Constitutional Identity, Legal Autonomy, and Sovereignty

Introduction

Mr Costello, a Green Party deputy to the lower House of Ireland’s legislature, challenged Ireland’s proposed ratification of the Canada–EU Comprehensive Economic and Trade Agreement (CETA), a mixed agreement requiring ratification by both the EU and all of its member states. ¹ A Supreme Court panel of seven judges, each delivering a judgment and with different majorities forming on each issue, ultimately held in Mr Costello’s favour. By a majority of 4:3, the Supreme Court held it would be unconstitutional for Ireland to ratify CETA. Three of the majority judges also held, however, that ratification could be constitutionally permissible if the Irish Arbitration Act 2010 were amended to allow the High Court to refuse to enforce CETA Tribunal awards that were inconsistent with either Ireland’s constitutional identity or Ireland’s EU law obligations. While not themselves considering this ‘constitutional cure’ necessary, the three minority judges agreed that it would solve what the majority considered to be the constitutional problem.

Costello has potential implications for European constitutionalism on three different levels. Most immediately, it has complicated efforts within the EU to secure ratification of CETA, contributing to the political debate over the desirability of investor-state dispute settlement. Second, the majority judgments recalibrate the balance between Ireland’s constitutional commitments to national sovereignty and international relations, potentially reducing the range of international obligations that can be undertaken on Ireland’s behalf by the government and/or legislature without need for constitutional amendment. ² This recalibration may impede EU ratification of other mixed agreements in the future. Finally, with the concept of ‘constitutional identity’ receiving its first judicial endorsement in Ireland, Costello raises at least a question about whether the Irish courts might assert an identity-based competence to circumscribe the scope of EU law in Ireland.

There is a mismatch, however, between the significance of Costello’s implications and the coherence of its reasoning. The endorsement of the constitutional cure by only three of the majority judges materially qualified the basis on which they held ratification unconstitutional, such that there was no ground of unconstitutionality that attained majority support. Moreover, there were significant inconsistencies within and between those three majority judgments with no coherent explanation provided of how a CETA award in the future might infringe Irish constitutional identity, such that ratification of CETA was now unconstitutional. The concept of constitutional identity articulated by two of the judges lacked definition and, notwithstanding its role in the constitutional cure, received no substantive consideration from the other five judges. Partly for these reasons and notwithstanding some academic speculation to the contrary, the more significant implications of Costello are unlikely to materialise. The Government will remain empowered to conduct Ireland’s international relations post-Costello in much the same way as pre-Costello; constitutional identity will not be deployed to ground restrictions on the competence of the Court of Justice. Costello’s greatest relevance will be its immediate consequences, the Supreme Court having rendered Irish ratification of CETA a near insoluble conundrum.

I begin by outlining how investor-state dispute settlement functions under CETA, its critique and how the Court of Justice assessed that critique in Opinion 1/17. The following section outlines the Irish constitutional background. The next section provides a detailed account of the Supreme Court’s judgment and the constitutional cure in an attempt to identify constitutional propositions that may have implications for future doctrinal development. The case note then gauges the implications for Ireland’s ratification of CETA before assessing whether the endorsement of constitutional identity will affect Ireland’s membership of the EU.

Canada–EU Comprehensive Economic and Trade Agreement

Investor-state dispute settlement under CETA

CETA seeks to open the economies of the EU and Canada to each other through a wide range of measures designed to ease cross-jurisdiction investment. As a mixed agreement, it requires ratification by both the EU and its member states. Those parts of CETA falling within the exclusive competence of the EU have been provisionally applied since 2016. ³ Not yet applied is Section F of Chapter Eight, relating to investor-state disputes. Investors are protected against discriminatory treatment and state measures that could diminish the value of their investment. CETA repeatedly asserts, however, the right of the parties to regulate within their territory to achieve legitimate public policy objectives, such as environmental protection. ⁴ Section F establishes Tribunals for the resolution of investor complaints. A foreign investor aggrieved by a domestic measure can challenge that measure under the domestic law of the investment country. However, the investor can also elect to take a CETA case directly to a CETA Tribunal, either instead of domestic proceedings or after those proceedings have concluded or been withdrawn. ⁵ Although disputes would be submitted to Tribunals operating under the rules of international investment arbitration agreements, CETA grafts on several features relating to qualifications, membership and remuneration of Tribunal members that render the Tribunals closer to an international judicial mechanism than ad hoc investment arbitration.

The subject matter of CETA disputes is the effect on investors of domestic measures, but CETA jurisdiction lies outside domestic law. The Tribunals interpret the CETA Agreement itself but are explicitly precluded from interpreting domestic law, which they must instead treat as a matter of fact, ⁶ following prevailing interpretations adopted by courts or authorities of that party; nor can they rule on the validity of any domestic law or measure. Tribunal awards of compensation can, however, be enforced under either the New York Convention ⁷ or the Washington Convention, ⁸ creating legal obligations within the domestic system. The New York Convention, implemented in Irish law by section 24 of the Arbitration Act 2010, allows refusal of enforcement where it would be contrary to public policy. ⁹ Section 25 of the Arbitration Act 2010, faithfully implementing Article 54 of the Washington Convention, appears more favourable to investors, requiring that the pecuniary obligations of an award under that Convention ‘shall, by leave of the High Court, be enforceable in the same manner as a judgment or order of the High Court to the same effect’. The Supreme Court majority in Costello considered that these enforcement mechanisms, with an investor free to lodge their claim under the rules of either convention, rendered CETA awards near-automatically enforceable in the state.

