1 Introduction

The development of democracy in Turkey over the past 20 years has alternated from a hopeful awakening after the turn of the millennium to a continuous regression in the last decade (Esen and Gumuscu 2021; Haggard and Kaufman 2021; Diamond 2021; Ozdogan 2021; Akboga and Sahin 2021). This is a challenge not only to democracy and opposition, rule of law and separation of powers, but also for its conceptual, analytical coverage. Various approaches are in use within political science, comprising competitive or electoral authoritarianism (Schedler 2006; Levitsky and Way 2010; Esen and Gumuscu 2016; Aytac 2021), hybrid regimes (Diamond 2002; Bogaards 2009; Bayulgen et al. 2018; Öney and Ardag 2022), illiberal democracy (originally Zakaria 1997), democratic de-consolidation (Foa and Mounk 2016; Waldner and Lust 2018; Akyuz and Hess 2018; Ágh 2019), and a deep state conspiracy (Filkins 2012; Gingeras 2010, 2011; Kaya 2009; Söyler 2013, 2015; Skowronek et al. 2021). With regard to the cases of Russia and Turkey, there is at least agreement that the term democracy (as a noun) is no longer appropriate. However, there is disagreement as to which concept should be used, especially against the background of comparative questions.

More recently, a number of publications applying Ernst Fraenkel’s (1898–1975) concept of the dual state have been added (Arato 2000, 2002, 2016; Bracken 2012; Ciftci 2017; Göztepe 2018a; Köker 2020; Lübberding 2016; Tekin 2018, 2022; Suntrup 2020). For instance, Tekin (2022) emphasises the Weberian roots of this concept and requests it to be taken as an ideal type. He points at the “selfreferential primacy of the prerogative state” and that “the normative state functioning only when the prerogative state remains silent” (ibid., p. 4). The concept of the dual state was originally developed by the Jewish German Ernst Fraenkel in the 1930s, writing the manuscript during his time as a lawyer in Berlin. In 1937, the first version (‘Urdoppelstaat’, Fraenkel 1937) was completed. Before his emigration via London to the USA in autumn 1938, he managed to send the manuscript abroad. A few years later, it was expanded, translated and published in English (Fraenkel 1941). It is not only a case study of the Nazi state during the 1930s, but also an abstract analytical concept, as the subtitle indicates (“A contribution to the theory of dictatorship”).Footnote 1 Although Fraenkel’s analysis was not intended as an ideographic case study.Footnote 2 However, this does not guarantee for its applicability as an analytical scheme to present-day authoritarian regimes. In the scholarly literature, however, there is growing interest in Fraenkel’s dual state as a conceptual framework for cases of rising authoritarianism (e.g. Sakwa 2010, 2011 and Plaggenborg 2006 for the case of Russia under Putin; Tunander 2012 for Sweden). Among the idiosyncrasies of the Russian case is that the rule of law, including the differentiation of public and private law, has never fully unfolded (Sakwa 2010, p. 189) and the ‘neoliberal’ dimension typical of Turkey was missing until the unleashing of the markets in the 1990s. Despite the idiosyncrasies of both cases, the intermingling of ‘rule of law’ and ‘rule by law’ observed by Sakwa seems relevant for Turkey, too. A case study on the strengths and limitations of its application to current developments in Turkey could thus also pave the way for similar studies on comparable cases, in Europe particularly Russia and Hungary. We will return to this in the conclusion.

The application of Fraenkel’s distinction between a routinely working ‘norm state’ and a sphere that suffers from prerogative interventions appears tempting to be tested in cases such as Turkey, but harbours some pitfalls. It is noticeable, for example, that these applications are often quite unstructured, a mere subsumption. The problem, however, is less about proving that an authoritarian state disintegrates into spheres of different degrees of norm binding or different modes of functioning than conducting a detailed analysis of the interdependencies and how this process of disintegration works. Furthermore, the focus on the legal binding of state action and its erosion remains often unclear. Against this backdrop, this article attempts to overcome these weaknesses by applying the three stages model developed by Dreier (2016) and Meierhenrich (2017) to the case of Turkey.

In comparison to the alternative approaches mentioned above, this approach does not claim to be ‘better’, but rather complementary. It focuses on mechanisms of the dissolution of the rule of law by the executive, highlighting this as an advantage of Fraenkel’s interpretation scheme. The assumption is that the concept can shed new light on the internal division and fractures within state institutions and the process of extending populist political authority, which according to Fraenkel was characterised by a mixture of party raison d’être and (neo)liberalFootnote 3 functional imperatives.

