalence of constitutional prohibitions against secession, if both the mother and the seceding states reach agreement, we expect the international community to accept their agreement. Why not then have only a constitutional prohibition against unilateral secession? This Article argues that the prevalence of constitutional prohibitions stems from principled as well as strategic considerations. The Canadian Supreme Court’s canonical decision, which interprets constitutional silence as tacit permission for Quebec to secede via a constitutional amendment, may thus have only limited value for the overwhelming majority of world constitutions that are not silent about secession.40
This Article proceeds as follows: Part I defines secession and discusses its potency in determining world history. It further explains why there is a rise in secessionist activity in the twentieth and twenty-first centuries. This Part also explains why international law has strategically left ambiguous the question of the right to secede outside colonial or alien rule and its effect on constitutional law. Parts II and III discuss the indirect powerful constitutional tools—the ban and eternity clauses—countries use to prevent secession. They also argue that secession reveals the ways in which both tools complement each other. They further explain that democracies conceal their struggles against secession because secession poses a serious constitutional paradox to democracies. Part IV argues that many countries engage in doublespeak, declaring territorial integrity inviolable yet setting procedures for territorial change. There are also a few countries that permit secession but only through means intended paradoxically to frustrate secession. Part V concludes by suggesting that the prevalence of the total prohibition on secession may stem from strategic as well as principled constitutional law considerations. It argues that constitutions typically require secessionists to resort to extra-constitutional means because secessionists challenge the very identity of the constitution-making body. As such, they may not rely on the constitutional amendment process to achieve their goals. Constitutions’ treatment of secession reveals that popular sovereignty is a territorial concept. It then explains why the Canadian Reference re Secession of Quebec case has limited bearing as a precedent that other countries may follow.
I. THE CHALLENGE OF SECESSION
This Part defines secession and argues for its importance in understanding the development of world history and politics. It further explains the factors contributing to the rise in secessionist activity in the twentieth and twenty-first centuries. It then explains the strategic legal considerations in regulating secession by both international and constitutional law and why states seek to resist secession.
A. Mapping Types of Secession
Internal autonomous arrangements within existing states that fall short of the creation of a new state are not considered secession. Secessions can be categorized into several types. The classic form of secession occurs when a group of people creates a new state in part of the territory of an existing state. As such, the classic secession involves the withdrawal of both citizens and land from an existing state.41 The fact that secession involves both elements is important since a democratic society may recognize a particular people’s right to self-determination but deny their claim to a specific territory. In such cases, the society would oppose that people’s right to secede. A recent example of an attempt to achieve classic secession was the Scottish referendum on independence in September 2014. If the Scottish people had answered “yes” to independence, it may have led to secession of Scotland from the United Kingdom.42 Some international lawyers define secession as a unilateral act alone (distinguished from “negotiated independence”),43 but for constitutional law purposes, it is important to include cases of consent between the rump state and the seceding state.44
A second type is irredentist secession, in which a portion of the population in an existing state wants to secede with part of the territory in order to join another existing state, typically a neighboring state. Irredentist secession usually occurs when the majority of the people in the seceding area belongs to the same ethno-national community as the majority in the neighboring state or when the seceding area once belonged to the neighboring state and the secessionist movement wants to restore the previous territorial distribution.45 Russia tries to portray its 2014 annexation of Crimea as a case of irredentist secession, under which the majority of Russian people living in Crimea wanted to join Russia. But, to most Western eyes, this seems like a case of forceful annexation of Crimea in contravention of international law.46 Some scholars do not treat the irredentist case as part of the secessionist phenomenon,47 but cases may be blurred where some of the secessionists desire an independent state and others annexation to a neighboring state.48 Thus, classic secession may also involve irredentist tendencies.
A third form of secession occurs when an existing state dissolves and new states are formed in its place. The dissolution may reflect an internal agreement among the entities in limine, or it may be imposed by superpowers. For example, when Czechoslovakia divided into the Czech and Slovak republics in 1993, the dissolution resulted from an internal agreement between the leaders and in opposition to the will of both populations.49 During WWII, Nazi Germany and the Soviet Union partitioned Poland as a result of an external agreement between the two.50
In the first two forms of secessions, the classic and the irredentist, the original state remains in a reduced form, while in the third type of secession, the original state completely dissolves. But what all three forms of secession share in common is the combined challenge to the rump state over control of both territory and citizens. This is where secession differs from mere amendments to boundaries of states that do not involve the transfer of citizens to foreign control. I argue that when there is a combined challenge of withdrawal of both citizens and territory, it requires a new self-constituting act on the part of the departing as well as remaining populations.51 In contrast, for international law purposes, these cases may be distinguished as existing states, in contrast to new states, which would typically not need to renegotiate their international agreements with the world at large nor need to seek the world’s recognition for their existence.
B. Shaping World Events
Secession continuously shapes world history, whether by dispersing people, as in the Old Testament Tower of Babel,52 or by fragmentation, as shown in the fall of the Roman Empire—the largest Empire of classical antiquity—in the fourth and fifth centuries, which left the Western and Eastern Roman Empires;53 whether by defiance, as in the Dutch Republic’s secession from the Spanish Empire in the sixteenth century,54 or by revolution, as in the secession of the American colonies from Great Britain.55
In fact, to a great extent, secessionist struggles were the opening shot of both WWI and WWII. The immediate trigger of WWI was the assassination of Archduke Franz Ferdinand, the presumptive heir to the Austro-Hungarian throne, by a Serbian secessionist student in Sarajevo, the capital of Bosnia. The Austro-Hungarian Empire controlled Sarajevo while secessionist forces wanted to secede and join the neighboring state of Serbia.56
The same is true with regard to WWII. Among the first major steps taken by Nazi Germany to create a Greater German Reich was to annex Austria (Anschluss). Within a month of Nazis’ annexation of Austria, the Nazis held a plebiscite that was manipulated to reflect approval by over 99% of the vote.57 Hitler next targeted the Sudetenland area of Czechoslovakia, arguing that the majority of its inhabitants were ethnic Germans who belonged with Germany.58 While the Munich Pact—signed by the leaders of Britain, France, Italy, and Germany—enabled Hitler to annex Sudetenland in exchange for peace, he did not stop there, as is well-known.
Nor is the immense power of secession a matter for history alone. After Russian armed intervention in Crimea at the end of February 2014, which some regard as the end of the post-Cold War era,59 the new Crimean authorities organized a referendum in Crimea.60 The new authorities included the self-appointed Prime Minister, Mr. Sergei Aksyonov—representing the Russian Unity Party, which received 4% of the popular vote in the 2010 Crimean parliamentary elections—and some Crimean legislators who had been allowed, by “unidentified” gunmen, to enter the building of parliament.61 According to the new Crimean authorities, “83.1 percent of the eligible population voted . . . and . . . the final result was 96.77 percent in favor of joining Russia and 2.51 percent against.”62 Just like Germany, Russia argued that annexation not only reflected the inhabitants’ will but was also justified because the majority of the Crimean population is Russian.63
While the West, including former U.S. Secretary of State Hillary Clinton, drew comparisons to Sudetenland, Putin relied on the precedent set in Kosovo.64 Putin portrayed Russia as the protector of the Russian minority in Ukraine, which happens to be the majority in Crimea.65 By obfuscating the details of its intervention in Crimea, Russia avoided international scrutiny for what many suspect was an illegal act under international law.66 That this was an illegal act may be supported by the fact that Russia militarily intervened in Crimea. That this may have been the willing act of the majority of the Russian population in Crimea may be supported by the peculiar history of Crimea’s status in Ukraine.
In 1954, the Soviet Union transferred the Crimean peninsula to Ukraine as a “gift” or a “donation.” The transfer had mere administrative implications because both Russia and Ukraine were then part of the USSR.67 However, Russia argues that it never intended to sever ties with Crimea, as happened when international borders were established in 1991 following Ukraine’s independence.68 Thus, from the Russian perspective, Crimea is a case of irredentist secession, and Russia was authorized to reverse the change that occurred over half a century ago.
