Costello v The Government of Ireland, Ireland and the Attorney General (Approved) [2022] IESC 44

1. This is a complex and important case, made more difficult by the level of abstraction necessarily involved, as the appeal does not immediately require the application of the law to a set of facts. The questions raised are to a large extent hypothetical and dependent on the evolution of thinking regarding the Comprehensive Economic and Trade Agreement (CETA).

2. This judgment does not purport to engage in an extensive and detailed analysis of the provisos of CETA, of the legal issues arising, nor of the authorities and literature surrounding the proposed ratification of CETA. I am in the happy position of being able to very gratefully adopt the detailed and eloquent judgments of my colleagues. Rather, I use this opportunity instead to make some observations regarding the issues in the appeal, and to identify my precise answers.

3. I agree with the High Court, and my colleagues in this Court, that the ratification of CETA by Ireland is not “necessitated” by EU membership, and adopt the reasoning of Hogan J. in regard to this ground of appeal.

4. I would allow the appeal on the grounds that CETA is capable of infringing upon judicial sovereignty to an impermissible extent, and I gratefully adopt the reasoning of my colleagues Dunne J. and Hogan J. with regard to that issue.

5. I do not consider that the ratification of CETA is impermissible on account of Article 15 of the Constitution, and I adopt and agree with the analysis of the Chief Justice and Dunne J. regarding this ground of appeal, subject only to the observations I make below at paras 80-85 regarding the power of the Joint Committees to interpret CETA.

6. In my view, the loss of judicial sovereignty is the central issue.

General observations regarding trade agreements

7. The Constitution does not envisage or demand a protectionist or isolationist approach to international relations or to trade. This is clear from the evolved notion of state sovereignty analysed in the judgments of my colleagues. The essence of this sovereignty is the ability and willingness to engage as an equal partner in international affairs, and sovereignty is thus not to be seen as either one-dimensional or a blunt refusal to be influenced by external actors. It involves an engagement with and openness to other states which can and does impact on domestic policy and thinking.

8. This appeal concerns the limits of that engagement, and whether the ratification of CETA is capable of resulting in consequences that infringe upon the sovereignty of the legislative or judicial organs of state provided for in Bunreacht na hÉireann.

9. The Irish State and the EU enjoy a long and fruitful diplomatic and trade relationship with Canada, the furtherance of which is intended to be beneficial to the domestic and intra-EU economies of the parties to the agreement. This appeal concerns one aspect only of the proposal to ratify CETA, that dealing with the jurisdiction of the arbitral tribunals, and the enforcement of their awards. The balance of CETA, concerning trade, investment and tariffs, has been in operation on a provisional basis since September 2017.

10. The purpose of Chapter Eight of CETA is to afford protection to Canadian investors and those of the Member States from discriminatory or arbitrary expropriation and unequal or inequitable treatment by actions of those states. Such an investor is a party, a natural person or an enterprise that seeks to make, is making or has made an investment in the territory of that state. An investment includes any asset which that investor owns, or controls directly or indirectly, and therefore can include a shareholding or a share in an investment vehicle or fund, which in turn has made such a direct investment.

11. CETA does not envisage a commercial arbitration based on contract where the parties choose to submit to arbitration a dispute regarding their contractual rights and obligations, and where that dispute, wherever and by whomsoever it is resolved, operates within the sphere of private law. The jurisdiction of the CETA Tribunal is different in that it does not arise by reason of an individual contract between an individual investor, the Irish State, Canada, a Member State of the EU, or the EU itself, that each of them will submit a dispute to arbitration. It is intrinsic to the scheme by which a dispute will be resolved by a CETA Tribunal that the covered investor will not be in a contractual relationship with the state against whom the claim is brought, or at least that such a contract is not a prerequisite to the commencement of the arbitral process, and that jurisdiction does not depend on the privity of contract between the parties to the dispute. Contractual arbitration takes place in a context where the limits of liability and the scope of obligations are confined by agreement and tailored to an agreed objective.

12. The causes of action under CETA are found in Chapter Eight and comprise procedural actions including claims based on manifest arbitrariness, targeted discrimination, abusive treatment but also substantive causes of action which require the making of findings of fact, such as a claim for damages on account of unfair and unequal treatment or for breach of a legitimate expectation.

13. Nothing in CETA will prevent a disappointed investor from mounting a claim in the Irish courts for a remedy arising from the same set of facts that could form the basis of a claim before the CETA Tribunal, should the investor be in a position to establish the elements of the claim that it has been unfairly treated, or was treated in a discriminatory or unjust fashion. Irish domestic remedies are sufficiently robust to meet a claim of an investor who, for example, reasonably expects a particular course of action on account of a representation or promise made at the time the investment was made, or who can show that a domestic state measure is discriminatory, or amounts to an unfair expropriation of its assets. The disappointed investor has a range of remedies available at national level, and a claim by an investor will be heard by an independent and skilled judiciary, and in a legal system which operates in conformity with high principles of the rule of law in a democratic state. The Constitution demands nothing less.