The selection of Turkey as a case study, as opposed to countries like Hungary (see Bogaards 2018) has the advantage of not being directly influenced by the EU (for Poland see Ágh 2019). In the case of Turkey, the accrual of capacity for prerogative, unchecked executive authority began long before the failed coup attempt of the Gülen movement (Hendrick 2013; Ebaugh 2012) in July 2016, the subsequent state of emergency, and the constitutional amendments (new presidential system) that became effective with the presidential oath of 09 July 2018. These developments further strengthened the authoritarian features of the polity at the price of checks and balances (Göztepe 2018a; Yilmaz 2019; Seufert 2018). In other words, this article aims to test strengths and weaknesses of Fraenkels dual state concept and its recent amendments with the case of Turkey.

It proceeds as follows: Section two provides the conceptual framework. Following Dreier and Meierhenrich, three expansion stages of the dual state are distinguished. Section three explores political reforms in Turkey under the question of a possible expansion of dual-state structures along these three subcategories. Finally, section four summarises the benefits and limitations of applying Fraenkel’s categories to the case of Turkey under the rule of Erdoğan. It also classifies the current constitutional crisis.

2 Analytical framework

This chapter first points out that dual state structures are much older and were maintained by different power constellations. Recent developments are therefore only a last step in a much older tradition of dual sovereignty. Subsequently, the analytical differentiation of the dual state added by Dreier and Meierhenrich is introduced as conceptual framework (2.2). Finally, alternative approaches and methodological issue are addressed (2.3).

2.1 A brief history of dual sovereignty in Turkey

Authoritarianism and a division of sovereignty between a civilian and a military sphere are eminent phenomena in Turkey (Karaveli 2018). Historically, dual state structures have often been the rule rather than the exception (Bezwan 2008; Yilmaz 2019). The concept of the deep state (or parallel state) differs from a dual state as a coexistence of norm-bound and politically motivated, arbitrary sphere of action. The former term seeks to identify autonomous structures in the security apparatus and administration that are uncontrolled by the state authorities and is popular in conspiracy theories. In Turkey, rumours of the invisible workings of a deep state beyond the official structures are rife (Filkins 2012). Such clandestine networks have been prevalent in the security sector of Turkey. However, they do not align with what Fraenkel called the prerogative state, which involves an extra-legal prerogative authority. In contrast, the deep state refers to an illegal but tolerated autonomization of security organs and administrative units (see Sect. 2.3). The specific civil-military divide came under pressure with the rise of political Islam in the 1990s, as this movement provides an alternative framework for informal coordination, albeit without unified quality. The AKP emerged in 2001 as the ‘moderate’ arm of this movement.

After the ruling elite’s falling out with the Gülen movement in 2013 and the coup attempt of July 15, 2016, authoritarianism has further intensified, and checks and balances were further undercut. Judicial and societal control mechanisms of state actions were weakened, informalisation and lack of transparency strengthened. The state of emergency, first approved on July 20, 2016 and extended until July 19, 2018, as well as the 2018 constitutional reform, have further promoted the expansion of diverse forms of prerogative state authority.

Historically, prerogative measures were closely linked to the rule of military-bureaucratic elites who saw themselves as guardians of a Kemalist ideology (Kilinç 2016). The content and the vehicles differed fundamentally from those of fascist ideology, for example (Plaggenborg 2012, p. 124). However, ‘Kemalizm’ (also known as Atatürkçülük) was not a coherent, closed ideology, but rather a collection of principles and opinions that people with very different worldviews could subscribe to. According to the party program of 1931, its basic principles “were republicanism, secularism, nationalism, populism, statism and revolutionism (or reformism)” (Zürcher 2017, p. 183).

The military-bureaucratic complex as guardian of Kemalism alongside the democratically elected governments since 1950, led to the prevalence of a dual sovereignty, “in which the elected government must be subordinated to the prescribed Kemalist program of secularism and national homogenization as defined and enforced by the military-civilian bureaucracy.” (Yavuz 2009, p. 210) This state within a state reserved the authority to take measures, even to the point of deposing elected governments. This also explains Erdoğan’s and the AKP’s cautious, tactical approach in the first legislative periods. With the election of Abdullah Gül as president in 2007 and the Ergenekon and Sledgehammer trials in 2008, which were conducted by pro-government administrative cadres due to alleged coup plans by the military (Seufert 2014). The latter’s influence on politics was gradually diminished. As a result, the political discourse also changed, and for instance a moral language influenced by the global human rights discourse took hold (Yavuz 2009, p. 267).

2.2 Fraenkel’s dual state and its subcategories

Fraenkel basically saw the state of the Weimar Republic under the pressure of the Nazis split into two areas of functioning: a norm state in which a legal-rational order was preserved, but also changed according to the ruling ideology, and a prerogative state in which actions became arbitrary respectively motivated by political or ideological causes. Both are alternative, irreconcilable forms of state actions (Normenstaat and Maßnahmenstaat; Fraenkel 1941, p. 147), based on two incompatible forms of rule (legal-rational versus charismatic rule) according to Weber. These forms do not allow for lasting stability. The prerogative state is characterized by highly irrational actions and arbitrary decisions. “One of the basic propositions of Max Weber’s works is that a rational legal system is indispensable for the operation of a capitalistic economic order. […] But we must then resolve the paradox of a capitalistic order continuing within a system under which there is no possibility of rationally calculating social chances.” (Fraenkel 1941, xiv) This ‘paradox’ can be resolved, for instance, by a self-restraint of the prerogative state that is aware of its deficits in comparison to the norm state and its stabilizing effect on the economy, at least initially. Despite the arbitrariness of the prerogative state (as consequences of the charismatic elements), the norm state is needed for the predictability of economic transactions, for instance. Without it, it would be hard to keep a capitalist economic order based on the division of labour permanently stable or even on a growth path.

