A Legality and Legitimacy Framework for Analysing (Unconstitutional) Constitutional Amendments

Abstract

This introduction to our special issue on ‘Constitutional Legitimacy and Amendments’ presents a framework for a more nuanced understanding of how constitutional change is contested, moving beyond the conventional notion of ‘unconstitutional’ constitutional amendments. We advocate for a clearer distinction between legality and legitimacy when analysing contestation over constitutional change, arguing that focusing exclusively on legality without addressing legitimacy risks oversimplifying constitutional debates and overlooking questions of broader political and social acceptance. We identify three grounds on which the legitimacy of a constitutional amendment may be challenged: lack of representativeness (an amendment is illegitimate if it is not representative of the will of the people); lack of justice (an amendment is illegitimate if it is unjust in a significant manner); and bad faith (an amendment is illegitimate if it is motivated by ulterior motives). We also outline three ways in which legality and legitimacy intersect: (1) legally valid amendments may still face challenges regarding their legitimacy; (2) formally illegal constitutional changes may still be perceived as legitimate; and (3) even amendments that are both legal and legitimate may still require legitimacy to be established through means other than procedural and substantive commitments to legality. By recognising that challenges to constitutional amendments often involve claims of both legality and legitimacy, our framework expands the analysis of constitutional amendments beyond claims based on constitutional identity or unamendability and contributes to a better understanding of how such amendments are contested.

Introduction

Constitutions change over time, but how they change varies significantly. The formal amendment process is perhaps the most common, though there are other modes, including ‘periodic replacement of the entire document’, ‘judicial interpretation’, and ‘legislative revision’.¹ Formal amendments require that the constitution be changed through procedures specified in the constitution, and in many instances help strike a balance between constitutional flexibility and rigidity. Even though most constitutions do not impose any specific substantive limits on the amendment powers, questions persist as to whether there should be such limits to constitutional amendment powers, and if so, where those limits should lie and who should enforce them. This has spawned various doctrines – such as ‘unconstitutional constitutional amendments’,² ‘destruction of constitutional identity’,³ and ‘constitutional dismemberment’⁴ – which scholars use to delineate between valid and invalid constitutional amendments. Courts around the world have engaged with and, at times, applied the doctrine of unconstitutional constitutional amendment as a judicial doctrine to check political actors seeking to substantially change the constitution.⁵

Constitutional change implicates normative claims about the legitimacy, legality, and nature of such change, and the effect that it may have on a constitution. Constitutional doctrines that impose substantive limits on constitutional amendment powers build on the assumption that while some amendments may be procedurally valid, they are substantively invalid. It is, however, not always clear whether the claim is that these amendments are constitutionally illegal, or that they are constitutionally legal but illegitimate, or that they are constitutionally illegal and constitutionally illegitimate. This is because assessments of constitutional amendments and enquiries into their constitutional limits do not clearly distinguish between questions of legality and questions of legitimacy.⁶ If anything, there is a tendency to conflate the two. Doyle further notes that there is even a tendency to mix moral and conceptual claims when addressing limits on the amendment powers; a moral objection is that an amendment ‘fails to respect the value of constitutionalism, understood as the value of constraining governmental power’, while a conceptual objection posits that some acts fall outside the concept of an amendment⁷ because they go too far. This critique does not draw distinctions between legality and legitimacy, but demonstrates the multilayered conceptions that impact our basic intuitions about the permissibility of some types of constitutional amendments.

As a starting point, when a constitutional amendment complies with the constitution's procedural requirements, it is prima facie constitutional and legal. At the same time, procedural validity can also provide the basis for constitutional legitimacy; for example, some scholars suggest that democratic or deliberative amendment procedures produce more legitimate outcomes.⁸ This means that a procedurally valid constitutional amendment could also be legitimate. Yet, the implicit claim under doctrines of unconstitutional constitutional amendments is that mere procedural validity is insufficient. A conceptual distinction is necessary to explain resistance against certain types of constitutional amendments and advocacy for institutional endurance. When judges strike down unconstitutional constitutional amendments, this must be on the basis that they are constitutionally illegal. But does it matter whether the amendment is constitutionally legitimate, and if so, in what way?

The symposium grapples with this intellectual intersection between constitutional amendments on the one hand and the interconnected concepts of legality and legitimacy on the other. How does and should legitimacy fit into our assessment of whether a constitutional amendment is legally valid? Admittedly, this is further complicated by the idea that constitutional amendments may be considered legitimate by one criterion and illegitimate by another. In addition, legality is also sometimes viewed as a criterion for legitimacy.⁹ The manner in which amendment rules function, the role of the public and participation mechanisms, the intervention of judges and judicial review, and the time of the change have a direct impact on how the legitimacy of amendments is perceived, established, and understood. Furthermore, how have competing conceptions of the legitimacy of constitutional amendments been resolved (or not)? Where reformers have turned to formally illegal procedures to amend the constitution, how were they able to validate these amendments as legitimate (or not), and how has this mattered?