14. By way of example, a claim for damages on account of a breach of a legitimate and binding representation may be maintained in domestic law under the doctrine of legitimate expectation. As a matter of domestic law, a person may maintain a claim for damages who can show that a public body has represented that it would follow a certain practice, from which a reasonably entertained expectation arose that the public body would act in a particular way in the future, and where it is shown that it would be unjust to permit the public authority to resile from it. This doctrine in Irish law is subject to public policy considerations to avoid the courts tying the hands of the Oireachtas.

15. CETA also makes provision for a claim by a disappointed investor under a general rubric of “legitimate expectation”, and the parameters and limitations of the claim are broadly similar to those in Irish law as noted in Article 8.10.4 of CETA:

“When applying the above fair and equitable treatment obligation, the Tribunal may take into account whether a Party made a specific representation to an investor to induce a covered investment, that created a legitimate expectation, and upon which the investor relied, in deciding to make or maintain the covered investment, but that the Party subsequently frustrated.” (emphasis added)

16. Article 8.9.3 of CETA goes on to say;

“For greater certainty, a Party’s decision not to issue, renew or maintain a subsidy:

(a) in the absence of any specific commitment under law or contract to issue, renew, or maintain that subsidy; or

(b) in accordance with any terms or conditions attached to the issuance, renewal or maintenance of the subsidy,

does not constitute a breach of the provisions of this Section.”

17. The same limitations are not apparent in other international investor protection treaties or agreements. By way of illustration, an International Centre for Settlement of Investment Disputes (ICSID) case, Cube Infrastructure Fund SICAV and ors v Kingdom of Spain Case, ICSID Case No. ARB/15/20 (19 February 2019) involved an investor tribunal considering whether Spain had created a legitimate expectation by enacting legislation that created a special regime of benefits and incentives. The tribunal held that there was no need for a direct or specific commitment for a legitimate expectation to arise in a highly regulated industry. A different decision was reached in Blusun and ors v Italian Republic, ICSID Case No. ARB/14/3 (27 December 2016) where a specific commitment was held to be to be necessary.

18. The CETA formulation of “legitimate expectation” is a considered and welcome limitation on the elements of the cause of action.

19. I would observe that, notwithstanding the availability of the CETA dispute resolution mechanism, it is probable that investor claims against the Irish State, or by an Irish investor against Canada, are likely to continue to be prosecuted in domestic courts, on account of their efficiency, their predictability in the application of established, understood and broadly similar legal principles, and because of their relative cost effectiveness. I agree therefore with the Chief Justice that the existence of the CETA Tribunal will not exercise a “gravitational pull” towards the resolution of investor disputes between a Canadian investor and the Irish State, and indeed as is apparent from the Vattenfall litigation (Vattenfall AB and Οthers v. Federal Republic of Germany (II), ICSID Case No. ARB/12/12), discussed in more detail below, the resolution of the dispute by arbitration may be less speedy, and potentially more costly, than its process through the Irish Courts.

20. Nonetheless, the constitutional question arising here is the potential that, by providing a different route to a legal remedy, the CETA Tribunals could impermissibly impact upon national sovereignty. While the precise parameters of any likely infringement may be more clearly assessed at the point at which the operation of the CETA Tribunal in an individual dispute comes to be considered, I with respect disagree with my colleague MacMenamin J.’s view that the challenge is premature. Although these issues present in an abstract way, I consider that this Court must here grapple with those issues, as the constitutional role of this Court compels an answer to the ratification question now before us.

The loss of judicial sovereignty?

21. The concern articulated by the present appeal is that proceedings for damages or restitution may be maintained before a CETA Tribunal when an action regarding the same facts is capable of being maintained at the level of domestic courts. It is precisely here that the first difficulty arises with regard to the impact of the creation of CETA Tribunals on domestic judicial sovereignty.

22. It is true, as noted by the trial judge, that the rights and obligations created by CETA operate in international law only, and may not be asserted in domestic courts, but only in the arbitral tribunals established under CETA. CETA principles and rights may not be invoked directly in the domestic legal system of a Member State. This is provided for in Article 30.6;

“1. Nothing in this Agreement shall be construed as conferring rights or imposing obligations on persons other than those created between the Parties under public international law, nor as permitting this Agreement to be directly invoked in the domestic legal systems of the Parties.

2. A Party shall not provide for a right of action under its domestic law against the other Party on the ground that a measure of the other Party is inconsistent with this Agreement.”

23. Butler J. expresses this eloquently at para. 154 of the High Court judgment ([2021] IEHC 600):

“[T]he fact that CETA is expressly framed so as not to have direct effect within the legal systems of the parties and the fact that the CETA Tribunal is separate from and outside the judicial systems of the parties means that disputes arising under CETA which the CETA Tribunal may determine are non-justiciable as a matter of Irish law.”

24. It is true too that a claim may not be maintained before a CETA Tribunal unless the investor withdrawals an extant claim in the domestic court, or if no such claim has commenced, waives the entitlement to commence such: Article 8.22.1(f) and (g). But the fact that a claim has concluded before a national court does not, as is discussed below, mean that a CETA claim may not thereafter be maintained by an investor.