Stefan Breuer (2004, p. 40) has pointed at ambiguities in Fraenkel’s opening sentence that the constitutional document of the Third Reich was the so-called ‘Reichstag Fire Decree’. On the one hand, this made a clear distinction between the Third Reich and the Weimar Republic. On the other hand, it explicitly referred to Article 48 (2) WRV and had been signed by a duly elected Reich President. This suggests that Fraenkel “already saw the double state laid out in the Weimar Republic” (ibid.). He also considered Max Weber’s insistence on a directly elected Reich President a mistake, as it incorporated a bridgehead for the counter-revolution into the WRV and could legally pursue its abolition (ibid.).Footnote 4

In addition to the distinction between a normative and a prerogative sphere of state, Fraenkel’s analysis also includes observations on the institutional effects of this differentiation as well as on the origin of the dual state from a community ideology (Meierhenrich 2017, lxi). However, in this article the focus is on the differentiation of three emanations of the prerogative state. Based on Dreier (2010, 2016), Meierhenrich (2017, lxiii) has outlined three subcategories of the prerogative state, namely its transgressive (1), restrictive (2), and constitutive (3) powers.

ad 1: These are forces in the state that undermine or override the functioning of the norm state. According to Fraenkel’s functional perspective, they can occur in a variety of institutions, but are most frequently observed in those that are most susceptible to political influences, such as appointments and promotions across administration and judiciary. However, not all areas of the judiciary are easily influenced by the executive. Long-term imprisonment without charge and sentence, confiscation of property without proper authorisation, and the destruction of professional existences without recourse to the courts are examples of this type of encroachment.

ad. 2: This point refers to anticipations of the political will within the norm state’s institutions. In areas and among individuals where the rule of law is less established, there can be a tendency to anticipate and obey the prevailing political will, partly out of fear of personal consequences, but also out of inner conviction. Dreier (2016) distinguishes between self-restriction and external restriction. An example of the latter is to exempt parts of the security apparatus from judicial control.

ad 3: Finally, the ideology or expectation of the prerogative state can become so internalised by actors within the norm state that it largely overrides the neutral application of the law. This is particularly evident in discretionary decisions, where actors not only exercise self-restraint but actively adopt and apply ideological principles, which has repercussions in the legal system itself. Legal norms are no longer interpreted in the spirit of the constitution, but in the spirit of the prevailing political will. Barriers and other self-protective mechanisms of the law can be more easily overcome, leading to the erosion of judicial independence or adaptation to the dominant political. Chapter three applies these categories to the case of Turkey and aims to show the mechanisms of intrusion of political power into the legal system.

2.3 Alternative concepts and methodological issues

The literature on the state and political regime and internal structures in Turkey is already extensive and divers. However, the theoretical embedding of these works is less extensive.Footnote 5 One of the strengths of Fraenkel’s approach is its ability to link individual practices to larger structures and developments, allowing for predictions about the dynamics of the normative or prerogative structures themselves. Fraenkel’s overall concern is the collapse of law binding and the corrosion of law itself from a perspective of what nowadays is called historical institutionalism and structural-functional analysis (Thelen 1999; Lane 1994; Peters 2019). His historical background is the problem of the institutionalisation of statehood and the rule of law since the 18th century. He focuses on ‘initial choices’ of dual sovereignty such as constitutional ambiguities and loopholes or on ‘critical junctures’ (Gerschewski 2021) and significant interruptions such as the ‘Reichstag Fire Decree’ of February 1933. The idea of path dependency can also be found in his works. This theoretical background can contribute to a more structured analysis of developments ‘from within’. We do not argue that Fraenkel’s approach provides an overall ‘better’ theoretical framework, but rather a different one that sheds light on neglected aspects of alternative approaches. In the case of Turkey, historical institutionalism means, for example, taking into account long-term developments in dual sovereignty and authoritarianism (Karaveli 2018). Exclusively present-day focused analyses tend to attribute blame to individual actors and tend to neglect the historical and structural environment as a conditioning factor. Competing approaches such as elective authoritarianism often lack this historical, structural background of explanation. However, Fraenkel’s analytical scheme has been criticized for being time-bound and under-complex for today’s cases, too (Breuer 2004). On the other hand, its recent extension and application illustrate the usefulness of this approach. It must be borne in mind that it is not an alternative, but a complementary to currently prevailing approaches in comparative authoritarianism. Its focus is on the law-binding of government and administration and the challenges of its enforcement or erosion.