This article, along with the other contributions in this special issue, which emerged from an accompanying symposium, argues that greater conceptual clarity between the legality and legitimacy of constitutional amendments is necessary to better understand the complexities surrounding debates for and against some constitutional amendments. This article argues that failure of judges and scholars to engage with legitimacy questions obscures the true nature of constitutional debates. In other words, we need to untangle the questions of legality from those of the legitimacy of constitutional amendments. Rather than providing lengthy summaries of the seven articles included in this special issue, we will first offer a concise overview of their diverse explorations into the legality and legitimacy of constitutional amendments across a wide range of jurisdictions. In the course of our substantive interventions in this introductory article, we will then discuss these contributions in more detail, highlighting how they fit into the broader scheme of scholarly analysis of the relationship between legality and legitimacy in constitutional amendments.

First, Francisca Pou Giménez proposes internal and external models of legitimacy to analyse constitutional amendments, particularly in a society undergoing autocratic regression. Next, Ming-Sung Kuo emphasises the internal point of view in determining the legitimating ideologies of constitutional amendments, contrasting Taiwan's emphasis on legal continuity with China's promotion of programmatic socialism. Toon Moonen analyses the legality-legitimacy bifurcation through the lens of two sets of constitutional activity in Belgium, one involving ‘implicit constitutional amendments’ and the other the use of a ‘temporary amendment’ procedure. Adem Kassie Abebe interrogates the legality-legitimacy intersection from the perspective of African countries with hegemonic-party or de facto one-party systems. Kevin Tan examines how breaks in legality often have to rely on revolutionary legitimacy to bring about constitutional change by employing the case study of Singapore. Pui-yin Lo's examination of Hong Kong's central-local competition with the administration in China highlights how competing claims of legality and legitimacy are intertwined. The last article brings us back to the function of the judiciary: Tamar Hostovsky Brandes and Yaniv Roznai return to the theory of unconstitutional constitutional amendment and examine how it limits judicial resistance to revolutionary changes that undermine core features of the constitutional order, particularly under conditions where the judiciary itself is under autocratic attack.

Legality and Legitimacy of Constitutional Amendments

To be sure, the two concepts of legality and legitimacy are at some level closely intertwined. Indeed, legality may be regarded as a subset of legitimacy. After all, legal legitimacy is one of Fallon's three conceptions of legitimacy, the other two being sociological and moral legitimacy.¹⁰ According to Fallon, legal legitimacy is a ‘legal concept’, whereby ‘legitimacy and illegitimacy are gauged by legal norms’.¹¹ For him, this means that ‘[t]hat which is lawful is also legitimate’.¹² Legal legitimacy is to be distinguished from sociological legitimacy,¹³ which concerns whether the law is accepted as deserving of respect or obedience, and from moral legitimacy, which is rooted in the moral justification of claims of authority made in the name of the law.¹⁴ Fallon's reference to legal legitimacy is more to distinguish it from other possible perspectives of legitimacy than to determine the relationship between legality and legitimacy. For instance, Fallon makes an interesting point that the legal legitimacy of the (American) Constitution ‘depends much more on its present sociological acceptance (and thus its sociological legitimacy) than upon the (questionable) legality of its formal ratification’.¹⁵ Accordingly, he emphasises, ‘[o]ther fundamental elements of the constitutional order, including practices of constitutional interpretation, also owe their legal legitimacy to current sociological acceptance’.¹⁶

Despite the importance of legitimacy questions for constitutional amendments, debates over constitutional amendments have tended to focus on conceptual claims about the proper nature of the constitution (eg, its constitutional identity) and the scope of the amendment powers. A constitutional amendment is unconstitutional, the argument goes, because it does not meet the conceptual definition of what an amendment power could achieve.¹⁷ This in turn becomes an inquiry into what a constitution is and whether a constitution so amended can still be the same constitution or, in some more extreme claims, still be a constitution. But this conceptual discussion is often unsatisfactory. One still needs to justify why the original constitution (and its supposed entrenched identity) should take precedence. Scholars who argue in favour of such constraints on amendment powers draw on theories of constituent power, arguing that the original constitution is an exercise of constituent power, which is normatively prior and superior to subsequent acts of constitutional change. Roznai provides a useful distinction between primary constituent power and secondary constituent power, whereby the amendment powers are only secondary constituent powers and must therefore be subordinated to the essential features of the original constitution, which resulted from an exercise of primary constituent power.¹⁸ This conceptual analysis does not, however, take us far enough; it does not answer the temporal question of why subsequent generations of ‘the People’ lose their primary constituent power or, more specifically, why they could not delegate their primary constituent power to an elected body to amend the constitution. After all, for all the grandiose claims of the initial drafting of some constitutions as an act of ‘We, the People’, the actual document is often drafted by a small group of elites, who may or may not be representative of the actual people.

