The principle of state liability for breach of EU law by national (superior) courts in Köbler (‘Köbler liability’) might be described as the highpoint of a line of CJEU jurisprudence, which gradually and progressively extended the reach of the primacy of EU law from the declaration of rights to the enumeration of specific EU law remedies.See T Tridimas, ‘Liability for Breach of Community Law: Growing Up and Mellowing Down?’ (2001) 38(2) Common Market Law Review 301-32 [noting how rights-remedies are part of constitution-building through out-working primacy and direct effect]. Our understanding of state liability, in particular, is complicated by the fact that while the conditions of liability are autonomous and European, they are minimum standards of protection that must be implemented by Member State courts, and are often coloured according to the principle of national procedural autonomy.Up to a point, i.e. Rewe effectiveness and equivalence. See Case C-120/78, Rewe-Zentral AG v Bundesmonopolverwaltung für Branntwein  EU:C:1979:42. A Köbler liability action provides individuals with a remedy for the ‘manifest breach’ of their EU law rights, and a corresponding procedural duty on Member States’ courts to give effect to this rights-remedy regime.As Varga clarifies, the national court has a procedural duty to apply EU law rules, and this may give rise to a right to a compensatory remedy where a right is conferred on an individual, see S Varga, The Effectiveness of Köbler Liability in National Courts (Hart Publishing: London 2020) 17-19. This remedy may arise as a result of judicial non-application and, more controversially, misapplication of EU law. The right to compensation for judicial error of law, or indeed potentially male fides, is a novel remedy for a large number of jurisdictions, which traditionally resolve judicial error of law by judicial review. It is therefore apt to cause a certain degree of friction between national courts which must give effect to their domestic constitutional and remedial order while, at the same time, reconcile their case-law with their judicial function as ‘agents of compliance’ within the EU legal order.See ibid 5-7, where Varga discusses the constitutional context. This friction is reflected in the way in which the principle of procedural autonomy is applied in practice and, in particular, in the relationship between res judicata or the finality of judgments and the Köbler case-law before national courts. In practical terms, holding a court liable in tort for breaching EU law also involves courts censuring their colleagues on the bench, which may be perceived as scuttling the hierarchy of domestic courts.
Varga, in the first part of her book and against this complex background, sets herself the ambitious task of examining the extent of Köbler liability’s ‘real impact’ on Member States’ legal orders de lege lata from the ‘bottom up’ and in context.ibid 2. The context, as Varga sees it, is other national remedies, and the structure of other EU law remedies. Her analysis is timely given that it is almost twenty years since Köbler was decided. Her starting point is to review Köbler liability’s criteria for recovery, namely manifest breach and causation.But curiously not conferral of a right. Perhaps she agrees with Dougan who argues that it is under-specified and decides that it does not merit in-depth treatment. See M Dougan, ‘Addressing Issues of Protective Scope within the Francovich Right to Reparation’ (2017) 13 European Constitutional Law Review 124-65. However, the issue of conferral of a right resurfaces, indirectly, at a later point when Varga discusses the scope of Köbler liability in the context of the refusal to make a preliminary reference, see Varga (n3) 191-212. It seems that the high threshold of manifest breach and the problem of concurrent liability of several branches of government make it very difficult for claimants to successfully obtain a remedy.See ibid 59-81 on causation both in EU law theory and Member States’ courts practice. An important additional problem here is the duty to mitigate, which might be better called in this context the duty to exhaust domestic remedies. This holds true whether the relevant ‘procedural’ breach relates to inter alia the direct effect of directives, conforming interpretation or the obligation to make a preliminary reference. The noteworthy exception is for breaches related to the failure to apply established ECJ case-law. In other words, the ECJ itself sets a high bar for bridging the gap between a procedural breach of rights and a substantive action for damages.See ibid 23-46. Varga, then, sets out to examine national case-law and the way in which the inauspiciously high threshold requirement has been interpreted by Member States’ courts in practice. A great diversity is apparent among the EU-28.ibid 83-117. Varga groups Member States’ approaches into three categories: regimes compatible with EU law,Denmark, Latvia, Spain, Sweden, Belgium, Bulgaria, Germany, France, Lithuania, the Netherlands, Austria, Portugal, Finland, the United Kingdom, Italy, and Poland. Member States that refuse to recognise Köbler liability,Czechia, Hungary, Ireland. and Member States without any sign of accommodation or refusal.Cyprus, Croatia, Estonia, Greece, Luxembourg, Malta, Romania, Slovakia, and Slovenia. Even in those ‘regimes’ compatible with Köbler liability, its practical impact is limited. Varga seems to attribute this to the high threshold of recovery. For example, in France, while Köbler liability is now recognised, and to a degree