25. The circumstances of the Vattenfall litigation illustrate this, albeit the claim was made under the Energy Charter Treaty (“ECT”). Legislation passed in the German parliament in 2011 proposed the abandonment of the use of nuclear energy in the country by 2022. The legislation made no provision for compensation to investors. Vattenfall, a Swedish investor in nuclear power plants in Germany, alleged that the operational lifetime of the power plants had been unlawfully abridged, and commenced proceedings in the German domestic courts, and at the same time commenced an investment arbitration against Germany under the ECT. The German Constitutional Court ultimately held that the state was obliged to preserve the legitimate expectation of investors, and further legislation was then enacted which, in turn, required the approval of EU Commission as it potentially infringed state aid rules.

26. The international arbitration claim against Germany under the ECT was for €4.7 billion in compensation on account of the acceleration of the phase-out of nuclear energy. The arbitration was lengthy and, ironically, because it is often perceived or expected that an arbitration will afford an investor a speedier remedy than court proceedings, the arbitration continued for many years until it was ultimately settled.

27. As my colleague Hogan J. noted, the litigation in Vattenfall illustrates a matter of deep concern regarding the interplay between an arbitral tribunal and the Court of Justice, but for present purposes the litigation also illustrates a number of factors in international investment litigation: the fact that the same set of circumstances can result in litigation in two fora, an international arbitration and a claim in a domestic court, and the fact that a legitimate policy decision to phase out nuclear power for considered public health and safety reasons following the Fukushima nuclear disaster exposed the German state to a claim for very large damages by a disappointed investor. The first of these factors is the one I now turn to examine as it concerns judicial autonomy. The second factor relates more to the different question in these proceedings concerning regulatory autonomy or what is sometimes called the risk of legislative “chill”.

Choice of jurisdiction is for the investor

28. CETA appears to envisage that, in claims by Canadian investors, the EU will itself be the respondent in most claims before an arbitral tribunal, as the Treaty of Lisbon vested in the EU competence over trade agreements and foreign direct investment. It is apparent that the ratification of CETA is intended to further the desire expressed in Article 207(1) of the TFEU for “uniform principles” in the conduct of this “external action.”

29. However, it is clear from the jurisprudence of the CJEU that CETA involves mixed competences or is a mixed agreement. While in practice it might be the case that most claims would be brought directly against the EU, or against Member States and the EU jointly, the fact remains that the investor tribunal under CETA will have the jurisdiction to make awards against the Irish State without the involvement of the EU as a defendant or respondent in those proceedings.

30. CETA envisages an investment arbitration court, the Investment Court System (ICS), which will be staffed with full time adjudicators or judges. The mode of appointment of the arbitrators and the requirements of impartiality and independence have hitherto not been found, or at least has not been common, in investor dispute arbitration under Investor-State Dispute Settlement (ISDS) to date when a separate arbitral tribunal is constituted for each case and that tribunal is not linked to any centralised legal authority. This is a welcome and positive choice, and allays some fears expressed by some academic writers concerning the make-up of the arbitral bodies. The ICS structure is designed to deal to an extent with the arguments concerning the independence and impartiality of CETA arbitrators.

31. The primary difficulty I apprehend with the CETA Tribunals is that a Canadian investor has a choice of jurisdiction: it may pursue an action in the CJEU, in the Member State, or before the CETA Tribunal. Proceedings which have not concluded in the Member State must be discontinued before a claim may be submitted to CETA, but there is no restriction on the right of an investor who unsuccessfully litigates in Ireland from then submitting a claim to CETA. The dispute then, on the same facts, will come to be adjudicated upon in two different fora and under two different standards with potentially quite different results.

32. If a Canadian investor seeks to resolve a dispute relating to a measure of the Irish State, the framework provided by the CETA dispute resolutions system would allow a dispute to bypass the Irish legal system, and seek relief in regard to that measure which is also within the jurisdiction of the Irish courts.

33. The CJEU in its judgment in Slowakische Republik v. Achmea BV C‑284/16, EU:C:2018:158), was concerned with this question, which it formulated as one concerning the possible impact on the autonomy of EU law. The Bilateral Investment Treaty (“BIT”) dispute resolution mechanism under consideration in this case was competent to give a final and binding interpretation of EU law. The Court ruled that the intra-EU investment agreement (CETA is not one such) was not in conformity with EU law because it could call into question the autonomy of EU law as the tribunal might be called upon to rule on the basis of domestic law or international agreements applicable between Member States, but nonetheless could not make a referral to the CJEU under Article 267 TFEU. The CJEU in Opinion 1/17 (ECLI:EU:C:2019:341), subsequently confirmed that the investment tribunals under CETA, because they are precluded from interpreting EU law, did not subtract from the autonomy of the CJEU (para. 122).

34. The Court of Justice in Opinion 1/17 concluded that a claim brought to the CETA Tribunal under CETA rules, laws or principles are not justiciable in the courts of the Member States because a litigant could not invoke those principles, rules, or laws in a domestic court. This is true insofar as it goes, but CETA creates an additional or alternative forum in which that litigation can happen.