C. Explaining the Rise in Secessionist Activity
While secession is part of human history, a few cumulative factors may explain the world rise in secessionist activity in the twentieth and twenty-first centuries. First, in the past, borders were drawn according to states’ ability to defend themselves from external attack or as a result of allocations of territories between colonial powers.69 The colonial powers distributed land among themselves without taking full account of the identity of the people inhabiting the particular territories. They thus tore nations apart. When the U.N. recognized the right to decolonization after WWII, it was granted according to pre-defined territories drawn by the colonies. Thus, it did not remedy colonial injustice. This international doctrine of uti possidetis, which means “leav[ing] the place as one received it,” underlies the mismatch between today’s boundaries and ethnic divides.70
With the rise of nationalism in the twentieth century, secessionists have sought to redraw the boundaries according to identity classifications, such as religion, ethnicity, language, culture, shared history, and the like. In fact, most modern secessionist attempts are conducted by ethnic minorities within an existing state.71 Moreover, secession is most threatening to multinational countries, where cultural-ethnic minorities reside in the same geographical area and are not disbursed throughout the country.72
Second, secessionists believe that their desire for self-determination is more feasible today than in the past. While people once felt the need to unite under one state to enjoy efficiencies of scale in terms of both national security and economic activity, today secessionists believe that even small states may survive and prosper.73 Globalization has made international trade available to economies of every size, and these economic possibilities have prompted many secessionist movements in European countries—primarily the Catalans in Spain and the Scots in the U.K.—to talk of “independence in Europe.” Some even argue that the E.U.’s success demonstrates that a monetary union may coexist with independent states. The existence of the E.U. may encourage secessionists to splinter states and attempt to join the E.U., as the Czech Republic and Slovakia did, and Scotland attempted to do.74 At the same time, the fading memory of WWII, the collapse of the Soviet Union, and the end of the Cold War have led to the belief that the world is a safer place for new independent countries than in previous eras. It is further assumed that the spread of democracy in recent decades translates into fewer wars between states, which in turn feeds this feeling of relative international security even for small, vulnerable entities.75
Third, secessionists feel that they can exploit the language of human rights to promote their agenda. After WWI, President Wilson promoted the idea of a nation’s right to self-determination to promote world peace. U.S. presidents emphasized this self-determination principle in international treaties following WWII in the context of decolonization.76 The United Nations Charter explicitly includes decolonization as part of the U.N. agenda.77 In 1960, the U.N. General Assembly adopted Resolution 1514 titled the “Declaration on the Granting of Independence to Colonial Countries and Peoples.”78 The “external” right to self-determination in the form of statehood exists in cases of colonial or alien occupation.79 At the same time, international bodies promoted the imperatives of social and cultural rights, devolution, and autonomy arrangements within existing states to protect minority rights. This is a manifestation of the “internal” right to self-determination within democratic states.80 Secessionists have used the international community’s prioritization of both self-determination and group/minority rights to push these agendas even further. Secession is thus demanded even outside of decolonization processes and even when the parent state makes every effort to accommodate group rights, as was evident in Quebec.81
Fourth, immigration policies may affect secessionist movements as well. This is especially true in the E.U., with its open-border policy among member states.82 Western countries in general have seen an increase of migrants since 1945. As mobilization has become easier and globalization has expanded, so too has migration in all regions of the world. In 1960, 76 million people migrated; by 2005, there were 191 million migrants.83 It is typically assumed that immigrants do not lead secessionist movements, since immigrants have no established connections to their territorial destinations.84 But, this observation does not capture the entire story. An influx of immigrants may threaten the dominance of indigenous inhabitants over a particular region and thus strengthen the desire of indigenous or local secessionists to secede.85 The desire to maintain the “British” way of life and control immigration may be one of the main factors explaining the British people’s decision to leave the E.U. in the 2016 Brexit referendum.86
The assertion that immigrants do not lead secessionist movements should also be refined. When Crimea became part of Russia in the eighteenth century, 99% of the population was Crimean Tatars. Through Russian migration, their percentage declined to 34% at the beginning of the twentieth century. In 1944, Stalin deported masses of Tatars, under the pretext that they collaborated with German troops. By the time the Soviets handed over Crimea to Ukraine in 1954, 99% of the population was Russian.87 In fact, Russia intentionally moved its population to various parts of the USSR to have better control over the territory.88 According to the 2001 Ukraine census, Russians comprised roughly 58% of the population in Crimea.89 To the extent there was an identifiable support at all,90 it was the Russian population that supported the Crimean secession in 2014, while the remaining Crimean Tatar population opposed and even banned the referendum.91 Similarly, in 1975 the Turkish Cypriot administration proclaimed the existence of the Turkish Federated State of Cyprus. Most of the Turkish Cypriots resettled in the North, and Turkey encouraged additional Turkish settlers to move there to affect a change in the composition of the population.92 In 1983, this northern part declared independence as the Turkish Republic of Northern Cyprus.93 That shifting populations may lead to a change in the nature of the populace controlling the territory is one of the reasons international law treats population transfers by conquering powers as a war crime.94
D. Opposing Secession
If there is such a strong demand for secession, why not allow it? Secessionists pose a threat simultaneously to the interests of two different groups—the majority population of the rump state who may object to the secession of one of its parts, and the minority population who will be caught in the region after it gains independence. The parent state may have legitimate objections to secession, such as its need to control the region to protect economic resources, to preserve legal and social order, to protect human rights, and to defend itself. Secession may sometimes threaten the very viability of the remaining state.95 The minority may object to secession for fear of the persecution that may follow or because it prefers to remain part of the parent state. Since secession poses a threat to crucial interests of the parent state and its citizens, it is typically accompanied by violence and even civil war.96
Moreover, while secessionists often promote their cause in the name of their right to ethnic-national–self-determination, the resulting new state’s boundaries are usually not drawn according to the traits of the populace. Instead, they are usually drawn according to earlier sub-national administrative borders.97 Thus, ironically, ethnic groups may become new minorities in the newly created state.98 The legitimacy for secession is thus weakened. Secessionism may, in fact, become an ongoing process where countries liberated through secession become subject to secessionist demands as well.99 Thus, for example, Bangladesh seceded in 1971 from Pakistan—which itself had won independence in 1947 after being ruled by Britain.100 East Timor seceded from Indonesia in 2002, which itself had been a Dutch colony until WWII.101 This potentially endless process of secession may threaten the idea of self-determination and statehood.102
To prevent this endless cycle of secessionism, secessionist movements seem to have their own interpretation of the international norm of protection of the “territorial integrity” of the state, which is codified in the U.N. Charter.103 They do not see the international norm of territorial integrity as a barrier to their claim to independence since they are not member states yet and are thus not obligated under the Charter. The International Court of Justice (ICJ) in its 2010 Kosovo case legitimated this one-sided interpretation of the U.N. Charter. It stated: “[T]he scope of the principle of territorial integrity is confined to the sphere of relations between States.”104 Non-state parties that seek independence should only avoid “unlawful use of force or other egregious violations of norms of general international law, in particular those of a preemptory character (jus cogens).”105 In addition, they should not violate specific international law provisions that apply to them as a result of decisions of the U.N. bodies.106
At the same time, the secessionists rely on the existence of the norm of territorial integrity to protect their own state once they have seceded. In fact, the very existence of the norm makes their aspirations for independence worthwhile since it increases the likelihood that they will be able to maintain their independence from external and internal attacks.107 This exposes the double standard that secessionists typically employ. For example, the Quebec government has demanded Canada’s recognition that the region enjoys the right to secede but opposes the secession of its minorities, who have expressed their will to remain part of Canada if Quebec achieves independence.108 For these reasons, secession disrupts the world order and may threaten global stability, and secessionists’ claims to legitimacy may be contentious.
E. Regulating Secession
How does the law deal with secession? Secession is traditionally thought to be regulated solely under international law. But, in fact, it lies at the intersection between the constitutional law of a given country and international law. Secession remains part of the “internal affairs” of the state so long as the state does not abrogate its duties under international law to properly treat the minority wishing to secede. But, if the state mishandles the affair, the situation might deteriorate into a conflict that places it in the province of international law. A great conflict of wills between the state and the secessionists regarding the application of international law is implied in secession. While those wishing to secede may seek external intervention and the world’s recognition that a change in the boundaries of an existing state is warranted, the parent state may ferociously defend its territorial integrity.109
Given that national boundaries are at stake, it may be surprising that international law is vague on when secession is justified and how the international community should view “newly” self-declared states. International law clearly legitimizes secession only in the case of decolonization or alien domination.110 It is not even clear whether international law recognizes a remedial right to secede when the parent state persecutes minorities or seriously abuses their human rights.111 In 2009, Serbia led the General Assembly to ask for an advisory opinion of the ICJ regarding the question: “Is the unilateral declaration of independence by the Provisional Institutions of Self-Government of Kosovo in accordance with international law?”112 The ICJ distinguished the issue before it from that decided by the Canadian Supreme Court in the Quebec case. The Canadian case dealt with the question of whether there is a right under international law to unilaterally secede.113 The ICJ defined the issue before it as whether Kosovo’s unilateral declaration of independence violated international law.114
Nor did it discuss the legal consequences of the unilateral declaration of independence.115 The ICJ did state that “[d]uring the second half of the twentieth century, the international law of self-determination developed in such a way as to create a right to independence for the peoples of non-self-governing territories and peoples subject to alien subjugation, domination and exploitation.”116 This language is identical to article 1 of the Declaration on the Granting of Independence to Colonial Countries and Peoples. Article 1 states, “The subjection of peoples to alien subjugation, domination and exploitation constitutes a denial of fundamental human rights, is contrary to the Charter of the United Nations and is an impediment to the promotion of world peace and co-operation.”117 This language is thus open to conflicting interpretations. One possible interpretation is that when people do not have a right to vote for the governing institutions, they are subject to alien domination and are entitled to an external right to self-determination.
Another possible reading of this language is in light of its historical context. The Declaration on the Granting of Independence to Colonial Countries and Peoples also guarantees territorial integrity,118 as does the U.N. Charter.119 Thus, the external right to self-determination that it promises is limited to “colonial countries and people” recognized as such when the declaration took place. The U.N. has an official list of seventeen Non-Governing Territories, that all but one date back to 1946, the year after the U.N. Charter was adopted.120 In addition, the external right to self-determination, under this interpretation, extends to people who became subject to belligerent occupation after the adoption of the U.N. Charter and as a result of a violation of the norm of territorial integrity. The ICJ explicitly stated that it is not deciding whether, outside decolonization or alien domination, there is a right under international law to secede.121
Rather than recognizing a remedial right to secede, international law is geared toward establishing internal protections for minorities as a way to avoid secession. International law defines what countries should do, rather than the consequences of failing to do what they should.122 International law thus forces democracies to face an untenable dilemma. If they fail to accommodate group rights of minorities, they may face charges that they are violating human rights—which in turn may lead to secessionist claims for remedial secession to end the injustices inflicted by the parent state. On the other hand, if democracies foster minorities’ separate identities by promoting group rights, including an official recognition of the minorities’ language, culture, religion, etc., they may actively support the separation of these minorities from the existing state. Democracies may find themselves providing the very resources and fostering the separate identities that allow minorities to embark on secession. For example, ironically, in Eastern Europe, secessionist movements arose when the protection of rights improved.123 This is a lose-lose situation for the states, and the solution requires a delicate, almost unachievable, balance: states must enable group identity to be formed without enabling minorities to overreach their status or undermine the unity of the state.124 This may partly explain this Article’s findings that even consociationalist and consensus democracies ban secession.125
It is also unclear under international law which groups should be “entitled” to secede in non-decolonization contexts.126 Should the right to secede be limited to groups sharing ascriptive characteristics (that is, characteristics of “being” rather than those of “achieving,” such as ethnicity),127 or should the right be widened to include majorities that share only the will to secede?128 International law fails to define what “people” are entitled to self-determination.129
But can international law afford to be vague on non-decolonization secession? In contrast to decolonization, a phenomenon that peaked in the late 1950s and early 1960s, contiguous secession involves territory that is no more than 100 miles away from the mother state. Contiguous secession peaked in the early 1990s with the breaking apart of the Soviet Union and Yugoslavia.130 Decolonization accounts for only about 35% of the total number of secessionist movements since 1931, the rest being contiguous.131 When considering the breadth of the contiguous secessionist phenomenon, it seems that there is a demand for clear international law on the subject. On the other hand, the fact that international law is vague about non-decolonization secession may indirectly hinder successful movements and minimize world unrest. We know that the success rate of secessionist movements is highly correlated to the nature of secessionism: while decolonization enjoys a 77% success rate, only 16% of contiguous secessionist movements succeed.132
Thus, strategic considerations are as important as substantive normative considerations, both of which invite discussion among constitutional law scholars. While many share the assumption that secession should be prevented, these scholars dispute the best approach to prevention. Sunstein believes that constitutional law should outlaw secession;133 others, like Norman and Weinstock, believe that setting an arduous procedure for secession in the constitution may be a more effective way to prevent secession.134 A constitutional secession clause may set substantive requirements as a precondition for achieving secession, or mere procedural hurdles, or it may be a hybrid of both.135 The intermediate position of Jackson applauds silence as a means of maintaining strategic vagueness on the subject of secession.136
II. BAN ON POLITICAL PARTICIPATION
The prevailing scholarly view is that most constitutional democracies simply ignore secession, and scholars argue whether this is the desirable approach to the subject.137 Contrary to this widespread view, I argue that the overwhelming majority of world constitutions, including many prominent constitutional democracies, set robust indirect mechanisms to prevent secession. The two most powerful tools used are: (1) the prohibition against secessionist political parties from participating at elections; and (2) the grant of eternity status to the unitary or federal nature of the state, including its territorial integrity. Furthermore, it is important to study both mechanisms in tandem to appreciate the extent of democracies’ fight against secession.
