The subject of this book is understanding constitutional amendments, which are very significant, as well as unusual, national events. These events alter the way politics is run in the country. From the demand side, they must be modifications that cannot be addressed by ordinary legislation; from the supply side, they must clear obstacles specified by the constitution itself – namely, the amendment rules of constitutions. While some from both law and political science perspectives consider these rules fundamental (and consequential), others consider them of very little use. To the average reader, it may seem surprising that experts dispute the significance of amendment rules. If institutions are important, the institutions that regulate the change of rules should be the most important of all institutions. So, let me start by establishing this contradiction in the literature.
I.1 Are Constitutional Amendment Rules Important?
Arguing for the importance of amendment rules more than a century ago, John Burgess, a lawyer and one of the founders of American political science, dedicated four chapters of his most important book to the “amending clause” because it “describes and regulates … amending power. This is the most important part of the constitution” (Burgess Reference Burgess1890: 137; emphasis mine). This understanding is echoed by Herman Finer (Reference Finer1949), who argues that “we might define a constitution as its process of amendment” (127), as well as Richard Albert (Reference Albert2019), who states, “No part of a constitution is more important than the rules we use to change it” (2).Footnote 1 A similar position is taken by Dixon and Holden (Reference Dixon, Holden and Ginsburg2012), for whom “constitutional amendments play a number of important functions in a constitutional democracy.”
On the other side of the coin, Ginsburg and Melton (Reference Ginsburg and Melton2015), in an article entitled, “Does the constitutional amendment rule matter at all? Amendment cultures and the challenges of measuring amendment difficulty,” analyze 790 current and previous constitutions and conclude that “the institutional variables are never statistically significant and, often, they do not even have the sign one would expect” (711). In his book on constitutional amendments, Albert (Reference Albert2019) summarizes his findings this way: “I show also that rankings of comparative amendment difficulty have a fatal flaw: they either ignore or fail to account for nontextual sources of amendment ease or difficulty … I ultimately arrive at two conclusions: first, that studies of amendment difficulty are doomed to failure; and second, that in any case they may not be worth the effort” (32). Marshfield (Reference Marshfield2018) writes, “A better measure of constitutional flexibility is a constitution’s actual amendment rate because this presumably captures both the formal barriers to amendment contained in the amendment rules as well as cultural attitudes regarding formal amendment” (80). Versteeg and Zackin (Reference Versteeg and Zackin2016) agree, saying, “The measure [of constitutional entrenchment] does not rely on formal amendment rules because these rules are mediated so dramatically by political norms” (661).
I will discuss these arguments later in the book, but I want to inform the reader right from the beginning that I side with the first approach on institutions, believing that institutions in general, constitutions in particular, and amendment rules even more so certainly matter. I will explain how we can use the texts to measure constitutional rigidity. I will also show that constitutional rigidity affects the frequency of amendments, as the first side of the literature suggests (although I will slightly modify the expectations and explain why).
However, I do not stop with the argument and evidence that constitutional rigidity affects the frequency of amendments. I go one step further and consider the difference between the expected (on the basis of rigidity) and actual frequency of amendments, calling this time inconsistency. Time inconsistency is the discrepancy between the intentions of the founders and the real development of the constitution (thinking from the collective point of view of the country, it is like a change of mind over time). I then study time inconsistency further as an independent variable. So, instead of arguing that constitutional rigidity does not matter at all and replacing it with amendment frequency as some of the literature does, I use it to determine time inconsistency and explain the reasons behind why some constitutions get revised more frequently than they were designed to (that is, have higher time inconsistency than others). Finally, I argue that higher constitutional rigidity leads to higher judicial independence of constitutional courts and possibly higher independence of constitutional interpretations of supreme courts, I explain the reasons for this, and I provide empirical evidence. Therefore, the argument in the book is not only that constitutional rigidity matters but also how and why it matters.
I will start my discussion by explaining why I study institutions and why I focus on amendment rules. None of these choices are uncontroversial, and I will address different points of view in this introduction. Throughout the book, I will present a theory of constitutional rigidity based on game-theoretic principles and then corroborate it with empirical evidence from the constitutions of all democratic countries. I will present case studies as well as aggregate results so that the reader will understand that the mechanisms described in the theory are in play and that the overall expectations of the theory are corroborated.
I.2 Importance of Institutions
This book presents an institutional approach to constitutions and their amendments. It is neither the first time such an approach has been taken nor the only way that one can study constitutions. However, my particular argument regarding the significance of institutions is both novel and theoretically rigorous given its game-theoretic foundations.
