Part III - Specific Institutions

9 Democratic Design and the Twin Contemporary Challenges of Fragmented and Unduly Concentrated Political Power

Stephen Gardbaum

There are (at least) four key values or principles of democratic governance. These are: (1) effective and responsive government, (2) stable government, (3) accountable government, and (4) representative and deliberative legislative bodies. Given the trade-offs among them, democratic polities cannot achieve all of these values equally but they are expected to attain at least a “minimum core” of each and to aim at balancing or perhaps jointly optimizing them.

This goal faces both a general problem and a more specific contemporary one. The general, and long-standing, problem is the central role and importance of political parties in modern democracies. Because political parties and their leaders compete to occupy two of the major governance institutions (the executive and legislature) and exercise public power, they can concentrate such power where the same party controls both and also disperse it where it does not, regardless of the formal or constitutional relationship between these institutions.¹ In this way, concentration of power threatens the values of continuously accountable government (i.e., not only at elections) and a genuinely deliberative legislature and, in so doing, increases the chances of various types of “misrule.”² On the other hand, dispersal of power risks undermining the values of effective and (sometimes also) stable government. For this reason, we cannot think of state institutions alone in analyzing or designing systems of democratic governance.

This general background problem or complexity is exacerbated by specific features of the contemporary political party systems in many democracies today. Party systems should not only be thought about in terms of numbers – a single, dominant, two-party or multiparty system – or the type of political regime in which they operate – presidential versus parliamentary parties³ – but also in terms of certain pathologies to which they are vulnerable. So, whether and to what extent a political party system is polarized, fragmented, or subject to hyper-partisanship also affects the difficulty of balancing and reconciling the four values. Polarization and hyper-partisanship can render both effective and accountable government, as well as deliberative legislative processes, harder to achieve because there is less, or no, overlapping middle ground. A fragmented party system makes effective and stable government less likely, as it is more difficult to obtain and sustain a governing majority.⁴ All three features undermine the political center and the types of consensus building and accommodation that tend to be important for the optimization of all four values. They also help to create the type of alienation from democratic politics “as usual” that has fueled various types of populism over the past decade.⁵

Both the general and the special problems can and do arise in all democratic regime types, of which there are at least six (and not only three), depending on the combination of (a) form of government and (b) political party and electoral systems. These are two party/majoritarian presidential, parliamentary, and semi-presidential systems and multiparty/PR versions of each.⁶ In all cases, the operation of both “ordinary” party politics and the special consequences of polarized, fragmented, and/or hyper partisan party politics complicates the task of balancing the four key values of democratic governance and skew polities toward either the fragmentation or the undue concentration of political power.

A recent strand within political science and constitutional scholarship has identified “semi-parliamentarism” as a new and alternative democratic regime type and also argued for its superiority to existing ones.⁷ The precise nature of this claimed superiority (as well as some of the institutional details) varies somewhat among its main proponents and could also benefit from being further developed but, at least implicitly or in part, the claim is that it better balances the four values.⁸ Semi-parliamentarism is presented as a distinct variation on two-party and multiparty parliamentary regimes. It is defined by the absence of the direct election of the chief executive and the existence of two directly elected and co-equal legislative chambers in which only the first can dismiss the cabinet in a no-confidence vote, while the second has veto power over legislation that cannot be overridden by an ordinary or absolute majority of the first. Importantly, its final key feature is the incorporation of different voting systems for each of the two chambers, to try and ensure that the governing party does not control both. Whereas for pure parliamentarism in either its two-party or multiparty versions, reconciling the first two values with the second two is notoriously difficult, it is potentially achievable with semi-parliamentarism.

I agree that semi-parliamentarism is a promising regime type. In this chapter, my primary aim is to explore whether the insights of its proponents can be adapted to suggest versions of non-parliamentary democratic regimes that better reconcile and optimize the four values and address the specific challenges of political party polarization, fragmentation, and hyper-partisanship. In other words, my focus is not on the question of which regime type is superior overall but rather on how to maximize the potential benefits of the semi-parliamentary model through ambitious, but not wholesale or root and branch, design reforms in the face of current democratic challenges. Pragmatically, given the well-known “stickiness” or path dependence of forms of government,⁹ ruling out these potential benefits to the roughly two-thirds of non-parliamentary democratic polities seems like a waste. Specifically, I will argue that semi-parliamentarism’s core feature of “symmetrical” and “incongruent”¹⁰ bicameralism is detachable from parliamentarism and that, with suitably customized modifications and reforms, is available in presidential and semi-presidential versions that may similarly reduce the contemporary pathologies of party systems and better balance the underlying values of democratic governance than existing regimes of these types. In so doing, all three adapted forms may also address some of the causes, and resist some of the consequences, of democratic backsliding in general and authoritarian populism in particular. The secondary aim of the chapter is to consider whether the design features of these versions that involve political parties and voting systems, rather than institutional powers and relations, should be constitutionalized and, if so, which.

9.1 The General and Special Problems That Political Parties Pose to Optimizing the Four Key Values of Democratic Governance

It is widely accepted that democratic governance seeks to promote at least four key values or principles. The first is effective and responsive government. Political parties and their leaders campaign during elections not simply to occupy public office but to offer voters a meaningful choice of policies on issues that matter to them that they will seek to put into effect if elected. This capacity to bridge, translate, and aggregate voters’ policy preferences into governing and legislative agendas is perhaps the central function and justification of political parties in a democracy.¹¹ Being elected to power and obtaining the relevant majority support legitimizes one policy agenda over another and, ceteris paribus, this is what a democratic government is expected to act on: elections have consequences. The ability of a government to effectuate the policies for which it was elected, as well as deal with ongoing and unexpected issues as they arise, is the hallmark of a functional democratic polity; the inability to do so is a sign of dysfunction.¹² The perceptions that democratic governments have been dysfunctional and/or more responsive to the interests of various elites than ordinary voters have, of course, been one of the main factors driving populisms of left and right over the past decade.

Government stability during the course of an election cycle is a second key value. Ongoing fragility or frequent turnover undermines the kind of mid-term planning that effectiveness requires and distracts voters and politicians by elevating office over policy. It also renders polities vulnerable in the face of new and unexpected crises that may arise. Obviously, too much stability is also problematic – if not always or necessarily inconsistent with democratic governance¹³ – as periodic (rather than frequent) turnover is another hallmark of a functional democratic polity. This, in turn, is partly driven by the third value of continuous governmental accountability, as without periodic turnover, party-state fusion¹⁴ and entrenchment across all institutions risks impunity and the inability to meaningfully question those in power. Putting into practice this third value is one of the key functions of democratic legislatures, although it is shared with other actors, public and private, including free and independent media outlets. The other key function of democratic legislatures is to promote the fourth value, by directly representing a broader range of voters and political positions than the executive and bringing these to bear in inclusive and collective deliberation over legislative priorities and content.

These four values exist in some tension with each other and inevitably involve some trade-offs in practice, as no single democratic polity could maximize all of them. A realistic normative goal is rather to balance or jointly optimize them in a way that ensures at least a “minimum core” of each is achieved, even if certain polities afford greater weight to some than others. Indeed, each of the three widely adopted modern forms of democratic government can be thought of as designed to achieve such a balance, albeit with different emphases resulting from the particular allocation of powers and functions between the executive and legislative branches. As referenced in the introduction, the general problem is that these modern forms of government were designed (or evolved) in ways that focused only on institutional relations and either ignored or were openly hostile to political parties.¹⁵ But the rise and role of modern political parties changes a great deal about how these forms operate in practice and complicates their actual ability to balance the values. And this general problem has been exacerbated in recent years by the particular nature of many party systems in democracies as polarized, hyper-partisan, and/or fragmented.

Starting with the general problem, institutions are occupied by leaders and representatives of political parties rather than by individuals per se, and this means that parties can effectively merge what is intended to be separate, as well as separate what is intended to be fused. This way in which parties can function like a sort of political holding company or conglomerate was well captured by Maurice Duverger in his classic work on the subject:

Officially Great Britain has a parliamentary system … in practice the existence of a majority governing party transforms this constitutional pattern from top to bottom. The party holds in its hands the essential prerogatives of the legislature and the executive … Parliament and Government are like two machines driven by the same motor – the party. The regime is not so very different in this respect from the single party system. Executive and legislature, Government and Parliament are constitutional facades: in reality the party alone exercises power.¹⁶

So, despite the classic British separation of power between King and Parliament, or its modern version between the King’s ministers in parliament and its ordinary members, as early as Bagehot it was recognized that the “efficient secret of the English Constitution” is “the close union, the nearly complete fusion of the executive and legislature powers” in the cabinet,¹⁷ stemming from the existential need of the government to retain the confidence of parliament. With the subsequent introduction or evolution of modern political parties, party discipline, and control over their legislators, the effectiveness and stability of Westminster-style parliamentary governments was increasingly achieved at the expense of genuine accountability to, and deliberativeness of, the legislative body. But even where executive and legislative branches are designed to be more separated and independent than in modern parliamentary systems, where and when the same political party controls both, a broadly similar concentration of power occurs with analogous effects on the third and fourth values.¹⁸ By contrast, where and when different parties control these two branches, or there is no majority party in the legislature, then accountability of the executive to the legislature and the latter’s independence to deliberate over proposed bills are often achieved at the expense of effective governance, due to the resulting gridlock and/or fragmentation of power.

This general background problem for reconciling the values of democratic governance is exacerbated by certain specific features of contemporary party systems. As referenced above, party systems should not only be categorized by the number of parties – single, dominant, two-party, multiparty, etc. – or by the regime they operate in – presidential versus parliamentary parties – but also by whether political parties are polarized, hyper-partisan, and/or fragmented. These are, obviously, distinct but overlapping pathologies. Although parties typically occupy different spaces on the relevant policy and ideological spectrum(s), polarization refers to a situation where the major parties or blocs are close to the opposite poles and far apart in their basic platforms and orientations, leaving the center of the spectrum relatively vacant. Hyper-partisanship generally references the way that parties and their supporters interact with, and treat, each other: do they engage in “hardball,” eschew cooperation and accommodation, act as if unconstrained by practical norms of bi-/multipartisanship, treat opponents as enemies or traitors, maximize the use of power for partisan ends. Although such hyper-partisanship is more likely to occur where polarization exists, it can happen without (for example, where one party breaks away from another, where the major parties cluster around a similar space on the spectrum and need to distinguish themselves, or where parties are personality rather than policy based) and need not happen with. A fragmented party system is one in which either (a) popular support is divided among several or many parties, without any one party or coalition of parties achieving majority or clear plurality support, or (b) such division takes place within, rather than between, the major parties.

These features of many contemporary democratic party systems are making the task of achieving and reconciling the four values significantly more difficult. Polarization and hyper-partisanship can render the task of forming and maintaining a government where no single party has a majority more complex and time-consuming, undermines the accountability and deliberative functions of the legislature where one party controls both branches, and makes gridlock worse during divided government. Fragmentation undermines effective and stable government, making it more difficult to obtain and sustain a governing majority and legitimate authority. Fragmentation and splintering of political power in general, and of party systems in particular, may currently be the most challenging problem bedeviling democracies around the world and has several causes.¹⁹ These include alienation of ordinary voters from the mainstream center-left and center-right parties that have mostly governed since the end of World War II for a mix of economic and cultural reasons, a realignment of party politics away from the traditional left-right axis based on socio-economic position and educational level, and the communication revolution that has enabled new parties, individual politicians, and even single citizens to bypass traditional parties and media outlets to reach mass audiences via social media, etc.²⁰

9.2 How the General and Special Problems Arise in All Widely Adopted Democratic Regime Types

Although the three basic and most common forms of democratic government were, in principle, intended to achieve and balance all four values, albeit in different ways, once the operative effects of electoral and party systems are taken into account, reconciliation is more difficult. Let us briefly see how and why for each of the six major democratic regime types, looking first at the “general problem” and then superimposing the contemporary special one.

In practice, the promotion and reconciliation of all four values has been hardest to achieve in parliamentary systems. In theory, as with the other forms, this is not so. The partial fusing of executive and legislative powers bolsters the effectiveness and stability of government, while still retaining full political accountability to a representative and deliberative legislature. But factoring in the impact of the electoral and party systems substantially changes the equation and balance. In two party, Westminster-style parliamentary systems resulting primarily from the majoritarian (and usually first past the post) voting system, effective and stable government is achieved at the expense of both genuine (as distinct from formal) political accountability to the legislature and inclusive, collective deliberation of the contents of legislative proposals. As indicated in the Duverger quotation above, this is due to the power and control of the typical governing (i.e., majority) party. Because of the necessary party discipline resulting from the “sink or swim together” political logic of the single election for both the executive and legislature, this regime type standardly concentrates power in the governing party, of which the prime minister is the leader, so that it typically controls the legislature through its majority. This, in turn, means that its survival is more or less assured (unless it acts in ways that cause a rebellion among its backbench members), political accountability is mostly reduced to somewhat theatrical exchanges with the official opposition party, and “government bills” that dominate the timetable are ordinarily steamrollered through the legislative process. This “ordinary” concentration of power in a majority party and its leadership (which has on occasion been referred to as an “elective dictatorship”)²¹ has been extended and abused by authoritarian populist regimes in parliamentary systems, such as those led by Orban and Erdogan (pre-2017), to further consolidate and entrench their power by undermining all independent institutions and sources of power.

With recent fragmentations of party systems, and the resulting greater likelihood and experience of coalition or minority governments,²² legislatures have become somewhat more independent of government control, leading to greater political accountability, representativeness (through the greater influence of smaller parties), and deliberation. But, as reflected in the chaotic period in the United Kingdom before Brexit occurred, this was very much at the expense of effective and stable government.

In other words, the United Kingdom at this time looked more like the second type of parliamentary regime, the multiparty one resulting from having a proportional representation election system. Here, and especially where there are not two blocs formed by allied parties, the traditional difficulty of reconciling the four values is the converse of the two-party version. Without a single majority party, effective and stable government can be difficult to achieve, in some cases notoriously so, but, on the other hand, the lesser concentration of power and its greater dispersal among parties may lead to a more independent and representative legislature with more scope for holding the executive accountable (including parties withdrawing support from a coalition government) and assembling ad hoc (rather than preordained) legislative majorities on particular bills. With polarization and/or fragmentation, the risks to effective and stable (coalition or minority) government are that much greater and the probabilities of ad hoc majorities for accountability or legislative purposes are smaller.

The presidential form of government, invented out of necessity in the United States, was designed to create effective and stable government through the direct election²³ of a legislatively irremovable single-person executive for a fixed term of office, while a separated, independent, and more representative legislature would engage in executive oversight and have the institutional freedom to exercise its major, legislative, function in a deliberative manner. Under majoritarian, two-party presidential systems, the reality has long been “separation of parties, not powers”: either a unified government where the same party controls both branches, with a high concentration of power and significant control of legislative outcomes, or a divided government with different parties in control of the two branches and the resulting risk of legislative gridlock.²⁴ As with two-party parliamentary systems, which they resemble,²⁵ unified presidential governments are often effective and stable but at the price of legislative accountability and deliberativeness. Divided governments are frequently ineffective, if stable, due to legislative paralysis, although oversight of (the often-increased reliance on) presidential unilateral authority is typically robust.

With polarized, hyper-partisan parties, the concerns about overly concentrated power in a unified government tend to be even greater, as bipartisan accommodation and restraints disappear, and the gridlock resulting from divided government is that much deeper and insurmountable. In this regime type, fragmentation tends to occur within, rather than between, parties so that even unified governments may be ineffective and unresponsive, as presidents find themselves unable to fulfill their legislative agendas due to internal opposition, as prominently recently in the first terms of Presidents Trump and Biden.

Apart from the United States and the Philippines, all other countries adopting pure presidentialism employ PR for legislative elections, as here the claims of representation that this voting system maximizes appear to trump the less relevant governance benefits of majoritarian systems. And yet such benefits turn out to be highly relevant as multiparty/PR presidential systems often suffer from the absence of a presidential party, or any party, majority in the legislature resulting sometimes in ad hoc support for presidentially sponsored bills in the absence of the more continuous coalitions needed to sustain the executive in parliamentary systems but sometimes in paralysis and ineffective government.²⁶ This is one of the well-known recipes for the “Linzian nightmare”²⁷ of presidential coups in Latin America and elsewhere. Although such ad hoc majority-building may suggest the potential for more independent, deliberative legislative processes and presidential oversight, achievement of these values is frequently undermined by weakly institutionalized, more personality-focused presidentialist parties, as compared with at least mainstream parliamentary ones. With fragmentation in particular, these particular pathologies of multiparty presidentialism tend to increase.

As the newest widely adopted form of democratic government,²⁸ semi-presidentialism also in theory promotes all four values.²⁹ A directly elected president who is not politically removable by the legislature ensures stability for the full length of the fixed term, even where the prime ministerial government that is fully accountable to parliament falls and changes beforehand. There are, as it were, two paths to effective government as either of the chief executives, or better both working together, can provide it. At the same time, having two chief executives produces less concentrated power than the “executive personalism”³⁰ of the fully presidential model and, at least vis-a-vis the president, a more separated legislature to engage in both oversight and deliberation of executive legislative proposals.

As with the other forms, however, factoring in electoral and party systems often renders achievement and reconciliation of these value a more complex and difficult task. The two party/majoritarian version of semi-presidentialism risks the most highly concentrated political power of all, where the president’s party controls the legislature, for here a president is effectively (although not formally) also the head of a parliamentary party and government. In earlier work, I have referred to this possibility as “super-presidentialism.”³¹ In this scenario, effective and stable government comes at an even higher cost in terms of accountability between elections and legislative deliberateness than in majoritarian parliamentary systems.

Where the prime minister is from the other major party, this resembles the situation in such parliamentary systems, with the exception that here, the effective leader of the opposition is the more powerful figure of the president. The timing of presidential and legislative elections, whether or not they are simultaneous, tends to be key to the probability of these two outcomes,³² as it is with unified or divided government under pure presidentialism. With polarization, such “cohabitation” risks further undermining the effectiveness of this regime type,³³ and fragmentation can leave even a newly elected president with reduced legitimacy,³⁴ as well as heading an internally divided majority party, although the constraint of sustaining the parliamentary government means that these divisions are likely to play out less than in pure presidentialism. A Senator Manchin veto may doom a presidential legislative policy or nominee but not the party’s hold on governmental office itself.

Finally, multiparty semi-presidentialism raises a risk to effective government that the two-party version rarely does; namely, fragmentation of power to the extent that neither presidential nor prime ministerial authority can be sustained. This risk, which characterized the Weimar Republic for the final half of its existence,³⁵ in a sense combines that of both other multiparty regime types. On the other hand, where this risk does not materialize and where the party system is more parliamentary than presidential in nature, legislatures may be in a better position to fulfill their accountability and deliberative functions. Again, polarization and hyper-partisanship may undermine the bases for necessary inter-party alliances and agreements, and the contemporary fragmentation of political power makes the risk of this regime type even greater.

9.3 Semi-parliamentarism

As noted above, parliamentary systems of both two-party and multiparty versions have traditionally found it hard to reconcile the four values, with each version prioritizing two different ones in ways that risk failing to achieve the “minimum core” of the other two. This is perhaps even more pronounced in the two-party, Westminster-style version because the normal majority party required for effective and stable government combined with the sink or swim political logic of the single election typically enables it to control and dominate the legislature and the legislative agenda. The resulting loss of genuine political accountability has been a major reason that many parliamentary systems have enhanced the legal accountability of government by establishing forms of judicial review for the first time.³⁶

In the last few years, an alternative to the standard two types of parliamentary regime has been proposed that it is claimed better reconciles and optimizes the four values.³⁷ This alternative has been labeled “semi-parliamentarism.” It is generally based on, although a modification of, the closest real-world examples, at both the national and state levels in Australia, and its main proponents have advanced a couple of different versions or sub-types. As stated above, the four defining components of this regime type are: (1) the absence of the direct election of the chief executive, (2) two directly elected legislative chambers, (3) only the first chamber can dismiss the cabinet in a no-confidence vote, and (4) both chambers have equal legislative power and exercise of the second chamber’s veto cannot be overridden by an ordinary or absolute majority of the first.³⁸ In addition to these purely institutional arrangements, the key to the claimed advantage of this regime type in better reconciling the values is the requirement of different voting systems for the two chambers in an attempt to ensure that the same party does not control or have a majority in both: specifically, a majoritarian voting system for the first chamber, to promote effective and stable government, but a PR voting system for the second chamber, to promote accountability and legislative deliberativeness.