A. The Misperception Regarding Militant Democracy and the Prevalence of a Ban on Secessionist Political Parties
For democracies, one of the most important sources of legitimacy is majority rule, by which every adult citizen has one vote, and by and large, the majority determines election results. If citizens wish to change the governing law, they are expected to use political means, rather than resort to violence. Citizens may petition their representatives, hold public demonstrations, or even run for office to change the governing law from within.
At the same time, it is widely accepted among democracies that they must take a more militant stand to protect democracy from internal threats. This lesson of WWII138 is being tested in Europe today, as extremist representatives compete in elections at all levels of government.139 Democracies learned that, when every citizen and party may run for office, democratic processes might ultimately enable those who seek to destroy a government’s democratic character to win office. As Joseph Goebbels, the Reich Minister of propaganda in Nazi Germany, famously said, “[t]his will always remain one of the best jokes of democracy, that it gave its deadly enemies the means by which it was destroyed.”140 As a result, it is customary among constitutional democracies to include a constitutional ban on political parties that threaten democracy itself.141 In fact, many European states have banned at least one political party at some point since WWII.142
The general perception is that militant democracy is used only to protect democratic values. In fact, the oxymoron of militant democracy—banning political participation in the name of democracy—has been accepted because militant democracy is thought to be used sparingly to defend only the democratic nature of the system. However, an examination of 192 constitutions of both democratic and non-democratic countries reveals that one-hundred and three countries (54%) have a ban on political participation of secessionist political parties.143 When looking at democracies and semi-democracies, sixty-two countries (47% of democracies and semi-democracies) have such a ban in their constitution.
These one-hundred and three countries may be divided as follows. Fifty-six countries have an explicit ban on political parties that threaten the territorial integrity of the state or national unity and sovereignty. Among these countries are Brazil, Bulgaria, France, Germany, India, Portugal, Romania, Turkey, and Ukraine.144 Thirty-four countries prohibit political parties that are regionally, religiously, and/or racially-based or require them to be nationalist in scope and character. Among these countries are the Czech Republic, Macedonia (the former Yugoslav Republic), Poland, and Sweden.145 For example, Ghana’s Constitution provides that “[e]very political party shall have a national character.”146 For registration purposes, the party must satisfy the following requirements:
a. there is ordinarily resident, or registered as a voter, in each district of Ghana, at least one founding member of the party; b. the party has branches in all the regions of Ghana and is, in addition, organised in not less than two-thirds of the districts in each region; and c. the party’s name, emblem, colour, motto or any other symbol has no ethnic, regional, religious or other sectional connotation or gives the appearance that its activities are confined only to a part of Ghana.147
Members of the national executive committee must also be “chosen from all the regions of Ghana.”148 Thirteen countries protect territorial integrity in their constitution, and require political parties to respect the constitution, compared to countries that allow the promotion of non-violent agendas to amend the constitution. Or they may ban political parties that pursue aims that are forbidden under criminal law. Thus, these parties may not seek to change criminal laws that protect the territorial integrity of the state.149 Among these countries are Argentina, Italy, and Spain.150
B. The Practice of Banning Secessionist Political Parties
Even when scholars acknowledge that a few countries include an explicit constitutional ban on secessionist political parties, they argue that these textual provisions are treated as dead letters. Scholars point to the existence of numerous secessionist political parties around the world as proof that democracies allow for secession. They theorize that it is unjustified for a democracy to ban political parties to prevent secession, if secessionists pursue their goals in democratic and peaceful ways. Taking this to its logical conclusion, scholars suggest that, since democracy allows for secessionist political parties, there is no reason not to establish a constitutional procedure for achieving secession.151
But the reality is different. In recent years, the banning of secessionist political parties has occurred in major European and Asian countries. It should also be noted that regional political parties are often the ones leading the secessionist movements on the national level,152 and they are the ones that are typically banned.153
Democracies selectively target and ban political parties because of their secessionist agendas, under the pretext of preventing the promotion of undemocratic values such as racism, violence, and terrorism. Even when the constitutional text of a given country allows for the banning of a political party based on secession, authorities often prefer to justify their actions in the name of a broader cause of protecting democracy. In this way, democracies obfuscate, and even use subterfuge, in their fight against secession. This is also where the banning of secessionist parties differs from the more general phenomenon of banning political parties to protect democracy. When your only motive is to protect democracy, there is no need to conceal your motivations.
I open with recent examples of democratic countries that have banned secessionist parties without having an explicit, direct constitutional provision that allows for a ban to protect the state’s territorial integrity. In these cases, it is easier to understand why a democracy would want to conceal its aim to ban secessionist parties per se. These examples deal with Spain and Belgium, who are both struggling with activist secessionist movements. The Spanish Constitution provides that political parties’ “creation and the exercise of their activities are free in so far as they respect the Constitution and the law.”154 It further states, in a different section, that it “is based on the indissoluble unity of the Spanish Nation, the common and indivisible homeland of all Spaniards.”155 Thus, the Spanish Constitution indirectly prevents political parties from pursuing the disintegration of the state’s unity. Based on this constitutional provision, and an amended Law on Political Parties (enacted in 2002), the Spanish Supreme Court banned the Batasuna (which means “unity” in the Basque language) from participating in Spanish or European parliamentary elections in 2003,156 after the Batasuna had continuously participated for twenty years preceding the ban.157 The Batasuna party was targeted because it shared the objectives of the Basque separatist ETA insurgents, whose militant activity had been escalating.158 Since it was difficult to find real evidence against the Batasuna, the party’s dissolution turned on its refusal to condemn acts of violence by the ETA.159 In fact, the amended Spanish law allows for the banning of political parties inter alia for “excusing assaults against life . . . legitimizing violence . . . giving express or tacit political support, legitimizing terrorist actions or excusing and minimizing their significance.”160 Later, the Spanish authorities banned successor parties to the Batasuna.161 In 2017, the Spanish central authorities quashed Catalonia’s secessionist forces, after the Catalan Parliament unilaterally declared independence. They did this by dismantling Catalonia’s legislature and executive branches and calling for renewed elections; all while jailing central figures of the Catalan independence parties on charges of sedition and rebellion and misuse of public funds.162
The Belgium Constitution does not explicitly provide for the banning of political parties.163 Yet, in 2004, the Belgium courts prohibited funding and media access for the Vlaams Blok (Flemish Block)—a secessionist party supporting the independence of Flanders. The courts held that Vlaams Blok’s agendas were racist because the party’s associated non-profit organizations promoted anti-immigration agendas found discriminatory.164 The prohibition was based on a 1981 anti-racism statute that makes incitement to racism and segregation illegal.165 The Vlaams Blok was the second largest faction in the Flemish Parliament at that time, winning around 24% of the vote in the year it was restrained.166 Even before the judicial decision, there was agreement among the other factions never to include the Vlaams Blok in a coalition government because of its extreme right wing agenda.167 The Court based its decision inter alia on the fact that the party did not disapprove of the discriminatory speech of its local branches.168 The Court did not discuss more recent evidence that the party tried to distance itself from extremists.169
Even more interesting are the cases in which the nation’s constitution explicitly and openly allows a ban of secessionist political parties, but officials nonetheless use the pretext of protecting democracy. For example, the Bulgarian Constitution of 1991 grants freedom of association, but organizations’ activities “shall not be contrary to the country’s sovereignty and national integrity, or the unity of the nation.”170 In 2000, the Bulgarian Constitutional Court banned the United Macedonian Organization Linden-Pirin, which advocated that Bulgaria’s Pirin region should belong to Macedonia, because the party’s separatist agenda threatened the state’s security, even though it was a marginal, non-influential organization.171
The Indian Constitution grants the constitutional right to form associations subject to reasonable restrictions, which may be enacted by law “in the interests of the sovereignty and integrity of India.”172 In India, a politician who campaigns “on the ground of his religion, race, caste, community, or language or the use of, or appeal to, religious symbols”173 may not run for office and, if elected, may be removed. This prohibition is actively enforced.174 As a result of the ban, any secessionist movement in India that is based on community identity—and this is the flag secessionist movements raise—may be banned based on racism rather than separatism. Thus, for example, the Indian government imposed bans on the activity of Sikh separatist movements in the 1980s.175 In Kashmir, the central government has intervened in the conduct of elections, dissolved assemblies, arrested elected politicians, and imposed president’s rule for prolonged periods to combat separatism.176
The Turkish Constitution includes an explicit textual clause in its preamble stating that “no protection shall be accorded to an activity contrary to Turkish national interests, Turkish existence and the principle of its indivisibility with its State and territory.”177 The Constitution further explicitly authorizes the banning of parties that threaten the territorial integrity of the State.178 Implementing these constitutional provisions, the Turkish Constitutional Court has repeatedly upheld bans on political parties advocating Kurdish separatism or independence on the grounds that they threaten state unity and public order, violate the territorial integrity of the state, and refuse to denounce the PKK—the Kurdistan Workers’ Party—in spite of its terrorist attacks.179 In fact, “half of the 18 parties banned since 1980 [and until 2007] have been Kurdish.”180 In these cases, the Court did mention separatism, but only as one cause of the ban among many.