The Investor-State Dispute Settlement critique

Investor-State Dispute Settlement is one of the most controversial features of international economic law. The broadest critique is that the possibility of investment arbitration causes regulatory chill, leading governments and legislators to avoid policies that serve the public interest for fear of having to compensate foreign investors. ¹⁰ Promoting domestic compliance with agreed international norms and protecting investors from potentially inefficient or biased legal systems in the host country is scarcely controversial, but investor-state dispute settlement is criticised for features that may support an over-emphasis of investor interests at the expense of the public good. In this regard, criticisms include the discrimination against domestic investors who cannot access arbitration, the incentive for arbitrators to find against states in order to encourage future complaints and hence remuneration for themselves, the private law character of arbitration addressing core public law concerns, the lack of any objective standard for balancing economic rights with public goods, and the lack of state control over arbitral bodies. ¹¹ The CETA system – seen by the EU as an embryonic investor court system to replace investor-state dispute settlement – partially responds to some of these criticisms, with greater institutionalisation of the arbitration process and greater weight for public good as a ground for restriction of investor protections. ¹²

The Court of Justice consideration of CETA

The Court of Justice protects the autonomy of the EU legal order both internally, configuring relationships between the Union and the member states and between the EU institutions, and externally, controlling the external actions of both the Union and the member states.¹³ The Court of Justice’s concern to protect its own role as the authoritative interpreter of EU law contributed to its conclusion that both the Draft Accession Agreement of the EU to the ECHR¹⁴ and a bilateral investment treaty between the Netherlands and Slovakia¹⁵ were incompatible with EU law. CETA was carefully designed to avoid this problem, contributing to the Court of Justice’s conclusion in Opinion 1/17 that Section F did not infringe the autonomy of the EU legal order.¹⁶ Some commentators, however, also detect a greater openness on the part of the Court to international engagement.¹⁷

The Court accepted that independent arbitration mechanisms, giving binding interpretations of international law, were probably necessary for international agreements to function.¹⁸ CETA’s careful separation of international and EU law – in particular that CETA Tribunals are precluded from interpreting or applying EU law¹⁹ – avoided the structural infringement of autonomy that had been fatal in previous cases.²⁰ The Court also required, however, that CETA could not have the power to make awards that ‘might have the effect of preventing the EU institutions from operating in accordance with the EU constitutional framework’.²¹ Academic commentators have characterised the Court as either introducing or re-emphasising a dimension of substantive autonomy alongside the structural autonomy that had dominated in earlier cases.²² Whereas structural autonomy assesses whether a particular competence overlaps an EU competence, substantive autonomy makes a more rounded assessment of the impact of an international agreement. Rapoport describes it as a broad concept of autonomy aimed at ensuring that ‘the agreement will not prevent the institutions from functioning in accordance with their constitutional framework and, more specifically, from freely determining which public interests the EU intends to protect and in what way’.²³

The Court perceived a risk that CETA Tribunals might determine the public interest did not support a measure complained of, creating ‘a situation where, in order to avoid being repeatedly compelled by the CETA Tribunal to pay damages to the claimant investor, the achievement of that level of protection needs to be abandoned by the Union’.²⁴ The Court, however, emphasised the CETA entitlement of the parties to legislate in the public interest, concluding that CETA Tribunals had ‘no jurisdiction to declare incompatible with CETA the level of protection of a public interest established by EU measures’ adopted following the EU’s democratic process and subject to legal review by the Court of Justice itself. This conclusion has been characterised as demonstrating ‘a degree of trust’ in relation to how CETA Tribunals will function²⁵ and as giving Tribunals the ‘benefit of the doubt’.²⁶ While CETA Tribunals may show deference to measures deemed by the parties to be necessitated by the public interest, they are likely to assert the competence to determine the CETA-compatibility of such measures. Perhaps the Court of Justice’s assertion is best understood as a jurisdictional shot across the bows of CETA Tribunals: ‘as long as the CETA Tribunals refrain from calling into question the level of EU public interest protection, [the Court of Justice] will not internally challenge international law through the narrative primacy of EU law’.²⁷

The Court’s Opinion and related academic commentary anticipate many of the issues that were to arise before the Irish Supreme Court. The careful construction of CETA to ensure that the CETA Tribunals’ interpretative role did not overlay that of the Court of Justice ensured no structural infringement of legal autonomy. Nevertheless, the existence of CETA had the potential to threaten the operation of the EU’s law-making and adjudicatory processes. The Court enhanced its protection of substantive autonomy to assess that threat, albeit ultimately concluding that the threat was not excessive. Some have criticised the Opinion as insufficient to protect against the regulatory chill likely to be effected by CETA,²⁸ prioritising free and fair trade over the rule of law.²⁹ The opposing view is that Opinion 1/17 properly recognises the need for reciprocity in international relations and neutrality in international adjudication,³⁰ thereby enabling the EU to enter into international trade agreements.³¹ While conducted in the somewhat different language of sovereignty, the judgments of the Irish Supreme Court in Costello echo these concerns.

