On 24 November 2022, the European Parliament adopted a resolution calling for the coordinated withdrawal of the EU and its Member States from the Energy Charter Treaty (ECT).1 The ECT is a multilateral agreement negotiated and concluded after the collapse of communist regimes in Eastern Europe. Its primary purpose was to encourage and protect foreign investments in the energy sector in those countries while they were transitioning to a more market-based economy. However, the ECT has come under intense criticism because its investor-state dispute settlement (ISDS) system has allowed foreign investors to litigate against states’ climate and environmental protection measures. Most recently, Italy was required to pay over 250 million euros in compensation for denying a UK-based ‘Rockhopper’ application for an oil drilling concession in the Adriatic Sea under the ECT.2
As a result, Germany, France, Spain, Poland, the Netherlands, Slovenia and Luxembourg announced their intention to withdraw from the Energy Charter Treaty. Italy had already withdrawn from the ECT in 2015. Together, these countries represent more than 70 per cent of the population of the EU. Initially, these announcements did not result in the Commission changing its position. In October 2022, it maintained that ‘the Commission is not preparing a coordinated withdrawal’ and that ‘the EU will remain a party to the ECT in its own right’.3 The Commission had invested significantly in reforming the ECT rather than withdrawing from it. However, the Commission changed its position following a resolution of the European Parliament and after it failed to secure a qualified majority in the Council to proceed with its strategy to ‘modernise’ the ECT by renegotiating parts of the agreement, including provisions that would partially phase out the protection of fossil fuels.4 In an unpublished non-paper addressed to its Member States, the Commission observed that ‘a withdrawal of the EU and Euratom from the Energy Charter Treaty appears to be unavoidable’ and that a coordinated withdrawal from the ECT by all Member States, the EU and Euratom is the best option going forward.5 Yet, as the non-paper suggests, not all Member States announced their intention to withdraw from the ECT.
This raises the question of how, under EU law, the withdrawal by the EU or its Member States from a multilateral agreement, to which both the EU and its Member States are parties, should occur. There are no provisions in the EU treaties that prescribe how withdrawal by the EU from an international agreement should take place, let alone how its Member States can withdraw from a mixed agreement. Furthermore, the ‘mixed’ nature of the ECT raises several further legal questions and complications. The ECT covers areas that fall, to a large extent, within the exclusive competence of the EU and also (rather crucially) contains an ISDS. These agreements cannot be established ‘without its Member States’ consent’.6 Can the EU and its Member States withdraw from a mixed agreement without formally coordinating such an action or is there a duty under EU law to coordinate such a withdrawal? What would be the extent of a commitment to coordinate a retreat? And suppose the EU decides to withdraw unilaterally. What would the legal consequences be for its Member States under EU law, given that they would be a party to an agreement partially within EU exclusive competence? Vice versa, what would be the legal consequences for the EU or a Member State should they withdraw from the ECT?
This article will argue that under EU law, the EU and its Member States can withdraw from a multilateral agreement, such as the ECT, independently of each other. While loyalty obligations may require both the EU and its Member States to coordinate their actions under a mixed agreement closely, they can still withdraw. The autonomous nature of EU law, following Opinion 1/19, does not allow Member State decision-making to prevent the EU from taking decisions that fall within its competence.7 Similarly, as the European Court of Justice (the Court) maintains, under EU law, the EU and its Member States are only ratifying those parts of the agreement that fall within their competence – Member States are entitled to cease exercising their powers through withdrawal.
This power to withdraw unilaterally is not without complications. First, it may affect the ability of others to remain a party to the ECT under EU law. Second, where not all Member States and the EU withdraw, ‘incomplete mixity’ raises several complicated questions under international law, particularly in relation to international responsibility and ISDS mechanisms. Therefore, a coordinated withdrawal by all Member States and the EU is the preferred policy option because it avoids these legal risks relating to international responsibility and ISDS provisions.8
This article will further argue that the procedure for the EU to withdraw mirrors the decision-making process for ratifying such an international agreement, which often involves the consent of the European Parliament by (at least) a qualified majority in the Council. This makes decision-making at the EU level (in relation to withdrawal) a comparatively heavy procedure, as most Member States allow the executive to take such decisions.
This article proceeds as follows. Section 1 first outlines the nature and reasons for the ECT as a mixed agreement. It then continues in Section 2 by setting out the EU withdrawal procedure, including whether the EU can only legally withdraw from the ECT if it coordinates with all Member States. Section 3 outlines the legal consequences for the Member States consequent to a unilateral withdrawal by the EU. Section 4 turns to a unilateral withdrawal by its Member States and the obligations of its Member States under the duty of loyalty in relation to withdrawal. The last section analyses the legal consequences for the EU of Member State withdrawal.
2. Division of powers on investment and the ECT as a mandatory mixed agreement
Before we assess how and under what conditions the EU and Member States can withdraw from a mixed agreement such as the ECT, it is worth recalling why the ECT was concluded as a mixed agreement in the first place and what type of mixed agreement the ECT is. Therefore, we start by outlining several features related to the content and nature of the ECT.