harmonised with French administrative law’s faute lourde requirement, no claimant has yet been able to surmount the sufficiently serious breach requirement. This is replicated in almost all other ‘European-friendly’ regimes.Finland seems to be the outlier with five awards of Köbler damages to date, see Varga (n3) 103. But while some Member States’ courts have more or less successfully reconciled their judicial practice with Köbler liability,ibid 87. In all ‘European-friendly’ jurisdictions, a period of bedding down can be observed. In some jurisdictions, some doubts as to the compatibility with EU law remain even if on the whole it appears that Köbler liability is accepted (eg Spain). others have refused outright to apply Köbler liability. Notable are Hungary, Ireland,With respect, I think Varga reads too much into Cronin v Dublin City Sheriff  IEHC 685 (Ní Raifeartaigh J) discussed at ibid 113. In that judgment, the plaintiff was seeking, erroneously, to rely on Köbler as grounds to overturn a judgment, and not to seek damages. The plaintiff’s claim was entirely misconceived, and Ní Raifeartaigh J was right to dismiss the Köbler aspect of the claim, and instead address the issue based on national procedural rules on res judicata. Varga is correct, however, to state that the Court seems to misunderstand the import of the CJEU’s ex officio jurisprudence, and, one might add, CJEU jurisprudence on res judicata, when coming to its conclusions. and Czechia, where the principles of res judicata and the finality of judgments apparently continue to prevail.Apparently, because in Hungary the jurisprudence is mixed, and for Ireland (n15). It should be added that the Irish Supreme Court has recognised Köbler liability since McGrath v Irish Ispat Limited  IESC 43 (unrep., SC 2006). It seems that outright rejection is only manifest in Czechia, see Varga (n3) 109-10. The remaining jurisdictions have not considered Köbler liability to date and, as such, it is difficult to draw any clear conclusions. But whether in European-friendly jurisdictions, or in those jurisdictions in which the reception has been lukewarm or hostile, the basic point is that Köbler liability has had a negligible impact in practice.See ibid 117-18 for conclusions. To date,Or, rather, as Varga clarifies, her book covers case-law in the period from 2003- 2018, see ibid 5. while national courts have considered Köbler liability on 60 occasions, damages have only been awarded five times across the EU-28 (at the time of writing).ibid 4.
Varga then attempts to explain why this may be the case, by drawing on ‘context’. For Varga, context denotes the alternative remedies available to claimants in Member States’ legal orders. In several Member States, at least, national procedural autonomy provides what she regards as functionally equivalent remedies for error of law.See ibid 220-1, where Varga includes a helpful annex, including an overview of ‘remedies under national law offering a comparable level of protection of individual rights to that provided by the Köbler principle in the event of breach of EU law by Member State courts.’ In some Member States, for example, non-tortious avenues of recourse exist such as the possibility of a re-trial, de novo, where a final judgment is incompatible with EU law. This allows for a judgment that is, otherwise, res judicata to be re-opened, and is usually contingent on a heightened standard of review.See ibid 125, Denmark, Malta, Finland, Sweden, the UK, and Lithuania. In some jurisdictions, this right to a re-trial in exceptional circumstances is recognised in Member States’ procedural codes.ibid, namely Croatia, Romania, and Slovakia. However, even in those Member States that allow for re-trial, it is qualified in certain respects,Eg limited to administrative law only. and there are at least fifteen Member States that exclude re-trial to a large extent or entirely in EU law matters.Varga (n3) 126. Varga states that re-trial offers, in certain respects, a higher level of protection that Köbler liability, and is better from the point of view of the rule of law and because it does not require the claimant to demonstrate a tort.ibid 125. She shows some evidence that, at least in certain Member States, judiciaries view re-trial and state liability as alternative remedies.ibid 145, notably Finland and Lithuania. In other Member States, certain states give priority to re-trial and others to liability claims. It should be noted that in this discussion Varga is sometimes referring to the general position on re-trial as distinct from the specific issue of whether re-trial is available with regard to EU law matters.ibid 146, where Varga notes the ‘theoretical’ possibility of the application of the remedy to EU law claims. It is thus difficult, in my view, to extrapolate from the availability of re-trial in EU law matters in some states, wider conclusions about its contextual position as an EU-law compatible alternative remedy to Köbler liability, as Varga seems to do. The same might be said about ‘constitutional complaints on the lawful judge principle’.ibid 148ff. In this context, this principle holds that in circumstances where a court fails to make a preliminary reference, the right to a lawful judge or, more broadly, the right to a fair trial, is violated. The consequence of finding a violation is the annulment of the relevant judgment. The remedy is, in essence, a breach of the right to an effective remedy according to Member States’ constitutional law applied in the context of EU law. There are two conditions: the first is that