I find semi-parliamentarism to be an innovative and promising democratic regime type that has the potential to offer a superior version of parliamentarism to the two existing ones. Part of its merits are the internal resources it brings to bear for addressing the contemporary challenges of polarization, hyper-partisanship, and/or fragmentation that render the balancing of values more difficult. The composition and legislative powers of the second chamber create incentives for multipartisan, ad hoc, issue-specific negotiation and accommodation among represented parties that might temper polarization. In a sense, it also deals with the contemporary problem of fragmentation of power by both channeling and celebrating it. So, the attempt is to contain the inter-party version in the first chamber through a majoritarian voting system that over-rewards the two leading parties but to increase the number of parties (and so representation) in the second chamber through PR.

Among its principal expositors, Steffen Ganghof argues that semi-parliamentarism is clearly superior to presidentialism because it provides the same benefits of separated powers between the executive and (here, the second chamber of the) legislature but without the unnecessary costs of “executive personalism.”³⁹ Although he is less clear or categorical about this, it also appears to be superior for him to both pure parliamentarism (because of the absence of separated powers) and semi-presidentialism (because most of the costs of executive personalism survive, in terms of a directly elected, irremovable chief executive, despite the existence of a second, parliamentary one).⁴⁰ Tarunabh Khaitan also suggests that (his “moderated parliamentarism sub-type” of) semi-parliamentarism is superior to the alternatives because it “combines the most attractive elements of each” while still yielding a stable regime.⁴¹ But he also makes the more modest claim that moderated parliamentarism is one way to optimize what he has identified as four constitutional principles relating to political parties,⁴² as well as to balance governmental effectiveness and accountability.

For reasons of space, my aim here is not to discuss whether or not Ganghof’s arguments for the superiority of semi-parliamentarism over all three widely adopted forms of government are compelling. For what it is worth, I believe that his conception of the separation of powers is somewhat narrow and constrained; it seems to imply, for example, that pure parliamentary systems lack this value altogether.⁴³ I also think the contrast between presidential and (modern) parliamentary systems in terms of the “executive personalism” of the former – that is, executive power being located in a single person versus a collectivity – is overdrawn in the modern era in which prime ministers are no longer simply “first among equals” and the office has become “presidentialized” in many countries.⁴⁴ But I do think his work is insightful and illuminating, so much so that it is worth exploring whether the principles and institutional ingenuity of semi-parliamentarism are perhaps exportable to other regime types. Similarly, in terms of Khaitan’s (more modest) claim that his version of semi-parliamentarism is one way to reconcile effectiveness and accountability, I want to ask whether there might be other ways and, in particular, whether the insights of the semi-parliamentary model that he has helped to develop are capable of being adapted for other regime types. Moreover, would some of the inherent features of semi-parliamentarism that might reduce current party pathologies be exportable to these others? Are there any independent design features that might be helpful in this regard?

9.4 Does Semi-parliamentarism Suggest There May Be Ways to Better Balance the Four Values in Non-parliamentary Regimes?

As we have seen, from the perspective of optimizing the four values, semi-presidentialism and (even more) presidentialism tend to achieve stable government for the duration of the presidential term – at least, absent the “Linzian nightmare” scenario – but, depending on political party strength and alignment, risk significant underperforming on the others.

Starting with presidentialism, a bicameral legislature with the same equal powers and separate voting systems as under semi-parliamentarism would likely, with a few additional modifications, have similar potentially beneficial effects. Effective and responsive governance would be bolstered by more or less ensuring that the president’s party has a majority in the first chamber for the full duration of the president’s term. This could be achieved by combining (a) a majoritarian voting system with (b) simultaneous executive and first chamber legislative elections and (c) making the first chamber term the same as that of the president (for example, four years). Importantly, this latter feature obviously rules out midterm legislative elections that often create divided presidential government and gridlock. To the extent that ineffective and unresponsive government is driving the current alienation from the more mainstream or centrist parties, this would help to address the problem. Given the presidential party’s likely majority in the first chamber, a presidential legislative veto – as a check on a potentially hostile, runaway, or “all-powerful” legislature – is likely unnecessary,

As with semi-parliamentarism, PR elections for the second chamber will likely enhance multipartisan, and so potentially more effective, executive oversight and accountability, as well as overall legislative representation of voters’ preferences and the deliberativeness of legislative processes. With likely no presidential or any single party majority, the prospects of both rubber stamping and continuous institutional gridlock are much reduced and the incentives for multipartisan, ad hoc, issue specific negotiations and accommodations for presidentially sponsored (and other) bills greatly increased. PR thus appears to be key to the values-optimizing goals of an “incongruent” second chamber. Other institutional permutations, such as different (and possibly staggered) terms and being elected at a different time from the simultaneous presidential and first chamber elections, seem less central than under semi-parliamentarism itself,⁴⁵ although certainly could be considered as potential supplements.

As for addressing and reducing the special problems, a number of inherent and possible additional features of such a regime hold promise. To push candidates away from the extremes and so reduce polarization (as well as increase responsiveness and representation), presidential elections should, as they do almost everywhere, require a direct national majority of voters. As the PR voting system for the second chamber is likely to result in the existence of more than two main political parties and presidential candidates, an instant or two-round run-off system would therefore need to be employed for presidential elections. Although not of course guaranteed, this increases the chances of an anti-polar majority at the second round, as we have now seen three times in France. The method of selecting party candidates for presidential (as well as legislative) office should similarly not reward more extreme or outlier positions, as, for example, the US system of party primaries can do by effectively bypassing both more knowledgeable party insiders and less motivated/partisan, ordinary party voters.⁴⁶

At the legislative level, this type of presidential regime again inherently counters polarization and hyper-partisanship by reducing the risk of general gridlock and creating incentives for issue-specific coalitions engendering multipartisanship and accommodation. Additionally, in terms of the majoritarian voting system for the first chamber, ranked choice voting (or a run-off system) is to be preferred to first past the post, for, as a true majoritarian rather than plurality system, it increases the prospects of less extreme positions and candidates. Again, party primaries that tend to be dominated by more motivated, partisan, and extreme party members/voters are part of the current problem and should either not be the method of selection at all or replaced by a “top four”⁴⁷ or equivalent method, to counter this effect. Independent redistricting commissions are also important, as having competitive general elections, rather than only primaries due to partisan gerrymandering, is key to reducing polarization.

In terms of fragmentation, again in a sense PR in the second chamber is designed to channel and reap its systemic benefits as compared with some of the costs of a two-party system that we have seen. If the risk or reality is of an overly fragmented second chamber exists, the common technique of capping it by employing voting thresholds for seats can be instituted. To the extent that the current fragmentation of parties and party authority is the result of the greater independence of legislative representatives stemming from the communications revolution and the potential for individual following and fundraising it has created,⁴⁸ this independence could be reduced in at least two ways. First, a closed list PR system for the second chamber would make individuals more dependent on the party, for their ranking on the list, and in this way replace older, intra-party seniority control systems. Second, whether through law (where possible) or internal regulation, funding and campaign finance rules for all elective offices could also render politicians more dependent on party and less on their own, individual fundraising.

With semi-presidentialism, the goal of achieving greater optimization of the four values revolves around attempting to avoid both the undue concentration that occurs when a president is effectively also the head of a parliamentary government, as party leader and the undue fragmentation where no party or stable bloc has a legislative majority. Essentially the same set of institutions, powers, voting systems, and reforms as with presidentialism just discussed would increase the probability of such an intermediate outcome. With simultaneous first chamber and executive elections, same terms of office, and a majoritarian (preferably ranked choice or two round) voting system, the result is likely to be presidential control,⁴⁹ but the key difference from the “super-presidential” scenario is the likely absence of a presidential, or any single, party majority in the directly elected and co-equal second chamber, where PR is employed. Combined with most of the other features and reforms discussed above for presidentialism, the potential result is to support effective and responsive government but without either undue concentration of power or built-in gridlock, enhance accountability to, and the representativeness of, the legislature, to create incentives for issue specific deliberation and accommodation across parties, and thereby also lower the risk of extreme fragmentation.

9.5 The Role of Constitutional Law

Democratic regime types are constituted by the combination of form of government (institutional powers and relations) and the operative voting and party system. This is why, for example, two-party and multiparty parliamentary systems are distinct types (or sub-types), generally furthering opposite values of democratic polities. It is a traditional blind spot of constitutional scholars to focus on the first part of the combination only, without the second. But a constitution should ordain and establish a regime type and not merely a form of government and so should presumptively include the key party and voting variables of the chosen type. Not only does this render it more difficult for elected officials to deliberately change or undermine a given democratic regime, but it also protects that regime from more unintended, serendipitous, or gradual alterations in its constituent parts.

As we have seen, the versions of presidentialism and semi-presidentialism that may better balance the four values and address some of the pathologies of contemporary party systems contain a distinctive set of institutional powers and relations that should presumptively be constitutionalized, as per the modern norm. To recap, these include direct national election of the president by a majority of the popular vote; two co-equal and directly elected legislative chambers, with the first having the same term as the president and a simultaneous or nearby election day, whereas members of the second chamber have different (and possible staggered) terms and/or are elected at a different time.

The key party/voting feature that is co-constitutive of these distinct democratic regime types is the employment of different voting systems for the two chambers, a majoritarian system – and preferably ranked choice voting – for the first, and PR – presumptively the closed-list variety – for the second. It is key, in significant part, because the respective party systems, two governing party contenders in the first house and multiparty in the second, largely follow from this choice. Accordingly, this should also be constitutionalized. Otherwise, depending in part on the legislative tie-breaker rule employed,⁵⁰ it might be possible for the first chamber to repeal an ordinary statute PR requirement for the second. Given the design requirement of single-member constituencies for the first chamber, the polarizing possibility of artificially eliminating competitive general elections through partisan gerrymandering should be minimized by also enshrining independent districting commissions in the constitution. This would likely have a knock-on effect of reducing partisan incentives for holding party primaries that select more extreme candidates, so that constitutional regulation of parties in this regard may be unnecessary.

9.6 Conclusion

This chapter represents an initial exploration of the possibilities of incorporating variations of political party and electoral systems into constitutional design for the purpose of addressing the twin contemporary democratic challenges of fragmented and unduly concentrated political power. It generalizes from the insights of proponents of semi-parliamentarism to suggest that its core promising feature of “symmetrical and incongruent” bicameralism can usefully be adapted to create equivalent versions of presidentialism and semi-presidentialism. By seeking ways to render all types of democratic government more effective and responsive, the chapter aims to counter the fragmentation and dysfunction that is helping to drive the appeal of more polar political forces, including authoritarian populisms of left and right. At the same time, it also seeks to curb the overly concentrated power that single party (or alliance) control of executive and legislative branches risks. For not only does this undermine the democratic values of accountability and legislative deliberation, but such concentrated power has been exploited and abused in recent years by various authoritarian populist regimes, both those with and without clear electoral majorities. In these ways, the aim is also to address some of the causes and consequences of such regimes. Obviously much remains to be done in terms of filling in the details.

10 Courts as Constitutional Rule-Makers for Elections and Parties Some Comparative Evidence

Aziz Z. Huq

10.1 Introduction

After declaring independence in 1960, the people of Côte d’Ivoire have faced the legacy of a long and difficult struggle to achieve democratic self-rule. Impeding their path was thirty-plus years of single-party rule by President Félix Houphouët-Boigny, successful and failed coup attempts, and a bloody civil war that ripped the country apart along a north-south line. Against this fractious backdrop, the presidential election of 2010 pitted incumbent Laurent Gbagbo against his former minister Alassane Ouattara and other candidates. When a runoff between Gbagbo and Ouattara was called in the challenger’s favor (54.1 percent to 45.9 percent), Gbagbo appealed to the national Conseil constitutionnel, alleging fraud. The Conseil, a judicial body patterned on the French model, threw out 600,000 votes and declared Gbagbo the winner. This decision struck many impartial observers as fraudulent and counter democratic.¹

A year later, the Awami League government of Bangladesh repealed the Thirteenth Amendment of that nation’s Constitution. This provision had created a system of non-party “caretaker” regimes prior to national elections, headed by a former chief justice, charged with the impartial management of polls. Under these caretaker regimes, Bangladesh experienced three elections (in 1996, 2001, and 2008) “widely lauded as free and fair.”² The later polls in 2014 and 2018, without a caretaker regime, were marked not just by declining competitiveness and captured legislative bodies but also sharp increases in “legal and extralegal measures to silence critics, weaken the opposition, and create a culture of fear.”³ In effect, the involvement of a judicial actor – albeit in a nontraditional, and arguably non-judicial role – allowed democracy to thrive.

In both Côte d’Ivoire and Bangladesh, an actor identified with the higher judiciary – either an apex court or its former chief – exercised a constitutionally defined power over the administration of national elections and parties. In the first case, judicial involvement arguably derailed fair democratic choice. In the second, a judicial actor’s involvement enabled a fairer measuring of popular judgment. At minimum, therefore, these cases caution against a rush to judgment for or against a robust judicial presence in the constitutional law of parties and elections. A global answer to this question quickly meets powerful counterexamples.

More modestly, this chapter explores theoretical and analytic foundations of these questions. It first aims to clarify some theoretical premises of this significant constitutional design choice. It then develops an analytic taxonomy of potential judicial tasks in managing elections with an eye toward democratic stability. This is complemented with an enumeration of potential risks. I make no claim that this theoretical and analytic ground-clearing yields sharp prescriptions about the role of courts in protecting democracies, although it does clarify the stakes of that functionality. This chapter offers, in the end, a crisper, less obstructed vantage upon the choices at play, rather than a decisive prescription.

Four other limits to the chapter’s scope need emphasis. First, I focus on the role of apex courts (broadly defined) in setting out basic ground rules for democratic choice. I do not address the interaction between ordinary litigation and election management. Almost any kind of jurisdiction can yield litigation that somehow “affects” an election. When a visiting session of the Lefortovo court sentenced the late Alexei Navalny to nine years’ imprisonment in March 2022 on fraud charges,⁴ for example, it was clearly influencing the possibility of democratic competition in Russia – perhaps in deep ways that were not evident at the moment of judgment. But it was not defining or applying the ground rules for elections or parties as such.

Second, and relatedly, there are many other aspects of intergovernmental relations and legislative process that might be thought essential to a functioning democracy beyond elections and parties. Gardbaum, for example, rightly singles out the legislative failure to hold an executive accountable as a political-process failure.⁵ Keeping with the more limited focus of this volume. I will home in more narrowly on the immediate accoutrements of democratic choice.

Third, I am concerned with courts’ function in democratic contexts, broadly defined to include some instances of competitive authoritarianism but not purely authoritarian ones. (The latter do run elections of a sort. And when de facto independent courts in non-democratic contexts are allowed, these do seem to mitigate the risk of illegal manipulation.)⁶

Finally, I focus on constitutional, not statutory, rules and institutions. The last distinction is a touch artificial. Constitutions can be entrenched to greater or lesser extents than statutes. Entrenchment can vary within a single document, for instance, via targeted eternity clauses. Whole constitutions can be self-consciously styled as “temporary,”⁷ just as institutions – think of Bangladesh’s caretaker governments – can be “intermittent.”⁸ To assume a stable variation in the entrenchment of statutes and constitutions, therefore, is unjustified. My use of the term “constitutional,” therefore, should be understood to connote a relatively high degree of entrenchment compared to other state institutions in a given polity.

10.2 Theoretical Coordinates

What role should apex courts play in the constitutional law of elections and parties? Is theirs a function other bodies can serve equally well? And if there’s choice between courts and alternative constitutional institutions, what considerations should guide designers? In the first instance, these questions sound in constitutional theory, oriented to offer guidance on the different functions necessitated by a commitment to democracy. Often, constitutional theory implicitly adopts the position of an imaginary constitutional designer. It implicitly assumes that theoretical and empirical work can, in tandem, at least narrow the range of plausible design choices. What is “optimal” may depend heavily on political and economic context, so specific prescription can be infeasible. A minimal “Hippocratic” insight into the range of feasible options that “do no harm” may be the best normative constitutional theory can do.⁹

It is a mistake to think that a democratic constitution must be composed solely of democratic parts. To the contrary, institutions insulated from partisan politics play necessary roles in fostering the persistence over time of democratic choice.¹⁰ A polity in which police and election officials responded to every whim of elected actors would not be democratic very long. Recent work by Gardbaum and Khaitan builds on this basic premise. Both start from the Madisonian intuition that constitutions invest individuals with legal and political authority, which can be used to advance or undermine the structural presuppositions of future popular choice under a stable constitutional frame. Even if some policy choices inevitably commit future actors,¹¹ a democratic constitution at least requires that electors retain the ability to change their minds at least about who is in power – although perhaps this is not alone enough¹² – so as to yield a tolerable measure of uncertainty and hence rotation of political office. To the extent it is conceptualized as intertemporally durable, democracy demands constraints at odds with elected actors’ incentives.

Gardbaum’s theory of “comparative political process theory” identifies the defense of “institutional structures and political processes within which democratic politics operates” from “erosion, corruption, and capture” as a distinctive design problem.¹³ Echoing Hans Kelsen’s celebration of the judiciary as the “guardian” of constitutional norms,¹⁴ Gardbaum isolates “judicial review” as the institutional vessel for that defense.¹⁵ Khaitan echoes Gardbaum’s concern about the instability of non-self-executing constitutional norms. In contrast to Gardbaum’s focus upon courts, though, he posits a need for “guarantor institutions” (or “fourth branch” bodies) with legislative and executive functions. He criticizes the “lawyerly blinkers” that lead scholars to “ignore” actors other than judges. The latter have both expressive and also material capacity (which courts are said to lack). They act either before or after a constitutional norm is violated.¹⁶ Khaitan’s point has empirical resonance: by 2019, some 64 percent of states had some kind of independently managed electoral system.¹⁷

Written constitutions have in the last fifty years taken up both Gardbaum’s judicial path and Khaitan’s fourth-branch proposal. Figures 10.1 and 10.2 plot the percentage of constitutions in force at any given time that contain respectively, a court explicitly tasked to resolve election disputes, or some form of election management body.^* They show how the adoption of courts preceded that of election commissions globally by about twenty years and, further, how neither model is obviously numerically dominant. The choice between courts and fourth-branch institutions (if imagined as a matter of substitutes and not as possible complements) hence remains a live one.

Figure 10.1 Percentage of extant constitutions with courts engaged in electoral supervision.

Figure 10.2 Percentage of extant constitutions with commissions engaged in electoral supervision.

Both accounts leave questions open. Gardbaum, for example, does not address the possibility, illustrated in the Côte d’Ivoire case, that courts can be instruments of entrenchment. It thus does not explain how to stymie judicial capture. Khaitan also invites inquiry into how to create institutions insulated from ordinary politics, how to maintain such insulation, and to how to respond in the event of capture. He further leaves open the question of whether (or which) guarantor institutions are “normatively desirable” at all.¹⁸ If the Côte d’Ivoire case invites the thought that courts, acting alone, cannot alone be assumed to be faithful guardians in a Kelsenian mold, a different study of election commissions in four Central American countries paints a rather dismaying picture of slow decay and partisan capture.¹⁹ More generally, a designer may be uncertain about the nature of future threats to democracy (which might bare differently on courts as opposed to commissions), the risk that political leaders will find ways to capture or subvert “guarantor” bodies, and her own ability to memorialize norms in ways that have consistent effect over time. Even as tremendous work has gone toward elucidating the forms of democratic failure writ large, in short, there is relatively little systematic work on how or why institutional guardians of democracy founder.

Current theoretical work further leaves open how to taxonomize the distinctive tasks involved in electoral guarantorship and the related puzzle of how to parcel out those tasks among different bodies. In part, this gap abides because there is no clear sense of what, quite specifically, “democratic maintenance” work on parties and elections involves. With a more fine-grained taxonomy of such tasks in hand, it may well be possible to discern whether and when institutional parallelism overlap, or even monopoly control by one body is desirable, or at least not a serious error.

