1

Amendment Power, Constituent Power, and Popular Sovereignty

Linking Unamendability and Amendment Procedures

YANIV ROZNAI^*

THE THEORY OF constitutional unamendability identifies a fundamental distinction between primary constituent (constitution-making) power and secondary constituent (constitution-amending) power. The latter is limited by unamendability and the former—perceived as the people’s democratic constitution-making power—is unlimited by unamendability. This chapter develops the distinction by supplementing it with a further one, between various shades of secondary constituent powers along a ‘spectrum’; a theoretical construct that links constitutional amendment procedures and limitations which ought to be imposed upon constitutional amendment powers. According to this spectrum theory, constitutional systems are polymorphic: the more similar the democratic characteristics of the amendment powers are to those of the primary constituent power, the less it should be bound by limitations; and vice versa: the closer it is to a regular legislative power, the more it should be fully bound by limitations. This examination is an important step towards a better understanding of the nature and scope of constitutional amendment powers.

I.INTRODUCTION

In earlier writings, I have claimed that constitutional amendment powers are increasingly conceived as limited (explicitly or implicitly) and that this unamendability rests on the theoretical distinction between primary and secondary constituent powers.¹ This chapter makes an additional move forward towards a theory of constitutional unamendability. The first section summarises the presupposition that the people are free to change even unamendable elements of the Constitution; yet this power resides not in the amendment power but in the exercise of the primary constituent power, a concept strongly linked to popular sovereignty. The second section describes and analyses the current trend of prescribing a constitutional process for exercising primary constituent power. This method, it is argued, is a fallacy. Following the two previous sections, the third section sketches a spectrum of constitutional amendment powers. It considers whether the amendment power is equally limited in jurisdictions where the amendment process attempts to imitate the re-emergence of the primary constituent power, by incorporating elements such as referendums and constitutional conventions, as in jurisdictions where it is more similar to regular legislative power. It argues that the more similar the characteristics of the secondary constituent power are to those of a democratic primary constituent power (‘demanding amendment power’)—that is, inclusive, deliberative, and time-consuming—the less it should be bound by limitations, and vice versa. This examination links limitations upon amendment powers and the amendments’ procedures. The framework which contextualises this chapter’s approach is both comparative by seeking to identify the character of actual existing constitutional arrangements of constitutional amendments, and theoretical by offering a characteristic explanation of these arrangements and evaluating how they work against their own internal logic.²

II.UNAMENDABILITY AND CONSTITUENT POWER

A.A Three-Track Democracy in a Nutshell

Once a constitution is constituted and the primary constituent power has accomplished its task, a constitutional organ—the amendment authority—is granted with the legal competence of revising the Constitution.³ What happens to the constituent power? Since constitutions are embedded within the idea of populism—the liberty of people to shape and reshape their society,⁴ the presupposition is that the people always retain the power to establish and change their constitutional order.⁵ As Carl Friedrich notes, ‘no matter how elaborate the provisions for an amending power may be, they must never … be assumed to have superseded the constituent power’.⁶ Therefore, constituted organs, including the amendment process, do not consume the primary constituent power which is neither exhausted nor bound by the existing constitutional limitations—including explicit or implicit unamendability.⁷ Put differently, the secondary constituent power is limited by unamendability. But as a delegated authority, it can be set aside, just as it can be created. The authorising primary constituent power thus remains in the constitutional background and can re-emerge to take its role.⁸

Unamendability thus limits the delegated amendment power but it cannot block the primary constituent power—the ‘sovereignty at the back of the Constitution’—from its ability to amend even the basic principles or structure of the constitutional order.⁹ Consequently, what unamendability means is that certain amendments establishing a ‘new constitution’ or a new ‘constitutional identity’ cannot be achieved through the regular amendment procedure but require a different constituent process.¹⁰ Unamendability should therefore not be viewed as blocking all the democratic avenues, but rather merely as proclaiming that one such avenue—the amendment process—is unavailable. In order to legitimately achieve the sought after constitutional change, other procedures ought to be used.¹¹ The power to change unamendable principles does not reside within the constitutional amendment procedure rather it is appropriately part of the primary constituent power of the sovereign people, from which all legitimate authority springs.¹²

And so, within a constitutional democracy, one has to recognise three kinds of powers: primary constituent power, secondary constituent power (amendment power), and constituted power (eg legislative power).¹³ The legislative track is the sub-constitutional track of political life in which the legislature and executive bodies enact, enforce, and implement political decisions through ordinary legislation. The amendment track is the track of constitutional politics, through which the bodies entrusted with the authority to amend the Constitution, may enact, add, annul, or amend constitutional provisions. This track usually involves different—and separate—processes and organs apart from the legislative track. Finally, the primary constituent track is the third track of democracy which is not part of the ordinary constitutional politics. It is unbound by prior constitutional rules and may even create a new constitutional order. Hence, primary and secondary constituent powers are related but distinct powers, as the latter is a power established in the Constitution, which is bound by explicit and implicit limitations.¹⁴ The Constitution cannot restrict the primary constituent power, which resides outside of it and can ‘exercise its authority de novo’.¹⁵

B.Primary Constituent Power and Popular Sovereignty

The understanding of the people’s primary constituent power to constitute and re-constitute the Constitution is compatible with commitment to democratic self-government. The people conceive themselves as a single sovereign in order to attribute the Constitution to a single act of will.¹⁶ However, the very idea of the people as holders of the primary constituent power is perplexing and has given rise to heated debates within constitutional theory.¹⁷ The idea of a ‘power of the people’ acting in an identifiable, comprehensible, and unmediated way in order to constitute for themselves a constitution is a kind of a myth, a fiction.¹⁸ Even if ‘the people’ are recognised as holders of the primary constituent power, the people do not necessarily take part in the constitution-making process,¹⁹ and with old constitutions there is no ‘we the people’ as the constitution-makers are long gone.²⁰ Since ‘the people’ simply cannot speak directly as a whole, the will that we attribute to ‘the people’ ought to be revealed through some kind of representation and deliberation processes.²¹

For the theory of unamendability, regardless of how historically accurate is the story we tell ourselves about ‘the people’ as constitution-makers, ‘the very plenipotentiary scope of the people as the normative constitutional author provides the limits on the normal amending apparatus’.²² I do not elaborate here on the various complexes of the constituent power concept, an exercise which deserves a separate treatment of its own.²³ I only briefly point to the relationship between constituent power and popular sovereignty which is necessary for the further development of the chapter.

‘The dirty little secret’ of contemporary jurisprudence, as Roberto Unger describes it, is the discomfort with democracy and fear of popular action.²⁴ Popular sovereignty’s understanding of the primary constituent power challenges this discomfort as it allows the people’s primary constituent power to re-emerge in ‘extraordinary moments, [when] politics opens up to make room for conscious popular participation and extra-institutional, spontaneous, collective intervention’.²⁵ Constituent power, Lior Barshack writes, belongs to the transcendent and absent corporate body. ‘The people’, as the sovereign in a democracy, remains a transcendent entity, which appears only during special constitutional moments such as declarations of independence, revolutions, and referenda. In these instances of communitas, the communal body is dissolved of structural boundaries as it descends back into the social; the transcendence is collapsed into the immediate presence.²⁶ These extraordinary moments, which carry a temporary character, define the corporate group and within them all generations are present.²⁷ Importantly, for Barshack, ‘it is not the democratic principle of popular sovereignty that places popular will above constitutional procedure, but the fact of sovereign incarnation and the concomitant relaxation of all principles. The communal body wields supreme legislative power whenever it is enacted, in democratic as well as non-democratic contexts…’.²⁸ It is the suspension of political parties and a high degree of communal body’s involvement that allows sovereignty to step forward in constitutional moments such as referendums.²⁹

In contrast with this transcendental conception, according to Akhil Amar, through Article V, the people, who are the source of the Constitution, delegated the amendment power to ordinary government, without limiting themselves. Article V thus supplements but does not replace popular sovereignty. It is non-exclusive as the immanent people retain their reserved and inalienable right to revise the Constitution themselves, by a majority of voters, via referendum or special constituent convention, even outside of the amendment process.³⁰

These conceptions resemble one another in one important sense: in Barshack’s conception of a corporate body, the transcendence collapses into the immanent present in extraordinary temporary constitutional moments. Likewise, Amar’s immanent conception of popular sovereignty is manifested in special constitutional moments. It is therefore not necessary, for the sake of the argument advanced here, to choose between these two conceptions. Yet, both accounts correspond with Sheldon Wolin’s conception of democracy as a political practice that involves the manifestation of popular sovereignty. Democracy, Wolin advocates, is not a ‘form’, but rather a representative moment in the nation’s life that dictates the substance of the Constitution. In these rare and episodic ‘moments of commonality … through public deliberations, collective power is used to promote or protect the well-being of the collectivity … power returns to “the Community” and agency to “the People”’.³¹ If we believe that the source of political authority rests with the people; that ‘the ability to engage in constitutional change is a fundamental act of popular sovereignty’,³² we also need to bridge the gap between the imaginary people and the real people. ‘Constitutionalism’, Carl Friedrich writes, ‘and more especially democracy, presupposes an active group of citizens who are ready to assume responsibility and become the “constituent power”’.³³ Therefore, recent scholarship has called for the development of democratic tools in order to ‘return the epicenter of sovereignty to the people’, especially in constitutional decisions.³⁴ Ali Riza Coban calls for the design of legal arrangements that would ensure a maximum level of the people’s democratic participation during constitution-making.³⁵ Likewise, Joel Colón-Ríos argues that significant constitutional changes require a process that endeavours to reproduce and maximise popular participation and democratic openness at the same degree presented during constitution-making moments.³⁶ Therefore, he claims, in order to acquire democratic legitimacy, fundamental constitutional changes, episodical by their nature, should take place through the most participatory process possible, which allows citizens the opportunity to propose, deliberate, and decide upon such changes. Participatory mechanisms such as elected constituent assemblies whose outcomes shall be approved in a popular referendum facilitate the exercise of constituent power.³⁷ This understanding allows the mythical people to re-activate its constituent power.³⁸

Indeed, the modern conception of primary constituent power is strongly associated with the notion of popular sovereignty and the recent proliferation of referendums is an indicator of the broader trend towards engaging the people themselves in constitutional matters.³⁹ According to studies of late, among current existing constitutions more than 40 per cent were publicly ratified by referendums and many others involved different forms of popular participation in the constitution-making process.⁴⁰ As Cheryl Saunders notes, ‘popular participation is a distinctive feature of constitution-making in the 21st century’.⁴¹

There are many familiar difficulties associated with popular mechanism such as referendums: the designation of the individuals who qualify to participate; the drafting of the ballot question; the lack of knowledge of the voters; fear of tyranny of the majority; and the historical associations of abuse of plebiscites.⁴² Moreover, referendums do not necessarily truly express the people’s will due to political elites or interest groups’ manipulations of the people or control of the political agenda which cast doubts on the ability of referendums to allow a genuine exercise of deliberative democracy.⁴³ Therefore, some, such as Antoni Ninet, claim that: ‘the legitimacy and validity of the Constitution requires not only popular ratification, but also real (or true) democratic involvement. A constitution made through ordinary parliaments and representatives is unacceptable’.⁴⁴ In other words, popular participation in ‘constitutional moments’ should be throughout the constitutional norms-creating process and not be limited to a solely ‘yes’ or ‘no’ vote in a referendum.⁴⁵ This understanding refuses to reduce primary constituent power to a mere acclamation—a ‘soccer-stadium democracy’, in the words of Holmes.⁴⁶ This resembles Christoph Burchard’s criticism of Schmitt’s conception of constituent power for ‘there is no discourse, no rational consideration, only irrational masses cheering or booing’.⁴⁷ Process matters.⁴⁸ It is the manifestation of ‘we the people’, not simply ‘oui the people’.