Any human interaction can be studied as a game. Von Neumann and Morgenstern (Reference Von Neumann and Morgenstern1944) argue that games can be presented equivalently either in extensive or in normal form and that the normal form lends itself to more theoretical developments while the extensive form helps with the identification of the equilibria of particular games. I will use the normal form, which is composed of three elements: (1) the players participating, (2) the strategies of each player, and (3) their payoffs (when each one of them selects one strategy). Table I.1 presents a very simple game with two players (A and B), three strategies for Player A, and two strategies for Player B. At the intersection of the different strategies, the table presents the payoffs of the players (named P for Player A and Q for Player B). So, for example, P11 and Q11 are the payoffs of A and B when they use strategies A1 and B1 respectively.
Table I.1 Two-person game (Player A with three and Player B with two strategies)
| B1 | B2 | |
|---|---|---|
| A1 | P11, Q11 | P12, Q12 |
| A2 | P21, Q21 | P22, Q22 |
| A3 | P31, Q31 | P32, Q32 |
This game becomes more realistic when each player has more strategies, enlarging from 3 × 2 to m × n strategies, but it would still have two players. Note that making more strategies available to one player requires additional responses for the other and, of course, more corresponding payoffs in their intersection. In addition, one can consider adding a third player, which would require one additional dimension in our table. It would now be a three-dimensional object, and it would require more strategies for each of the previous two players and, of course, many more payoffs. Although it would be conceptually the same as a two-person game, it would be difficult to represent on a book page. One can continue complicating the interaction by adding more players, each one with their strategies, requiring more responses (strategies) from the other players and more payoffs for the possible outcomes. I have managed to present an intellectually simple but realistically very complicated situation as a game (a triplet of players, strategies, and payoffs). Yet, from this description, the word “institutions” is absent. How would the introduction of institutions modify this account?
First, institutions modify the strategies of the players. It is institutions that tell us when different players move; for example, unless there is a proposal on the floor of a parliament, there can be no vote, or unless there is a restriction of circulation for some reason, a citizen cannot be fined for being out late. Permissions and prohibitions are some simple examples, but institutions may also specify more complicated patterns. For instance, participation in an auction requires that the participants do not know the bids of their competitors (in game-theoretic terms, they move simultaneously, which is specified by the rules of the auction). This situation would be very different if some of the players knew the bids of others (that is, if they moved after the others had submitted their bids). Again, there are institutional structures that specify the sequence of moves among different players and, as a result, the strategies of the corresponding players. In cases in which we have an institutional structure that determines how exactly a procedure is going to work, the strategy space of the players is going to be seriously affected. In this book, we will analyze the amendment procedures of all democratic countries, and we will see that some of these countries describe several alternatives and endow certain actors to engage in one of the alternatives. Each one of these procedures generates a whole strategy space for the corresponding players, and the existence of multiple alternatives expands the strategy spaces even more. Chapter 2 will examine exactly how we can calculate the constitutional rigidity that procedures like these engender.
Second, institutions determine the payoffs of the actors. For example, prohibited moves are associated with negative payoffs (i.e., the penalties incurred when one violates the prohibiting rule). Similarly, there are “incentives” for desirable actions, helping players make the corresponding choices. If the payoffs are well-known among all the players, they can be used as hints about the effort they will undertake in order to achieve the desired outcome. If the first prize in a competition is accompanied by a ten-million-dollar check, the competition will become more fierce. One can anticipate how hard a politician will try to reach a particular goal if the achievement is a major requirement for his reelection strategy, and so on.
Third, institutions can specify the players of the game. It is not the case that every individual is entitled to participate in every game. State regulations decide whether only voters registered in the corresponding party can participate in the Republican and Democratic primaries. Laws that specify that every elementary school instructor must report bodily injuries on a child affect the participants in the injury protection game with the goal of impacting the final outcome of the game (reducing injuries). A similar strategy was used by the state of Texas for the prevention of abortion by enabling any person that observes any facilitation of abortion to report it. Again, an expansion of players will have dramatic effects on the equilibrium outcome of the game.
Clearly, institutions affect each of the items that determine the triplet of a game: players, strategies, and payoffs. Given that every interaction that involves this triplet is a game, I have demonstrated that institutions affect every human interaction in multiple ways.
Of course, I did not demonstrate that institutions are the only factor that affects human interactions. Other candidates could be preferences, ideologies, interests, cultures, and so on. I will consider the effect of the most general of these concepts: preferences. I consider it more general than the other concepts because it is more systematic and less idiosyncratic. People have tried to provide theories about collective organizers of preferences like interests, ideologies, and even cultures. The analyses cover how these concepts form and what their effects are; for example, workers can be argued to have more progressive ideologies, and they may prefer left-wing politicians. However, none of these arguments is expected to work all the time and determine the preferences of any particular individual.
In Chapter 3, I will address some cultural arguments that have been proposed to explain constitutional amendments and explain their insufficiencies. In Chapters 4 and 5, I will analyze specific cases of failed amendments (Italy and Chile) and successful amendments (Mexico) as my alternative approach to grafting politics to the institutional framework of the book.
