1. Introduction: Domestic judges: from ‘EU judges’ to ‘ordinary judges of international law’
Since the Simmenthal case,Footnote 1 and despite the fact that the European Court of Justice (ECJ) has never formally defined it,Footnote 2 the expression according to which the domestic judge is the ordinary judge of Community law has acquired a specific meaning. While the literature in English usually adopts the simpler formula of ‘national judges, as European Union (EU) judges’, the translation of the French version, ‘juge de droit commun’, into ‘ordinary judge’ is useful in that it is more precise than setting national judges ‘as international law judges’. However, both versions reflect the same concept. As developed through case law, the phrase has come to refer to several obligations incumbent upon the national judge in the course of his duties as the judge of EU law. It is commonly understood to mean that the judge must:
1. under certain conditions, directly apply the EU provision;
2. set aside or leave unapplied the inconsistent national provision;
3. so far as it is possible to do so, interpret the national provision in accordance with the EU one, and, finally;
4. hold the state responsible for any violation of EU law.
Although the list is not exhaustive,Footnote 3 it corresponds to what is usually referred to as the different forms of invocability of EU law by private parties before national courts.
Whereas it has acquired this rather technical meaning in EU law, the phrase ‘ordinary judge’ has a common meaning within some legal systems such as the French one. It refers to a court which has a general competence, as opposed to courts which have a specialized jurisdiction conferred upon them by specific legislation. Therefore, it is the scope of the competence and, more particularly, its general character that justifies the characterization of ‘ordinary judge’. Under EU law, the phrase quickly took on another significance – the focus is on the effectiveness of the role of the domestic judge in respect of EU law, rather than on competence. Indeed, before the Simmenthal judgment, nobody questioned the competence of the domestic judge to adjudicate disputes relating to Community law. Rather, doubts subsisted as to the powers the judge had to ensure compliance with Community law. The existence of a specialized court for matters of EU law, i.e. the ECJ, could suggest that the responsibility incumbent upon a domestic court to ensure the compliance with Community law was minor. However, the case law following Simmenthal has proven beyond doubt that this responsibility of the domestic court seen as an ordinary or even primary court of EU law, is as important as that of the ECJ. As is the case concerning EU law, it is obvious that the domestic court holds general jurisdiction to apply international law. This is all the more significant as private parties seldom have access to ‘specialized’ courts, i.e. international courts. Yet what remains unclear is the extent to which the domestic judge has a similar responsibility for ensuring the effectiveness of international law. An assessment of the practice in some jurisdictions shows that national judges are increasingly behaving like ordinary judges of international law. This evolution is then comparable to that experienced by Community law, except that for the latter, an external influence, the ECJ, imposed upon the domestic judge his status of ordinary judge of Community law. In contrast, for international law, it is domestic judges themselves who spontaneously behave as ‘worldly judges’.Footnote 4 In one case, it is a formal and compulsory method, while in the other, it is an informal and voluntary approach.
While parallels between the way EU law and international law are applied by domestic judges have already been drawn,Footnote 5 this has never been done concerning the potentialities of the phrase ‘national judges, as EU judges’. The related idea of the domestic judge being a ‘natural judge’ of international law has been raised but without drawing any parallels with EU law.Footnote 6 Yet this analogy seems useful since academics specializing in EU law have long studied the technical meaning of this principle. Whereas the evolution of EU law has, at least in its beginnings, borrowed from international law,Footnote 7 conversely, nowadays it is the evolution of international law in the domestic legal order which increasingly mirrors the relationship established between EU law and domestic law. Professor Leben wrote in 1998 that ‘Community law is successful international law’ and that it constitutes a ‘horizon’ for international law.Footnote 8
Although the concern about parallelism with the reception of EU law by national judges implies an exclusive focus on treaty law, the role played by national judges in the establishment of customary international law is such that the designation of ‘ordinary judge’ may be relevant. The notion could be extended far beyond its equivalent in relation to EU law and give rise to a variety of meanings. However, the analysis must be limited, in a comparative perspective, to the precise meaning of the expression concerning the relationship between national law and EU law. In this respect, the expression ‘ordinary judge of Community law’ was originally used solely to refer to norms which not only benefitted from the principle of the primacy of EU law over domestic law, but which also fulfilled the criteria for direct effect. Indeed, the Community provisions involved in the Simmenthal case, the first case to justify the designation of ‘ordinary judge’, were granted direct effect. Thus, they could be directly applied by national judges in preference to contrary national provisions. Not only would the contrary national provision be excluded (invocability of exclusion) but it would also be replaced by the EU norm (invocability of substitution).Footnote 9 In this regard, while these concepts are often taken as synonyms, direct effect (as direct applicability) may be considered as the consequence attached to self-executing provisions. This begs the question of the relevance of the envisaged transposition to international law when the latter does not contain as many self-executing norms as EU law. Whereas the ECJ has granted – and thereby imposed – direct effect to EU law with a notorious liberalism, the recognition by the domestic judge of the self-executing character of international treaties is subject to more caution. We could then conclude that the domestic judge can admittedly be the ordinary judge of international law, but he/she could only be so in relation to provisions endowed with direct effect. The scope of the research would then be substantially limited. Nevertheless, although the principle of direct applicability has long been considered an exception with regard to treaty law, it seems that there is a slight evolution as a result of both the nature of the norms of international law invoked and the behaviour of domestic judges, who may now more readily grant international norms direct effect than they would have done in the past. This is the first interest of this article. A second merit of the question lies in the fact that, in EU law, the theory of ordinary judge has quickly expanded through case law beyond the principle of direct effect. This is how, for instance, the ECJ has imposed on national judges the doctrine of consistent interpretation where the EU provisions at issue are not directly applicable. It is also in this legal context that the theory of the state obligation to make good any damage caused by its violation of EU law when the provisions involved are not endowed with direct effect was initially developed.