Such conceptual arguments can become circular and obscure the true nature of the disagreement. It might thus be productive to recast the debate around claims of legality and legitimacy. This means that the challenges to and defences of a constitutional amendment could be understood as a mix of legality and legitimacy claims. Legality is here defined as compliance with legal rules, which admittedly generates a thin form of legitimacy. The core of the challenge to such procedurally valid constitutional amendments has to be that this itself is not sufficient. Here, the two concepts of constitutional legitimacy identified by Harel and Shinar are instructive: a constitution could be regarded as legitimate either because it is representative of the people, or because it is seen as just, or both.¹⁹ Harel and Shinar point out that a representative constitution ‘gains its legitimacy from the fact that it is a reflection of who the citizens are or what they desire or judge to be true’. The core is, therefore, self-governance – that the constitution reflects the ‘will of the people’.²⁰ This can be an empirical will of the people (identified in a particular foundational moment) or an objectification of the ‘real’ values of the people (found in culture or history). By contrast, reason-based legitimacy regards the constitution as legitimate because its content is ‘just, rational or grounded in reason’.²¹

Building on this framework, we propose three distinct legitimacy claims that should inform the evaluation of a constitutional amendment's constitutionality. The first is a claim that the amendment is illegitimate because it is not representative of the will of the people (the representativeness claim); the second is a claim that the amendment is illegitimate because it is unjust in a significant manner (the justice claim); and lastly, there is a claim that the amendment is illegitimate because it is motivated by ulterior motives (the bad faith claim).

On the representativeness claim, an amendment may be challenged as illegitimate because it fails to reflect the representative will of the people. This argument resonates most with key conceptual arguments for unconstitutional constitutional amendments. If the amendment goes against the original will of the people, then that amendment violates the basic features of the original constitution and must therefore be regarded as unconstitutional. While not often couched in legitimacy terms, there is a legitimacy claim embedded in this conceptual argument, and that is that the original constitution has its source in a constituent power that reflects the people's will. Subsequent amendments are somehow less representative as they are not an exercise of the same constituent power. Accordingly, entanglements around representativeness of constitutional amendments are ultimately linked to the persistent question of why constitutions drafted by earlier generations should be obeyed, even in the face of present democratic choices.²² This is particularly relevant when we recognise the existence of amendment powers themselves as sources of constitutional legitimacy where the possibility of constitutional change allows us to view constitutions continuously as a ‘manifestation of the consent of the governed’.²³

Secondly, the justice claim is one that is more likely to be articulated outside the judicial arena but is nonetheless implicit in judicial decisions. Here, the claim that the amendment will create a less just constitution may be intertwined with the conceptual argument that the amended constitution is ultimately not the same constitution. The original constitution is portrayed as a just constitutional order that is ‘destroyed’ by the amendment, creating a new order that is less just than the one before. Tracing the justice claim in judicial decisions, we see that judges sometimes draw on external normative arguments of justice to justify exercising powers of review over procedurally compliant constitutional amendments. When unable to take recourse to specific constitutional provisions, judges may identify ‘logically antecedent presuppositions’²⁴ that give constitutions their legal status, either by interpreting them through the constitution's structure, through claims of constitutional identity, or through long-standing social and political practice.²⁵ A variation of this claim suggests that a constitution that does not commit to a (defined) idea of constitutionalism is no constitution: amendments that undermine this idea are therefore, necessarily unconstitutional. Lastly, justice claims can be founded on sources external to the constitution: international law, transnational norms, and natural law arguments may all provide foundations in support of claims that an amendment creates an unjust order.²⁶

Thirdly, the bad faith claim is one that is more commonly found in critiques of abusive constitutional practices. To be sure, Landau's definition of abusive constitutionalism as the use of constitutional amendments (and replacements) to undermine democracy²⁷ focuses on outcomes rather than the intent of the actors. On this basis, Landau's abusive constitutionalism could be understood as illegitimate in terms of justice rather than bad faith. That said, there is significant overlap between the two, particularly in arguments about abusive constitutional amendments. Some amendments are illegitimate because the ulterior motives of the government proposing the change taint the nature of the amendment. In other words, some facially neutral amendments may be challenged as illegitimate precisely because they are understood from the perspective of a proposer acting in bad faith. The bad faith claim attains particular relevance when opponents of an amendment seek to distinguish between amendments that are procedurally legal but produce illegitimate outcomes, as Roznai and Brandes suggest in their contribution to this special issue. This may require reading several amendments together as part of a larger project to undermine existing constitutional structures and norms that limit power.²⁸ Similarly, Abebe notes in his contribution to this issue that while procedural safeguards have sometimes prevented ‘self-serving constitutional changes’, the success of such alterations through formal amendment processes requires us to consider alternative means of evaluating and establishing the legitimacy of constitutional amendments.²⁹

Mapping Out Configurations of Legality and Legitimacy

There are multiple ways in which legality and legitimacy may intersect.³⁰ The articles in this special issue map out situations in which jurisdictions have engaged in constitutional change that implicates questions of legality and legitimacy in a variety of ways. Conceptually, we categorise these into three possible configurations reflected in the papers in this special issue: legally valid but illegitimate change; legally invalid but legitimate change; and finally, change that is considered both legal and legitimate, but with reference to the distinction between the two concepts.

