On the legal implications of a ‘permanent’ constituent power

Abstract

This article examines the development of the doctrine of the ‘permanent constituent power’ in Mexico. This doctrine reflects a long tradition in constitutional theory according to which the exercise of constituent power is a one-time event: once a constitution is adopted, there will be no legal mechanism in place for the exercise of the people’s original constitution-making authority. This view is nonetheless in tension with a notion that has also been historically embraced by liberal constitutionalism: that the people has an inalienable right to alter the form of government. The constitutional provisions that reflect that idea, we will see, can have important implications in terms of the nature and scope of the amending authority and, at the same time, point toward alternative mechanisms for the exercise of constituent authority. By closely examining the operation of those kinds of provisions in the Mexican constitution, we seek to illustrate a tension central to the liberal constitutional tradition and to suggest a way out of it. In so doing, we aim to draw some lessons from the Mexican case that can contribute to current discussions about constituent power and fundamental constitutional change in liberal constitutional orders.

I. Introduction

There is a long tradition in constitutional theory according to which the exercise of constituent power is a one-time event. Once a constitution is adopted, it is said, the legally unlimited force that brought it into existence is exhausted. That means there will be no legal mechanism in place for its future exercise and that it must be treated by the legal system as if it never existed. Put differently, once exercised, constituent power ceases to be a relevant juridical category. All that is left is the competence to amend the constitution according to its own rules. This view, nonetheless, operates in tension with a notion that has historically been embraced by liberal constitutionalism: the idea that the people have an inalienable right to alter the form of government.¹ It is thus perhaps not surprising that, at least in some jurisdictions, the idea that the exercise of constituent power is a one-time event has been challenged successfully. This is generally true in the Latin American region.² The Colombian Constitutional Court, for example, has grounded the doctrine of unconstitutional constitutional amendments on the view that only the people, acting through a special (and potentially extra-legal) constitution-making body, have the authority to replace the existing constitution if they determine such a course of action desirable.³

Based on this view, those changes that alter the fundamental content of a constitution (i.e. changes that amount to the creation of a new constitutional order) are outside the scope of the ordinary amending power and fall under the exclusive jurisdiction of the constituent people. Indeed, some Latin American constitutions explicitly distinguish between the people’s (original) constituent power and the ordinary power of constitutional reform, placing substantive limits on the latter.⁴ Mexican constitutional jurisprudence is somewhat of an outlier in this respect. Legal thinking in Mexico largely embraces the traditional ‘one-time event’ approach, and constitutional scholars frequently insist on the ephemeral nature of the original constituent power.⁵ However, there is also an influential idea, which maintains that once the original constituent power is exhausted, a permanent constituent power emerges. According to it, under the Mexican Constitution of 1917, that permanent constituent power is held by those public authorities in which Article 135 places the power of constitutional reform: a two-thirds majority of the Federal Congress acting in concert with a majority of the state legislatures.⁶

This idea has had major legal implications in Mexico. It has resulted, for example, in the judicial rejection of the doctrine of unconstitutional constitutional amendments and in a lack of differentiation between ordinary amendments and changes that alter the constitution in fundamental ways. In this article, we will first examine the development of the notion of the permanent constituent power in Mexican constitutional theory and the impact it has had on the country’s constitutional doctrine. We will then consider whether the notion of an unlimited power of constitutional reform is consistent with the text and the overall structure of the Mexican Constitution, particularly in light of its adoption of the previously mentioned idea that the people possess an inalienable right to alter the form of government. Contrary to the prevailing view, we will argue that there are good reasons to conclude that the amending power in Mexico is subject to substantive limits. Building on this analysis, we will reflect on different means through which the kind of power that falls outside the scope of the amending authority could be exercised within the legal order – that is, without a break in legal continuity. Those alternative means, we argue in the final section of the article, do not always need to take the form of a special constitution-making body convened under the rules of the constitution, but include what we will call the ‘horizontal’ exercise of constituent power.

Those exercises of constituent power may be advanced by individual citizens and social movements and will (in order to be successful constituent exercises) always result in formal or informal constitutional change (or in the prevention of major constitutional change advanced by the ordinary institutions of government). They include the use of courts to protect the exclusive jurisdiction of the constituent subject from the amending authority and the exercise of political rights in ways that trigger formal or informal constitutional transformations. The conflicting ideas of constituent power as a one-time event and of the people as having the right to create a new constitution at any moment can coexist in liberal constitutional orders because of the possibility of those horizonal exercises. That possibility offers additional support to the notion that, given that the ordinary amendment procedure is not the only means to produce changes into the constitutional order, subjecting it to substantive limits would not deprive the people of its sovereign power of constitutional change. In fact, the horizontal exercise of constituent power, given its ‘bottom-up’ nature, will usually have a higher claim to democratic legitimacy than the exercise of the power of constitutional reform by the ordinary institutions of government.

Given its focus on the Mexican case, our approach may appear parochial at first glance. But the Mexican Constitution is not the only one characterized, on the one hand, by a tension between a people attributed with the right to alter the form of government and, on the other, by a constitution that does not seem to provide a mechanism for its exercise. Indeed, the argument developed in this article can advance our understanding of the implications of an idea reflected in many constitutions around the globe: that the constituent people never give away their sovereign authority over the constitutional order. The constitutional provisions that reflect this idea, as we will see, can have important implications in terms of justifying the imposition of limits on the amending authority and, at the same time, can open the way for the notion that changes falling outside the scope of the amendment power can be adopted through alternative mechanisms. By closely examining the operation of that kind of provisions in the Mexican Constitution, we seek to illustrate a tension central to the liberal constitutional tradition and to suggest a way out of it. In this sense, our objective is not only to offer a critical overview on the relationship between the constituent and the amending power in Mexican jurisprudence. Rather, we also aim to draw some lessons from the Mexican case that can contribute to current discussions about constituent power and fundamental constitutional change in liberal constitutional orders.

II. The theory of the permanent constituent power

The question of the place of constituent power after a constitution is adopted has long occupied political and constitutional theorists. A key, and still influential, answer to that question is found in the work of Raymond Carré de Malberg, one of the main jurists of the French Third Republic. We will introduce the notion of constituent power through a brief discussion of Carré’s ideas, nonetheless remaining conscious of the different views and debates around the concept that are present in the literature.⁷ Carré de Malberg argues that from a certain perspective, the theory of constituent power lies in an extra-legal terrain: when used to explain the creation of a state’s first constitution – the constitution that brings a new state into existence – the exercise of constituent power is entirely a matter of fact rather than of law.⁸ This is also the case when new constitutional content is created in violation of the formal amendment rules of an already existing constitution.⁹ According to Carré de Malberg, the justification of that kind of action can result in interesting moral and political debates, but does not present any real legal questions.¹⁰ Once a constitutional order is in place, the exercise of that type of power can only be seen as a political fact, as a one-time event that resulted in the creation of a legal system. That does not mean, however, that Carré de Malberg thought constituent power was altogether irrelevant from a juridical point of view: in the context of an already existing constitutional order, one could observe what he called the ‘regular and pacific’ constituent power: the power to modify the existing constitution according to its own rules.¹¹

Carré thus understood the amendment authority as a ‘juridical’ type of constituent power which does not resist legal regulation.¹² Modern constitutions, he maintained, have generally operated according to this view, which is why they contain carefully designed amendment mechanisms. As a result, when a constitution is to be amended or replaced, the exercise of an extra-legal power becomes unnecessary: ‘the intervention of the organs that the constitution – the same constitution that will be amended or replaced – prescribes in advance for the regular and pacific exercise of the constituent power’ is sufficient.¹³ Those organs – such as a legislative supermajority or a special assembly – are legally authorized to introduce any content into the constitutional order, as long as they follow the applicable legal procedures.¹⁴ Academics and courts that advance this approach distinguish, like Carré de Malberg, between a revolutionary (original) constituent power and a legally (derived) regulated one.¹⁵ But what characterizes Carré de Malberg’s approach is that both notions of constituent power (the ‘original’ and the ‘derived’ one) are seen as capable of producing the same results.