The aim of the article is to proceed from observation of French and other domestic case law in order to test the assumption that the domestic judge is an ordinary judge of international law. Even though the reception of international law by French judges will be the primary focus of this article, regular glances into other jurisdictions will widen the scope of the study and allow a preview of the direction of the way the law may evolve. Therefore, this article will establish a twofold comparison: on the one hand, it will apply in a systematic manner the different implications of the domestic judge being an ‘ordinary judge of EU law’ to international law; and, on the other hand, it will draw parallels between the way in which French courts ensure the effectiveness of treaty law and the stances adopted by other domestic courts. To that effect, the phrase ‘the domestic judge, as an ordinary judge of international law’ will be broken down into different forms of invocability of treaty law. The first and most established type of invocability is the direct effect theory which tends, in spite of the evolutions it has experienced, to run counter to the suggested thesis in that it appears as an obstacle to a greater effectiveness of international law (section 2). Other, more recent forms of invocability introduced in France and other countries tend, on the contrary, to justify the transposition to international law of the phrase stemming from the Simmenthal case (section 3).
2. Resistance to international law through the theory of direct applicability
It is the combined effect of the principles of primacy and direct effect which enabled the ECJ in the Simmenthal case to turn the domestic judge into the ordinary judge of Community law. However, whereas treaty provisions generally benefit in France from the same principle of primacy over legislative and infra-legislative norms that governs EU law, it is difficult to assert that they enjoy direct applicability to the same extent. Indeed, direct applicability appears as the main obstacle to describing the domestic judge as an ‘international law judge’. Even though several elements point to an evolution in the French judge's treatment of international law, there are still numerous signs of reluctance towards a generalization of the direct applicability of international norms which would be comparable to the generalization which occurred in respect of EU norms.
2.1. An incipient evolution
The theory of direct applicability, which originates back to an 1829 decision of the Supreme Court of the United StatesFootnote 10 and a 1928 opinion delivered by the Permanent Court of International Justice (PCIJ),Footnote 11 was established at a time where the ‘contract type’ treaties were predominant. The sole purpose of these treaties, which dealt mainly with peace and borders, was to regulate the relations between the contracting states. Accordingly, the opinion of the PCIJ establishes the principle of the non-presumption of direct applicability.Footnote 12 The direct applicability of treaties was then only an exception to the ‘well-established principle’Footnote 13 according to which treaties only regulate the rights and obligations of contracting states, excluding those of private parties. Additionally, as any other exception, this particular one was subject to restrictive conditions. It appears that the Permanent Court only conceived the possibility of direct applicability if ‘the very object of an international agreement, according to the intention of the contracting Parties, [was] the adoption by the Parties of some definite rules creating individual rights and obligations and enforceable by the national courts’.Footnote 14 Apart from the criteria of precision and comprehensiveness of the rules, namely ‘enforceable by the national courts’, which are more objective, the criterion that stands out in this definition is that which, being eminently subjective, rests upon ‘the intention of the contracting Parties’. Even though this opinion of an international court confirms the applicability of international treaties before the domestic judge, the implementation of this principle nevertheless pertains to his/her exclusive competence. Hence the direct applicability of international treaties is submitted to conditions which may vary significantly from one national judge to another. In contrast, under EU law, the criteria identified by the ECJ in the Van Gend en Loos case must be strictly applied as part of EU law. Moreover, while the ECJ undoubtedly drew its inspiration from the criteria established by the PCIJ, it did so as to better depart from them, especially with respect to the subjective criterion of intention of the parties to which it never resorted;Footnote 15 the only criteria established by the Van Gend en Loos case are the clarity, precision, and unconditionality of the provisions involved.Footnote 16
In France, the conceptualization of direct applicability is very recent because until the GISTI case in 1990, delivered under the influence of Advocate General R. Abraham,Footnote 17 domestic judges refused to interpret international treaties and thereby referred the matter to the ministry of foreign affairs. What can be observed in the first cases relating to the direct applicability of international treaties, and, more precisely, in the adjoining opinions of the advocates general, is a reference not only to the objective criteria of precision and comprehensiveness of the provision, but also to the subjective criterion relating to the addressee of the provision. Thus, in the GISTI case of 1997, although the Advocate General R. Abraham indirectly refers to the case law of the ECJ by using the expression ‘direct effect’ instead of the more traditional ‘direct applicability’, he first proceeds to an examination of the subjective criterion of the addressee of the provision.Footnote 18 One can nonetheless already discern an intention to draw a parallel with the criteria applicable to Community law, suggesting that the more restrictive approach endorsed by the PCIJ could be abandoned in favour of a conception more favourable to international law, inspired by the treatment reserved for EU law in domestic law.