Köker (2020) provides a good synopsis of the strengths and weaknesses of theory-based studies of Turkey in terms of de-democratization and authoritarianism. He concludes, for instance, that the pattern of system change in Turkey almost exactly overlaps with the concept of ‘competitive authoritarianism’ as described in the literature (Levitsky and Way 2010). Its application to Turkey by Esen and Gumuscu (2016) highlight the increasingly uneven playing field such as politicised public institutions and biased media. However, they do not provide a structured approach to analysing the encroachment of the administration and institutions (polity) by party politics. Therefore, the results hardly go beyond what is already known: Elections becoming increasingly unfair, civil liberties are under pressure, and the overall playing field is tilted towards the incumbent party.

Levitsky and Way (2010) do not include Turkey and put special emphasis on international leverage and linkage mechanism by respectively from Western countries. Their focus is on long-term historical processes and driving factors, independent from individual leaders. “At the international level, linkage to the West (with the partial exception of EU-led integration) is less the product of elite decisions than of geography, economic development, colonialism, and long-standing geostrategic alliances. Similarly, at the domestic level, strong coercive and party organizations are rarely the product of short-term crafting or institutional design.” (ibid., p. 83). In fact, there is a tendency in Turkey studies to focus on domestic factors such as personality, leadership, and party dominance. In this respect, the approach of Levitsky and Way rightly point to international conditions of autocratization also for Turkey; a perspective that is largely missing in Fraenkel. On the other hand, his approach offers more detailed instruments for explorative single case studies. The results of Levitsky and Way remain descriptive; theoretical generalisations remain vague. This is not uncommon for ex post facto designs; forecasting capability is not their best. For Köker, the weight and persistence of the formal democratic elements in the Turkish case are a crucial factor for any predictions. This shifts constitutionalism and rule of law into the centre and requires a specific theoretical framework that is capable of analyzing this dimension of authoritarianism. Köker traces this process using concepts of Carl Schmitt (2015; see also Krumm 2018; Şahin 2017) and Ernst Fraenkel (see also Arato 2000, 2002) but does not carry out a detailed analysis bases on recent innovations in this concept.

Another alternative is the deep state concept. Firstly, it is worth noting that the deep state concept is only applied to a few countries, including Turkey and the USA.Footnote 6 Söylers (2015) study of the Turkish deep state suggests path-dependency as an integrative theorem, combining inductive methods with structure-agency theory. He locates the concept of the deep state at a meso-level with a special focus on informal structures revolving around the security sector. The historical-institutional approach certainly overlaps with Fraenkel, while the focus on the security sector and the development of informal structures (Söyler 2013, 2015) is its main difference. Problems of legal binding are particularly evident in the often opaque security sector, but not limited to it. Kaya (2009) aims to test Mancur Olson’s theory of collective action against the Turkish deep state, thus, applying a public-choice perspective and focusing on rent-seeking coalitions and exclusionary processes. This approach does not focus on the erosion of rule of law and legal binding. From Fraenkel’s perspective, the development of informal, gang-likeFootnote 7 structures in this sector and possibly overarching ‘state capture’ represent only a section of the comprehensive problem of rule of law improvement. Furthermore, the ‘substantivisation’ of the deep state concept makes informality, secretive actions, and public-private networks seem more tangible. This can create a deceptive certainty of theoretical comprehension. Both the deep state and dual state are not unitary actors; however, Fraenkel’s concept better captures the internal fractures, contradictions, and interdependencies of these different spheres of the state. These conceptual differences do not preclude the deep state concept from achieving a (selectively) higher level of theory—at the expense of embedding it in legal and social theory. This may be one of the reasons why the deep state concept as a ‘dependent variable’ has not become common in comparative politics (however, this also applies to the dual state concept). Both have so far only been used for single case studies, as in the present case. In the case of Turkey, deep state structure declined in the late 1990s, coinciding with the rise of political Islam as an alternative framework of informal coordination.

Regarding methodology, in Fraenkel’s time issues of case study design were not yet known. This gap in the original concept must be filled subsequently.

It would be tempting to label such theory applications as crucial case studies, following Eckstein (1975), who argued that crucial cases should fit well with a theory “if one wants to have confidence in the theory’s validity or, conversely, must not fit equally well any rule contrary to that proposed.” (ibid., p. 118). John Gerring (2007), for example, has pointed at difficulties of this kind of study. Thus, the aim of this article is not to prove or disprove Fraenkel’s concept with a single case study.