35. I agree therefore with Dunne and Hogan JJ. that the vesting in the arbitral tribunal by CETA of jurisdiction to determine investor disputes against the State has the potential to remove from the Irish courts the jurisdiction they currently enjoy to resolve and determine those claims. Whilst the resolution by the CETA Tribunal would involve the application of different laws and principles, the subtraction or removal of jurisdiction arises because the arbitral tribunal will have the power to adjudicate upon state actions that occur within this State and to make findings of breach of obligations by the Irish State arising from the same facts that might ground proceedings in the Irish courts. As I turn now to examine, the possibility of a real, and in my view impermissible, breach of judicial sovereignty arises because the CETA award is “almost automatically” enforceable within the State, so that there may be cases where the award flies in the face of a final decision of the national court, or of a principle fundamental to our laws.

Automatic enforcement

36. It is intrinsic to CETA investor protection that the Member States will recognise and enforce any final awards of the CETA Tribunal. This factor is tied to the fact that CETA arbitral awards against the State will be enforceable in Ireland in a manner correctly described by Butler J. as “almost automatic”. The range of available defences is narrow, and enforcement may be denied in a limited range of circumstances such as lack of notice to a relevant party.

37. This factor is critical to an understanding of the place of an arbitral award in the domestic legal order. It is possible by reason of the fact that an award of an arbitral tribunal is automatically enforceable in the Irish courts, that the consequence of an arbitral award could be that the Irish High Court, Court of Appeal or Supreme Court, might be compelled to enforce an arbitral award that ran directly contrary to a finding already made by a court of last resort in Ireland. It is not in my view fanciful to imagine that an investor who fails in an Irish Court to obtain an award of damages on account of an unlawful state action, such as occurred in Cromane Seafoods Ltd v Minister for Agriculture [2016] IESC 6, [2017] 1 IR 119, might seek and obtain damages before a CETA Tribunal which would then come to be enforced in Ireland. The Irish court hearing an application for enforcement could not in the light of the provisions of CETA and the Arbitration Act 2010, as amended, refuse to enforce the award notwithstanding that in doing so it would contradict a decision of the Irish court on the same facts.

38. The CJEU in Opinion 1/17 was concerned with one question only, the compatibility of the dispute resolution mechanism in CETA with the autonomy of the EU legal system. It was not concerned with the law or constitutional order of the Member States, nor with the enforceability of the award of such Tribunal, and the fact that a CETA award will be enforceable under the New York Convention or the Washington (ICSID) Convention under the Arbitration Act 2010, is central to the correct understanding of how the CETA award will operate in Irish law.

39. I agree with the High Court (para. 150 of the judgment of Butler J.) that the CETA Tribunal will exercise the function of the administration of justice. However, I depart from her reasoning in that my view is that the ratification of CETA will have the effect of detracting from the jurisdiction of the Irish Courts. I consider that it is the combined effect of the scope and effect of the jurisdiction of the CETA Tribunal and the fact of automatic, or “almost automatic”, enforcement that makes the jurisdiction in a CETA Tribunal problematic.

40. The fact that leave is required to enforce an arbitral award under the Arbitration Acts 2010 does not in itself give the Irish courts power to refuse enforcement save in the very narrow sphere of statutory defences that are available and have been interpreted in a number of cases including that of Kelly J. Brostrom Tankers AB v Factorias Vulcano SA [2004] 2 I.R. 91. One cannot in my view read the decision to ratify CETA in isolation from the Act of 2010 and the fact that either the New York or the Washington Conventions is envisaged as the enforcement mechanism. This is not to say that the Act of 2010 is constitutionally frail, because s. 24 and 25 provide no more than a procedural enforcement mechanism. But in its present form, the Act of 2010 does not offer a sufficient safeguard to permit an Irish court hearing an application to refuse enforcement, where to do otherwise would risk the domestic court enforcing an award that conflicts in a material respect with a decision already made in a domestic court of final instance, or with a fundamental principle of our laws. The constitutional structure for which Article 34 provides is not protected in such circumstances.

Domestic law is a matter of fact

41. Thus far, I have concluded that the CETA arbitral provisions are capable of impacting upon the jurisdiction of the Irish Courts in two respects. A further difficulty presents as CETA expressly provides that Canadian law, EU law, and the law of an individual Member States, must be treated as a matter of fact, and precludes the CETA Tribunal from interpreting that law or making any determination on the legality of such law. Article 8.31.2 of CETA provides that “the Tribunal shall follow the prevailing interpretation given to the domestic law by the courts or authorities of that Party”, which is understood to mean that it does not engage with the reasoning of the domestic court, the principles behind the impugned statute or common law.

42. In Opinion 1/17, the CJEU examined the meaning of Article 8.31.2, finding that in relation to European Union law, “in order to assess whether there is an infringement of [CETA] … the Tribunal will have to confine itself to an examination of EU law as a matter of fact and will not be able to engage in interpretation of points of law.” (para. 76). The Court went on to hold that the decision of CETA will, because of Article 30.6, have no direct effect in domestic legal systems, because they “operate within legal orders that are wholly separate” (para. 77).

43. In Opinion 1/17 at para. 122, the Court held;

“It follows that the power of interpretation and application conferred on that Tribunal is confined to the provisions of the CETA and that such interpretation or application must be undertaken in accordance with the rules and principles of international law applicable between the Parties.”