Irish constitutional background

Article 29 of the Constitution commits Ireland – a ‘sovereign, independent, democratic state’ per Article 5 – to the ‘ideal of peace and friendly co-operation amongst nations’, affirms the principle of pacific settlement of disputes by international arbitration or judicial determination, and accepts generally recognised principles of international law. The Government is empowered to conduct Ireland’s external relations, but international agreements – other than those of a technical and administrative character – must be laid before the Dáil (lower House) and require its approval if, as with CETA, they involve a charge upon public funds. No international agreement becomes part of domestic law unless enacted by the Oireachtas (Parliament). This framework reflects a simple division between the external and the domestic: the Government acts internationally; the Dáil and the Oireachtas become involved only where international obligations incur or require domestic consequences, whether financial or legal.

The supranational EU could not be accommodated within this framework. The Constitution was amended in 1972 to authorise Ireland’s membership of the European Communities. In Crotty v An Taoiseach, assessing a challenge to ratification of the Single European Act, the Supreme Court held that this authorisation encompassed treaty changes that did not alter the essential scope or objectives of those Communities. ³² But Title III of the Single European Act, concerning co-operation between member states in the field of foreign policy, exceeded that authorisation by moving from the economic to the political realm. A majority of the Court considered that ratification of Title III would unconstitutionally fetter the freedom of the Government to conduct foreign policy on behalf of the state. The minority dissented on the basis that Title III was too aspirational and vague to constrain the Government in any meaningful way. In Pringle v Government of Ireland, the Supreme Court rejected a challenge to Ireland’s ratification of the European Stability Mechanism Treaty, which committed Eurozone member states to provide funding to rescue another member state that faced insuperable financial difficulties.³³ A majority emphasised that Article 29.4 envisaged Ireland undertaking treaty obligations, necessarily constraining future freedom of action. The Crotty limitation, the majority reasoned, concerned not the undertaking of international law obligations per se but rather the transfer of a policy-making competence to an international organisation.

The Crotty and Pringle judgments establish that the Government, Oireachtas and Dáil cannot approve Treaties that abrogate aspects of sovereignty, understood as competences otherwise assigned by the Constitution to domestic organs of government. This analysis reflects a structural protection of sovereignty: unconstitutionality consists in the assignment of a competence to an international organisation that overlaps with a competence assigned to an Irish organ of government, without attention to the substance of how that competence might be exercised. The difference between the majority and minority in Crotty can be understood as the majority emphasising the structural overlap of Title III with a constitutional competence rather than the question of whether this overlap would substantively affect the capacity of the Government to exercise its foreign affairs power. The fact that Crotty concerned the sole external constitutional competence risks obscuring the underlying constitutional principle, but the same sovereignty logic underpinned the 1972 decision to seek constitutional authorisation for accession to the Communities as well as subsequent decisions to amend the Constitution prior to ratifying later Treaties. Assigning a competence to make laws that apply in Ireland or judicial determinations that have force in Irish law would unconstitutionally interfere with competences assigned to the Oireachtas or the Irish courts respectively. Investor-state dispute settlement does not fit easily within the structural sovereignty framework established by Crotty and Pringle, still less the simple dichotomy between internal and external action endorsed by the original constitutional text. The challenge facing the Supreme Court in Costello was whether to scrutinise CETA solely for discrete structural infringements of sovereignty in line with Crotty and Pringle or – analogously to the Court of Justice – develop a substantive account of sovereignty that could allow for a more rounded assessment of the challenge posed by CETA to national governance systems.

Before addressing the Supreme Court’s decision, it is helpful to outline some general features of constitutional litigation in Ireland that shaped the Supreme Court’s approach. Ireland operates a dispersed system of constitutional review, where constitutional issues are raised through litigant-initiated proceedings in the ordinary courts. While the President may refer a Bill to the Supreme Court for a binding decision on its constitutionality prior to signing – a power that has only been exercised 16 times – there is no procedure for pre-ratification review of treaties. Instead, proceedings must be initiated by private individuals in the High Court, under rules of standing that allow measures to be challenged where they affect everyone in general but no-one in particular. Constitutional review of treaties is infrequent, Crotty and Pringle being rare but important exceptions. The constitutional implications of investor-state dispute settlement had never before been considered by the Irish courts, notwithstanding that the state had previously ratified treaties that involved investor-state dispute settlement. Because ratification can entail international law obligations, the Irish courts are prepared to anticipate the likely effect of a legal measure, an approach in tension with their general preference for assessing the measure’s actual effect in a concrete case. This anticipatory character was especially apparent in Costello, where much of the concern lay with how CETA Tribunals might interpret the balance between investor protection and the public interest as determined by contracting states.