10.3 A Taxonomy of Judicial Responsibilities Related to Elections and Parties

Keeping a system of competitive political parties and electoral choice in good working order is no simple matter. It requires attention to many different legal and constitutional mechanisms. There are many sharp corners during any representative process at which a hazardous and destabilizing political judgment can throw the democratic project off track. Election-related activity often begins when popular discontent with an incumbent regime bubbles over into oppositional associational action by the public. Parties are formed; old ones dissolve, merge, or reboot. Platforms are drafted, voters courted. Coalitions must be formed. Candidates or parties demand a line on a ballot. Voting happens, without or without irregularities. Counting follows, and an outcome is reached – or not. And so on and – one hopes – on, again and again: The promise of democracy resides in its endless capacity for iterative revision, of messing up, and starting again. From the ex ante perspective of the constitutional designer, a choice must be made about which of these diverse moments falls under constitutional regulation.

This section organized the heterogeneous range of such possible platforms for election regulation in a constitution into four broad categories. Setting these out, I avoid the assumption that each one must be subject to judicial supervision. Rather, there is a wide range of possible forms of choices about what to regulate and also how. Hence, a specific problem might be under a court’s supervision, regulated by a constitutional (or statutory) fourth-branch body, left without any regulation at all, or even insulated from state regulation via negative constitutional rights against the state. I bracket this choice of regulatory instruments and instead offer a taxonomy of moments in the democratic process in which judicial intervention can be imagined. A four-prong taxonomy, in my view, captures the range of potential issues in a tractable way. To the extent feasible, I illustrate this taxonomy with examples from outside the American context: The latter is relatively familiar and well-studied, and in many ways “exceptional.”²⁰

a. Entry rules for voters, candidates and parties. Constitutional regimes for voters, candidates, and parties tend to point in different directions. Respecting voters, constitutions tend to be inclusive. Article 42 of the Ghanaian Constitution, for example, declares that “every citizen of Ghana of eighteen years of age or above and of sound mind has the right to vote and is entitled to be registered as a voter for the purposes of public elections and referenda.”²¹ Until a 2012 Supreme Court decision, however, adult prisoners in Ghana were not permitted to vote. As in South Africa and Kenya, a constitutional judgment redrew the electorate’s boundaries to draw in the incarcerated.²² By way of counterpoint, the Indian high court has invalidated a statutory prohibition that negated felon disenfranchisement only for legislators.²³ The right to vote might also be implicated by practical hurdles, such as registration deadline, residency rules, and identification demands.

Voters are numerous and candidates, to some extent, replaceable. In contrast, parties tend to be fewer and far less fungible. As Schattschneider famously said, “political parties created democracy, and modern democracy is unthinkable without them.”²⁴ So constitutional regulation of parties has higher stakes for democratic stability. The risks of both their regulation and their freedom are acute. One implication is the allocation of constitutional rights to associate to parties, which appear to have generally positive effects.²⁵ Another is the possible prohibition of parties that would otherwise command meaningful public support. I focus on the latter here.

Starting with the German Basic Law, constitutions have included “militant democratic” party bans as prophylactics against anti-democratic political formations (as discussed at more length in the Introduction to this volume). Article 21(2) of Germany’s Basic Law, their locus classicus, flatly bans political parties that “seek to impair or abolish the free democratic basic order or endanger the existence of the Federal Republic of Germany.”²⁶ A recent survey of legislative and constitutional party-ban provisions in Europe found around thirty-six examples, excluding prohibitions on parties on ideological grounds (65 percent), for violent activity (56 percent), as a means to protect the political order (44 percent), or for undemocratic internal party functioning (5 percent).²⁷ These provisions are largely enforced by courts, which rely for evidence on parties’ platforms, leaders’ statements, and members’ activities.²⁸ Administrative agencies play a larger role within individualized lustration regimes, addressed below.²⁹ It is worth considering whether this forum choice influences the way in which democratic threats are evaluated. An administrative process, for example, might focus less on ostensible policy ambitions (courts’ core concern) and more on empirical evidence of, say, party linkages to foreign, anti-democratic forces (e.g., the Russian government) or paramilitary formations.

A variant on “militant democracy” bans is the use of rules setting a threshold, or floor, of votes before a party obtains any legislative representation. Thresholds of 3 to 5 percent are common in party-list proportional representation systems.³⁰ In 1952, the German Bundesverfassungsgericht invalidated a Schleswig-Holstein law that imposed a 7 percent threshold, citing grounds of political equality and the need for a compelling justification for thresholds greater than 5 percent. It subsequently invalidated vote threshold rules that disfavored smaller parties from the former East Germany in the wake of reunification. The Bundestag responded by amending the election law to ameliorate the constitutionally flawed threshold rule.³¹ The net result of these cases is that changes to Germany’s election framework that seem likely to fence out smaller parties receive close judicial scrutiny, at the same time as the basic law raises the specter of judicial exclusion. This implies a judgment that courts, and not legislatures alone, should manage party disqualification. Germany is not alone in this practice. In 1998, the apex court of Italy invalidated a threshold that disfavored the representation of linguistic minorities in the Trentino-Alto Adigo region.³²

One reason to favor judicial settlement of individual and party bans is that such prohibitions implicate considerations of due process that are familiar to judges and well-suited to resolution by courts. The law is singling out a person or association for a handicap not imposed on similarly situated actors. Courts may be able to draw on deeper jurisprudential resources to elaborate fitting procedural safeguards and evidentiary thresholds. The existence of a familiar template for judicial involvement lowers the expected adoption and error costs associated with courts’ intervention. The same, however, may well not be true for challenges to vote thresholds – which instead implicate questions of judicial competence in respect to more synoptic judgments about election systems in the round.

b. Electoral system review. The cases concerning the validity of vote thresholds in proportional representation systems are a reminder that constitutional courts can exercise jurisdiction over abstract (otherwise known as “facial”) challenges to different elements of electoral systems. These cases present the question of conformity between a constitutional norm and the verbal contents of election law, bracketing any inquiry into how that law is applied on the ground. Generally, such challenges are adjudicated before an election. They seem less common than more specific charges of fraud or malfeasance in a particular election, a class of jurisprudence that I take up below. But the two categories obviously overlap.

A common question for review involves the carving up of a polity into geographic districts for representational purposes. In 1961, Ireland’s High Court struck down a districting plan characterized by “grave inequalities.”³³ Between 1986 and 2010, France’s Conseil Constitutionnel invalidated several electoral districting statutes with large deviations from an equal population baseline. It twice instructed the French legislature to redraw all the nation’s districts.³⁴ Other electoral system features can also be subject to judicial review. In 2008, for example, Indonesia’s Constitutional Court upheld a mandatory gender quota for party lists but invalidated a vote threshold for party-list candidates.³⁵ In 2013, Italy’s Constitutional Court struck down a 2005 election law because it used closed party lists and assigned a majority party a “bonus” in seats. These features, reasoned the Court, violated Article I of Italy’s Constitution (popular sovereignty), Article 3 (equality before the law), Article 67 (representation as national, not local); and Article 48 (freedom of the vote). In an unusual further step, the Italian court issued a remedy in the form of a new, purely proportional, electoral system using a preference voting system. This system served the interest of neither party coalition. Their leaders Matteo Renzi and Silvio Berlusconi quickly negotiated an alternative scheme.³⁶

Systemic challenges can also be lodged in international judicial bodies. Since 1986, the European Court of Human Rights has also permitted challenges to European law concerning elections to the European parliament.³⁷ And in 2016, the African Court of Human and People’s Rights ordered Côte d’Ivoire to amend its electoral commission law to bring it into compliance with an impartiality principle in the African Charter on Democracy, Elections and Governance. The international court, though, stepped in only after a substantial challenge had been filed and rejected by the nation’s high court.³⁸

A final relevant class of cases involves the legal environment in which elections are run. Constitutions might extend a positive constitutional protection to speech, spending, and conduct constituting a campaign. Alternatively, they can impose negative restraints on how and when campaigns are conducted. Litigation of such rights that yields judgments of general scope will shape the electoral environment. They may indirectly change outcomes, but the frequency and identity of outcome-dispositive voting rules is hard to evaluate.

c. Election integrity claims. Specific complaints about election integrity seem more common than abstract challenges. A “surprisingly large” number of elections are closely fought.³⁹ Emotions are likely to run higher during such heated contests, with accusations of impropriety (or perhaps even its appearance) being more common. Across Anglophone democracies, moreover, partisan identification – both negative and also positive – appears to play an increasingly large role in national politics.⁴⁰

The judicial reversal of a national election result on grounds of specific illegality or malfeasance seems rare. (The invalidation of subnational elections, by contrast, is more common, as Camby illustrates).⁴¹ The example of Côte d’Ivoire’s 2010 poll seems exceptional, not illustrative. Hence, when presidential polls in Taiwan in March 2004 led to an incumbent victory by 0.22 percent, the loser sought and obtained a court-ordered recount that reduced, but did not flip, the margin of victory.⁴² In 2013, the Kenyan Supreme Court declined to set aside Uhuru Kenyatta’s first-round victory in a presidential race against Raila Odinga.⁴³ Four years later, in 2017, after yet another closely fought election between Kenyatta and Odinga, the same Court accepted the latter’s complaints of election irregularities, nullified the poll, and ordered fresh voting within sixty days. Because Odinga refused to participate in this second round – citing concerns about the election commission’s integrity – judicial intervention did not ultimately change the election’s outcome.⁴⁴ In a variant on this dynamic, international courts in Africa, rather than their domestic counterparts, have been reviewing election disputes in Burkina Faso, Nigeria, Kenya, and Tanzania.⁴⁵

The logical limit-case of these possibilities is Bangladesh’s caretaker government regime. Recall that this looked to a former chief justice to administer government during the run-up to an election.⁴⁶ In effect, it assumed that the risk of illegal self-dealing by the incumbent government would be unacceptably large. Rather than adjudicate challenges after the fact, it imposed a blanket prophylactic remedy of control by judicial personnel. The judicial character of the caretaker regime, indeed, seems to have been important to its success.

Such widespread practice aside, there is also good reason to doubt that judges generally do a good job catching or preventing irregularities. A threshold reason is their lack of relevant expertise in recondite and technical matters of election administration. A second problem is impuissance. The path of the 2017 Kenyan election suggests that even when a court perceives irregularities, it may lack the political capital to force a change in election outcomes. The latter also illustrates the possibility of tension, even conflict, between a judiciary and a fourth branch institution with different views of a vote’s legality – a point to which I return below.

On the other hand, the power to consider specific allegations of fraud or malfeasance in respect to a particular election is distinct from the power to consider whether, in the abstract, a legal framework contained in statutes and regulations comports with constitutional norms. The two genres of litigation usually turn on different evidentiary grounds and demand different kinds of judicial competencies. The distinction between specific and abstract review, though, should not obscure a more profound connection between these two genres of litigation: A well-designed and well-functioning legal structure for parties and elections minimizes the opportunity for strategic action by powerful state actors intended to derail an anticipated electoral defeat. Post hoc judicial review of specific objections might usefully identify vulnerabilities in the electoral framework, providing the legislature with the information necessary to “patch” the system. Alternatively, such review might have a corrosive “moral hazard” effect: Elected officials know that they can appeal, after the fact, to the bench in cases where something goes awry. Ex ante, therefore, they have less urgent reasons to anticipate and fix vulnerabilities in an electoral system. The exercise of judicial review, which is often assumed to be the sine qua non of legality, hence creates a dynamic unravelling of the rule of law around elections. A possible way to mitigate the force of such incentive effects is the judicial use of a “penalty default” remedy. This is a judicial solution disfavored by all – perhaps akin to the Italian constitutional court’s ruling in 2013 – that forces hegemonic partisan actors to revise the election law rather than compete under an intolerable disposition.

Ginsburg and Elkins offer a more minimalist justification of the judicial role in settling close elections.⁴⁷ Likening a closely contested election to the game-theory model of “chicken,” they posit that a judicial resolution can operate as a “self-enforcing focal point” that mitigates the risk of overt conflict even if it is not a particularly accurate factfinder. This role, they note, need not be played by a court; the latter is just a “convenient third-party [that] draws on the imagery of a neutral dispute resolver.”⁴⁸ This logic, though, assumes a relatively narrow temporal focus on the moment after a close election. It does not account for the risk of strategic false claims by a likely loser that an election is closer than it really is: Again, the problem can be characterized in terms of moral hazard, with the risk of upstream strategic action rising as the expected accuracy of the third-party adjudicator decreases.

d. “Exit” rules for candidates and parties. Just as law can impose front-end hurdles to entering the political process so too can law create chutes for expelling individuals or groups that pose a threat to the democratic process. Democratic exclusion varies not only by whether it operates at a granular, individual level or a coarser group level but also whether it hinges on past “bad” conduct or a prediction of future harm. I have addressed party bans above because they can be thought of as ex ante barriers to participation. I hence focus on individual disqualification here because they commonly hinge on a finding of “bad” past action.

At one end of this spectrum is the familiar impeachment remedy. Almost all (90 percent) of constitutions with a presidency speak to impeachment. Crimes and constitutional violations are the most common bases for removal. A lower legislative chamber usually begins an impeachment by a supermajority vote, and ex post judicial review is often, albeit not always, available. Between 1990 and 2018, impeachment was proposed at least 210 times in 61 countries, against 128 different heads of state. But only ten removals were ultimately carried out, some of which involved judicial review of a legislative removal decision.⁴⁹ Impeachment is not the only pathway for exclusion. Israeli courts have adopted an aggressive program of removing officials and blocking appointments based on a judge-made concept of “good character,” deeming officials ineligible from remaining in or holding office if currently under indictment.⁵⁰ On the other side of the ledger, lustration rules might be subject to constitutional challenge rather than simply being enforced. In 2003, for instance, Indonesia’s Constitutional Court invalidated a prohibition on the election of Communist Party members.⁵¹

Given their stakes, disqualification disputes unsurprisingly can spill into larger questions of institutional power. India provides an example. After the Allahabad High Court invalidated the election of Indira Gandhi on corruption-related grounds, the Indian parliament enacted a constitutional provision making election disputes involving the prime minister and speaker non-justiciable. In 1975, the Indian Supreme Court invalidated that amendment, citing its “basic structure” doctrine.⁵² Hence, a discrete dispute about one politician become the basis for a more sweeping ruling about judicial power.

e. Summary. Apex courts are in practice involved in a wide range of election and party-related disputes. In some instances, they perform tasks assigned by a constitution (e.g., enforcing rights, resolving factual claims of irregularity, and disqualifying parties or candidates). At other times, they clarify the meaning of the organic law (e.g., in abstract challenges to the constitutionality of election-related provisions). A minimal conclusion from this survey is negative: Given the heterogeneity and numerosity of potential disputes, there is unlikely to be a single “right answer” to the question of when courts should be involved, or not, in the constitutional regulation of parties and polling. The world is too complex for a single, simple solution of that kind.

10.4 Institutional Choice in the Constitutional Supervision of Elections and Parties

If there is no “one size fits all” institutional choice covering all of these examples, the question arises whether there are reasons to assign some tasks to courts and some to fourth-branch bodies. The choice is nonbinary: Courts might be either substitutes or complements for fourth-branch bodies, such as election commissions. Indeed, as Figure 10.3 shows, a nontrivial number of constitutions contain both election-specific judicial powers and also an election commission. This overlap creates the possibility of both cooperation and conflict.

Figure 10.3 Proportion of extant constitutions with both judicial and fourth-branch electoral bodies.

The case studies discussed above push in different directions as to this choice. The Côte d’Ivoire example points toward skepticism about courts and hence an emphasis on nonjudicial bodies. In South Africa, the Constitutional Court and the Public Protector have worked in tandem in respect to high-level corruption relevant to the potential disqualification of the prime minister.⁵³ Or an election commission might be subject to judicial oversight, as was the case in Kenya in 2017. That oversight can be intense or mild. In Taiwan, for example, courts “respect[] the regulatory authority” of the Election Commission by deferring to its judgment.⁵⁴ The Bangladesh example further hints that judicial actors can be embedded within an election management infrastructure to sustain constitutional norms. There is, in short, a quite wide array of potential relationships between courts and other bodies: cooperative, adversarial, entangled, or dominating.

There is good reason for resisting a strong presumption in favor of courts or fourth-branch bodies. To see this, consider three obstacles confronted by any constitutional body charged with maintaining democratic hygiene in and around elections: All three afflict both courts and non-judicial bodies in roughly equal measure. First, an institution with sufficient insulation from contemporaneous partisan pressures must get off the ground. Not only must that institution be designed with robust insulation from the very beginning, but that design feature must prove enduringly effective. The initial constitutional design of either a judicial or a fourth-branch body can be flawed. In a study of election management bodies, Michael Pal identifies gaps in their organic documents that facilitate capture.⁵⁵ Similarly, the American federal judiciary was designed with two assumptions in mind: that the nation’s Senate would be a nonpartisan body and that the supply of potential judges would be limited enough that selection could not be used to partisan ends. Both assumptions faltered within a decade of ratification. As a result, the American system for appointing federal judges is directly partisan, and partisan projects infuse the operation of the national judiciary.⁵⁶ One sort of democratic success, that is, engendered a different kind of democracy-related failure in another element of the Constitution.

Good design requires constitutional founders who eschew a partisan cast of mind and who are capable of anticipating evolving and as-yet unanticipated strains on the democratic process. Obviously, these won’t always be to hand. Perhaps the most obvious circumstances in which good design will come to the fore arise after a widely maligned (e.g., a fascist or apartheid) regime has collapsed, and there is intense public and geopolitical pressure for a fresh start.

Second, even a well-designed court or fourth-branch body institution can be subject to overweening informal (or illegal) pressures that compromises their good operation. For the past three decades, the Indian judiciary has exercised almost complete control over the formal processes for appointments to the higher bench. This insulates it from direct partisan pressure. The Supreme Court, indeed, invalidated a 2014 statute creating a National Judicial Appointments Commission. Yet the judiciary has recently turned a “blind eye” to controversial steps by the Bharatiya Janata Party (BNP) of Prime Minister Narenda Modi, such as a controversial campaign finance law, extrajudicial detentions in Jammu and Kashmir, and the construction of a Hindu temple on the site of the Babri Masjid. Judicial quiescence appears to be the result of “subtle” administrative measures, such as slow-walking promotions and background checks and resisting calls for better funding.⁵⁷ More troubling allegations of corruption on the part of a new chief justice and blackmail by the BNP suggest that even most robust legal defenses of institutional independence might be vulnerable to circumvention by sufficiently unscrupulous political operatives.⁵⁸ Similarly, Zambian Chief Justice Mathew Ngulube was forced to resign in 2002 after it was found out that he had received bribes from incumbent President Fredrick Chiluba while deliberating on an election-related petition.⁵⁹ Where an incumbent resorts to extralegal measures to undermine democratic choice, the institutional choice between guardianship bodies will have limited salvific effect. What is more important is the strength of the larger supervisory regime of criminal law pertaining to political corruption.

It is not clear whether courts or fourth-branch bodies can best resist this pressure. Tushnet⁶⁰ has suggested that the modal adjudicative tasks of an apex court often force it to take positions that have inescapable partisan connotations. Tasking judges with election-related dispute resolution may amplify the risk they will be tarred as “party-political”.⁶¹ And Tushnet’s concern clearly extends, as he recognizes, to fourth branch institutions. And it may be that apex courts are more vulnerable because their ordinary docket has high political stakes. Or it might be that an apex court has a sufficiently large load of non-partisan-coded cases that it has a greater capacity to absorb or deflect “party-political” criticism. It is hard to know which way this consideration cuts in advance.

