Constitutional Referrals by Ordinary Courts: A Platform for Judicial Dialogue and Another Toolkit for Judicial Resistance?

Constitutional referrals by ordinary courts – Judicial dialogue between constitutional courts and ordinary courts – Toolkit for judicial resistance – Novel taxonomy of features and functions of constitutional referrals – Separation of powers approach – Democratic decay – Self-defence mechanism – Mix of methodological approaches – Case study on Czechia

which it can be conducted. 2I argue that the referrals by ordinary courts to constitutional courts can serve as a platform for such a dialogue.Moreover, I explain how this mechanism forms a unique toolkit in the hands of the judiciary vis-à-vis other branches of power, which is of special interest in a time of democratic backsliding.
A referral process in which one court has the ability to call upon another, particularly a superior court, for interpretations of law can be observed in various, mostly European, jurisdictions. 3 Perhaps the most popular is the preliminary ruling procedure in the EU. 4 However, this EU procedure derives from the constitutional referrals mechanism previously adopted in several national judicial systems of the member states, where ordinary courts might query a constitutional court. 5n contrast to the preliminary ruling procedure before the European Court of Justice 6 or the horizontal dialogue among international courts, 7 dialogue between constitutional courts and ordinary courts is considerably under-researched.Internal constitutional referrals have been analysed in more detail only by de Visser in her general comparative study 8 and by other authors in writings on constitutional systems of particular countries. 9Similarly, this mechanism has been overlooked by scholars dealing with abusive constitutionalism and democratic decay. 10No one has yet viewed this mechanism through the lens of judicial resistance.Therefore, this article partially fills this gap and delivers the first more general study of this mechanism.
In this article I analyse constitutional referrals both theoretically and empirically, from a general perspective as well as in in-depth case study.I conceptualise this mechanism and provide a novel taxonomy of its main features (the first section) and functions (the second section).I frame this mechanism from the perspective of judicial dialogue on constitutional matters and argue that this mechanism might serve as an exclusive self-defence mechanism of the judiciary against other branchesas a vehicle for judicial resistance in the era of democratic backsliding.This original perspective contributes to the topical debate on judicial dialogue as well as judicial resilience.
To step out of the theoretical level, in the third section I demonstrate these claims on the single case study of Czechia.Czechia has not yet been subject to challenge of the same magnitude as its regional counterparts; however, clear signs of fragility and susceptibility to democratic backsliding might be observed.Thus, Kosař and Vyhnánek refer to Czechia as being in a state of 'democratic careening'. 11The healthy constitutional dialogue between courts and the Rights', 29 Penn State International Law Review (2011) p. 759; J. Ferejohn  institutional self-defence of the judiciary can therefore play a key role there.This is even more so given the intriguing evolution of the relationship between ordinary courts and the Czech Constitutional Court in the last 30 years. 12Furthermore, in the light of the current cases before supranational courts involving Hungarian and Polish judges, claiming their right to independence, 13 Czech judges can be considered pioneers in this respect, as they have already been successful at the national level.

U  :   
The mechanism of constitutional referrals exists in every European state which has a specialised constitutional jurisdiction. 14Its essence lies in concrete constitutional review (i.e.review of a law in the context of an actual case) initiated by ordinary courtsbut only initiated.The final say belongs to a constitutional court that provides an answer on the constitutionality of an impugned statute or its particular provision. 15Consequently, the referring court has to decide the case at hand in accordance with the constitutional court's ruling. 16hat said, the dynamics of this mechanism vary depending on the legal (institutional), political and societal context within which it operates.What matter are, in particular, the other powers of the respective constitutional court and access of individuals to any kind of constitutional review.Take, on the one hand, systems with individual constitutional complaint (e.g.Germany, Austria, Switzerland, Poland, Czechia) or recurso de amparo (Spain), the French model with the long-term dominance of the parliamentary minority in access to constitutional review, recently seasoned by specific constitutional referrals that Pfersmann calls 'indirect constitutional complaints', 17  Italian model where constitutional justices are mostly 'judges of judges' 18 and where individual constitutional complaint is completely absent. 19In systems with direct access of individuals to constitutional review (constitutional complaints), referrals by ordinary courts amount to a rather supplementary part of constitutional reviewthe vast majority of cases before these constitutional courts are brought by individuals, not courts. 20On the other hand, in the Italian model constitutional referrals by ordinary courts represent the most important way of activating constitutional review.Below, I show how the political context (e.g.constitutional courts' capture) or the socio-legal context (e.g. a complicated relationship between constitutional courts and ordinary courts in transitional democracies) shapes the functioning of constitutional referrals.
Although the dynamics of this mechanism differ across jurisdictions, some general conceptual remarks are in order.Using an institutional perspective, let us look in detail at six underlying features of the constitutional referrals mechanism as introduced above.
First, the precondition for constitutional referrals is the presence of a body to refer toa constitutional court or another supreme court with a monopoly of constitutional review. 21Thus, the mechanism of constitutional referrals, as perceived in this article, works only in the centralised model of constitutional review. 22Even though some sort of internal preliminary reference in constitutional matters is conceivable also in the so-called diffuse model with decentralised constitutional review, that goes beyond the scope of constitutional referrals as understood in this text. 23Centralised constitutional courts were mostly established as guardians straddling law and politics, organised separately from the rest of the court system. 24Constitutional courts in countries adhering to the modern German model are closer to the ordinary judiciary, while in the French model the constitutional court (council) inclines more towards the political sphere.Nonetheless, mainly because of the extension of human rights and the influence of supranational courts (especially the European Court of Human Rights and the European Court of Justice), both ordinary and constitutional courts are far from being truly separate. 25Rather, they retain their specifics, unique virtues and shortcomings, but can be deemed part of the judiciary in a broader sense. 26econd, the whole mechanism is triggered by an ordinary courteither any court (e.g. in Germany, Poland, Belgium or Czechia) 27 or merely a court of a specific level (for example, in Bulgaria power to refer is reserved for the Supreme Court of Cassation and the Supreme Administrative Court alone; and the French model has already been mentioned). 28Accordingly, we can distinguish: (a) the model with unlimited referrers (all ordinary courts); and (b) the model with limited referrers.Ordinary courts are generally empowered to interpret and apply ordinary law and, at the same time, they are not endowed with the power to strike down statutes incompatible with constitutional requirements. 29The mechanism of constitutional referrals, however, gives ordinary courts an opportunity to pronounce, albeit to a limited extent, on the issue of constitutionality.As will be explained below, ordinary courts play a pivotal role in these proceedings since they embody gatekeepers 30 sifting through the caseload and, in a similar vein, 'agenda-setters' raising the reference.