The Supreme Court decision in Costello

The ground of unconstitutionality

While the core political critique of investor-state dispute settlement focuses on regulatory chill, the constitutional translation of this critique did not find favour with the Supreme Court in Costello. A majority held that legislative autonomy was not infringed by the potential chilling effect of a CETA Tribunal ruling, because legal freedom to legislate was not affected. ³⁴ A majority also rejected claims that democracy and sovereignty were infringed by the power of the CETA Joint Committee, consisting of Canadian and EU representatives, to adopt interpretations of CETA that would bind arbitral tribunals. ³⁵ The only constitutional objection to CETA that mustered the support of a majority concerned the way in which CETA Tribunal awards could become enforceable in the state through the mechanisms provided by the Washington and New York Conventions, as implemented by the Arbitration Act 2010.

Article 34 of the Constitution provides that justice shall be administered in courts established by law and that the decision of the Supreme Court is final and conclusive in all cases. These features of the constitutional system grounded, for the majority judges, two concerns with the CETA system. First, subtraction from the jurisdiction of Irish courts by enabling disputes to be resolved by CETA Tribunals instead of Irish courts. ³⁶ Second, subversion of the authority of the Supreme Court by allowing its final decisions to be overruled. ³⁷ While these points could also be made against other international treaties to which Ireland is committed, such as the ECHR, the majority distinguished CETA because of the near-automatic enforceability in the state of CETA awards. One majority judge offered the following vivid image:

[I]n the context of CETA, the [Arbitration Act 2010] has … been conscripted into service as a means of giving effect to the awards of CETA Tribunals: in this respect the Act serves as a sort of makeshift pontoon bridge by which a CETA Tribunal award is enabled to cross that legal Rubicon from the realm of international law into an enforceable judgment recognised as such by our own legal system on a more or less automatic basis. ³⁸

Together, these features of the CETA system amounted to an intrusion on Ireland’s juridical sovereignty. While the subtraction of jurisdiction and potential overruling of Supreme Court decisions might be tolerable if the consequences remained entirely in the international law sphere, the reach-back of those consequences into the domestic sphere – through enforcement under the Arbitration Act 2010 – was an unconstitutional infringement of Ireland’s juridical sovereignty.

The constitutional cure

While a majority of four judges agreed that CETA ratification was unconstitutional for reasons connected with juridical sovereignty, three of those judges endorsed – while one rejected – a ‘constitutional cure’ for solving this constitutional problem. Ratification of CETA could constitutionally proceed if the Oireachtas were to amend the Arbitration Act 2010 to allow the High Court to refuse to give effect to awards where it considered:

(a) the award materially compromised the constitutional identity of the State or fundamental principles of our constitutional order, or

(b) the award materially compromised our obligation (reflected in Article 29.4.4 of the Constitution) to give effect to EU law (including the Charter of Fundamental Rights and Freedoms) and to preserve its coherence and integrity. ³⁹

One of the majority judges rejected the proposed constitutional cure on the grounds that it would amount to a breach of CETA and in any event would not solve the juridical sovereignty problem. ⁴⁰ The three minority judges, however, approved the constitutional cure as a means of curing the constitutional frailty that the majority perceived in CETA. ⁴¹ Accordingly, six judges supported the constitutional cure, although only three of those judges considered it necessary.

The Irish Supreme Court has never before gone to such lengths to suggest how constitutional defects could be addressed. ‘Constitutional identity’ had never previously been mentioned in any Irish judgment prior to Costello.⁴² The constitutional cure was not the subject of legal argument before the Supreme Court and only two judges, one of whom has since retired, considered what constitutional identity entailed.⁴³ Based on those two judgments, constitutional identity appears to encompass the following elements: democracy; ⁴⁴ separation of powers; ⁴⁵ non-abrogation of the Constitution through treaty-making; ⁴⁶ fundamental constitutional values and/or principles; ⁴⁷ state sovereignty; ⁴⁸ the protection of fundamental rights.⁴⁹ This list appears almost as broad as the Constitution itself.

The reference to Ireland’s EU law obligations touches on the question, addressed by the Court of Justice in Opinion 1/17, of whether a CETA Tribunal could determine that an EU provision adopted in the public interest breached CETA. As we saw above, the Court of Justice, in finding that the EU’s legal autonomy was secure, emphasised CETA Tribunals’ lack of jurisdiction to declare the level of protection of public interest established by EU measures to be incompatible with CETA. The Irish Supreme Court judge who proposed the constitutional cure considered it possible that a CETA Tribunal would interpret its powers differently from the Court of Justice and reject the lawfulness of an EU law or measure,⁵⁰ leaving the Irish High Court faced with a conflict between its obligation to enforce a CETA award and its obligation to uphold EU law. It is difficult, however, to see how the presence of a statutory power to refuse enforcement in this context could alter the obligations on the Irish courts. If there were an EU law obligation not to apply a CETA award because it was inconsistent with EU law, then an Irish judge would have to refuse enforcement, irrespective of whether explicitly empowered by Irish law to do so.

First critique: the subtraction and overruling grounds

The majority judges do not consistently commit to the proposition that CETA subtracts from the jurisdiction of the Irish courts and allows for the overruling of final Supreme Court judgments. Two judges considered that the CETA system was problematic on both counts; ⁵¹ one judge considered that there was a subtraction problem but not an overruling problem; ⁵² while another judge considered that there was an overruling problem but not a subtraction problem. ⁵³ The inconsistency is telling because both critiques rest on the same presupposition: that CETA Tribunals would resolve the same dispute as would arise in the Irish judicial system. If the CETA jurisdiction is different from the jurisdiction of the Irish courts, then a CETA award could never overrule a Supreme Court judgment and the existence of a CETA jurisdiction would not subtract a jurisdiction under Irish law. True, the election of an investor for arbitration under CETA means that a particular disagreement will be litigated as a CETA dispute rather than a dispute of Irish law. But this election no more subtracts a jurisdiction from the Irish courts than any decision to choose arbitration or indeed a decision not to litigate a grievance in the first place.