A democratic primary constituent power must be committed to popular sovereignty which may exercise itself in forms such as special constitutional assemblies and constitutional referenda.⁴⁹ These forms must have a special character, that is, separate from other public functions, thereby replacing revolution with peaceful means incorporating actual, deliberate, free choice by society’s members.⁵⁰ While it is true that ‘in the end, there can be no precise algorithm specifying the conditions for defining a people capable of exercising constituent authority’,⁵¹ an exercise of primary constituent power should be inclusive, participatory, and deliberative.⁵² After all, the word constituere, Kalyvas reminds us, marks the act of founding together, jointly.⁵³

III.THE CONSTITUTIONALISATION OF PRIMARY CONSTITUENT POWER

A.The Fallacy of Prescribed Constitution-Making Procedures

In the previous section, I noted that even after the establishment of a constitution, the primary constituent power, the ‘sleeping giant’ so to speak,⁵⁴ can wake-up and be exercised. Being external to the pre-existing constitutional order, this power cannot be bound by the prior and existing constitutional rules. Consequently, constitutions cannot (and most do not) regulate its emergence. Constitutions, in other words, ‘contemplate their amendment but almost never their replacement’.⁵⁵ Nonetheless, there is often a desire to institutionalise the primary constituent power within the constitutional form and some constitutions attempt to regulate the re-emergence of the primary constituent power.⁵⁶ For instance, the Constitution of Venezuela of 1999 states that ‘The original constituent power rests with the people of Venezuela. This power may be exercised by calling a National Constituent Assembly for the purpose of transforming the State, creating a new juridical order and drawing up a new Constitution’.⁵⁷ Other constitutions prescribe a procedure for their ‘total reform’, ‘complete revision’, or ‘adopting a new constitution’.⁵⁸

Since by its nature the primary constituent power is unbound by any constitutional rules, this method, as Schmitt claimed, seems to be a fallacy: ‘no constitutional law, not even a constitution, can confer a constitution-making power and prescribe the form of its initiation’.⁵⁹ If constituent power is considered superior and external to positive law, law cannot prescribe it. Therefore, Richard Kay claims that these attempts are not only ‘paradoxical’, but might also ‘be dismissed as rhetorical decoration’.⁶⁰ Instead of their dismissal, it might be more valuable to regard these provisions not as constituting, but rather as recognising or declaring, existing powers; hence different from amending provisions.

If a new constitution is constituted through these mechanisms, what are the implications for constitutional theory? Two plausible answers; either this process was an exercise of the instituted secondary constituent power,⁶¹ or it was indeed the primary constituent power that played the constitution-making role, yet it simply decided to act according to the existing procedures rather than being obliged by them.⁶² Carl Friedrich gives the example of the constitutional reform that occurred in Switzerland in 1874 when the Constitution of 1848 was ‘entirely overhauled and democratized’ through the ordinary amendment procedure. Constituent power, according to Friedrich:

manifested itself through the amending power; but that does not mean that it is identical with it; in fact even to say that it manifested itself through it is something of a misstatement. It would be more accurate to say that the group which might otherwise develop into the constituent power manifests itself through, acts through the amending power.⁶³

Since the primary constituent power has extra-juridical dimensions, it cannot be fully regulated or stipulated legally.⁶⁴ ‘Like it or not’, Kay correctly notes, ‘a true constituent authority must act without the comfort of legal authorization’.⁶⁵ This, however, does not mean that a constitution cannot stipulate the means by which a new constitution would be constituted. It only means that by its nature, primary constituent power does not have to abide by it, although it can act accordingly if it so wishes.⁶⁶ These mechanisms can be viewed not as containing primary constituent power, but rather, simply as vehicles for its exercise.⁶⁷

Not to be mistaken, these mechanisms may carry benefits. Exercising primary constituent power behind ‘a façade of legality’ may serve significant political interests,⁶⁸ and may enhance legal certainty, continuity, and the legitimacy of the new document.⁶⁹ Therefore, the Venice Commission ‘strongly endorse[d]’ the use of a legal procedure even for the adoption of ‘new constitutions’, as such a procedure would ‘strengthen the stability, legality and legitimacy of the new system’.⁷⁰ Finally, regulating a relatively demanding constitutional–replacement process, would obviate abusing the amendment process or ignoring any constitutional procedures or guaranties whatsoever in the name of pouvoir constituant.⁷¹ Since the people’s will is divided, such legal rules may create ‘organized, complex procedures of deliberation and voting’, which would preserve the primary constituent power’s credibility.⁷²

B.We The ‘Limited’ People?

The German Basic Law of 1949 is an interesting example of an attempt to legally positivise the primary constituent power.⁷³ The final article of the Basic Law anticipated the Basic Law’s own destruction: ‘This Basic Law will lose its validity on the effective date of a constitution that has been chosen by the German people in a free decision’.⁷⁴ Contrary to some of the examples noted earlier, this provision—while recognising the people’s constituent power—did not stipulate conditions or procedures for its exercise, which might be seen as a confirmation of its extra-legal character.⁷⁵ This particular constitutionalisation of primary constituent power raises interesting questions as to the theory of unamendability.

Famously, the German Basic Law includes in Article 79(3) an unamendable provision which prohibits amendments affecting the division of the Federation into Länder, human dignity, the constitutional order, or basic institutional principles describing Germany as a democratic and social federal state.⁷⁶ Is the emergence of a new primary constituent power, constitutionalised by Article 146, restricted by Article 79(3)? Some claim that the unamendable principles would bind future constitution making while others remark that Article 146 is a legal manner with which to overcome the unamendable provision.⁷⁷ In the Lisbon Case, the Federal Constitutional Court expressly left this question unanswered:

It may remain open whether, due to the universal nature of dignity, freedom and equality alone, this commitment even applies to the constituent power, i.e. to the case that the German people, in free self-determination, but in a continuity of legality to the rule of the Basic Law, gives itself a new constitution.⁷⁸

I agree with Jo Murkens that Article 79(3) addresses only the amendment power, whereas Article 146 foresees a new constitution adopted by the primary constituent power, which by its nature cannot be bound by the rules of the prior constitution. The new constitution-drafter may take Article 79(3) into account, but that would depend on its own ‘goodwill’, rather than on the nature of the unamendable provision as a legal obligation.⁷⁹ Article 79(3) guarantees are unamendable, not eternal (as they are often wrongfully referred to).⁸⁰ Consequently, although the primary constituent power is constitutionalised within the German Basic Law, Article 79(3) is unable to bind later generations when exercising their primary constituent power.⁸¹

One often noted example to the claim that the exercise of constituent power is hardly bound by limitations, is the 1962 Amendment to the French Constitution over the form of presidential elections which passed through a referendum initiated by President de Gaulle despite its violation of the amendment procedure.⁸² The amendment was challenged before the French Constitutional Council, which held that it had no competence to review laws passed by the people in a referendum since they are a direct expression of national sovereignty.⁸³ This reflects the idea of le peuple-roi—‘the people’ is the new sovereign king who always retains the power to revise the Constitution. As one advocate stated in 1849 before a Versailles court, ‘the people never violate the constitution’.⁸⁴ It is also compatible with Barshack’s supposition that ‘the fuller the sovereign presence, the more relaxed the constitutional structure and the formal procedure that governs the referendum’.⁸⁵ Since this extra-constitutional change is to be regarded as a ‘constitutional violation’,⁸⁶ its authoritative legitimacy could only be granted retrospectively.⁸⁷

This postulation raises a thorny question: what are the implications of an amendment process which includes the primary constituent power’s characteristics of directness and speciality, like popular referendums or elections for special constituent assemblies? Here the people are part of the instituted amendment power. Are the people in that capacity limited? A positive answer would subordinate not only decisions of the people’s representatives, but also those of the people themselves to the judiciary. It seems that there are two prevailing approaches to this challenge.⁸⁸

According to the first approach, when the amendment power is exercised by the people it is unlimited. This is the general approach of Irish jurisprudence. Due to the Christian character of the 1937 Constitution, there is a heated debate within Ireland on whether natural law sets limits to the amendment power.⁸⁹ In its jurisprudence, the Irish Supreme Court rejected the claim that natural law is superior to the Constitution, holding that the people, not God, are the creator of the Constitution and the supreme authority. Hence, amendments made by the people in a referendum which expresses their will become the fundamental and supreme law of the land, and cannot be reviewed or nullify by courts.⁹⁰ According to this approach, even if the primary constituent power rests with the people, albeit essentially anarchic and lawless, it may choose, so to speak, to be exercised within the constitutional framework of constitutional amendment.

According to the second approach, the amendment power is limited even when exercised directly by the people. The people in that capacity of inclusion in the amendment process represent a legal organ of the state. Since there can be no sovereign within the constitutional order, the people’s power is necessarily limited.⁹¹

An example of this approach comes from Switzerland, where 100,000 people eligible to vote have the right to propose constitutional revisions (Volksinitiative). Following a Volksinitiative, according to which asylum seekers who enter the state unlawfully would be deported immediately and without the option of appeal, the Federal Council proposed that the Federal Assembly invalidate the Volksinitiative, which it did accordingly on 14 March 1996. According to the Federal Council, respect for international law fundamental norms, which include the peremptory norm of non-refoulement, is inherent to the Rechtsstaat principle of ‘rule by law’.⁹² In 1999, Switzerland granted explicit constitutional recognition to the proposition that international law jus cogens norms limit constitutional reforms, whether total or partial.⁹³ Therefore, in Switzerland, even when the people are directly involved and the Constitution allows for its total revision, this faculty is still limited. According to this approach, ‘the people’ may be regarded in two distinct capacities: as a source of absolute power (primary constituent power) and as a constitutional organ established by the Constitution for its amendment (secondary constituent power).⁹⁴

Stephen Tierney’s analysis appears lucid. Tierney claims that one has to distinguish between referendums that operate wholly within existing constitutional structures, thereby internal to the normative constitutional order, and referendums that transcend the existing order, which are external to the Constitution and include the power to ‘bring about a new order’.⁹⁵ Tierney urges us to be cautious when referring to ‘people sovereignty’ simply due to an exercise of a referendum, and to pay attention to the mode of that exercise and the role that the people have played within it.⁹⁶ On this account, when the people have a role within the amendment process, such an exercise is ‘constrained to operate within mainstream representative democracy, subordinate to the constitutional rules and subject to constitutional institutions, including courts’.⁹⁷ It is only when the people ‘act as original constitutional authors, bringing a clear break in the old order; the referendum manifests the “people’s” direct democratic capacity to act as the supreme source of constitutional law in foundational constitutional acts’.⁹⁸ Within this dilemma, Claude Klein remarks, lies ‘the crux of the problem of the theory of the amending power’.⁹⁹ In the next section, I deviate from the dichotomy of the two approaches mentioned above and offer a more subtle account of a spectrum of constitutional amendment powers.