I.3 Structure of Institutions
Although individuals may be located in different institutional environments at different times or places (they may, for example, operate under different rules in their family vs. at school or when they are at work vs. during the weekend), they are subject to the laws of their country at all times and, in addition, they can conform to the requirements and enjoy the benefits enumerated in their constitution. The laws of the country are expected to be in agreement with the constitution, and if they are not, they are invalidated by court decisions. Every constitution specifies the criteria that any legislation has to satisfy in terms of the procedure that engenders it and in terms of the general content that it should (or should not) have.
The fact that the constitution describes the conditions under which different laws or other types of decisions are valid means that it generates a constitutional equilibrium. The term equilibrium is game theoretic and implies that all the actors involved are selecting among their strategies the ones that are best responses to each other – a concept introduced by J. Nash (Reference Nash1951). As a result, no actor would modify their action if they had the opportunity.
As an example, the French Constitution (Article 34) specifies the areas that the Parliament can legislate. Any area not included in this article does not belong to the jurisdiction of the Parliament but to the government. This provision is unusual from a comparative perspective where parliaments have the authority to legislate on any issue of their choice. But, as a result, the French constitutional equilibrium is different from that of other countries in this respect. Similarly, different constitutions specify under what conditions war is declared (is it the jurisdiction of a president? Of government? Of parliament?) a referendum is proclaimed (who asks the question? Who triggers it?), and so on. It is obvious that different constitutions generate different constitutional equilibria.
From the point of view of this book, it is interesting to identify the formal conditions that every piece of legislation has to fulfill according to the constitution. For example, the constitution may define a whole hierarchy of laws.Footnote 2 If this is the case, there are different constitutional equilibria depending on the required majorities for different levels of laws. In addition, the constitution may generate different tiers of constitutional articles: It may prohibit the alteration of certain articles,Footnote 3 and it may make some articles more difficult to modify than the rest of them.Footnote 4 In each one of these cases, a different type of constitutional equilibrium is generated because of the difference in the regulating institutions.
Having established what a constitution is and how it generates a constitutional equilibrium, I now turn to the question of constitutional change. In order to simplify things, I will divide the issues into three levels.
At the first level, there are modifications inside the constitutional equilibrium. The constitution specifies rules for how each particular game is to be played. How is a law to be voted for? What happens if a particular law is defied? Who decides if there was a violation of the legal order? How does the nation go to war? How does it stop the war engagement? For some of these questions, the constitution may not provide specific answers, but it may delegate the decision to particular agents. For example, it may include the phrase “according to the law,” inviting the legislature to make more specific decisions, or the phrase “according to an executive order,” giving jurisdiction on the subject to the executive. The two most important delegations that the constitution makes “in equilibrium” are for legislation (for the legislative procedure to be engaged) and for judicial interpretation (for the judiciary to apply the laws [statutory interpretation] or the constitution [constitutional interpretation] in particular). It goes without saying that every constitution specifies the exact scope and rules for these procedures.
At the second level, there are modifications outside the constitutional equilibrium (amendments). The constitution uses the provision of amendment rules in order to specify how specific provisions may be replaced. Amendment rules may divide the articles of the constitution into different categories, or it may not. This is a significant distinction for what follows. The constitution may specify that certain articles cannot be amended or can be amended by a more stringent procedure than the rest of the constitution. If there is only one amendment procedure, then this procedure has to be followed any time the constitution is considered inappropriate or insufficient. If there are multiple amendment procedures, then it has to be certified that the appropriate amendment procedure is used. Usually, it will be the judiciary that is enabled explicitly by the constitution or by convention to decide whether the procedure used is the appropriate one. This will be a procedural decision: The court will decide that the specific modification cannot occur because the article is unamendable or belongs to a more stringent amendment category than the one classified by the political system; therefore, the amendment procedure in use is not appropriate.
At the third level, there is the possibility of the replacement of the whole constitution. This may occur under a diversity of conditions, such as through a general agreement (e.g., recently in Chile), under a popular revolution (e.g., the French Revolution of 1789), or because of a military coup (e.g., the Greek colonels in 1967). The most interesting part of this category is that we do not know the payoffs of the different actors; therefore, we cannot describe it as a game until long after it is over. For example, the military coup in Greece failed after seven years, and democracy was restored in 1974. The leaders of the coup went from running the government to being in jail for life. Were the seven years in power worth life imprisonment? To determine this, we will argue whether the seven years of interruption of the democratic regime was a long or a short period, and, according to the answer, we will make a different assessment of the payoffs and analyze the outcomes.
Table I.2 summarizes the argument. I want to emphasize here that the distinctions I am making are procedural and not substantive. It may be that a new constitution is a replica of an older one, but the procedures followed were not specified by the previous constitution. For example, Chile recently underwent two failed attempts to replace (as opposed to amend) the Pinochet Constitution through constitutional conventions (a procedure not included in the constitution).Footnote 5 The same logic applies to amendments, the significance of which varies from extremely important (e.g., slavery in the US) to completely insignificant (e.g., most constitutional amendments in Mexico). We will divide the amendments according to significance in Chapter 2, but here we will focus solely on their rules of adoption (they all followed the amendment rules of the corresponding constitution – that is, they belong to Level 2 of Table I.2).