In his opinion on the GISTI case of 1997, Advocate General R. Abraham declared, in what seems to be a wish rather than an observation, that the Conseil d’État is more and more open to direct effect.Footnote 19 In an integrationist thrust, he even went as far as to assert that French law, which he categorically describes as monist, had opted for a presumption of direct effect to which specific exceptions were ascribed that should be interpreted strictly. One may see in this assertion an intentional departure from the reasoning held by the PCIJ. Although such an understanding surely confirms the hypothesis according to which the domestic judge is the ordinary judge of international law just as much as he is the ordinary judge of EU law, it does, however, not seem to be supported by the practice of French courts, which still display caution, or even reluctance, to recognize direct effect. They have nevertheless irreversibly initiated an evolution of the theory of direct effect that will hopefully continue.
Indeed, the case GISTI-FAPIL delivered by the Conseil d’État on 11 April 2012 has generally been welcomed as an improvement by the administrative court in its openness to international law and its application of the theory of direct effect.Footnote 20 This case clarified for the first time the criteria to be applied by an administrative judge in order to determine whether a treaty provision is directly applicable. In the past, French courts would generally only conclude in a laconic manner that such or such provision of a treaty was or was not directly applicable. It was however clear from the pre-2012 case law that the administrative judge often resorted to the wording of the provisions, relying on phrases such as ‘the contracting states undertake to . . .’, to conclude that a treaty only created rights and obligations to states. While an isolated case from 2008 implicitly initiated a new trend,Footnote 21 the supreme administrative court incorporated it in a recital of the GISTI-FAPIL case. Indeed, after laying down the two relevant criteria for the determination of the direct effect of a treaty provision, namely that the stipulation must not have ‘for sole purpose to regulate the relations between [s]tates’ (i) and that it must be unconditional (ii), the Conseil d’État specifies that the absence of these conditions ‘should not be implied from the sole fact that the provision defines the contracting [s]tates as the recipients of the obligations it creates’.Footnote 22 In other words, it excludes the possibility that the wording of the provisions could play a role in the identification of these two criteria, which must be examined only in light of ‘the intention expressed by the parties and the general spirit of the treaty invoked, as well as its content and terms’.Footnote 23
With regard to the French ‘judicial’ judge – as opposed to the administrative one – after showing a clear reluctance between 1993 and 2005 to directly apply the Convention on the Rights of the Child, and even though he/she has not yet clarified his/her own criteria, it should however be noted that he/she has also initiated an evolution in favour of the theory of direct effect. It is indeed particularly tempting to consider the ‘judicial judge’ as the ordinary judge of international law when the Cour de cassation automatically raises the inconsistency of a national rule with a treaty provision as vague as the ‘right to work’ provided for by Article 6 of the International Covenant on Economic, Social, and Cultural Rights, to which it has nonetheless granted direct effect.Footnote 24 In doing so, the supreme judge seems to have swept aside the allegedly programmatic feature of this international convention, whilst ignoring the drafting options of the Covenant which may give the impression that member states are the only addressees.
