Richard Albert’s new book, Constitutional Amendments: Making, Breaking, and Changing Constitutions, is a pathbreaking contribution to comparative constitutional law and the burgeoning literature on constitutional change, in several different senses. The book is a comprehensive guide to issues in designing amendment rules, such as the choice of single versus tiered- or multi-tracked rules for change, and the role of time in amendment. It also breaks new ground by covering issues of amendment that I have never seen nor honestly even thought about, such as the chapter on how to codify an amendment – at the end of the constitution or interspersed with the text – and relatedly on whether amendments erase the original constitutional text or leave it intact. This is a fascinating issue, and one with significant variation. To my knowledge, Albert’s book provides the first treatment of it.

Finally, the book is conceptually significant. Albert’s concept of a dismemberment is a key contribution. The idea is that some things look like an amendment, but really are not because of their scope or effect, and further that this distinction should be embedded in the design of amendment rules. Constitutions should thus work off a spectrum and use more demanding rules for certain kinds of changes – what Albert calls a “multi-tracked” amendment rule and what Ros Dixon and I have elsewhere called a tiered constitutional design.

In this post, I want to elaborate on a set of challenges to Albert’s project, which ultimately I think stem from its formalism. He seeks to build a universalistic theory of amendment, covering democracies and non-democracies alike. Thus, his concept of a dismemberment is a neutral one – it covers any constitutional change that undoes a core part of the current constitutional order, regardless of the direction of change or the nature of the starting point. Authoritarian constitutions can be dismembered just as liberal democratic ones can. From a normative perspective, too, the design recommendations are the same: all constitutions should take steps to protect their core values and identities, regardless of what those values and identities are. And perhaps more importantly, those seeking to change those values should follow the rules laid out in the constitution, rather than evading or ignoring them. Albert’s project is thus not a liberal democratic one, but a broader, more universal one, which seeks to guide the design of amendment rules in all constitutions.

But from the perspective of scholars, designers, and practitioners of constitutionalism, such a value-neutral conception of the rules of constitutional change is normatively problematic, and maybe impossible. It labels as the same thing – as a dismemberment – the destruction of a liberal democratic constitution via abusive forms of constitutional change that, say, undermines judicial independence or consolidates executive power and changes that move in the opposite direction, by democratizing an older, authoritarian constitution as in the post-Communist world. Albert’s design recommendations and recommendations for reformers make sense in the former case to protect against democratic erosion, but probably not in the second case when their import would be to safeguard or prolong authoritarianism.

In fact, Albert’s book does adopt a normative stance: it privileges formalism, as he makes clear in the book’s last chapter. Albert does not dismiss virtues of informal change, but he says that change should, as much as possible, reflect the formal rules. Albert rightly points out that adhering to the written rules of constitutional change is often a normative good. It usually increases stability, enhances legitimacy, and provides a focal point for actors seeking to carry out constitutional projects. But at the same time, the benefit of following the formal rules is not intrinsic, but rather is conditional on its relationship with other values. And the relationship between amendment rules and those broader values is complex. There are plenty of examples in recent history – Hungary, Turkey, Venezuela, and Ecuador, for example – where packages of amendments that certainly or arguably followed the formal rules of constitutional change did immense damage to the liberal democratic order. On the other hand, if you have an authoritarian starting point, insisting on adherence to existing amendment rules might prolong that regime. Stability is not an unmitigated good, but a conditional one.

Furthermore, the concept of adherence to the formal rules is itself often difficult to discern. Constitutional design is of necessity incomplete. The tiered or multi-track amendment rules that both Richard and I favor are good examples. These rules, as they are normally written, continue to leave an enormous amount of discretion in courts and other political actors. As an example, consider a 2014 Ecuadorian case that determined the route needed for Rafael Correa to eliminate presidential term limits. Ecuador has a state-of-the-art, three-tiered design, differentiating amendment from a partial reform and a replacement. The question in the case was whether completely eliminating term limits would impact the “fundamental structure of the state” or “restrict constitutional rights and guarantees.” Changes affecting these aspects of constitution required more demanding procedures, such as referenda or constitution-making process. There are good arguments of course that eliminating presidential term limits would at least impact the fundamental structure of state, and also maybe restrict constitutional rights and guarantees. But the Ecuadorian Constitutional Court was dominated by supporters of President Correa, who by that point had been in power for about eight years and had carried out other reforms to increase power over judiciary. The court predictably said that the baseline method of constitutional change was acceptable. Correa’s efforts to remain in power eventually failed, but they were blunted by street protests rather than constitutional design. In this case, even a state-of-the-art amendment rule failed to prevent a president from eliminating term limits, and from doing plenty of other things to consolidate power. Formal adherence to the rules may not be that much of a defense of the legal order, even when you have a good design.

The incompleteness of a formalistic conception of constitutional change also underlies my disagreement with Albert on one of the most significant developments in the contemporary theory of constitutional amendment – the judge-made unconstitutional constitutional amendment doctrine, as it has developed in a large number of countries such as India and Colombia. Richard is both here and elsewhere extremely critical of the doctrine, largely on the grounds that it clashes with the rule of law. He prefers a tiered or multitrack amendment rule, which creates greater predictability and less tension with democracy. But many constitutions around the world are not ideal designs, and in those systems the default rules of constitutional change are themselves ripe for abuse by would-be authoritarian actors. The case for the unconstitutional constitutional amendment doctrine, in those contexts, is that it lets judges do ex post what the design has not done ex ante: protect sensitive parts of the liberal democratic constitutional order against the risk of abuse of democratic erosion. The judicial derivation of an unconstitutional constitutional amendment doctrine, as an inference from the text or from the nature of constitutionalism, may in some sense be a deviation from the formal rules, but also a potentially useful one. Take the famous Colombian case where the Constitutional Court used a wholly judge-made version of the unconstitutional constitutional amendment doctrine to stop Alvaro Uribe from potentially seeking a third consecutive term in office, after he had already amended the constitution to allow two consecutive terms. The formal rules of the constitution made it easy for Uribe to amend the constitution any way he wanted, and thus the Court’s decision plausibly helped to prevent an erosion of democracy.  

In short, a theory of amendment rules that purports to be value-neutral, or that privileges stability or formalism above all else, won’t always do the things we want amendment rules to do. Richard Albert has given scholars and constitutional designers an invaluable guide to thinking through the wide range of issues connected to the design of amendment rules. His conceptual vocabulary of a dismemberment helps us to distinguish between forms of change, and many of his normative recommendations – most importantly creating multi-track or tiered amendment designs rather than simple, single-track rules – are major improvements to current practice. But they ultimately make sense against a liberal democratic starting point, and in the service of the protection of liberal democratic constitutionalism.

David Landau is the Mason Ladd Professor and Associate Dean for International Programs at Florida State University College of Law. You can reach him by email at dlandau at law.fsu.edu.