Table I.2 Three levels of rulemaking
| Level of rulemaking | Types of actions |
|---|---|
| 1. Inside the constitutional equilibrium | Legislation Statutory and constitutional interpretation by courts |
| 2. Outside the constitutional equilibrium (but within the constitution) | Constitutional amendments (respecting amendment rules if these rules are uniform and there is no distinction among articles) Unconstitutional constitutional amendments (if there are layers of amendment and the rules of amendment are not respected) |
| 3. Outside the constitution | New constitution |
Cases inside the constitutional equilibrium also vary in significance. Just like judicial interpretation, legislation may be significant or it may not be. For example, the constitutional interpretations of abortion (Roe v. Wade) or desegregation (Brown v. Board of Education) are more significant than most laws the political system of the US has produced. However, they are still constitutional interpretations – that is, they move inside the constitutional equilibrium. Similarly, Obamacare and Trump’s tax cut law are much more significant than other laws produced by the US political system, but they are still laws – that is, they are decisions within the constitutional equilibrium.
Finally, an example that moves from outside of the constitutional equilibrium to inside of it is the issue of women’s rights. Some fifty years after the original introduction of the Equal Rights Amendment (ERA) in 1923, both houses passed it with the required two-thirds majority in 1971 and 1972. However, the amendment never cleared the hurdle of approval by three-fourths of the states as required by Article V of the US Constitution. Because of this, there has been no ERA amendment in the US Constitution. On the topic of equal rights, Lilly Ledbetter was hired by Goodyear as a supervisor and was receiving, without knowing, a significantly lower salary than her male counterparts. When she found out, she sued the company for sex discrimination, but she lost her case in the Supreme Court because the legal limit to present the case was 180 days from the date of the discriminatory case (the day she received her first paycheck). The statutory interpretation of the Supreme Court was correct and demonstrated a movement within the constitutional equilibrium. In response, the first legislative act of the Obama administration was the Lilly Ledbetter Fair Pay Act, which enabled an actionable complaint within 180 days of any discriminatory act (therefore, the last paycheck received) and, consequently, enabled the plaintiff to win. This piece of legislation contradicted the Supreme Court decision and also represented a move within the constitutional equilibrium. We will see in subsequent chapters that the court can overrule the political system by issuing constitutional interpretations, which are also movements within the constitutional equilibrium. The latest such constitutional interpretation was the decision about presidential immunity. Despite the fact that, in the opinion of the court minority, many constitutional scholars, and the author of this book, this interpretation significantly modifies the equality under the law principle (that is, the common understanding of the constitution), it cannot be modified by legislation.Footnote 6
In all discussions, what qualifies as law, statutory/constitutional interpretation, or amendment is defined alone by the procedures followed rather than by the significance of the outcome. In other words, I follow the procedure – the decision-making rule – with the same respect, dedication, and tenacity that a prosecutor or an economist follows the money in order to understand the motives of different actions. Throughout this book, I focus on Level 2 of Table I.2 – that is, constitutional amendments. I consider the totality of amendments without confusing them with judicial interpretations or legislation, no matter how significant these latter decisions are. In addition, I independently assess the significance of different amendments and study the adoption of insignificant, significant, and fundamental amendments as a function of the amendment rules. This is a defining characteristic of this book since the procedures determine the institutions that are involved in the analysis as well as the decisions that are taken. I will refer to this approach as “follow the decisions.” No matter how significant a law is (e.g., the Franklin D. Roosevelt legislation or the Ledbetter Fair Pay Act), it is not a constitutional amendment, and no matter how insignificant a constitutional amendment, the reason that its content was adopted through the (more difficult) constitutional amendment procedure is that legislation to that effect would have been considered unconstitutional. Each one of these decisions follows a different procedure, and that is why they have a different name. I will follow the decisions.
I.4 Constitutional Moments and Unconstitutional Constitutional Amendments
The importance of following the decisions (that is, of defining terms on the basis of procedures rather than substantive significance) is not universally recognized. There are two different procedures that constitutional lawyers consider as being fundamental extensions of constitutional law: the first is the theory of “Constitutional Moments” presented by B. A. Ackerman (Reference Ackerman1991), and the second is the theory of “Unconstitutional Constitutional Amendments” whose birth is currently in progress (Dixon and Landau Reference Dixon and Landau2015, Yap Reference Yap2015, Roznai Reference Roznai2017, Dixon and Uhlmann Reference Dixon and Uhlmann2018, Torres-Artunduaga and García-Jaramillo Reference Torres-Artunduaga and García-Jaramillo2020). Here is how Albert puts it:
Briefly stated for now, Bruce Ackerman’s theory of constitutional moments uncovers the dualist foundations of the U.S. Constitution to show how leaders have transformed constitutional meaning without a corresponding alteration to the constitutional text, while the basic structure doctrine, first articulated by the Indian Supreme Court, enforces implicit limitations on the power to amend the codified constitution. Each idea has disrupted how we understand the forms and functions of constitutional amendment, each has caused us to rethink the very meaning of constitutionalism and how it translates democracy, legitimacy, and sovereignty into law, and each continues to generate important scholarship critiquing, applying, and extending it both inside and out of the domestic context from which it emerged.