Instead, the article argues that Fraenkel’s scheme and its elaboration by Dreier and Meierhenrich allow for a more systematic exploration of the erosion of the rule of law. This calls for a theory-oriented case study design. The innovation of Dreier and Meierhenrich is to understand the dual state not as static but as dynamic, thereby strengthening the process dimension. The dual state is no longer understood solely in substantive terms, but also in heuristic terms, which aligns with concepts of process tracing (Bennett and Checkel 2015). This strength becomes clear when it comes to the classification of the current constitutional crisis in the scheme (3.3).

3 Adaptation of a prerogative state under Erdoğan?

During Erdoğan’s first term in office (2002–2007), there were initially positive trends in democracy, the rule of law and rapprochement with the EU (Öniş 2009). However, with the crisis of the 2007 presidential elections and the change in the electoral mode, these processes increasingly came to a halt. However, since the rupture with the Gülen movement in 2013 (less due to the Gezi Park protests) and its coup attempt in 2016, this development has taken on a new quality. This chapter discusses the establishment of prerogative powers along the subforms outlined above.

3.1 Transgression: Counteracting the norm state

The government’s repeated emphasis on strengthening the rule of law through new reforms, most recently in a new ‘civilian’ constitution, would seem to argue against a breach of the rule of law in Turkey. The need for reform of the judicial system is implicitly acknowledged through various ‘reform packages’ and ‘human rights action plans’, but these reforms are primarily aimed at attracting foreign investment (Grigoriadis and Kamaras 2008; Önder 2016; Bozkurt-Güngen 2018) and have little practical effect. For example, the state of emergency that followed the coup attempt in July 2016 (which itself was declared in accordance with the constitution) vastly expanded the scope for arbitrary dismissals and arrests through presidential decrees. By the end of 2017, 159,506 people had been arrested under emergency decrees and at least 152,000 civil servants and employees had been dismissed, many of them teachers and academics. “Over 4200 judges and prosecutors were dismissed through executive orders of the High Council of Judges and Prosecutors while the Constitutional Court dismissed two judges. An additional 22,474 people lost their jobs due to closure of private institutions, such as foundations, trade unions and media outlets.”Footnote 8

As the dismissals were carried out by naming individuals in the annex to the decree, the affected persons were cut off from legal recourse before both the administrative courts and the Constitutional Court (Göztepe 2018a, p. 530; Yesil 2021). The same practice was applied in a ‘temporary’ override of Article 83(2)(1) of the Constitution to waive the immunity of 138 of the 550 members of Parliament through ‘transitional Article 20’ (gecici madde). With 373 votes in favour (67.8%) from AKP, CHP (Republican People’s Party) and MHP (Nationalist Movement Party), the approval on 20 May 2016 was just above the required two-thirds majority. MPs from all parties were affected by the measure, but from the pro-Kurdish HDP, 50 of its 59 MPs lost their immunity and could be charged under strict anti-terrorism standards. Many of the charges against HDP deputies stemmed from speeches they had made in favour of Kurdish autonomy, for example.Footnote 9 Erdoğan had argued, among other things, ‘My people do not want to see deputies in parliament who have committed crimes’ (Mertens 2016). During the parliamentary vote on the 2017 constitutional amendment, 11 of the 59 HDP deputies were in prison on terrorism charges. In the constitutional referendum on 16 April 2017, 51.4 per cent approved the amendments, with a turnout of 85.4 per cent (Krumm 2018).

Arbitrary counteraction to the normative state also includes the fact that prisoners are not released despite such rulings by courts, including the European Court of Human Rights (cases of Osman Kavala, Selahattin Demirtaş),Footnote 10 or are immediately re-imprisoned with a new arrest warrant after a short release, such as the journalists Ahmet Altan and Nazli Ilicak, who were sentenced to life in solitary confinement without the possibility of parole after the 2016 coup attempt.Footnote 11 After going through the courts, the Court of Cassation overturned the sentences in July 2019 and ordered a retrial, followed by conditional release in November. Just a week later, a new arrest warrant was issued for Altan on the grounds of a flight risk. The issuance of a new arrest warrant with amended charges is the usual procedure to counter release orders from the higher courts.

Another example of the flouting of norms is the new presidential palace in Ankara, built in a forest area declared a nature reserve by Atatürk. Several court orders to stop construction, which began in 2011, were simply ignored. Moreover, a decision by the Supreme Administrative Court in May 2015 declaring the construction illegal and retroactively cancelling the building permit had no practical consequences (Akyol 2018, p. 241).Footnote 12 The examples show that the normative state is easily undermined by the politically ‘captured’ areas of the judiciary and administration ignoring its decisions. In this process of counteraction, it can be helpful to refer to decisions made at lower levels that have already been politically ‘captured’. In the hierarchy of law, the constitution binds the legislature, and the laws enacted bind the administration and the judiciary. The actors of the prerogative state can act at various points to politically manipulate this chain of legitimation. From the executive’s point of view, the problem is that, against the background of a constitution and jurisprudence that say otherwise, constant violations and countermeasures are still visible and criticisable as violations of norms. The political momentum is therefore clearly aimed at rewriting the constitution itself. The Constitutional Court is probably the last major obstacle to the supremacy of political will throughout the state. It is denounced, for example, as a remnant of the parliamentary system.