Third, it is well known that a verbal specification of a norm may be an imperfect proxy for an underlying constitutional value. Over time, a specific formulation may do an increasingly bad job of tracking that value. A more general verbal formulation, in contrast, may require new judgments over time as to how best to apply it across new and different circumstances over time. The negative right to speak free of governmental coercion, for example, is plausibly thought to be necessary to a well-functioning democracy. But that negative right may become increasingly irrelevant if it is possible for government to crowd or drown out critical, adversarial speech.⁶² As a result, a court or guarantor institution tasked with maintaining a healthy democratic public sphere must revise radically how “democratic debate” is realized to avoid obsolescence at the hands of new digital media technologies. Again, a potentially paradoxical dynamic can be traced: A court that has successfully defended a negative-rights account of the democratic sphere may be so beholden to that history, intellectually or as a matter of institutional pride, that it is unable to recognize a decisive evolution in the sociopolitical environment. Blinded by success, it is driven to failure.⁶³

There is, in short, a complex blend of countervailing pressures at play across a range of constitutional design decisions. It is, in particular, difficult to see an overwhelming case for either courts or fourth-branch bodies. Both courts and non-judicial bodies are vulnerable to threshold design flaws and illegal influence. Perhaps the longer historical pedigree of judicial independence offers a more secure psychological and social foundation for resisting drift and corruption; or perhaps the bien-pensant myth of judges who stand above the fray makes their partisan allegiance more difficult to discern and critique. In respect to the third concern raised above, it is easy (but probably wrong) to assume that courts are worse than legislative or executive bodies at updating principles’ legal entailments under new circumstances. The Canadian Charter of Rights and Freedom has famously been viewed as a “living tree,” capable of fruitful adaption to new environments.⁶⁴ Max Weber’s idea that bureaucracy can be hindered by an “iron cage” of rigid doxa and habit has grown into cliché – but holds more than a grain of truth. The dominance of either “living tree” or “originalist” metaphors in the high-level legal culture of a polity may have a greater impact than the abstract choice between judge and bureaucrat.

To the extent that even a rough cut at the question of institutional choice is appropriate, Tushnet’s suggestion that overlapping guarantor institutions may well be ex ante desirable seems a good starting point.⁶⁵ Rather than trying to read the tea leaves to discern the political future, a designer might write in multiple, seemingly redundant bodies so that at least one does not fail when confronted by an unexpected challenge.⁶⁶ Yet even this guidance requires caveats: Where particularized election disputes are concerned, a plurality of forums for contesting an outcome may induce forum-shopping, inconsistent judgments, and even outright conflict because of the lack of a focal-point resolution.⁶⁷ Where an election-related dispute requires a quick answer therefore – usually because operative power hangs in the balance – it is likely better to avoid overlap and second-guessing. Perhaps this will lead to a higher rate of erroneous decision-making, but that may be better than a higher-rate of election-related conflicts.

10.5 Conclusion

Apex judiciaries have played a significant part in administering elections in many jurisdictions. Success stories as well as cautionary tales abound. To an extent, the courts are likely capable of being replaced by fourth-branch bodies. But the design of the latter raises many of the same questions about “independence” as a judiciary’s design. Reinventing the wheel might do little to improve the quality of democracy. For this reason, there may be some small reason to prefer courts in a jurisdiction where de facto judicial independence has been achieved, and more reason to look to fourth-branch institutions where that public good remains an elusive and unrealized goal.

11 The Durability and Dynamism of American Indian Constitutional Reform

Elizabeth H. Reese

11.1 Introduction

There are hundreds of self-governing tribes in the United States. Not all are democracies. Not all have constitutions. But because of cultural and colonial forces, the constitutional democracy is undoubtedly the most common form of tribal government. The constitutional provisions that structure the democracies of American Indian tribes are not only strikingly different from one another, they are by in large far more in flux than the constitutions of United States federal and state governments that colonized them. Within their constitutional structures, tribes are trying out different representative structures, different qualifications for their elected leaders, different restrictions on the right to vote, different practices for voting, and granting different parts of their governments the power to oversee elections.

As this chapter will discuss, many tribes are dynamically remolding their constitutional structures. In doing so they are demonstrating what I suggest is a notably high tolerance for reform or failures that catalyze reform. Nowhere is this dynamism more on display than in the different ways that tribes now politically order themselves – in the reshaping and restructuring of their representative democratic institutions. The takeaway from this chapter is that these tribal governments are carefully experimenting with democracy and that they view the need for institutional change as a moment of growth rather than a failure in their practice of self-government. Reforms have become an almost natural – if not celebrated – part of perfecting their government structure.

Part of this story of different attitudes to constitutional change is undoubtedly tied to complex origins of many tribal constitutions as well as a humility about the role of law and institutions within the identity of a nation. The two tribes profiled in this chapter demonstrate how tribes have survived and avoided democratic crises – not by having constitutional structures that are durable but by having constitutional cultures that accept the need for change and value resilience.

11.2 Uninspiring Constitutional Origins Stories

With a few exceptions,¹ the hundreds of self-governance documents that are tribal constitutions began as post-contact creations, written under complex political circumstances often involving varying degrees of pressure, or even oversight from the United States. Most prominently, under the Indian Reorganization Act (IRA) of 1934, Indian tribes were encouraged to adopt, by majority vote, written constitutions that were generally reminiscent of the US federal constitution.² The IRA not only made it clear that adopting tribal constitutions was the clear policy preference of the United States but also created an incentive for tribal governments to adopt a constitution: power.

The IRA specified that: “In addition to all powers vested in any Indian tribe or tribal council by existing law, the constitution adopted by said tribe shall also vest in such tribe or its tribal council the following rights and powers: To employ legal counsel; to prevent the sale, disposition, lease, or encumbrance of tribal lands, interests in lands, or other tribal assets without the consent of the tribe; and to negotiate with the Federal, State, and local governments.”³ While the right to retain counsel and negotiate with other governments were arguably rights that tribes already had or were exercising, the right to prevent the further sale of their lands without their consent loomed large against the backdrop of recent federal policies that resulted in the dramatic loss of tribal lands. Ultimately, 181 tribes accepted the IRA, while 77 rejected it.⁴ Today, around 60 percent of the 574 federally recognized Indian tribes have a constitutional system with origins in the IRA.⁵

Some tribes who did not adopt IRA constitutions nevertheless adopted written constitutions as a result of efforts from within the tribe to transform their government systems. This, of course, did not happen in a vacuum. The pressures of American colonialism and norms of American democracy had a tremendous effect on tribal decision-making. Some tribes chose American-style constitutional democracies more out of a desire to win favor or approval from the United States than out of genuine admiration for the system or a belief that it was best suited to their project of self-governance. For that reason, even tribes that seemed to have independently chosen constitutional democracies often harbored seeds of structural instability or doubts about the suitability of their constitutions for their own people.⁶ Many tribal governments’ initial constitutions were not an accurate, or even workable, expression of a tribe’s political culture or priorities.⁷ That is why the wave of tribal constitutional reforms in the last fifty years is both expected and vital.⁸

Because of these initial circumstances surrounding the adoption of their written constitutions, constitutional reform has become a highly conscientious reckoning for tribes. A reckoning about their past, present, and future identity as Nations.⁹ Many, if not most all of these tribes fit the description of postcolonial constitutionalism developed to describe the experience of many other newly independent nations in Africa and Asia who adopted nominal constitutions that largely reflected the constitutional order of their former colonizer and then entered into “a dialectical process involving an ongoing struggle between absorption and rejection of the former colonizer’s” institutions, political culture, and identity.¹⁰

What the constitutional design and reform process of American Indian tribal governments brings to the fore, however, is just how important and transformative the dialectical process itself is for political communities who are struggling not only with their relationships to their colonial past (and present), but with their connections to their precolonial past and the character of their hopes for a postcolonial future. Telling a story of political and cultural continuity throughout all these changes is a vital part of what holds these nations together. There is no sense that tribes have failed as nations if they did not get it right the first time or need to revisit their constitutional structure. Tribal political identities and pride in their self-governance persist despite moments where it was clear that their current government structure was failing them. Instead, there is comfort with recognizing and accepting the need for change – a comfort that has made them remarkably resilient.

The source of this comfort is no mystery. It developed as a reaction to the number of forced changes that tribal governments have experienced already and will undoubtedly continue to experience going forward. Throughout all these changes, these tribes describe their nations as political survivors. These reforms are not simply questions of what form of political ordering can best safeguard rights or express the interests of the people but who the people are now compared to who they were and who they hope to become.¹¹ No matter what path they choose, the narrative of continuity and survival in the face of change is an animating part of the tribal constitutional reform process. Even when the underlying reforms are quite drastic and could be characterized as re-foundings, they are, importantly, not viewed as such.

11.3 Stories of Reform

The remainder of this chapter showcases two examples of tribal constitutional reform concerning democratic rights and orderings and how the constitutional culture of continuity I described above got them through the democratic reform process, including constitutional crises provoked by factionalism. It presents how these tribes – the Cherokee Nation and the Citizen Potawatomi Nation – have remade the structure of their democratic institutions from the laws delineating citizenship, voting, and office holding to the configuration of their representative institutions themselves.¹² Specifically, it describes how both tribes pulled themselves out of periods of democratic crisis by reforming their constitutions.

These two tribes have notable differences. While they are both presently located in Oklahoma, they are originally from very different parts of the United States and had quite different precolonial political cultures. They are very different sizes – Citizen Potawatomi has approximately 37,000 citizens,¹³ while Cherokee Nation has 430,000 citizens.¹⁴ Although both are constitutional democracies today, Cherokee Nation became one long before the IRA, while Citizen Potawatomi is an IRA tribe.

11.3.1 Cherokee Nation

The Cherokee originally occupied territory that extended across seven states in what is now the American South. After a series of treaties between the Cherokee and various southern colonies and, then, eventually, the United States, the tribe had lost a great deal of land but secured promises that their remaining lands and the sovereignty of their borders would be protected thereafter.¹⁵ But the United States broke these treaty promises against the mounting pressure for land in the south and consistent attempts by the southern states to claim authority over Cherokee lands.¹⁶

Many of Cherokee Nation’s leaders hoped that by adapting to American ways, they could bolster the legitimacy of Cherokee Nation’s government and its land claims in the eyes of the United States – and maybe even prevent forced removal out west. The Cherokee developed a written form of the Cherokee language that was used to print a tribal newspaper, and in 1827, adopted a written constitution that established a republican form of government for the Cherokee.¹⁷ But removal came despite these changes. Congress, at the urging of newly elected President Andrew Jackson, passed the Indian Removal Act of 1830, and the Cherokee signed the Treaty of New Echota, which relinquished all their original territory in exchange for lands in present-day Oklahoma.

The 1827 Cherokee Nation Constitution created a three-branch system of government with a legislature, executive, and judiciary.¹⁸ The Cherokee legislature, known as the General Council, was composed of two separate bodies, a Committee and Council.¹⁹ Two representatives to the Committee and three representatives to the Council, were each selected by the same eight districts of the Cherokee nation that sent delegates to the constitutional convention.²⁰ The General Council was given the power to select the executive branch leader, the Principal Chief,²¹ and to make the laws governing Cherokee elections after the initial election of 1828, other than the requirement that elections be conducted via voice vote.²² The right to vote was extended to all free male citizens over eighteen²³ and eligibility for General Council was further limited to those over twenty-five.²⁴ Eligibility for the office of Principal chief was limited to men over thirty-five, and – like the United States Constitution – who were natural born citizens.²⁵ Though the constitution did not have a Cherokee Nation residency requirement, it specified that citizens who move away from the Nation and “become citizens of any other government” would lose the rights and privileges of Cherokee citizenship unless they petitioned the General Council for readmission.²⁶

After removal, the Cherokee Nation was reunited with the Western Cherokee who were already settled out west. The two governments passed a formal act of union²⁷ and held a new constitutional convention. The constitution ratified in 1839 was quite similar to the one that preceded it.²⁸ The legislature, now called the National Council, was tasked with dividing the new lands of Cherokee Nation into a new set of eight districts and given the power to add one or two more “if subsequently it should be deemed expedient.”²⁹ They also kept the practice of voice voting.³⁰ The largest change was to the method of selecting a Principal Chief. The office was now chosen by an at-large vote instead of by the National Council.³¹

The 1839 Constitution governed the Cherokee for about sixty years before the Curtis Act dissolved tribal governments in Indian Territory in 1907 as part of the preparation for transitioning the territory into the State of Oklahoma.³² The tribe elected a Chief for the first time since the Curtis Act in 1971, and his first order of business was for the Cherokee Nation to draft a new Constitution³³ for the reconstituted nation that would grow from a mostly defunct entity that funneled resources to members, into robust government once again.³⁴

The Cherokee Constitution of 1975 enacted substantial changes to the political arrangements of Cherokee Nation. These changes reflected the vastly changed political circumstances of the Nation, and a very different relationship to the United States than the one the Nation had in the nineteenth century. It did away with the bicameral legislature, instead providing for a single Council composed of fifteen council members who were all elected at-large.³⁵ This constitution also included the first membership definition, limiting Cherokee citizenship to those persons with ties to the Dawes Commission rolls.³⁶ The 1975 constitution lowered the age requirement for Principal Chief to thirty and a Cherokee “by blood” requirement was added to the qualifications for both Council and Principal Chief.³⁷ Voice voting was also eliminated, and secret balloting was adopted instead.³⁸

The Council retained the power to regulate elections,³⁹ however, the most notable change to democratic governance was the inclusion of an entirely new article dedicated to popular referendums, initiatives, and amendments. The people now retained the power to enact referendums, legislation, and constitutional amendments by majority vote and to independently place them on the ballot after obtaining the signatures of 5, 10, or 15 percent of registered Cherokee voters – calculated based on the “total number of votes cast at the last general election for the officer receiving the highest number of votes at such election” – respectively.⁴⁰ The constitution specified careful rules for the inclusion and presentation of measures in a popular election, ensuring that overlapping subject matters or measures included on one ballot could not confuse voters.⁴¹ Crucially, the people also retained the exclusive power to call for a constitutional convention and the possibility of a constitutional convention was required to be submitted to the people at least once every twenty years.⁴²

An unconventional change was a few instances of self-imposed subornation to the US federal government. The 1975 Constitution adopted US Constitutional supremacy, stating that “the Cherokee Nation is an inseparable part of the Federal Union. The Constitution of the United States is the Supreme law of the land; therefore, the Cherokee Nation shall never enact any law which is in conflict with any Federal law.”⁴³ Moreover, the 1975 Constitution required that all constitutional amendments be approved by the president of the United States before they could take effect.⁴⁴

After the passage of this new constitution, a series of amendments and Cherokee tribal court cases further tweaked or clarified the Nation’s democratic structure. In 1987, the people amended the constitution via referendum to do away with at-large voting for the Tribal Council and return to the old system of representation by legislative districts. The Tribal Council was directed to “establish representative districts which shall be within the historical boundaries of the Cherokee Nation of Oklahoma” that were to be “apportioned to afford a reasonably equal division of tribal membership among the districts.”⁴⁵ Some voters were taken aback by the Council’s decision to require citizens to register in their new districts for the 1995 election and initially challenged the Council’s power to institute such a requirement under its new constitutional authority.⁴⁶ The Nation’s Judicial Appeals Tribunal upheld the Council’s actions as a valid exercise of their newly specified powers.⁴⁷

The Cherokee Courts also upheld legislation implementing district-specific residency requirements on Cherokee political candidates as contained within the Council’s power to create districts rather than exceeding its power by creating another affirmative qualification for office not otherwise specified in the constitution.⁴⁸ Councilmen also brought a lawsuit challenging the Council’s decision to pass a law delaying the reapportionment that was necessary after 1994 until 2002 and successfully struck down the district maps that would have been in place in 1999. Council was ordered to draw new maps reapportioned to reflect a constitutional right to equal representation contained within the 1987 constitutional amendments.⁴⁹

During this period of judicial volleys and reform, however, the Cherokee Nation suffered through a widely publicized constitutional crisis.⁵⁰ In the 1995 general election, none of the candidates for Principal Chief received a majority of the votes, and so a runoff was scheduled between the top two candidates, George Bearpaw (38 percent) and Joe Byrd (29 percent).⁵¹ However, ten days before the runoff, two petitions were filed in Cherokee Courts claiming that Bearpaw was ineligible to serve as Principal Chief since he had an Oklahoma felony conviction from twenty years prior.⁵² The Tribal Election Commission disqualified Bearpaw and effectively handed the office to Byrd after setting aside all votes Bearpaw received in the runoff.⁵³ Based on this initial election conflict, battle lines were drawn, distrust was deeply ingrained, and a long shadow was cast over Byrd’s administration.

Once he took office, Byrd’s political opponents accused him of financial misconduct. Rather than cooperating with the investigators, Byrd delayed turning over documents, forcing the Cherokee Courts to order a search of his office. Byrd responded by first firing all the Cherokee marshals involved in the search and then – once he had been charged with obstruction of justice – leaning on his supporters on Tribal Council to impeach all three members of the Judicial Appeals Tribunal.⁵⁴ The marshals and justices kept showing up to work until Byrd sent officers to confiscate the marshals weapons and close down the courthouse. The tribal prosecutor who protested the shutdown was arrested for resisting and then Byrd fired the tribe’s other prosecutor – leaving the courthouse entirely empty as of August 1997.⁵⁵ Outraged, the minority of the Tribal Council who opposed Byrd’s actions started boycotting Council meetings in April 1998, denying them a quorum to conduct business for over a year until the courts were reopened.⁵⁶

What saved the nation from this political crisis was an unlikely provision of their constitution and a decision that predated the crisis. In the same 1995 general election that led to the Bearpaw–Byrd runoff, the question of a constitutional convention was on the ballot, as article XV, section 9 of the constitution requires must happen at least once every twenty years.⁵⁷ The Cherokee people had voted overwhelmingly for a constitutional convention, and after languishing for several years as a lower priority for the tribal council, the opportunity for reform coincided with the political crisis.

After consulting with outside experts on how to avoid letting the divisions take over the reform process, the Rules Committee decided to create a commission that would be composed of two representatives appointed from each branch and a seventh member chosen collectively by the first six.⁵⁸ Byrd loyalists would control the executive appointments, and the judiciary was staunchly anti-Byrd after his attempts to oust them. With the Tribal Council splitting their two members between Byrd and anti-Byrd factions that divided the body, the selection process ensured that the commission would be evenly divided. To further give credibility to the commission, the commissioners agreed to hold open meetings, took an oath of political neutrality, to act only upon unanimity, and to promise not to hold political offices in the new government. The Council attempted to limit the commission’s authority to an advisory role for the Council, but the commission successfully preserved its independence by threatening to walk away entirely.⁵⁹ The commission’s enabling legislation thus made it an entirely “independent commission” with the authority to put either a new constitution or set of amendments directly on the ballot.⁶⁰

The commissioners set about their work, holding twenty meetings across Cherokee nation and in major US cities where a large number of Cherokee citizens lived.⁶¹ The commission cataloged suggestions for reform by topic and frequency, accepted written testimony, and ultimately created an eight-hundred-page record of testimony from Cherokee citizens.⁶²

The next task was selecting delegates for the convention. The commission decided on seventy-nine delegates composed of the seven commissioners, eight delegates each selected by the current branches of government, twenty-four delegates selected from a pool of those citizens who testified at the hearings, and twenty-four chosen by lottery from a pool of citizens who had either submitted a written request to serve as a delegate or had given oral or written testimony.⁶³ The constitutional convention was called to order on February 26, 1999. After nine days of deliberation – the records of which capture around three-thousand pages of discussion – a new constitution was adopted by the convention on March 6, 1999.⁶⁴

The new constitution, notably, created term limits for both the executive and legislative officers,⁶⁵ as well as staggered terms for the appointed judiciary to avoid court-stacking.⁶⁶ It also addressed the removal problems that had so quickly escalated in the recent political crisis. The Tribal Council kept its power to remove all elected and appointed officials, but this power was now limited to specific for cause failings such as willful neglect or a felony, requiring a two-thirds majority vote for removal and only after a trial before the Tribal Council that afforded the accused officials due process.⁶⁷ The judiciary, moreover, was given an additional removal protection in the creation of a Court of the Judiciary – made-up of seven persons using the same appointment process as the constitutional commissioners – that had the power to recommend removal of judicial officers.⁶⁸ On top of all of these new removal protections, the people were given the power to separately recall any elected official.⁶⁹

To address the reality that 40 percent of Cherokee Nation citizens now lived outside the boundaries of the reservation, the 1999 Constitution created two new Council seats,⁷⁰ which would be chosen at-large by the off-reservation members of Cherokee Nation.⁷¹ A separate election commission was also created to independently administer elections.⁷² Finally, the 1999 Constitution removed the incorporation of the United States Constitution and the requirement that the president approve any amendments to the Constitution.

