Delineating Eternity Clauses

Eternity clauses are provisions in the constitution that impose substantive immuni­ties against the amendment procedure.^[273] They are only by name eternal, inasmuch as they were introduced by the constituting will and the reconstitution of the consti­tuting will, i.e., constitution-making, is required for their content to be altered.^[274]

Moreover, they are also not eternal inasmuch as they are subject to imperfect enforcement.

Looking into the existing classification schemes, three contributions can be found that are specific to absolute amendment constraints.^[277] First, eternity clauses may be divided according to which provision they entrench. If different charac­teristics of the constitution influence the life of citizens and politicians differently, then in order to understand how they will be adhered to one needs to know their content. Friedman proposes such a division.^[278] Specifically, he divides eternity clauses to entrenchments of the character of the government, the spirit or principles of the constitution and the character of the country.

Second, Schwartzberg by looking into unamendability from the perspective of the difficulty of amendment and time proposes four categories of entrenchment, namely formal, time-unlimited (eternity clauses) and time-limited (sunset clauses) entrenchment; de facto (prohibitive amendment process) entrenchment; and implicit entrenchment.^[279] This categorization also suffers from inconsistencies. Specifically, the same amendment constraint could in one legislative period lead to de facto entrenchment, when in a next period due to changes in the power of the winning coalition does not.

Lastly, Albert proposed two classifications.^[280] The first scheme is based on two dimensions, namely the goals of the constraint and the degree of the constraint. Like the Friedman scheme, unamendability is broken down with the use of subjective valuation into preservative entrenchment, transformational entrenchment, and rec- onciliatory entrenchment, according to its goals.^[281] Moreover, according to an objective measure, i.e., the degree of constraint, it is divided into three stages, namely conventional, heightened, and indefinite entrenchment. The second scheme is refining the same temporal and formal characteristics of Schwartzberg, but also adds qualitative characteristics (substantive and procedural entrenchment).^[282] This scheme is very precise and utilizes objective criteria, which could facilitate legal discourse. Nevertheless, it suffers from the extent of its complexity.

The merits of the scheme proposed here are multiple. It captures all absolute constraints, namely both in the big-C, i.e. the constitution-as-form, as well as the small-c, i.e., the constitution-as-a-function without increasing complexity.^[283] It is based on objective criteria and not the aspirations of the framers, which are to a great extent unobservable.

The first criterion of Table 1 asks whether constraints in the amendment process are explicitly included in the constitutional master text. This dimension captures to a great extent the saliency of the amendment constraint. First, the writeness of formal clauses makes them significantly less volatile or interpretable in their application compared to informal constraints. Moreover, since even nonexperts can

know the content of a codified constitution and perceive it as a unity, such con­straints will be more costly to disregard. In that sense, it refines the formality criterion, to the extent that it is interested only in constitutional formality. The second criterion looks into whether the constraints imposed poses an absolute hurdle, i.e., a hurdle which can be overcome only with the reconstitution of the constituting will, or a limited hurdle, which is lifted after certain conditions are fulfilled. Absolute constraints make the option of a nonrevolutionary change unavailable. Subsequently, they directly increase the benefits of revolutionary change. Limited constraints, on the other hand, simply increase the cost of non­revolutionary change.^[284]

On the first cell of the matrix, we find true substantive constraints. They are the taboos of the constitutional order. They are provisions that are absolutely entren­ched themselves, i.e., insusceptible to a double amendment procedure.^[285] Even though, rank-ordering the difficulty of changing constitutions is a mean task and many have given up,^[286] it seems straightforward to consider self-entrenched eternity clause as the strictest conceivable constraint.

On the bottom left cell, we find ultra-constitutional substantive constraints. One vehicle to introduce unamendability is through the lack of an amendment proce­dure. According to contractarians, there is a distinction between those consenting and those that are enabled to act within their capacities because they acquired consent. The former is the constituting power and the latter is the constituted power.^[287] In this setting, the power of the constituted actors to amend is not self-evident. On the contrary, amending the constitution is a bridge built by the constituting power to allow for adaptations by the constituted powers, i.e., it is a delegated constituting power.^[288] Subsequently, the lack of an amendment procedure is to be interpreted as no power to amend. However, legal interpretation has allowed even constitution with no amending formula,^[289] to get amended. Just to give an example the British North America Act of 1867, which outlined the system of government in Canada, contained no amendment procedure. Nevertheless, it was amended at least 20 times with various acts until 1982, when an amendment pro­cedure was introduced. In other words, the lack of a procedure only through interpretation leads to an amendment constraint. Subsequently, constitutions which include no amending power and constitutions which prohibit amendment are not to be treated the same.