Usually, though, ordinary courts are only the 'secondary' agenda-setters, as the initiating proceedings emanate from a case brought to a court by individuals  whose role in these proceedings should also not be overlooked. 31That holds true especially in the French model, where only parties may raise the question of unconstitutionality and, after the scrutiny of two ordinary courts (the court hearing the case at hand plus the Cour de Cassation or the Conseil d'etat, respectively the civil/criminal and administrative supreme courts), 32 the question is 'only' transmitted to the Constitutional Council.In other words, ordinary courts lodge questions that are not their own and the original parties are also the primary parties, even though they are not themselves empowered to transmit the referral (question). 33Thus, Pfersmann refers to this model as 'indirect constitutional complaint' by individuals. 34However, it is still important to bear in mind that dialogue takes place between two courts which somehow adopt the party's question.
Third, as the review arises out of the regular adjudication of a case between parties, the constitutional referrals mechanism represents a type of concrete judicial review. 35This type of review consists of the verification of whether a law to be applied in a particular judicial case is actually constitutional.It encompasses both referrals by ordinary courts (objective concrete review) and incidental review of legislation proposed by the parties during constitutional complaint proceedings (subjective concrete review). 36However, as Sadurski points out, this kind of concrete review may sometimes be 'contaminated by abstractness'. 37onstitutional courts occasionally examine references by ordinary courts without really engaging with the facts of the case at hand and merely juxtapose a contentious law with the constitution in an abstract fashion.Hence, a formal 'concrete' review may virtually transform into an 'abstract' one. 38At any rate, a specific case can sometimes reveal a constitutional problem which is not apparent in abstracto, and therefore a concrete review is a key way of protecting fundamental rights. 3931 In his recent book Pavone illuminates how lawyers cooperating with their clients profoundly impact on EU law.It is arguable whether these conclusions can be applied mutatis mutandis to constitutional referrals.See T. Pavone, The Ghostwriters (Cambridge University Press 2022).
Fourth, a reference by an ordinary court to a constitutional court can, in various jurisdictions, take the form of either a question (the question model) or a motion (the motion model).The former is the European Court of Justice's modelcourts may (or shall) ask the European Court of Justice questions concerning the interpretation or validity of a law. 40In the realm of constitutional referrals, that applies in a similar fashion.Ordinary judge do not have to be convinced that a law in question is unconstitutional; it suffices that they seriously doubt its constitutionality, and a constitutional court ultimately clarifies the matter.Italy and France, for example, employ this model. 41On the other hand, in a less frequently used model, 42 which we can observe in Germany, Czechia or Slovakia, 43 for instance, a court submits a motion detailing reasons why it deems a pertinent law unconstitutional.Thus, in the motion model, a court 'concludes'and does not 'doubt' (leastways from a formal point of view)that a law is unconstitutional.
Fifth, following up on the previous aspect, reference may be compulsory or optional.For example, in Belgium, Germany and Czechia, if certain conditions are met the ordinary court shall refer. 44Such conditions may appear as either substantive regarding the interpretation of a law in question (ex officio) or procedural, usually because of a request by the parties. 45Although an ordinary court itself assesses the fulfilment of the necessary criteria, this setting, at least formally, does not permit discretion and thus can be referred to as obligatory.Failure to refer in such circumstances amounts to a violation of the law (the constitution).On the other hand, for instance, in Spain and France, 46 even if the specific criteria are met, the ordinary court only may refer the case to the constitutional court.In other words, the constitution lays down a margin of appreciation for ordinary judgesit is up to them.Although the line between the two is sometimes fluid, the discretionary model versus the compulsory model can be another typology.Finally, the term 'constitutional' indicates a benchmark for review.Preliminary references can only be made concerning the potential unconstitutionality of statutes or their provisions which are to be applied, not for assistance to deal with any constitutional issue that ordinary courts encounter. 47But what does 'unconstitutionality' mean?In the majority of countries, the grounds for initiating the preliminary reference procedure correspond to those for abstract review and include an entire constitutional order as such. 48Nevertheless, in France, for example, the frame of reference consists only of 'rights and liberties guaranteed by the Constitution'. 49The concrete review instituted by the preliminary ruling mechanism is thus narrower than the abstract one. 50In any case, a constitutional court has the final say regarding ambiguous questions about standards for reference (and review).
These six essential features of the constitutional referrals mechanism do not just perform an introductory functionit pays to keep them in mind, as these features permeate the rest of the article.They help us better to grasp the mechanism's functions as well as the specific mechanism in the single case study.

F         
The mechanism of constitutional referrals can serve dozens of functions and there is little point in trying to list them all.However, this section strives to tackle at least some of the most important functions in respect judicial dialogue.First, though, it is necessary to explain an approach to this issue.