Amendments inevitably vary in their degrees of legality and legitimacy, with questions of legality influencing perceptions of legitimacy and vice versa. Constitutional amendments can fit into any one of the three possible configurations, demonstrating the complexity of constitutional change and the variety of ways in which it may be carried out. Moonen's article in this special issue, for instance, points out that Belgium's constitutional history demonstrates that the practice of implicit constitutional amendments has generated illegal changes that have nonetheless been broadly accepted as legitimate. On the other hand, a temporary amendment procedure adopted in 2012 created an apparently legal but deeply contested and potentially illegitimate alteration to the constitutional framework.³¹ As Moonen notes, while legitimacy and legality are disparate concepts, one path to legitimacy is through compliance with formal amendments rules, but mere compliance, as evidenced by his examples, is insufficient. Untangling the connection between legality and legitimacy is thus fundamental to a better understanding of constitutional change.

Legally Valid Amendments, Illegitimate Change

Can an amendment that is validly enacted in accordance with a constitution's stated procedure nonetheless be invalid, and if so, what is the basis of this illegality? Doctrines of implicit unconstitutional constitutional amendments grapple precisely with this question. Roznai traces the origins of arguments in favour of such implicit constitutional unamendability to debates around Article V of the Constitution of the United States, which contained two constraints on amendments, the first relating to slave-trading and the second to depriving states of equal representation in the Senate.³² Arguments about whether an amendment could be allowed even if it implicitly ‘destroyed’ the constitution itself developed in this context, just as German scholars developed the concept of ‘constitutional identity’, which was used to construct explicit limits (ie, an eternity clause) and justify implicit limits on constitutional amendments.³³ Indian courts drew from German scholarship to develop the doctrine of ‘basic structure’ in 1973: this doctrine allowed judges to review and strike down constitutional amendments that they found inconsistent with features they identified as core to the Indian Constitution, even if they were legally enacted in compliance with formal amendment requirements.³⁴ The ‘basic structure’ doctrine has since migrated to several jurisdictions in Asia and beyond, to jurisdictions including Argentina, Greece, South Africa, Switzerland, Tanzania, and Bangladesh, and has been a subject of debate in jurisdictions where it has met with limited or no acceptance.³⁵

The invalidation of legally enacted amendments presents a challenge for constitutional scholars who consider the source of constitutional legality to be the constituent power that enacted it. Roznai argues that this apparent conflict can be resolved by understanding the power to amend constitutions as drawing from a ‘secondary constituent power’ which is subject to procedural and substantive constraints.³⁶ Theories of constituent power have come under criticism, although they still have their fair share of defenders.³⁷

One set of criticisms centres upon the descriptive implausibility of the people as a unified body exercising constituent power. Stone, for instance, disagrees with Roznai that limitations on amendment powers flow by necessary implication from delegated constituent power to enact amendments.³⁸ Stone's argument rests, in part, on the observation that the original constitutional text itself is not necessarily an expression of constituent power, and as scholarship on the drafting of constitutions increasingly demonstrates, it may represent elite consensus rather than public consent. This is also a critique offered by Yap, who suggests that it may never be possible to constitute a sufficiently representative assembly, and moreover, that a constituent assembly to replace a constitution is itself one way in which abusive constitutionalism can occur.³⁹ Further, some scholarship on constitution-drafting lends credence to the claim that it may not be fair to assume that the original constituent power was in fact a product of popular will, as opposed to elite consensus.⁴⁰ Consequently, Colon-Rios contends that the doctrine of unconstitutional constitutional amendments is vulnerable to claims that it produces illegitimate outcomes. He instead makes the case for greater participation in amendments, such as the Colombian requirement of a constituent assembly for major constitutional reforms.⁴¹

As constituent power provides a limited justification for unconstitutional constitutional amendments, scholars have since turned to constitutional politics. Barak, Issacharoff, Fombad, and Dixon and Landau all point to the judicial invalidation of constitutional amendments as a safeguard against authoritarianism, populism, and other forms of undemocratic politics that can lead to a retrenchment of constitutional guarantees.⁴² Examples of such ‘abusive constitutionalism’,⁴³ ‘stealth authoritarianism’,⁴⁴ or ‘populist constitution replacement’⁴⁵ abound, with scholarship from around the globe demonstrating how constitutional amendments are often used to consolidate power.⁴⁶ Indeed, part of the prevalence of the doctrine of unconstitutional constitutional amendments can be attributed to the normalisation of judicial involvement in political questions, resulting in the acceptance of the role of judges in invalidating constitutional amendments.⁴⁷ Accordingly, Preuss argues that ‘the institution best suited to verify an unconstitutional constitutional amendment is the constitutional court’, noting that judicial review of legislation and amendments bears sufficient parallels to suggest that a court is best equipped to undertake this task.⁴⁸ Such arguments draw on common justifications of judicial review to hold that courts perform anti-majoritarian functions in reviewing amendments, as they do in reviewing legislation.⁴⁹