The idea that the exercise of the original constituent power is a one-time event and that a materially unlimited “constituent power” is held by the institutions attributed with the amendment authority has been particularly influential in French constitutional thought.¹⁶ It has also been embraced by Mexican courts, through the concept of the “permanent constituent power”. This concept originates in the work of Felipe Tena Ramírez, whose ideas we will examine in some detail below. Tena taught constitutional law for many years at the Universidad Nacional Autónoma de México, and was also a judge at the Supreme Court of Justice from 1951 to 1970, having clerked at the court since 1946. His main work, Derecho Constitucional Mexicano, was originally published in 1944, and by the time of his death in 1994, it was on its 28th edition.¹⁷ In the opening pages of that book, Tena maintains that the existence of constitutional supremacy depends on two conditions: the separation between the constituent and the constituted powers and the presence of a written and rigid constitution.¹⁸ Since state organs (i.e. the constituted powers) are regulated by the constitution, the constitution must have been adopted by an entity different from (and superior) to them.¹⁹ That entity, according to Tena, has traditionally been called the constituent power.

The constituent subject does not exercise any ordinary governmental functions; it rather issues the fundamental laws that govern them. Operating under the one-time event view of constituent power referred to above, Tena considers that once a constitution is adopted, the constituent power ‘disappears from the juridical realm of the state, and it is substituted by the state organs it has created.’²⁰ For Tena, in countries like the United States, which have a rigid constitution (that is to say, a constitution unchangeable by the constituted powers acting in their ordinary capacity), the will of the constituent subject is protected by the courts every time they declare a legislative or executive action unconstitutional.²¹ Only the organ authorized to alter the constitutional text through a special procedure (a legislative super-majority) can “legislate” in a way that contradicts a rigid constitution. In this sense, from a strictly legal perspective, constitutional rigidity is an exclusively formal quality: any content, whether it is “constitutional” or not from a subject matter point of view, can be made more difficult to change and legally superior to ordinary law.

Most written constitutions, however, are not only rigid but contain provisions of a truly “constitutional” nature, which are related to the structure of the state, to the separation and limitation of the different state organs and to the recognition of certain rights. Following the constitutional tradition of the time, Tena called that content ‘material’. He referred to the dogmatic, organic and super-structural components of the constitution in the material sense: the dogmatic component included the recognition of fundamental rights; the organic referred to the organization of the competences of the different public organs; and, in a federation such as Mexico, the super-structural component established the division between the federal and state powers.²² This material content, of course, could very well be contained in a rigid or a flexible constitution, or in a written or unwritten one. In the context of a rigid constitution, the question is whether the written provisions that transform the material content into positive constitutional law are somehow protected from the institution authorized to formally amend the constitutional text – or, put differently, whether the material constitution, regardless of whether it has been codified or not, is beyond the competence of the constituted powers.

Carl Schmitt, writing some years before Tena (and whose work, as we will see shortly, Tena had read), famously answered those questions affirmatively: the material constitution – which he called the ‘constitution in the positive sense’²³ – could only be altered by the constituent subject itself, not by the ordinary amendment power. For Schmitt, the material content of a constitution referred to the fundamental political decisions of the constituent power. In the context of the Weimar Constitution, Schmitt maintained that those decisions were reflected in the constitutional text’s adoption of democracy as a form of government (and its rejection of monarchy),²⁴ of a federal structure of government, of parliamentarism and of the institutions of the ‘bourgeois Rechtsstaat with its principles, fundamental rights and the separation of powers’.²⁵ These decisions fell outside the competence of the amending power and thus could not be altered through the constitution’s amendment rule. This was the case even though the Weimar Constitution lacked eternity clauses. Indeed, even if a constitution contained a rule authorizing its “total revision”, those fundamental political decisions were, for Schmitt, nonetheless outside the scope of the amending power.²⁶

The fact that the Weimar Constitution did not (and could not) regulate the exercise of constituent power was a manifestation of its very nature: the constituent power was ‘prior and above every constitutional procedure.’²⁷ Constituent power, for Schmitt – who here rejects the one-time event approach discussed above – was not exhausted in an act of constitutional creation, but continued to exist ‘alongside and above the constitution’ and could therefore act at any moment in legally unanticipated ways.²⁸ Tena reached a different conclusion, one that paid particular attention to the text of the Mexican Constitution of 1917. His argument took the following form. The people exercise their constituent power when a constitution is adopted, an act that would usually take place through an extraordinary constitution-making body, such as a constituent assembly. Once that happens, constituent power is extinguished and the relevant constitution-making body disappears. At that very moment, the constituted powers are born. These newly created institutions must act in accordance with the relevant constitutional norms. These institutions are not sovereign – that is, they lack constituent power²⁹ because their faculties are ‘enumerated and restricted’.³⁰ Tena argued that, in theory, the distinction between the constituent and the constituted power was simple and clear, but in practice it was not so. Article 135 of the Mexican Constitution, he maintained, blurred it considerably as it established an organ, comprising the federal congress and state legislatures, capable of altering the constitution without specifying its limits or competencies. To the extent that it could alter the constitutional text, this organ ‘must participate in some way in the sovereign function: its function is therefore, constituent.’³¹ Since it survives the author of the constitution, Tena wrote that the organ created by Article 135 ‘merits the name of the Permanent Constituent Power.’³² The recognition of a permanent constituent power (PCP), according to Tena, should not result in a confusion between constituted and constituent entities: both the federal congress and the state legislatures remained, by themselves, ordinary constituted authorities.³³ When acting jointly, these entities would exercise constituent functions, but in that capacity they would be unable to engage in ordinary legislative activities. That is to say, the federal law-making power would remain in the exclusive hands of the federal congress, and the state law-making power in the exclusive hands of the relevant state legislature.³⁴

According to the text of Article 135, when acting as a constituent body, the task of the federal congress and state legislatures is to make reforms or additions (reformas o adiciones) to the constitution. For Tena, this meant that the PCP ‘did not have the power to totally abrogate the established constitution and to replace it with another one’ because that would go beyond the explicit authorization of the aforementioned provision.³⁵ This statement seems to bring Tena closer to Schmitt, in the sense that a change in the material constitution could be seen as amounting to the creation of a new one and therefore as falling outside of Article 135. Tena discussed Schmitt’s thesis at some length, not without noting that the idea that there might be material limits to the power to amend a constitution was first developed in Mexico by Emilio Rabasa, more than a decade before the publication of Schmitt’s Constitutional Theory in 1928.³⁶ After Rabasa, Tena explained, other Mexican jurists had accepted the thesis that the organ created by Article 135 lacked the competence to modify ‘certain basic principles of the Constitution.’³⁷

This approach was indeed reflected in foreign constitutions containing clauses that placed certain provisions outside of the scope of the amending power, or that simply referred to an unspecified fundamental content that could not be altered.³⁸ Other constitutions explicitly conferred (through clauses that referred, for example, to the possibility of a ‘total revision’) an unlimited power of constitutional reform to the organ authorized to alter the constitutional text.³⁹ Contrary to Schmitt, Tena thought that the latter should be understood as instances where the constituent power has deliberately authorized a constituted organ to alter the fundamental principles in which the constitution rests. And such a decision had to be respected by the legal system.⁴⁰ But the Mexican Constitution of 1917 was part of a group of constitutions that lacked eternity clauses and, at the same time, contained a general amendment power with no specific competencies – it did not contain a “total revision” clause.⁴¹

In order to determine the scope of the PCP in Mexico, Tena first established that Article 39 of the constitution recognized the people’s unlimited power of constitutional change. Article 39 reads as follows:

National sovereignty rests essentially and originally in the people. All public power derives from the people and is instituted for their benefit. The people have at all times the inalienable right of altering or modifying the form of its government.