The inconsistency within and between the majority judgments on the subtraction and overruling grounds derives, I suggest, from the simple fact that CETA Tribunals do not resolve the same dispute as the Irish courts. In this regard, the Chief Justice’s dissent was correct:

The decisions of the Irish courts are final and conclusive, and Ireland has juridical sovereignty, as a matter of Irish law. That principle is not breached, at least per se, by the possibility of proceedings at the level of international law even concerning the same subject matter, or for that matter by an arbitration based on consent.⁵⁴

The Chief Justice was also correct to note that CETA Tribunals did not withdraw jurisdictions from the Irish courts because CETA could not determine issues of Irish law.⁵⁵

If we reject, as I suggest we must, the claims that CETA subtracts from the jurisdiction of the Irish courts or allows judgments of the Supreme Court to be overruled, then the juridical sovereignty ground reduces to the sole concern that CETA awards are near-automatically enforceable in the state. But it is difficult to understand why that level of enforceability should be constitutionally problematic. The only context in which a CETA award could come for enforcement before the Irish courts would be where the state refused to meet its international law obligations to comply with the determination of a CETA Tribunal. But Article 29.2 of the Constitution commits the state to ‘the principle of the pacific settlement of international disputes by international arbitration or judicial determination’. Article 29.3 commits Ireland to ‘the generally recognised principles of international law as its rule of conduct in its relations with other states’, which must include the principle of pacta sunt servanda. It is problematic to ground a conclusion that CETA ratification would be unconstitutional on the proposition that a future Irish Government might – in breach of the principles in Article 29 – renege on its international law obligations to the other CETA parties to respect a CETA award in favour of a Canadian investor.

Second critique: reconciling the unconstitutionality with the cure

The endorsement of the constitutional cure raises two related problems for the majority position. First, it divides the majority judges in two, depriving any ground of unconstitutionality of majority support. The proposed constitutional cure does not involve general judicial scrutiny of CETA awards prior to enforcement. Rather, enforcement can be refused only on defined grounds: constitutional identity and compliance with EU law obligations. The constitutional cure implies that the constitutional ill was not foreclosure of judicial scrutiny per se but rather the risk that an award might be enforced that offended constitutional identity. For three majority judges, judicial scrutiny of awards was only instrumental to the goal of protecting constitutional identity. For the fourth majority judge, however, judicial scrutiny itself was required – merely allowing judges to refuse enforcement of awards that offended constitutional identity would not address this concern. Only this judge can properly be characterised as holding CETA ratification unconstitutional on juridical sovereignty grounds.

This division on the Supreme Court means that no constitutional objection to ratification of CETA had majority support. For one judge, the enforceability of CETA awards infringed juridical sovereignty. For three judges, that enforceability created an intolerable risk that constitutional identity might be infringed.⁵⁶ And for three judges, there was no constitutional problem. This 1:3:3 division makes it difficult to gauge the immediate implications of Costello and impossible to formulate any legal principle for which Costello can be cited as authority.

Second, it is difficult to glean from the majority judgments any clear sense of how an actual CETA award might infringe constitutional identity. We saw that constitutional identity appears an all-encompassing term – perhaps as broad as the Constitution itself. The judge who proposed the constitutional cure suggested that constitutional identity would be infringed by an award ‘at odds in some material way with the legislative and juridical autonomy of the state’. ⁵⁷ Given that a majority held that the CETA system did not infringe on legislative sovereignty in a constitutionally impermissible fashion, however, it is difficult to envisage how an award could go so far as to infringe constitutional identity on this basis. And it is difficult to envisage a juridical sovereignty problem with CETA awards that could affect some such awards, but not all of them. But the three majority judges who proposed the cure must have accepted that some CETA awards could legitimately be enforced within the state. If there were a general constitutional defect with the CETA system such that no award could ever be enforced, the constitutional cure would allow Ireland subscribe to CETA in theory while ensuring that no CETA award would ever be enforced. We can safely assume that no such judicial sleight of hand was intended by the majority judgments. But if general constitutional concerns about CETA – its democratic character, limitation of legislative choice, intrusion on domestic judicial systems – should not lead to the non-enforcement of CETA awards, what specific problem in a particular case might infringe constitutional identity? One judge considered it problematic that enforcement of a CETA award could be required even if based on what the domestic courts considered to be an erroneous understanding of national law. Such a scenario could amount to an infringement of judicial sovereignty, but it would be a remarkably hypothetical basis to support the conclusion that ratification of CETA would be unconstitutional. ⁵⁸ One is driven to the conclusion that three majority judges held CETA ratification unconstitutional now, without any clear sense of the future constitutional problem that they were seeking to prevent.