To analyse and classify the functions of constitutional referrals more systematically, I employ Barber's approach to the separation of powers, 51 recently elaborated by Kavanagh. 52They both distinguish negative and positive approaches to the separation of powers (emanating from so-called negative and positive constitutionalism).The negative approach focuses on building friction and seeks to keep the powers that be constrained by each other.On the other hand, the positive account emphasises the efficiency of exercising powereach branch has its specific expertise, skills and virtues and is designed to utilise these virtues for the wellbeing of the state as such. 53For example, the judicial branch exists to resolve legal disputes and to interpret andwhere neededto adjust the law. 54Through these functions and thanks to their legal expertise, skills and a certain immunity from electoral pressures, courts specifically contribute to the efficient state advancing the wellbeing of its members. 55s the constitutional referrals procedure represents a subcategory of judicial review, it amounts to a tool of the judiciary vis-à-vis the legislature as well as the executive. 56Thus, it is worthwhile viewing this mechanism from the separation of powers perspective.As explained above, being well aware of the special place of constitutional courts within European constitutionalism, I consider constitutional courts to be part of the judicial branch in a broader sense, and regard the judiciary as bifurcated into 'constitutional' and 'regular' arms. 57rdinary and constitutional courts are partners in a collaborative enterprise in promoting, tempering and defending a constitutional order. 58rawing on that, I divide the functions of the mechanism of constitutional referrals into negative and positive ones.Negative functions encompass: first, ensuring that the legislature does not overstep constitutional boundaries; and, second, protecting the rights of individuals in specific cases.This overlaps with two of de Visser's four principal purposes of constitutional adjudication. 59n the other hand, the positive functions comprise: first, ensuring the uniform interpretation of the constitution; second, securing the internal coherence of the legal order of the state; and, third, integrating ordinary courts into constitutional dialogue. 60Now I elaborate on both categories in more detail.

Negative functions of constitutional referrals: the sword against the other branches of power
The negative functions embrace: (1) ensuring that the legislature does not overstep constitutional boundaries; and (2) protecting the rights of individuals in specific 54 Barber  cases.The latter function is crucialwell-entrenched right-based theories maintain that efficacious protection of fundamental rights justify, to a great degree, the very existence of judicial review. 61Even if numerous countries enable individuals to challenge a law (to be applied to their case) before a constitutional court via the constitutional complaint procedure, access to a constitutional court usually requires that remedies provided by an ordinary judiciary have been exhausted. 62Accordingly, an ordinary court's reference to the constitutional court is able to hasten the protection of an individual's rights and facilitate access to justice.Moreover, as mentioned above, a concrete case can occasionally reveal a constitutional flaw that might not be apparent in an abstract review, since such a flaw fully manifests itself only during the application of a respective norm.
On the other hand, the first functionkeeping the legislature in checkis the pivotal objective of abstract judicial review. 63Nonetheless, I will explain the considerable importance of this function in respect of constitutional referrals as well.As has been outlined, the negative approach to the separation of powers stresses instruments capable of curbing abuse of power and thwarting looming arbitrariness. 64In this vein, each institution has its swords (sanctions or threats that one institution can actively use against another) and shields (immunities that serve to protect institutions from each other). 65Constitutional review embodies a sword which, to paraphrase Madison, keeps the legislature 'in its proper place'. 66owever, abstract review usually requires activity of a different branch to be launched and concrete review through a constitutional complaint depends on an action by individuals.Only concrete review via constitutional referrals proceedings is a type of constitutional review exclusively assigned to the judiciary.Indeed, it initially needs a case to have been brought by individuals, but the procedural framework, which is usually open to all, gives rise to a large number of cases and also allows for a sort of strategic litigation. 67onstitutional referrals can thus be viewed as an institutional self-defence mechanism. 68 resistance, by which I mean 'set of techniques, tools and practices which courts or individual judges can use to prevent, avert, stay or punish imminent political attacks on judicial independence'. 69sing this sword, the judiciary (the constitutional and ordinary judiciary acting together) may autonomously (vis-à-vis the legislature and the executive) defy legislative acts which, for instance, deprive courts of some powers or otherwise interfere with their independence.Yet, in addition to the judiciary being able to defend itself, using constitutional referrals it might also counter other acts jeopardising constitutional democracy.
Therefore, I deem this mechanism to be one of those that are, at least potentially, able to postpone or even resist the rule of law and democratic decay. 70tilising their respective specific virtues, the constitutional and ordinary judiciaries in collaboration can act as (at least) a 'useful speed bump or stop sign' against authoritarian initiatives. 71As they are closer to the everyday problems of individuals, ordinary courts are better positioned than constitutional courts to identify swiftly threats that are minor at first sight but that can have far-reaching damaging effects on the rule of law and democracy. 72But, most importantly, they wield the 'activate button' to launch the review before the constitutional court.Yet, they lack the power to strike down the law and their rulings often have only limited effects (especially when we look at the lower courts).On the other hand, constitutional courts have the power to quash the law, the potential to have a far-reaching impact because of high media attention and the legitimacy to counter democratic dysfunctions. 73But they need their judicial fellows (ordinary courts/judges) to flag up the issue and, notably, to trigger the whole review procedure.
In jurisdictions without constitutional referrals by ordinary courts, the ordinary judiciary is usually entitled to strike downor, at least, not applythe unconstitutional law in question.In the model with diffused judicial review, constitutional issues are raised and adjudicated on during ordinary litigation and there is no procedure specifically designed to protect constitutional principles. 74hus, individual constitutional protection may come sooner and more easily, but large-scale systemic effects (lying at the core of this negative function) arrive much later, when the case filters through the net to the apex courts (or often to the 69 Šipulová, supra n. 10, at p. 159

Constitutional Referrals by Ordinary Courts
Supreme Court).That is also the case with mixed systems (a mix of diffused and centralised review) without constitutional referrals. 75n the other hand, the relevance of the constitutional referrals mechanism is growing in countries where individual constitutional complaint is lacking, and the only way to activate the constitutional court from the bottom is via referral by ordinary courts (see above, mainly the Italian and French model of constitutional review).However, in systems where individuals also have access to constitutional review, the way via constitutional referrals is more straightforward and can speed up the countering of democratic dysfunction.Moreover, constitutional referrals can be seenalongside the general protection of constitutional valuesparticularly as an exclusive tool of the judiciary in its resilience, that is, in protecting particular values linked with courts (primarily judicial independence).