According to Tena, that right included the ability to alter the constitution in fundamental ways. Even authors who thought that there were fundamental principles that could not be modified by the amending power, Tena noted, reached that conclusion precisely because they thought that only the people could modify them (acting, for example, through a special constitution-making body). The question, however, was, ‘What are the means by which the Mexican people can exercise their inalienable right of altering or modifying the form of government?’ The answer, he maintained, could not be that it would be exercised directly, through a referendum, because the constitution did not recognize that mechanism.⁴² It could be argued that the constituted powers could convene a Constituent Congress, and that that entity would serve as the means through which the people could exercise its Article 39 right. However, such a power was not recognized by the constitutional text and, in fact, all past Mexican constituent congresses had been illegally convened.

Naturally, he said, there could not be a legal doctrine according to which fundamental changes to the constitution were to be made illegally.⁴³ Tena’s solution was thus to ‘admit that the constituent organ created by Article 135 is the only entity invested with full sovereign power to make reforms or additions to any parts of the Mexican Constitution.’⁴⁴ Therefore, when the correct process is followed, no aspects of the material constitution escape the competence of the PCP. Placing the material content of the constitution outside the scope of Article 135, Tena thought, would in fact negate the inalienable right of the people to live under any form of government they wanted. ‘The grammatical meaning of words,’ he wrote, ‘must not act as a barrier that forces a people into a dilemma that seems to lack a solution.’⁴⁵ It is true, as noted earlier, that Article 135 does not authorize the PCP to formally replace the constitution with an entirely new one (only to make “reforms” or “additions” to it), but Tena maintained that the same goal could be achieved by simply “reforming” the existing constitution in the desired way by introducing any content – that is, the content that the PCP would want, including content that for all practical purposes would amount to an entirely new constitutional text.⁴⁶

The PCP was thus the ‘organ, the voice, and the will’ of the sovereign people.⁴⁷ Without it, the people would not be free and their inalienable right to change their form of government would be negated. This did not mean that Tena thought this was a desirable arrangement. He criticized the fact that the individuals who formed part of the PCP (that is, federal and state legislators) would not necessarily have been elected to engage in constituent functions.⁴⁸ Moreover, Article 135 did not give the people the right to be consulted either before or after the PCP engaged in constituent activity. Despite that concern, Tena did not favour referenda,⁴⁹ but instead preferred a system where constitutional amendments were proposed by one congress and approved by a successive one, with the intervening election serving as a mechanism of public consultation.⁵⁰ Such an arrangement would also limit the number of constitutional amendments: the notion that the amending power (the PCP) is not subject to material limits, he argues, is consistent with the idea that frequent reforms should be rejected.⁵¹

Tena was aware that his justification for the recognition of an unlimited PCP amounted to a negation of the ‘right to revolution’, but that conclusion was inescapable.⁵² The right to revolution could not be given a juridical status, and in fact the constitution explicitly rejected it. Article 136 thus stated:

This Constitution shall not lose force and effect, even if its observance is interrupted by a rebellion. In the event that a government, whose principles are contrary to those that are sanctioned herein, is established through any public disturbance, as soon as the people recover their liberty, its observance shall be re-established …

Even though that article refers to “rebellion” and not “revolution”, Tena saw it as a clear rejection of the illegal alteration of the principles protected by the constitution, which for him was the very definition of revolution. In his view, a revolution – as opposed to a mere rebellion – always involves the ‘violent suppression of the constitutional foundations of the State, and not simply the rebellion against government officials without touching the principles of the constitution.’⁵³

Although he defended the notion of a legally unlimited amendment power, Tena thought the constituent power itself, even when exercised during a revolution, would be subject to certain (seemingly non-justiciable) limits.⁵⁴ These were, for example, limits of a political nature and limits that emerged from international law and the non-observance of which may lead to consequences that the constituent subject may not be willing to accept.⁵⁵ But the most important limit was the obligation of creating a constitutional order; if such a limit was not respected, no exercise of constituent power would have taken place. In other words, to the extent that certain notions, such as the separation of powers, were accepted by the relevant society as part of the very idea of “constitution”, they needed to be respected by the constituent subject. This is why Tena wrote that, ‘Nothing escapes the competence [of the PCP], as long as a constitutional order, which includes the principles that the historical conscience of the country and of the time considers essential for the existence of a Constitution, subsists.’⁵⁶ As we will see in the next section, by adopting the notion of the PCP, the Mexican Supreme Court has, in our view, incorrectly rejected the idea of material limits to the amendment power. Indeed, it has effectively placed the amendment power above the law – that is to say, as not subject to the possibility of any kind of legal scrutiny.

III. The judicial sanction of an unlimited amending power

The adoption of Tena’s concept of the PCP has allowed the Mexican Supreme Court to answer in the negative the question of whether there are material limits to the amending power. This question had been brought to the court several times since the mid-1950s, but it was not until 1999 that the judges fully engaged with it.⁵⁷ The concept of the PCP has permeated both legal academia and constitutional doctrine, and has remained influential despite the otherwise changing political conditions in Mexico. Indeed, the first recorded use of the term ‘permanent constituent power’ by the court was on 8 June 1953,⁵⁸ shortly after Tena’s judicial appointment. During the decades that followed, it became part of the court’s constitutional language. To this day, it appears in thousands of Supreme Court opinions.⁵⁹ It is used interchangeably with that of the amending power and has become a key element in the development of a doctrine that bars the possibility of reviewing the substance of constitutional amendments. As we will see below, this doctrine has resulted in a materially and procedurally unlimited amending power.

Even though the notion of the PCP was adopted early on by the judiciary, its main legal effect was not seen until 1999, when the Supreme Court examined for the first time the scope and limits of the amending authority. Prior to 1999, the court had been asked to review the constitutionality of constitutional amendments, but the claims were dismissed on the basis of the idea that the constitution is the fundamental norm from which the validity of the rest of the legal system stems, and thus cannot be subject to judicial scrutiny.⁶⁰ In other words, that constitutional law cannot be unconstitutional. In those cases where the court has examined the scope and limits of the amending power, Tena’s influence is undeniable.⁶¹

We may say that Tena’s PCP is based on three fundamental premises: (1) as long as the correct process is followed, there are no material limits to the PCP (the sufficiency of the process premise); (2) the federal congress and state legislatures exercise a constituent function when acting jointly as the organ created by Article 135 (the joint constituent function premise); and (3) Article 135 is the only means through which the people may legally exercise the sovereign power referred to in Article 39 (the exclusivity premise). The Supreme Court largely followed Tena’s views in the Amparo en Revisión 1334/1998,⁶² where the plaintiff argued that an amendment that barred him from running for a position he previously occupied in the federal district was invalid due to a series of procedural irregularities, and that it would otherwise violate his political right to run for election.⁶³ The judges unanimously decided that individuals had standing to challenge constitutional amendments when these affected fundamental rights. However, largely adopting Tena’s PCP first fundamental premise, those amendments would only be subject to a procedural, and not substantive, review (e.g. an amendment violating a fundamental right could be declared unconstitutional only if the amending process was not correctly followed). Since it unanimously concluded that none of the procedural requirements of Article 135 were violated, the Supreme Court did not grant the relief requested by the plaintiff.