An obvious limitation to judicial constraints on amendments relates to change that occurs incrementally or that produces shifts in constitutional commitments. A separate body of scholarship considers the use of constitutional amendments by political forces as a deliberate tool to increase executive control and diminish constitutional protections of rights, institutions, and principles. Evidence from Hungary, for instance, shows that multiple and successive constitutional amendments, enacted in formal compliance with the constitutional procedure, can nonetheless generate illegitimate outcomes, beginning with the rewriting of the amendment rules themselves.⁵⁰ A substantial set of studies from South America and Africa, focusing particularly on the extension of presidential term limits, demonstrate that validly enacted amendments may permit executive authorities to extend and expand their control over democratic institutions by allowing indefinite control over executive governments.⁵¹ Validly enacted constitutional amendments could be utilised to pack courts to ensure deferential (or abusive) judicial review and to remove the check of unconstitutional constitutional amendment holdings, as in the case of Turkey or as unsuccessfully attempted in India.⁵² Recent amendments to the Mexican constitution, which significantly restructure courts and limit judicial independence, were technically in compliance with amendment procedures, but arguments that their passage was achieved illegitimately and that they serve illegitimate ends persist.⁵³

In such cases, judges faced with individual changes may be unable or unwilling to find unconstitutionality, despite the cumulative effect of such changes creating a condition of unconstitutionality.⁵⁴ In Tamar Hostovsky Brandes and Yaniv Roznai's contribution to this symposium, this conundrum is examined with reference to constitutional amendments in Israel, noting the particular difficulty presented when courts have to examine the ‘first slice of the salami’ and do not have the benefit of looking back at the cumulative effect of several changes in order to determine their combined impact.⁵⁵ Brandes and Roznai argue that judges should incorporate an approach of ‘militant democracy’, examining a series of conditions that allow them to act as gatekeepers against such changes, including evidence of future changes to the overarching constitutional order and the critical importance of the change being implemented. This extended account of unconstitutional constitutional amendments obliges courts to protect democracy itself, with the caveat that this should not extend to blocking changes in their own institutional structure.

Advocates of judicial review of constitutional amendments propose ways in which such review can be controlled to address legitimacy concerns.⁵⁶ A common strategy is to limit the scope of judicial review. Dixon and Landau, for instance, argue that unconstitutional constitutional amendments are one of a series of remedies that can raise the threshold of public engagement with reforms in order to address challenges of abusive constitutionalism; other methods include tiers of amendment with different thresholds and presidential term limits.⁵⁷ Under this model, judicial review is invocable only when judges ‘are confident that amendments, or packages of amendments, pose a substantial threat to a set of values closely associated with the democratic order’, and engagement with courts across borders acts, in turn, as a second measure of protection in helping judges establish when to engage in review.⁵⁸ Yap, on the other hand, makes the case for a standard of manifest unreasonableness, evident from amendments that would ‘substantially destroy’ the previous constitution.⁵⁹

Finally, contextual approaches, as generally proposed by Doyle, Neo, and others, suggest that the degree of review can only be determined by reference to the specific circumstances of the amendment. The shape of a constitution, the role of majorities, party politics, constitutional culture, and other factors may play a significant role and prevent a one-size-fits-all solution.⁶⁰ As Jacobsohn contends, references to the circumstances of an amendment may reflect the idea that a constitutional identity is discernible and enforceable from the text of a constitution.⁶¹ Here, arguments for a universally acceptable standard of review are neither necessary nor desirable; instead, the development of contextually appropriate standards of review provides various and diverse justifications for engaging in review of validly enacted amendments.

Arguments such as these, which seek to justify the incorporation of judicial review of constitutional amendments, must surmount the counter-majoritarian difficulty to a greater extent than in debates over the review of legislation.⁶² Constitutional amendments present a way for democratic majorities to set a new constitutional agenda, even by legislatively overruling judicial decisions. When judges in turn strike down a constitutional amendment, they need to be able to legitimate their decisions. Relying on conceptual arguments about constitutional identity, we argue, is not sufficient. Courts should articulate more clearly the legitimacy deficiency of such a constitutional amendment – whether this is a lack of representative legitimacy or justice legitimacy, or an act in bad faith – in addition to claims about the legality of the constitutional amendment.

Consequently, while judicial review of constitutional amendments can provide an essential safeguard against constitutional change that is legal but not legitimate, the doctrine has its limits. Francisca Pou Giménez focuses on this in her article for this issue, examining ways in which we can distinguish and establish legitimacy in the context of amendments.⁶³ Identifying, among other things, the unconstitutional constitutional amendments doctrine as a means of establishing ‘internal legitimacy’, Pou Giménez notes that contemporary constitutional developments demonstrate that courts can and have used these means to reify illiberal, undemocratic, and iniquitous changes, and that they are particularly ineffective as a counter majoritarian check to such change when it occurs through incremental means, or when it draws on the apparent content-neutrality of such methods. As an alternative, Pou Giménez advocates for a ‘self-standing, external legitimacy’ that draws on sources outside the constitutional text and instead turns to the origin, procedure, content, and justification of amendments measured by yardsticks outside constitutional provisions. Her argument locates the distinction between legality and legitimacy in contemporary constitutional practice, noting the increasing distance between constituent power (which animates constitutional amendment rules) and actual amendment practice.