I.4.1 Constitutional Moments
According to Ackerman, besides the “classical” system of amendment determined by the constitution (which is institutional and formal), there is also a “modern” system (which is revolutionary and informal). He states, “Here the decisive constitutional signal is issued by a President claiming a mandate from the People. If Congress supports this claim by enacting transformative statutes that challenge the fundamentals of the preexisting regime, these statutes are treated as the functional equivalent of a proposal for constitutional amendment” (Ackerman Reference Ackerman1991: 268; see also Amar Reference Amar1994). This “modern” system bypasses Article V of the US Constitution. According to Ackerman, this informal procedure has been used in adopting the constitution, the Fourteenth Amendment, and the New Deal legislation.Footnote 7 I do not know what inferences can be made from unique historical events like the first two, but the adoption of the New Deal legislation is a matter of constitutional interpretation by the Supreme Court, which was evaluated first as unconstitutional and subsequently as consistent with the US Constitution.Footnote 8
The court is permitted to make a constitutional interpretation no matter what its position may be while staying within the constitutional equilibrium (Table I.2). The fact that it moved towards progressive positions also does not mean it will do that all the time or even ever again. The Dobbs decision on abortion in 2022 showed that the court moved against the political system (since all three legislative actors were democrats) as well as the will of the voters. Consequently, what happened during the New Deal was not a constitutional amendment but a change in the opinion of the Supreme Court with respect to the constitutionality of the New Deal legislation. The reverse occurred with respect to abortion rights when the court changed its interpretation in a conservative direction. Whenever such a change occurs (regardless of the reasons), we have a different constitutional interpretation, which is a move within the constitutional equilibrium. Therefore, it is not a constitutional amendment.
This is not the only blow that the extraconstitutional procedure of constitutional moments suffers from. The fact that the ERA failed to pass because of not satisfying the requirements of Article V indicates that the proposal presented by Professor Ackerman cannot be the basis of a constitutional amendment: if there are several procedures and one of them is easier than the others (because of the omission of the approval of three-quarters of the states), then no political actors would ever use the difficult procedure again in order to change the constitution.Footnote 9 Consequently, if there were alternative procedures (as Ackerman claims), ERA proponents would have never used Article V and thereby failed.
An alternative interpretation of Article V would be that it is an insurmountable obstacle as opposed to one that can be bypassed. Stohler et al. (Reference Stohler, Bateman and Woodward Burns2022) argue that Article V has become an instrument of constitutional position-taking, presenting the almost 12,000 unsuccessful amendment proposals that have been made in the history of the US as evidence. This position respects the rigidity of the US Constitution, exploiting historical analysis.
1.4.2 Unconstitutional Constitutional Amendments (UCA)
The term unconstitutional constitutional amendments indicates that a constitution constrains the people’s constituent power. Some of the amendments undertaken may be against the constitution. There may be two different reasons for this: the first is procedural (it may be that some of the specified conditions were not fulfilled – for example, a required referendum may have been fraudulent), and the second may have been a matter of content (for example, there may have been violations of eternity clauses in the constitution). Under these conditions, the court can judge that a particular amendment violates the eternity provision of an article (for example, if the regime type is protected by an eternity clause, then significantly modifying the powers of an institution may be a violation of this clause). If an amendment is prohibited preemptively on the grounds of unconstitutionality, the move is within the constitutional equilibrium; if it passes and is invalidated later by the court, then we are in the area outside the constitutional equilibrium. Germany, South Africa, and India are countries which are contemplating the case of unconstitutional constitutional amendments and are enabling the courts to make a judgment (Albert Reference Albert2009).
However, arguments have been made about courts making unconstitutional constitutional amendment decisions even if the constitution does not include appropriate provisions. For example, in Colombia, Alvaro Uribe, who was elected as president in 2002, tried to change the constitution that limited presidents to serve only one term in office. Because Uribe maintained a very high approval rate until the end of his presidency term, he sought and received approval of a constitutional amendment that allows presidents to serve two consecutive terms. While the amendment was challenged both on procedural grounds and as an unconstitutional “substitution of the constitution,” the Constitutional Court upheld the amendment (Dixon and Landau Reference Dixon and Landau2015). However, when Uribe sought another extension for the presidential term limit – to serve for three consecutive terms – the court blocked the attempted constitutional change. In this second re-election case, “the majority of the court struck down the proposed referendum, both on the grounds that the procedures for approval had been unconstitutional and on the grounds that the amendment constituted a substitution of the Constitution” (Dixon and Landau Reference Dixon and Landau2015).