44. Advocate General Bot’s opinion in this case was that the limited scope of review or appeal meant that the Tribunal must interpret the parties' domestic law “as little as possible.” (para. 150).

45. But that proposition cannot be said to operate as a universal restriction. While CETA nominally prohibits CETA Tribunals from engaging in interpretation of domestic law, the Court of Justice acknowledged at para. 131:

“[T]he CETA Tribunal, when it is called upon to examine the compliance with the CETA of the measure that is challenged by an investor and that has been adopted by the investment host State or by the Union, will inevitably have to undertake, on the basis of the information and arguments presented to it by that investor and by that State or by the Union, an examination of the effect of that measure.”

46. It is unclear how this “examination” would work in practice. If an investor challenges a newly-introduced legislative measure, and chooses to bypass the domestic legal process as permitted under CETA, there will be no “prevailing interpretation” of domestic law available to a CETA Tribunal.

47. The case of Southern Pacific Properties (Middle East) Limited v. Arab Republic of Egypt, ICSID Case No. ARB/84/3 (Decision on Jurisdiction - 14 Apr 1988), acknowledged this lack of clarity when called to resolve a dispute involving an Egyptian measure, stating;

“As to [Article 8 of Law No. 43 of 1974] itself, the Claimant's contention that this provision of municipal law should be treated as a "fact" is not helpful. The Parties are in fundamental disagreement as to what Article 8 means and the Tribunal therefore must interpret Article 8 and determine its legal effect in relation to the Washington Convention.” (para. 58).

48. Thus, if parties to an CETA arbitration are not in agreement as to the meaning of a domestic measure, it will be necessary to resolve this dispute, and this in my view necessarily involves the interpretation of domestic measures. The CJEU in Opinion 1/17 which held that the power of interpretation was confined to the provisions of CETA did not deal with this separate question of the interpretation of domestic legal provisions when there is no apparent “prevailing interpretation” of domestic law, or where there is disagreement as to what it means.

49. The fact that the meaning of national measures may be contested before a CETA Tribunal is considered in CETA Investment Law, edited by Bungenberg and Reinisch. At pages 733-734, the authors explain this difficulty:

“Considering domestic law as fact suggests that it will be up to the disputing Parties to submit relevant evidence such as domestic case law. Hence, the Tribunal would not be obliged to base its appreciation of domestic law on material beyond the sources provided by the disputing Parties, Consequently, when deciding questions of domestic law, the Tribunal may also take into account the burden of proof which is for the disputing Parties to discharge.”

50. The authors note that there may be “exceptional circumstances” in which the Tribunal would have to depart from the interpretation provided by the respondent state, such as where a state puts forward a manifestly incorrect interpretation of its domestic law, but the true difficulty it seems to me is that if there is a contest as to meaning, the tribunal will have to resolve that dispute and itself come to a view as to the meaning of a domestic provision. In turn that meaning informs an award that comes to be enforced in the domestic court, even if the meaning attributed to a domestic legal provision is manifestly wrong.

51. The problem is compounded by the power of the appellate tribunal to review awards on the grounds of “errors in the application or interpretation of applicable law” as well as on grounds of “manifest errors in the appreciation of the facts, including the appreciation of relevant domestic law”. This will invariably require the appellate tribunal to consider the meaning of a particular provision of domestic law, and to “appreciate” its nuances (Article 8.28.1). That means that the arbitral tribunal and in particular the appellate tribunal, could not be said to be engaged in the exercise of “neutrally” applying domestic law as a matter of fact, particularly in the context of a common law jurisdiction where the understanding and application of the law invariably, and of necessity, requires argument and the distinguishing of different threads in every principle.

52. In his long article, Beyond the Pledge: The Imperfect Legal Framework for Enforcing Awards of the CETA Investment Court against the European Union (McGill Journal of Dispute Resolution Vol. 7 (2020-2021) No. 4), Leo Butz makes some observations regarding the meaning of the requirement in CETA that national law be considered “as a matter of fact”. I agree with his observation that all consideration of law necessarily entails some form of interpretation (at p.115). The meaning or proper interpretation of Irish domestic provisions when these are either contested, or not yet established, is to be gleaned from a reading of case law and judgments, the import of which is often not possible to discern except in its application to a particular case. Opinion 1/17 affords little assistance in regard to this more difficult interpretative method.

53. Whilst Article 8.31.2 clarifies that “any meaning given to domestic law by the Tribunal shall not be binding upon the courts or the authorities of that Party” and is consistent with the fact that CETA cannot be directly invoked in the domestic legal systems, and complements the prohibition of parallel domestic court proceedings pursuant to Article 8.22.1(f) and (g), the fact that this interpretative power is one capable of being engaged is a direct encroachment on domestic law and a subtraction of the power currently vested in the Irish courts as the sole interpreter of domestic law, precisely because the CETA award is enforceable irrespective of the correctness of the underlying interpretation of domestic measures.