While this trend may be accounted for by the change in the nature of international law under which the individual has an increasingly central position, it cannot be denied that there is also a change in the attitude of the domestic judge. Indeed, in relation to the Convention on the Rights of the Child, only this ‘psychological’ evolution of judges explains that the same provisions had been deprived of direct effect for years before the emergence of an undeniable shift in two 2005 cases.Footnote 25 In French law, the European Convention on Human Rights (ECHR) is a unique case since most of its provisions are now endowed with direct effect. This is more than enough to allocate to the domestic judge the label of ‘ordinary judge of the European Convention of Human Rights’.Footnote 26 Is it possible to extend this designation to the context of international law generally in spite of the restricted number of treaty provisions which are granted with direct effect? The number of instruments containing directly applicable provisions according to French courts remains limited. While most examples of direct applicability are drawn from multilateral treaties dealing with human rights such as the Convention on the Rights of the Child, the Covenant on Economic, Social and Cultural Rights or ILO conventions, it is often overlooked that provisions of many bilateral agreements, and particularly those of tax and establishment conventions, are endowed with direct applicability. One may highlight the privileged position enjoyed by the latter category of treaties, the violation of which must be raised by the courts on their own motion.Footnote 27
2.2. An ongoing reluctance
Despite the privileged status of some economic conventions and significant improvements regarding human rights treaties, direct effect of international norms clearly remains the exception rather than the rule. A study undertaken by the Conseil d’État emphasizes that ‘international norms creating rights or obligations for private parties . . . remain rare, because only a few provisions of conventional international law are of a directly applicable nature’.Footnote 28 Even though the clarification made for the first time in 2012 about the criteria of direct applicability was an important step forward, there remains a doubt as to how liberal a use the judge may make of this methodology in the future. These criteria may be interpreted very loosely in abstracto, but one must be cautious and wait to see how they will be applied in practice by the administrative judge. The new methodology has given rise to some expectations regarding certain international treaties such as the European Social Charter or the Aarhus Convention which are often set aside for lack of direct effect,Footnote 29 but it quickly appeared that the change in vocabulary in the GISTI-FAPIL case would not necessarily entail a fundamental change in the attitude of the Conseil d’État in relation to international law. In fact, even though it resorts to the two criteria identified in that case in its reasoning, it still concludes that most provisions of the European Social Charter are not directly applicable even though they are clearly aimed at protecting the rights of individuals.Footnote 30 In other situations, it simply refers back to the identified criteria and merely resorts to a laconic formulation which presupposes the lack of direct effect of the Charter,Footnote 31 notwithstanding that other national jurisdictions such as the Dutch Supreme Court have long granted direct effect to some provisions of this text.Footnote 32 Besides, the criterion of the unconditional character of the provisions is sufficiently flexible to grant the judge with a significant discretion without making any difference to the situation that was prevailing prior to the GISTI-FAPIL case. It has even been announced that ‘the Conseil will often have to hide behind the same old formulations in order to deny direct effect’.Footnote 33
We are therefore far from the position of EU law before domestic courts. Concerning domestic courts, besides the fact that direct effect is assessed authoritatively by the EU judge who imposes it on national judges in cases where they do not recognize it spontaneously, the three criteria applied, namely precision, clarity, and unconditionality, have an objective character which stems exclusively from the drafting of the norm. Thus, while the intention of the parties, resulting from the case law of the PCIJ, remains a tool for the assessment of the criteria of direct applicability before national judges, it was not used by the Community judge in the Van Gend en Loos case relating to EC norms in domestic law, nor in the Demirel case relating to treaties that are binding on the EU.Footnote 34
The situation of international law before French courts is also very different from its situation before some foreign courts. To some extent, the Conseil d’État seems to have, by using these two criteria, matched the American case law, as specified in the Restatement (Third) of the Foreign Relations Law of the United States at paragraph 111: for a treaty provision to be self-executing, (i) its implementation must not depend upon the intervention of national legislation; and (ii) individuals must be the addressees of the provision. Interestingly enough, the criterion which offers flexibility by lending itself to a most subjective assessment is not the same in the two jurisdictions. While the French judge tends to rely on the addressees’ criterion to deny direct applicability, American courts frequently claim that ‘treaty drafters intended the treaty to be implemented by legislation rather than judicial action’.Footnote 35 Be that as it may, one can conclude that although it is formally monist, the American system has distinguished itself over the last years because of its material, or jurisprudential, dualism resulting from national judges’ reluctance to incorporate international law into domestic law.Footnote 36 This tendency has been called ‘nationalist’.Footnote 37
With regards to other foreign jurisdictions, it seems that the greatest openness to international law is usually made possible through the abandonment of the criterion pertaining to the object of the treaty or provision. In other words, the very subjective question of whether the object of the relevant provision is to regulate the relations between states or rather to grant rights to individuals is not raised. As is the case with EU law, the starting point appears to be a presumption that rights have been created in favour of individuals. Thus, the courts of the Netherlands and Luxembourg only take into consideration the necessity or absence thereof to adopt additional national measures, i.e. the criterion of unconditionality, including where an agreement imposes obligations incumbent exclusively upon states.Footnote 38 Similarly, Spanish judges tend to set as criteria for direct applicability the precise, complete, and unconditional aspects of the provision.Footnote 39 Whereas the relevance of the phrase ‘ordinary judge of international law’ with reference to the French judge may be put into question, the expression makes sense when applied to other national courts. By interpreting the theory of direct effect in a liberal way, those courts behave as ‘ordinary judges of international law’.