This last change, however, put the Nation in a bit of a legal bind. Article XV, Section 10 of the 1975 Constitution plainly required that no amendment or new Constitution shall become effective without the approval of the president of the United States or his authorized representative. The draft 1999 constitution had been submitted to the BIA, but the Nation spent years in an elaborate and, at times, contentious back and forth over its approval. At one point the Nation sought more limited approval for simply an amendment removing the presidential approval requirement. Talks eventually broke down in 2002, leading the Nation to give up and set elections for removing BIA oversight on May 24, 2003, and then the new constitution on July 26, 2003.⁷³ The 1999 Cherokee Constitution passed – narrowly – via popular vote on July 26, 2003.⁷⁴

Though the constitutionality of the new constitution’s adoption without presidential consent was challenged, the Nation’s Supreme Court ultimately held that the people had the “inherent sovereign power” to “remove the self-imposed requirement” of presidential approval,⁷⁵ and thus, “the 1999 Constitution of the Cherokee Nation became effective on July 26, 2003.”⁷⁶

Per the requirements of the 1999 Constitution, the Council has been tasked with drawing and redrawing the fifteen legislative districts. On one occasion, their map was struck down by the Cherokee Courts since the districts deviated by 22.8 percent from equal representation.⁷⁷ On another occasion, the districts were upheld when they came close to 10 percent deviation, as the court acknowledged both deference to the Council where possible and that the 10 percent rule was persuasive guidance.⁷⁸

Since 2003, the Nation has continued to have disputes over elections, but none have resulted in a similar crisis. In perhaps one of the best tests of the new Constitution’s institutions, history closely repeated itself, but the nation avoided the political turmoil and crisis that followed the Byrd–Bearpaw crisis. In 2019, just days before the election, the Cherokee Nation Supreme Court upheld the Election Commission’s decision, just a few weeks previously, to disqualify one of the leading candidates for Principal Chief for accepting illegal campaign contributions.⁷⁹ The candidate, councilmen David Walkingstick, had run a divisive campaign that criticized the Nation’s ties to the Democratic Party⁸⁰ and though he claimed the election was “stolen,”⁸¹ nothing else came of it. And more recently, in the 2021 election, a candidate for Tribal Council was disqualified by the Election Commission for accepting improper campaign contributions but did not contest the election even though he lost by only seven votes.⁸² The Nation even survived a Bush–Gore style election in 2011, where the election was so close that the winner changed several times in various recounts⁸³ and so the Supreme Court threw out the election and ordered a new one after deciding that it would be “impossible to determine the results with mathematical certainty.”⁸⁴

Through all of these election showdowns that tested the current formulation of Cherokee Nation’s institutions, they have held strong. And the legacy of Cherokee Nationhood remains one of change, continuity, and resilience all the way back to their original constitution. As their current Principal Chief, Chuck Hoskin, describes: “As the U.S. was growing around the Cherokee Nation … we changed how we governed ourselves …. We developed a written constitution. We elected a chief. We centralized in strength and our government so we could deal with what was happening around us.”⁸⁵

11.3.2 Citizen Potawatomi Nation of Oklahoma

The people who currently make up the Citizen Potawatomi Nation were once part of a confederation of tribes that included the Anishinaabe Nations of the Potawatomi, Ojibwe, and Odawa that originally controlled a large territory in the Great Lakes region of the United States.⁸⁶ However, by the late eighteenth century, white settlers and US treaties had eroded the Confederacy’s land base such that the United States was able to forcibly remove the three Nations.⁸⁷ In 1838, the Potawatomi were removed by the US military on what is called the “Trail of Death” from northern Indiana to Kansas.⁸⁸ In 1861, the Potawatomi in Kansas were offered a treaty that would require them to surrender tribal membership in exchange for US citizenship and private land ownership in Oklahoma.⁸⁹ Those who took the deal and moved to Oklahoma are now the Indians who make up the Citizen Potawatomi.

The Citizen Potawatomi’s original 1938 constitution was based on the Oklahoma Indian Welfare Act⁹⁰ model – an Oklahoma-specific IRA. It left all major decisions up to a “Council,” which was composed of all members over the age of twenty-one who attended a meeting on the last Thursday in June.⁹¹ The meetings were chaotic and unproductive, and so in 1985 the tribe passed a constitutional amendment that redefined its supreme governing body as the entire voting electorate allowed for referendums by ballot. This reform allowed absentee voters to participate for the first time, a vital change given the diaspora of their population.⁹²

In 1930, the vast majority of Indian people continued to live in rural reservation communities, with barely 10 percent of the total Native population living in urban areas.⁹³ However, as a result of federal relocation policies⁹⁴ and high rates of Native military service in World War II, by 1980 nearly half the population of Native people lived in urban areas.⁹⁵ The Citizen Potowatomi were no exception to this phenomenon. The Dust Bowl in the 1930s also contributed to their citizens moving away from Oklahoma, such that by the 2000s, two-thirds of the Citizen Potawatomi Nation’s citizens moved far from the tribe’s lands in Oklahoma and lived throughout the rest of the United States.⁹⁶ By this time, the tribe had politically stabilized and was experiencing rapid economic growth.⁹⁷ However, the Nation struggled with low voter turnout – especially from absentee voters – and a general sense of “apathy” about the nation’s government.⁹⁸ The Nation “decided to take the government to the people” and held meetings throughout the United States to solicit input from the large populations of off-reservation Citizen Potawatomi.⁹⁹ The result was a decision to completely overhaul the government in order to reengage the nation’s off-reservation population.

In 2007, the Nation amended its constitution to include a legislature comprising sixteen legislators.¹⁰⁰ Eight of these legislators are chosen from new legislative districts drawn to equally represent citizens who live outside the state of Oklahoma, while five are chosen from districts within the state of Oklahoma and three belong to the at-large elected executive officials.¹⁰¹ The tribe recognized that there was a risk that out-of-state constituents would be incentivized to undervalue in-state interests (such as land purchases) and so divided the legislature’s weighted representation. Although out-of-state citizens make up two-thirds of the population, they receive only half of the seats on the legislature.¹⁰² Similarly, an Oklahoma residency requirement ensures that executive positions are tied to in-state interests, although they are elected by Nation-wide popular vote.¹⁰³

In 2017, Citizen Potawatomi conducted its first ten-year constitutionally mandated redistricting to reflect population growth.¹⁰⁴ Overall voter participation for the nation doubled as a result of these changes in the structure of the legislature. Moreover, the Nation saw significant increases in other forms of civic engagement and cultural participation. With these reforms, the Nation also achieved geographic participation parity, with out-of-state voters comprising 67 percent of the vote.¹⁰⁵

The scope and impact of the Citizen Potawatomi system make it unique even compared to the handful of international examples of expatriate voting districts or other tribes such as the Cherokee Nation, which also provides for off-Reservation representation. Unlike in these cases, extraterritorial resident citizens are far from the exception. The Citizen Potawatomi extraterritorial districts make up half of their legislature. As such, Potawatomi legislative restructuring was not only a way to increase voter turnout and grant more of a voice to extraterritorial citizens, it was also a bold statement about how the Citizen Potawatomi view citizens across the country. These citizens are an important part of the Nation, as opposed to citizens who left the Nation by leaving its geographic borders.¹⁰⁶ Citizen Potawatomi pulled themselves out of a period of dysfunctional governance and voter apathy, like the Cherokee, through not only simple swift constitutional fixes but also constitutional changes that involved elaborate processes and rigorous processes that engaged their entire electorate.

11.4 Conclusion

The choices that the two tribes have made reveals the strange boon that the chaos of colonialism has been for them. There have been countless disruptions or interference with their ability to freely self-govern, and yet, the tribe’s political identity and will to self-govern persisted. The undeniable limitations and even outright failings of some of their governing documents allows for tribal constitutional reform efforts to not only be easier to achieve but seems to have allowed for tribes to think more creatively about their reforms than they would have if they were more strictly wedded to their prior version of democratic ordering. This complex postcolonial dialectic of institution building through constitutional reordering seems to be, in short, liberating for tribes who can contemplate and implement creative solutions to even their most basic political orderings problems.

The constitutional structure of representation and elections, in other words, has become one of the most important sites for American Indian tribal nations to set the boundaries and substance of their identities as nations. As Citizen Potowatomi Chairman Barrett puts it: “If you’re not in the constitution-fixing business, you’re not in economic development, you’re not in self-governance; you’re not sovereign.”¹⁰⁷ And as Cherokee Nation Principal Chief Hoskin describes, the resilience of tribal governments is a key part of their identity as nations: “We were torn apart politically; our economy was destroyed, [and y]ou would predict it would take decades or generations to come back if they ever did. But what happened next is I think is one of the most remarkable stories in the history … of this country and the world. We rebuilt within about a decade. We invested in a system of law and justice. We have to remember, and we have to remind people … what the Indian nations have gone through.”¹⁰⁸

A key takeaway from these experiences, is the value of seeing the usefulness of failure – even in a democracy or a constitution – and the durability of the commitment to self-governance as a continual process rather than a singular ideal that can be perfected and captured in a single document.

12 Eternity Clauses and Electoral Democracy

Silvia Suteu

Whether in the ‘old’ key of militant democracy or in the newer one of democratic backsliding, the question of how constitutions can insulate against the erosion of democratic institutions remains ever fresh. Much has changed in this landscape, however. Experiences with populists in power and authoritarian takeovers the world over have cast doubt on long-standing certainties. The faith in courts and constitutional review as preeminent tools of legal protection of democracy and fundamental rights has been shaken by the reality of captured courts and eroded judicial independence. With it, too, the belief that detailed constitutional bills of rights would reign in arbitrary power.

The search for legal institutions to uphold and strengthen democracy’s foundations has instead turned to other horizons. One of these, explored in this chapter, is the turn to eternity clauses and the prospect that constitutional unamendability could act as a stronger barrier against democratic erosion through otherwise legal means. The hope is a familiar one: that when faced with procedurally legitimate constitutional amendments that undermine or even ‘dismember’ the constitution, substantive hurdles should remain in place that sanction these amendments as illegitimate and unconstitutional.¹ We saw this hope raised and swiftly dashed in Hungary after the Orban government embarked on constitutional change to entrench its hold on power and before the disempowerment of the Hungarian Constitutional Court. During that period, some hoped the Court would embrace an unconstitutional constitutional amendment doctrine to allow it to prevent, delay, or at least signal the authoritarian takeover veiled in legality.² More recently, we witnessed Kenyan courts embrace the idea of substantive limits on constitutional amendment and even consider embracing a basic structure doctrine to block the president-initiated Building Bridges Initiative (BBI) package that would have transformed the Kenyan Constitution.³ We have also seen calls for the unconstitutional constitutional amendment doctrine itself to be adapted to the realities of our populist/authoritarian times, such as by renouncing judicial self-restraint in reviewing amendments and adopting a more holistic interpretation of their cumulative effects.⁴

Doubts have remained, however, including expressed by this author, as to whether unamendability is indeed the answer to democratic backsliding, or whether it is itself salvageable from the clutches of populists and authoritarians in power.⁵ That scepticism has been grounded in the ambivalent operation of unamendability in practice, whether as a bargaining tool during constitution-making processes or when enforced judicially. This reality includes the propensity of eternity clauses to entrench partisan hold on power as well as to essentialise political identity. As we will see, this ties into the complex relationship between eternity clauses and electoral democracy. The tension between unamendability and democracy has of course received ample attention. Comparatively underexplored has been the particular type of democracy eternity clauses seek to protect, how that relates to the specific constitutional context in which they are adopted, and how this more specific understanding of democracy influences the unconstitutional constitutional amendment doctrines developed by local courts. In particular, the relationship between unamendable democratic commitments and the electoral arena is ripe for close examination.

This chapter seeks to fill this gap. It explores the link between eternity clauses and electoral democracy by looking at two instances of applied unamendable democracy: party bans, whether direct or indirect, and the protection of parliamentary mandates. Both types of interventions are operated in the name of guarding democracy, whether against anti-democratic forces, as in the case of party bans, or against weakening core democratic institutions, as in the case of parliamentary mandates. These two approaches are illustrated via a range of case studies: the ban of anti-democratic parties in Germany; bans of ethnic, separatist, and religious parties in Turkey; indirect unamendability and its chilling effect on party competition in Israel; and the judicial protection of parliamentary mandates as unamendable in Czechia. These are indeed meant to be illustrations of the problems I discuss, rather than to be taken as prototypical examples. The wide range of democracy type covered provides insights into the very different understanding, enforcement, and effects of unamendability in consolidated, transitional, and hybrid democracies.

Underpinning this work is the belief that unamendability cannot be adequately understood, and its propensity as democratic defence evaluated, divorced from the constitutional politics within which it is embedded. As part of that politics, questions of electoral balance of power, health of the party system, and politicisation of court intervention must be faced head on. Doing so engenders scepticism about unamendability as an unquestionable ally in the fight to protect democracy. I hope to show that unamendability’s propensity to be misused and to lead to distorted outcomes is greater precisely in those contexts where it is most likely to be adopted: incomplete or fragile democracies seeking to entrench a path towards consolidation (hence also my choice of case studies). I also argue, however, that the bluntness and open-ended nature of unamendability risks having a chilling effect on electoral democracy in stable democratic contexts as well. Thus, we should not merely assume that eternity clauses and the judicial doctrines surrounding them will be democracy-enhancing. When we instead investigate their operation in context and across time, including by evaluating their effect upon the electoral arena, we find them to sometimes misfire or even backfire as democratic defences.

12.1 Unamendability and Electoral Competition: Party Bans

One could say the very essence of eternity clauses is to protect democracy from its enemies. We can view such provisions as a prime legal embodiment of the ethos of militant democracy: a constitutional democracy should be able to defend itself against those who seek to undermine its very foundations, including against those who seek to do so via constitutional amendment. In language that has now become the norm, eternity clauses would thus be viewed as prime weapons against ‘abusive constitutionalism’.⁶ They would thus complement other measures, such as electoral thresholds, designed to prevent the fragmentation of parliamentary politics as led to the downfall of the Weimar Republic, as well as party bans, aimed at preventing anti-democratic forces from even operating on the electoral arena.⁷ A recent attempt at classifying the constitutional elements of militant democracy listed unamendability alongside other tools such as term limits, loyalty oaths, the right to resist, emergency provisions, and civilian control of the military, to be deployed depending on the nature of the threat.⁸

The logic of party bans overlaps with that of unconstitutional constitutional amendment doctrines. Just as the latter seeks to prevent otherwise procedurally sound amendments when they substantively undermine democracy and the rule of law, so too party bans seek to prevent not just parties that advocate or engage in violence but also those that threaten ‘a “legal” anti-democratic takeover of the state apparatus’.⁹ The hope is that when such subterfuge is afoot, there remain legal resources for courts to intervene before the whole democratic edifice is taken down.

However, we can find more direct connections between party bans and eternity clauses. One of the distinctions between rationales for party bans traces a shift from Weimar-inspired bans to a ‘legitimacy paradigm’.¹⁰ The former are aimed at parties that seek to abolish democracy wholesale and have been enforced against Nazi, fascist, communist, and, more recently, Islamist parties. The latter seeks to justify the proscription of those parties that ‘threaten certain elements within the liberal constitutional order, such as commitment to equality and non-discrimination, the absolute commitment to a nonviolent resolution of disputes or secularism’.¹¹ This has led to bans on ethnic and religious parties, which have assumed the place of ideological parties in the postwar period.¹² However, the logic of Weimar – the fear that mass parties gone awry would destroy democracy wholesale – does not apply neatly to religious and ethnic parties, particularly in a pluralist, multicultural society.¹³ Challenges to political identity are vaguer, more diffuse, and therefore more elusive than frontal attacks on democratic institutions, and the danger of essentialising identity – itself a catch-all concept – may be inherent in such bans.¹⁴ Eternity clauses, especially those insulating state characteristics such as the form of government, territorial integrity or unity, official language or religion/secularism, are precisely aimed at defining such a political identity and placing it beyond the reach of political contestation.¹⁵

The move away from the Weimar paradigm, then, elevates the risk that party bans be abused for partisan purposes.¹⁶ One example would be government self-entrenchment against political opponents, the latter recast as enemies of liberal democracy and as such eliminated from political competition. There is also a heightened danger that party proscription follows a process of ‘securitisation’, such as bans in the name of protecting ‘national communities from challenges to core identities and values’.¹⁷ Religious, ethnic, and regional parties, whose banning may compound discrimination already experienced by the communities they represent, are most likely to be cast as ‘existential threats’ to the state and as a consequence would also see legitimate avenues for political expression and contestation closed off.¹⁸ Insofar as the status quo is the baseline against which unlawful party ideology and behaviour is to be measured, parties organised precisely to contest that status quo become pariahs by default.¹⁹ As we will see, party bans in conjunction with constitutional unamendability compound these dangers and judicial oversight may not in fact act as the neutral safeguard some have hoped it to be.²⁰

I will discuss three instantiations of these different rationales: party bans in the name of a democratic principle enshrined in an eternity clause, illustrated by Germany; bans of ethnic, separatist, or religious parties in the name of unamendable secularism or territorial integrity and unity, such as in Turkey; and indirect restrictions, where parties are not banned outright but prevented from standing for elections for alleged breaches of state ideology, as in Israel. Insights from other national contexts are brought in where relevant. These examples show how unamendability has been deployed to reinforce democracy not just at a high level of abstraction or in its minimal understanding but in response to locally specific evaluations of democratic threats, sometimes with the effect of significantly skewing the electoral arena. In some cases, such as Germany’s, courts have been astute at modulating the forcefulness of their intervention over time, balancing the threat posed by a given party against the anti-democratic effects of its ban. In other instances, however, such as Turkey’s, courts have adopted a much more rigid approach, reinforced by an extensive eternity clause. In others still, such as Israel’s, party bans have reinforced an ever more exclusionary notion of citizenship.

I will show that the entrenchment of democracy through eternity clauses, whether explicit or implicit, is not always limited to minimal conceptions of democracy. Nor, indeed, is it always interpreted by courts and other constitutional actors as leaving room for the contestation of a single, sometimes exclusionary conception of democracy. In practice, such interpretations have led to party bans and interventions in parliamentary politics that at least in some instances have silenced reasonable disagreement and reduced electoral competition. As such, and like other types of eternity clauses, there is a dark side to democratic unamendability of which it is imperative we remain vigilant.²¹ Put differently, the constitutional entrenchment of democratic commitments to the point of rendering them unamendable may yet undermine rather than strengthen those same commitments.

12.1.1 Anti-democratic Parties

The most straightforward case for a democratic defence involving prohibiting parties would seem to be that involving those organisations advocating for democracy’s very demise. Germany’s Basic Law is often analysed as the epitome of a constitution that embraces militant democratic goals. It does so, among other means, by enshrining the democratic principle (Article 20(1)), which it then renders unamendable through the Ewigkeitsklausel in Article 79(3). Importantly, however, the Basic Law for the first time recognised political parties as constitutional actors and enshrined their protection as well as their duties in Article 21. Thus, they are recognised as opinion formers but also required to be internally democratic, publicly accountable financially, and – according to Article 21(2) – they are subject to the Federal Constitutional Court which can rule them unconstitutional should ‘their aims or the behaviour of their adherents, seek to undermine or abolish the free democratic basic order or to endanger the existence of the’ state. Article 21 therefore aims to walk the tightrope between extending constitutional protection to parties, in direct response to the perceived failures of Weimar and its poorly institutionalised party structure,²² while at the same time requiring them to abide by the democratic rules of the game under threat of unconstitutionality.

There have been two successful party ban cases in Germany: the Socialist Reich Party (the party-heir to the Nazis) and the Communist Party, both in the immediate post-war years.²³ In banning both parties, the Constitutional Court laid out its test in such cases as involving assessing a party’s internal structure and public actions and statements and opting for the ban only when the party seeks to topple supreme fundamental values of the free democratic order that are embodied in the Basic Law. The Court did not, in the two cases, rely on Article 79(3) for its determination. The link to the eternity clause was indirect, insofar as the democratic principle under threat was unamendable. Its unamendability signalled its centrality to the constitutional order, as well as its non-negotiable status.