IV.THE SPECTRUM OF CONSTITUTIONAL AMENDMENT POWERS

According to the dichotomy described above, there is a binary constitutional code of constitutional amendments: an amendment originating via the constitutional process through the secondary constituent power, which is limited (even if includes the people), and an amendment that is constituted in a constitutional moment through the re-emergence of the primary constituent power and thus unlimited. Constitutional systems are more complex than this. They are polymorphic. Just as in materials science, a solid material can exist in multiple forms, so does the amendment power. Not only do constitutions have different procedures, mechanisms, and actors involved in constitutional amendments, but also a same constitution might incorporate different procedures for amending different provisions and principles. This section argues that constitutional amendment powers ought to be regarded not in a binary manner (limited/unlimited), but rather as a spectrum of scope. The more similar the characteristics of the secondary constituent power are to those of a democratic primary constituent power (‘demanding amendment power’), the less it should be bound by limitations, and vice versa. The closer it is to a regular legislative power (‘facile amendment power’), the more it should be bound by limitations and judicial scrutiny. This calls for an examination of the link between amendment procedures and the limitations that ought to be imposed upon amendment powers and their judicial enforcement.

A.Demanding and Facile Amendment Powers

There is no single unified method or process for amending constitutions. Constitutions differ between dissimilar degrees of amendability. Some are ‘flexible’ in that the amendment process is relatively easy, such as ordinary legislative majorities, and some are more ‘rigid’ in that they require high barriers, such as super-majority threshold in parliament, a higher quorum than ordinary legislation, time delays, state ratification in federal systems, or additional requirements such as constituent assemblies, intervening elections and popular referendums.¹⁰⁰ These hurdles not only make constitutional change more difficult than that of ordinary law, but also reflect the notion that the Constitution is a special kind of law.¹⁰¹

As described in the first part IIA of this chapter, the amendment power is situated in a grey area between the ordinary legislative power (constituted power) and the extraordinary constituent power.¹⁰² In this grey area, a spectrum of amendment power exists. Some amendment procedures are facile amendment powers in the sense that the amendment process is similar (or relatively similar) to the ordinary legislative process in terms of the organs involved, and the temporal and procedural constraints. Others, which significantly deviate from the ordinary legislative process with regard to these features are demanding amendment powers in the sense that their exercise resembles (or almost resembles) a constitutional moment—nearly an invocation of the primary constituent power.

The most facile amendment powers are those in which a simple legislative majority is enough to bring about constitutional amendments.¹⁰³ An amendment power more demanding than the ordinary majority is the one that requires a qualified majority in Parliament for the adoption of amendments. Supermajority requirements expand the range of interests to be taken into account and promote careful considerations, deliberation and persuasion.¹⁰⁴ Requirement of multiple readings in Parliament, time delays between the initiative and the first debate in Parliament or between the readings are additional techniques for making the amendment process longer and more difficult thereby making the amendment power more ‘demanding’.¹⁰⁵ But all of these procedures are still ‘facile’ from the perspective of a democratic constituent power in the sense that they exclude the people from the process.

These procedures of ordinary and qualified majorities in Parliament may be reinforced, for instance with a requirement of a popular referendum, intervening elections, convening a special constitutional convention or a combination thereof.¹⁰⁶ When constitutions require intervening elections for amendments, this is a demanding amendment power provided that the subject matter is put as a central issue in the elections’ agenda thereby asking the people’s opinion of it, and allowing enough opportunity for public and political debates on the proposed amendments. It is inclusive, deliberative, and attempts to minimise abuses of the amendment power—as the amending authority would not necessarily be the same before and after the amendment’s enactment. Demanding amendment powers are also those amendment procedures that require elections to a special constituent assembly. This process is inclusive of the people (through the elections, at the very least) and supplies a deliberative setting.¹⁰⁷ The US Constitution is an interesting case. On the one hand, comparatively, Article V is unusually onerous as its hurdles are very demanding and time-consuming.¹⁰⁸ These consensus requirements arguably label it as a demanding amendment power.¹⁰⁹ On the other hand, this procedure is so demanding that some commentators note that ‘from the perspective of … the constituent power … the banishing of sovereignty from the internal life of the republic was perhaps too successful. … it is almost impossible to legally change the American constitution’,¹¹⁰ but this cumbersome procedure does not directly involve the people.¹¹¹

The argument I advance is simple: demanding amendment powers should be awarded wider scope than facile amendment powers. This idea is compatible with existing constitutional arrangements, which utilise a ‘constitutional escalator’. As noted earlier, some constitutions incorporate different procedures for amending different constitutional subjects.¹¹² These procedures may include, again, increased qualified majorities, referendums, intervening elections, or convening special constituent assemblies. To mention some examples, certain principles may only be amended by a referendum in Belarus (1994, Article 140), Estonia (1992, Article 162), Latvia (1992, Article 77), Lithuania (1992, Article 148), Singapore (1963, Article 5), Serbia (2006, Article 203), and Vanuatu (1980, Article 86). A different procedure exists in the Russian Constitution in which amendments to fundamentals of the constitutional system require conveyance of a Constitutional Assembly (1993, Article 135).¹¹³ This is also often the case when the Constitution allows for a ‘total revision’ of the Constitution or its replacement with a new one. For example, the Austrian Constitution requires a popular referendum for its total revision (1920, Article 44[3]).¹¹⁴ In Spain, a total revision of the Constitution or amendments to certain basic provisions demand a more robust process, including the dissolution of Parliament and a subsequent approval by a referendum (Article 168(1)).¹¹⁵ In Costa Rica, a general amendment of the Constitution can be effected only by a constituent assembly (1949, Article 196), and in Bolivia, the total reform of the Constitution or that which affects its fundamental premises, can take place through a constituent assembly, put into motion by a popular referendum (2009, Article 411(1)). In line with my argument, except with the Russian Constitution, none of the above-mentioned constitutions includes explicit unamendable provisions.¹¹⁶

The rationale behind this ‘constitutional escalator’ is clear; those provisions that are deemed more fundamental or protection-worthy enjoy a special protection from hasty changes through heightened requirements.¹¹⁷ This ‘selective rigidity’ mechanism was recently advocated for by constitutional scholars, such as Richard Albert and David Landau, for allowing greater protection to the core parts of the democratic order, thus reducing the possibilities of abusing the amendment process, while simultaneously allowing relatively easily amendments of non-fundamental principles. The two scholars also emphasise the importance of the temporal dimension; that is, intervening elections or time delays before the adoption of constitutional amendments, which resist the ability of powerful political forces to abuse their power and take advantage of their temporary popularity to amend the Constitution in a way that would damage democracy.¹¹⁸ From the perspective of constituent power theory, the ‘constitutional escalator’ is not only a practical safeguard for better protecting certain constitutional principles or institutions but means for generating legitimacy, as I elaborate in the next section.¹¹⁹

B.Linking Amendment Procedure and Unamendability

Demanding amendment processes are devices that aim to imitate constitutional moments in which the primary constituent power is incarnated. Through formal mechanisms, such as referendums and summoning constituent assemblies, they aim to create an environment in which the people are ‘awaking’, in a sense, to resume their role as constituent authors.¹²⁰ As Xenophon Contiades and Alkemene Fotiadou explain:

The people are traditionally considered to have spoken during the exercise of the pouvoir constituant. Amending formulas may be described as replications of the constitutional moment where the pouvoir constituant was exercised, being attempted simulations of that primordial, constitution-making function.… This original constitution-making process is embellished with great symbolic force, the reproduction of which during every constitutional revision would be unfeasible. Yet, desire to somehow preserve the spirit of that moment is often apparent in constitutional arrangements that risk sacrificing practically for symbolism.¹²¹

Since primary constituent power is unlimited by constitutional limitations, a process which imitates the re-emergence of primary constituent power should not be considered as equally bound to restriction as facile constituent power, solely activated by one constitutional organ in the same procedure as in ordinary politics. This is because, as Barshack clarifies:

Theoretically inelegant as this result may be, the binding force of constitutional procedure varies in every constitutional moment in proportion to the intensity of sovereign presence.… When the communal body asserts itself in the amendment of a constitution as intensely as it was involved in its original adoption, it is hardly bound by constitutional procedure at all and hardly subject to judicial review over the constitutionality of the amendment.¹²²

Therefore, from the perspective of constituent power, it is not merely the protection of principles or the fear of abuse which justify ‘constitutional escalator’, that is, allowing the more fundamental principles to be amended through more heightened procedures, it is also the notion that ‘the more exuberant the sovereign presence, the less bound is the collective body by […] the non-amendability of certain constitutional principles…’.¹²³

Different constitutional procedures can aim to create fuller or hollower bodies of ‘popular sovereignty’. It is in this way that we may understand the unique requirement in the Lithuanian Constitution that more than three-fourths of the electorate must participate in a referendum for amending Article 1 of the Constitution, according to which ‘Lithuania is an independent democratic republic’.¹²⁴ The spectrum of amendment powers links the amendment process and limitability of the amendment power; the more demanding the amendment power, the less limited it should be.

Climbing along the escalator of amendment powers depends upon the various procedures included in the amendment process, which enable inclusive, participatory and deliberative involvement of the people. Yet, the primary constituent power itself is unbound by any constitutional procedures. Thus, at a certain point, with the appearance of the ‘genuine’ primary constituent power the ‘constitutional escalator’ becomes a ‘constitutional roller-coater’ so to say. This suggests that to some extent demanding amendment powers might paradoxically be more difficult to exercise than primary constituent power.¹²⁵ The normative requirement that in order to acquire a wide scope of action, secondary constituent powers must be demanding, in the sense of their inclusiveness, participatory and deliberative processes, surely implies the assumption that a legitimate exercise of primary constituent power should indeed conform to such requirements.

This theoretical construct of a spectrum may be supported by two cumulative rationales.

The first rationale is normative. Owing to the democratic nature of the primary constituent power, amendments that are enacted through demanding amendment powers carry greater legitimacy. An author-based theory of legitimacy considers constitution as ‘respect-worthy’ in light of their maker, and the people are the most ‘legitimate’ authors of democratic constitutions. True, when the people are involved in the amendment procedure, they act in their capacity as an institutional organ. Nonetheless, when the people are involved they act as a ‘legitimation elevator’ of the constitutional change.¹²⁶ For example, referendums maximise legitimacy,¹²⁷ and special assemblies carry a higher degree of ‘popular legitimacy’ than ordinary legislatures.¹²⁸ A notable example is the South-African exceedingly participatory constitution-making process which enjoyed a high level of legitimacy.¹²⁹

Importantly, the binding power of constitutions does not rest solely on ‘procedural legitimacy’, but on other substantive factors as well. However, if the people’s involvement indicates that the current generation accepts the constitutional framework and that the Constitution reflects its values, there is a greater claim for the Constitution’s ‘democratic legitimacy’.¹³⁰ I am not claiming that a democracy cannot function without strong popular involvement,¹³¹ rather that since ‘inclusiveness is the contemporary mechanism for ensuring that a constitution actually is an exercise of the constituent power’,¹³² then participation and inclusiveness of the people during exceptional moments of constitutional change increases the Constitution’s democratic legitimacy.