The amendment power as direct successor of the original constituent power

A few years later, in the Controversia Constitucional 82/2001, the Supreme Court (in an eight to three decision) concluded that the amending power was not subject to any type of judicially enforceable limits.⁶⁴ In that case, more than 300 Indigenous municipalities challenged the Indigenous Peoples Rights Reform of 2001, which was adopted in the context of the peace accords seeking, among other things, to put an end to the popular uprising led by the Ejército Zapatista de Liberación Nacional. The claimants argued that Article 135’s procedural requirements had been violated and that their right to be consulted (derived from Article 6 of the ILO-Indigenous and Tribal Peoples Convention of 1989, no. 169) had not been observed. The court dismissed the case on non-justiciability grounds, concluding that when acting under Article 135, the Federal Congress and the State Legislatures are the ‘direct successor’ of the original constituent power.⁶⁵ In line with Tena’s second premise (the joint constituent function premise), the court determined that ‘the challenged acts … come from an organ that is the direct successor of the [o]riginal [c]onstituent power, vested with powers that would correspond to the latter, but that before disappearing … [the original constituent power] expressly transferred and deposited in a special [complejo] organ that the doctrine calls “Reforming Power” and that is described in article 135’⁶⁶

In the view of the court, this special status is reflected on the amending authority’s ‘accruing’ nature – the fact that it comprises different organs (i.e. each of the chambers of the federal congress and a majority of state legislatures). Such an amending power, according to the court, contains the necessary features to control itself. Its ‘self-controlling’ nature explained the absence of an explicit constitutional authorization to review constitutional amendments.⁶⁷ This same reasoning was employed, and perhaps taken a step further, when deciding the Acción de Inconstitucionalidad 168/2007, an abstract review case in which several political parties challenged the electoral reform of 2007. That amendment, among other things, established a system to distribute air time for political campaigns and banned the acquisition of airtime by private individuals for electoral purposes (Article 41).⁶⁸ The claimants argued that the air time distribution system disproportionately advantaged larger parties and unduly harmed newly created ones by decreasing the chances of fair political competition.⁶⁹

In a seven to four decision, the court dismissed the case, arguing that constitutional amendments could not be subject to abstract review for three main reasons. First, since the function of the organ created by Article 135 is not an ordinary but a sovereign one, it is above the constituted powers and not subject to judicial oversight. Following Tena, the court determined that after the constitution is adopted, the original constituent power disappears and, in accordance with the exclusivity premise, the PCP becomes the exclusive ‘depositary’ of the sovereign function of ‘modifying and adding’ constitutional laws.⁷⁰ In the view of the court:

Once the basic foundational task is fulfilled, the original [c]onstituent [p]ower disappears and the newly created political organization will act through the constituted powers … In this context, it may be argued that the body in which the amendment function is placed … is the only and direct depository of the sovereign function of modifying and adding norms to the Fundamental Law [and as such] it is above the three branches of government.⁷¹

Second, it ruled that, in the absence of an explicit authorization to review the constitutionality of constitutional amendments in abstract review, the court was incompetent to do so. Instead, the competent subject to ensure the observance of the procedural limits established in the constitution was the “self-controlling” organ created by Article 135 itself.⁷² The court thus determined that once the process regulated by Article 135 is put in motion, it ‘cannot be challenged through any judicial review mechanisms’.

The question of whether the specific procedural requirements have been respected can only be examined by ‘the federal legislative organ at the moment when the ratification is officially declared, in its role as part and last instance of the constitutional amendment process.’ For the court, the amending authority is ‘responsible not just for the development of the process, but also for reviewing that the constitutional requirements are observed so that a reform or addition may become part of the Fundamental Law.’⁷³ Third, and as to the specific question regarding the possibility of subjecting constitutional amendments to judicial review, the court expressed that abstract review mechanisms were only applicable to laws in the strict sense (i.e. ordinary legislation), not to constitutional amendments.⁷⁴ Largely following Tena’s exclusivity premise, the court maintained that

the Constituent power cannot be subject to juridical limits or restrictions, because according to Article 39 of the Constitution, the people exercises its sovereignty through the Original Constituent power and, thereafter, by the Permanent Constituent power, which is comprised by the Congress of the Union and the State Legislatures.⁷⁵

Moving away from Tena’s PCP?

By 2008, the conception of the amending power that follows from the second and third premises of Tena’s theory (the joint constituent function and the exclusivity premises) was consolidated in the court’s jurisprudence. The organ created by Article 135 was thus understood as the only means available for the people to exercise their unlimited power of constitutional change. However, in late 2008 the court issued a judgment that seemed to move away from Tena’s conception. That development took place in the Amparo en Revisión 186/2008 (Amparo Empresarios), where a group of businessmen challenged the previously mentioned electoral reform of 2007, arguing that banning the acquisition of airtime for political purposes violated their freedom of speech.⁷⁶ In a six-to-four judgment, the Supreme Court reversed the lower court’s decision that the judiciary lacked jurisdiction to review constitutional amendments. Before sending the case back to the lower court, it laid out a theory of limits to the amending power, establishing a clear distinction between the constituent and the constituted authorities, a distinction that ran against the logic of Tena’s PCP.

In the court’s view, and contrary to the reasoning in the Controversia Constitucional 82/2001, the amending power could not simply be identified with the constituent power without sacrificing the legal principle of constitutional supremacy in favour of the “political” principle of popular sovereignty. For the court, in this decision, an unlimited constituent power is a force external to the constitutional order that can only manifest through revolution.⁷⁷ The amending power, on the other hand, is a constituted power bound to formal and material limits. The formal limits consist in the strict observance of the process established in Article 135. The material limits, on the other hand, are implicit and would have to be identified by the Supreme Court on a case-by-case basis. They would nonetheless always ‘be related to guaranteeing rights and protecting the separation of powers.’⁷⁸ The interpretative shift in this judgment can perhaps only be explained by the presence of new members in the court, specifically Justice José Ramón Cossío, who penned it. This approach, however, was abandoned within a few years.

Back to the non-justiciability logic

In March 2011, the Amparo en Revisión 2021/2009 (Amparo Intelectuales) was decided. In that case, sitting en banc, the court (whose membership had changed – three judges had been replaced since the last time it had to decide on this question)⁷⁹ again rejected the possibility of judicial review of constitutional amendments. The case also dealt with the electoral reform of 2007. As in the Amparo Empresarios, the claimants argued that the ban on the acquisition of airtime unlawfully restricted their freedom of speech. This time, in a seven-to-four decision, the court dismissed the case on non-justiciability grounds. Instead of discussing the scope and limits of the amending power, the majority focused on a procedural rule regarding the effects of decisions handed in Amparo proceedings. According to the judges, a ruling in favour of the claimants (i.e. declaring a constitutional amendment unconstitutional) would go against the inter partes effect of Amparo rulings – that is, that rulings can only have a legal effect on the parties to the case at hand. Seven months later, on 11 October 2011, the Second Chamber of the Supreme Court dismissed five additional cases against the same amendment following the same reasoning.⁸⁰ By issuing these five decisions, the Second Chamber met the legal requirement in Mexico for the establishment of a binding precedent for lower courts in Amparo. ⁸¹ From then on, lower courts would be bound to dismiss any individual complaint challenging a constitutional amendment.⁸²