the CJEU be recognised as a lawful judge; the second, that individuals have locus standi before constitutional courts. To date, five Member States only have recognised the CJEU as a lawful judge.ibid 150. This greatly weakens Varga’s argument that it is an alternative remedy insofar as she means to draw a general conclusion about its potency as a contextual alternative remedy to Köbler liability within the EU. However, as Varga notes, another and more general path is the ECtHR Article 6 jurisprudence that states that there is an obligation on Member States to state reasons for the refusal to make a preliminary reference.See ibid 151-56 for analysis. This jurisprudence is ambivalent, but appears, in recent years, to extend a little beyond Cilfit; it is not sufficient simply to state one of the Cilfit grounds – an explanation for its invocation is required, which will vary with context and the nature of the dispute. This may be viewed as an alternative or indirect non-compensatory remedy in the context of refusal to make a preliminary reference. But if it is an alternative remedy, it is rather uncertain and limited in scope.And the ECtHR case-law it is not particularly consistent, see J Krommendijk, ‘“Open Sesame!”: Improving Access to the ECJ by Requiring National Courts to Reason their Refusals to Refer’ (2017) 42(1) European Law Review 46-62, at 54-56 esp.
Perhaps because her contextual analysis, which argues that alternative remedies explain the lack of recourse to Köbler liability, is inconclusive, and (at least in my view) unconvincing, Varga, in her penultimate chapter – the second part of the book – turns from analysing the law de lege lata to attempting to re-imagine Köbler liability de lege ferenda. Because it is largely a paper tiger, it should not be considered a compensatory remedy aimed at vindicating EU law rights.See Varga (n3) 177-82 esp., where she canvasses the arguments for Köbler liability as a deterrent tool in the arsenal of the CJEU. However, there is no definite statement that it is a deterrent tool, as distinct from a compensatory remedy, until 215. Even so, Varga says this is its ‘primary function’. However, her analysis, which stresses procedural autonomy, belies this statement, and there is no extensive analysis of the alternative perspective. I take it that she sees it as its exclusive function. But, then quō vādis, Köbler liability? Varga seems to accept the deterrence or ‘sanction for the disobedient state’C Harlow, ‘Francovich and the Problem of the Disobedient State’ (1996) 2(3) European Law Journal 199-225. explanation of Köbler liability as the only plausible explanation left. Once this explanation is accepted, Varga then considers that its central role should be in relation to the refusal to make a preliminary reference. In other words, Köbler liability should not be considered from the point of view of a specific (individual) rights-remedy requirement, but as a means to sanction states for their failure to engage in ‘judicial dialogue’, which is ultimately apt to undermine the effectiveness of the EU legal order. Buttressing this argument requires some legal casuistry, and Varga skilfully obliges. Her starting point is that Article 267(3), TFEU – the refusal by national courts to refer a case via preliminary reference – does not confer a right to damages for individuals,Varga (n3) 201. and should instead be understood as a procedural obligation on Member States to make a reference. However, when Article 267(3) is read against Article 47 of the Charter’s declared ‘effective judicial protection’ requirement, it might justify liability for refusal to refer a question. As is well-known, Article 47 of the Charter concerns fair and effective remedies and, Varga speculates, a failure to refer a question may be considered to impugn the fairness of proceedings. Drawing on ECJ case-law, but also on analogies from ECtHR jurisprudence,ibid 207, where Varga notes that art 52(3) provides that the level of protection afforded by the Charter should be at least equivalent to that provided by the ECHR. Varga concludes that there are certain circumstances in which the failure to state reasons should be considered a violation of Article 47 of the Charter. It apparently follows from CJEU case-law that where ‘fundamental procedural rights’ are violated, compensation for non-material losses should be awarded and constitute a sufficiently serious breach.ibid 207-8. Therefore, Varga fashions a potentially novel means of recourse in Köbler liability for failure to give reasons above the rather deferential Cilfit criteria. Notwithstanding Varga’s broad Köbler scepticism, here she hints at a residual role for Köbler liability when the Member State courts not only misapply EU law, but actively disengage from it. It seems in the end that, aside from such circumstances, Varga’s ultimate conclusion is tant pisKöbler liability in other circumstances:ibid 216 esp. effectiveness and equivalence can, and do, fulfil the task that Kobler liability perhaps might do, and Köbler liability is better considered as a big, deterrent stick when the ‘sincere co-operation’ between Member States and the CJEU breaks down.See Varga’s findings at ibid 213-219. The paucity of Member State case-law seems, in her analysis, to serve as a ballast for her argument that Köbler liability,See ibid 216 on what the lack of successful claims ‘shows’. in the final analysis, has very little to do with compensating individuals for the violation of their EU law rights, and everything to do with ensuring that the EU legal order is effective.