Banning political parties to oppose secessionism has been used at least twice in Ukraine. The Ukrainian Constitution prohibits the establishment of political parties “if their programme goals or actions are aimed at the . . . violation of the sovereignty and territorial indivisibility of the State.”181 Within days of Ukrainian independence from the Soviet Union in 1991, a legislative committee banned the Communist Party of Ukraine and confiscated its assets, thus substantially diminishing its ability to return Ukraine to Russia’s rule.182 After the recent Russian annexation of Crimea, Ukrainian authorities banned the extant Communist Party and brought charges of treason against its leaders on the grounds that they supported Russian annexation and were supporting the terrorist tactics of Eastern Ukrainian separatists.183
These examples of bans of secessionist political parties are even more significant in light of the fact that bans on political parties are a last resort tool that should rarely be employed. Moreover, the effect of the constitutional prohibitions on secessionist political parties goes beyond the cases that involved actual exercise of the ban. Rather, the very existence of the prohibition affects politics and public discourse in ways that are difficult to measure. Parties may be deterred from openly pursuing secessionist agendas. The authorities may target individuals and groups by the use of criminal law that protects the territorial integrity of the state. The prohibition is intended more to chill formation of such organizations than to deal with them after they exist. And even when the state does not apply the ban and turns a blind eye to secessionist organizations, these organizations know that they exist at the mercy of authorities rather than by right.
C. The Difficulty to Identify the Ban on Secessionists
These are examples of a handful of prominent democratic countries that banned secessionist political parties in recent years. Why is it so difficult to identify this phenomenon? Why do scholars often believe no ban on secessionist political parties takes place in democracies? Democracies seem to have succeeded in creating an acoustic separation, where the secessionists understand that they are persecuted for their secessionist activities while the wider audience believes that they are banned because of their racist or militant agendas.184
While scholars point to the fact that there are dozens of secessionist political parties across the democratic world,185 in fact, there is an ongoing game between democratic states and secessionist forces. In many constitutional systems, the ban leads to a game of “whack-a-mole,” in which the ban is tailored to the circumstances, allowing a rebirth of the same secessionist party under a different name.186 Thus, for example, in Turkey, the Peace and Democracy Party (BDP) succeeded the Democratic Society Party (DTP) after the DTP was closed for its alleged connections with the PKK.187 Similarly, the Belgium far-right Vlaams Blok party “rechristened itself as the Vlaams Belang [Flemish interest]” one week after the judicial decision to restrain it.188 In fact, this phenomenon is so widespread that Angela Bourne and Fernando Bértoa called it a “lapsed ban” phenomenon, which means “a ban that has been undermined by the failure of the state to prevent a successor from taking on the mantle of a banned party.”189 At times, the ban extends only to campaigning, while other political activity is allowed. Thus, for example, in India there are multiple political organizations that promote secessionist causes with varying degrees of openness about their aims.190 Bans may be applied for limited periods, after which secessionist activity re-emerges. In this chaste process, democracies try to sap secessionists’ energy and resources by confiscating assets, with an endgame of coercing these secessionist forces to soften and rephrase their agendas.191
At the same time, democracies do not target every secessionist political party. Democracies try to target secessionist political parties that are neither too small to be bothered with nor too big to be dismantled. Thus, in some constitutional democracies, the ban on political participation is dormant as far as secessionists are concerned. For example, Germany’s Basic Law provides that a court may ban a political party if its aims or behavior seek “to endanger the existence of the Federal Republic of Germany.”192 Nonetheless, the Christian Social Union (CSU), which seeks to strengthen Bavaria’s autonomy within Germany and even establish it as an independent state, has been allowed to participate at regional elections since the 1940s, without so far posing a serious threat to Germany’s federation.193 Systems ban only secessionist political parties that they feel threatened by and allow others to flourish. Thus, while Batasuna was banned, there are other secessionist political parties still operating in Spain. The same is true for Belgium.194 Allowing non-threatening secessionist parties to exist provides a steam valve for their frustration within the confines of regular politics.195
In addition, it is a rather easy task for democracies to justify their ban against secessionist political parties based on democratic values alone. One may argue that the secessionist phenomenon is racist by definition in the sense that the secessionists often seek separate treatment from others based on ethnicity. Usually, secessionist political parties will also maintain some kind of relationship with their militant counterparts, which helps to cast the parties as terrorists. Overall, democracies are reluctant to admit that they are against secessionist parties per se.
D. The Roots of Militant Democracy Theory
In fact, banning secessionist political parties is not a phenomenon of the twenty-first century. It dates back to the beginning of the twentieth century. In May 1918, almost all leaders of Sinn Féin were arrested by British authorities for allegedly conspiring with the Germans, which was never proven.196 Northern Ireland banned Sinn Féin between 1956 and 1974, covering the period of the Irish Republican Army border campaign.197 Later, the U.K. restricted the broadcasts of Sinn Féin and its members, thus restricting free speech to protect territorial integrity.198 Similarly, in France, the Constitution states that “France shall be an indivisible . . . Republic.”199 Political parties must “respect the principles of national sovereignty.”200 A statute from 1936—which survived the establishment of the Fifth Republic—authorizes the President of the Republic to dissolve groups inter alia if their goal is the dismemberment of the territorial state.201 Over the years, France banned more than 100 organizations, including some separatists.202 In fact, the Conseil d’Etat has on a few occasions affirmed the dissolution of secessionist political parties.203 When France occupied Saarland after WWII, the French High Commissioner banned the Democratic Party of the Saarland (DPS) because of its aspirations to reunite with West Germany.204 Germany banned the Communist Party in the 1950s, not only because it perceived communism as an ideological threat to democracy, as commonly perceived, but probably also as a bulwark against irredentist secessionist forces. The effects of the ban on the potential for German reunification were the subject of thorough oral arguments before the German Federal Constitutional Court.205
In fact, while democracies disguise their fight against secessionist political parties, the origins of the concept of “militant democracy” have always been tied to the war against secessionism. When Karl Loewenstein advocated for militant democracy to fight against fascism in 1937, from his exile in the United States, he explicitly discussed the fact that such an approach is warranted to protect against “the political activities of foreigners or alien emissaries on the national territory.”206 He gave the example of the militant democracy used by Czechoslovakia to protect against irredentist secessionist forces in Sudetenland.207 Similarly, the European Convention on Human Rights from 1950 allows for the restriction of freedom of expression to protect territorial integrity.208
III. ETERNITY CLAUSES AND THE “UNCONSTITUTIONAL CONSTITUTIONAL AMENDMENT”
So far, I have shown that, contrary to the prevailing perception that democracies allow for secessionist political parties and even forbid banning them, democratic constitutions commonly authorize bans on secessionist political parties. I have further offered an interpretation of the practice under which democracies camouflage their exercise of the ban, not admitting openly that they are targeting secession per se. This Part examines the tool of “eternity,” “intangibility,” or “immutable” clauses, and the accompanying judicial doctrine of the “unconstitutional constitutional amendment” as a mechanism to combat secession.
A. The Use of Eternity Clauses
Another lesson democracies attribute to WWII is that they should set limits on the power to amend the constitution.209 Arguably, the Nazis could overtake power in a legal way and transform the Weimar Republic from a democratic to a totalitarian system because there was no concept of an unalterable democratic core to prevent this type of transformation.210 Such limits on the amendment power are known as constitutional “eternity clauses.” Eternity clauses grant absolute entrenchment status to certain constitutional values and rights to ensure that they remain eternal, and are not amended.
To protect this eternal status, courts developed the doctrine of the “unconstitutional constitutional amendment.” This accompanying judicial doctrine means that, even if eternity clauses are violated by a constitutional amendment, the courts may declare such an amendment unconstitutional, and, thus, invalid. The eternity clauses may be codified expressly or read implicitly into the constitutional document by the constitutional courts.
Scholars customarily associate eternity clauses with the protection of constitutional rights. A well-known example that stems directly from the lessons of WWII is found in Germany’s treatment of human dignity as eternal and unamendable under its Basic Law.211 It is less known that eternity clauses originally developed as a way to protect the federal or unitary structure of the state, and in particular, prevent secession.212 States treated their territorial integrity as an existential value and a prerequisite for attainment of all other constitutional values and rights. Only later did eternity clauses become a tool also used to protect constitutional rights. In fact, the German Basic Law itself protects the federal structure of Germany from being amended.213 Thus, the German Constitutional Court ruled in 2016 that Bavaria could not hold a referendum on independence, as secession is not allowed under the Basic Law.214 The Italian Constitutional Court issued a similar decision in 2015 preventing Veneto from holding a consultative referendum on independence based on the reasoning that such a referendum would violate the constitutional principle of territorial integrity and unity of the state.215
Contrary to conventional wisdom, constitutional democracies often treat their federal structure, unitary status, and territorial integrity as unamendable. These constitutional clauses are intended to deter secession, and if deterrence fails, to trigger the “unconstitutional constitutional amendment” doctrine. Forty-three constitutions, which represent 22% of world constitutions, include an explicit eternity clause protecting the territorial integrity of the state from amendment. These include the constitutions of Brazil, France, Germany, Italy, Portugal, Romania, Turkey, and Ukraine.216 Sixty-five constitutions do not have an explicit eternity clause but nonetheless identify in their preamble and/or fundamental principles and state duties the imperative to protect territorial integrity and/or national unity as a permanent, eternal, and unchangeable value.217 These include the constitutions of Argentina, Australia, Bulgaria, China, Czech Republic, Cyprus, Finland, Luxembourg, Macedonia (former Yugoslav Republic), Mexico, Norway, South Africa, and Spain.218 The Cambodian Constitution, for example, not only declares that its territorial integrity “shall never be violated,”219 but provides that any treaty or decision of the legislature to the contrary shall be annulled.220 Another example is Ecuador, which provides that its territory is “unalienable”221 and “[u]nder no circumstances shall the exercise of autonomy allow for secession from the national territory.”222 Sri Lanka’s Constitution prohibits any individual, including officials and Members of Parliament (MPs), from promoting separatism. Anyone who violates this prohibition not only loses his office but may forfeit his property and civic rights, including the right to a passport and the right to pursue an occupation that requires a license from the state.223 The Tanzania Constitution clarifies that “No person shall have the right to sign an act of capitulation and surrender of the nation.”224 Overall, 108 constitutions, which represent 56% of world constitutions, rigorously protect territorial integrity to the point of declaring it an unamendable value. When looking at democracies and semi-democracies, sixty-four countries (roughly 48% of democracies and semi-democracies) treat territorial integrity as eternal.