One might argue that this theory has an inner contradiction. On the one hand, it is argued that whether a primary constituent power is a genuinely ‘sovereign’ exercise can, to a large extent, only be judged retrospectively, that is, if a vast majority of the people accepts the exercise of power as legitimate, then one might argue that the state experienced a ‘constitutional moment’. On the other hand, it is claimed that the presence of demanding participatory and deliberative ex-ante procedures that include the people will increase the likelihood of social legitimacy. The alleged problem with characterising amendment powers as ex-ante facile or demanding (based on the constitutional text), is that it overlooks the backward-looking view of whether the people conceived of the exercise of power as legitimate. In other words, if the legitimacy of the primary constituent power is determined on an ex-post basis, can the position of types of constitutional amendments along the spectrum be determined on an ex-ante basis? The response to this claim is based of course on the distinctions between the two types of powers. A decision whether there was a legitimate exercise of primary constituent power must be judged ex-post facto since this is most commonly the result of illegality or extra-constitutionality (such as amending the Constitution outside the amendment procedure or amending unamendable provisions).¹³³ This act cannot be judged upon standards of legality hence the need for legitimacy claims. Actions by the secondary constituent powers, in contrast, already carry a certain legal basis since they are brought about through the constitutional amendment procedures. There is a lesser need for ex-post considerations when one can infuse, so to say, within the constitutional mechanisms, procedures which would assist in elevating the democratic legitimacy of the amending exercise ex-ante.

The second rationale is a practical one. Amendment procedures aim, inter alia, to provide mechanisms for deliberations and to subject the Constitution and proposed amendments to critical scrutiny.¹³⁴ Presumably, the more deliberative, multi-institutional and prolonged the processes of amendments are, the less the likelihood of abuse of the amendment power.¹³⁵ This echoes Jon Elster’s argument that special constituent assemblies should make constitutions, rather than ordinary legislatures, since the latter are more likely to be influenced by group and institutional self-interests while due to the irregularity of the former, they are presumed to be impartial bodies insulated from short-term political bargaining.¹³⁶ Interestingly, a similar argument has been made with regard to direct democracy.¹³⁷ A referendum device may provide a safeguard against an executive which controls the Parliament and can abuse the amendment power for its short-term political interests.¹³⁸ In other words, the referendum plays a role both as fostering a wide civic participation and assisting the check over the executive power in constitutional change.¹³⁹

Some argue that exceptional popular mechanisms, especially in the absence of ordinarily political institutions, have actually aided charismatic leaders to impose authoritarian constitutions.¹⁴⁰ Others point out that popular inclusiveness in a ‘we the majority’ form risks minority rights,¹⁴¹ which might even justify robust judicial review of direct democracy.¹⁴² In contrast, others claim that ‘direct democracy and the protection of minorities are not mutually exclusive’.¹⁴³ The question is not whether mechanism of direct democracy risk minority rights, but whether such mechanisms increase or reduce such risks compared to other representative mechanisms, but this remains indecisive.¹⁴⁴

Importantly, this chapter does not engage with ordinary legislative referendums, but with constitutional referendums taking place in exceptional moments. The two types of referendums are different.¹⁴⁵ Moreover, a stand-alone referendum does not engender demanding amendment power, as I have described it. Since a demanding amendment process is meant to be an inclusive, deliberative, and time-consuming, the referendum should be an additional platform to the political process, perhaps together with a special constituent assembly. The risk of abuse of the amendment power arises especially with facile amendment powers, where the amendment body is the same body that decides the everyday political decisions. The identity of bodies causes the mingling of longer-range issues of constitutional planning with short-term interests of political power.¹⁴⁶ Blount, Elkins, and Ginsburg remark that ‘higher levels of participation are presumed to function like supermajority rules, restricting the adoption of undesirable institutions and protecting prospective minorities in the democratic processes that are established’.¹⁴⁷ This is not necessarily a matter of difficulty,¹⁴⁸ but of inclusiveness and deliberations. As Christopher Eisgruber notes, ‘by establishing a separate and difficult track for some political issues, the Constitution may focus public attention upon those decisions and improve deliberation about them’.¹⁴⁹ An inclusive and deliberative process aims to improve the quality of the constitution-making/amending outcome.¹⁵⁰

Accordingly, inclusive and deliberative amendment procedures which allow time for public and institutional deliberations reduce the possibility of abuse and enhance the legitimacy of the endorsed constitutional change.¹⁵¹ Such procedures, which imitate the invocation of primary constituent power should be given a greater margin of change: ‘the more significant a change may be, more likely is it that the procedure for affecting it would be more arduous, difficult and demand a higher level of deliberative legitimacy’.¹⁵² Importantly, even demanding amendment powers still act as limited constitutional organs. Yet, this limitability is not one-dimensional but moves along the spectrum of the amendment powers: ‘not all amending routes are equal. The closer amending routes are to the citizenry, the more freedom there is to alter the foundational terms of our political life’.¹⁵³ Amendment processes must be linked to unamendability:

The nature of the amending power as well as the level of deliberative legitimacy which the amending process imbibes in itself jointly [determines] … what is allowed to be changed and what is not through an amendment. So it is not that the limits on amending power are sketched without any reference to the procedure of amendment. Both the nature and procedure of amendments are critically important to truly understand what these limits may be.¹⁵⁴

Considering the idea of a spectrum of amendment powers, William Harris’s constitutional theory carries much force. Similar to my argument, Harris distinguishes between the people’s unlimited constituent power and the instituted amendment authority which is a constitutional agent, holding its power in trust and thus bound by limits.¹⁵⁵ Harris links the scope of constitutional amendments with the ‘wholeness’ of the collective people as the source of the constitutional authority. The more amendments seek to influence people’s rights or large scale revisions, the more necessary it is to seek popular approval through ‘sovereignty-reinforcing’ mechanisms such as special constituent conventions. Put differently, there is a reciprocal relationship between the ‘wholesale’ of the sought change and the ‘wholesale’ of the required sovereignty’s presence, which is identified through the criteria of ‘wholeness and deliberateness’.¹⁵⁶ The spectrum of amendment powers advances Harris’ theory by recognising a systematic scale of amendment powers based upon the proximity to a democratic primary constituent power.

C.The Spectrum of Amendment Powers and Judicial Review of Amendments

The ‘spectrum of amendment powers’ is an analytical description of various amendment procedures. This explanatory theoretical model carries with it normative aspects: First, a constitutional design aspect directed at constitution-makers and urging them to design constitutional amendment-rules in an escalator way. In doing so, the more basic constitutional principles would be amendable in a more participatory process, which is time-consuming, deliberative and inclusive. This is in contrast with the less foundational provisions of the Constitution, which should be amendable relatively easily.

Secondly, it is aimed at the judiciary. The intensity of judicial scrutiny of constitutional amendments should be connected to the amendment process. True, since we are dealing with secondary constituent powers, judicial review of amendments should be based on its compliance with the still-binding constitutional rules laid down explicitly or implicitly.¹⁵⁷ But even once amendment powers are conceived as delegated authorities, holding their power in trust, fiduciaries’ duties are enforced with different degrees of strictness, depending on the nature of the specific relationship under consideration.¹⁵⁸ As I elaborate in this chapter, not all amendment powers are similar. A spectrum of amendment powers exists. Some are facile amendment powers, which resemble ordinary legislative powers, some are demanding amendment powers, which allow for popular participation and deliberation and involve a great deal of time. One can expect (although this is not always the case) that the easier the amendment process, the higher the rate of formal amendments.¹⁵⁹ An extremely facile secondary constituent power, where a dominant executive/legislature may control the amendment process, coupled with short-term political interests and temporary majorities, increase the fear of abuse of the amendment power.¹⁶⁰ It is here where the strictest judicial oversight is required.¹⁶¹ Even Lester Orfield, ordinarily an antagonist of unamendability, states that ‘undoubtedly, where a simple majority is required, it is not an especially serious matter for the courts to supervise closely the amending process both as to procedure and as to substance. But when so large a majority as three-fourth has finally expressed its will in the highest possible form outside of revolution, it becomes perilous for the judiciary to intervene’.¹⁶²

Therefore, the legitimacy of judicial review of amendments depends, at least in part, on the amendment process: ‘the fuller the enactment of sovereignty, the less justiciable the sovereign action’.¹⁶³ The more the amendment is the product of inclusive and deliberative demanding amendment powers, which enjoy a high degree of democratic legitimacy and minimise risks of abuse, the less intensive the judicial review of amendments should be, and vice versa.¹⁶⁴ Of course, this is merely a call for judicial restraint not the elimination of judicial oversight over actions of the amendment power. The Constitution’s substance is at least as important as the procedure by which amendments are enacted, and even demanding amendment powers are still limited constitutional organs.

It has been remarked that jurists ‘must either learn to trust the amending process or repose their faith in non-elected judges’.¹⁶⁵ I claim that these two variables—judicial scrutiny and the amendment process—are mutually reinforcing. Since the spectrum of amendment power is linked to the question of unamendability, the mirror picture of the spectrum of amendment power is the spectrum of intensity of judicial scrutiny and restraint, which should be exercised over constitutional amendments.

V.CONCLUSION

The basic theoretical presupposition behind the theory of unamendability is the distinction between the people’s primary constituent power and the instituted secondary constituent power.¹⁶⁶ In contrast with the primary constituent power, the secondary constituent power is limited by unamendability so that certain constitutional decisions require the re-emergence of the primary constituent power. They force ‘the real sovereign to return from its retirement in the clouds’¹⁶⁷ in extraordinary constitutional moments. Amendment procedures often try to imitate these constitutional moments by creating extraordinary procedures, such as referendums and constituent assemblies, which would distinguish these constitutional politics from ordinary ones.¹⁶⁸ As Contiades and Fotiadou write in this volume:

In recent years, there is a growing tendency to involve the people in constitutional amendment through processes enhancing deliberation and popular participation. This trend envisages constitutional amendment as a participatory process aimed at creating a bridge between constituent and constituted power.¹⁶⁹

These are demanding amendment powers. This chapter argued that the more an amendment process contains inclusive and deliberative democratic mechanisms, the closer it resembles the people’s primary constituent power. Congruently, since primary constituent power is unlimited by constitutional limits, demanding secondary constituent powers, which present a fuller presence of peoples’ sovereignty—while still limited—should be allowed greater latitude of constitutional changes. This postulation finds support in comparative constitutional amendments in the form of ‘escalator amendment procedures’. As the puzzle surrounding constitutional unamendability concerns a deeper conflict between substantive versus procedural approaches to constitutionalism; the former focuses on the Constitution’s fundamental principles and the latter on the Constitution’s procedures,¹⁷⁰ the theory of spectrum of amendment powers links amendment procedures to unamendability thereby harmonising substance and procedure.¹⁷¹

^*An earlier version of this chapter was presented at the IACL-BC Workshop on Comparative Constitutional Amendment at Boston College Law School on 15 May 2015. I would like to thank the organisers of the workshop—Richard Albert, Xenophon Contiades and Alkmene Fotiadou for inviting me to present the paper and the participants for their valuable comments, especially to Zoran Oklopcic for his insightful discussion. I would also like to thank Aharon Barak, Andreas Kalyvas, Conor Gearty, Gábor Halmai, Grégoire Webber, Kim Lane Scheppele, Lawrence Sager and Oran Doyle for helpful conversations and Alexander Somek, Bartosz Marchiniak, Martin Loughlin, Michael Wilkinson, Neil Walker, Richard Kay, Thomas Poole and the anonymous reviewer for useful remarks on earlier drafts. An elaborated argument appears in Yaniv Roznai, Unconstitutional Constitutional Amendments—The Limits of Amendment Powers (Oxford University Press, 2017).