I wish to contribute three, related criticisms. Varga’s monograph, from the outset, largely eschews theory. Instead of theorising about Köbler liability, and in particular its relationship to Francovich state liabilityAlthough Varga indirectly addresses this issue in her discussion of causation at ibid 59-81. and, more broadly, state liability’s place in the overall architecture of EU law – well-trodden ground according to Varga – the added value of her book is that by pursuing a ‘bottom-up’ doctrinal and contextual analysis, she is better placed to judge ‘the real impact of Köbler on national remedies.’ibid 2. This ‘law in action’ approach is presumably meant to expand our understanding of Köbler liability in a way that reaches beyond what one might call EU law ‘coherentism’,R Brownsword, ‘Law and Technology: Two Modes of Disruption, Three Legal MindSets, and the Big Picture of Regulatory Responsibilities’ (2019) 14 The Indian Journal of Law and Technology 1-39. For a critique of ‘coherentism’ in an EU law context, see HW Micklitz, ‘The (un)-Systematics of (Private) Law as an Element of European Legal Culture’ in G Helleringer & KP Purnhagen (eds), Towards a European Legal Culture, (Baden-Baden: Nomos 2014) 81-114. namely the unfolding of the internal logic of the law of integration.P Pescatore, Law of Integration: Emergence of a New Phenomenon in International Relations, Based on the Experience of the European Communities (Leiden: AW Sijthoff 1974). In a sense, EU law ‘rhetoric’ is being tested against Member State judicial practice. From the latter perspective, Köbler liability is a mirage, because the grounds of liability effectively make recovery impossible. From this finding, she concludes that whatever the law in books says about the compensatory nature of the remedy, practice tells us otherwise, and as such it should be re-imagined in a more modest, and delimited way in the light of its truer role as an arm of EU law effectiveness in extremis. It is trite to say that the EU legal order is functionalist, and that its liability law is ‘regulatory-instrumentalist’ in tenor.Brownsword (n42). But it is rather odd to argue from the principle of effective judicial protection in the Charter of Fundamental Rights to buttress this conclusion. Charter rights, more typically, might be considered as concerned with the protection of individual rights and the rule of law. A deeper engagement with the role of the Charter as a means to effect classical rule of law, or otherwise, would have been beneficial. In other words, EU law is perhaps a bit more complex than Varga seems to intimate.
Varga seems to infer, relatedly, that since Köbler liability is a second-best solution to alternative remedies, whether Member State-based, or, one might add, based on EU law, the best explanation of Köbler liability is one exclusively based on ensuring the effectiveness of EU law.Varga might disagree with my characterisation of her argument here. However, see (n32). This is certainly true if one stresses the procedural aspect of the remedy – imposing a duty of co-operation on Member State courts as ‘agents of compliance’, and the way in which it has arisen alongside Francovich state liability and other measures that limit the procedural autonomy of Member States as arising from the outworking of the primacy of EU law. However, it is a non sequitur to argue that because it is secondary, and concerned with effectiveness, it is therefore solely concerned with compliance qua instrumental goal. It is entirely possible to argue that the EU is Janus-faced – not simply a functional legal order, but also concerned with more classical notions of creating an objective legal order of rights and remedies.See J Krommendijk, ‘Is the Light on the Horizon? The Distinction between “Rewe Effectiveness” and the Principle of Effective Judicial Protection in Article 47 of the Charter after Oriizonte’ (2016) 53 Common Market Law Review 1395 [distinguishing a rights-based and effectiveness-based understanding of effective judicial protection]. In this light, Member States’ reliance on other, perhaps sub-optimal and non-compensatory avenues, or their outright dismissal of Köbler liability, in some circumstances without adequate justification, appear less about providing alternative and effective remedies, and more about the violation of the EU law rights to a specific remedy in practice by failing to adequately integrate EU law into their domestic legal orders.See Varga (n3) 188, where she seems to doubt that it requires the provision of a specific remedy citing M Dougan, ‘The “Disguised Vertical Direct Effect of Directives’ (2000) 59 Cambridge Law Journal 586-611. In other words, it is one thing to say that the remedy is difficult to apply in practice, and to provide empirical evidence to support this argument; it is quite another to say that it should not be applied in practice, except in an attenuated way, and instead say that other remedies, often non-compensatory, provide ‘a comparable level of protection’.Varga (n3) 220. Incidentally, it is not entirely clear how ‘comparable’ is measured. Are non-compensatory remedies really comparable to compensatory remedies? They are if the focus is entirely on the compliance, but not if on compensation.