Legal theorists consider unconstitutional constitutional amendments to be a way of preventing anti-democratic or “abusive” constitutional rules (Dixon and Landau Reference Dixon and Landau2015). Dixon and Landau identify conditions under which such a decision, which invalidates the decisions of elected representatives by non-elected courts, is permissible or, indeed, necessary (such as in the case of Colombia).Footnote 10
Similarly, Ginsburg and Huq (Reference Ginsburg and Huq2018) argue that nonelected officials such as judges play a key role in protecting the democratic order from erosion. However, there is no guarantee that judges will have the interests of democracy at heart. For example, in Honduras, under extreme pressure from the Honduran president, the Constitutional Court ruled that the presidential term limits were unconstitutional on a flimsy pretext, claiming that they violated international human rights (Cassel Reference Cassel2009). This argument was made despite the Honduran Constitution being written explicitly to preserve presidential term limits and to prohibit any changes to the constitution to remove these limits. Indeed, there was a specific clause in the constitution that presidential term limits were strictly unamendable (Landau et al. Reference Landau, Dixon and Roznai2019a). In Bolivia, the Constitutional Court used the same pretext to rule in 2017 that the clause preserving term limits in the Bolivian Constitution was unconstitutional despite the fact that, only a year before, an attempt to amend the constitution to remove presidential term limits failed in a referendum (Wolff Reference Wolff2020).
The countries where this doctrine becomes very significant are the ones where democracy is at stake. In these cases, legal theorists are trying to both protect democracy and limit the power of the judges. For example, Dixon and Landau (Reference Dixon and Landau2015) introduce the narrowness of the doctrine and transnational constitutionalism as ways to limit the judiciary’s overuse of power. Narrowness can be achieved either by identifying and protecting a small set of particular institutional provisionsFootnote 11 or by ensuring only a narrow set of constitutional provisions are unamendable. Transnational constitutionalism uses transnational constitutional law as a reference for judicial consideration of institutional practices and jurisprudence.
Both ideas represent a fundamental change in the way we think about constitutions, but, in my opinion, they are not valid. Both ideas violate the three different levels of analysis of political and legal change discussed in Section I.3. They confuse decisions made inside the constitutional equilibrium (legislation and judicial interpretations) with decisions made outside this equilibrium (constitutional amendments), and they create confusion where there is a need for clarity.
I.5 How to Bypass Constitutional Moments and Unconstitutional Constitutional Amendments
I have now presented legal arguments that transcend the distinction between actions inside (legislation, judicial interpretation) and choices outside the constitutional equilibrium. In this book, I will maintain this distinction and will analyze only the choices outside the constitutional equilibrium – that is, the constitutional amendments.
I.5.1 The Written Text
First, I will base my whole analysis on the written text of the constitution of the countries I will examine due to the importance of this text. Second, in keeping with the principle of scripta manent (written words remain), I will avoid studying the historical conditions under which the amendments were accepted as historians may disagree about the events and the legitimacy of the procedures. For example, while there is debate in the US about the conditions under which the Fifteenth Amendment was adopted, nobody disputes that it was adopted and that it is very significant. There is a similar event in France: The Constitution of 1958 prescribed an indirect mode of election for the president of the republic. In 1962, General De Gaulle asked the people to approve his plan for direct election of the president by referendum. He did it on his own, violating Article 11 of the constitution, which requires a “recommendation by the government” or “a joint motion of the two houses.” An overwhelming majority of French constitutional scholars declared the initiative unconstitutional. However, De Gaulle proceeded with his plan, and the referendum was approved. All the critics reversed their position with the argument that the people erased any irregularity through their approval.Footnote 12 If historical questions about the legitimacy of adopted procedures exist for the US and France, they probably exist for many other countries. However, nobody disputes the text of the constitution or the success of the amendment (or its significance). As a result, the text of the constitution rules.