This negative feature is akin to the way Polish and Hungarian judges at the supranational level have resisted pressure from their (former) populist governments.They have submitted external references to the European Court of Justice or individually to the European Court of Human Rights. 76If the constitutional court in question has not yet been captured (as it has been in the case of Poland and Hungary), 77 the constitutional referrals mechanism poses a similar framework for resistance, but at the domestic level.Certainlysimilarly to other tools of judicial resiliencecourts alone cannot save democracy.Yet, they can play a crucial role in hindering erosive processes. 78onetheless, there is a threat that the judiciary will abuse this sword to advance its own interests, e.g. by defying various accountability mechanisms or never allowing its salaries to be cut.Hence, this tool should be utilised prudently and rarely.Here, it holds true more than anywhere else that judicial review must be responsivecautious of the degree of legal support and political justification there is for the ruling. 79Failure to be so risks undermining public trust in the judiciary, resentment by a society which views judges as an untouchable elite group, and thus prepares a breeding ground for politicians aiming to capture the judiciary. 80I will return to this issue again at the end of the case study.

Positive functions of constitutional referrals: consistency and integration
Through their positive functions courts live up to their role in the separation of powers as they use their expertise in the interpretation of the law and power to elucidate authoritatively what the law actually is. 81They are well placed to deal with constitutional uncertainties and rectify some deficiencies which might otherwise lead to legal turmoil. 82Accordingly, via constitutional referrals, the judiciary can: (1) secure the consistent interpretation of the constitution; (2) prevent contradictions in the legal system; and, finally, (3) publicly deliberate on constitutional issues.
The first function targets the future and represents a manifestation of a prospective effect of judicial decision-making.A constitutional court assesses the constitutionality of a challenged statute (its provision) and sets the record straighteither striking down the statute (its provision) or providing a guideline for a constitutionally conforming statutory interpretation.Constitutional courts are vested with a monopoly on these matters (contrary to the diffuse model of judicial review) and have more time to engage in deep contemplation of the fundamental issues involved. 83Therefore, it should be guaranteed that they deliver uniform and sufficiently reasoned interpretations of the constitution capable of becoming widely accepted.That fosters legal stability and certainty and is one of the underlying principles of the rule of law. 84his function is further amplified in systems with the question model, where the fact that the ordinary court doubts the constitutionality of a law suffices for it to submit a case to the constitutional court. 85By not requiring an ordinary court's assurance that a pertinent law is unconstitutional (the motion model), the way to a constitutional court is more permeable, which should enable a constitutional court to pronounce on more ambiguous issues.The limits of this function lie in the fact that a precondition for a ruling of a constitutional court is a decision by an ordinary court to refer.If an ordinary court chooses to resolve the question of its own accord (notwithstanding whether it is in accordance with the law or not), it will deprive a constitutional court, at least temporarily, of the opportunity to express its final opinion.In that case, ordinary judges must be careful that they do not alter the division of labour that the system is based on. 86nsuring the uniform interpretation of the constitution is closely linked with the second functionsecuring the internal coherence of the state's legal order.Also, this function reinforces the stability of the law and legal certainty.Since ordinary courts have to apply a statute unless the constitutional court declares it unconstitutional, that prevents (at least in the long run) the situation where a particular law is applied by one court and not applied as unconstitutional by another. 87Indeed, that is not the case absolutely, as a temporary inconsistency among ordinary courts can occur while the referred case is pending before the constitutional court.Nonetheless, it is a small price to pay for the fact that an ordinary court is not compelled to applyfrom its point of viewan unconstitutional provision.By striking a justifiable balance between this limited autonomy of the ordinary courts and the centralisation of a power to quash laws, the coherence of the legal order is undoubtedly better safeguarded by use of the mechanism of constitutional referrals than without it.
Finally, the significant function of the mechanism of constitutional referrals is its capability to integrate the ordinary courts into constitutional dialogue.Even though ordinary judges' day-to-day business consists of the interpretation and application of non-constitutional law, they are at the front line in dealing with the norms and principles emanating from the constitution.This is, according to Sadurski, 'the standard rationale' for granting the power of concrete judicial review to ordinary courts. 88Thus, the procedure of constitutional referrals puts ordinary courts in dialogue with constitutional courts and allows them to pronounce on a given constitutional issue.This will be thoroughly investigated in the following section.

Constitutional referrals as a platform for judicial dialogue
Constitutional referrals provide a unique procedural framework within which ordinary courts communicate directly with a constitutional court.Communication in which a court stands on at least one side is referred to as judicial dialogue. 89ince dialogue through constitutional referrals takes place between two courts in a special hierarchical relationship within one state, in Rosas' categorisation, it can be classified as a domestic vertical dialogue. 90lthough the subject matter of such dialogue may encompass various legal questions, the crux of the dispute revolves around the issue of constitutionality.Therefore, dialogue through this mechanism might be seen as a type of constitutional dialogue. 91This notion has gained currency in North American scholarship to portray interactions between political institutions and the constitutional judiciary, 92 and, in time, it has travelled across the ocean.In accordance with Groussot, I use the term to include interactions specifically within the judicial branchbetween constitutional courts and ordinary courts. 93evertheless, constitutional referrals virtually create a dialogic triangle between (1) ordinary courts, certifying the matter to the constitutional court, (2) the legislature, whose acts are subjected to constitutional scrutiny; and (3) the constitutional court itself. 94Each vertex of such triangle is given the opportunity to pronounce on the constitutional matter in question, each in accordance with its role in the separation of powers.As Alec Stone-Sweet aptly remarks, by ongoing interactions of all these three actors, the construction of constitutional law is driven. 95hus, the constitutional referrals are part of discursive constitutionalism, 96 of a 'continuing colloquy' among the judiciary, political institutions and society at large on the requisites and demands of constitutional democracy. 97As Ferreres Comella points out, the 2008 reform in France, which established the constitutional referrals procedure (question prioritaire de constitutionalité), 98 has facilitated the emergence of richer constitutional conversations. 99Driven mainly by the so-called Melki saga, this 'legal revolution promising to bring human rights back to its homeland' prompted the French debate (not only) on judicial and constitutional politics.Since the centralised model of judicial review otherwise furnishes ordinary courts with only limited opportunities to pronounce on these matters and to communicate directly with a constitutional court, 101 judicial dialogue on constitutional matters represents an important function of constitutional referrals.In European constitutionalism, the role of constitutional courts in constitutional interpretation 'tends to be closed and exclusive'. 102That said, thanks to its unique institutional design straddling the political and legal spheres, constitutional courts wield great deliberative potential. 103Within this procedure, ordinary and constitutional courts can talk directly to each other about constitutional issues, while this talk is formalised, public and thus approachable for at least a professional audience.It is an institutionalised legal discourse 104 on fundamental values, principles and rights within the judicial branch of power.To underline it in Habermas' spirit, such public discourse constitutes a procedural ground of liberal democracy. 105rdinary courts differ from constitutional courts in many respectsthey encounter problems of 'laws in action' on a daily basis and are closer to the commonplace issues of individuals.On the other hand, a constitutional court is able to bring a broader perspective as well as profound expertise to constitutional law and human rights.Although they each have their pros and cons, their dialogue merges their virtues together.This might be described as 'cross-fertilisation' within the domestic legal order. 106By exchanging the arguments of actors from distinct contexts, the dialogue via constitutional referrals is able to promote better-reasoned judgments and thus rationalise legal outcomes. 107That is crucial, as communication is an essential weapon in a court's arsenal; lacking a purse and a sword that is actually robust (vis-à-vis other branches), the judiciary relies on persuasive legal argument. 108herefore, the difference between ordinary courts and a constitutional court is a strong case forand not againsttheir conversation on constitutional matters.