Another account of thicker procedural legitimacy can be found in Abebe's article on the illegitimacy of constitutional amendments in a number of African states, which offers insight into the intersection of legality and legitimacy of constitutional change in hegemonic-party or de facto one-party systems. He argues that the legitimacy of constitutional amendments should be measured by the extent to which such change results from a process aimed at achieving broad consensus, beyond the usual reliance on standard amendment procedures. The idea of a ‘broad consensus’, he observes, is a concept endorsed by the African Charter on Democracy, Elections and Governance. Abebe laments that amendment procedures have allowed dominant groups to push through capricious, self-serving, and generally regressive/abusive⁶⁴ reforms without securing a sufficiently broad consensus. For him, such broad consensus is achieved not only through these formal procedural proxies, but should also be manifested through cross-party support. Abebe notes that this is necessary in light of the unusual political circumstances of many jurisdictions in Africa, where dominant groups often have supermajority control in the legislature. Thus, he suggests that amendments should be regarded as legitimate only if they can be shown to have the support of a majority (or a specified proportion) in a majority (or a specified number) of political groups in parliament, rather than merely a supermajority of the total membership – a threshold often easily met by dominant groups in Africa. Abebe's core argument is that no single political group, regardless of its electoral dominance, should be able to unilaterally alter the fundamental constitutional rules of the democratic game. He sees this as ‘an inclusive re-imagining of majoritarianism’.⁶⁵

Abebe's intervention shows that we need to broaden the notion of legitimacy from a procedural to a substantive one, rooted in social and political acceptance. There is no doubt that the legitimacy of the constitutional amendments in this regard may be difficult to enforce judicially, but the issue remains extremely salient from the perspective of how successful such amendments could be implemented politically and whether this would contribute to constitutional legitimacy as a whole. Indeed, the legitimacy of constitutional amendments can contribute to, or even undermine, the stability and endurance of constitutions.⁶⁶

Illegal Amendments, Legitimate Change

If a legally valid constitutional amendment can be illegitimate, then it is equally plausible that an illegal constitutional development could nonetheless be viewed as legitimate. For instance, Ackerman's theory of constitutional moments suggests that a constitutional change using formally illegal procedures could be legitimated by overwhelming expressions of popular approval.⁶⁷ Along the same lines, the Fifth French Republic has lived its own encounter with illegality. Derosier points out that the 1962 referendum on direct presidential election was a misuse of Article 11 of the French Constitution, but the Conseil constitutional declined to exercise jurisdiction to determine whether the referendum was legal and whether the people voting in the referendum were acting lawfully precisely because of the broad popular support.⁶⁸ As he notes, the Conseil constitutionnel had no legitimacy to oppose the legal revolution of the people.⁶⁹ Determining whether an illegal constitutional change can still be regarded as legitimate is challenging and has been attempted through several lenses: through judicial review, with limited success; by considering the conditions that permit constitutional change; and through evolutionary, sustained public assent.

Assessing the sociological legitimacy of constitutional change, whether from the representative, justice, or bad faith perspective, may help to understand why some judicial review interventions have undermined judicial legitimacy. While the seminal decision in Kesavananda Bharati v State of Kerala ⁷⁰ is often celebrated for having set out the basic features doctrine, it was in fact a retreat from a prior case in which the Supreme Court had asserted a stronger stance against the power of Parliament to amend the constitution. In IC Golaknath v State of Punjab,⁷¹ the Supreme Court held that an amendment to the Constitution constitutes ‘law’ within the meaning of Article 13(2) of the Constitution of India, and that this provision prohibits the state from making ‘any law which takes away or abridges the rights conferred by this Part III [on fundamental rights]’. The actual decision was an important compromise; the Supreme Court conceded that the constitutional amendment concerning agrarian land reforms and redistribution of property effectively abridged fundamental rights, but did not invalidate it on practice grounds, noting that this might create a ‘chaotic situation’ as the land reforms had already been implemented. Instead, the Supreme Court chose to create a prospective rule that prohibited any further amendment abridging fundamental rights, reading it into the text of Constitutional provisions.