3.2 (Self)Restriction: Withdrawal of control capabilities

The next stage is an expansion of self-imposed and external restrictions on judicial control options. The individual constitutional complaint introduced in 2012, for example, is a counter-argument to the increasing restriction of judicial control options (Göztepe 2018b). However, this substantial progress is undermined by a) the government’s influence in the selection of judges and b) in critical cases by simply ignoring the decision. The latter became clear in 2018, for example, when the Constitutional Court criticised the long pre-trial detention of two journalists. The relevant criminal courts ignored the ruling on the grounds that it had not yet been published in the Official Gazette.Footnote 13 These (rare) critical rulings have led to a political debate on the role of the Constitutional Court, with nationalist parties even calling for its abolition. The Turkish Constitutional Court usually acts in a very restrained manner, for example when it ruled in 2017 that it had no jurisdiction to review emergency decrees (“lack of jurisdiction”)Footnote 14. Even before that, appeals against curfews in the south-east of the country were not accepted.

The predominant aim of the executive is not to expand the scope of control, but to protect critical areas of state activity from judicial scrutiny by means of legislation.Footnote 15 A law passed by the Turkish parliament in 2016 grants immunity to soldiers in the fight against terrorism. A few days earlier, “border guards had been accused of shooting at refugees trying to cross into Turkey from Syria, killing children and women as well. The Turkish government denied the allegations”Footnote 16 The law passed on June 23, 2016 requires permission from the military or political leadership to open investigations against soldiers in the fight against terrorism.Footnote 17

The immunisation of parts of the security apparatus from judicial control also falls into this category. This development has accelerated considerably since the Gezi Park protests (May 2013) and the break with the Gülen movement at the end of 2013 (Zürcher 2017, p. 353). Prosecutors close to Gülen had already revealed secret negotiations by the intelligence service (MIT) with the PKK on behalf of the government in 2012. “When the public prosecutors issued a warrant for Fidan’s arrest, the government immediately reacted and, the same day, parliament passed a law that exempted the MIT from prosecution and took the public prosecutor involved off the case.” (ibid., p. 357) A corresponding intelligence committee of the parliament does not exist. The conflict escalated further in 2014 when prosecutors and police chiefs close to Gülen, in defiance of a government order, stopped and publicised arms shipments from Turkish intelligence to Syrian rebels near the border. In June 2019, five of the defendants were sentenced to more than 20 years in prison for publishing secret state documents. The resulting message is clear: restraint is the order of the day.

Furthermore, Emergency Decree 696 of 24 December 2017 granted impunity to civilians who participated in the lynching of coup plotters on the night of the coup and afterwards.Footnote 18 The emergency decrees could not be reviewed by the Constitutional Court due to the state of emergency. Unlike the Nazi Law on “Measures of State Self-Defence”Footnote 19 of July 3, 1934 (“Röhm-Putsch”), Emergency Decree 696 initially legalized the non-organized actions of civilians on the coup night and thereafter as well. According to § 1 of the decree, anyone “who decides to act against coup efforts, such as the coup attempt of July 15, 2016, terrorist efforts, or any type of continuation of these activities, (…) is excluded from legal, administrative, financial, and criminal prosecutions.”Footnote 20 The decree has been criticised for its vague wording, which could in future give carte blanche not only to individuals but also to (paramilitary) groups, provided they can be linked to the defence against a terrorist threat. In addition to removing the possibility of judicial oversight, the state of emergency was used to purge the civil service through dismissals by decree and the closure or expropriation of businesses against which there was no legal recourse.

3.3 Constitution: Incorporating the courts into the prerogative state

Incorporating the courts into the prerogative state occurs, for example, through the creation of special courts and through the self-degradation of existing courts to executive bodies of the prerogative state (Dreier 2016, p. 362). In Turkey, however, a counterexample are the special courts for serious political crimes that were abolished in a 2014 judicial reform. The pending cases were transferred to the pre-existing jury courts. A separate military jurisdiction was abolished only in 2017. The state security courts were abolished as early as 2004, and some of their tasks were transferred to the criminal courts. In 2005, a new regional appellate body (istinaf mahkemesi) was introduced for criminal and civil courts, but started working only in 2016. Respect for the independence and impartiality of the judiciary is also mentioned as a basic value in a current judicial reform.Footnote 21