Risk of discordance

54. Another problem also presents from the decision of the ISCID Tribunal in Vattenfall, mentioned above. There, the Tribunal was considering the implications of the CJEU decision in Achmea that the BIT investor dispute mechanism was incompatible with EU law. The Tribunal decided that it did have jurisdiction, notwithstanding the view of the CJEU in Achmea, holding that the principles of EU law were not applicable to the interpretation and application of the ECT because the applicable law was that of that Treaty itself and Article 25 of ICSID.

55. At paragraph 167 of the Decision on the Achmea Issue, ICSID Case No. ARB/12/12 (31 August 2018), the Tribunal said:

“EU law and the ECJ judgment [in Achmea] may not be “taken into account” for the purpose of the so-called harmonious interpretation of Article 26 ECT that would exclude intra-EU investor-State arbitrations.”

56. The law of the ECT prevailed, notwithstanding the view of the CJEU that the ECT provisions could not be invoked in an intra-Member State dispute. Butz anticipated this problem and notes the uncertainty regarding what should happen if an investment court bases an award on an erroneous understanding of CJEU case law, or in circumstances where there is not yet an established interpretation of a provision of EU law that is relevant as a matter of fact to the decision of the investment court.

57. The Vattenfall ruling shows that the Tribunal considered its jurisdiction to be wholly independent of, and to operate outside, the law of the EU, and wholly in the international sphere, to the effect that the findings of the CJEU could not and did not influence the result of the jurisdictional question.

58. This is a most disquieting result, especially when seen against the backdrop of the almost automatic enforcement in domestic law of a potentially large award by a CETA Tribunal, under an erroneous interpretation of Irish law, and where the Irish courts have not afforded an opportunity to interpret those laws. The net effect is that a CETA Tribunal can make an award that entirely ignores Irish domestic law, save and insofar as it must consider this as a “matter of fact’ and where the tribunal itself must resolve any dispute as to the meaning of that “fact” which inherently entails some form of interpretation of domestic law.

Exhaustion of domestic remedies

59. There might be concern that the approach of the majority of this Court regarding the CETA Tribunal might impact on the State’s accession to the ECHR and the jurisdiction of the European Court of Human Rights (the “ECtHR”). The decisions of the ECtHR, whilst not enforceable in domestic law, will be, and historically have been, accepted by government and remedies have been afforded to litigants who are successful in the Strasbourg court. The awards bind the State at an international level but not in domestic law.

60. A significant element, which is both procedural and substantive, of the jurisdiction of the ECtHR to hear a complaint against a state is that the litigant has exhausted domestic remedies. No such requirement exists for an applicant before the CETA Tribunal. The requirement to exhaust domestic remedies respects the fundamental role of the national judicial systems (and domestic law) in the operation of the Convention. But the matter goes further than this and the Strasbourg Court is to be seen as being engaged in a dialogue with the national court, and has available the reasoning and detailed treatment of the facts and legal principles underlying the claim. This not only has the effect of crystallising the problem, and discarding any undergrowth of irrelevant or secondary facts, but the Strasbourg Court has the benefit of the assessment, reasoning and arguments of the domestic courts, and will have regard to the reasoning of the domestic court, even if this is only to disagree, or find that for reasons of principle it must depart from that reasoning, or agree only in part. The dialogue between the domestic court and the Court in Strasbourg is a mutual dialogue of reasoning and the mutual respect for the separate sovereignty of the other.

61. The CETA Tribunal will have no such dialogue. It is to treat domestic law as a “matter of fact”, which must mean that it does not engage with the reasoning of the domestic court, the principles behind the impugned statute or common law. Law operates within the realm of reasoning and a reasoned discourse is an essential element of its legitimacy.

62. It may seem almost ironic that the feature of the CETA dispute resolution mechanism that led the CJEU to conclude that the sovereignty of the EU legal order would not be impacted by the existence of the dispute resolution mechanism for which CETA provides, is that important feature of the relationship between the ECtHR and the Irish courts which affords respect for sovereignty.

63. The practice of the Strasbourg Court is not to be seen as one of procedure only, and the requirement of the exhaustion of domestic remedies is a substantive element of the jurisprudential basis of that Court’s decision making. The Convention in its operation recognises the sovereignty of domestic courts. The dialogue that exists is seen as fundamental to the system, both because national authorities, including courts, must respect the Convention, but also because the Convention in turn respects national courts and the role of national judges. The principle of subsidiarity is an express principle European Union law but is also embedded in Convention thinking (see paper entitled Subsidiarity: a two-sided coin? - Jean-Marc Sauvé, Speech at Seminar to mark the official opening of the ECtHR 30 January 2015).

64. The Court in Strasbourg does have a power to interpret domestic law although it will use that power very sparingly. In S.A.S. v France App no. 43835/11 (1 July 2014) the Court observed:

“national authorities have direct democratic legitimation and are, as the Court has held on many occasions, in principle better placed than an international court to evaluate local needs and conditions. In matters of general policy, on which opinions within a democratic society may reasonably differ widely, the role of the domestic policy-maker should be given special weight” (para. 129).