An alternative framing: the minority approach

The artful construction of investor-state dispute settlement under CETA to address specific issues previously identified by the Court of Justice means that it gets very close to infringing structural sovereignty in several respects, without actually crossing that line in any one respect. The analysis in Costello was ill-served by identifying separate democracy, legislative sovereignty and juridical sovereignty grounds of challenge. Rather, these grounds of challenge required holistic consideration to assess whether – in substance – ratification of CETA would excessively intrude on the capacity of Irish constitutional organs to exercise their constitutionally assigned competences in the interests of the common good. The Chief Justice’s dissent adopted this approach.

Recognising that the public law context of CETA awards put them in a very different context from private arbitration disputes, the Chief Justice rejected the state’s structural defence of CETA as exercising a separate jurisdiction in a separate domain. ⁵⁹ In language redolent of the Court of Justice’s protection of substantive legal autonomy in Opinion 1/17, the Chief Justice observed that the state’s defence failed to address the possibility that CETA might ‘significantly constrain the exercise by the organs of government of their powers, and the performance of their duties, under the Constitution’. ⁶⁰ He reasoned that the Government could not exercise its treaty-making power in such a way as to affect adversely the autonomy of the Irish legal order, a conclusion that required a careful analysis of how the CETA system would actually function. ⁶¹ For the Chief Justice, CETA did not create ‘an impermissible chilling effect whereby the institutions of the State would be precluded, or indeed, dissuaded from regulatory measures of general application and which were not plainly discriminatory or arbitrary’. ⁶² Notwithstanding the lack of any obligation to exhaust domestic remedies, he considered that the Tribunals appeared intended to be the last resort for disappointed investors where relations were terminally ruptured. Furthermore, the scope of the Agreement was limited to Canadian investors and investments, the remedies were limited and the grounds so general that they would normally lead to clear invalidity if the same measure were challenged in Irish law.

Given that there was no majority support for any one ground of unconstitutionality – one judge being concerned with juridical sovereignty per se and three judges being concerned with judicial oversight as a protection against anticipated infringement of constitutional identity – the Chief Justice’s framework is of more than academic importance. It represents an innovative framework for protecting the substantive sovereignty of the Irish constitutional order – i.e. the actual ability of Irish constitutional organs to make democratically accountable and legally reviewable decisions in the public interest as they perceive it – complementary to the structural sovereignty that the majority judges sought to protect. It is difficult to claim that CETA, notwithstanding the legitimate concerns over investor-state dispute settlement, would so compromise the substantive freedom of the state to make its own choices that it should lie outside the international relations competences constitutionally assigned to the Government, Dáil and Oireachtas. But such a view would provide a far more coherent basis on which to deem ratification unconstitutional.

Implications for Ireland’s international relations

The immediate challenge posed by Costello relates to Ireland’s ratification of CETA. Initial newspaper reports suggested that the Government would seek to implement the constitutional cure, securing amendments to the Arbitration Act 2010.⁶³ CETA refers to the New York and Washington Conventions, which Ireland has already ratified. While the New York Convention allows enforcement of arbitration awards to be refused on grounds of public policy, Article 54 of the Washington Convention requires Ireland to enforce the pecuniary obligations imposed by an award as if it were a final judgment of an Irish court. In Micula v Romania, the UK Supreme Court considered it arguable that Article 54 allowed scope for some additional defences against enforcement, in certain exceptional or extraordinary circumstances which were not defined, if national law recognised them in respect of final judgments of national courts and they did not directly overlap with those grounds of challenge to an award which were specifically allocated to Convention organs under articles 50–52 of the Convention.⁶⁴ But the UK Supreme Court also considered the opposite view equally arguable: it would be inconsistent with the scheme of the Convention for a national court to refuse to enforce an award on the ground that, if it had been an ordinary domestic judgment, giving effect to it would be contrary to a provision of national law.⁶⁵ Under this analysis, the purpose of Article 54 is to secure a mechanism for enforcement; it would thwart this purpose if enforcement could be made contingent on substantive compliance with national law simply on the ground that judgments of national courts must also comply with national law. It was not necessary for the UK Supreme Court to decide between these two interpretations of the Washington Convention for the purpose of deciding the case before it, the Court noting that the difference could only be resolved authoritatively by the International Court of Justice.⁶⁶

In Costello, the judge who proposed the constitutional cure considered that it was open to the Oireachtas ‘to build on what was said in Micula and to spell out in legislative form the defences to the enforcement of such a final judgment of the CETA Tribunal in the manner tacitly contemplated, for example, by Article 54(1)’.⁶⁷ But the viability of this suggestion depends on the correctness of the first interpretation of Article 54 considered arguable by the UK Supreme Court. It behoved the Irish Supreme Court at least to canvas the alternative interpretation considered equally arguable in Micula before proposing the constitutional cure. Leonard suggests that the constitutional cure goes far beyond the sort of defences permitted by the Washington Convention.⁶⁸ Berman has argued against the first interpretation of Article 54(1) canvassed by the UK Supreme Court.⁶⁹ In his view, the most plausible reading of Article 54(1) is that it requires courts of contracting states to ‘subject enforcement of [Washington Convention] awards to no more onerous procedures than they follow in enforcing local judgments’. The alternative of subjecting such awards to substantive national law tests: (i) places too little weight on the fact that the Washington Convention – unlike the New York Convention – does not identify any substantive grounds for non-enforcement; (ii) subjects the core enforcement obligation to disparate requirements depending on where enforcement is sought; (iii) places too little weight on the fact that the drafters made execution of judgments, but not enforcement, subject to the laws of the jurisdiction where enforcement was sought; and (iv) is inconsistent with the impetus of the Washington Convention being to minimise the role of courts at post-award stage.⁷⁰