However, these seemingly positive developments contrast with another practice. The transfers and dismissals of judges and public prosecutors, which have intensified since the corruption investigations of 2013 and the attempted coup in 2016, are illustrative of this.Footnote 22 The Council of Judges and Public Prosecutors (HSK) became a political instrument of the government through multiple restructuring (Gümüş 2016; Krumm 2018). Immediately after the coup attempt in 2016, two members of the Constitutional Court were arrested on suspicion of links to the Gülen movement and dismissed from all offices by Emergency Decree 667 (‘Decree on measures to be taken under state of emergency’). The pre-trial detention, not the dismissal itself, was condemned by the European Court of Human Rights on April 16, 2019.Footnote 23 Özbudun (2019) observed that the Constitutional Court recently increased its self-restraint. “This shift can be explained as part of the political regime’s drift towards competitive authoritarianism and the governing party’s (AKP) capturing almost total control over the entire judiciary.” (ibid., p. 278) Given this context, the current push for a new ‘civil’ constitution aims to align the still unpredictable constitutional court, as it currently derives its legitimacy from the constitution rather than from the will of the people.

The character of emergency decrees as prerogative laws is expressed for instance in Article 1 of Decree 667 of July 22, 2016: “The aim of this Decree Law is to establish measures that must necessarily be taken within the scope of attempted coup and fight against terrorism under the state of emergency declared throughout the country by the Decree Law of the Council of Ministers dated 20 July 2016 and numbered 2016/9064, and to determine procedures and principles relating to these measures.”Footnote 24 Article 3 (“Measures related to the members of the judiciary and those considered as members of this profession”) grants judicial bodies, including the Constitutional Court, the authority to decide with an absolute majority on the dismissal of its own members accused of links to terrorist organizations.Footnote 25 A mere ‘connection’ of members of the Constitutional Court to structures, organizations, or groups considered by the National Security Council to be involved in activities against the national security of the state is sufficient for dismissal (§ 35).Footnote 26

Shortly after the end of the state of emergency on July 19, 2018, parliament passed an anti-terror law (‘Fight against terror in the normal state’), initially limited to three years, which made many regulations of the state of emergency permanent. The law also contributes to the aforementioned politicization of the provincial governors by extending powers such as restricting freedom of assembly from the state of emergency. Similar to the state of emergency, outdoor demonstrations are prohibited after dark. “With the new, vaguely worded argument that they ‘must not complicate the everyday life of citizens in an extreme and unbearable way’, the state gains another opportunity to prevent assemblies.”Footnote 27

The presidentialization of the government system that came into force with the parliamentary and presidential elections in June 2018 also contributes to the expansion of state competence. As part of the constitutional reform, directorates such as the Karayolları Genel Müdürlüğü (KGM, General Directorate for Motorways) were directly subordinated to the President. This arrangement is reminiscent of the special powers directly under the ‘Führer’ in Nazi Germany (Frei 2001, p. 142), characterized by an “often grotesque juxtaposition of norm and prerogative state” (Frei 2001, p. 143). Another example of the (self-)incorporation of courts is their participation in building up a political ‘prisoner pool’ in 2017/18, including individuals such as Andrew Brunson, Deniz Yücel, and Peter Steudtner, to strengthen political bargaining power vis-à-vis foreign governments.

The arbitrariness of the state was also reflected in the decision of the High Electoral Commission (YSK) to repeat the election for the mayor of Istanbul on March 31, 2019, following a complaint by the ruling party. The YSK found the argument convincing that not only state employees but also private school teachers and bank employees had supervised the ballot boxes, contrary to the requirement for state employees only. Critics of this decision emphasized that while the mayoral election (which was lost by the AKP) was repeated, the district mayor and municipal council elections, held under the same conditions (in which the AKP was successful), were not repeated. However, this example also illustrates that such interventions can backfire, as the opposition candidate won the repeat election with 54.2%.

In this context, the EU foreign affairs representative referenced the YSK’s practice in southeastern Turkey of denying elected mayoral candidates office and instead giving it to the runners-up (AKP candidates).Footnote 28 In autumn 2019, the interior minister dismissed HDP mayors in major cities in the southeast on suspicion of supporting the PKK, transferring their mandates to the provincial governors appointed by the interior minister.Footnote 29 At the end of 2020, the attorney general applied to the Constitutional Court for a party ban on the HDP, but the court is still deliberating on the matter. These examples illustrate the dynamics of transgression and incorporation of the judiciary into prerogative authority. The presidentialisation and the projected new constitution can easily be placed within this context of superseding prerogative power. A new constitution would offer the opportunity to formally legitimize the loosening of legal constraints and to remove remaining veto players such as the Constitutional Court. This project is also used for agenda-setting and to divert attention from issues such as the continuing economic crisis.

