12 - Eternity Clauses and Electoral Democracy

Summary

This chapter explores the link between eternity clauses and electoral democracy by looking at two instances of unamendable democracy: party bans, both direct and indirect, and the protection 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. Whereas such measures are adopted in the name of protecting democracy, the analysis here indicates that courts will not always strike the right balance between safeguarding and unduly narrowing democratic commitments. In some cases, they may even unintentionally undermine multipartyism itself or significantly influence electoral outcomes. Thus, the bluntness and open-ended nature of unamendability risks having a chilling effect on electoral democracy in both fragile and more stable democratic contexts.

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.