Many countries command their armies to protect territorial integrity as a constitutional duty and often treat a potent threat to territorial integrity as justifying an emergency regime. It is also very common to require the heads of state—typically the executive branch, but sometimes the judicial branch—to take an oath to defend the territorial integrity of the state. The Indian Constitution even makes it an explicit oath of office of ministers, parliamentary candidates, MPs, justices, comptrollers and auditors general, candidates to state legislatures, ministers of state, members of state legislatures, and judges of high courts.225 Constitutions sometimes even explicitly state that it would be high treason for the president to allow the territorial integrity of the state to dismember.226 Some constitutions explicitly entrust the executive branch to dismantle regional assemblies and governments if they act against the territorial integrity of the state.227 This is the interpretation offered by the Spanish central government to article 155 of the Spanish Constitution in its struggle against Catalonia’s separatism. The Spanish central government acted upon this interpretation when it dismantled the Catalan governing institutions and called for early, extraordinary elections at the end of 2017.228 All these measures, like the ban, are intended to prevent secession from gaining momentum to the point that courts will need to apply the “unconstitutional constitutional amendment” doctrine.
A full forty-four constitutions, which represent 23% of world constitutions, even impose a constitutional duty upon the citizen, and not only the state, to protect the territorial integrity of the state.229 This is a special feature, since we traditionally perceive constitutions as imposing duties on the government while bestowing rights on the people. Further, if every citizen has a constitutional duty to protect the territorial integrity of the state, it is much harder to advance a goal of separation from the state. Citizens will find it difficult to rely on other constitutional rights, such as freedom of speech and association, to promote separation when these rights conflict with their constitutional obligation to protect territorial integrity.
Moreover, the lack of an explicit eternity clause did not prevent courts from implying territorial integrity as an unamendable constitutional value. The American Civil War was fought over the protection of the Union no less than the abolition of slavery, and its aftermath may stand for the proposition that in the United States, secession amounts to an “unconstitutional constitutional amendment.”230 In fact, Chief Justice Chase seemed to have found a constitutional text that prohibits secession. In Texas v. White, based on the Constitution’s preamble, he stated “It is difficult to convey the idea of indissoluble unity more clearly than by these words. What can be indissoluble if a perpetual Union, made more perfect, is not?”231 He further suggested that territorial integrity was part of (what we would term today) the basic structure of the Constitution. “The Constitution, in all its provisions, looks to an indestructible Union, composed of indestructible States.”232 He also seemed to rely on Article IV of the U.S. Constitution to empower the national authorities to prevent secession by guaranteeing to each state a republican form of government.233 This was no mere dicta. It led to the majority decision that the Court enjoyed jurisdiction over the case since Texas never ceased to be part of the Union, though it was not yet represented in Congress during the time the case was heard (the Reconstruction era).234 The Court held that Texas never achieved secession though a majority of its population supported secession in a referendum that led to its departure from the Union and its membership in the Confederation during the Civil War.235 In India, Justices of the Supreme Court identified territorial integrity as unamendable.236 There are Swiss constitutional scholars who suggest that the federal structure is so fundamental to Switzerland that courts should recognize it as an unamendable value.237
We can trace the idea of unamendability also in the U.K. The Acts of Union between the English and Scottish Kingdoms that created Great Britain in 1707 stipulated that the Union, being irrevocable and indissoluble, will endure “in all time coming.”238 The Union with Ireland Act 1800 contemplated that the Union between Great Britain and Ireland will “have effect for ever.”239 There is judicial support for the proposition that these Acts should be treated as enjoying special constitutional status, such that they could not be repealed implicitly, requiring full accountability of parliament for any deviation from them.240 Furthermore, Dicey, the most renowned British constitutional scholar of the nineteenth and early twentieth century, treated the Acts of Union as amounting to a constitutional document that may only be amended via a national referendum.241 In fact, the Union with Ireland did not last, and was amended via extra-constitutional politics.242 Further, since the 1970s, it is an established practice in the U.K. to hold referenda before effecting major change in the U.K.’s relationship with its constituent parts or with the E.U., if the change is likely to affect sovereignty.243 If Britain is viewed as based on the principle of parliamentary sovereignty, it means that issues affecting national sovereignty cannot be amended via a regular constitutional amendment, but require instead an extra-constitutional act of the People.
B. The Doublespeak of Democratic Constitutions
If territorial integrity is protected under the “unconstitutional constitutional amendment” doctrine in various countries, why did scholars interpret constitutional declarations of indivisibility to allow for regular constitutional amendment?244 We find the same doublespeak and underhanded usage of eternity clauses as we did with bans on secessionist political parties. Many constitutions contain seemingly contradictory provisions: on one hand, they protect the inalienability of the territory, the entrenched nature of the federal system, or the indivisibility of its unitary nature; and on the other hand, some of them include clauses that set arduous procedures for territorial change. For example, the Ukrainian Constitution article 2 states, “The territory of Ukraine within its present border is indivisible and inviolable.”245 Furthermore, article 157 states that “[t]he Constitution of Ukraine shall not be amended, if the amendments . . . are oriented toward the liquidation of the independence or violation of the territorial indivisibility of Ukraine.”246 Nonetheless, article 73 declares that “[i]ssues of altering the territory of Ukraine are resolved exclusively by an All-Ukrainian referendum.”247 This presents a contradiction in terms.
Similarly, article 2 of the Spanish Constitution proclaims, “The Constitution is based on the indissoluble unity of the Spanish Nation, the common and indivisible homeland of all Spaniards.” Article 8 further requires “the Armed Forces . . . to guarantee the sovereignty and independence of Spain and to defend its territorial integrity.” But article 168 enables amendments to these provisions through a special process—one that is also used for a total revision of the Constitution.248 This stance of the Spanish Constitution to allow revision of every article but to make the revision of some articles impossible to achieve has been termed “political schizophrenia.”249 This doublespeak phenomenon is found in forty-two constitutions or 22% of world constitutions, including Bulgaria, Finland, France, Luxembourg, Spain, and South Africa.250 All of these countries have a ban on secessionist political parties and/or eternity clauses, and they also provide a procedure for territorial change, thus the doublespeak.251
The number of countries engaging in doublespeak is higher were we to also include constitutional provisions that provide procedure for changes in boundaries (as distinguished from territories).252 Additionally, many more countries engage in doublespeak if we include those that impose constitutional duties on state organs to protect territorial integrity on the one hand and provide a procedure for territorial change on the other hand.253
The same ambiguity arises from judicial decisions. As we have seen, the constitutions of both Ukraine and Spain include eternity clauses preventing secession. In both countries, the constitutional courts intervened before secessionist referenda were held in the seceding area and declared them unconstitutional because only the national political bodies could hold a national referendum on the subject. Secession was a national rather than a regional matter.254 In both cases, the courts knew that secession was not supported at the national level, and the requirement to hold a national referendum effectively meant that secession could not be brought about by constitutional means.255
These court decisions do not amount to an application of the “unconstitutional constitutional amendment” doctrine because no amendment was adopted and none was abolished. Moreover, the courts refrained from stating that territorial change could not be brought about by amendment. Rather, the decisions reflect the courts’ understanding that it is better to intervene early in the process of secession. By the time the “unconstitutional constitutional amendment” could be applied, the seceding area would no longer be subject to the jurisdiction of the constitutional court of the parent state. In a successful secession, the new state would be bound only by its own new legal system, established through revolution. In the Ukrainian case, even though the court’s decision preceded the referendum, it was too late in the game.
It is uncommon to find a court invalidating secession based on the doctrine of the “unconstitutional constitutional amendment.”256 Secessions do not usually occur through the staid process of a constitutional amendment; they are usually achieved by force and in illegal and extra-constitutional ways.257 And, if there is no applicable constitutional amendment passed according to the procedures prescribed in the constitution, then there is no amendment that can be declared unconstitutional. In this context, as in others, the “unconstitutional constitutional amendment” doctrine serves more as a threat than a reality. Yet the doctrine’s existence may suggest to secessionists that they have no way to achieve their aims other than by forceful, extra-constitutional means.
C. On the Nexus of Eternity Clauses and Bans on Political Parties
What theoretical basis may be offered to support the concept of eternity clauses? Why differentiate between the amending power and the constitution-making power? Two theories may be offered to answer these questions. Under one theory, the amendment power is inferior to the original constituent power that created the constitution in the first place. Constitutional amendment power is a constituted power and a derivative power from the constitutional text. In contrast, the original constituent power is above the constitutional text and is not limited by it. Constitutional theories usually add that, while the amending power is entrusted to representative bodies, the original constituent power belongs to the People alone. Thus, we want the People, rather than their representatives, to decide the most basic features protected under eternity clauses.258
Under another theory, even if both the amending power and the original constituent power belong to the same body, any change to the most fundamental characteristics of the constitution must be done by replacing the constitution as a whole. This will guarantee that the change is not done lightly, but with full accountability. It will lead to greater awareness of and deliberation by the People about the consequences of the change.