65. Under the Convention, the national authority must have the fullest opportunity to address a Convention complaint before that complaint can be admitted to the ECtHR. In contrast, the CETA Tribunal has no such principle, no “margin of appreciation” can have an effect on the result, there is neither procedural nor substantive recognition of the reasoning of the domestic court, nor of the reasons behind domestic law.

66. The CETA Tribunal has no dialogue or engagement with the reasoning of the domestic court, or the reasons behind the principles of domestic law in the relevant area.

67. The possibility of different results from domestic courts and a CETA Tribunal resulting from the same dispute, and the resulting tension is troubling, but what is more troubling is the fact that the CETA Tribunal would not engage with the reasoning and principles operating in domestic law, and are not engaged in the type of dialogue envisaged in the ECtHR where competing interests and a recognition of the sovereignty of national courts affords a more complex approach to sovereignty than the rather blunt system established in CETA.

68. In an article published in 2014, “Study on Investor States Disputes Settlement (ISDS) and Alternatives of Dispute Resolution in International Investment Law” (at pp. 91-92), Hindelang suggests that a more societally acceptable approach might be to require that an investor do exhaust local remedies before proceeding to arbitration, subject to the provision that the “local remedies rule” could be relaxed where an argument can reasonably be made that the domestic system falls short of certain criteria previously specified in an investment instrument.

69. CETA does not encourage the use of domestic courts, albeit the potential problem created by parallel claims in domestic and international fora is dealt with in part only by the requirement to waive domestic claims before pursuing arbitration. The absence of such a positive approach to domestic law to my mind means that the sovereignty of domestic courts is not sufficiently acknowledged or recognised in CETA.

Preservation of the Right to Regulate/Chill and Article 15 of the Constitution

70. CETA recognises the right of contracting parties to regulate for legitimate public policy objectives. The text of CETA is an improvement on that found in other international treaties, where no such express reservation is usually found. Any measure in the form of a law, regulation, rule, procedure, decision, administrative action, requirement practice, or any other form of measure by a state actor is one capable of giving rise to a claim by a qualifying investor. In turn, Article 8.9.1 of CETA is an affirmation of this right to regulate by domestic authorities, and this in itself is a recognition of sovereignty and is unconditional. It does not mean that damages are not recoverable by an investor arising from such regulation, but rather means that the economic interest of the investors are balanced with the sovereign rights of states to regulate.

71. In Opinion 1/17, the CJEU was considering inter alia whether the fact that the CETA Tribunal will in essence be balancing the business interests of an investor against the public interest in regulation undermines the constitutional framework of the EU (para. 137). The Court was satisfied, in particular because of Article 8.9, that the constitutional framework of the EU was preserved because the investor tribunals could not “call into question the level of protection of public interest determined by the Union following a democratic process” (para. 156). The CJEU went on to say that the tribunals have no jurisdiction “to call into question the choices democratically made” (para. 160), and notwithstanding that the investor tribunal can award damages on account of the taking of those sovereign steps.

72. The CETA Tribunals could award compensation, but they could not direct a state or the EU, to amend, withdraw or not enact legislation. The CJEU considered that the Member State’s capacity to operate autonomously was therefore not infringed.

73. Much of the critical literature on CETA has as its focus the risk that an award made by a CETA Tribunal could have a chilling effect on domestic legislation, and concerns are raised as to the real risk that environmental legislative action, especially in the context of climate change and the implementation of Paris Treaty carbon obligations, might be delayed or abandoned. “Chill” in this context is not merely the direct effect on domestic law of an award to a disappointed investor on account of the enactment of the provision. “Chill” can also be anticipatory, as is illustrated by the fact that New Zealand lawmakers are said to have delayed the planned imposition of plain packaging for tobacco products pending the completion of the Philip Morris investor claim against the Australian government (Philip Morris Asia Limited v. The Commonwealth of Australia, UNCITRAL, PCA Case No. 2012-12).

74. The fact that a CETA claim can result in the award of compensation where Irish domestic law would not, for example when no mala fides can be shown in the regulatory decision, also can impact on national policy making because the arbitral tribunal has no obligation to consider the reasons for, or the reasonableness of, a domestic measure.

75. Take but one example, a government who propose to enact environmental legislation should be in a position to ascertain the likely risks of a claim arising from a complete ban on the cutting or the using of turf in electricity generating stations. This is a live topic in the context of environmental protection, and a legislature contemplating the enactment of such legislation may feel itself constrained in its decision-making if it is not readily apparent on examination whether there exists a possible risk of a claim by a Canadian investor in a domestic bog which could result in the State being liable to pay compensation as a result of the discontinuation of cutting or the burning of turf. This is precisely what happened following the decision of the Supreme Court of the Netherlands in The Netherlands v. Urgenda (ECLI:NL:HR:2019:2007), where German coal suppliers mounted a claim under the ECT against the government of the Netherlands for breach of their legitimate expectation that they could expect to continue to supply coal to Dutch coal burning stations. That claim remains unresolved, but it is a useful illustration of the type of claim that could be brought, and of the possible impact on legislative and executive decision making, or the domestic implementation of EU obligations.