In this fashion, ordinary courts and a constitutional court are collaborators in upholding constitutionality. 109The review of legislation via constitutional referrals is carried out on two levels -first, by a preliminary check by an ordinary court and, second, by a constitutional court with its final and binding assessment.Thus, constitutional review can be viewed as a dialogic.Even though Tushnet and Roach used the term 'dialogic judicial review' to address relations outside the judicial branch, between the judiciary and the legislature, 110 we can stay within it and develop its intra-judicial aspect.Ideally, the two courts will deliver their opinions on the constitutional issue and respond to each otherthis interplay then provides the outcome.
The opportunity of the ordinary courts to make an initial assessment and to pronounce on the matter of constitutionality represents a specific manifestation of the principle of subsidiarity, encapsulating the fact that fundamental rights must, even in a centralised model of constitutional review, first and foremost be protected by lower-level courts. 111Apart from accelerating the protection of individuals' rights, such setting might contribute to the 'constitutional cultivation' 112 of ordinary courts, which is important notably in jurisdictions with short experience of liberal constitutionalism. 113oreover, dialogue is, as Groussot observes, not only a means of communication but also a medium of power. 114By way of dialogue via constitutional referrals, ordinary courts can assert their view on the interpretation or validity of a law and try to persuade a constitutional court.Subsequently, lower courts can use this mechanism to 'leapfrog' the ordinary judicial hierarchy, especially if they do not agree (or anticipate that they will not agree) with the interpretation of higher ordinary courts. 115However, it could also be the other way round -Dyevre, for example, argues that the new concrete review mechanism in France empowered the Constitutional Council andat the same timeposed a threat to the influence of the Cour de cassation.
In any event, healthy constitutional dialogue requires two preconditions: mutual respect and the active engagement of both actors in communication.If a constitutional court does not respect the ordinary courts, it will disregard or even mock the arguments in the reference.And vice versa, if an ordinary court does not respect a constitutional court, it will be reluctant to submit the reference.Last but not least, if an ordinary court does not want to engage in constitutional debate, there will be no dialogue.In that case, not only will an ordinary court deprive the community as a whole of vibrant deliberation on a constitutional matter,116 but it will also fall short of its role in the separation of powers by not utilising all its available virtues.
Furthermore, it should be reiterated that the mode of dialogue also depends on the broader institutional framework in the particular state (apart from the institutional design of constitutional referrals proceedings), as well as on its political and socio-legal context.The (non-)existence of an individual constitutional complaint in particular is a gamechanger.If it is not present, constitutional referral is often the only way in which the constitutional court can pronounce on the certain issue.Therefore, its approach to the ordinary courts should be all the more respectful and open. 117o sum up, constitutional referrals offer a platform for constitutional dialogue between ordinary courts and a constitutional court.It is an important function of this mechanism, as it allows ordinary courts to pronounce on constitutional matters and directly communicate with a constitutional court.Such a judicial dialogue can rationalise and legitimise the outcome of legal disputes.However, it does not work without active courts with mutual respect.

E :     
In this section I will examine the theoretical claims using the example of Czechia, a country with centralised constitutional review, following the Germany prototype.Czechia belongs to the Central and Eastern Europe region, where many countries have experienced democratic backsliding and has a still relatively fragile democracy in a state of 'democratic careening',118 albeit spared from grave democratic excesses. 119Czechia has also undergone an intriguing evolution of the relationship between the the Czech Constitutional Court and the ordinary courts since the establishment of the new, post-Communist state in 1993.Mainly in the first decade, the Czech Constitutional Court promoted an anti-formalist type of judicial interpretation and regularly criticised the excessive textual positivism deeply embedded in the post-Communist perception of the application of laws. 120Keeping this in mind, it is no surprise that several clashes between the purposive interpretation and value-laden reasoning of the Czech Constitutional Court and the judicial formalism of the Supreme Court have arisen over time.The most famous one, when the Supreme Court refused to respect the case law of the Constitutional Court, went down in history as 'the war of the courts'. 121his term is also used to describe the skirmish between the Constitutional Court and the Supreme Administrative Court. 122he Czech Constitutional Court is entrusted with both abstract and concrete judicial review, but only ex post (with the marginal exception of international treaties assessment).An abstract assessment of constitutionality can be initiated by the President of the Republic or a group of members of Parliament, whilst concrete review might be triggered either by individuals and a Panel of the Constitutional Court in connection with the constitutional complaint procedure or by an ordinary court when it encounters a potentially unconstitutional provision in the course of deciding any case (i.e. the mechanism of constitutional referrals). 123Thus, abstract and both objective and subjective concrete review exist in Czechia.