Beyond Tena

The influence of Tena’s concept of the PCP in Mexico’s constitutional jurisprudence is evident. The court’s adoption of Tena’s theory, however, is incomplete. As we mentioned at the beginning of this section, Tena’s PCP relies on three fundamental premises: (1) the sufficiency of the procedure premise; (2) the joint constituent function premise; and (3) the exclusivity premise. In the previously mentioned cases, the court did three things in relation to Tena’s PCP theory: it explicitly adopted the second premise, implicitly accepted the third and effectively rejected the first. The cases explicitly adopted the joint constituent function premise by determining that the body created by Article 135 is the direct successor of the original constituent power and that, when acting together, the federal congress and state legislatures are not subject to judicial oversight.⁸³ They also presented the adoption of constitutional amendments as a sovereign function not subject to any type of external control performed by the ordinary constituted powers. For the court, the amending power ‘constitutes … a sovereign function not subject to any type of external control … its very guarantee is to be found in the integration [conformación] of the organ and in its constitutional function.’⁸⁴

The judgments implicitly adopted the third premise by presenting Article 135 as the only means for the constituent power to legally manifest. The court determined that, ‘The Constituent power cannot be subject to legal limits or restrictions because, according to article 39 of the constitution, the people exercises its sovereignty through the Original Constituent power and, afterwards, through the Permanent Constituent power,’ which rests in the organs created by Article 135.⁸⁵

Nevertheless, the cases have also rejected the first premise of Tena’s theory. That is to say, the Supreme Court has effectively negated the possibility of judicial review of constitutional amendments for procedural errors. Although the court has determined that the amending power is bound by the procedural limits found in Article 135, it has also concluded to have no competence to review (not even for procedural violations) a norm identified as an amendment by the relevant organs. The only entity with the competence to do so, according to the court, is the “self-controlling” constitutional amendment body itself.⁸⁶ In this sense, it would not be too much of an exaggeration to say that the Supreme Court conceives the federal congress acting together with a majority of the state legislatures in the manner in which constituent assemblies are frequently understood – that is, as final arbiters of their own actions and powers.

The court’s deployment of the concept of the PCP has thus gone beyond Tena’s own views, resulting in an amending power that is both procedurally and materially unlimited. In the words of the court:

The constituent power cannot be subject to limitations or restrictions that stem from the legal order because according to Article 39 of the constitution, the people exercise their sovereignty through the original constituent power and then through the PCP … Once constitutional provisions have been enacted by the constituent power or the PCP, that sovereignty is placed in the constitution, which is precisely what justifies the obligation of preserving constitutional supremacy because it is the latter what protects sovereignty. If sovereignty manifests through the exercise of constituent power (original or permanent), limiting its actions means limiting sovereignty, which would undermine its very essence.⁸⁷

IV. Exercising constituent power

It is now time to challenge Tena’s views about the nature of what he calls the “permanent constituent power”. In so doing, we will also show that the reasoning of the court in those cases where the doctrine of unconstitutional constitutional amendments was rejected is flawed: it does not fully consider the interplay between Articles 39–41 and 135 of the Mexican Constitution. As we noted earlier, while focused in the text of the Constitution of 1917, our argument is not exclusive to the Mexican constitutional order. The Mexican case illustrates a tension present in the liberal constitutional tradition – that the people always retain the right to give themselves a new form of government even when the exercise of constituent power is understood as a one-time event. In this respect, by focusing on the Mexican case, our intention is not only to contribute to the understanding of the PCP in that country’s constitutional doctrine but, more generally, to the theoretical question about the place of constituent power after the adoption of a constitution.

The very posing of that question, in a sense, seems to negate the one-time event approach. However, what that approach (or the exhaustibility thesis) holds is that once a constitution is established, the legal order will not provide a means for the future manifestation of constituent power, not that the people forever surrender its constitution-making jurisdiction. In other words, that since there are no legal means to exercise it, constituent power is, from a legal perspective, exhausted. The notion that the people never gives away their sovereignty has in fact been positivized in many national constitutions. It is expressed by those constitutional provisions that identify the people (or ‘the nation’) as sovereign and then proceed to explain the ways in which that sovereign power will be exercised under the established constitution.

These are provisions such as that contained in Article 3 of the Constitution of Colombia, which establishes that, ‘Sovereignty resides exclusively in the people from whom public power emanates. The people exercise it directly or through their representatives, within the limits established by the Constitution.’ ⁸⁸ Similarly, Article 2 of the Constitution of Cameroon establishes that, ‘National sovereignty shall be vested in the people of Cameroon who shall exercise it either through the President of the Republic and Members of Parliament or by way of referendum.’⁸⁹ Or consider Article 14 of the Constitution of Brazil: ‘Popular sovereignty shall be exercised by universal suffrage, and by direct and secret vote, with equal value for all, and, as provided by law.’⁹⁰ Or Article 1 of the Constitution of Italy, which provides that, ‘Sovereignty belongs to the people and is exercised by the people in the forms and within the limits of the Constitution.’⁹¹ Each of these provisions, of course, has its own history and legal status in the jurisdiction in which it exists, and every one of them could be subject to a separate study. We include them here just to show that the tension present in the Mexican constitution (constituent power as a one-time event and the people as sovereign under the constitutional order) is also present in other jurisdictions.

The equivalent of those provisions in the Mexican Constitution is found in a combination of Article 39 (‘National sovereignty rests essentially and originally in the people’) and the first clause of Article 41 (‘The people exercise their sovereignty through state organs (los Poderes de la Unión), in those cases that fall within their competences’). All these provisions identify the people as the source of sovereignty or as the entity in which sovereignty resides. However, instead of suggesting that the exercise of sovereignty only takes place when a constitution is adopted, they seek to channel it through the legal means provided by the constitution itself – either ‘directly’ (Colombia), ‘by referendum’ (Cameroon), ‘by universal suffrage’ (Brazil), ‘in the forms and within the limits of the Constitution’ (Italy) or ‘by state organs in those cases that fall within their competences’ (Mexico). The provisions that recognize popular sovereignty as the source of the constitution also reflect the background assumption present in all liberal constitutional orders that, even if the exercise of constituent power is a one-time event, the people have a right to give themselves a constitution.⁹² Article 39 of the Mexican Constitution explicitly embraces that notion, referring to the people’s ‘inalienable right to alter the form of government’.

Despite the content of Articles 39 and 41, Tena (and later the Mexican Supreme Court) thought that the only constitutional way in which the people could exercise their power of constitutional change was through the state organs authorized to act by Article 135. In presenting his argument, Tena put special emphasis on the constitutional text: the constitution placed the amending power in the federal congress and state legislatures and, at the same time, did not establish explicit limits as to the content of an amendment. The very text of Article 39, however, presents a problem for Tena’s interpretation: if the exercise of constituent power by the people is a one-time event (and, from the perspective of the constitutional order, would be forever channelled through Article 135), why does the constitutional text describe the people’s right to modify or alter the form of government as inalienable?

If an “inalienable” right is one that cannot be delegated,⁹³ then the constitution implicitly establishes a limit on the amending power (the PCP for Tena) – that is, the amending power cannot touch the form of government. Indeed, after recognizing the people’s ‘inalienable right to alter or modify their form of government’ in Article 39, the Constitution of 1917 identifies in Article 40 (in a section titled ‘Of Sovereignty and the Form of Government’) the elements that comprise it and attributes them to a prior decision of the people (acting through their representatives, as the Constitution of 1917 was not subject to popular ratification): ‘It is the will of the Mexican people to constitute themselves in a republic that is representative, secular, and federal, comprised by sovereign and free States in all internal matters.’⁹⁴ These are the same elements that Tena associated with the material constitution. If the people’s right to alter their form of government cannot be delegated to the amending authority, then an amendment abolishing federalism or establishing an official religion would be unconstitutional. It would be contrary to Articles 39 and 40.