Under both theories, the foundational idea is that constitutional amendment may lead to a partial change of the constitution, not a replacement of it. The amending power may only “amend” but not “destroy” the constitution. A revision of its most basic identifying features or basic structure requires that a new constitution be adopted, since, without them, the constitution would lack coherence.259
There is a deep connection between the two mechanisms that constitutional democracies use to prevent secession. While banning secessionist political parties obstructs attempts to amend the constitution through the legislature, the eternity clause with its concomitant “unconstitutional constitutional amendment” doctrine is intended to void secessionist constitutional amendments that do get through the political branches. The former prevents secession a priori, and the latter frustrates secession after the fact. The ban on political participation at elections is the front guard of the constitutional system. It prevents those challenging the basic values and structure of government from even reaching power. But, if the ban fails and those contenders reach political power, the “unconstitutional constitutional amendment” doctrine serves as the rear guard preventing their constitutional amendment from being valid. Both are mechanisms that express an absolute commitment to the territorial integrity of the state.260
In fact, when reflecting on these two mechanisms, it becomes apparent that countries that adopt one of them (ban or eternity) may legitimately implicitly infer the existence of the other in their constitutional system. This is important because often a constitution may embody only one of the mechanisms. Thus, for example, only thirty-seven out of 192 constitutions (19%) have both a ban and an explicit unamendability clause to protect territorial integrity.261 Sixteen of the thirty-seven states are democratic or semi-democratic. The number may rise to seventy-seven (40%) constitutions if we include not only explicit unamendable clauses but also declarations of unamendability.262 Forty-three out of the seventy-seven states are democratic or semi-democratic.
Roughly 70% of the constitutions have at least one of the mechanisms. When looking at democracies and semi-democracies, forty-three democracies and semi-democracies have both mechanisms (32%), while eighty-three democracies and semi-democracies (62%) have at least one of the mechanisms. France, Germany, Italy, Turkey, Ukraine, and Spain are among the democracies/semi-democracies that adopted both a ban on political participation and unamendability to protect territorial integrity.
Why do I claim that these unconventional weapons are twinned in this way? The two mechanisms are both justified under the theory of militant democracy and as a prominent lesson of WWII. Though the literature identifies militant democracy with the ban on political parties alone, Karl Loewenstein, in his seminal articles, advocated the use of both mechanisms. He advocated the ban: “Calculating adroitly that democracy could not, without self-abnegation, deny to any body [sic] of public opinion the full use of the free institutions of speech, press, assembly, and parliamentary participation, fascist exponents systematically discredit the democratic order and make it unworkable by paralyzing its functions until chaos reigns.”263 In the same article, he later also hinted at unamendability to deal with the same problem: “Constitutions are dynamic to the extent that they allow for peaceful change by regular methods, but they have to be stiffened and hardened when confronted by movements intent upon their destruction.”264 But scholars equated militant democracy with a ban on political participation alone since Loewenstein mentioned unamendability only briefly and in passing.
Both tools share common historical roots. They became increasingly prevalent around the world after WWII as an antidote to anti-democratic forces. Also, they use the same method of absolute entrenchment. The ban serves eternity clauses, since typically to amend a constitution one needs the consent of both elites (through the legislature) and the People (through a referendum or the like). If the system blocks the reformers’ access to the legislature, they will also typically not be able to embark on constitutional amendment. The reverse is also true: while theoretically a system may opt to amend the ban on political parties, the amendment could and should not run against the basic values of the system as expressed in eternity clauses.
The two mechanisms are also problematic in similar ways. Systems that resort to these tools too often or too casually may undermine their own democratic legitimacy. How democratic is a country when it prevents peaceful competition by citizens to change policies through prescribed procedures?265 The same is true with regard to eternity clauses. Supreme constitutions with the accompanied power of judicial review raise a counter-majoritarian difficulty. Why should a constitution, adopted in a bygone era, prevail over the will of the current majority as it is properly expressed in legislation? Why should a judicial interpretation of the constitution override the representative branches’ interpretation?266 To these penetrating questions, constitutional systems provide a ready answer. The supreme authority of the constitution—protected by the judiciary—maintains the supremacy of the will of the People, which cannot be eroded by regular legislation.267 If the People disagree with the courts’ interpretation of the constitution, they may amend the constitution to express their will. Thus, the constitutional amendment process resolves the counter-majoritarian problem, or so the claim goes.268 But, under eternity clauses and the doctrine of the “unconstitutional constitutional amendment,” the amendment path is closed, and thus the counter-majoritarian difficulty intensifies.
Since both constitutional tools—the ban and an eternity clause—raise pressing and self-sabotaging legitimacy challenges, constitutional systems are wise to treat these tools as last resorts. They should serve as a potent threat against extreme forces but rarely be utilized. Just like nuclear weapons, their force is in their threat rather than their deployment. Even weapons of self-defense can destroy the user.
Understanding both tools as a mirror of one another may constrain courts when developing either an implied ban or an implied eternity clause. The courts may develop an implied tool to serve only the values identified in the explicit tool already provided for in the constitutional system. Understanding them as complementary may also offer a textual basis for courts when developing an implied tool as long as the other tool is explicitly present in the constitutional system. This may address some of the criticisms raised against the courts for developing an implied ban or an implied eternity clause.269 Of course, a constitutional system may avert these threats and opt not to expand its militant tools.270 But, if a system does decide to imply the existence of one of the tools, it has strong theoretical justifications to do so if the other tool exists in its system. It will also develop a coherent identity if the values protected under both mechanisms are the same.
The data may reveal another important nexus. The general ratio between Proportional Representation (PR) and First Past the Post (FPTP) election systems is 60% to 40%, respectively.271 However, when looking at the countries that do not protect territorial integrity, the ratio is reversed: 58% (FPTP) to 42% (PR).272 This may suggest that PR countries feel a greater need to protect territorial integrity than do FPTP countries. PR systems must confront the challenge of extreme political parties and instability of government rule in parliamentary systems.273 To meet these challenges, they use not only constructive votes of no confidence and electoral thresholds, they also enable carryover of pending bills from a previous legislature, rather than treat elections as requiring discontinuity of parliamentary debate.274 They, furthermore, adopt militant democracy tools. There may be a tradeoff between the type of an election system and the need to resort to militant democracy tools.275 While militant democracy is traditionally perceived as a substantive doctrine, this Article reveals its “structural” nature as a tool in greater need in proportional representation election systems.
Similarly, when examining the democratic countries enumerated in Lijphart’s study,276 we find that “consensus” democracies protect territorial integrity no less than majoritarian democracies. The majority of the countries that may serve as an archetype of consensus democracies ban secession. Only two are silent on the topic.277 In fact, only three of the twelve in Lijphart’s group of majoritarian democracies prohibit secession via a ban and/or unamendability. Almost half are silent on the topic, with the United Kingdom allowing it.278 In the unitary/consensus democracies group, seven of the twelve countries adopted a ban and/or unamendability to fight secession.279 Of the five federalist countries that have mixed features, two are silent and three ban secession.280 If we take out the majoritarian democracies, thirteen out of twenty-five “consensus” democracies ban secession via the unconventional constitutional tools. An examination of Lijphart’s list of existing consociationalist democracies that are members of the U.N. reveals that the overwhelming majority of these countries prohibit secession via a ban and/or unamendability as well.281 When juxtaposing consensus and/or consociationalist democracies with majoritarian systems, the former may even protect territorial integrity more rigorously than the latter to compensate for their policies of accommodation.
D. Constitutional Paradox
When it comes to the tools used to combat secession, why do constitutional democracies create this gap between what is said and what is done? Why do they have a ban on the book, yet use a different rationale? Why do they obfuscate their stance on secessionism? Democratic states are hampered by a constitutional paradox. Their existence and legitimacy are grounded on the consent of the governed, which lends support to their right to self-determination and self-rule.282 Yet, when secessionists argue for their right of exit and entitlement to separate existence based on the same principles, democracies deny their claims. Otherwise democracies would undermine their own unity and form, perhaps even their own existence.283 The dilemma is whether denying the right to secession is a greater compromise of democratic values than is compromising the state’s territory and cohesion.284
Moreover, democracies use the most unconventional constitutional weapons to fight secession. These weapons are traditionally justified to protect democracy, but in the secessionist context they are used to protect existential needs of the state. To justify the use of unconventional constitutional weapons to protect the territorial integrity of the state, democracies disguise the fight as a fight for democracy.
Another limitation on the use of these constitutional tools is the “peer pressure” of the international community. For example, the European Court of Human Rights overturned decisions of both Turkish and Bulgarian courts that banned secessionist political parties, based on its policy that a secessionist agenda alone does not provide legitimate grounds for a ban, unless the parties’ agendas are antidemocratic or they are associated with violence.285 The court, in its 2005 decision regarding Bulgaria, stated:
In a democratic society based on the rule of law, political ideas which challenge the existing order without putting into question the tenets of democracy, and whose realisation is advocated by peaceful means must be afforded a proper opportunity of expression through, inter alia, participation in the political process. However shocking and unacceptable the statements of the applicant party’s leaders and members may appear to the authorities or the majority of the population and however illegitimate their demands may be, they do not appear to warrant the impugned interference. . . . It is of the essence of democracy to allow diverse political programmes to be proposed and debated, even those that call into question the way a State is currently organised, provided that they do not harm democracy itself.286
The European Court of Human Rights’ attitude encourages states to propound rationales other than self-preservation when banning secessionist activities. Otherwise, they may be accused of totalitarian impulses, operating behind a façade of democracy only to retain power and oppress the opposition.
IV. SUPPOSEDLY ENABLING MECHANISMS: REFERENDA
So far, this article has shown that constitutional democracies often prohibit secession in their constitutional document, yet conceal their fight against secession because of the constitutional paradox that secession poses to them. This Part discusses those countries that supposedly permit secession and even set the procedures to achieve it in their constitution.