¹See, eg, Y Roznai, Unconstitutional Constitutional Amendments—The Limits of Amendment Powers (Oxford University Press, 2017); Y Roznai and S Yolcu, ‘An Unconstitutional Constitutional Amendment—The Turkish Perspective: A comment on the Turkish Constitutional Court’s headscarf decision’ (2012) 10(1) Int’l J Const L 175; Y Roznai, ‘Unconstitutional Constitutional Amendments—The Migration and Success of a Constitutional Idea’ (2013) 61(3) Am J Comp L 657; Y Roznai, ‘The Migration of the Indian Basic Structure Doctrine’ in M Lokendra (ed), Judicial Activism in India—A Festschrift in Honour of Justice V. R. Krishna Iyer (Universal Law Publishing Co, 2012) 240. For a typology of limitations on constitutional amendment powers see O Doyle, ‘Constraints on Constitutional Amendment Powers’, in this volume.

²See, eg, M Loughlin, ‘Constitutional Theory: A 25th Anniversary Essay’ (2005) 25(2) Oxford Journal of Legal Studies 183, 186; S Tierney, Constitutional Referendums—The Theory and Practice of Republican Deliberation (Oxford University Press, 2012) 2.

³For the classical distinction between constituent and constituted powers see EJ Sieyès, ‘What is the Third Estate?’ in Political Writings (Hackett Publishing Company Inc, 2003) 136. For a contextualisation of Sieyès’ doctrine, see T Pereira, ‘Constituting the Amendment Power: A Framework for Comparative Amendment Law’, in this volume.

⁴RD Parker, ‘“Here, the People Rule”: A Constitutional Populist Manifesto’ (1993) 27 Val U L Rev 531, 583.

⁵J Wilson, Commentaries on the Constitution of the United States of America (J Debrett, J Johnson and JS Jordan, 1792) 38–39; WW Willoughby, An Examination of the Nature of the State—A Study in Political Philosophy (BiblioBazaar, 2009) 215–18; WT Han, ‘Chain Novels and Amendments Outside Article V: A Literally Solution To a Constitutional Conundrum’ (2010) 33 Hamline L Rev 71, 79.

⁶CJ Friedrich, Constitutional Government and Politics—Nature and Development (Harper & Brothers Publishers, 1937) 117. See also T Murphy, ‘Constituent/Constituted Power’ in J Protevi (ed), Edinburgh Dictionary of Continental Philosophy (Yale University Press, 2006) 105.

⁷The US Constitution itself was adopted in violation of the Articles of the Confederation. See generally RS Kay, ‘The Illegality of the Constitution’ (1987) 4 Const Comment 57; B Ackerman and N Katyal, ‘Our Unconventional Founding’ (1995) 62 U Chi L Rev 475; J Frank, ‘“Unauthorized Propositions” The Federalist Papers and Constituent Power’ (2007) 37(2–3) Diacritics 103.

⁸On the amendment power as a delegated authority see Roznai, Unconstitutional Constitutional Amendments (n 1); AR Amar, ‘Philadelphia Revisited: Amending the Constitution Outside Article V’ (1988) 55 U Chi L Rev 1043, 1054–58; AR Amar, ‘The Consent of the Governed: Constitutional Amendment Outside Article V’ (1994) 94 Col L Rev 457, 458–500.

⁹CV Keshavamurthy, Amending Power Under The Indian Constitution—Basic Structure Limitations (Deep & Deep Publications, 1982) 80–1. This approach was accepted, for instance, by the Brazilian Federal Supreme Court, which held that the constitution’s ‘immutable provisions’ impose limits on the secondary constituent power, but do not subordinate the primary constituent power itself. See AZ Melo, ‘A limitação material do poder constituinte derivado’ (2008) 8(1) Revista Mestrado em Direito 31, 48.

¹⁰E Smith, ‘Old and Protected? On the “Supra-Constitutional” Clause in the Constitution of Norway’ (2011) 44(3) Israel L Rev 369, 375–76; SK Guha and M Tundawala, ‘Constitution: Amended it Stands?’ (2008) 1 Nat U Jur Sci L Rev 533, 537.

¹¹V Jackson, ‘Unconstitutional Constitutional Amendments: A Window into Constitutional Theory and Transnational Constitutionalism’ in A Wallrabenstein, P Dann and M Bäuerle (eds), Demokratie-Perspektiven, Festschrift für Brun-Otto Bryde zum 70 (Mohr Siebeck, 2013) 47.

¹²AW Machen, ‘Is The Fifteenth Amendment Void?’ (1909–1910) 23 Harv L Rev 169, 170; WF Harris II, The Interpretable Constitution (Johns Hopkins University Press, 1993) 193. Of course, the assumption that constituent power resides in ‘the people’ carries a contingent character as it is the consequence of a certain political belief which may change in different periods and places.

¹³cf S Weintal, ‘The Challenge of Reconciling Constitutional Eternity Clauses with Popular Sovereignty: Toward Three-Track Democracy in Israel As a Universal Holistic Constitutional System and Theory’ (2011) 44(3) Israel L Rev 449; A Kalyvas, ‘Carl Schmitt and The Three Moments of Democracy’ (1999–2000) 21 Cardozo L Rev 1525; S Prateek, ‘Today’s Promise, Tomorrow’s Constitution: “Basic Structure”, Constitutional Transformations and The Future of Political Progress in India’ (2008) Nat U Jur Sci L Rev 417, 458–61.

¹⁴See Roznai, Unconstitutional Constitutional Amendments (n 1); U Baxi, ‘Some Reflections on the Nature of Constituent Power’ in R Dhavan and A Jacov (eds), Indian Constitution—Trends and Issues (NM Tripathi Private Ltd, 1978) 122, 136.

¹⁵C Saunders, ‘The Constitutional Credentials of State Constitutions’ (2011) 42 Rutgers L J 853, 870. Of course, this idea of free and unbound constituent power carries its own dangers, as identified by H Arendt, On Revolution (Penguin Group, 1965) 163. See the debate in J Braver, ‘Revolutionary Reform in Venezuela: Electoral Rules and Historical Narratives in the Creation of the 1999 Constitution’ in this volume; JZ Benvindo, ‘“Revolutionary Reform” and the Seduction of Constitutionalism’ in this volume.

¹⁶M Loughlin, Foundations of Public Law (Oxford University Press, 2010) 224; S Levinson, Constitutional Faith, 2nd edn (Princeton University Press, 2011) 193–94.

¹⁷See, eg, M Loughlin and N Walker (eds), The Paradox of Constitutionalism: Constituent Power and Constitutional Form (Oxford University Press, 2007); Z Oklopcic, ‘Three Arenas of Struggle: A Contextual Approach to the Constituent Power of “the People”’ (2014) 3(2) Glob Const 200; Z Oklopcic, ‘Constitutional Theory and Cognitive Estrangement: Beyond Revolutions, Amendments and Constitutional Moments’ in this volume.

¹⁸See debates, eg in ES Morgan, Inventing The People: The Rise of Popular Sovereignty in England and America (WW Norton, 1988) 14, 58; B Darby ‘Amending Authors and Constitutional Discourse’ (2002) 25(2) Dalhousie L J 215, 221; FI Michelman, ‘Constitutional Authorship By The People’ (1998–1999) 74 Notre Dame L Rev 1605, 1628; E Oxman, The Constitution of the People (PhD Dissertation, Princeton University, 2009); Benvindo (n 15).

¹⁹DJ Galligan, ‘The Sovereignty Deficit of Modern Constitutions’ (2013) 33 O J Leg Stud 1, 9–11.

²⁰DA Strauss, ‘We the People, They the People, and the Puzzle of Democratic Constitutionalism’ (2013) 91 Texas L Rev 1969.

²¹RS Kay, ‘Constituent Authority’ (2011) 59 Am J Comp L 715, 739; See also Jackson (n 11) 65. Of course, the relationship between constitutional law and popular sovereignty differs between different constitutional systems which have diverse relationship with ‘the people’ in their constitutional design and constitution-making. See Pereira (n 3).

²²Harris (n 12) 193.

²³See generally M Loughlin, ‘The Concept of Constituent Power’ (2014) 13(2) European Journal of Political Theory 218; JI Colón-Ríos, ‘Five Conceptions of Constituent Power’ (2014) 130 Law Quarterly Review 306; M Spång, Constituent Power and Constitutional Order (Palgrave Macmillan, 2014); Y Roznai, ‘“We the People”, “Qui, the People” and the Collective Body: Perceptions of Constituent Power’ in G Jacobsohn and M Schor (eds), Comparative Constitutional Theory (Edward Elgar, 2017, forthcoming).

²⁴RM Unger, What Should Legal Analysis Become? (Verso, 1996) 72.

²⁵A Kalyvas, Democracy and the Politics of the Extraordinary: Max Weber, Carl Schmitt, and Hannah Arendt (Cambridge, Cambridge University Press, 2008) 7.

²⁶L Barshack, ‘The subject of ideals’ (2000) 4(1) Cultural Values 77, 82; L Barshack, ‘Notes on The Clerical Body of The Law’ (2003) 24(3) Cardozo L Rev 1151, 1155, 1164; L Barshack, ‘Political Theology and the Authenticity of Modern Experience’ (2012) 5 Jerusalem Rev Leg Stud 12, 19.

²⁷L Barshack, ‘Time and the Constitution’ (2009) 7(4) Int’l J Const L 551, 553, 557, 566–67. On time and constitution-making/change see S Ranchordás, ‘Constitutional Sunrise’, in this volume.

²⁸L Barshack, ‘Constituent Power as Body: Outline of a Constitutional Theology’ (2006) 56 Uni Toronto L J 185, 202.

²⁹ibid 212–13.

³⁰AR Amar, ‘Of Sovereignty and Federalism’ (1987) 96 Yale L J 1425; Amar, ‘Philadelphia Revisited’ (n 8) 1054–58; AR Amar, ‘The Consent of the Governed’ (n 8) 457–500; AR Amar, ‘Popular Sovereignty and Constitutional Amendment’ in S Levinson (ed), Responding to Imperfection: The Theory and Practice of Constitutional Amendment (Princeton University Press, 1995) 89, 89–101.

³¹S Wolin, ‘Fugitive Democracy’ in S Benhabib (ed), Democracy and Difference: Contesting the Boundaries of the Political (Princeton University Press, 1996) 31, 39, 41. See also S Wolin, Politics and Vision (Princeton University Press, 2004) 602–03.

³²M Schwartzberg, Democracy and Legal Change (Cambridge, Cambridge University Press, 2009) 6.