However, an apparent weakness in this argument is that, as noted above, Article 41 of the constitution establishes that ‘the people exercise their sovereignty through state organs (los Poderes de la Unión) in those cases that fall within their competences.’ If sovereignty includes the power to alter the form of government, and if the fact that popular sovereignty is exercised through state organs is an instance of delegation, then the “inalienable” nature of the right recognized by Article 39 does not make it non-delegable. Note, however, that Article 41 does not simply establish that popular sovereignty is exercised through state organs, but also that the power of such organs depends on the specific competences assigned to each. An essential element of the act of establishing a form of government is, in fact, the distribution of competences to different institutions. That is, by creating or altering the form of government, a sovereign (i.e. someone who determines its own competence as well as the competence of others) determines which entities will exercise the legislative, judicial and executive powers, as well as who can amend the constitutional text. This means, of course, that the organs called to exercise particular competencies cannot exercise others or expand their own. It would thus be odd to say that the entities given a particular competence – that is, the state organs authorized to amend the constitutional text – are also vested (or delegated) with the competence to produce the only type of constitutional change that the constitution identifies as the people’s “inalienable” right.⁹⁵

Even if one were to accept that the state organs in which Article 135 places the amending power can exercise, by delegation, the right to alter the form of government, the inalienable nature of that right must at least mean that such delegation is always revocable. The question then would be whether Article 135 is the exclusive means of legally exercising it. Tena answers that question in the affirmative. For him, subjecting the amending power to material limits would run contrary to the principle of popular sovereignty. In other words, since the rule of constitutional change contained in Article 135 is the only legally recognized means for the people to adopt the constitutional forms it wants, any attempt to limit the amending power would directly deprive the people of their sovereignty. This is the kind of argument that, in a US context, Akhil Amar has challenged in several works. For Amar, the people’s right to amend the constitution outside the amendment rule (Article V) is the ‘most undeniable, inalienable, and important, if enumerated, right of the People.’”⁹⁶

If such a right exists, then the adoption of the US Constitution by the Philadelphia Convention in 1787 would have been legal (assuming, that is, that the convention had a special claim to act in the name of “the people”), even if inconsistent with the rule of change contained in the Articles of Confederation and with the amendment rules contained in state constitutions. In support of this argument, Amar refers (among other things) to the records of the Philadelphia Convention and to the statements contained in the constitutions of several states. For example, while some delegates at the Philadelphia Convention, such as Daniel Carroll, argued that the ratification rule contained in Article VII of the proposed constitution conflicted with the amendment procedure of their state’s constitution, James Madison maintained that the same ‘problem’ existed in all states (even in states where the constitution contained no mode of change), but since ‘the people were in fact the fountain of all power, … by resorting to them, all difficulties were got over. They could alter constitutions as they pleased.’⁹⁷

That view was also reflected in Article 7 of the Constitution of Pennsylvania (1776), which states that, ‘The community hath an indictable, unalienable and indefeasible right to reform, alter, or abolish government in such manner as shall be by that community judged most conducive to the public weal.’ Similarly, the Constitution of Massachusetts (1780) maintains in its Article VII that, ‘The people alone have an incontestable, unalienable, and indefeasible right to institute government; and to reform, alter, or totally change the same, when their protection, safety and prosperity and happiness require it.’ For Amar, these provisions (which, as the reader will note, have key similarities with Article 39 of the Mexican constitution) are not only rhetorical statements, but had legal implications at the time of the founding as well as for the current meaning of Article V of the US Constitution:

Although Article V might at first seem to be the exclusive mode of amendment, it nowhere says so explicitly. To be sure, it provides one mode of amendment, and without it, even two-thirds of Congress and three-fourths of the state legislatures would be powerless to amend the Constitution on their own. In the absence of any explicit authority from the People in the Constitution, ordinary organs of government would not have been presumed competent to alter the document on their own without popular ratification. Thus, without Article V’s express (though always revocable) delegation from the People, even an ‘amendment’ unanimously adopted by both Houses of Congress and every state legislature would have no higher status than ordinary federal and state statutes … Put another way, Article V makes constitutional amendment by ordinary governmental entities possible and thus eliminates the necessity of future appeals to the People themselves. However, future appeals to the People remain sufficient, as a general matter, to effect constitutional change … The Constitution is a set of grants of authority from the People themselves to their government agents for limited purposes. The corollary of these limited grants of power is that government agents generally possess no authority – no ‘sovereignty’ – other than that expressly or impliedly delegated by the People.⁹⁸

Unlike the US Constitution, the Mexican Constitution of 1917 explicitly embraces the view that the people retain the right to alter or modify the form of government, and not in the preamble but in Article 39. Although developed in a very different context, we think the core proposition of Amar’s argument – the continuing existence of the people’s right to alter the form of government outside of the constitution’s amendment rule – also applies in Mexico as well as in other jurisdictions that rest on a similar conception of constitutionalism. This is a view that has, for example, been accepted by some Latin American courts, perhaps most notably in Colombia in 1990.⁹⁹ However, our argument differs from Amar’s in that he seems to assume that the only way to exercise popular sovereignty (i.e. constituent power) is via a constitutional amendment adopted through some popular act outside of Article V. In contrast, and as will be argued in Part IV, we think there are other possible ways of exercising constituent power outside of a constitution’s amendment rule and that do not necessarily involve the formal alteration of the constitutional text. We will examine those alternative possibilities in the next section.

For now, the point is that the existence of those possibilities (i.e. the formal alteration of the constitutional text outside the amendment rule and informal exercises of constituent power that do not entail a formal constitutional amendment) removes the main obstacle that Tena and the Mexican Supreme Court identified for the adoption of the doctrine of unconstitutional constitutional amendments – that the legal impossibility of exercising the popular right recognized in Article 39 through means other than Article 135 would deprive the community of the (legal) ability of giving itself whatever constitution it wants. This, by itself, does not mean that there can’t be any other reasons recommending against understanding the organs operating under Article 135 as subject to substantive limits, but those reasons would not be based on popular sovereignty grounds. Indeed, if one accepts the interpretation we offer here, changes that lie beyond the jurisdiction of the amending authority would nonetheless be within the reach of the people, acting under the constitutional order. The idea that the constituent power must be understood solely as an extraordinary jurisdiction exercised during revolutions or breaks in legal continuity is thus to be rejected. Constituent power rather describes the fact that there is nothing that can permanently bind a political community to the constitution under which it lives: a people will always retain the authority to replace their constitution with a new one.

They may do this with the assistance of determinate organs that are convened to deliberate and draft a new constitution, but they never fully relinquish their authority to do so through other (legal) means.¹⁰⁰ It is in this sense that the popular right to modify or alter the form of government is inalienable: there must be ways of exercising constituent power within the constitutional order. Once constituent power is understood in this way (neither as a one-time event nor as a necessarily extra-legal power), a whole range of new (constituent) possibilities emerge. Social movements, political parties, and different groups that desire constitutional change become potential constituent actors. This approach also leads to a different (and in our view better) understanding of Article 39 of the Mexican Constitution and of similar provisions in other constitutional orders. In recognizing the people’s inalienable right to alter the form of government, provisions such as Article 39 negate the idea that the original constituent power disappears in the constitution-making act. An inalienable right is a right that, even if it can be delegated, can never be surrendered.