The conditions for the recognition of direct effect therefore seem to be the main obstacle to the effectiveness of international law within domestic law. As was put by A. Nollkaemper, ‘direct effect more often than not functions as a shield’.Footnote 40 Direct effect would then be to formally monist systems what incorporation is to formally dualist systems. Indeed, disregarding some exceptions,Footnote 41 the debate surrounding the direct effect of international norms is non-existent in systems which require the incorporation of international law into domestic law. This is the case for instance in the United Kingdom, Australia, Canada, Malta, Israel, Ireland, Denmark, Norway, Sweden, Iceland, and the Czech Republic.Footnote 42 Can we then go as far as saying that the question of the effectiveness of international law through domestic courts never arises in respect of these systems? Does it ensue from this that in all of these legal systems the national judge can never be said to be the ordinary judge of international law? In other words, is direct effect the only channel through which the national judge may progressively become the ordinary judge of international law? This is certainly not the case if the analogy with EU law is to be furthered. It is undoubtedly in relation to directly applicable norms that the concept of ‘ordinary judge of Community law’ has emerged, but it is in light of other forms of invocability that it has acquired its fullest meaning.
3. Remedies for the direct effect obstacle
Despite being initially the only method of ensuring the effectiveness of EU law, direct effect has lost its ‘monopoly’Footnote 43 to the benefit of other techniques which tend just as much, although in an indirect way, to ensure the prevalence of EU law by the national judge. Just as direct effect, these other techniques contribute to the transformation of the domestic judge into the ordinary judge of EU law. The two main techniques are that of consistent interpretation and the action for damages against the state for violation of EU law. A third, more controversial technique, is called the invocability of exclusion. These three techniques can be presented as remedies for the hurdle of direct effect. The question therefore arises of whether, in the absence of a more liberal recognition of direct effect, international law may benefit from these forms of invocability which would ground the assertion that the domestic judge is also the ordinary judge of international law.
3.1. Invocability of consistent interpretation
The obligation to interpret domestic law so that it is consistent with EU law, including when the latter is not directly applicable, has been imposed on national judges by the ECJ since 1984.Footnote 44 The obligation of consistent interpretation has then been extended to international agreements binding on the European Community.Footnote 45 The ECJ has thereby distinguished the principle of direct effect, which it does not recognize in the WTO (World Trade Organization) agreement on intellectual property at issue, from the obligation of consistent interpretation incumbent upon domestic judges. The invocability of consistent interpretation has clearly been applied in this case as a remedy for the lack of direct effect of the provisions concerned.
As a remedy to the obstacle of direct effect in monist systems, the theory of consistent interpretation is often dated back to the US Supreme Court's case law, through what is called the Charming Betsy canon.Footnote 46 Many courts belonging to formally monist systems use this technique. This is particularly so in the Netherlands,Footnote 47 Spain, Greece, Finland, and Romania.Footnote 48 The situation in Spain deserves special attention because the duty of consistent interpretation is prescribed for by the Constitution as Article 10.2 requires provisions on the fundamental rights and freedoms recognized by the Constitution to be interpreted in light of international treaties ratified by Spain, and this, in spite of the principle of constitutional supremacy over treaty law. A fortiori, the technique of consistent interpretation is also mandatory with regards to infra-constitutional norms, as it is frequently resorted to by national judges, not only in relation to norms that are devoid of direct effect, but also in relation to treaties which have not been published according to the rules.Footnote 49 Other constitutions opt for similar options.Footnote 50
The theory of consistent interpretation must however not be confined to monist systems. It also has a role to play in formally dualist systems in order to regulate the relations between international and domestic law. Thus, some national courts resort to this tool in order to take into consideration international norms which either have not been formally incorporated into national law or while incorporated seem to be in conflict with national norms. It then appears that a large number of domestic judges belonging to formally dualist systems resort to the theory of consistent interpretation in a way that is much more favourable to international law than do many of the judges belonging to monist systems. In so doing they adopt a rather monist approach. Certain cases are quite emblematic as it is the Constitution itself which imposes the use of this technique in the interpretation of any domestic norm. The South African Constitution is a notable example.Footnote 51 In these states, domestic tribunals tend to have a strong presumption of the consistency of domestic law with international treaty provisions which they would not otherwise be able to apply directly. Thus, following the case of Baker v. Canada (Minister of Citizenship and Immigration) determined in 1999,Footnote 52 several decisions by the Canadian Supreme Court have confirmed that the values and principles which are reflected in treaty law and binding on Canada should be taken into account when interpreting national norms, including when the former have not been incorporated into domestic law.Footnote 53 Thus, a system may be formally dualist and materially or jurisprudentially monist. The British courts have also used this technique with respect to Article 15 of the United Nations Convention Against Torture which relates to the inadmissibility of evidence obtained as a result of torture.Footnote 54 Similarly, the Israeli Supreme Court held that in accordance with the ‘“presumption of compatibility” between domestic law and provisions of international law’, it was ‘required to interpret statutes – as well as powers acquired by government authorities – in a manner which complies with the provisions of international law’.Footnote 55 Other examples may be found in relation to other jurisdictions such as the High Court of New ZealandFootnote 56 or the Supreme Court of Bangladesh which held that even though it had not been incorporated into domestic law, the United Nations Convention on the Rights of the Child should be taken into consideration in cases where there is a lack of certainty or a gap in national law.Footnote 57 The Indian Supreme Court adopted, in its interpretation of the Constitution, similar reasoning with respect to Article 9, paragraph 5 of the Covenant on Civil and Political Rights which relates to the obligation to compensateFootnote 58 whereby it explicitly relied on a similar interpretation adopted by the Australian Supreme Court on the Convention of the Rights of the Child. This tendency of jurisdictions based in formally dualist systems to incorporate, by way of interpretation, treaties relating to human rights into the domestic order has been rightly coined ‘creeping monism’.Footnote 59
Along this quite generalized trend of domestic courts considering themselves as the true ordinary judges of international law, a French exception seems to subsist. Thus, while Canadian and Australian judges overcome their own formal dualism by resorting to the theory of consistent interpretation, the French courts do not seem to live up to their apparent monism by rejecting any form of invocability other than direct effect. It may only be a matter of time, since the French judge has often needed more time than other national judges when it comes to the relationship between the national and the international legal order. This was obvious in regards to Community law, as it took longer for the French judge to accept its fundamental principles of primacy and direct effect than for other national judges based in dualist systems. Besides, the French judge is familiar with the technique of consistent interpretation for he applies it routinely to interpret legislative norms in light of constitutional norms which are not directly applicable and to interpret, at the request of the ECJ, national norms in light of EU law where the provisions of the latter do not comply with the established criteria for direct effect.