76. A government deciding to close a coal or turf burning power generator, or acting on an EU requirement to do so, may later be faced with an unintended and unanticipated consequence, such as war or financial catastrophe, for which no provision was made in that decision. All decisions involve the closing of certain doors, and choices made which either expressly or implicitly preclude or are capable of precluding other choices.

77. Whist my colleague Hogan J. makes an eloquent argument regarding the potential impact upon legislation of a claim under CETA by a Canadian investor, I consider that the “chill” effect does not of itself impact upon legislative autonomy. I agree with the Chief Justice and with my colleague Dunne J. that that the risk of a direct or indirect chilling effect on legislation does not make the ratification of CETA impermissible because the power to make legislative choices is not impaired. CETA does not make laws for the State so as to evoke the provisions of Article 15 of the Constitution.

78. The ratification of CETA does not for this reason in my view result in the making of laws for the State nor is there a direct impact upon the powers of the Oireachtas, or indeed of executive decision-making, within its permissible realm.

The Joint Committees

79. The role of the Joint Committees is two-fold: to interpret CETA: Article 26.1.5(c); and the power to “adopt interpretations of the provisions of the Agreement, which shall be binding on” the CETA investor tribunals: see Article 26.1.5(e).

80. The fact that the Joint Committees may provide a binding interpretation of a provision of CETA involves the loss of democratic control over the text of CETA and is troubling for that reason.

81. The power of amendment has been invoked on foot of a recent request by the Federal Government of Germany in regard to the meaning of “fair and equitable” treatment in Article 8.10, which has resulted in a proposal that could, if it were adopted, limit the scope of a challenge to state action to combat climate change and which inter alia requires the CETA Tribunal to “take due consideration” of the climate neutrality objectives of the parties to the Paris Agreement, and in a manner that “allows the Parties to pursue their respective climate change mitigation and adaptation measures”.

82. The request from the German Federal Government was for a “more precise definition” of the concepts of “indirect expropriation” and “fair and inequitable treatment” of investors. A draft text of a binding interpretation has been agreed between the Commission and the German Federal Government which is proposed to be submitted to the Member States for agreement. The modified text improves the respect for the sovereignty of the Member States and the EU both in regard to the legislative and judicial functions. For example, the CETA Tribunal may take notice of the existence of a remedy at national law which was not availed of by an investor, and this answers some of my concerns arising from the absence in the current text of a positive encouragement to avail of domestic remedies. Further, the more restrictive scope of the grounds of challenge are capable of addressing the perceived risk of legislative “chill” in climate legislation.

83. The concern is that the interpretative function is one that is in practice capable of altering the text of CETA and the mutual obligations of the parties without any democratic input from the sovereign states, and I agree with my colleague Hogan J. that this has an impermissible impact upon national sovereignty.

84. The fact that the meaning of the text of CETA may be determined by the Joint Committees illustrates the lack of democratic accountability in the ongoing analysis of scope and meaning of the text of the treaty.

85. I do not say that the interpretative powers means that the Joint Committees can make legislation. Rather, the problem I perceive with this broad power of the Joint Committees is that the interpretative function can result in a change to the text. Whilst I agree that the purpose of the establishment of the Joint Committees, as the Chief Justice notes in his judgment, may be to prevent overreach by an arbitral body, and to limit the risk the possibility of inconsistent or unduly expansive interpretations, the powers of the Joint Committees are not constrained in their exercise by this objective. The realpolitik of the inter-EU relations may, and almost always would, restrict or limit an overreach, but the potential exists that the interpretations could involve an alteration of the text of the Treaty without full agreement by the Member States, and without domestic oversight.

Possible way forward

86. I have carefully considered the analysis of Hogan J. that proposes the amendment of the Arbitration Act 2010 and I agree with him, and my five other colleagues that a suitable amendment may be, subject to the views of the Government and of Oireachtas, capable of providing a solution to a constitutional challenge. By providing a power in the Irish Courts to review in certain cases a decision of the arbitral tribunal before enforcement, the Irish Courts may be capable of protecting the constitutional identity of the Irish courts.

The questions raised in the appeal

87. I propose therefore the following answers to the questions identified by Dunne J. in her judgment:

1. Is ratification of CETA necessitated by membership of the EU? No.

2. (a) Is ratification of CETA incompatible with Article 34 in the basis that it impermissibly withdraws disputes from the jurisdiction of Irish Courts? No.

(b) Is ratification of CETA incompatible with the finality of decisions of the Irish Courts under Article 34? Yes.

3. Is ratification of CETA incompatible with the legislative sovereignty of the State under Article15.2? No.

4. Is Ratification of CETA incompatible with the democratic nature of the State under Article 5? Yes.

5. Would amendment of the Arbitration Acts permit ratification of CETA? Yes.

Result: Allow appeal and set aside order of the High Court

	[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	Supreme Court of Ireland Decisions
You are here: BAILII >> Databases >> Supreme Court of Ireland Decisions >> Costello v The Government of Ireland, Ireland and the Attorney General (Approved) [2022] IESC 44_4 (11 November 2022) URL: http://www.bailii.org/ie/cases/IESC/2022/2022IESC44BakerJ.html Cite as: [2022] IESC 44_4

Supreme Court of Ireland Decisions