In November 2023, the Supreme Court of Appeals issued criminal complaints against the nine Constitutional Court members who voted in favour of the release of the sentenced Workers’ Party (TİP) lawmaker Can Atalay over alleged links to the 2013 Gezi Park protests. The Constitutional Court had previously ruled for his release because of his infringed right to be elected. Instead of releasing him, a local court had forwarded the case to the Supreme Court of Appeals’ Chief Public Prosecutor. “In a unanimous and historically significant 23-page ruling, the Court of Appeals has ruled to disregard a Constitutional Court decision for the first time, refusing to comply with the violation decision and release order regarding Atalay. Instead, they have mandated that Atalay’s case be forwarded to the parliament for deliberation on his parliamentary membership’s termination.”Footnote 30 However, the politicization of the judiciary can also be traced back to the supporters of the small alliance partner MHP, which was particularly outraged by the Constitutional Courts order in March 2023 to unblock the bank accounts of the pro-Kurdish opposition party HDP.

4 What future for the dual state in Turkey?

This article has demonstrated that Fraenkel’s concept of a dual state is applicable to the case of Turkey. Furthermore, it has expanded upon existing dual-state literature by utilizing the three subcategories proposed by Dreier and Meierhenrich to analyze the interaction between the normative and prerogative spheres of the state. These subcategories should be viewed as a heuristic scheme rather than a sequence of stages. This becomes clearer when countries such as Russia and Hungary are taken into account. The differences to the Russian case (regarding the use of state of emergency powers, for instance) have already been pointed out at the beginning. Sakwa’s observation from 2010 appears transferable, suggesting that two political “orders operate in parallel, and neither system is allowed to work within the full capacity of the logic of its own dynamics.” (202) Each of the two spheres competes and interferes with the other, “leading to a crisis of governance and the rampant proliferation of institutional nihilism. […] Ironically, the deeper the crisis, the greater the emphasis on administrative governance, but this is more an indication of weakness than strength, and only exacerbates hybridity and, thus, crisis.” (ibid., p. 202).

In Hungary, the Orbán government carried out constitutional reform at the beginning of its tenure, while the recent case in Poland shows that authoritarianism is not a one-way street. In Turkey, the ongoing project includes an attack on the constitution and its court. However, one should not assume determinism in these cases. Events such as the introduction of the individual constitutional complaint in 2012 and results of the local elections of 2019 illustrate the scope of possible developments. The temporal dimension is one of the issues when applying Fraenkel to current cases. More conceptual work is needed to modernize Fraenkel’s concept. Present-day ‘dual states’ do not automatically and swiftly lead to self-destruction. They are apparently able to keep the different functional imperatives in abeyance for longer. The current constitutional crisis in Turkey is a good illustration of competing forces within the institutional structures.

Some judicial reforms up until the introduction of the individual constitutional complaint in 2012 were certainly progressive, and the repeat election in Istanbul in June 2019 shows that not all transgressions are accepted by the electorate. This illustrates limitations of the dual state approach for current cases. The observation of taking the coup attempt as an opportunity to further streamline the judiciary and administration does not require much theoretical effort. Regular elections still seem to provide a chance of replacing ruling elites, contradicting hard authoritarianism. The press and the media are under increasing pressure or streamlined, but by no means synchronized as in the Nazi era; civil society is still surprisingly well organized (Yabanci 2019), compared to Russia, for instance. Furthermore, even the updated versions of the dual state do not take into account factors such as the format and polarization of the party system (Haggard and Kaufman 2021) or political culture.

Nevertheless, Fraenkel’s analytical scheme can help to better understand the functional logic and legitimacy claims that erode rule of law and thus democracy today. For instance, the use of plebiscitary forms of legitimation is neglected in many studies, but is on Fraenkel’s agenda. Plebiscitary and populist forms of politics, he argues, are based on the tacit presupposition of a unified popular will, which is assumed a priori to be identical with the general interest or will (Fraenkel 1991a, p. 154). This is reminiscent of the rhetoric of a unified national will (“milli irade”), not only in Turkey’s presidentialization discourse. From this point of view, minority and special interests are disruptive factors of an imaginary ‘will of the people’ as the supreme authority (which is somewhat reminiscent of the Nazi ‘Volksgemeinschaft’). The plurality of empirical wills should therefore be either negated or neutralized by relegating them to a pre- or apolitical space (Fraenkel 1991a, p. 154 f.).

As a neo-pluralist, Fraenkel was anti-monist, which became particularly clear in his rejection of Carl Schmitt’s anti-parliamentarism and praise of democratic dictatorship (Fraenkel 1991b; Wildt 2009). Fraenkels later works focus on how particular interests, derived from majority decisions, are declared to be the general interest or ‘will of the people’.Footnote 31 Conversely, it becomes clear why there is no acceptance for legitimacy claims by institutions (e.g. the rule of law, constitutional court) beyond this imagined will to which everything must be subordinated. The appeal to a homogenous-monist, supposedly general will is present for instance in the government’s discourse on the unity and strength of the nation. However, such claims have a much older tradition and can be traced back to the Young Turks and Ataturk (Karaveli 2018). Fraenkel’s categories provide a theoretical framework for analysing the internal functioning, dynamics, and contradictions of such hybrid legitimacy claims, based on a prerogative sphere of seemingly effective governance and the remnants of the norm state and its democratic façade.