³³CJ Friedrich ‘Military Government and Dictatorship’ (1950) 267 Annals of the American Academy of Political and Social Science 1, 4.

³⁴AAI Ninet, ‘From Popular Sovereignty to Constitutional Sovereignty?’ (2010) Buffalo Legal Studies Research Paper 2011-007 <ssrn.com/abstract=1587350> accessed 1 August 2016.

³⁵AR Coban, ‘The Continuity Problem of the Constituent Power’ in A Scherzberg, O Can and I Doğan (eds), Verfassungstheorie und Verfassungsgebung: Überlegungen anlässlich der Diskussion um eine Verfassungsreform in der Türkei (LIT Verlag, 2012) 49, 56, n 13.

³⁶J Colón-Ríos, ‘The Second Dimension of Democracy: The People and Their Constitution’ (2009) 2(2) Baltic J L & Pol 1.

³⁷J Colón-Ríos, Weak Constitutionalism—Democratic Legitimacy and the Question of Constituent Power (Routledge, 2012); Colón-Ríos, ‘The Legitimacy of the Juridical: Constituent Power, Democracy, and the Limits of Constitutional Reform’ (2010) 48(2) Osgoode Hall L J 199; Colón-Ríos, ‘Notes on Democracy and Constitution-Making’ (2011) 9(1) The New Zealand J Pub & Int’l L 17; Colón-Ríos, ‘The Counter-Majoritarian Difficulty and the Road Not Taken: Democratizing Amendment Rules’ (2012) Canadian J L Juris 53. See also AC Hutchinson and J Colón-Ríos, ‘Democracy and Constitutional Change’ (2011) Theoria 43, 52–53.

³⁸Colón-Ríos, ‘The Legitimacy of the Juridical’ (n 37) 240.

³⁹S Tierney, ‘Whose Political Constitution? Citizens and Referendums’ (2013) 14(12) German L J 2185, 2194; Tierney ‘Sovereignty and Crimea: How Referendum Democracy Complicates Constituent Power in Multinational Societies’ (2015) 16(3) German L J 523, 525.

⁴⁰See, eg, J Blount, ‘Participation in Constitutional Design’ in T Ginsburg and R Dixon (eds), Comparative Constitutional Law (Edward Elgar, 2011) 49; Z Elkins, T Ginsburg, and J Blount, ‘The Citizen as Founder; Public Participation in Constitutional Approval’ (2008) 81 Temple L Rev 361. See also M Versteeg, ‘Unpopular Constitutionalism’ (2014) 89(3) Indiana L J 1133, 1142 (‘today, 34% of all constitutions require ratification by popular referendum, while in 1950, only 7% did’).

⁴¹C Saunders, ‘Constitution Making in the 21st Century’ (2012) 4 Int’l Rev L. See also Kay (n 21) 746; JA Lenowitz, ‘Why Ratification? Questioning the Unexamined Constitution-making Procedure’ (PhD Thesis, Columbia University, 2013) 119.

⁴²D Butler and A Ranney, ‘Theory’ in D Butler and A Ranney (eds), Referendums Around the World: The Growing Use of Direct Democracy (DC, AEI 1994) 11, 17–21; Friedrich (n 6) 115–17.

⁴³For a discussion of interest groups’ roles in referenda see O Doyle and D Kenny, ‘Constitutional Change and Interest Group Politics: Ireland’s Children’s Rights Referendum’, in this volume. I thank Mark Tushnet for emphasising that due to elite’s control of referendums parliamentary setting is more deliberative than referendums. Indeed, referendums are traditionally considered as ‘elite driven’. See, eg, A Lijphart, Democracies, Patterns of Majoritarian and Consensus Government in Twenty-One Countries (Yale University Press, 1984) 203. However, there are examples of failed referendums in which governments were unable to secure the outcome of referendums they have initiated. See D O’Brien, ‘Formal Amendment Rules and Constitutional Endurance: the strange case of the Commonwealth Caribbean’, in this volume.

⁴⁴A Ninet, Constitutional Violence: Legitimacy, Democracy and Human Rights (Edinburgh University Press, 2013) 14.

⁴⁵Hutchinson and Colón-Ríos (n 37) 53; Coban (n 35) 73.

⁴⁶S Holmes, The Anatomy of Antiliberalism (Harvard University Press, 1996) 49.

⁴⁷C Burchard, ‘Carl Schmitt on Democracy and International Relations’ (2006) 19 Leiden J Int’l L 9, 13.

⁴⁸See R Levy, ‘“Deliberative Voting”: Reforming Constitutional Referendum Democracy’ (2013) Pub L 555, 566 (proposes to improve deliberative voting in constitutional referenda); EJ Leib, ‘Can Direct Democracy Be Made Deliberative?’ (2006–2007) 54 Buffalo L Rev 903, 911–12 (‘deliberative democracy offers a potential blueprint that can help us devise a better way to undertake direct democracy’); S Tierney, ‘Constitutional Referendums: A Theoretical Enquiry’ (2009) 72(3) Modern L Rev 360, 382 (it ‘seems intuitively plausible that a referendum, carefully tailored to meet the specificities of a particular society, can help bring a populace together in a deliberative, constitutional moment’.)

⁴⁹cf Barshack (n 28) 190; B Ackerman, ‘The New Separation of Powers’ (2000) 113(3) Harv L Rev 633, 665–67.

⁵⁰D Conrad, ‘Limitation of Amendment Procedures and the Constituent Power’ (1970) 15–16 Ind Y B Int’l Aff 347, 404–10; Prateek (n 13) 454.

⁵¹Kay (n 21) 742. Since there is no clear mechanism for the exercise of primary constituent power, the decision whether a constitutional norm was indeed a manifestation of the primary constituent power can only be given retrospectively, such as by a judiciary’s legitimation of the extra-constitutional or the recognition of the governed. See M Schwartzberg, ‘Should Progressive Constitutionalism Embrace Popular Constitutionalism? (2011) 72(6) Ohio State L J 1295, 1303, 1314.

⁵²An important aspect is therefore the maintenance of freedoms such as freedom of speech, free and fair election, freedom from arbitrary arrest, and freedom of assembly and association. See Guha and Tundawala (n 10) 543; D Conrad, ‘Constituent Power, Amendment and Basic Structure of the Constitution: A Critical Reconsideration’ (1977–1978) 6–7 Delhi L Rev 1, 12; K Chryssogonos, ‘Popular Involvement in Constitution-Making’ (2008) 20(4) Revue Européenne de Droit Public 1299–1316; Friedrich (n 6) 116.

⁵³A Kalyvas, ‘Popular Sovereignty, Democracy, and the Constituent Power’ (2005) 12(2) Constellations 223, 235.

⁵⁴I borrow this term from H Cyr, Canadian Federalism and Treaty Powers: Existential Communities, Functional Regimes and the Canadian Constitution (LL.D Thesis, Université de Montreal, 2007) 45, n 74.

⁵⁵D Landau, ‘The Importance of Constitution-Making’ (2011–2012) 89 Denver University L Rev 611, 616–17. See also C Winterhoff, Verfassung—Verfassunggebung—Verfassungsänderung. Zur Theorie der Verfassung und der Verfassungsrechtserzeugung (Mohr Siebeck, 2007) 150–51; Kay (n 21) 745.

⁵⁶Kay (n 21) 725–26; Coban (n 35) 56; M Tushnet, ‘Constitution-Making: An Introduction’ (2012–2013) 91 Texas L Rev 1983, 1988; D Landau and R Dixon, ‘Constraining Constitutional Change’ (2015) 50 Wake Forest L Rev 859.

⁵⁷Art 347. Arts 348–349 of the Constitution further regulate this process.

⁵⁸See, eg, Argentine Const (1994) Sec 30; Nicaraguan Const (1986) Arts 191–193; Swiss Const (1999), Arts 138, 193, and Bulgarian Const (1991) Arts 158–162. See Kay (n 21) 725; ‘Report on Constitutional Amendment’, European Commission for Democracy Through Law (Venice Commission) (Venice, 11–12 December 2009) 13 <http://www.venice.coe.int/webforms/documents/CDL-AD(2010)001.aspx>, accessed 1 August 2016.

⁵⁹C Schmitt, Constitutional Theory (J Seitzer tr, Duke University Press, 2008) 132.

⁶⁰Kay (n 21) 727–28.

⁶¹See, eg, Kalyvas (n 53) 228; R Guastini, ‘On the Theory of Legal Sources’ (2007) 20(2) Ratio Juris 302, 305: ‘in no legal system can constituent power […] be regarded as an institutionalized source: otherwise, it would not be “constituent but “constituted.”’

⁶²Lenowitz (n 41) 87.

⁶³Friedrich (n 6) 118.

⁶⁴R Cristi, ‘The Metaphysics of Constituent Power: Schmitt and the Genesis of Chile’s 1980 Constitution’ (1999–2000) 21 Cardozo L Rev 1749, 1758, 1765; M Goldoni and C McCorkindale, ‘Why we (still) need a revolution’ (2013) 114(12) German L J 2197, 2214–15; PW Kahn, ‘Political Time: Sovereignty and The Transtemporal Community’ (2006) 28(1) Cardozo L Rev 259, 268.

⁶⁵Kay (n 21) 735.

⁶⁶Tushnet (n 56) 2006.

⁶⁷cf Tierney (n 2) 133.

⁶⁸Kay (n 21) 733. On the use of positive law argument for justifying revolutions, see RS Kay, ‘Legal Rhetoric and Revolutionary Change’ (1997) 7 Caribbean L Rev 161.

⁶⁹Coban (n 35) 56.

⁷⁰‘Report on Constitutional Amendment’ (n 58) 15.

⁷¹Of course, such mechanisms themselves are also open for abuses. D Landau, ‘Abusive Constitutionalism’ (2013) 47(1) UC Davis Law Review 189; Landau and Dixon (n 56).

⁷²Chryssogonos (n 52) 1299–1316.

⁷³C Möllers, ‘“We are (afraid of) the people”: Constituent Power in German Constitutionalism’, in Loughlin and Walker (eds) (n 17) 87, 97–98.

⁷⁴Grundgesetz (1949) Art 146. At the moment of unification, Art 146 was not invoked and the Basic Law was instead amended and was kept in place. See Kay (n 21) 727. In the acts of the reunification of 23.09.1990, Art 146 was amended as follows: ‘after the union and freedom of Germany have been finalized this constitution shall be valid for all the German people and will continue to be valid until the day when a new constitution is accepted by the free will of the German people’.

⁷⁵Coban (n 35) 58.

⁷⁶See H Goerlich, ‘Concept of Special Protection for Certain Elements and Principles of the Constitution Against Amendments and Article 79(3), Basic Law of Germany’ (2008) 1 Nat U Jur Sci L Rev 397.

⁷⁷See debates in Kay (n 21) 727; Möllers (n 73) 97; JEK Murkens, From Empire to Union: Conceptions of German Constitutional Law since 1871 (Oxford University Press, 2013) 173–75; PE Quint, The Imperfect Union: Constitutional Structures of German Unification (Princeton University Press, 1997) 49.

⁷⁸The Lisbon Case, BVerfg, 2 be/08 of 30 June 2009, para 217 <http://www.bundesverfassungsgericht.de/entscheidungen/es20090630_2bve000208en.html> accessed 1 August 2016.