More recently, the German Constitutional Court changed its approach to party bans. This was seen in the 2017 attempt to ban the Neo-Nazi Nationaldemokratische Partei Deutschlands (NPD), which the Court declined to do on account of the party’s perceived electoral insignificance and the strength of German democracy against such threats.²⁴ The Court no longer found it sufficient for a party to be shown to pursue anti-constitutional aims; proof of its potential to be successful would now also be required: ‘a presumption that the criterion of “seeking” has been met only if there are specific weighty indications suggesting that it is at least possible that a political party’s actions directed against the goods protected under Article 21(2) GG may succeed (potentiality)’.²⁵ Given the extreme nature of a ban, the Court would henceforth impose one only if the political party has sufficient means to exert influence due to which it does not appear to be entirely unlikely that the party will succeed in achieving its anti-constitutional aims, and if it actually makes use of its means to exert influence.²⁶

In other words, even while the criteria developed in the 1950s cases might have otherwise led to a ban, the Court balanced this against the perceived consolidation of German democracy, which was deemed robust enough not to need to go down the more militant route of a ban.

Importantly, the 2017 judgment for the first time clarified the relationship between Article 21(2) and the eternity clause in Article 79(3). Insofar as specifying the meaning of “free democratic basic order” in the former, the Court explained that ‘its regulatory content cannot be defined by means of general recourse to Art. 79(3) GG but is limited to those principles which are absolutely indispensable for the free democratic constitutional state’; instead, the Court anchored its meaning in ‘the principle of human dignity (Art. 1(1) GG), which is specified in greater detail by the principles of democracy and the rule of law’.²⁷ The Court thus explained that it would read Article 21(2) in a more limited manner, concentrating ‘on a few central fundamental principles which are absolutely indispensable for the free constitutional state’, invoking the importance of the political will–formation role of parties.²⁸ The content of the eternity clause goes beyond this minimal conception of democracy, the Court said, such as by protecting republicanism and federalism.²⁹ Given that ‘constitutional monarchies and centralised states can also be in accordance with the guiding principle of a free democracy’, the Court would not ban parties on account of challenging these unamendable features of German democracy.³⁰

This aspect of the 2017 judgment has been termed ‘surprising’ and ‘certainly not warranted by the case at hand’.³¹ The narrower interpretation of Article 21(2) aimed at aligning of German law with European human rights law in this area. The European Court of Human Rights (ECtHR) has assessed party bans to include both acceptance of democratic contestation of the current dispensation of state principles and structures³² and a higher ‘imminent threat’ standard for assessing the danger posed.³³ With this move, however, the German Constitutional Court has been viewed as selecting a ‘core of the core of the Grundgesetz’ that in practice might allow a party to advocate unconstitutional change that could only be achieved by violating the eternity clause – a ‘stunning’ result.³⁴

The 2017 judgment also had a series of important consequences. In doctrinal terms, it means the German Federal Constitutional Court has now added a timing, contextual element to its assessments of party ban requests. Thus, the substantive test of whether the party opposes the democratic order is now complemented by a ‘risk calculation’ test that looks at the potential of that party to realise its goals.³⁵ It has been argued that the Court created a new category of party in Germany: one that engages in anti-constitutional activity but lacks the potential to realise its aims.³⁶ Article 21 has also been amended to enable the removal of funding from this new category of ‘anti-constitutional but not unconstitutional’ parties,³⁷ with the Federal Constitutional Court retaining sole competence to decide on such funding stripping. The practical effects of this change, as we know from the literature on indirect party bans achieved via restrictive regulation, may yet amount to a de facto ban.

These changes have been controversial, with some viewing Article 21’s amendment as introducing a form of party differentiation that breaches the principle of party equality in German constitutional law.³⁸ The German eternity clause comes back into the picture insofar as it insulates from amendment the principle of democracy enshrined in Article 20(1), which could be seen as preventing such unequal treatment among parties. Following this line of interpretation might even lead to a finding that the amended text of Article 21 amounts to ‘unconstitutional constitutional law’.³⁹ At the very least, the 2017 judgment introduced an element of uncertainty regarding the interpretation of the core of the Basic Law. Uncertainty also now exists about the application of the new standard for determining when a ban is to be imposed, insofar as the Court left open the questions of how many seats should a party have or how close to power should it be before it is deemed dangerous are now open-ended questions.⁴⁰

12.1.2 Ethnic, Separatist, and Religious Parties

A more complex case is that of parties said to be organised along ethnic or separatist lines, whose purported threat to the democratic state would amount to their challenging of its territorial makeup, as well as that of religious parties, whose attack on secularism has been viewed as an attack on state foundations. Turkey is infamous for its rich experience with both types of party bans. According to one study, there have been twenty-seven party bans in Turkey between 1961 and 2019, banning either Kurdish separatist parties (said to breach unamendable territorial integrity) or parties seen to promote political Islam (said to breach unamendable secularism).⁴¹ Another study looking at the 1983–2015 period, found Turkey overrepresented among European party bans with sixteen out of fifty-two (31 per cent).⁴² A recurrent feature of Turkey’s democratisation process,⁴³ party bans have not been limited to electorally insignificant actors. They have included parties with significant parliamentary presence and even part of ruling government coalitions.⁴⁴

One might wonder about the relevance of including an ‘incomplete democracy’ such as Turkey’s in this analysis. However, party ban studies have found such ‘incomplete democracy bans’ to be the largest category of party bans (at least in Europe), especially when it comes to sub-state nationalist parties.⁴⁵ Turkey’s example is also illustrative for how bans on salient parties come about and their effects, whether we consider national (as in the case of the Welfare Party) or sub-national salience (as in the case of Kurdish parties). Especially when considering party success at the sub-national level, we find similar considerations present when banning parties in Germany, Spain, Belgium, or Greece.⁴⁶ Moreover, in terms of the link to unamendability, a democratising context such as Turkey’s is fertile ground to test eternity clauses’ ability to protect fragile democratic gains and foster abidance to constitutional democracy.

Turkey’s constitution contains several unamendable provisions. Article 4 renders unamendable the republican character of the state as well as Articles 2 and 3, which in turn entrench, among others, the state’s democratic and secular character as well as its territorial integrity. These principles are embedded in the constitutional text also outside these two provisions, however. Thus, secularism is further protected by the preamble that mandates ‘that sacred religious feelings shall absolutely not be involved in state affairs and politics as required by the principle of secularism’, by Article 13 as a ground for rights limitations, and by Article 14 on the abuse of rights, among others. Article 68 of the Turkish Constitution explicitly requires party statutes and programmes to respect the independence of the state, its indivisible territorial and national integrity, human rights, equality and the rule of law, national sovereignty, and the principles of the democratic and secular republic. The Turkish Constitutional Court thus had a rich textual panoply on which to construct not only its unamendable constitutional amendment doctrine⁴⁷ but also its party ban case law.

When it comes to bans on separatist parties, the case of Halkin Emek Partisi (HEP), the People’s Labour Party,⁴⁸ is instructive. The court found the Kurdish party, having promoted Kurdish political and cultural rights, to have threatened the unity of the nation-state and thus to be in breach of several constitutional provisions, including unamendable ones. The preamble of the Turkish Constitution declares ‘the eternal existence of the Turkish Motherland and Nation and the indivisible unity of the Sublime Turkish State’, while unamendable Article 3 declares the state, with its territory and nation, ‘an indivisible entity’ and the national language Turkish. Other constitutional provisions also mention territorial integrity, such as Article 14 on the prohibition of abuse of fundamental rights.

Roznai and I have argued elsewhere that declaring territorial integrity unamendable tends to occur in the face of internal contestation of the constitutional dispensation and of external threats to the state’s boundaries, illustrating our argument with the (sadly, since all too relevant) example of Ukraine.⁴⁹ But whereas courts will be powerless against the latter, they can and have exercised their interpretive powers to operationalise unamendable territorial integrity internally. In the words of one author, the Turkish court invoked constitutional text and the history of post-Ataturk Turkey to find that ethnic or language groups would be denied minority status on account of its incompatibility with national unity: ‘The state was unitary, the nation was a whole, and arguments to the contrary could only be seen as unwarranted foreign influences intensified by the rhetoric of human rights and freedoms.’⁵⁰ In the party ban literature, Turkey’s would be an example of the ‘legitimacy paradigm’ casting Kurdish parties as an ‘existential threat’ to the state for demanding cultural and territorial accommodation.

It should be noted that bans on Kurdish parties coexist with other measures that limit political representation in practice. A 10 per cent electoral threshold for gaining seats in Parliament was in place until 2022, having been introduced by generals after the 1980 coup in a bid to address political fragmentation. This unusually high threshold curtailed the political representation of not only the Kurdish community but wider Turkish society insofar as it precluded the parliamentary voice of numerous smaller (mainly leftist) parties. By one study, as many as a quarter of voters were disenfranchised as a result of the 10 per cent threshold.⁵¹ Its effects also extended to increasing the share of parliamentary seats allocated to the AKP, which enjoyed repeat absolute and even supermajorities that in turn allowed it to push through constitutional reform. For example, the AKP held over two-thirds of seats on only 34 per cent of the votes cast following the 2002 election, when some 46 per cent of votes had been redistributed. The threshold has only been reduced to 7 per cent in 2022 following a split in the Nationalist Movement Party (MHP), the AKP’s traditional coalition partner, which would have seen it remain outside parliament had the 10 per cent bar stayed in place.

Interestingly, two candidates in the 2002 election whose party did not enter parliament lodged an application with the European Court of Human Rights alleging that the threshold of 10 per cent imposed nationally for parliamentary elections interfered with the free expression of the opinion of the people in the choice of the legislature, relying on Article 3 of Protocol No. 1 to the European Convention on Human Rights. The ECtHR Grand Chamber disagreed and accepted the Turkish Government’s justification of the threshold as aimed at ‘avoiding excessive and debilitating parliamentary fragmentation and thus of strengthening governmental stability’.⁵²

An even more famous instance of a party ban in Turkey was the prohibition of the Refah Partisi, the Welfare Party.⁵³ The threat the Constitutional Court identified to the democratic system was said to be the party’s embrace of Shari’a law, contradicting the unamendable secularism enshrined in the Turkish Constitution. The court defined secularism as ‘a way of life that has destroyed the medieval scholastic dogmatism and has become the basis of the vision of democracy that develops with the enlightenment of science, nation, independence, national sovereignty, and the ideal of humanity’.⁵⁴ The court proceeded to defend this understanding of secularism as reinforcing the protection of religion itself, insofar as by separating it from politics, religion ‘is saved from politicization, saved from being a tool of administration and kept in its real respectable place which is the conscience of the people’.⁵⁵ The same logic was later invoked in the even more famous Headscarf decision.⁵⁶ There, the Turkish Constitutional Court invalidated an amendment meant to abolish the ban on headscarves in universities on grounds of equality and the right to education in the name of secularism, said to be an essential condition for democracy and ‘a guarantor of freedom of religion and of equality before the law’.⁵⁷

The court invoked the language of militant democracy and stayed silent on the political implications of banning what by then had become the most electorally significant party in the country, in power for two years. In fact, according to some observers, reducing electoral competition had been precisely the point, revealing the Constitutional Court’s own political bias in favour of secularist elites.⁵⁸ Even on the face of the judgment, we find its discussion of the notion of democracy it was defending to have been limited and unsystematic⁵⁹ and its assessment of democratic threats black and white. The court’s reasoning left no room for democracy’s inner tensions and only saw it as ‘a formal category, an abstract entity in need of protection’.⁶⁰

On this occasion again the European Court of Human Rights endorsed the Turkish Constitutional Court’s decision. In its own highly contested Refah Partisi case, the ECtHR accepted the militant democratic argument once more.⁶¹ It found Shari’a to be incompatible with the fundamental principles of democracy, legal pluralism meant to implement it to undermine individual rights, and the possibility of recourse to force to gain political power – read into the ambiguity of jihad – to justify forceful state action, including a party ban. The fact of Refah’s being in power was actually read as even more reason to intervene, insofar as the ECtHR saw it as making the party more likely to implement its agenda.

Turkey’s experience is illustrative of the ways in which eternity clauses can underpin party bans with at times far-reaching effects on electoral democracy. Democracy itself may be part of the Turkish Constitution’s unamendable core, but it is a particular understanding of it: certainly secular and, via unamendable territorial integrity and official language, also nationalist and majoritarian. The constitution works in tandem with other tools to restrict access to the electoral arena, such as electoral thresholds. Interestingly, the Constitutional Court has adopted an expansive reading of the reach of these unamendable provisions, applying them to party ban cases and not only unconstitutional constitutional amendment cases. Moreover, we see that appeals to supranational standards of human rights protection reinforced the Turkish Court’s reading of the constitution and militant defence of it. The ECtHR was concerned with showing due regard to Turkey’s history of political fragmentation – when the 10 per cent electoral threshold was challenged – and to the rigid understanding of secularism that justified banning even a governing party that had not taken steps to implement an Islamist agenda. In so doing, both the national and the supranational court narrowed the scope of what Turkish democracy could mean, ironically contributing to the erosion of multi-party democracy in the country over the long term and facilitating the political dominance of the AKP.⁶²

12.1.3 Indirect Party Bans

It is not always the case that parties are restricted from the electoral arena through an outright legal ban. Instead, they may be prevented from standing for elections or accessing public funding indirectly, such as through restrictions on ideological commitments. This is arguably the case in Israel, where parties that would seek to challenge the Jewish and democratic definition of the state are not permitted to stand for elections. From one perspective, this could be added to the examples of anti-democratic party bans discussed above. However, the particular Israeli situation warrants a separate examination: not only are we dealing with an incompletely codified constitutional system, where the constitutional basis for restricting political parties is thus less clear-cut, but the country’s ethno-religious definition and political division make it a unique case.

One may be sceptical from the outset as to whether the question of unamendability even arises in Israel. Given the Israeli system’s incomplete constitutionalisation via a series of Basic Laws, all arguably open to amendment by the Knesset, one might think unamendability foreign to Israeli legal thought or judicial practice. Moreover, the Israeli Supreme Court has recognised the Knesset as sitting not only as a legislative assembly but also as a constituent body.⁶³ This would seem to suggest its legislative powers limitless, including in the constitutional realm. However, already in the famous Bank Mizrahi judgment, the Supreme Court indicated that only another Basic Law could alter a previously enacted one and also that certain constitutional values would operate as limits on the Knesset’s constituent power.⁶⁴ Later case law clarifying that those limits embodied the Jewish and democratic nature of the state.⁶⁵

Thus, even in the absence of a formal eternity clause, it has been argued that Israel does exhibit a form of implied unamendability. Aharon Barak has claimed Israel to be an example of a narrower form of unamendability, one operating in the absence of a textual eternity clause but whose object was the Jewish and democratic nature of the state as laid out in the country’s Declaration of Independence.⁶⁶ Consequently, while a future Israeli constitution might expand the scope of unamendability to include things like judicial review or independence, until the process of constitutionalisation is completed, Barak has argued, only the state’s definition would amount to a substantive limit on Basic Laws.⁶⁷ More recently, Mazen Masri has argued that two forms of unamendability operate in the Israeli system: one concealed, through controlling the composition of the Knesset, and one unwritten and judicially created.⁶⁸ Like Barak, he views these as resulting in the entrenchment of the definition of the state. Additionally, however, Masri finds unamendability also to be a vehicle through which to embed a hierarchy among citizens and to reinforce favourable status for certain groups.

Understanding how this form of unamendability has impacted the electoral and parliamentary arenas in Israel requires a trip back in time. In the 1965 Yeredor case, the Supreme Court upheld the electoral disqualification of the Socialist List, a principally left-wing Arab list.⁶⁹ The ground invoked was that the party did not respect the ‘fact’ of Israel’s founding as an eternal Jewish state, fulfilling the right to self-determination of the Jewish people.⁷⁰ The party’s programme was seen as sharing premises with that of the previously banned Al-Ard pan-Arab movement. The state saw the movement as a threat to Zionism and as such to its own existence and the Central Elections Committee agreed, despite no formal statutory basis to block the party’s candidacy; the Supreme Court nevertheless endorsed the Committee’s decision.⁷¹ The Weimar experience and the concept of militant or defensive democracy, respectively, were invoked by the majority justices in their opinions. Neither the unlikely electoral success of the party nor its emphasis that it was contesting the Jewish but not the democratic nature of the state factored into the decision.⁷²

While the Supreme Court later adopted a narrow interpretation of this judgment, the case already at the time raised the question of whether a judicially created ‘supra-constitution’ had emerged.⁷³ Others see it as creating an implicit eternity clause whose effect is pre-emptive, by screening in advance ideas that can enter the Knesset.⁷⁴ The implications are especially significant given the Knesset’s double role as ordinary legislature and constituent assembly. It has thus been argued that in Israel, revolutionary amendments are neutralised before even entering the constituent arena, insofar as their very initiators are precluded from even attempting to enter its gates.⁷⁵

The Knesset adopted Amendment 7A to The Knesset Basic Law in 1985. It enshrined in law the self-defensive understanding of Israeli democracy, which ‘allows Israel to ban a list or candidate who supports armed struggle against the state of Israel, who negates the existence of the state of Israel as the state of the Jewish people, or incites to racism’. In 1988, the extreme right-wing anti-Arab party Kach was prevented from contesting elections on this new legal basis, as a racist party. Despite the passing of Amendment 7A, no party has been banned on the second ground since 1965.⁷⁶ The Israeli Supreme Court in later cases distinguished the Yeredor case as an extreme measure, whereas in other instances – such as the attempted ban of the Progressive List for Peace – it declined to find evidence of the impugned party seeking the dissolution of the state.⁷⁷ The Supreme Court attempted to ground the definition of the state in universalist, liberal values, with its former Chief Justice Barak equating Jewish values not with religious values but with Western democratic principles.⁷⁸ In 2002, the Knesset amended 7A and changed this second ground for party proscription to a ban on negating ‘the existence of the state of Israel as a Jewish and democratic state’, part of the growing shift towards entrenching a particular view of state identity in law.

Kedar, reconstructing the origins of the catchphrase ‘Jewish and democratic’ in Israeli legislation, refers to it as having been born ‘almost inadvertently’.⁷⁹ The language was introduced during negotiations for the 1992 Basic Law: Human Dignity and Liberty and Basic Law: Freedom of Occupation to placate fears that religious practices previously accommodated would now risk being found discriminatory.⁸⁰ Thus, it was a matter of political compromise, as part of ‘a fight over the division of political power between the state and the religious establishment’.⁸¹ The expression was seen as ambiguous enough to appease different sides in the debate and the courts remained reluctant to give it effect.⁸² The task of constitutionally defining state identity and building consensus could be once more relegated to another day.

However, developments since 2014 moved away from this universalist logic. With the passing of the Basic Law: Referendum that year, strict restrictions were placed on governmental action affecting the territory of Israel, which in turn limits possible routes to a peace agreement.⁸³ Then in 2018, the Basic Law: Israel as the Nation State of the Jewish People was adopted. In addition to listing state symbols, recognising Hebrew as the official language (with Arabic afforded a special status) and recognising Jewish settlement as a national value, the law controversially declared the Jewish nature of the state without making reference to its democratic character, or indeed to a principle of equality. Many saw the 2018 law as entrenching the state definition as well as the erosion of equality rights of both individuals and non-Jewish groups in Israel.⁸⁴ At the very least, it ‘create[d] the impression that the Jewish character of the state takes precedence over democracy’.⁸⁵ This impression was hard to escape given that the law’s supporters had expressed hope it ‘would end the Supreme Court’s increasing bias, since the 1990s, in favour of human rights and democracy and against the states’ Jewish identity’.⁸⁶

Fifteen petitions were lodged with the Supreme Court challenging the Basic Law as an unconstitutional constitutional amendment. In 2021, the Supreme Court rejected the petitions and found the legislation compatible with the other Basic Laws of Israel.⁸⁷ It emphasised the declaratory nature of the law and rejected its interpretation as discriminatory in light of other guarantees of individual rights in Israeli law. The only Arab judge on the Court was the sole dissenting voice. In his view, the law contradicted the state’s democratic nature and undermined equality by ignoring Arab and Druze citizens.