The idea of material limits on the amending power naturally follows from this view. In the specific context of Mexico, and as discussed earlier, those material limits would at the very least include the form of government as defined in Article 40. As noted above, the possibility of alternative means of constitutional change that could be used to produce those modifications removes the main obstacle that Tena and the Supreme Court have identified for the adoption of the doctrine of unconstitutional constitutional amendments. Once those alternative means are recognized, the idea that there is a distinction between the amending and the constituent power, and that the former is subject to substantive limits, is no longer susceptible to the typical democratic/popular sovereignty objection. In a certain way, the availability of alternative ways of exercising constituent power is a manifestation of the old view that, while the state can only do what it is authorized to do by law, individuals – and, in this case, the entire community – can do whatever is not legally prohibited. In what follows, we examine what those alternative means may look like.

V. The horizontal exercise of constituent power

The popular right to alter or modify the form of government is formally recognized by the Mexican Constitution. The prohibition of ‘rebellion’ in Article 136 has to be understood in light of Article 39 – that is, as consistent with the people’s right to change its form of government.¹⁰¹ The fact that the constitutional text does not provide a specific means for the exercise of this right outside Article 135, as Amar would argue, cannot be understood as preventing the people from exercising it through a different process: the right is an inalienable one. The exercise of this popular right outside of Article 135 would involve a change in the constitutional order that does not take place under the constitution’s amendment rule, but one that, to the extent that it is implicitly sanctioned by the constitution, cannot be accurately described as entailing a legal revolution. In this respect, the Constitution of 1917 itself contradicts Tena’s and the Supreme Court’s thesis about the disappearance of the original constituent power. What distinguishes an exercise of constituent power from other types of political practices is thus not its relationship to the constitutional order (i.e. its legality or extra-legality) but the type of change in which it results.

An apparent problem with this approach is that it seems to rest on an unrealistic premise: that the people can somehow act outside any pre-established procedure and decide to change their form of government. In a sense, this was Tena’s view: any appeal to an extra-legal constituent power had no basis in either law or reality. This is not necessarily the case if one embraces the idea that while an actual collective act of the people is not possible, different political practices will come closer to (or move further away from) it. For instance, there is always the possibility of convening, through formally correct legal procedures, a Constituent Assembly which, when compared to the Federal Congress and State Legislatures, could be seen as a superior means for the popular will to manifest. Such a process would involve a referendum (triggered by a duly enacted law) that asks the electorate whether they wish to convene a Constituent Assembly.¹⁰² An affirmative result would then trigger a special election of the members of the assembly, followed by a final referendum ratifying a new draft constitution or a radically amended one.¹⁰³

Alternatively, the amendment rule itself could be modified in anticipation of the creation of a new constitution in order to allow for the convention of a constituent assembly.¹⁰⁴ The very possibility of exercising the original constituent power through a Constituent Assembly means that Article 135 is not the sole means available to alter the constitutional text. As explained in the previous section and to the extent that such a mechanism is subject to democratic election and control, it would provide additional support for the judicial adoption of the doctrine of unconstitutional constitutional amendments. In the rest of this section, we nonetheless wish to focus on a third way in which constituent power may be exercised outside of Article 135 or, more generally, outside of a constitution’s amendment rule. Under this approach, constituent power is not associated with the act of a single, unified people, always acting either through revolution or through an extraordinary assembly. Rather, it manifests through the acts of individuals and/or social movements within the constitutional order who seek to alter, or in some cases protect, a country’s fundamental laws. These horizontal constituent episodes, as we will see below, can involve both collective and individual acts.

The possibility of the horizontal exercise of constituent power provides alternative means of constitutional change. Since it originates from the bottom of the legal order, it may also be seen as enjoying a higher level of democratic legitimacy than the ordinary amendment rule and, on occasion, of an organized act of constitution-making that takes place through an officially convened Constituent Assembly. The horizontal understanding of constituent power rests on the view that constitutional change often occurs in seemingly ordinary and sometimes invisible ways (as opposed to the traditional understanding that links it to extraordinary moments).¹⁰⁵ In our view, the Mexican Constitution, as well as similar constitutions, offers opportunities for ordinary citizens to change and re-create their fundamental laws that are not necessarily apparent. There are at least three ways in which the horizontal exercise of constituent power may take place: through individual constitutional challenges, through the exercise of rights of assembly, and through the exercise of the right to vote.

Judicial review and the responsive use of constituent power

Judicial review (specifically the individual constitutional complaint mechanisms available in many constitutional orders) may function as a horizontal means to exercise constituent power. This is true even when constitutional review by unelected judges can be subject to strong democratic objections.¹⁰⁶ That is to say, our point is not that judges engage in the exercise of constituent power, but that individual citizens may instrumentally resort to the courts to protect the constitutional order (and, by implication, the people’s exclusive right to replace the constitution) from the ordinary amendment authority. In Mexico, the individual constitutional complaint mechanism is the previously mentioned Amparo. ¹⁰⁷ According to Article 103 of the Mexican Constitution, the Amparo allows individuals to challenge the constitutionality of laws, acts or omissions that infringe constitutional and human rights recognized both by the constitution and international treaties signed by Mexico; laws or acts issued by the federal government that violate or restrict the sovereignty of the states; and state laws or acts that invade the jurisdiction of the federal government.¹⁰⁸

If the form of government, as protected by Articles 39 and 40 of the Mexican Constitution, is beyond the competence of the amending power, a hypothetical constitutional amendment that touches upon the secularity of the state or the distribution of federal/state competences would be justiciable in Amparo for being issued by an incompetent authority.¹⁰⁹ The structure of the Amparo in Mexico is in line with the interpretation we advance in this article, in that it not only gives individuals the possibility of challenging governmental actions and laws that violate rights, but also those that may affect the sovereignty of states and the jurisdiction of the federal government. That would be the case, for example, of an amendment that gives the federal executive the power to directly appoint state governors (abolishing popular elections for those offices).

Or consider an amendment that seeks to extend the parliamentary term to 20 years. That would involve an important modification to the form of government, affecting its representative and democratic nature. Suppose that an individual in Mexico, or in another country having a similar institutional arrangement, argues that such an amendment goes beyond the competence of the amending power as it seeks to do something that only the people can do – that is, transforming the form of government. Regardless of the final decision of the court,¹¹⁰ that individual would have engaged in what may be called a responsive use of the constituent power. They would be making known to the state that the amending authority has attempted to fundamentally change the constitution – that is, to exercise a competence that is exclusive to the constituent people. If the court decides in favour of the complainant, an informal yet important constitutional change would have taken place. In this case, it would mean that extending the parliamentary term beyond a certain number of years would have become ultra vires the amending authority.

This type of responsive use of constituent power is not exclusive to amendment proposals. For example, every time an otherwise unconstitutional ordinary law goes unchallenged (or the court refuses to declare its invalidity – for instance, under a political question doctrine), there is an informal (and not necessarily fundamental) change to the constitution. In this way, when an individual complaint mechanism is used to challenge unconstitutional laws, the claimants can be understood to be acting on behalf of the constituent people, protecting the constitution from the ordinary legislature.¹¹¹

Rights of assembly and constituent activities

The exercise of rights of assembly (understood here as including the right to peaceful protest and freedom of expression and association) has a close connection to constituent power. For example, they can result in the convocation of an extra-parliamentary constitution-making body. A recent example is Chile’s constituent process. One should remember the pictures and footing of the massive protests in that country that made international news in October 2019¹¹² and that triggered a political agreement that paved the way for the adoption of a constitutional amendment to regulate different aspects of a constitution-making process.¹¹³ But assembly rights can also serve as means for what we have called the horizontal exercise of constituent power. This would be the case of their exercise by grassroots movements that advocate for policy changes that could result in constitutional change, or by individuals engaged in the creation of a political party that promotes the adoption of a new constitution.