The key provision of the Czech Constitution stipulating the constitutional referrals procedure is Article 95(2), which states: 'Should a court reach the conclusion that a statute which should be applied in the resolution of a matter is not compatible with the constitutional order, it shall submit the matter to the Constitutional Court.' 124 Using the typology outlined in the first chapter, it is (1) the model of unlimited referrers, as all ordinary courts can refer; (2) the motion model, since ordinary courts have to 'reach the conclusion'and not just have 'doubts' (leastways from a formal point of view)that a law is unconstitutional; (3) the compulsory model because, if certain conditions are met, the ordinary court shall refer.Even though there is formally no discretion, it is up to the ordinary court to determine whether it finds that the conditions for referencei.e. the conflict of a statute to be applied with the constitutional orderhave been met.
In my empirical study I seek to explore whether this mechanism truly serves as a platform for dialogue on constitutional issues or is instead just a series of utterances by two actors who do not listen to each other.Beyond the very existence of dialogue, I aim to reveal some nuances of interaction via this mechanism related to the positive and negative functions explained above.To do so, I have analysed all the submissions by the ordinary courts and the subsequent rulings of the Czech Constitutional Court 125 issued during the period from January 2014 to December 2022.These amount to 70 submissions by ordinary courts and 70 subsequent rulings by the Constitutional Court.My study thus covers the so-called 'Third Constitutional Court' (also labelled as Zeman's Constitutional Court), the era of the Czech Constitutional Court with Justices appointed by the former President, Miloš Zeman. 126This era can be referred to as the first 'normal' one, that is to say, dealing with quite ordinary questions of an established constitutional democracy as opposed to controversial transitional issues (The First Constitutional Court) or sovereignty issues linked to access to the EU (The Second Constitutional Court).This sample thus allows me to isolate the most significant outlier variables.Moreover, this era is interesting, given the questions raised in the previous section, as it overlaps with the rise of the populist leader, Andrej Babiš, the Prime Minister from 2017 to 2021 and very influential Minister of Finance from 2014 to 2017. 127efore delving into the analysis itself, a brief data appetizer on constitutional referrals in Czechia is in order.The vast majority of the Czech Constitutional Court's docket consists of individual constitutional complaints (around 95 per cent).It is not surprising that constitutional review linked with individual complaint is the most common type of judicial review before the Constitutional Court.However, among all 'qualified' petitioners empowered to propose the annulment of the statute (i.e.not individuals in connection with their constitutional complaint), ordinary courts are the most frequent and most successful petitioners.From the establishment of Czechia in 1993 to the end of 125 Since I focus on the content of judicial opinions, I exclude rulings without any substantive argument.Therefore, the dataset encompasses only judgments on merits (nálezy), quasi-substantive decisions (usnesení) that dismiss a petition as manifestly ill-founded, and decisions that dismiss a petition because the court was not authorised to refer.For an explanation of procedure before the Czech Constitutional Court, see Kosař and Vyhnánek, supra n. 12, at p. 159.
126 Justices of the Czech Constitutional Court are appointed by the President of the Republic for 10 years (reappointment is not prohibited, but the new president, Petr Pavel, has declared that no one will be reappointed).All of the President's nominees must be approved by a simple majority in the Senate (the upper house of Parliament). 127For more details of the Czech Constitutional Court context, see Smekal et al., supra n. 10.
2022, the Constitutional Court delivered 334 decisions within the constitutional referrals procedure and the number of submissions decreases year on year. 128ow, let us move to the analysis.It is two-tieredit has macro and micro levels. 129cro level: constitutional dialogue in the majority of cases At the macro-level I conducted a content analysisall judicial opinions were hand-coded based on the codebook attached in the Appendix (online).As to the submissions of ordinary courts, I focused on the constitutional argument of ordinary courts and looked into whether or not the ordinary courts in their reasoning argued by constitutional principles, referred to the case law of the Czech Constitutional Court or the European Court of Human Rights or to the constitutional scholarship.This stems from the assumption that greater contemplation on constitutional matters requires more than merely working with the text of the constitution itself.On the other hand, regarding the rulings of the Czech Constitutional Court, I examined whether or not the Court had responded to the constitutional arguments of the ordinary court, i.e. whether it dealt with these arguments, took them into consideration and provided answers to them: in other words, whether the Court entered the dialogue.For purposes of this article, I label this a discursive style of reasoning. 130fter analysing the data through descriptive statistics (frequencies and crosstabs), I found that the ordinary courts argued on constitutional principles such as the separation of powers, equality, judicial independence or proportionality, in 63 per cent of cases.They referred to the case law of the Czech Constitutional Court or the European Court of Human Rights in 70 per cent of cases and to scholarly writing addressing the pertinent constitutional issue in only 17 per cent of cases.Leaving aside the literature references, generally not very frequent in judicial reasonings in Czechia, 131 the constitutional principles argument and references to the case law appear together in more than half of all submissions, and in only 17 per cent of cases do we find neither of them.On the other hand, the discursive style of reasoning (i.e. the explicit responsiveness of the Czech Constitutional Court to the arguments of the ordinary courts) is present in 64 per cent of all that Court's rulings, in 71 per cent of those that followed submission with constitutional argument.However, the results suggest that there is no strong link between constitutional argumentation by ordinary courts and a likelihood of discursivity at the Czech Constitutional Court.In other words, it does not appear from the data that arguing constitutional principles, referring to the case law or to constitutional scholarship, would increase the chances of ordinary courts receiving an explicit response to their arguments in their submissions.The Czech Constitutional Court is sometimes discursive even when deeper constitutional law arguments are lackingand vice versa.