Although the technique of invocability unquestionably contributes to the transformation of the domestic judge into the ordinary judge of international law or into an ‘internationalist’ judge, it is limited as it must not constitute an interpretation contra legem. In other words, the technique of consistent interpretation can only be resorted to in so far as the national provision does not openly conflict with the international provision. Thus, other forms of invocability may be more effective.
3.2. Invocability of damages
The right to an action for damages before the domestic court for a violation of EU law, stemming from the Francovich and Bonifaci case,Footnote 60 first emerged as a remedy for the lack of direct effect. Subsequently, and upon the delivery of the Brasserie du Pêcheur and Factortame judgments,Footnote 61 the ECJ drew a distinction between direct effect and invocability through damages by rejecting that they could be mutually exclusive. However, even today, the possibility of bringing an action for damages in order to obtain compensation from a state for violations of its EU obligations remains an option for the aggrieved applicant. This is only possible in situations where the applicant cannot directly invoke the provisions at issue either because of their lack of precision or unconditionality, or because of the horizontal nature of the dispute in question.Footnote 62 However, given the direct applicability of most of the provisions of EU law, the action for damages now appears to be more of a complementary mechanism than an alternative per se. The same is not true with respect to international law, where, given the restrictive criteria for direct effect, the action for damages remains a convenient alternative at least on a theoretical level.
The Francovich and Bonifaci case has been followed by several supreme jurisdictions of the EU.Footnote 63 Before the French judiciary, a right to obtain compensation for violations by the state of its international obligations emerged almost at the same time as when the action for damages was crafted by the ECJ. In fact, the administrative judge recognized in the 1990s a fault-based liability triggered by the introduction of administrative acts that are inconsistent with international obligations.Footnote 64 The groundbreaking Gardedieu judgment completed the picture in 2007 by creating a new form of state liability. This results from the inconsistency of a legislative act with an international obligation on the ground of a sui generis type of responsibility as distinguished from the two classical fault-based and no-fault liability systems.Footnote 65 In both cases, the breach of an international obligation, be it by the administration or the legislator, will trigger an obligation incumbent upon the state to make good the resulting damage. The analogy between this case law and the Francovich and Bonifaci case, which substantiated the notion of ‘ordinary judge of EU law’, has already been made.Footnote 66 While one may read into this case law an implicit condition of direct applicability of the relevant treaty obligation,Footnote 67 such a criterion does not appear explicitly in any of the relevant cases. Consequently, nothing indicates clearly whether compensation intervenes as a remedy for the absence of direct effect or, on the contrary, if it is conditional upon the direct effect of the norms invoked.