⁷⁹Murkens (n 77) 175. See also Coban (n 35) 59.

⁸⁰See Y Roznai, ‘Unamendability and the Genetic Code of the Constitution’ (2015) 27(2) Eur Rev Pub L 775.

⁸¹Goerlich (n 76) 404; UK Preuss, ‘The Implications of “Eternity Clauses”: The German Experience’ (2011) 44(3) Israel L Rev 429, 440, 443.

⁸²L Jaume, ‘Constituent Power in France: The Revolution and its Consequences’, in Loughlin and Walker (eds), (n 17) 67, 82–83; J P Derosier, ‘The French People’s Role in Amending The Constitution’, in this volume; Tierney (n 2) 130–36.

⁸³CC Decision No. 62-20 DC, 6 November 1962 Referendum Act, Rec., 1962, 27. See D Baranger, ‘The Language of Eternity: Constitutional Review of the Amending Power in France (Or the Absence Thereof)’ (2011) 44(3) Israel L Rev 389, 392–94; Derosier (n 82).

⁸⁴Cited in F Lieber, On Civil Liberty and Self-Government (JB Lippincott, 1859) 388.

⁸⁵Barshack (n 28) 212–13.

⁸⁶See P Carrozza,’ Constitutionalism’s Post-Modern Opening’ in Loughlin and Walker (eds), (n 17) 169, 174. Whereas the factual tends to have a normative power, an effective exercise of constituent power does not mean that the extra-constitutional action was constitutional under the previous constitutional rules. See Y Roznai and S Suteu, ‘The Eternal Territory? The Crimean Crisis and Ukraine’s Territorial Integrity as an Unamendable Constitutional Principle’ (2015) 16(3) German L J 542, 569, n 161. Therefore, Derosier (n 82) is correct in his contribution to this volume when he refers to the 1962 amendment as a legal revolution. Nevertheless, one may argue that the employment process may establish a constitutional precedent for a new amendment procedure to be availed of in future occasions. cf B Ackerman, We the People: Transformations (Harvard University Press, 2000) 415.

⁸⁷Kalyvas (n 523) 231.

⁸⁸K Gözler, Pouvoir constituant (Ekin Press, 1999) 102.

⁸⁹See A Kavanagh, ‘Unconstitutional Constitutional Amendments from Irish Free State to Irish Republic’ in E Carolan (ed), The Constitution of Ireland: Perspectives and Prospects (Bloomsbury Professional, 2012) 331; Y Roznai ‘The Theory and Practice of “Supra-Constitutional” Limits on Constitutional Amendments’ (2013) 62(3) Int’l & Comp L Q 557, 566–69; R O’Connell, ‘Guardians of the Constitution: Unconstitutional Constitutional Norms’ (1999) 4 J C L 48, 61–66; A O’Sullivan and PCW Chan, ‘Judicial Review in Ireland and the Relationship Between The Irish Constitution and Natural Law’ (2006) 15 Nott L J 18, 32; O Doyle, ‘Legal Validity: Reflections on the Irish Constitution’ (2003) 25 Dublin University L J 56, 65–67.

⁹⁰See O’Sullivan and Chan ibid; Roznai ibid; Kavanagh ibid; GJ Jacobsohn, ‘An Unconstitutional Constitution? A Comparative Perspective’ (2006) 4(3) Int’l J Const L 460, 469. Doyle (n 1), in this volume, notes that ‘The Irish rule that the Government cannot spend public money supporting just one side of a constitutional referendum is a majoritarian process-constraint, designed to ensure that the real “will of the people” is ascertained’.

⁹¹O Beaud, La puissance de l'état (PUF, 1994) 437–38; O Beaud, ‘Le Souverain’ (1993) 67 Pouvoirs 33, 37.

⁹²See E de Wet, ‘The Prohibition of Torture as an International Norm of jus cogens and Its Implications for National and Customary Law’ (2004) 15(1) Eur J Int’l L 97, 101–05; A Zimmermann, ‘Is It Really All about Commitment and Diffusion? A Comment on Commitment and Diffusion: How and Why National Constitutions Incorporate International Law’ (2008) U Ill L Rev 253, 258.

⁹³Swiss Const (1999), Arts 193(4), 194(2). See also G Biaggini, ‘Switzerland’ in D Oliver and C Fusaro (eds), How Constitutions Change—A Comparative Study (Hart Publishing, 2011) 303, 316–17.

⁹⁴C Klein, ‘Is There a Need for an Amending Power Theory?’ (1978) 13 Israel L Rev 203, 213. See also Note, ‘The Faith To Change: Reconciling The Oath To Uphold With The Power To Amend’ (1995–1996) 109 Harv L Rev 1747, 1759.

⁹⁵Tierney (n 2) 12.

⁹⁶ibid 13; Tierney (n 48) 360.

⁹⁷ibid 13. See similarly Lenowitz (n 41) 85; Smith (n 10) 376.

⁹⁸ibid 14. See also Tierney (n 48) 364.

⁹⁹Klein (n 94) 213.

¹⁰⁰See R Albert, ‘The Structure of Constitutional Amendment Rules’ (2014) 49 Wake Forest L Rev 913; M Andenas (ed), The Creation and Amendment of Constitutional Norms (BIICL, 2000); Oliver and Fusaro (n 93); X Contiades (ed), Engineering Constitutional Change: A Comparative Perspective on Europe, Canada and the USA (Routledge, 2012); ‘Report on Constitutional Amendment’ (n 58); DS Lutz, ‘Toward a Theory of Constitutional Amendment’ (1994) 88 American Political Science Review 355, 363–64; J Elster, Ulysses Unbound. Studies in Rationality, Precommitment and Constraints (Cambridge University Press, 2000) 100–04.

¹⁰¹cf JE Lane, Constitutions and Political Theory (Manchester University Press, 1996) 114; AV Dicey, ‘The Referendum’ (1894) 23 National Review 65, 69.

¹⁰²I elaborate on this in Roznai, Unconstitutional Constitutional Amendments (n 1).

¹⁰³See EV Schneier, Crafting Constitutional Democracies: The Politics of Institutional Design (Rowman & Littlefield, 2006) 223–24.

¹⁰⁴R Ku, ‘Consensus of the Governed: The Legitimacy of Constitutional Change’ (1995) 64(2) Ford L Rev 535, 542; LG Sager, ‘The Birth of Logic of a Democratic Constitution’ in J Ferejohn et al (eds), Constitutional Culture and Democratic Rule (Cambridge University Press, 2001) 110, 132–33; J Ferejohn and L Sager, ‘Commitment and Constitutionalism’ (2003) 81 Texas L Rev 1929, 1957.

¹⁰⁵See ‘Report on Constitutional Amendment’ (n 58).

¹⁰⁶ibid; Schneier (n 103) 224.

¹⁰⁷J Elster, ‘Deliberation and Constitution Making’, in J Elster (ed), Deliberative Democracy (Cambridge University Press, 1998) 97.

¹⁰⁸A Vermeule, ‘Second Opinions and Institutional Designs’ (2011) 97 Virginia L Rev 1435, 1438; R Dixon, ‘Partial Constitutional Amendments’ (2011) U Penn J Const L 643, 651–54.

¹⁰⁹cf LG Sager, ‘The Incorrigible Constitution’ (1990) 65 N Y U L Rev 893, 951–53.

¹¹⁰A Arato and J Cohen, ‘Banishing the Sovereign? Internal and External Sovereignty in Arendt’ (2009) 16(2) Constellations 307, 317. Critics have argued that the amendment procedure is the main democratic defects of the US Constitution. See S Levinson, Our Undemocratic Constitution: Where the Constitution Goes Wrong (And How We the People Can Correct It) (Oxford University Press, 2006) 159–66; SM Griffin, ‘The Nominee is … Article V’ (1995) 12 Const Comm 171.

¹¹¹Interestingly, in contrast with the US Federal Constitution, direct democracy in the form of referenda is inherently part of US States’ constitutional tradition. See J Goossens, ‘Direct Democracy and Constitutional Change in the US: Institutional Learning from State Laboratories’, in this volume.

¹¹²See, eg, South African Const. (1996), Art 74; Canada Constitution Act, 1982, Part V. See H Corder, ‘The Republic of South Africa’ in Oliver and Fusaro (eds), (n 93) 261, 270; AC Hutchinson, ‘Constitutional Change and Constitutional Amendment—A Canadian Conundrum’ in Contiades (ed) (n 100) 51, 64–66; R Albert, ‘Amending Constitutional Amendment Rules’ (2015) 13(3) Int’l J Const L 655; R Albert, ‘The Difficulty of Constitutional Amendment in Canada’ (2015) 53(1) Alberta L Rev 85.

¹¹³See Coban (n 35) 57–66.

¹¹⁴See A Somek, ‘Constitutional Theory as a Problem of Constitutional Law—On the Constitutional Court’s Total Revision of Austrian Constitutional Law’ (1998) 32 Israel L Rev 567.

¹¹⁵See V Ferreres-Comella, ‘A Defense of Constitutional Rigidity’ in P Comanducci and R Guastini (eds), Analisi e diritto (G Giappichelli Publisher, 2000) 45, 62, n 42; A Elvira, ‘Spain’ in Oliver and Fusaro (eds) (n 93) 281, 282–84; AB Ortega and IS Guijarro, ‘Constitutional Change in Spain’ in Contiades (ed) (n 100) 299, 302–08.

¹¹⁶This of course does not exclude the possibility of implicit unamendability. For example, in Resolución 2010–1331, the Supreme Court of Justice of Costa Rica (Constitutional Chamber) stated, in an obiter, that even the derived constituent power cannot amend the Constitution in a way that violates ‘the essence of fundamental human rights’.

¹¹⁷P Suber, ‘Amendment’ in CB Gray (ed), Philosophy of Law: An Encyclopaedia I (Garland Publishing Co, 1999) 31–32.

¹¹⁸Landau (n 71); R Albert, ‘Constitutional Handcuffs’ (2010) 42 Ariz St L J 663, 707–11; R Albert, ‘The Expressive Function of Constitutional Amendment Rules’ (2013) 59(2) McGill L J 225. For an argument that selective rigidity distorts the balance between flexibility and stability see RP Plato, ‘Selective Entrenchment against State Constitutional Change: Subject Matter Restrictions and the Threat of Differential Amenability’ (2007) 82 N Y U L Rev 1470, 1489–93.

¹¹⁹A Arato, ‘Multi-Track Constitutionalism Beyond Carl Schmitt’ (2011) 18(3) Constellations 324, 340.

¹²⁰Here, the myth of the people as holders of constituent power serves as the guiding narrative for constitutional design, even if only formally. See A Sajó, ‘Enthusiasm and Acquiescence in Constitution-making’ (2009) 3 Indian J Const L 24, 27.

¹²¹X Contiades and A Fotiadou, ‘Models of Constitutional Change’ in Contiades (ed) (n 100) 417, 430.

¹²²Barshack (n 28) 201.

¹²³ibid 201–02.

¹²⁴Lithuania Const. (1992) Art 2.

¹²⁵I thank Zoran Oklopcic and Alkmene Fotiadou for pointing these issues.