Scholars argue that constitutional provisions that authorize secession “are extremely rare.”287 Monahan and Bryant found only five existing countries that have adopted them: Austria,288 Ethiopia,289 France (regarding its overseas territories),290 Singapore,291 and St. Kitts and Nevis.292 The former Soviet Union,293 the former Czech and Slovak Federative Republic,294 and the former State Union of Serbia and Montenegro also allowed secession.295 Other studies add to the list the European Union296 and Canada because of its Supreme Court decision regarding Quebec that allows for secession by agreement.297 Some constitutional scholars argue that these rare examples demonstrate that secession is not alien to constitutional law and can be achieved through lawful procedures. They recommend the adoption of constitutional provisions that enable secession if the process conforms to rule-of-law principles.298
My examination of 192 constitutions reveals that there are many more countries that address changes of territory in their constitutions. Seven countries (3.65%), including Ethiopia, Liechtenstein, St. Kitts and Nevis, Sudan, Uzbekistan, Slovakia, and arguably South Africa, explicitly recognize the right to secede and provide the procedure for achieving it in their constitution. But, three of these countries (South Africa, Slovakia, and Uzbekistan) also declare their territorial integrity indivisible, and Uzbekistan and Slovakia further ban secessionist political parties.299 This list differs from the one provided by Monahan and Bryant.300 In addition, Bosnia and Herzegovina is unique in explicitly authorizing the Constitutional Court to decide “[w]hether an Entity’s decision to establish a special parallel relationship with a neighboring state is consistent with this Constitution, including provisions concerning the sovereignty and territorial integrity of Bosnia and Herzegovina.”301
Fifteen countries require the approval of both the executive and the legislature to bring about change in territory, leaving open to dispute whether they apply to secession. Territorial changes may not amount to secession if no transfer of population is involved. These provisions intend to clarify that, even though the executive is oftentimes in charge of foreign relations, when treaties or international agreements involve change of territory, the executive must obtain the legislature’s approval.302 Six countries require the approval of a supermajority, ranging from two-thirds to three-fourths of the entire legislature (not just those voting), for territorial change.303 Thus, twenty-one countries (11%) leave the decision regarding territorial change to their representative bodies.304
Thirteen countries require the approval of both the legislature and a national approval for territorial change.305 Seven countries require approval by enactment of law and a supermajority ratification in a national referendum.306 Seven countries require the approval through both legislation and a referendum of the “concerned” or “interested” population, which leaves open for interpretation whether they demand a national or a regional referendum.307 In total, twenty-seven (14%) countries require the cooperation of their representative bodies and the people to affect territorial change.
Comoros is candid enough to admit that its procedure for territorial change would not be enough without constitutional amendment,308 but this is true of the overwhelming majority of all the other countries discussed above, as they too provide for territorial integrity as an inviolable principle.309 Similarly, Costa Rica openly requires both “the approval of the Legislative Assembly, by a vote of no less than the three-quarters part of the totality of its members, and that of two-thirds of the members of a Constituent Assembly, convoked to [that] effect.”310 It is typically difficult to amend the constitutions in the various countries as amendments need to meet supermajority requirements of both legislative bodies and the people. Moreover, since the constitutional amendment at stake involves an inviolable principle it is not at all clear that the courts of the relevant country would not declare such an attempt unconstitutional.
Only a few countries have explicit constitutional clauses that allow secession, and typically they set such difficult procedural hurdles that secession is all but impossible to achieve under the provisions. They say “yes” to secession as a theoretical matter but mean “no” as a practical matter. It is doubtful whether these countries should and could be used to demonstrate that secession may align with constitutionalism.
In some cases, countries constitutionalize secession only to lure smaller countries to join their federation. Once they enlarge the federation, they renege on their promise to enable secession, and treat the state as unitary or indivisible. This explains the paradoxical situation in which a constitutional secession clause was found in non-democratic countries like the former Soviet Union,311 Burma,312 and China.313 Lenin was braggadocios enough to admit that the Soviet Constitution enabled secession to enlarge the bigger states, not to accommodate the smaller states.314 The Soviets left the secession clause intact, despite amending the Constitution a few times over the years,315 since it was clear to all that it was never intended to be acted upon.316 The Soviet constitutional secession clause did not enumerate the procedure for secession. It stated that “[e]ach Union Republic shall retain the right freely to secede from the USSR[,]”317 which could not even theoretically serve as guidance for the secession of the thirteen out of fifteen republics that led to the collapse of the Soviet Union. Gorbachev required 66% support in a referendum, but most secessionist referenda did not abide by this requirement.318 Today, the Russian Federation prohibits secession in its Constitution.319
The Chinese abolished the constitutional provision for secession once the Communist Party “convinced” ethnic nationalists in Chinese mainland and in surrounding territories to join the republic based on the possibility of secession.320 The current Chinese Constitution not only protects its inalienable territory, but its preamble also states, “Taiwan is part of the sacred territory of the People’s Republic of China. It is the lofty duty of the entire Chinese people, including our compatriots in Taiwan, to accomplish the great task of reunifying the motherland.”321 It thus makes clear its intention to (re)gain and maintain control over Taiwan.
In Burma, the right of secession appeared in the 1947 Constitution,322 but was repealed in 1974 with the imposition of the Constitution of the Socialist Republic of Burma.323 In Ethiopia, the secession clause was also intended to enable the formation of the Federation and prevent a repeat of an Eritrean-like secessionist struggle. However, all concerned understand that it is unlikely that the authoritarian regime will allow any state or group to actually secede based on the clause.324 In Sudan, two civil wars led to the inclusion of the secession clause in the Constitution as part of a Comprehensive Peace Agreement reached between the North and the South in 2005. South Sudan voted for secession in 2011 by an overwhelming majority of 98.8% with almost full participation of the electorate body.325
In other cases, the constitutional secession clause imposes such burdensome obstacles on the secession process that it becomes impossible to follow. The obstacles come in various forms. They may appear as a supermajority requirement, demanding that, even if a majority of the people approves secession in a referendum, the result will be void because the supermajority requirement is not met. This is not a mere theoretical problem, but has materialized time and again in different settings. For example, the Constitution of St. Kitts and Nevis enables secession of Nevis, but requires the support of two-thirds of the votes cast in Nevis at a nationally organized referendum.326 In 1998, 61.7% of the Nevis electorate approved secession, which meant the failure of secession.327 The supermajority requirement is also part of the constitutional law of Canada that requires a “clear majority” for secession.328
Another obstacle to secession may appear in the form of a requirement that the entire populace of the unitary or federated state approve secession, not just the people of the seceding area.329 The Slovakian Constitution, for example, requires the consent of the national populace.330 Very often, the national populace will have a different stand than the people in the region that seek to secede.
In addition, secession referenda are typically treated as merely consultative in the sense that they are binding if the result is negative, which then ends the secession process. But, if they are positive, they do not bind the representative bodies. In fact, on more than one occasion, although the secession referendum yielded a positive result, secession did not take place. Western Australia, for example, voted in favor of secession in a referendum, enjoying the support of two-thirds of the voters, but secession was never achieved.331 Referenda are the opening phase of a long process of negotiation, which may or may not culminate in an agreement on secession, as the Supreme Court of Canada openly admitted. If no agreement is met, secession will not be allowed from the constitutional law perspective of the rump state. This lack of agreement will also affect the willingness of the international community to recognize the formation of a new state that runs against the territorial integrity of the rump state.332 Typically, the central government of the mother state is put in charge of the secession process, deciding the timing and conduct of the referendum, the phrasing of the question, and the interpretation of its results.333 Moreover, in 79% of the countries, the culmination of the secession process requires a constitutional amendment that, if not struck down as unconstitutional by the courts, at least grants veto power to the central governing bodies.334
V. CONCLUDING LESSONS FOR COMPARATIVE CONSTITUTIONAL LAW
While scholars argue that constitutions by and large are silent about secession, and this silence may be interpreted as tacit permission to secede, this Article offers a very different interpretation of the constitutional landscape and practice. It argues that 79% of world’s constitutions ban secession. Moreover, the overwhelming majority of countries ban secession by using their most potent unconventional constitutional tools. Many constitutions even treat a consensual secession between the mother state and the seceding part as amounting to treason, and a violation of both constitutional and criminal law on the part of the leaders involved. While providing the unconventional constitutional weapons, constitutional democracies try to conceal their fight against secession for as long as it is possible in a given context because their attitude towards secession is laden with democratic paradoxes. To fare better in international tribunals and the international court of public opinion, they prefer to portray their fight against secessionists as an issue of national security, or a defense of citizens against extremities of racism or other bigotry. Once the confrontation becomes too messy to handle, however, the true nature of the struggle may be revealed to all, as in the 2017 Catalonian crisis.
Yet, if both the mother state and the seceding state reach agreement on secession, it will supposedly overcome these constitutional hurdles. Such an agreement would be extra-constitutional but effective. Neither side will challenge the secession in international tribunals, and the international community will most likely accept such agreement. In contrast, when secession is opposed, the international community is expected to assist the mother state by at least delaying or totally preventing the recognition of the new state, as the Canadian Supreme Court suggested.335 In fact, Nova Scotia in the nineteenth century, and Western Australia in the twentieth century, each gained clear majority support of the people in the region to secede from Canada and Australia respectively. Yet, the U.K.—their ruling colonial empire at the time—refused their request for recognition to protect reliance interests of the mother states’ populations.336
Why then do constitutions ban secession outright even though supposedly an agreement between the rump state and the seceding region will overcome such a prohibition? The answer may lie in both strategic and principled constitutional law considerations. Strategically, a total prohibition may serve as the most effective deterrent against secession. A prohibition may raise the stakes for secessionists and make the path to secession very difficult to achieve. The facts may offer support for this proposition. When secession is legal under international law, as in decolonization, its success rate is 77%. When secession is forbidden, the success rate is only 16%.337 Secessionists must feel strongly enough about their preferences to bear the costs of violating states’ constitutional and criminal laws. The total prohibition supplies the leaders of the mother state with the most potent weapons to quash the rebellion. If the mother state’s leaders fail to prevent secession, a total ban may give the mother state an advantage at the negotiation table to extract the best deal it can under the circumstances. The constitutional prohibitions matter as the transaction costs of reaching a separation agreement are so substantial that even the Canadian Supreme Court, while recognizing Quebec’s right to secede by agreement, admitted that negotiations might fail.338
But a complementary explanation of total prohibitions may lie in the fact that secession cannot truly be achieved via a constitutional amendment and might require a break with constitutional continuity by populations in the seceding and remaining territories. It is already common wisdom that secession requires at the minimum a self-constituting act by the seceding region. This is why the U.K. held a referendum on independence in Scotland in 2014. This is why the Canadian Supreme Court required the expression of a “clear majority” of the Quebec people before the rest of Canada started negotiating the terms of the divorce with Quebec. This is why Russia/Crimea held a referendum in Crimea on its annexation to Russia (after the fact).