Indeed, an individual reference to the constitutional principles, case law or literature in argument does not constitute dialogue and, likewise, a discursive style of reasoning does not necessarily indicate a meaningful conversation on constitutional matters.However, incorporating all of these elements together, the results provide a valid and reliable picture of the presence or absence of constitutional dialogue in the constitutional referrals procedure.Thus, in the majority of cases, a dialogue on constitutional matters between ordinary courts and the Czech Constitutional Court can be discerned.The data show that constitutional referrals often serve as a platform for inter-judicial constitutional dialogue, as opposed to just a series of unresponsive monologues.

Micro level: disrespectful dialogue and the judicial sword
At the micro level, I employ an in-depth qualitative analysis of these cases (contextual interpretation, idiographic study). 132From the dataset on the macro level, I have singled out three most intriguing cases, given the functions of constitutional referrals presented above.By conducting a nuanced analysis of these cases I uncovered some details and peculiarities hidden at the macro level. 133n particular, the macro-level analysis was not able to capture the quality and manner of interaction.Dialogue can be healthy, courteous, constructive and mutually enriching, but also disrespectful, condescending and even toxic. 134 constructive inter-judicial dialogue on constitutional matters was evident in the majority of cases in which ordinary courts presented sound constitutional reasoning and the Czech Constitutional Court explicitly responded to their arguments. 135Notwithstanding the formal result of the case, the mechanism of constitutional referrals in such circumstances fulfils its functions, the ordinary courts are integrated into constitutional issues, 136 the Czech Constitutional Court lives up to its role, and a professional audience at least can observe this institutionalised legal discourse. 137evertheless, one of the preconditions for healthy dialoguemutual respectis not always present or, maybe more accurately, is not visible to outside observers.An example is the 2020 case of the Supreme Administrative Court's submission, which seeks to annul the provision of the Code of Administrative Justice 138 setting the expiry of the time limit for bringing an action for an administrative authority's failure to act. 139Although the Czech Constitutional Court responded to the principal arguments of the petitioner, it did so in a condescending manner, rejecting the arguments of the Supreme Administrative Court as completely out of touch. 140For instance, the Constitutional Court stated that 'to accept the argument put forward by the Supreme Administrative Court would deny the very essence of the functioning of the judiciary in the Czech Republic'. 141Subsequently, the Constitutional Court derisively concluded that 'the Supreme Administrative Court, after twenty years [of validity of the impugned provision], out of the blue realised that the provision is unconstitutional'. 142The mere fact that three dissenting Justices of the the Constitutional Court agreed with the Supreme Administrative Court implies that the reasons provided by the petitioner were not completely pointless, as the Constitutional Court's contemptuous tone suggests.enriching exchange of opinions and ideas between two partners are sidelined, whilst hierarchy and conflict take the lead. 150Instead of integrating ordinary courts into contemplation on constitutional matters, the Constitutional Court in these cases stressed its exclusive superiority in constitutional interpretation and treated the ordinary courts as pupils who dared to propose such a thing.In other words, the Constitutional Court used the judicial dialogue as a medium of power. 151This represents a switch from interpretative constitutional pluralism to interpretative constitutional monism. 152uch an approach contributes nothing positive to the legal order.On the contrary, it can even discourage ordinary courts from using constitutional referrals.This approach might explain the downward trend in the number of constitutional referrals in Czechia over the last decade. 153In terma of repercussions, not only does it deprive the interested part of society of the public institutionalised deliberation on a constitutional matter, 154 but it also frustrates both other positive and negative functions of constitutional referrals.The potential to confine the legislature, protect human rights, remove contradictions from the legal order or consistently interpret the constitutionall will be, in the best-case scenario, delayed. 155he second case further displays an example of how constitutional referrals can serve as strategic litigation by judges.Judges are a specific group of litigantsalmost 6 per cent of analysed cases were brought before an ordinary court by judges.Judges attempted to raise their subjective right to judicial independence, and through this vehicle achieve a derogation from certain pieces of undesired legislation (on retirement age, salaries, disciplinary proceedings). 156The following, third, example clearly demonstrates this phenomenon.
In 2014, the Czech Constitutional Court again ruled in in the endless saga of judicial salaries. 157In this case, a judge of the Regional Court claimed for payment of the difference between his entitlement to salary before and after the amendment by which Parliament reduced judges' salaries as part of an austerity package responding to the effects of the economic crisis.Before the Municipal Court, the plaintiff (the judge) proposed that the case be referred to the Constitutional Court in order to annul the amendment.And the Municipal Court did so.
As hinted, it was not the first time the Czech Constitutional Court had dealt with cuts in judges' pay. 158The Constitutional Court has consistently reiterated that the independence of the judiciary ranks among fundamental principles of the Czech legal order, and the reduction in judicial salaries is per se a violation of such independence, as it could be abused as a form of governmental or legislative pressure upon the judicial branch. 159Unsurprisingly, judges are aware of this case law and know that the Constitutional Court has so far always sided with them. 160his was also the case here.The Constitutional Court, with reference to the previous case law, struck down the contested amendment. 161he judiciary thus achieved the annulment of the unwanted law without any outside involvement.A judge initiated the dispute, the first instance court created a petition, and the Czech Constitutional Court accomplished the work.That shatters traditional narratives that courts will not hear cases unless someone outside the judiciary wishes to initiate the proceedings and bring the case to court. 162Since judges do not have privileged status to challenge laws in abstract review procedure in Czechia, it turned constitutional referrals to the quasiabstract judicial review.Therefore, it works as a perfect institutional self-defence mechanism of the judiciary. 163The judiciary holds a sword that might be used but is alsoas is nearly everythingmisused vis-à-vis the legislative branch.As noted above, we can view constitutional referrals as a mechanism for defending against an amendment that threatens the rule of law at the onset of a surge of authoritarian populism. 164In other words, as a toolkit of judicial resistance in relation to abusive constitutionalism and democratic decay. 165onetheless, as stated above, there is a threat that the judiciary will abuse this sword to advance its own interests, e.g. by never allowing its salaries to be cut.And since even the road to hell is paved with good intentions, this tool should be utilised prudently and rarely.Failure to do so risks undermining public trust in the judiciary and fostering resentment in society, creating the sense that judges are an untouchable elite group, which thus becomes a breeding ground for politicians aiming to capture the judiciary. 166urthermore, when constitutional courts are captured by political forces, as has been the case in countries like Hungary and Poland, 167 the constitutional referrals process becomes futile for the purposes outlined above, leaving recourse to supranational courts as the only viable alternative. 168In contrast, in such an environment constitutional referrals can be abused as another tool of the ruling political majority.For instance, if a populist government wants to repeal a certain piece of legislation without overtly doing so itself, it can use constitutional referral submitted by a loyal ordinary-court judge.In doing so, it delegates this unpopular decision to formally independent bodies. 169In a similar vein, judges who have deferred to the interests of a populist government can abuse constitutional referrals as an instrument for so-called judicial populism. 170

C:   
This article has unpacked the mechanism of constitutional referrals by ordinary courts.Building upon the conceptual analysis, I have sketched the typology of constitutional referrals as: (1) the question and the motion model; (2) the discretionary model and the compulsory model; or (3) the model of unlimited referrers (all ordinary courts) and the model with limited referrers.Furthermore, using Barber's approach to the separation of powers, I have identified several negative and positive functions of constitutional referrals.I have presented two main argumentsthat constitutional referrals can serve as: (1) a platform for inter-judicial dialogue on constitutional matters; and (2) an exclusive tool of the judicial branch against the legislature.