¹²⁶IM Young, Inclusion and Democracy (Oxford University Press, 2000) 5–6. This is social and political legitimacy, not a legal or moral one. See generally RH Fallon, ‘Legitimacy and The Constitution’ (2004–2005) 118 Harv L Rev 1787.

¹²⁷Butler and Ranney (n 402) 14–15.

¹²⁸LE Miller, ‘Designing Constitution-Making Processes: Lessons from the Past, Questions for the Future’ in LE Miller and L Aucoin (eds), Framing The State in Times of Transition: Case Studies in Constitution Making (US Institute of Peace Press, 2010) 601, 612.

¹²⁹H Klug, ‘Participating in the Design: Constitution-Making in South Africa’ (1996) 3 Review of Constitutional Studies 18.

¹³⁰cf B Medina, ‘“Foundational” Originalism? On Jack Balkin’s Living Originalism’ (2013) 7(1) Jerusalem Rev Leg Stud 1, 8.

¹³¹See, eg, L Vinx, ‘The Incoherence of Strong Popular Sovereignty’ (2013) 11(1) Int’l J Const L 101, 124 (claiming that ‘The view that strong popular sovereignty is essential to democracy is false’).

¹³²Tushnet (n 56) 2000. See also Kalyvas (n 53) 237.

¹³³See, eg, M Tushnet, ‘Peasants with Pitchforks, and Toilers with Twitter: Constitutional Revolutions and the Constituent Power’ (2015) 13(3) Int’l J Const L 639, 642–43 (according to whom, successfully amending an unamendable principle through a supposedly unconstitutional amendment is an exercise of constituent power).

¹³⁴H Kong, ‘Deliberative Constitutional Amendments’ (2015) 41 Queen’s L J 105; Compare with CR Sunstein, Designing Democracy: What Constitutions do (Oxford University Press, 2001) 7, 240.

¹³⁵cf Landau (n 71) 226.

¹³⁶J Elster, ‘Forces and Mechanisms in the Constitution-Making Process’ (1995) 45 Duke L J 364, 370–71, 395; J Elster, ‘Legislatures as Constituent Assemblies’ in RW Bauman and T Kahana (eds), The Least Examined Branch: The Role of Legislatures in The Constitutional State (Cambridge University Press, 2006) 181, 185; J Elster, ‘Arguing and Bargaining in Two Constituent Assemblies’ (2000) 2 U Pa J Const L 345, 419–20. Of course, the construction of constituent assemblies established for constitutional revision can also be manipulated or abused as demonstrated in this volume by Braver (n 15).

¹³⁷cf TE Pettys, ‘Popular Constitutionalism and Relaxing the Dead Hand: Can the People Be Trusted?’ (2008) 86 Wash U L Rev 313, 350–51.

¹³⁸KM Pillai, ‘Amendability of Fundamental Rights under the Constitution of India’ in Dhavan and Jacob (n 14) 192, 199.

¹³⁹E Daly, ‘A Republican Defence of the Constitutional Referendum’ (2015) 35(1) Leg Stud 30; AV Dicey, ‘Ought The Referendum To Be Introduced Into England?’ (1890) 57 Cont Rev 489, 505.

¹⁴⁰W Partlett, ‘The Dangers of Popular Constitution-Making’ (2012) 38 Brook J Int’l L 1, 4; D Landau, ‘Constitution-Making Gone Wrong’ (2013) 64(5) Ala L Rev 923.

¹⁴¹HP Monaghan, ‘We The People[s], Original Understanding, and Constitutional Amendment’ (1996) 96 Colum L Rev 121. This might find some support in empirical studies. See, eg, BS Gamble, ‘Putting Civil Rights to a Popular Vote’ (1997) 41(1) Am J Pol Sci 245, 254, 261; B Frey and L Goette, ‘Does the Popular Vote Destroy Civil Rights?’ (1998) 42 Am J Pol Sci 1343; ZL Hajnal, ER Gerber and H Louch, ‘Minority and Direct Legislation: Evidence from California Ballot Propositions Elections’ (2002) 64(1) Journal of Politics 154; RT Moore and N Ravishankar, ‘Who Loses in Direct Democracy?’ (2012) 41 Social Science Research 646. But see JG Matsusaka, ‘The Eclipse of Legislatures: Direct Democracy in the 21st Century’ (2005) 124 Public Choice 157, 168 (arguing that the question of risk to minorities in direct democracy lacks ‘rigorous empirical work on this issue, and the work that does exist rests on flawed methodologies’).

¹⁴²JN Eule, ‘Judicial Review of Direct Democracy’ (1990) 99 Yale L J 1503.

¹⁴³W Marxer, ‘Forward’ in W Marxer (ed), Direct Democracy and Minorities (Springer Verlag, 2012) 7, 10. See also CA Johanningmeier, ‘Law & Politics: The Case Against Judicial Review of Direct Democracy’ (2007) 82 Indiana L J 1125, 1150.

¹⁴⁴JG Matsusaka, For the Many or the Few: The Initiative, Public Policy, and American Democracy (University of Chicago Press, 2008) 116; S Spadijer, ‘A Hardcover Case Against (Strong) Judicial Review of Direct Democracy’ (2012) 31 U Queensland L J 55.

¹⁴⁵See Tierney (n 2); Tierney (n 48) 360. Interestingly, referendums on constitutional revisions are less subject to substantive judicial review than referendums on ordinary legislation. See L Morel, ‘Referendum’ in M Rosenfeld and A Sajó (eds), The Oxford Handbook of Comparative Constitutional Law (Oxford University Press, 2012) 501, 522.

¹⁴⁶Conrad (n 52) 14–16; D Grimm, ‘The Basic Law at 60—Identity and Change’ (2010) 11(1) German L J 33, 40.

¹⁴⁷J Blount, Z Elkins, and T Ginsburg, ‘Does the Process of Constitution-Making Matter?’, in T Ginsburg (ed), Comparative Constitutional Design (Cambridge University Press, 2012) 31, 36.

¹⁴⁸A requirement of a popular ratification certainly adds another hurdle to the amendment process, but approvals of constitutional amendments via referendums is not necessarily more difficult than approving it in state legislatures. See DS Lutz, Principles of Constitutional Design (Cambridge University Press, 2006) 166–68; J Ferejohn, ‘The Politics of Imperfection: The Amendment of Constitutions’ (1997) 22 L & Soc Inq 501, 523.

¹⁴⁹CL Eisgruber, Constitutional Self-Government (Harvard University Press, 2007) 44.

¹⁵⁰cf Blount, Elkins and Ginsburg (n 147) 31, 54; AM Banks, ‘Expanding Participation in Constitution Making: Challenges and Opportunities’ (2008) 49 William & Mary L Rev 1043, 1050; K Samuels, ‘Post-Conflict Peace-Building and Constitution-Making’ (2006) 6(2) Chi J Int’l L 663, 668.

¹⁵¹Deliberative settings may also increase public trust in the constitutional reform process. See R Levy, ‘Breaking the Constitutional Deadlock: Lessons From Deliberative Experiments in Constitutional Change’ (2010) 34 Melbourne U L Rev 805, 838.

¹⁵²Prateek (n 13) 459.

¹⁵³Note (n 94) 1762–63.

¹⁵⁴Prateek (n 13) 472.

¹⁵⁵Harris (n 12) 167.

¹⁵⁶ibid 194–203.

¹⁵⁷On the authority to review constitutional norms and the proper standard of such a review see Y Roznai, ‘Legisprudence Limitations on Constitutional Amendments? Reflections on The Czech Constitutional Court’s Declaration of Unconstitutional Constitutional Act’ (2014) 8(1) ICL-J 29, 33–40. On judicial review of amendments see generally K Gözler, Judicial Review of Constitutional Amendments—A Comparative Study (Ekin Press, 2008); A Barak, ‘Unconstitutional Constitutional Amendments’ (2011) 44(3) Israel L Rev 321; G Halmai, ‘Unconstitutional Constitutional Amendments: Constitutional Courts as Guardians of the Constitution?’ (2012) 19(2) Constellations 182.

¹⁵⁸cf EJ Leib, DL Ponet and M Serota, ‘Translating Fiduciary Principles Into Public Law’ (2013) 126 Harv L Rev Forum 91, 93.

¹⁵⁹CM Fombad, ‘Limits on the Power to Amend Constitutions: Recent Trends in Africa and Their Potential Impact on Constitutionalism’ (2007) 6 University of Botswana Law Journal 27, 59; B Ackerman, ‘Constitutional Economics—Constitutional Politics’ (1999) 10 Constitutional Political Economy 415, 423. Amendments-rate depends on many other criteria. See R Dixon, ‘Constitutional Amendment Rules’ in Ginsburg and Dixon (n 40) 96; R Dixon and R Holden, ‘Constitutional Amendment Rules: The Denominator Problem’ in Ginsburg (n 147) 195. For a critical analysis of quantitative amendment metrics see X Contiades and A Fotiadou, ‘Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution’, in this volume and a response by JE Fleming, ‘Comment on Amendment-Metrics: The Good, the Bad and the Frequently Amended Constitution’, in this volume. See also T Ginsburg and J Melton, ‘Does the Constitutional Amendment Rule Matter at All? Amendment Cultures and the Challenges of Measuring Amendment Difficulty’ (2015) 13(3) Int’l J Const L 686.

¹⁶⁰Conrad (n 50) 415.

¹⁶¹PJ Yap, ‘The Conundrum of Unconstitutional Constitutional Amendments’ (2015) 4 Glob Const 114, 131 (‘we should … weigh the risks of any legislative abuse of the amendment process against the dangers of any judicial abuse that may follow from unelected judges enforcing a nebulous ‘essential features’ doctrine that can frustrate legitimate constitutional revisions designed to meet changing times’.)

¹⁶²LB Orfield, Amending the Federal Constitution (The University of Michigan Press, 1942) 120. See also LB Orfield, ‘The Scope of the Federal Amending Power’ (1929–1930) 28 Mich L Rev 550, 558.

¹⁶³Barshack (n 28) 198, n 24.

¹⁶⁴cf ‘Report on Constitutional Amendment’ (n 58) 46–47; Jacobsohn (n 90) 487; Prateek (n 13) 465–67; Yap (n 161) 131–5; C Bernal, ‘Unconstitutional Constitutional Amendments in the case Study of Colombia: An analysis of the Justification and Meaning of the Constitutional Replacement Doctrine’ (2013) 11 Int’l J Const L 339, 357.

¹⁶⁵R Dhavan, ‘The Basic Structure Doctrine—A Footnote Comment’ in Dhavan and Jacob (n 14) 160, 178.

¹⁶⁶Roznai, Unconstitutional Constitutional Amendments (n 1).

¹⁶⁷Harris (n 12) 198.

¹⁶⁸Tierney (n 48) 367 (referendums can be regarded ‘as some kind of reversal of the original act of transference, or at least as a temporary return of power to the people’).

¹⁶⁹Contiades and Fotiadou (n 159), in this volume.

¹⁷⁰JK Tulis, ‘The Demon at the Center’ (1988) 55(2) U Chi L Rev 548.

¹⁷¹cf ME Brandon, ‘The “Original” Thirteenth Amendment and the Limits to Formal Constitutional Change’ in Levinson (ed) (n 30) 216, 228.