Of course, this does not mean that those political practices will always be successful exercises of constituent power: they can indeed fail to result in the desired changes. Given their close connection to popular sovereignty, it is not surprising that assembly rights are generally seen to have a fundamental importance in a democratic society.¹¹⁴ For instance, the European Court of Human Rights [ECHR] has understood freedom of assembly as a fundamental right that ‘like … freedom of expression, is one of the foundations of … a democratic society’.¹¹⁵ According to the Constitute Project, freedom of assembly is mentioned in 182 of the constitutions today in force,¹¹⁶ and is also recognized in international law.¹¹⁷ The recognition of these rights, and their potential use for the eventual alteration of a constitutional order, illustrates one of the ways in which constituent power can be exercised from within the legal order, and reflects the notion that the people does not surrender its sovereignty in the act of creating a constitution. At all times, individuals retain the freedom to organize and push to bring about constitutional change.

In Mexico, the constitution protects the right to freedom of assembly, association and protest in Article 9.¹¹⁸ A good example of the potential of bringing about constitutional change through these rights is the case of President Andrés Manuel López Obrador and his party Movimiento de Regeneración Nacional (MORENA), which came into power in a landslide election in 2018. López Obrador ran for president twice before being elected.¹¹⁹ The first two times, he ran as a candidate for the Partido de la Revolución Democrática – PRD (the traditional centre-left party). After losing a very close election in 2012 to Enrique Peña Nieto, López Obrador decided to bet on a grassroots movement that originated in 2006 (the year of his first presidential defeat) and helped found a new party – MORENA. In less than six years, López Obrador and his movement went from being a fringe party to winning the largest landslide in Mexico’s recent history.¹²⁰ One of the key factors to explain López Obrador’s historic victory is that, in a context similar to what Jack Balkin calls ‘constitutional rot’,¹²¹ he presented himself as an agent of a ‘Fourth Transformation’ that would fundamentally change Mexico’s public life in ways reminiscent of key moments in the country’s constitutional history.¹²²

To be sure, the democratic pedigree of this kind of episode is by no means insurance against a potential mandate violation by the officials called to produce the relevant changes (which could act as what the literature has identified as policy switchers and/or rent seekers).¹²³ This, however, is an empirical question that can only be answered in hindsight, after a horizontal exercise of constituent power has taken place. Our aim here is only to highlight that liberal constitutional orders may offer possibilities for the horizontal exercise of constituent power through legal means. Both the Chilean and Mexican examples are illustration of social mobilization being channelled through law: first facilitated by assembly rights, and then by the legal possibility of founding a party (in the case of Mexico) and through a constitutional amendment that facilitated the convocation of a special constitution-making body (in the case of Chile). In both cases, constituent power was exercised outside of the ordinary institutions of government (outside of what Article 41 of the Mexican Constitution calls state organs – poderes de la union) but without a break in legal continuity.

The right to vote: The electorate as constituent power

The right to vote can also serve as a means for the horizontal exercise constituent power. For the right to vote to be understood in this way, a particular idea of representation must be accepted: voters elect representatives to act in line with their policy preferences and representatives are politically bound to respect that mandate.¹²⁴ Against this backdrop, the election of representatives, frequently (and perhaps rightly) seen as a manifestation of low-intensity forms of democracy, can work as a powerful tool through which ordinary citizens can participate in processes of constitutional change. Voting in favour of a political party or candidate, in some instances, can be much more than a routine exercise of the right to vote and instead constitute an exercise of popular sovereignty as important as, for example, voting in a referendum convening a constituent assembly. History offers plenty of examples of political leaders who have been able to mobilize the electorate in large numbers under the promise of major (formal or informal) constitutional transformations.

For instance, the 1983 presidential election of Raúl Alfonsín in Argentina (1983), with a voting turnout 83.3 per cent,¹²⁵ meant the restoration of constitutional democracy after the 1976–83 dictatorship. One should recall the footing and pictures of the ‘Ahora Alfonsín’ campaign closing rally on October 1983, which is said to have gathered over 400,000 people.¹²⁶ Another example is Ecuador’s 2006 election of Rafael Correa, who won with 65 per cent of the vote (with a voting turnout of 76 per cent)¹²⁷ and who, as part of his presidential campaign, explicitly promised to call for a referendum to adopt what would eventually become the Constitution of 2008.¹²⁸ Evo Morales’ election in 2005 (elected by 53.7 per cent of the vote with a voting turnout of 84.5 per cent) in Bolivia, a key part of a process of vindication of the Indigenous peoples of that country that also led to the adoption of a new constitution, is another example.¹²⁹

According to the Constitute Project, 136 constitutions in force today contain an expression in favour of universal suffrage. Particularly in line with the idea advanced in this article, the Constitution of Brazil (Article 14) explicitly recognizes ‘universal suffrage, and direct and secret vote’ as a means to exercise popular sovereignty. An important part of the democratic legitimacy of a liberal constitutional order should be seen as resting on the fact that the exercise of rights by members of the political community, sometimes individually (as in the responsive use of constituent power through judicial review) and sometimes acting together (as in the exercise of the assembly rights and the right to vote), can in the end result in major constitutional transformations. This means that an ‘exercise of constituent power’ is not only what happens at the moment a particular decision is transformed into constitutional law, but extends to political acts that may lead to future (formal or informal) constitutional change.

VI. Conclusion

We have shown that the concept of the permanent constituent power developed by Tena and later embraced by the Supreme Court of Mexico is in tension with an idea that not only finds expression in the text of Article 39 of the Constitution of 1917, but is also a key assumption of liberal constitutionalism: that the people have an inalienable right to alter the form of government under which they live. This is a tension that is not exclusive to Mexico, but rather pervades the liberal constitutional tradition, which rejects the possibility of extra-legal exercises of popular power while at the same time recognizing the popular right to recreate the state. For Tena, under the Mexican Constitution, such a right could only be exercised through Article 135, the constitutional amendment rule. Accordingly, if the amending power was made subject to material limits, the people’s right to alter their form of government would be negated. The Supreme Court of Mexico has relied on Tena’s theory to reject the doctrine of unconstitutional constitutional amendments and has actually gone beyond Tena in determining that the amending power cannot be subject to judicial review for material or procedural errors.

In contrast, we have shown that Article 39 is in fact consistent both with the notion that congress and state legislatures can alter the constitutional text (as long as they meet the formal requirements of Article 135) and with the idea that in the act of adopting the constitution, the people retained the right to change the form of government through non-specified (and not necessarily extra-legal) means. This means, on the one hand, that the judicial enforcement of limits on the ordinary amendment power would not negate the people’s right to alter the form of government and, on the other, that not all exercises of constituent power take place outside the constitutional order, as in the case of a revolution. Constituent power can be also exercised within the constitutional order and not only through the legal convocation of a Constituent Assembly. That is, judicial review of legislation and of constitutional amendments, as well as the exercise of rights of assembly and of the right to vote, can also serve as means for what we called the horizontal exercise of constituent power: protecting the constitution from unwanted change or engaging in political acts that result in formal or informal constitutional transformations.