The 2021 decision was an attempt by the Supreme Court to square the circle: neither to outright reject the petitions nor to find in their favour by, for the first time, striking down a basic law as unconstitutional. The Court attempted to neutralise the potentially discriminatory nature of the law via interpretation by emphasising its declaratory nature. There are certainly those who believe it would have been ‘an extremely unfortunate move’ for the Court to strike down the law as an unconstitutional constitutional amendment before it was ever applied.⁸⁸ However, when viewed as one piece of a larger puzzle, the 2021 decision does little to assuage fears that the 2018 law further eroded the purposeful ambiguity of the ‘Jewish and democratic’ definition of the state. We see instead the trajectory being an ever more exclusionary understanding of the state, one that fuses the democratic and Jewish characteristics thus making it impossible for the latter to be challenged without the former also being presumed attacked. Moreover, reforms were introduced in early 2023 to curb the Supreme Court’s powers of judicial review, introduce government control over judicial appointments, and give the Knesset powers to override Supreme Court rulings. If adopted, the prospect of a politicised Court far less inclined to walk the interpretive tightrope discussed above becomes near certainty.

12.2 Unamendability and Parliamentary Politics: Parliamentary Mandates

I wish also briefly to discuss another type of judicial intervention in electoral politics facilitated by unamendability: instances in which courts intervene to protect the electoral or parliamentary arena in the name of an eternity clause. The case I will discuss here is by now famous in the literature and concerns a Czech Constitutional Court decision from 2009.⁸⁹ This was the decision setting out the material core doctrine of the court, grounded in Article 9(2) of the Czech Constitution, which reads: ‘Any changes in the essential requirements for a democratic state governed by the rule of law are impermissible.’

But first, some background to the 2009 crisis resulting in the case. In March 2009, after four failed attempts, the parliamentary opposition succeeded in passing a no confidence vote against the Government. An early dissolution of the Assembly of Deputies and early elections was seen as desirable. The constitutional procedure to follow would have involved first proving that the legislative body was unable to function effectively, which in turn involved one of three scenarios: either three failed consecutive attempts at confidence votes in a new Government, or a failed vote on a government bill on which the Government had attached the issue of confidence, or, finally, when the Assembly had been adjourned for a longer period than permitted by the constitution (Article 35(1)). Even under these scenarios, the procedure is not automatic but merely empowers the President to act. The cumbersome procedure was seen as too time-consuming by all political sides, with cross-party consensus emerging that early elections were preferable.⁹⁰ As a consequence, an ad hoc constitutional amendment, Constitutional Act No. 195/2009/Coll. was adopted (with a 172:9 vote in the Assembly and 56:8 vote in the Senate) to procedurally pave the way for early parliamentary elections in October 2009.⁹¹ This would have shortened the mandate of the existing Parliament, given that normally elections would have been held in May 2010.

This was not, in fact, the first time such an ad hoc path was chosen to deal with a political crisis. In 1997/1998, a similar political compromise emerged in the aftermath of the breakdown of the ruling coalition Government. A constitutional amendment was passed then similarly to enable the running of early elections and resulting in the shortening of that parliamentary term by two years. In that instance, however, the amendment was not challenged before the Constitutional Court and the early elections proceeded as planned.⁹²

In 2009, however, one of the MPs standing to lose his mandate challenged the amendment before the Court, arguing that it violated his right to participate in the administration of public affairs and that any exception to his carrying his mandate to the full four-year term needed to be prescribed by the Constitution at the time of his election. He also challenged the nature of the Act in question, arguing it was not a real constitutional amendment because it violated the material core of the Czech Constitution. Specifically, he claimed it breached principles of non-retroactivity, generality, and predictability of laws, which come under the umbrella of respect for the rule of law.

The Constitutional Court agreed and voided the Act. Its decision involved several important steps. First, the Court had to establish its power to review constitutional acts, whereas the Constitution and Constitutional Court Act only stipulated its power to review the constitutionality of ‘laws’. However, relying on its constitutional role as guardian of the constitution (Article 83), the Czech Constitutional Court proceeded with its review. Second, it challenged the nature of the 2009 Act, calling it constitutional only in form and not in substance. Given that it referred to a specific rather than general situation, the Act was closer to an administrative act and in breach of principles of equality, non-arbitrariness, and right to an independent judge, in addition to the principle of separation of powers.⁹³ Third, the Court established a link to the Constitution’s ‘material core’ as enshrined by Article 9(2) by accepting the claimant’s rule of law arguments (while ignoring his rights-based claims). Additionally, the Court emphasised the irregular parliamentary procedure followed for adopting the amendment as itself evidence of the breach of the ‘material core’. It sought to ground its decision in both precedent and appeals to history. It thus cited case law having recognised ‘popular sovereignty, a right of resistance, and the basic principles of election law’ as ‘fundamental inviolable values of a democratic society’ and as such part of the ‘material core’ of the constitutional order.⁹⁴ The Court also invoked the Weimar experience together with Czech experience with communist semblance of legality to justify its intervention.⁹⁵

It should be noted that, hitherto, the enforceability and practical implications of the Czech eternity clause had been disputed. Some had seen it as purely declaratory or else directed to the Senate as the chamber responsible for revising legislation passed by the Assembly.⁹⁶ In its decision, however, the Court removed any doubt about the teeth of the eternity clause, declaring it ‘non-changeable … not a mere slogan or proclamation, but a constitutional provision with normative consequences’.⁹⁷ In stepping in to enforce it, the Court saw itself as guarding not just the rule of law but also the whole democratic order and the integrity of the Czech constitutional system.⁹⁸ It would go on to build its constitutional identity doctrine in later case law, all the while resisting calls to provide an exhaustive list of the elements constituting this constitutional ‘material core’.⁹⁹

The literature on unamendability has long debated such judicial self-empowerment when it comes to enforcing eternity clauses. What I wish to focus on here is rather the necessity and implications of the Czech Court’s intervention in the concrete case at hand. The proportionality of the Court’s intervention is dubious.¹⁰⁰ Clearly, its invocation of a Weimar-like threat signals the Court saw a real danger to parliamentary politics in the country. However, when looking at the political context surrounding the passing of the 2009 Act, it is difficult to conclude that Czech democracy had really been endangered to the point implied by the Court. The bicameral political consensus underpinning the adoption of the Act, as well as its support from both the prime minister and president, are evidence of wide agreement – among the same MPs that would stand to have their mandates shortened – that early elections were desirable. One could also argue that parliament itself choosing to cut short its term is far less likely to amount to an abuse of process than were it to have done the opposite and extend its mandate or were the curtailment to have occurred at the hands of the executive alone. Additionally, while the 1997/1998 precedent may not have completely excluded the possibility of unconstitutionality, it certainly undermined the existential threat rhetoric employed in 2009. Finally, ignoring the individual rights claims in the case also seems a weakness of the judgment.

The practical consequences of the 2009 decision were manifold. A new constitutional act was adopted in September 2009 that creates a route to early elections involving the self-dissolution of the Assembly by a three-fifths vote (Article 35(2)), thus rendering the amendment in general terms. However, to avoid another constitutional challenge, the new procedure was not relied upon in 2009 and the existing Parliament carried out its full term. It has been argued that the delay ‘hanged the Czech political landscape and probably also the victor of the elections’, seriously denting the vote share of Social Democrats – previously frontrunners in the polls – and allowing the rise of new ‘pro-business parties’ such as Public Affairs in the 2010 elections.¹⁰¹ The wide eternity clause was thus the hook on which the Czech Constitutional Court anchored its ‘material core’ doctrine. In 2009, the Court deployed the doctrine ostensibly in the name of protecting parliamentary politics against a Weimar-like threat. It did so, however, at a high political cost and against the wishes of all political actors.

12.3 Conclusion

This chapter has aimed to examine, through a selective range of case studies, the complex interplay between eternity clauses and electoral democracy. It has sought not to take at face value unamendability’s claim to be democracy enhancing, even where the unamendable provisions in question embody a militant ethos and explicitly aim to protect multi-party democracy. Party bans and election invalidations have the potential seriously to affect political competition and parliamentary democracy, so court intervention resulting in such measures deserves very careful scrutiny.

Germany’s example illustrates a seemingly clear-cut commitment to militant democracy that combines constitutional tools including an eternity clause and party bans. The early years of German post-war democracy saw it ban both Nazi and communist parties seen to seek to destroy the democratic constitutional order. As German democracy consolidated, the need for such drastic measures might be said to have decreased, in recognition of which the Constitutional Court stopped short of banning the Neo-Nazi NPD in 2017. However, the Court’s attempt to delimit the constitutional grounds for proscribing parties from the normative content of the eternity clause may have introduced uncertainty both about the standard for banning anti-constitutional parties and about the constitutional core of the Basic Law. As the rise of the AfD has shown, moreover, the threat of extremist parties – also well-versed in avoiding falling foul of constitutional rules¹⁰² – may yet test the German constitutional system’s militant democratic commitments.

In Turkey’s case, we have witnessed numerous party bans on grounds rooted in the constitutional eternity clause, with far-reaching implications for electoral politics. On the one hand, bans on Kurdish parties have operated in tandem with other rules, not least the long-standing 10 per cent parliamentary threshold, to preclude their ability to enter the electoral arena. On the other, the ban on (sometimes salient) religious parties would not appear to have weakened them, given that Erdogan’s AKP has been in power since 2003. However, a different interpretation is possible. As Rosenblum has argued, the separation of state and religion in Turkey, to which we can add the constitutional arrangements entrenching secularism, was ‘uniquely one-directional: government was protected from religion but not vice versa’.¹⁰³ As a consequence, no contestation of the balance between state and religion was possible, which in turn could be viewed as having provoked the politicisation of religion.¹⁰⁴

Israel’s trajectory is less typical for studies of unamendability, insofar as its constitution is fragmentary and its unamendable core must be pieced together from different legislative and judicial sources. Nevertheless, the definition of the state as Jewish and democratic is clearly part of this core. For decades, its ambiguity served to stave off conflict over political identity, not just between the Jewish majority and non-Jewish minorities but also with religious Jewish groups. The Supreme Court’s universalist, human rights–based approach during that period mitigated the exclusionary potential of this definition. In party ban cases, this meant developing a more restrained approach that upheld bans against racist parties but not those accused of denying the existence of the state as Jewish and democratic. The increasing polarisation in Israeli politics, however, and the entrenchment of this state definition in legislation culminating in the 2018 Nation-State Law have shifted the terms of the debate. It has made it much more difficult to defend this entrenched political identity as anything other than exclusionary, especially in the absence of a similarly entrenched equality guarantee.

Czechia’s experience reveals another side to the story of unamendability’s potential impact on electoral politics. In a context of serious political crisis but also rare political consensus, a political solution was found to pave the way for early parliamentary elections. The constitutional amendment it was enshrined in, however, was invalidated by the Czech Constitutional Court on the grounds that it violated the Constitution’s ‘material core’. A close reading of the decision reveals more concern with building the legitimacy of the Court’s unconstitutional constitutional amendment doctrine than sensitivity to the political context within which the invalidation would produce effects. There are serious reasons to believe the Court’s assessment of the Weimar-like threat to Czech parliamentary democracy was overblown. Moreover, while Czech political actors respected the judgment, its impact on the electoral balance of power was significant.

The examples above show that courts will not always strike the right balance between protecting and unduly narrowing democratic commitments. In some cases, they may even unintentionally undermine multi-partyism itself or significantly influence electoral outcomes. With its bluntness, unamendability may hinder rather than help bring nuance to these difficult decisions.

13 Monarchy and Democracy in Modern Malaysia

Yvonne Tew

13.1 Introduction

No country in the world has as many monarchs as Malaysia. Among the monarchies in the world today,¹ Malaysia is unique in its system of a rotating, elected monarchy. Within the Malaysian Federation, nine Malay Rulers are sovereign as the constitutional head of their respective states. Every five years, the King who serves as Supreme Head of the Federation – known by the title of the Yang di-pertuan Agong – is elected in a rotating system from among these nine Rulers. The Federal Constitution of Malaysia set up a Westminster-style parliamentary system, with a constitutional monarch as the head of state. The written document explicitly sets out the monarch’s powers, which limits the powers of the King in most areas of governance. As a matter of design, the constitutional text created at independence in 1957 contemplated the monarch as a constitutional figurehead with a largely symbolic role.

Not so in practice. In recent times, the contemporary monarch in Malaysia has emerged as a critical actor in the formation and functioning of the federal government. The monarch assumed a key role during a period of unprecedented democratic transition and disintegration in modern Malaysia.

In 2018, Malaysia experienced a historic national election that resulted in the ousting of the Barisan Nasional ruling coalition that had governed the country for more than sixty years since independence.² Malaysia’s political transition was hailed as a democratic breakthrough, ending decades of dominance by a single political coalition.³ Twenty-two months after being elected, however, the newly elected Pakatan Harapan government disintegrated following a series of political defections. A divisive battle for government leadership ensued among former prime minister Mahathir Mohamad, his supposed successor Anwar Ibrahim, and senior politician Muhyiddin Yassin. The King intervened directly into this government crisis by deciding to appoint Muhyiddin Yassin as the new prime minister, cementing a government turnover to a coalition that returned to power many from the previously deposed ruling political party.

In the months that followed, the monarch played a central role in the country’s governance. As the Perikatan Nasional government under Muhyiddin Yassin struggled to deal with the COVID-19 pandemic while maintaining its fragile grasp on power, the Agong refused the new premier’s bid to declare emergency rule. Later, in July 2021, the Agong publicly rebuked Muhyiddin’s administration for announcing that the emergency proclamations had been revoked without the monarch’s consent.

This chapter explores the central part the monarch played in facilitating the political transition and in the constitutional governance of Malaysia’s contemporary political order. Doing so raises broader questions about the role of the monarchy in a democracy and its institutional capacity to safeguard against incumbent capture or to accelerate democratic erosion. It begins in Section 13.2 by setting out the historical context for Malay kingship and its eventual transformation into a constitutional monarchy after Malaya’s independence in 1957 from British colonial rule. It examines the constitutional design of Malaysia’s federal parliamentary system and the design choices made regarding the ways in which the Federal Constitution structures the relationship between the monarchy and the other branches of government.

Section 13.3 tells the story of the monarch’s rise during Malaysia’s first change of government following a general election and the subsequent collapse of two successive governments. The political crises that occurred amidst these government transitions and during the COVID-19 pandemic have left the country’s political parties and electoral institutions highly fragile. Meanwhile, following the royal interventions between 2018 and 2021, the monarchy has emerged with an enhanced position in Malaysia’s contemporary constitutional order. Far from being an outdated or anachronistic institution, the story that emerges is that monarchy in Malaysia is not just surviving, it is thriving – a story that has broader resonance in other parts of Southeast Asia. Section 13.4 offers some concluding reflections on the role that the monarch can play in protecting or undermining constitutional democracy.

13.2 Of Sultans and Kings: History and Constitutional Design

13.2.1 Malay(sian) Kingship: Historical Background

Malaysia does not have one monarchy; it has nine. The origins of the nine Malay Rulers have their roots in the sultanate that presided over the famed port city of Malacca. The city rose to glory in the fifteenth century as one of the region’s preeminent trading centers, situated at the crossroads of the spice route between the East and West. Following the fall of the Malacca empire in 1511 to the Portuguese, sons of the Malacca sultan established empires in Perak and Johor, and other new sultanates modeled on Malacca’s emerged across the Malay peninsula.⁴ Malay ideas of kingship drew on influences from Islamic traditions infused with local customs and Hindu and Buddhist elements.⁵

The Portuguese colonial powers were followed by the Dutch, and then the British. Between 1874 and 1930, the British established a series of treaties with the sultans of the nine Malay states. This system of indirect rule required the state Rulers to act on the advice of a British Resident, except in matters relating to Malay religion and custom. After the Second World War, during which the Japanese occupied the Malayan states for a few years, the British sought to unify the nine Malay states, along with Penang and Malacca, into a unitary entity – the Malayan Union – in 1946. Vehement opposition to the Malayan Union consolidated into a rallying point for Malay political leaders to form the United Malays National Organisation (UMNO), the political party that would dominate Malaysian politics for the next seven decades. The political rebellion against the proposed Malayan Union was due to many reasons, but a main objection was that the sovereignty of the Malay monarchs over their individual states would be abolished by the creation of a unitary state.

The British relented. The Federation of Malaya was created in 1948, replacing the proposed Malayan Union structure with a federal system of government that preserved the powers of the Malay sultans as the Rulers of their respective states. This federal arrangement would form the basis for the design of the constitutional system put in place at the Federation’s independence.

In 1956, a constitutional commission consisting of five Commonwealth legal experts, chaired by Britain’s Lord Reid, began the constitution drafting process based on terms of reference that had been agreed on by representatives of the Malayan government, the Malay Rulers, and the British government.⁶ The draft constitutional text was modified in several aspects after scrutiny by a local working party and eventually ratified by the federal legislative council. The Constitution came into force on August 31, 1957, when the Federation of Malaya became fully independent. In 1963, when Sabah, Sarawak, and Singapore joined the Federation, it became the Federal Constitution of Malaysia.⁷

13.2.2 Constitutional Design and Constitutional Politics

The Malaysian Constitution provides for a constitutional monarchy, as well as a Westminster-style parliamentary system with a bicameral legislature, executive, and independent judiciary. It proclaims the Supreme Head of the Federation as the King, officially called the Yang di-Pertuan Agong⁸ – quite literally, “One who is Made Supreme Lord.” Federalism remains at the core of the design of the political system and the constitutional monarchy. The Constitution cemented the position of the Malay Rulers as the constitutional heads of their respective states. It guarantees the sovereignty of the Malay Rulers within their territories,⁹ and the rights and privileges of each Ruler as the head of their states and as the head of the religion of Islam in those states.¹⁰

Malaysia has a unique system of rotating, elected constitutional monarchy.¹¹ The position of the Yang di-Pertuan Agong (Agong) rotates every five years among the sultans of the nine states traditionally headed by Malay Rulers. Electing the federal constitutional monarch is a matter for the Conference of Rulers, a body comprising the nine Malay Rulers and the governors of the other four states.¹² That task is carried out typically once every five years or when there is a vacancy, which may occur due to death or, more uncommonly, resignation.¹³ In addition to the power to elect – and remove – the King, the consent of the Conference of Rulers is required to amend certain provisions of the Federal Constitution, such as those dealing with the status of the Rulers, the Conference of Rulers, the Malay national language, and the special position of the Malays and the natives of Sabah and Sarawak.¹⁴

The role of Malaysia’s Agong was loosely modeled on that of a constitutional monarch in a Westminster-style political system. Unlike Britain’s uncodified constitution, however, the powers of the Malaysian Agong are expressly set out in the Federal Constitution. The Malaysian Constitution vests executive authority in the Agong,¹⁵ but specifies that the King “shall act in accordance with the advice of the Cabinet or of a Minister acting under the general authority of the Cabinet, except as otherwise provided by this Constitution.”¹⁶ According to the Constitution, the Agong also possesses legislative authority as a constituent part of Parliament, which consists of the Agong and the two Houses of Parliament.¹⁷ The King is constitutionally empowered to summon, prorogue, and dissolve (or decline to dissolve) Parliament.¹⁸

The monarchy in Malaysia has always been bound up with constitutional politics. Post-independence, the Rulers clashed with the governments ruling their states in several episodes in the 1970s and 1980s; ultimately, tensions and conflicts between the Rulers and the federal government resulted in a constitutional confrontation.¹⁹ During the first Mahathir Mohamad administration, which spanned the 1980s and 1990s, the federal government pushed to amend the Federal Constitution to constrain the powers of the Rulers. Constitutional amendments passed in 1984 and 1994 limited the King’s power to refuse assent to laws passed by Parliament, removed the royal immunity of the Rulers from suit, and required the Rulers to act on executive advice.²⁰

Nonetheless, the Constitution specifies several areas in which the King “may act in his discretion,” including: “(a) the appointment of a Prime Minister; (b) the withholding of consent to a request for dissolution of Parliament.”²¹ Those who view the sovereign as occupying a largely symbolic position in a Westminster-style parliamentary system consider the monarch’s function to be primarily formal and ceremonial, with the head of state having no real discretion even in these circumstances.