While the meaning of Article 54(1) may not be fully settled, there is – at the very least – a significant risk that Ireland would be in breach of the Washington Convention and/or CETA if the Oireachtas legislated to implement the constitutional cure proposed by the Supreme Court. In that eventuality, international partners and commentators might well perceive not an unintentional breach on Ireland’s part but rather a deliberate decision to subject Ireland’s international law obligations to a test of national law compliance, reneging on the principle of pacta sunt servanda. Such a perception might be unfair, but the Government would want to minimise these risks by formally addressing the compatibility of any legislative amendments with CETA, whether through a declaration, reservation, or – if the other parties to CETA were to agree – a protocol. It is not immediately clear, however, why other parties would accept that CETA awards should be less enforceable in Ireland than in other jurisdictions. Any Irish plea for special treatment is more likely to intensify the general political debate that has led 11 EU member states to withhold ratification to date. In the best-case scenario for the Irish Government, Costello might shape the general investor court system which the European Commission wishes to see replace investor-state arbitration, and for which CETA is a prototype. In any event, it is unlikely that the Irish Government will proceed with the constitutional cure unless or until there is some shared understanding with the EU institutions, the other EU member states and Canada that the proposed amendments to the Arbitration Act 2010 would be acceptable.

The Government could instead propose a constitutional referendum to allow ratification of CETA. But persuading the general public of the merits of an international agreement is fraught at the best of times: investor-state dispute settlement would be a particularly difficult sell. Mr Costello himself is a Green Party deputy, currently supporting the three-party coalition government of Fianna Fáil, Fine Gael and the Green Party. His constitutional challenge to CETA ratification could just about be separated from the day-to-day business of government. But any campaign for referendum approval – almost certain to be opposed by the more left-wing opposition parties with a general election required by March 2025 – would destabilise the Green Party and by extension the coalition Government. The Government will avoid this route at all costs and, notwithstanding the difficulties identified above, seek some way to implement the constitutional cure or indefinitely defer resolution of the issue.

While the implications for CETA ratification are immense, Costello may prove to have few if any implications for the Irish Government’s treaty-making power. As we saw above, there is no principle supporting the finding of unconstitutionality that attracts majority support. The closest is the position of three judges that increased judicial scrutiny of CETA awards is necessary to reduce the risk of enforcement offending constitutional identity, but this is not a principle of broad relevance, particularly given the inability of those three judges to provide a coherent account of how such a risk could arise. Had they followed the fourth majority judge in rejecting the constitutional cure or had they framed the constitutional cure in terms of requiring prior judicial scrutiny on broad grounds such as public policy, then their position would have been internally consistent. Such a strong assertion of sovereignty would, however, have been more difficult to square with the internationalist commitments of Article 29, making it exceptionally difficult for Ireland to sign up to the sort of international agreements that most other countries, even jurisdictions with constitutional courts that jealously guard their own legal autonomy, accept. ⁷¹

The sovereignty language of the majority judges in Costello resonates with the sovereignty concerns in Crotty that have shaped 35 years of treaty-making, but the similarities end there. Crotty articulated a reasonably clear, albeit contestable set of limits on Ireland’s ability to participate in supranational organisations. Costello provides us with no similar theory for international agreements in the style of CETA. Those eager to assert limits on Ireland’s participation in international agreements – or the primacy of the people in deciding in such participation – can point to the ultimate holding in Costello. But that holding is unsupported by any coherent account of sovereignty, still less one supported by a majority of the Supreme Court. Once the travails over the ratification of CETA recede, Costello is unlikely to have a Crotty-like transformative effect on the conduct of Ireland’s international relations.

Constitutional identity and Ireland’s membership of the EU

For several decades, many constitutional courts across Europe have developed doctrines that reserve some scope to reject rulings of the Court of Justice that infringe on fundamentally important provisions of their national constitutional order. More recently, these doctrines have tended to be formulated in the language of ‘constitutional identity’ – national courts seeking some legitimacy from Article 4 of the Treaty on European Union, which commits the Union to respect member states’ ‘national identities, inherent in their fundamental structures, political and constitutional’.⁷² Hogan J, who proposed the constitutional cure referencing constitutional identity, was previously Advocate General at the Court of Justice and was unquestionably familiar with the EU debate over constitutional identity. Given that the phrase had never before been deployed in Irish constitutional law, some initial commentary on Costello canvassed the possibility that it might lead to the Irish Supreme Court deploying constitutional identity to place some outer limits on the applicability of EU law in Ireland.