It is argued that the action based on liability for violations of international obligations could appear as supplementary means for the judge to ensure the effects of international law in cases where it cannot be directly applied. Indeed, if the logic of the EU judge is to be followed, the invocability of damages, similarly to the invocabilities of interpretation and exclusion, would not be dependent upon direct applicability. This is because it is a corollary of the principle of primacy rather than of that of direct effect.Footnote 68
However, as applied by the French courts, state liability for breaches of international law are not subject to the same legal regime as that of its equivalent in the EU legal order. Therefore, while the French judge does not make any mention of the principle of direct effect in cases of state liability, neither does he indicate which other conditions must be fulfilled by the relevant provision of international law. The judge simply applies the administrative law regimes of fault-based liability in one case and sui generis liability in the other. As for the ECJ, since its first judgments it has submitted state responsibility for breaches of Community law to two main conditions: the object of the breached rule must be to confer rights on private parties; and the violation must be of a specific nature. No reference to such conditions can be found in the French case law relating to state liability for violations of EU law. The French judge merely defines the violation as fault-based and assesses the causal link with the damage suffered by the applicant. This apparently more favourable regime may be attributed to the influence exercised, in parallel with EU law, by the European Court of Human Rights, which would probably not be satisfied with a regime that, according to another condition set by the ECJ, only sanctions violations which go beyond a certain threshold.Footnote 69 The administrative judge thus endeavoured to create a regime that would apply to the whole of international law and which would not be specific to EU law. It is worth noting that the ECJ requirement that the provision involved grant rights to private parties amounts to one of the conditions set by the Conseil d’État with respect to direct effect. Whereas the European judge has carefully distinguished between the conditions for direct effect, specifically precision, clarity, and unconditionality, and those for invocability of damages, the French judge has merged these two sets of conditions by requiring that a provision be unconditional and grant rights to individuals to be directly applicable. It is as if all the efforts were focused on making direct applicability hard to get while not seeing the necessity to limit the recourse to state liability for breaches of international law by adding new conditions to the classical regime of public liability.
This may be because even though French case law relating to the principle of state responsibility for the violation of EU law is quite extensive, no implementation of this mechanism can, in practice, be found with respect to international law, understood as going beyond European law. Indeed, while the wording of the French judge points to state liability for breaches of international law generally, in practice this legal action is only resorted to in cases involving EU law or the law of the ECHR.Footnote 70 This is confirmed by the case law which followed and applied the Gardedieu judgment.Footnote 71 Even though one may agree with C. Broyelle that ‘Community law . . . has led judges to come up with new solutions which will benefit to the whole of international conventional norms’,Footnote 72 no example of an application of this mechanism to violations of other norms of international law can be found at present. Yet, ECHR law, just as EU law, benefits from an extensive direct applicability which relegates the action for damages to a complementary, if not secondary, function. What would truly allow for the expansion of the effectiveness of international law before domestic judges is the extension of the legal action for damages to other treaty norms.
There are some examples of domestic courts in other jurisdictions which have recognized a right to compensation following a violation of international law which differs from that stemming from the EU or the ECHR. The Belgian Cour de cassation held that any violation of a provision of international law endowed with direct effect is a fault which triggers the liability of the administration.Footnote 73 Although the invocability of damages is clearly dependent upon the existence of direct effect in this case, it must be recalled that the latter is not subject to the condition that the norm involved grant rights to individuals before the Belgian judge. The invocability of damages is then more favourable to individuals than the regime of liability created by the Francovich case law or the regime of direct effect in France. American courts seem to conceive the action for damages as a remedy for the breach by state authorities of a treaty provision that is self-executing and that creates individually enforceable rights such as Article 36 of the Vienna Convention on Consular Relations.Footnote 74 This may be analysed as one of the rare cases in which direct applicability differs from the doctrine of self-executing treaties. While direct applicability, with its exclusion and substitution effects, would have led to the suppression of the incriminating statements that had been introduced in breach of a treaty provision, the judge has opted for an alternative type of invocability where the action for damages provided that the self-executing provision also creates enforceable rights. A principle of state responsibility is also found in Greece where it seems to be applied to the whole of international lawFootnote 75 or in Austria where it is equally submitted to the direct applicability of the provisions at issue.Footnote 76 Generally speaking, depending on their public liability regimes, most monist systems can develop this type of invocability for breach of treaties which are adequately ratified. Indeed, the latter constitute an integral part of the rules of legality which, when breached, may entail a right to compensation under certain, more or less restrictive, conditions. However, for this invocability of damages for violation of international law to be as effective as that resulting from the Francovich case of the ECJ, its conditions ought to be distinguished from those of direct applicability. Indeed, the specificity of the EU invocability of damages is that it can fill in the gaps of the direct effect invocability. Besides, the specific conditions surrounding the granting of compensation in national law may also render utopic this remedy. For instance, some systems of domestic responsibility make it generally difficult to obtain a subjective right to compensation for a breach of an obligation perpetrated by the administration. Indeed, although one is always tempted to look for an international ground for invocability of damages in domestic law,Footnote 77 the latter depends on rules of national law which are generally constitutive of the regime of administrative responsibility.