Following that point, even when constitutional texts remain the same, imple­mentation allows the organic instrument to be adapted to the needs of times.^[290] Constitutions change also implicitly or unlawfully.^[291] However, to avoid overuse of this option, framers introduce independent formal institutions that are entrusted with the task of guarding constitutionalism or allow for the emergence of political conventions.^[292] Such constraints are driven by internal institutions^[293] in the sense of institutional economics or the amendment culture.^[294] Weintal based on the severity of the constraint proposed bundling absolute constraints created by the judiciary and constraints introduced by the constitution.^[295] Nevertheless, besides the saliency differences mentioned above, this bundling is for another reason unfortunate.

Moving on to the top right cell of the matrix, we find procedural constraints. Under this category belong quasi-procedural constraints (not absolutely entrenched substantive constraints) and true procedural constraints. The former are constraints not protected against the double amendment procedure with substantive nature.^[297] True procedural constraints impose barriers to change. An example of procedural unamendability is the prohibition to amend again the constitution within a certain period after an amendment (temporal) or the requirement of a certain quorum to vote on an amendment (requirements of process).

The economic literature on constitutional change claims that if the formal pro­cedure for amendment becomes too cumbersome unconstitutional change is more likely to occur.^[298] In the language of economics, there is a substitution effect. Assuming that politicians want to keep their constituencies happy by transferring rents through amendment, whether the change is done formally or informally affects the quality of the service politicians provide to their constituencies. Formal amendment is a superior good in that it satisfies the need for keeping constituencies happy in a more long-lasting way. When the price of the superior good becomes high (cost of formal amendment) or income is reduced (political support), political actors shift to the inferior good, i.e., informal change.

Finishing, on the bottom right cell lies what I will call factual unamendability. This protection originates from the improbability of overcoming even the simplest amendment constraints for certain provisions, due to the structure of the state and the political game.^[299] Certain institutions and structures within the constitution are so entrenched in the existence of a state that they enjoy a de facto amendment- proofness. Specifically, even in the absence of formal entrenchment, such provi­sions will enjoy a greater longevity due to the very low factual willingness of amending them. In other words, factual unamendability could be understood as the results of the preferences of a society and the structure of the political arena. Specifically, it captures a situation where the willingness to support a topic does not meet the willingness to provide of the politicians. Either none or only marginalized minorities are willing to campaign for changing such provisions. Subsequently, this demand is not sufficient to cover the costs of a political actor forwarding such a demand. This type of unamendability is not an actual constraint in the sense that if preferences change the relevant actors would be able to change the constitution. Nevertheless, prolonged structural unamendability might open the door to con­ventional substantive unamendability.

Constitutions entrenching Islam to that respect can serve as an extreme example. The constitution of Saudi Arabia of 1992 explicitly states that the constitution of the country is the Holy Quran and the prophet’s Sunnah (Article 1). One can interpret the clause such that the constitutional document is no longer a constitution, inas­much as it is not the supreme law of the land^[300] or that Article 1 is a nonbinding declaration, which has no place in the constitution. In either case, there is no explicit constraint to amend. Nevertheless, amending in discordance with such supra-constitutional rules—divine ones at that—is highly unlikely.

From the categorization, we directly see that true substantive eternity clauses bear significant comparative advantages against other forms of unamendability. First of all, they are written in a compact body of text. That means that they are on the long run protected from amendment through desuetude. Specifically, it is plausible that on the short run all political actors might agree upon breaching the social contract, for example, in times of emergency, however, unconstitutionality becomes unstable under the constant threat of a political actor invoking the eternity clauses and overturning the amendment. Moreover, citizens know their content and can form stable expectations. The textual nature of eternity clauses allows the People to directly monitor and hold accountable the politicians against violations. Of course, this effect can be muted when eternity clauses entrench vague legal terms, so that courts can decide the content of the constitution.^[301] This is, however, another consideration which will not be touched upon by the chapter.

3