All of the identified functions demonstrate the uniqueness of constitutional referrals, and the necessity of having healthy relationships between constitutional actors in any well-functioning pluralistic democracy governed under the rule of already been captured by political authorities (e.g.Hungary and Poland),178 it does not fulfil this purpose and the only option is to refer to supranational courts. 179On the other hand, in this context the mechanism can serve as another tool of the ruling power or as an instrument for so-called judicial populism. 180espite the exploratory nature of the empirical study, it sheds light on how courts work with this mechanism.Moreover, the theoretical part provides the initial conceptualisation of constitutional referrals and the novel framing of it through the lense of Barber's theory of the separation of powers.Therefore, the article represents the first but crucial step in a more thorough understanding of constitutional referrals.Drawing on the findings of this article, future research may focus on individual judges and their motivations to utiliseor, on the other hand, disincentives to avoidthis mechanism.A comparative study of multiple jurisdictions, using the conceptual framework presented, might be another way to proceed with follow-up research.Either way, the constitutional referrals pose a unique framework for institutionalised legal discourse and should be the focus of constitutional scholarship.Supplementary material.To view supplementary material for this article, please visit https://doi.org/10.1017/S1574019624000087

60
Cf, mutatis mutandis, G. Tridimas and T. Tridimas, 'National Courts and the European Court of Justice: A Public Choice Analysis of the Preliminary Reference Procedure', 24 International Review of Law and Economics (2004) p. 125 at p. 126.
and, on the other hand, the 12 For more detail see D. Kosař and L. Vyhnánek, 'The Constitutional Court of Czechia', in A. von Bogdandy et al. (eds.),Constitutional Adjudication: Institutions (Oxford University Press 2020) p. 119 at p. 136-174; Z. Kühn, 'The Constitutional Court of the Czech Republic', in A. Jakab et al. (eds.),Comparative Constitutional Reasoning (Cambridge University Press 2017) p. 199 at p. 205; Z. Kühn, The Judiciary in Central and Eastern Europe: Mechanical Jurisprudence in Transformation?(Martinus Nijhoff 2011); and D. Kosař, Perils of Judicial Self-Government in Transitional Societies (Cambridge University Press 2016) p. 117.
, supra n. 51, at p. 61; cf M.M. Shapiro, Courts: A Comparative and Political Analysis (University of Chicago Press 1986) Ch. 1. supra n. 51, at p. 241; for more details see also N. Barber, The Constitutional State (Oxford University Press 2012) Ch. 2. The Three Branches: A Comparative Model of Separation of Powers (Oxford University Press 2013) at p. 127.See de Visser, supra n. 3, at p. 98.According to the author, the other two purposes of constitutional adjudications are: resolving institutional disputes among governmental organs; and ensuring the integrity of political office and related processes.
62Cf Ginsburg, supra n. 35, at p. 38.63 . See e.g.R. Dixon and D. Landau, Abusive Constitutional Borrowing (Oxford University Press 2021); and A. Sajó, Ruling by Cheating (Cambridge University Press 2021).Roznai, 'Who Will Save the Redheads?Towards an Anti-Bully Theory of Judicial Review and Protection of Democracy', 29 William & Mary Bill of Rights Journal (2021) p. 327 at p. 341.

75
But see the case of Portugal, where individuals can appeal directly to the constitutional court against any diffuse review by ordinary courts: Cortes and Violante, supra n. 9. See also the judgment of the ECtHR 23 June 2016, No. 20261/12, Baka v Hungary.Poland's Constitutional Breakdown (Oxford University Press 2019) Ch. 3; P. Castillo-Ortiz, 'The Illiberal Abuse of Constitutional Courts in Europe', 15 EuConst (2019) p. 48; T. Drinóczi and A. Bień-Kacała, Illiberal Constitutionalism in Poland and Hungary (Routledge 2022).
'Constitutional Dialogues, Pluralism and Conflicting Identities', in M. Avbelj and J. Komárek (eds.),Constitutional Pluralism in the European Union and Beyond (Hart Publishing 2012) p. 319 at p. 320.See P. Pasquino, 'The New Constitutional Adjudication in France.The Reform of the Referral to the French Constitutional Council in Light of the Italian Model', 3 Indian Journal of Constitutional Law (2009) p. 105.See A. Dyevre, 'The Melki Way: The Melki Case and Everything You Always Wanted to Know about French Politics (but Were Afraid to Ask)', in M. Claes et al. (eds.),Constitutional Conversations in Europe: Actors, Topics and Procedures (Intersentia 2012) p. 309; and M. Bossuyt and 92See de Visser, supra n. 3, at p. 329.93 X. Groussot, 98 99 Ferreres Comella, supra n. 83, at p. 59. 100