Finally, a pitfall of the ‘bottom-up’ approach more broadly is that it appears at times that national law is in the driving seat in determining the meaning and scope of Köbler liability. Rather than answering the question of ‘the real impact of Köbler [liability] on national remedies’, her stated aim, Varga’s analysis seems to show the impact of national remedies on Köbler liability. This is evident in the discussion of the very possibility of judicial liability, which traditionally is entirely excluded in six Member States, and recognised with some reservations in the rest.ibid 80-82 and 82ff. Varga’s discussion of the well-known and uneasy relationship between EU law rights and national procedural autonomy is fascinating and comprehensive. But, as Varga acknowledges, procedural autonomy has its limits in effective judicial protection and by focusing on the national ‘reception’ of EU law, one might be forgiven for thinking that ultimately the meaning of EU law is decided on in a haphazard way by national courts. Thus, for example, Varga presents the Spanish legal position as providing judicial liability for ‘the improper administration of justice’, and states that the same holds true, ‘theoretically’, for breaches of EU law.ibid 84. Elsewhere, Varga notes how French law has adapted its faute lourde criterion in the light of Köbler in relation to failure to refer a case,And in other respects, see ibid 94-96. which evidences how French courts have ‘acknowledged that the state may incur liability based on the criteria established by in the ECJ [in Köbler]’.ibid 96 (emphasis added). Ireland appears entirely resistant to Köbler liability claims.See n15. Of the 28 jurisdictions surveyed, it is only in Belgium that a sort of entente has been reached between Köbler liability and domestic law, because the Belgian courts have harmonised, so to speak, the Köbler liability criteria and recovery in national law. Rather, the overall picture that emerges is one of grudging acceptance or reception of EU law, but one which is channelled through, and in some cases distorted by, national law particularities.See Varga (n3) 90. Similar stories can be told for eg Bulgaria, Germany, Portugal. But Belgium seems to accept EU law without hesitation. This is an interesting and valuable finding on its own terms, but in the absence of a broader theoretical approach to frame these findings, one which in particular shows exactly how compensatory and non-compensatory remedies are equivalent, it is difficult to endorse Varga’s positive argument that the violation of Köbler liability by Member State procedural or remedial rules is permissible when other domestic remedies provide for the equivalent and effective protection of EU law.See ibid 216, where Varga states that given the principles of equivalence and effectiveness, ‘I think that Member State remedial or procedural rules which are in violation of the Köbler principle must not be necessarily considered incompatible with EU law.’ Some reference to, and perhaps treatment, of the theoretical literature on conflict of laws,See eg C Joerges, ‘The Impact of European Integration on Private Law: Reductionist Perceptions, True Conflicts and a New Constitutional Perspective’ (1997) 3(4) European Law Journal 378–406. or hybridisation,See eg M Amstutz, ‘In‐Between Worlds: Marleasing and the Emergence of Interlegality in Legal Reasoning’ (2005) 11(6) European Law Journal 766–84. for example, might have strengthened this argument, and made it more convincing.
Overall, this book’s main achievement is to draw together 60 decided cases spanning 28 jurisdictions, which gives a rich overview of Köbler liability from the ‘bottom-up’. It is, in other words, a much-needed analysis of the law de lege lata, which will no doubt be an invaluable resource for EU law scholars working on the issue of judicial liability in EU law. For this reason, it should be applauded as a welcome, ‘bottom-up’ addition to EU law scholarship. However, it is my view that it perhaps pays too much heed to procedural autonomy on its own terms and does not have enough regard for the complexity of state liability in EU law, and, more broadly, the sui generis constitutional structure of the EU.