Yet, in recent times, Malaysia’s monarch has emerged to play a critical role in the formation and functioning of government in the country. As we will see, the Agong’s appointment of the prime minister and the monarch’s consent (or lack thereof) to emergency rule during the COVID-19 pandemic would turn out to be crucial turning points during Malaysia’s 2020 government transition and in subsequent political crises.

13.3 Constitutional Monarchy in Contemporary Practice

13.3.1 Formation of Government

The future of democracy in Malaysia was being decided in February 2020, so it appeared, inside a room in the palace. At half past two o’clock on a Wednesday afternoon, on February 25, the first wave of parliamentary members began arriving at the palace gates in Kuala Lumpur to be interviewed by the King. Over the course of that day and the next, they would meet one by one with the Agong. The palace announced that “his majesty himself” would interview each of the 222 members of the lower house of Parliament to determine which candidate the parliamentarians supported as prime minister.²²

Days before, the Pakatan Harapan government had collapsed, after less than two years in power following its democratic triumph in the 2018 national elections. Initially triggered by several key members defecting from the governing Pakatan Harapan alliance to join forces with rival political blocs, a confounding political drama played out over the last week of February. Mahathir Mohamad resigned as prime minister, the Agong named him as interim prime minister, and then Mahathir sought to resume the mantle of prime minister. A battle for the country’s premiership ensued between Mahathir Mohamad, the supposed premier-in-waiting Anwar Ibrahim, and Muhyiddin Yassin, a senior politician from Mahathir’s own political party. Days of confusion followed as parliamentarians switched support for the contenders vying for the prime ministership in a series of bewildering political twists and turns.²³

The King intervened at the heart of the political turmoil, seeking to exercise his constitutional power to appoint a prime minister. Never in the history of the country had the monarch’s constitutional task seemed so fraught. What followed was a high-profile two-day interview process at the palace, during which the King met with all the members of Parliament to determine who commanded the support of the majority of the house. At the end of the two days, though, the King professed himself unable to reach a resolution, declaring that he was not confident that a single parliamentarian had the majority support to form a new government.²⁴ The Agong then asked for various party leaders to nominate their candidates for prime minister, and met with the party leaders as well as with the Conference of Rulers.

Soon after, the King announced that he had determined, based on the representations of the party leaders, that Muhyiddin Yassin had the support of the majority and that he would appoint Muhyiddin to the premiership. Even as the palace made its announcement, Mahathir Mohamad claimed that he had majority support, releasing a list of 114 parliamentarians who had allegedly signed statutory declarations in his support.²⁵ Nevertheless, the King refused to grant him an audience. As things turned out, Mahathir never got to be in the room where it happened.

On March 1, 2020, Muhyiddin Yassin was sworn in as the new prime minister. He took power at the helm of a hurriedly assembled Perikatan Nasional governing coalition containing many politicians from the United Malays National Organisation, which had been the key party bloc in the previously deposed Barisan Nasional government.

Malaysia’s political crisis in 2020 resulted in an unusual government transition in a country that had hitherto experienced six decades of one-party rule and one (short-lived) democratic changeover when voters ousted the ruling Barisan Nasional coalition from governance for the first time.²⁶ Also unprecedented was the King’s role in deciding who to appoint as prime minister. The events brought to the fore constitutional issues and constitutional conventions that had never been tested at the level of the federal government.

Begin with the constitutional text. The King’s role in appointing a prime minister is laid out in Malaysia’s Federal Constitution; Article 43(1)(a) states that “the Yang di-Pertuan Agong shall … appoint as Prime Minister to preside over the Cabinet a member of the House of Representatives who in his judgment is likely to command the confidence of the majority of the members of that House.” Up to this point in Malaysia’s political history, which had been dominated for decades by a ruling coalition that controlled a vast majority of legislative seats, the monarch’s appointment of the prime minister in Malaysia had been merely a pro forma task.

Not so in 2020. The government upheaval of February 2020, with political shenanigans resulting in razor thin, constantly shifting margins of support for each of the vying prime minister candidates, raised questions that had never come to the fore at the federal level. How was the King to exercise his “judgment” as to who was “likely to command the confidence of the majority” in appointing the new prime minister? And what evidence could the constitutional monarch rely on in making that determination?

The Malaysian monarch decided to directly intervene into the midst of the political crisis. The Agong took the extraordinary step of interviewing each individual member of Parliament and then met with the leaders of the competing political parties to hear their representations – a move that the palace itself characterized as going “beyond the call of obligation.”²⁷

Malaysia’s case offers a striking example of a non-majoritarian institution playing an assertive role in influencing the government’s formation. Textually, of course, the monarch is constitutionally empowered to appoint the person who “in his judgment” is likely to command majority support in Parliament.²⁸ But as Asanga Welikala observes, “the king’s actions do raise a question that often arises in Westminster-style systems: a course of action that may be strictly legal may nevertheless be regarded as unconstitutional.”²⁹

Malaysia’s unusual government transition in 2020 implicates fundamental questions of constitutional design. Under the constitutional framework of a Westminster-style parliamentary democracy, the monarch’s role in appointing a prime minister is normally to affirm the outcome of the political process.³⁰

Of course, times of political crisis are not the normal circumstances under which a monarch appoints a prime minister.³¹ In extraordinary circumstances, the head of state may be pressed to draw on various sources of evidence to try to measure majority support.³²

In a situation of highly fraught political uncertainty, though, when there is serious dispute as to which candidate in fact commands majority support, there are strong arguments of democratic accountability and constitutional structure for allowing the political process to play out, rather than being resolved through premature royal intervention. When the palace announced the King’s decision, effectively sealing the premiership, Muhyiddin and Mahathir both claimed to command majority support, and Mahathir had produced a list of members of Parliament declaring their support for him.

Had the monarch not decisively named a prime minister, the contenders might have continued to battle it out in the wider political sphere through negotiations, compromise, or political horse-trading.³³ Or the contentious matter could have been resolved in Parliament through a vote of confidence (or no confidence) to determine majority support openly on the floor of the legislature.

Amidst the political crisis, the King was thrust with the circumstances to play kingmaker.

Still, a constitutional monarch intervening too precipitously into a political controversy may end up being perceived as overturning the country’s democratic mandate.³⁴ Royal assertiveness in the formation of government of the sort that resulted in the Malaysian government transition in early 2020 risks undermining, rather than promoting, democracy. The events that unfolded thereafter appeared to underscore this point. When Muhyiddin was appointed prime minister, Mahathir and the opposition parties immediately called for a vote of confidence to be held in Parliament – a call that was repeated throughout Muhyiddin’s premiership.

Yet, as the next section describes, that vote of confidence never happened.

13.3.2 Suspension of Government

On May 18, 2020, after a prolonged period of suspension, members of Malaysia’s Parliament convened for an unusual one-day parliamentary sitting.³⁵ Barely an hour after the parliamentarians had assembled in the lower house chamber, the meeting was over and Parliament adjourned yet again.

That one-day sitting was the first time that the Malaysian Parliament had convened since the Perikatan Nasional government, with Muhyiddin Yassin as prime minister, had come into power on March 1, 2020. The Agong opened the parliamentary sitting with a half-hour address. Although a vote of no confidence had been put forward against the newly appointed prime minister, the Speaker of the House of Representatives announced before the session that the motion had been dropped from the agenda.³⁶

Citing the COVID-19 pandemic, the government declared that no motions would be allowed during the parliamentary session and that the King’s speech would be the only order of business that day.³⁷ No debates or questions were permitted during the parliamentary sitting. It was a pro forma session held to satisfy – in form only – the constitutional requirement that no more than six months should elapse between parliamentary sittings.³⁸

In the months to follow, Prime Minister Muhyiddin Yassin proclaimed a state of emergency due to the coronavirus pandemic and suspended Parliament. For more than half of 2021, Parliament remained shut down. No motion of confidence in the government was permitted or voted on throughout Muhyiddin’s tenure.

To be sure, the Malaysian government’s move to bypass usual democratic procedures in the name of the COVID-19 pandemic was not exceptional when viewed through a global lens. Throughout the pandemic, national executives across the world exercised expansive powers in response to the health crisis,³⁹ generating debate over whether the pandemic has accelerated the rise of authoritarianism globally.⁴⁰ Much of the discourse on the role of government power during the COVID-19 pandemic focused on the accrual of power by elected executives worldwide.⁴¹

The Malaysian experience during the pandemic offers an example of the expansion of power by another executive branch actor: the monarch. As the Muhyiddin administration struggled to respond to the coronavirus pandemic, the King launched a series of interventions.

Consider the use of emergency powers. Prime Minister Muhyiddin Yassin first sought to invoke a national emergency in late October 2020. This push for emergency rule, which came several months after the one-day parliamentary session in May, after which time Parliament had been suspended, was justified by Muhyiddin as necessary to curb the spread of COVID-19 in the country. For many, the proposed emergency was simply another attempt by the premier to avoid a showdown in Parliament that might challenge his precarious position.

Stunningly, the Agong rejected the prime minister’s request to declare a state of emergency. This display of royal authority was unprecedented: never before had the constitutional monarch rejected a request made by the prime minister, tendered on the advice of the government – on a matter of national emergency, no less.⁴² Nor did the Agong hesitate to make clear his view on the matter, announcing that he saw no need for an emergency declaration and sternly calling on the politicians to end their politicking.⁴³

The Agong’s move was lauded by the public. At a time when citizens coping with pandemic restrictions were becoming increasingly disillusioned with the infighting among the political elite, the monarch’s refusal of Prime Minister Muhyiddin’s request boosted popular support for the Agong, who was widely praised for defending democratic values.⁴⁴

Earlier that month, the King had also refused to recognize opposition leader Anwar Ibrahim’s claim that he commanded a parliamentary majority such that he could unseat the incumbent Muhyiddin Yassin. Unconvinced by the evidence produced – Anwar had supplied the King with the number of legislators that he claimed supported him, but not their identities – the King had dismissed Anwar’s bid for the premiership.

Eventually, however, the Agong acceded to Prime Minister Muhyiddin Yassin’s request for a national state of emergency. On January 12, 2021, the King declared a state of emergency, which was set to expire on August 1. With the emergency decree, Muhyiddin Yassin’s administration gained broad powers, claimed as necessary to tackle the COVID-19 spread. The emergency also allowed for suspending Parliament and the holding of any elections. It was the first time in more than half a century – since the country’s devastating racial riots in 1969 – that a national emergency had been proclaimed.⁴⁵

In January 2021, Parliament was promptly suspended, and the legislature did not sit for more than half of the year. Opposition politicians denounced the declared emergency as a blatant attempt by the government to cling to power. Throughout the months-long state of emergency, there were repeated calls for Parliament to reconvene, including from the Agong, who, on at least three separate occasions, publicly reiterated his call for Parliament to resume proceedings as soon as possible.⁴⁶

Political tensions simmered and eventually reached a tipping point in July 2021. Faced with mounting criticism over the lengthy emergency rule and the government’s handling of the pandemic, the government eventually held a special session of Parliament at the end of July. During that session on July 26, Prime Minister Muhyiddin Yassin’s administration declared that all emergency ordinances had already been revoked – a sudden announcement that came as a surprise to the legislators, since the issue had not been debated in Parliament.⁴⁷

Almost immediately, the King publicly announced that he had not given his assent to revoke any emergency ordinances, asserting that he had agreed only to the proposal being presented to Parliament. In a rare rebuke to the administration, the Agong expressed his “great disappointment” with the announcement that the government had revoked the emergency ordinances, emphasizing that the statement was “inaccurate and had misled members of the house.”⁴⁸ The King excoriated Muhyiddin’s Cabinet, stating that the misleading statements had not only failed to uphold the rule of law but also ignored the Agong’s functions and powers as the head of state, which included the power to promulgate and revoke emergency ordinances.⁴⁹

The palace’s statement made no bones about the King’s conception of the role of the constitutional monarch:

His Majesty is aware of the necessity for His Majesty to act in accordance with the advice of the Cabinet as stated in Article 40(1) of the Federal Constitution. Nevertheless, the King emphasizes that, as the Head of State, His Majesty has the responsibility to advise and reprimand in the event of any unconstitutional action taken by any party.⁵⁰

In the aftermath of the palace’s damning rebuke, which fueled public outrage over his government’s handling of the emergency, Prime Minister Muhyiddin Yassin resigned on August 16. He conceded that he no longer had majority support in Parliament, ending his tenure after a turbulent seventeen months in power.

When a prime minister resigns, the Federal Constitution of Malaysia provides for the dissolution of Parliament, followed by elections, or for the King to appoint a new prime minister. The Agong ruled out holding a national election because of the COVID-19 pandemic. Instead, the monarch consulted with political party leaders, and the palace issued a call for members of Parliament to indicate their choice of a prime minister – through email, fax, or WhatsApp – to the palace.⁵¹

Yet again, the King was positioned as kingmaker. Soon after, the King announced that he was satisfied that that Ismail Sabri Yaakob – a leader from the United Malays National Organisation, which had been deposed in the 2018 national elections – had the backing of a (bare) majority of 114 of the 220 parliamentarians.

Five days after Muhyiddin Yassin’s resignation, Ismail Sabri was sworn in as Malaysia’s new prime minister on August 21, 2021, taking over as leader of the Perikatan Nasional alliance previously headed by Muhyiddin Yassin. It was the second time in less than two years that the leader of the government had assumed the premiership through royal appointment, instead of through an electoral outcome or a parliamentary vote of confidence.⁵²

13.4 Concluding Reflections: Constitutional Monarchy and Constitutional Democracy

On August 5, 2017, in a speech at a convention held in the administrative capital city of Putrajaya, Sultan Nazrin Shah, the state Ruler of Perak, put forward his account of the constitutional monarch in striking terms:

The King is not a rigid decorative ornament – without life – without soul …. It is a mistake to think that the role of a constitutional monarch is the same as that of a President, limited to what is written in the Constitution. The role of the Ruler goes above what is contained in the provisions of the Constitution.⁵³

That statement would turn out to be predictive. In the four years after the speech was delivered, the monarch would emerge as a critical actor in the creation and governance of the federal government. When the Pakatan Harapan government collapsed in early 2020, the King intervened into the political dispute by appointing a new prime minister. And, as the new administration sought to maintain its tenuous hold on power while struggling to manage the coronavirus pandemic, the Agong came to play a crucial role in declaring the start, and the end, of emergency rule in the country. These royal interventions at the federal level throughout 2018 to 2021 placed the monarch at the very heart of the nation’s constitutional politics.

Does an enhanced role for the monarch serve to protect or undermine constitutional democracy? Answers vary, unsurprisingly, depending on who you ask, and when that question is asked. For some, the constitutional monarch played a crucial part in “stabilizing” political crises and as a “check-and balance to the government in power.”⁵⁴ Others have decried the royal assertiveness as illegitimate intrusions into the democratic process – indeed, some have called the monarch’s actions a “royal coup”⁵⁵ – and voiced fears about “overreach by future monarchs.”⁵⁶ What seems undeniable is that the monarch’s position and powers have been augmented to an extent not seen since before Malaysia’s independence in 1957.

How and why has the monarch come to assume such an empowered position in contemporary constitutional governance?⁵⁷ Those viewing the King’s enlarged powers as “substantially departing from Westminster-modelled constraints” raise concerns over whether recent interventions have left “a lasting imprint on an ‘Eastminster Constitution’ with an enhanced monarchy.”⁵⁸ As Andrew Harding asks: “Do we live in postmodern world of constitutionalism in which even apparently defunct or declining institutions can take on new life, taking their place alongside both the familiar and the innovative?”⁵⁹

The revival of constitutional monarchy in contemporary Malaysia underscores the ways in which constitutionalism and practices of constitutional governance exist beyond the written text. In terms of constitutional design, the Malaysian experience highlights how the drafting of the written constitution and the structure it sought to put in place may produce effects unintended by the framers. When Malaysia’s Federal Constitution was created at the Federation’s independence from the British, the nation’s founding fathers envisaged a Westminster-model constitutional monarchy, with a symbolic role for the Rulers. It was clear, Kobkua Swannathat-Pian notes, that they “were not in favor of a return to a time where the Rulers had socio-political powers”; rather, it was their “ardent desire that the Rulers’ role would be limited to simply that of the symbolic kind.”⁶⁰ The framing of the constitutional text reflected these aspirations, with limited discretionary powers afforded to the King. Unlike Britain, Malaysia’s constitutional system accorded supremacy to a written text.⁶¹

Yet, there can be significant play in the joints when it comes to constitutional text and constitutional practice. Even in countries that accord primacy to the written constitutional document, like the United States, constitutionalism is recognized also to encompass the actual institutional practices of government.⁶² For emerging democracies in Southeast Asia – of which Malaysia is a case in point – institutional practices and constitutional conventions are far from being settled. In a setting in which constitutional norms are still in flux, institutional actors – like the monarch – have greater space to negotiate their role in constitutional governance. That’s especially so in times of political turmoil, during which the lack of established constitutional rules and political norms often create a lacuna that allows, or even compels, a monarch to step in.

In times of political or constitutional crisis, a monarch may take on a key role to resolve the conflict.⁶³ Tom Ginsburg describes the monarch’s function in this kind of situation as “crisis insurance.”⁶⁴ Examples of sovereigns who have acted to provide a “focal point” during times of crisis include King Juan Carlos I denouncing the 1981 Spanish coup d’etat attempt in a televised address, or Thai King Bhumibol summoning the prime minister general and the leader of the anti-government protest movement to the palace in 1992 for a royal audience that was broadcast on television.⁶⁵ This account captures part of the story, although Malaysia’s constitutional monarch has taken on a role that arguably goes beyond serving as a symbol of unity. The Agong emerged as the critical actor that directly facilitated the government’s turnover in early 2020 by deciding who to appoint as prime minister.

Recall the King’s unprecedented intervention in interviewing all the legislators individually to determine who commanded majority support during the government crisis in February 2020. “Politically neutral institutions such as the judiciary and monarchy attract attention in such circumstances as potential power-brokers,” Harding observes.⁶⁶ Of course, many established democracies in the West have also witnessed the breakdown of what have been thought to be well-established political and constitutional norms. In many constitutional systems, the spotlight has been on the rise of the courts as the institution empowered to decide. We see this, for example, in the United Kingdom Supreme Court’s invalidation of the Boris Johnson administration’s prorogation of Parliament in the lead up to Brexit at the end of 2019.⁶⁷

In a system of fragile (judicial) constitutionalism, as in Malaysia, it is the monarch that stepped into the breach by intervening in the immediate political moment. Turning the lens to other monarchies in Southeast Asia, think of the neighboring state of Thailand. Although described as following “in form” a “Westminster-style constitutional monarchy … in practice, the monarch [in Thailand] reserves the ultimate extra-constitutional powers to interpret, intervene, reject, or direct a course of action on the affairs of the state.”⁶⁸ On this account this kind of assertive monarchy reflects a Southeast Asian model of constitutional monarchy.⁶⁹ Some have used the term “Eastminster” to describe British postcolonial systems that have developed and deviated from a Westminster model.⁷⁰ But cautionary tales exist, and monarch that is not circumspect about guarding their legacy risks damaging its institutional legitimacy and authority. ⁷¹

The rules of democratic and constitutional engagement are not yet to be fully specified in regimes undergoing democratic transition(s),⁷² or after a decades-long equilibrium has broken down in a dominant party system, as occurred in Malaysia. In such circumstances, a non-electorally legitimated institution – like the monarch – that draws on its reserve powers to intervene in a stabilizing capacity may well renegotiate its role in the country’s governance.

Periods of instability or political transition may present a monarch with the opportunity to play a decisive role in modulating the transition process. A monarch’s enhanced function in times of crisis may prove dysfunctional at a later time, however. And in a political system has emerged from an initial period of instability, it may prove challenging to tie the hands of a monarch that has come to wield an expansive part in affecting the democratic character of a regime.