I will come back to this point in Chapter 2 to develop a theory that will translate the constitutional text of different countries into an index of constitutional rigidity and in Chapter 6 to defend this method against arguments that claim “in any case they may not be worth the effort” (Albert Reference Albert2019: 32). Here, though, I want to defend the written legal text as an achievement of humanity, going back to all monumental forms of civilization, from Hammurabi, to Moses, to Lycurgus in Sparta, and to Solon in Athens. Constitutions are the fundamental form of law in all contemporary societies. Even the people who consider them flawed and who disagree with important parts of the text agree that the only way to change them is through constitutional amendments (that is, working outside the constitution, but applying the rules of amendment themselves). For example, it may be disappointing that the US Constitution does not respect the fundamental democratic principle of “one person, one vote” for presidential elections. However, nobody would claim that we could respect this principle without changing the constitution. A reference to Dean Ely comes to mind here: “A neutral and durable principle may be a thing of beauty and a joy forever. But if it lacks connection with any value the Constitution marks as special, it is not a constitutional principle and the Court has no business imposing it. I hope that will seem obvious to the point of banality” (Ely Reference Ely1973: 949).Footnote 13
I.5.2 Focus on Democratic Countries
The reason that so many constitutional scholars go beyond constitutions and consider their own theories and principles legitimate is that they care about democracy and they see democratic values being disputed or even trespassed. In other words, in many countries, the constitution itself is not respected, and what is quite clearly written and generally understood (like the impossibility of standing for a presidential election twice in some countries) is violated by the relevant actors. This is the reason for searching for other principles as well as other authorities (such as foreign organizations or powers). I respect the intentions of these researchers as well as their efforts to create consistent theories. Still, I will adopt a different method in this book. Throughout, I will restrict the countries I study to democracies because in these countries the rule of law prevails and, as a result, the constitution is respected. Consequently, the written text is relevant and applicable.
An example of the democratic intentions of scholars focusing on nondemocratic countries is the work of Sadurski who, in studying countries of Central and Eastern Europe, comes to the conclusion of “relative insignificance of formal constitutional design” in one article (Sadurski Reference Sadurski2020a: 324) and that “formal institutions must be underwritten by norms which are by-and-large shared, and by common understandings about what counts as a norm violation, even if formal legal rules are silent about it” in another (Sadurski Reference Sadurski2020b: 59). Such conclusions/suggestions are not included or discussed in this book because of the focus on democracies.
As a factual example of disrespecting a constitution, consider the case of the military coup against Honduras in 2009 when President Manuel Zelaya was arrested and subsequently exiled by the Honduran military. Zelaya had been attempting to change the constitution to allow him to run for a second term as president by appealing to the people in a consultative referendum on whether to hold a constitutional convention. In the Honduran Constitution, this was not a formal avenue for constitutional change, and the Honduran Supreme Court issued an injunction to make him halt his attempts. Zelaya continued his efforts despite the injunction, and a group of Honduran elites, including the Supreme Court and the military, conspired to arrest him and deport him to Costa Rica. In this situation, neither side respected the constitution. Zelaya wanted to change the constitution illegally, and the Supreme Court, despite its claims that it did what it did to protect the constitution, did not respect the constitution when it conspired with the military to have him removed (Sosa Reference Sosa2015).
From a comparative constitutional point of view, if nondemocratic countries are included in the study, there is substantial noise that contaminates the results, so their reliability declines. One of the first experiments in statistics was conducted by Student, who, when asked whether milk was beneficial for infants, compared a small number of twins, one of which was offered milk and the other not, and claimed that his results were more reliable than other researchers who were comparing thousands of infants without being able to control for intervening factors (Student 1931). Whenever we can, we should also try to find “twins.” Obviously, in this book we are not performing an experiment like Student, but we are selecting countries that respect their constitutions – that is, democracies – when we study constitutional amendments in as much of an approximation as possible.Footnote 14 Therefore, this book will focus on the amendment texts of democratic countries and a series of consequences that these texts have for the political life of the countries.
The distinction of different changes occurring inside and outside the constitutional equilibrium that I follow in this book highlights the logic the existence of a constitution imposes on these changes. Instead of calling what most people call constitutional interpretation an “amendment,” or instead of calling an extraconstitutional interpretation constitutional, I use the existing terms and identify the reason for their use. This way, when I use existing datasets of constitutional amendments, I do not need to reevaluate them country by country (an enterprise appropriate for historical accounts) on the basis of new interpretations.
The restriction to democracies also affects the comparative part of the book, which uses statistical analyses. In the descriptive part of the book, when I present specific institutional structures or historical accounts, I am going to extend the universe of cases and speak of events that took place regardless of whether they happened in democratic countries or not or even whether they occurred in countries, states, or supranational organizations like the European Union. Finally, I will also not be restricted by time limits. In all cases, I will provide all the necessary information. These selection rules are to permit the necessary cases to be included if they are interesting and relevant to the issue under discussion.
I.6 Outline of the Book
This book is organized into eight chapters, beginning with a descriptive (Chapter 1) and theoretical (Chapters 2 and 3) account. Then, it moves to an individual country approach (Chapters 4 and 5) that focuses on the amendment provisions of several countries in detail in order to understand the institutional mechanisms proposed in the book. Finally, it moves to a comparative analysis in Chapters 6–8 where I provide evidence that constitutional amendment mechanisms define constitutional rigidity, which, in turn, decreases the frequency of constitutional amendments (Chapter 6), that these mechanisms affect time inconsistency of constitutions – that is, how often they are amended despite locking provisions (Chapter 7), and that constitutional rigidity is a necessary condition for the significance of the judiciary of a country (Chapter 8). I will now describe the contents of these chapters in more detail.