Public and academic outrage was immediate. While the judiciary had apparently upheld the fundamental status of rights, it was heavily criticised for ignoring the value of such agrarian reforms for ordinary Indians. Viewed as an anti-socialist and anti-egalitarian outcome, the Supreme Court's constriction of the legislature's amendment powers found little favour with the professional and lay public. The decision was called a ‘tragedy’ as the judiciary was seen as siding with landowners on agrarian reform; Reddy noted: ‘There was a flow of high-sounding rhetoric about the “transcendental” nature of the Fundamental Rights, but hardly a thought for the welfare of the “People of India” mentioned in the Preamble to the Constitution’.⁷²

Amendments that followed Golaknath asserted Parliament's power to amend any part of the Constitution, including provisions relating to fundamental rights, and were the subject of review in Kesavananda. While most scholars have focused on the judicial construction of the basic features doctrine as a meta-constitutional limit on the amendment powers derived from the preamble and directive principles, few have dwelt on the fact that the Court held that the amendment itself did not violate the basic structure of the constitution.⁷³ Kesavanada replaced the clear, articulated doctrine in Golaknath with an expression of more amorphous substantive limits on amendments that did not rest on textual arguments but on the unarticulated ‘basic structure’. Post-Kesavananda, Parliament is no longer explicitly restricted from amending any part of the Constitution, including the provisions relating to fundamental rights, unless they violate the judicially identified but undefined constitutional identity of the Indian Constitution.⁷⁴ The Supreme Court's articulation of the basic features in Kesavananda as potential substantive limits on Parliament's amending power was a weak compromise that arguably stemmed from the Court's failure to properly engage with the legitimacy of the initial agrarian reforms in Golaknath. Thus, despite its celebrated status, Kesavananda could be considered a judicial retreat, as the Court affirmed the 24^th Amendment passed by the Indian Parliament, which expressly excluded constitutional amendments from the scope of Article 13.⁷⁵

The refusal to invalidate amendments that were considered illegal but nonetheless enjoyed some measure of public legitimacy further demonstrates the challenges that courts face when assessing amendments. We have seen that constitutions may adopt complicated and multi-tiered rules of amendment for a variety of reasons: to facilitate greater public engagement, to limit the frequence of amendments in order build ‘veneration’ and commitment to constitutional guarantees, and to secure processes that are adequately consultative. Consequently, procedural review of amendment rules presents an entirely different set of challenges: courts can, and do, strike down amendments for failing to comply with these rules, even if the change these amendments seek to bring about may enjoy popular support, and otherwise remain in alignment with constitutional principles. Courts may find that deficits in legality may themselves generate illegitimacy: in some states, for instance, review may be restricted or limited to such procedural errors, allowing only ‘formalist resistance’ to constitutional amendments.⁷⁶ In this regard, Albert's typology of three kinds of procedural review could be instructive. According to Albert, amendments may be invalidated (1) for noncompliance with the outlined procedure; (2) because they exceed constitutionally established time limits for passage or ratification; and (3) due to errors in administering votes in procedures that require a vote.⁷⁷ Justifications for such formalist resistance may rest on an understanding of how constituent power is exercised to amend the constitution, requiring the court to defer to a properly enacted amendment as de facto legitimate.

Finally, even though scholarship on the review of amendments has focused on the ability of courts to enhance legitimacy by preventing amendments that retrench or limit constitutional guarantees, there is recognition that expanding judicial power comes with a specific set of risks. Dixon and Landau's work on abusive constitutionalism, and specifically abusive judicial review in courts, demonstrates that judges may, on occasion, actively facilitate the subversion of constitutional arrangements.⁷⁸ Sceptical accounts of judicial review suggest that judges may actively facilitate authoritarianism or populism, using legal frameworks (such as election laws, anti-corruption regimes, tax regimes, and free speech limits) to control opposition efforts.⁷⁹ Sadurski's work on the Polish Constitutional Court, Sanchez Urribari's work on the Venezuelan Supreme Court, Suteu's work on Hungry and India, and broader comparative studies, such as Versteeg et al's research on the judicial validation of amendments to term limits, all suggest that courts can and do play an active role in facilitating democratic decline.⁸⁰ A smaller subset of cases deals with instances where popular reforms are validly enacted with public support but are nonetheless blocked by courts engaged in review of constitutional amendments. Roznai and others argue that the Turkish headscarf ban is an instance of popularly supported reform that was struck down by courts despite its apparent legitimacy.⁸¹ Sethi and others argue that proposed reforms in India, enacted through a constitutional amendment that enjoyed popular, cross-party support, should similarly not have been struck down on review by the Indian Supreme Court, which invalidated legitimate public efforts to reform judicial appointments.⁸²

Legitimacy in constitutional change is also closely linked to the rate and frequency of amendments, which may establish not only the conditions under which change is possible, but also the consequences of such change for the endurance and stability of constitutions.⁸³ While rigid amendment rules, such as explicitly unamendable provisions (‘eternity clauses’) and other substantive restrictions on amendments,⁸⁴ are often invoked to protect core constitutional values and parts of constitutions against the demands of populist movements (see, for example, the protection of human dignity in Germany and the protection of rights in Brazil and Algeria), much research reveals that such rigidity also tends to prevent democratic efforts at change and to reify anti-democratic practices.⁸⁵ Entrenchment can prevent the correction of errors, the adaptation to changing social norms, and legitimate, deliberate legal formation.⁸⁶ Thereby, it can function as an anti-democratic barrier to legitimate change.⁸⁷ In Thailand, for instance, constitutional entrenchment of the monarchy and the judicial review of constitutional amendments have prevented changes to unpopular constitutional regimes imposed through a flawed constituent process.⁸⁸