With regard to dualist systems, it seems difficult to envisage that states would be subject to an obligation to compensate triggered by a breach of a treaty provision where the provision has either been formally incorporated into national law, and it is in this scenario a matter of responsibility for the violation of domestic law, or it has not yet been incorporated, and thus there cannot be a breach since there is no domestic obligation incumbent upon the state despite ratification of a text entailing an international commitment. There cannot exist any system of liability for breaches of international law. This is why the British legal system, which is formally dualist, had to adapt and create a new legal action derived from the Francovich case law and exclusively reserved to EU law.Footnote 78 However, a recent case by the Norwegian Supreme Court shows that even domestic judges belonging to predominantly dualist systems can be innovative where the need to give effect to international law is acknowledged. For example, a right to compensation for breaches of the EEE Agreement has been established even though neither agreement nor the national measure of incorporation provided for it. This has been understood as an intention to ensure a greater effectiveness of international law.Footnote 79
3.3. Invocability of exclusion
The main consequence flowing from the Simmenthal case for national courts is the obligation to set aside national law that is inconsistent with EU law. This is referred to as the ousting effect of the principle of primacy. However, this ousting effect was linked in this case to the direct effect of the provisions at issue. It was subsequently found that it could apply including in the absence of direct effect. This is called invocability of exclusion as opposed to invocability of substitution also entailed by direct applicability. The latter implies that the EU norm substitutes the national contrary norm while the former only allows to set aside the national contrary norm without replacing it by the EU norm. Even though this form of invocability has ‘never been accepted by the Court in a general and explicit way’,Footnote 80 it is possible to detect an evolution in this direction in its 2010 Kücükdeveci case on a directive which was not directly applicable.Footnote 81 National courts, and particularly French courts, applied this form of invocability early on while refusing to grant direct effect to some EU provisions, notably those contained in directives invoked to challenge an individual measure.Footnote 82 The technique displays an interesting potential with respect to international treaty law, the effects of which remain significantly limited by the principle of direct effect. Nevertheless, even though it has been strongly advocated by some authors and advocates general, resort to this means of ensuring the effectiveness of international law remains very occasional in France, if not non-existent. Its use is consequently strictly limited to EU law and there are no obvious examples of its application to general international law.
Advocates general R. Abraham and G. Dumortier have both argued for the same technique with different words. They recommended that the Conseil d’État accept, in the absence of direct effect, an invocability of exclusion, or of incompatibility, for the former and an ‘objective control’ for the latter, following the practice of the Belgian jurisdiction for example. The invocability of exclusion was notably presented by R. Abraham as an open possibility against legislative instruments but not against individual measures. In her opinion on the 2012 GISTI and FAPIL case, G. Dumortier relies on comparative law and highlights that ‘[s]tates having an objective approach to legality, such as the Netherlands, Belgium, or Spain are those in which the judge has been the most reluctant to subject any control to the condition of direct effect’.Footnote 83 In other words, in those countries, the judge has been open to an invocability of exclusion and has accepted to control domestic measures even where the international norms at issue were not directly applicable. In Belgium it seems that a treaty provision, even if devoid of direct effect, can also be used as an auxiliary norm of control in constitutional disputes pertaining to the competence of the arbitral Court, and now the Constitutional Court,Footnote 84 and as a norm of control in disputes relating to the legality of administrative measures, which fall within the competence of the Conseil d’ÉtatFootnote 85. This case law distinguishes between subjective direct effects and objective legality which binds the administration including where the treaty provision does not grant any right to private parties and is thus not directly applicable.Footnote 86 Similarly, Spanish judges follow a twofold procedure. They do not limit their examination of an international norm to its direct applicability, but take their reasoning further in cases where the criteria of precision and of unconditionality are not fulfilled, for the purposes of determining whether the national norm is compatible with the state's international treaty obligations.Footnote 87 The European legal order itself seems to recognize this type of limited invocability with respect to international norms which are binding on the EU but which do not have direct effect. This is referred to as the Fediol Footnote 88 and Nakajima Footnote 89 exceptions to the principle of non-invocability of the WTO agreements before the ECJ. Thus in these two cases, the ECJ recognized two situations in which, despite their lack of direct effect, the GATT agreements can be invoked before it to challenge the legality of an allegedly inconsistent measure. However, this limited invocability is restricted to measures of the EU which are designed for the implementation of the international norm at issue or which expressly refer to specific provisions of the international norm.
4. Conclusion
As is the case in the context of EU law, the national judge may well appear as the ordinary judge of international law. However, this characterization is only theoretical due to the exceptional character of the recognition of direct effect and the recourse to other forms of invocability. In fact, even though comparative law may in certain circumstances support this assertion, this does not accurately reflect the current position of the French judge, notwithstanding the notable emergence of some signs of real or desired evolution over the last years. Nevertheless, at least in formally monist systems such as France, in situations where all the theoretical tools are present to turn the national judge into the ordinary or primary judge of international law, it may be true that there is only a difference in degree rather than a difference in nature between the national judge as an EU judge and the national judge as an international law judge.Footnote 90 Although this expression remains for now better suited to EU law than to international law, if applied to domestic judges generally, it leaves room for development. The tools are there. They have been identified. They need only be invoked in a more systematic manner by applicants before domestic judges to further the change of attitude towards international law. Besides, the assessment of the technique of consistent interpretation shows that most ‘internationalist’ judges do not always come from formally monist systems. Contrary to what may be thought at first, the degree of effectiveness of international law does not depend upon the monist or the dualist option endorsed by a given system, but it actually depends on the quasi-informal attitude of the national judges to international law.