Chapter 1 provides descriptions of different amendment provisions, which range from simple ones, such as that a constitutional amendment can be decided by a simple majority in the parliament of a country, to more complicated provisions, such as when an amendment requires approval in both chambers of the parliament of a country by qualified majorities. Some other provisions discussed in this chapter include even more composite ones, such as parliamentary approval and referendum approval, or alternative paths, such as either qualified majority in parliament or simple majority and referendum. In addition, there may be time limits implemented to wait for certain procedures or to not exceed in others, double passages with or without an intermediate election, quorum requirements, and other provisions. All these procedures are so complicated that different actors may have disagreements as to which is appropriate, and they may be fighting for their opinions, or, even more frequently, they may be modifying the institutions in order to ensure their opinion prevails. All these descriptions are meant to persuade the reader that amendment procedures are not only very diversified across countries but are also very influential since the people involved disagree about them.
Chapter 2 presents the theory of the book and explains how this variety of institutions affects the constitutional rigidity of a country. I use two theoretical concepts: the win-set of the status quo (the set of outcomes that can defeat the status quo) and the core of the constitution (the features of the constitution that cannot be changed given the amendment rules and the preferences of the actors). These two concepts help us classify the different amendment provisions in terms of their difficulty since the constitutions that are easier to amend will have a smaller (or empty) core and a larger win-set of the status quo. Special attention is given to the institution of referendums for both theoretical and empirical reasons that will become obvious later on. This classification enables us to construct an index of constitutional rigidity that will be used in the empirical part of the book. The expectation is that higher constitutional rigidity will not only produce a lower rate of amendments but also a lower variance of this rate. Indeed, when a country has high institutional rigidity, amendments are very difficult, and there will be very few of them, but if it has low rigidity and amendments are possible, it does not mean that they will happen frequently.
Chapter 3 addresses different cultural theories that are presented as alternatives or complementary to my institutional approach. I argue that there are three different approaches that deal with constitutional amendments: one uses cultural analyses as idiosyncratic forces, one as a systematic omission, and one as an independent variable. I provide different reasons for why each approach needs further elaboration before they can be used in a comparative perspective.
Chapters 4 and 5 provide a concrete institutional answer to the cultural theories presented in Chapter 3. They demonstrate that political analysis complements the broad brush of comparative constitutional analysis provided in Chapter 2.
Chapter 4 analyzes the institutions of two different countries (Italy and Chile) that recently had either constitutional amendments or rejections of the whole constitution. First, I analyze their institutions and explain why they have high rigidity before focusing on the final step of the modification (whether amendment or whole constitution rejection) process, which involved a referendum. I then analyze the public opinion that led to a NO vote and present an institutional improvement to constitutional referendums, which are likely to become more frequent.
Chapter 5 examines the opposite situation (from Chapter 4), focusing on a case where constitutional amendments are very frequent: Mexico. I analyze the institutions in more detail and explain why Mexico in particular has less-rigid institutions than expected when applying broad comparative rules. I then show that the political conditions of the country have created oversized coalitions that clear all constitutional obstacles for amendment.
Chapter 6 applies the constitutional rigidity index presented in Chapter 2 to all 103 democratic countries. I argue that higher institutional rigidity produces not only lower amendment frequency but also lower variance, as anticipated in Chapter 2. I divide the over 900 amendments into three different categories, examine them according to their significance, and show that the higher their significance, the more important the institutional rules are for their adoption. Chapter 6 is the empirical demonstration that constitutional rigidity is significant in determining the frequency of amendments and that its significance increases with the importance of the amendment, unlike the minor importance or irrelevance that other analysts have claimed.
Chapter 7 uses constitutional rigidity as the basis for developing a measure of time inconsistency. Instead of adopting the idea proposed in the literature that amendment frequency depends on culture and is unrelated to constitutional rigidity, I use the difference between the actual frequency of amendments and the one expected on the basis of constitutional rigidity, calling this time inconsistency. I demonstrate that this difference depends on the size of the constitution and that longer constitutions are more time-inconsistent (locked and amended more frequently) than shorter ones. As a result, I expect longer constitutions to be more restrictive and have more negative economic outcomes than shorter ones – expectations that are then empirically confirmed.
Chapter 8 analyzes the role of the judiciary. I argue that the judiciary will have low independence when constitutional rigidity is low because it will not be able to interpret the constitution without fear of being overruled. However, high rigidity may generate high independence because, in this case, the judiciary will not be concerned about being overruled. Therefore, in case it disagrees with the executive, it will strike down existing legislation (exactly like how constitutional rigidity affected the frequency of amendments in Chapter 6). I expand on this idea because frequently in the social sciences relationships are presented as necessary conditions alone, and, therefore, the statistical tests should be different than the standard ones.
Open access