Finally, the transition of postcolonial states towards independence presents a particular case of illegal but nonetheless legitimate constitutional change. Kevin YL Tan's article for this special issue closely examines Singapore's Constitution, drawing a distinction between the establishment of legitimacy through legal continuity and a moment of rupture marked by decolonisation.⁸⁹ Evaluating Singapore's independence, merger with, and separation from Malaysia, as well as subsequent constitutional consolidation and amendment, Tan argues that when legal continuity between the old and new state cannot be established, legitimacy has to be established through other means. Tan notes that efforts to establish legitimacy were negotiated amid political realities: the utilisation of a referendum instead of electoral approval was a strategic choice in response to a thin Parliamentary majority, and subsequent acknowledgments of Parliament acting in a plenary capacity were invoked to justify the adoption of a new constitution and the break in legal continuity. Tan's analysis points to the temporality of legitimacy, especially in cases where it is established outside conditions of legality. In such instances, legitimacy can be created cumulatively, by multiple means and at multiple moments, both at the time of revolution and again thereafter.

Legal and Legitimate Amendments

Constitutional amendments that are both legally valid and legitimate should be the least problematic since there is a confluence of procedural legitimacy and substantive legitimacy. Such amendments would best realise the underlying objective of amendment powers, which is to ensure that the constitution continues to serve the interests of the people.⁹⁰ Substantive legitimacy here could arise as a mix of representativeness, justice, and/or good faith, and what is key is that these are evaluations that rely on internal contexts of each jurisdiction. In his contribution to this special issue, Ming-Sung Kuo⁹¹ astutely points out that we need to uncover the legitimating ideologies of constitutional amendments through an internal point of view. He observes that what is required is an excavation of persistent characteristics in substance, process, form, and style that are constitutive of different constitutional amendment enterprises. He notes the divergent dynamics of legality and legitimacy in China (formally known as the People's Republic of China, PRC) and Taiwan (also known as the Republic of China, ROC) and observes how the preservation of the formal identity of the 1947 ROC Constitution has been central to Taiwan's constitutional transformation.⁹² Thus, throughout Taiwan's formal constitutional revision, there is a meticulous adherence to the rules of amendment, which is aimed at providing a semblance of legal continuity and which indicates the pivotal role of legality in Taiwan's constitutional development. By comparison, China's 1982 Constitution reflects a more programmatic character of its socialist constitution, which sees constitutional change as instrumental to achieving its socialist goals. Kuo contends that modernity informs both China and Taiwan in maintaining the legitimacy of their respective constitutions, but through divergent constitutional roads. While the claimed legitimacy of constitutional amendment in China depends on how an amendment reflects the new political direction under socialism, since the constitution was conceived as a steering programme, Taiwan has persistently emphasised constitutional legality as an integral measure of constitutional legitimacy of constitutional revision across different regimes.

Pui-yin Lo's article in this special issue complicates the picture by highlighting how competing legality frameworks generate different claims of legitimacy.⁹³ Lo's analysis is situated in a hybrid or authoritarian regime operating in subnational contexts. He notes that the PRC's efforts to functionally alter the operation of Hong Kong's Basic Law through executive, legislative, and interpretative means that are consistent with the PRC Constitution but not necessarily with Hong Kong's Basic Law have not been addressed by courts resorting to doctrines of unconstitutionality. His article shows that in subnational territories, particularly, formal legality may encompass dual regimes, and establishing legitimacy through the framework of one such regime may not always be sufficient to fulfil conditions of popular consensus. Lo's and Kuo's articles both present questions about the establishment of legitimacy through means additional to procedural commitments to legality. Both articles also present cases where legitimacy is supposedly forged through a top-down vision of representation and justice. As Kuo points out, identifying legitimating ideologies in context may require us to address specific constitutional cultures and their conceptions of what counts as legitimate power.

Conclusion

The contributions to this special issue were brought together as part of a research project organised by Jaclyn Neo, Kevin Tan, and Richard Albert under the auspices of the National University of Singapore Faculty of Law's Centre for Asian Legal Studies⁹⁴ and the University of Texas at Austin. Contributors were initially invited to participate in a workshop in Singapore, but this became one of the casualties of the Covid-19 pandemic. We were able to convene online instead, but the project was significantly delayed. It bears stating therefore that this symposium would not have been possible without the diligence and patience of all the contributors.

The thinking behind this research project was to contribute to scholarship on constitutional change and amendments beyond judicial interpretation to interrogate more holistically the political and social contexts of constitutional amendments,⁹⁵ and specifically how these contexts shape ideas about legitimacy, as distinguishable from legality. Accordingly, the embrace of diverse jurisdictions⁹⁶ in this symposium serves to thicken our conceptual and theoretical formation beyond the usual suspects,⁹⁷ hopefully in ways that would spur further research. Our proposed framework of legality and legitimacy as intersecting claims will hopefully further illuminate this area of research beyond the conventional theories of unconstitutional constitutional amendments.