Barrett, while noting the differences between its evolution in Germany and Ireland, suggests that constitutional identity could still develop into a potentially formidable obstacle to European integration.⁷³ He maintains that ‘Constitutional identity seemingly involves the idea that there may be areas of EU law which must now be regarded as falling outside the immunity from constitutional attack granted by Article 29.4.6 of the Irish Constitution’.⁷⁴ He further suggests that Costello involves the idea that there are policy choices which ought not as a matter of interpretation to be regarded as having been authorised to be transferred to European level under Article 29.4.5. While Barrett’s suggestions are tentatively expressed, they amount to a significant overreading of Costello. Even if Costello is authority for the proposition that constitutional identity controls the enforcement of international awards in Ireland, constitutional identity can only play a role in other constitutional doctrines to the extent that constitutional provisions permit. The constitutional basis for Ireland’s membership of the EU does not permit of such extension.

Ireland amended its constitution by referendum to join the European Communities in 1973, and held subsequent referendums to insert specific constitutional authorisation for the Treaties of Maastricht, Amsterdam, Nice and Lisbon, as well as the Fiscal Compact Treaty. In addition to this authorisation for membership, Article 29 of the Constitution has been amended to state that no provision of the Constitution invalidates acts done or measures adopted by the state that are necessitated by the obligations of membership, nor prevents laws enacted, acts done or measures adopted by the European institutions from having the force of law in the state (the ‘necessitated clause’). The necessitated clause on its face precludes any constitutional jurisdiction to query the competence of the EU as interpreted by the Court of Justice, since doing so would prevent an act done by a European institution from having the force of law in the state.⁷⁵ One Supreme Court judge suggested obiter that the pro-life amendment to the Constitution in 1983, since it was adopted subsequently, might qualify the necessitated clause. But the necessitated clause has since been reinserted into the Constitution with the amendments to ratify the European treaties, foreclosing any possibility that it could be deemed a subordinate provision on temporal grounds. Given that the power of the people to amend the Constitution is subject to no substantive constraints,⁷⁶ it is difficult to defend any claim that the necessitated clause must be read in light of a concept of constitutional identity that qualifies the status of EU law within the Irish constitutional order. ⁷⁷

Some academic doubt has been expressed whether the absolutist reading of the necessitated clause would prevail where the Court of Justice ‘assumed a jurisdiction which objectively it did not enjoy’,⁷⁸ but it is difficult to find a basis within the Constitution for anything other than an absolutist reading of the necessitated clause. Furthermore, as Cotter observes, Hogan J’s acceptance of the ‘fact that CETA would enjoy immunity from constitutional scrutiny for the purposes of Article 29.4.6 if ratified by the EU’,⁷⁹ combined with his conclusion that ratification of CETA would infringe constitutional identity, illustrates that, even for Hogan J, constitutional identity cannot be deployed to question the applicability of EU law in Ireland.⁸⁰ Courts do, of course, sometimes adopt radically new directions and if the Irish Supreme Court were ever to impose a constitutional identity limit on the application of EU law in Ireland, Costello would likely be cited in support of that conclusion. But particularly with one of the two proponents of constitutional identity having retired from the Supreme Court, there appears little judicial enthusiasm for a constitutional identity doctrine. The chances of such a doctrine being developed to circumvent the necessitated clause appear slim. ⁸¹

Conclusion

The Supreme Court’s declaration that ratification of CETA, as presented, would be unconstitutional is significant. At the narrowest level, ratification requires the Government secure either the agreement of the other parties to qualify the enforceability of CETA awards in Ireland or a constitutional amendment by referendum. The former route is much more attractive for the Irish Government but may further fuel political unease across Europe, with concerns over investor-state dispute settlement potentially stalling CETA ratification. Beyond the immediate context of investor-state dispute settlement, the most significant potential implications of Costello are unlikely to be realised. There is no majority support for any rationale underpinning the Court’s conclusion and by extension no general principle that could apply in other contexts. The references to constitutional identity perhaps hint that the Irish Supreme Court could emulate some continental courts and review the scope of EU competence. But significant changes to other long-established constitutional positions would be required before constitutional identity could be deployed by an Irish court to this end, and there is no evidence that constitutional identity has the level of judicial support that would support such a radical development.

Assessing the constitutionality of investor-state dispute settlement required a difficult reconciliation of the Constitution’s sovereigntist and internationalist commitments, with little guidance to be gleaned from the framework designed for a more straightforward intersection of national law and international co-operation. In attempting that reconciliation, the Costello majority unpersuasively presented CETA as a structural interference with Ireland’s juridical sovereignty. The Chief Justice’s approach, applying a broader concept of substantive sovereignty but less strictly, allowed for a more convincing analysis of how not only CETA-ratification but also judicial preclusion of CETA-ratification would affect important constitutional values. It is possible that a future Supreme Court might reconstruct the reasoning of Costello to support a more judicially interventionist reconciliation of the Constitution’s sovereigntist and internationalist commitments. But such a reconstruction would owe more to the attitudes of a future court than the logic of Costello itself. For the time being, therefore, the significance of Costello lies in its implications for CETA ratification, not constitutional doctrine. Perhaps because of its subject-matter, its politically charged conclusion, its high constitutional language, its complexity, or just its sheer length, Costello has the look and feel of a significant constitutional case. But it is a diminishing echo of the chords of sovereignty first played in Crotty, not the start of a new movement.