But, my argument is that secession also requires an independent self-constituting act on the part of the remaining regions. While we typically treat popular sovereignty as a population concept, my argument is that “We the People” is a territorial concept composed of the combination of citizens and territory. Thus, the decision to cut the rump state’s size and relinquish responsibility for part of its citizens involves a redefinition of the sovereign body in the rump state. In that sense, secession amounts to an “annihilation”339 of the existing constitutional order and requires a new constitutional start.
When the Canadian Supreme Court in Quebec held that secession may be done via constitutional amendment and did not require a revolution, it simultaneously set the amendment process as composed of negotiations between the seceding part and the rest of Canada.340 Already the negotiations on separation rely on the consent of two separate “We the People” bodies whose birth are attributed to secession. The Canadian Supreme Court may call it a constitutional amendment to preserve the appearance of continuity, but its essence is revolutionary. The same is true of the U.S. Supreme Court decision in Texas v. White. Chief Judge Chase suggested that secession may come about through revolution or by consent of the remaining states.341 The consent of the remaining states is the consent of a new popular sovereignty body, which involves a redefinition of the basic norm of the rump state. When a constitutional amendment attempts to redefine who the constitution-making body is, it requires a new constitutional beginning. The legitimacy for this constitutional amendment comes from without rather than from within the constitutional order, even if the system tries to grant it the appearance of continuity and “constitutional amendment.” It amounts to the birth of a new republic.
In fact, even those who treat constitution-making and constitution-amendment powers as on par and belonging to the same entity,342 may have to come to terms with the fact that the one thing constitutional amendment may not do is redefine the constitution-making power. This must be done by definition through a new constitutional beginning, rather than through amendment.
This is why secession differs from mere massive immigration or even emigration waves or generational changes that involve redefinition of citizenry without the remaking of territory. These do not involve a redefinition of “We the territorial People.” The new members are granted citizenship rights on the social compact terms defined by the existing popular sovereign. This is also why secession differs from mere change of territorial boundaries without the loss of part of the citizenry body. Territorial change alone does not require “We the territorial People” to come to terms and accept that it is no longer responsible for people who used to be part of it. Relinquishment of citizenship on a massive scale, when coupled with the loss of territory, requires the remaining people to come to terms with the separation.
As such, secession requires a new beginning by two new People—the remaining population of the mother state as well as the seceding population.343 Each must engage in a self-defining act of constitution-making independent of the other in the sense that ultimately, a new beginning is a factual matter from a constitutional perspective.344 Thus, the consent of the seceding population in a referendum is insufficient to legitimize secession, as was attempted in Scotland in 2014.345 Nor is it enough to gain the people’s consent in a national referendum, as provided in various constitutions.346 The two new People must reach agreement from a constitutional perspective to avoid competing claims to sovereignty over people and territory.
Furthermore, the ICJ decision in the Kosovo case suggests that a unilateral declaration of independence may not necessarily violate international law (though it can, in defined categories).347 Moreover, the Kosovo case stands for the proposition that Kosovo did not violate international law because its declaration of independence was openly extra-legal in international terms.348 It did not try to rely on the interim international arrangements set in Kosovo through U.N. governing bodies’ decisions.349 These interim arrangements forbade unilateral acts by both Kosovo and Serbia, as is typical in interim arrangements, to address the concerns of the parties to the conflict.350 Rather, Kosovo’s unilateral declaration of independence clearly states that the interim international arrangements failed and that is why Kosovo unilaterally declared independence.351 The Kosovo case shows that, in international law as well, secession may require an extra-legal act to succeed and be accepted as legitimate. But it is not enough to avoid the condemnation of the international community. To gain the international community’s acceptance, the seceding part will typically need to reach agreement with the rump state.352
These principled constitutional considerations may explain the prevalence of total prohibitions on secession throughout the world. It should be clarified that, even if secession oftentimes cannot be achieved via constitutional amendment, that does not mean that secession necessarily requires violence or force. But it may require a “break with the past”—a break with constitutional continuity.
A total prohibition on secession may be less justified from a constitutional perspective when it does not challenge the identity of the popular sovereign body in a given territory. This occurs, for example, when the relationship between the central authorities and the secessionists is governed by a compact or treaty. Some hallmarks of treaty-like relationships (as contrasted with national enterprises) may include the veto right of each member state/region/province over any amendment of the constitutional document; and the right of nullification, so that each member may decide to prevent the application of a federal statute in its territory.353 In such cases, the withdrawal of a member may not prompt the need of either it or the former confederation to begin anew. This describes the situation in the European Union today,354 and it explains why Brexit is possible via a constitutional amendment, rather than through constitutional revolution.355 In fact, article 50 of the Treaty on European Union, which grants member states the right to secede, originated with the failed Treaty Establishing a Constitution for Europe.356 Under the Constitutional Treaty, the European Union attempted to further qualify its members’ veto power and move to a more manageable qualified-majority regime and integration (less rights to voice) while clarifying that each member has an explicit right to secede (more exit rights).357 Generally, such treaty-like relationships may exist during the early stages of the formation of a federal state,358 as was the case under the Articles of Confederation in the United States.359 It may also be true during the early phases of an acquisition of a new territory (depending on the circumstances).
The analyses in this Article differ from the stories democracies usually tell themselves about secession in another fundamental way. The Canadian Reference re Secession of Quebec case, which is the most renowned decision worldwide on secession, stands for the proposition that secession may be achieved via constitutional amendment. The Court explicitly rejected the proposition that secession must be done extra-constitutionally and in a revolutionary manner.360 Scholars came to expect that the same logic applies globally. They further assumed that most constitutions are silent about secession, as the Canadian Constitution is.
Yet, this Article reveals that the Canadian case must be distinguished on several grounds that limit its precedential nature for world constitutionalism. First, the Canadian Supreme Court interpreted constitutional silence on secession, unlike 79% of world constitutions that explicitly and even vehemently prohibit it.
Second, Canadian constitutional development has been an evolutionary story, even when the changes themselves were revolutionary. Canada evolved from the British North America Act of 1867 through the Westminster Act of 1931 to the Constitution Act of 1982, all without admitting the truly revolutionary nature of each of these milestones. These Acts transformed Canada from a group of colonies to a dominion in 1867, then to an independent state in 1933 with links to the U.K., and lastly to a full independent state in 1982 that “removed the last vestige of British authority over the Canadian Constitution.”361 In contrast, many other countries develop through “break with the past” transformations.362 In keeping with its evolutionary story, it is thus fitting that Canada chooses to treat Quebec’s possible secession as merely a function of constitutional amendments. This pretense serves an important purpose: to lower the stakes of secession and achieve the split in a more peaceful manner. Other states do not necessarily share this desire.
Third, Canada’s constitutional decision must be understood against the unique historical relationship between Quebec and the rest of Canada. Quebec never consented to the Charter.363
Fourth, on a related point, the Canadian constitutional decision must be understood against the unique Canadian federalist structure. Each province has a “nullification” right to prevent the application of substantial parts of the Charter in its territory via the override power.364 Most importantly, each province enjoys the constitutional power to initiate constitutional change. This was an important consideration in the Canadian Supreme Court’s ruling:
The Constitution Act, 1982 gives expression to this principle, by conferring a right to initiate constitutional change on each participant in Confederation. In our view, the existence of this right imposes a corresponding duty on the participants in Confederation to engage in constitutional discussions in order to acknowledge and address democratic expressions of a desire for change in other provinces.365
These powers of constitutional initiation and nullification define the identity of the constitution-making body in Canada. Both powers bring Canada closer to a compact model than countries with neither mechanisms, like the United States. This may explain the Canadian Court’s ruling that if the Quebec people express their will to secede by a “clear majority” on a “clear question” in a referendum, the remaining provinces should negotiate. Even without this consultative referendum (as referenda have no formal constitutional status in Canada), Quebec is entitled to initiate constitutional change under the Canadian Constitution.
The story of constitutions’ treatment of secession thus challenges some of our most basic understandings of democratic constitutionalism. Although the Reference re Secession of Quebec is viewed as a landmark case for comparative constitutional law, its precedential value may be limited to Canada’s unique constitutional context. Although militant democracy has been justified in the name of democracy alone, militant democracy might be more militant than democratic at least when it comes to secession. And eternity clauses might be about naked power no less than constitutional values. Although democracies are typically contrasted with authoritarian regimes, they do share some of their draconian tools when it comes to crushing secession. Even more telling is the fact that consociationalist or consensus democracies ban secession as much or even more than majoritarian democracies. Despite previous assumptions, secession might require revolution rather than mere constitutional amendment. Though we are accustomed to treating popular sovereignty as composed of population alone, constitutions’ treatment of secession may reveal that it is a territorial concept. This, along with strategic considerations shared also by international law, may explain the prevalence of absolute constitutional prohibitions on secession. The Texas v. White decision held that the state is composed of three major elements—people, territory, and constitutional arrangements.366 Secession reveals that when there is a combined challenge of withdrawal of both citizens and territory, it may require a reworking of the rump state’s third component: its constitutional arrangements. Secession may amount to a redefinition of “We the territorial People.”