First, the ECT is a multilateral agreement to which several, but not all, Member States are parties.9 The ECT is not a bilateral mixed agreement or a bilaterally structured mixed agreement to which the EU and its Member States are parties to one part and a third state or international organisation a party to another, such as the Comprehensive Economic and Trade Agreement (CETA). The fact that the ECT is a multilateral agreement means that the relationship between the EU and its Member States is slightly looser in terms of how the agreement can enter into force.10 A bilaterally structured mixed agreement generally only allows the agreement to enter into force after all Member States and the EU have ratified the agreement. However, under their termination provisions, even bilaterally structured agreements allow for unilateral withdrawal by a Member State or the EU.11
Second, the ECT is a mixed agreement because the agreement falls only partially within the exclusive competences of the EU and its Member States. The ECT is an example of an agreement where mixity is mandatory because neither the EU nor each of its Member States has the competence to conclude an agreement on their own. Most significantly, the ECT contains a part on ‘investment promotion and protection’ (Part III). That part establishes standards of protection of foreign investments in direct and portfolio investments in the energy sector. While the former is an exclusive competence of the EU (since the entry into force of the Treaty of Lisbon), the latter is a shared competence between the EU and its Member States.12 Moreover, Article 26 ECT establishes an ISDS available to foreign investors regarding investments that fall within the scope of the agreement protected by the standards under Part III of the ECT. According to the Court in Opinion 2/15, including an ISDS in an international agreement is a ‘competence shared between the European Union and its Member States’.13
However, the shared competence referred to in relation to ISDS differs from other shared competences, such as portfolio investments. In Opinion 2/15, the Court made clear that an ISDS cannot ‘be established without its Member States’ consent’ because an ISDS ‘removes disputes from the jurisdiction of the courts of its Member States’.14 This language suggests an important departure from the traditional reading of ‘facultative mixed agreements’ and ‘mandatory mixed agreements’.15 A mandatory mixed agreement must be concluded by both the EU and its Member States because it contains matters that fall within the EU’s exclusive competence and provisions that fall within the exclusive competence of its Member States. Under EU law, Member States and the EU must jointly conclude an international agreement if they want to assume all the rights and obligations of that agreement. They cannot do so without each other, as they would be acting outside their respective competences when concluding the agreement. A facultative mixed agreement, on the other hand, is only mixed because of the political choice by the Council not to exercise EU powers that are shared with its Member States. Instead, these competences are exercised by the Member States. Under EU law, the Member States, and not the EU, assume the rights and obligations under those parts of the agreement that fall within those shared competences.
Opinion 2/15 suggests that even though an ISDS is a competence shared between the EU and its Member States, international agreements containing ISDS agreements can only be legally concluded with the participation of Member States. Several commentators have relied on a subsequent ruling in the International Organisation for International Carriage by Rail (OTIF) to suggest that the EU alone can conclude agreements containing an ISDS.16 In OTIF, the Court sought to clarify its findings over the nature of the EU’s competences for portfolio investments by stating that the EU could exercise these shared competences alone in the conclusion of an international agreement.17 However, it did not make such a clarification in relation to ISDS agreements. There was no legal reason specific to the case in OTIF to make such a clarification with regard to portfolio investment only. Hence, one can only assume that the Court views the nature of competence concerning ISDS agreements differently from that of portfolio investments. What is more, ISDS agreements touch on a far more important constitutional issue than the division of powers in relation to portfolio investments. An ISDS removes disputes from the courts of its Member States and, in so doing, directly interferes with the preliminary reference procedure in the EU.18 This procedure is the ‘keystone’ of the EU’s judicial system because it allows the ECJ to work with national courts to oversee the correct and uniform interpretation and application of EU law by all Member States. Therefore, the conclusion should be that agreements containing an ISDS, such as the ECT, are mandatory mixed agreements.19 The Dutch government has already taken this view publicly. In a letter to the Dutch Parliament on the ECT, the Dutch government made clear that an ISDS is an exclusive competence of the Member States and that any international agreement containing an ISDS requires ratification by the Netherlands.20 Furthermore, the government made clear that it would not be possible to give ‘consent’ within the meaning of Opinion 2/15 through a Council decision.
Third, the ECT’s current text is incompatible with EU law, following Achmea, Komstroy and Opinion 1/17.21 First, the ECT does not contain a disconnection clause (a clause that ensures that an ISDS does not apply between investors from Member States and other Member States under the ECT). This results in ISDS tribunals claiming that jurisdiction under the ECT (over disputes) is contrary to EU law.22 Second, the ISDS provisions in the ECT do not contain the same safeguards present in CETA that the Court considered necessary (in Opinion 1/17) to protect the autonomy of EU law by safeguarding the jurisdiction of the Court in giving a definitive interpretation of EU law.23 Third, the investment protection standards in the ECT do not contain the same safeguards present in CETA that the Court considered necessary in Opinion 1/17 to protect the regulatory autonomy of EU law. Finally, the current ECT does not sufficiently circumscribe the definition of ‘fair and equitable treatment’ or contain provisions that seek to preserve public interest decision-making.24
This incompatibility means that, under EU law, the EU and its Member States cannot remain a party to the ECT as it currently stands. Instead, they must either renegotiate or withdraw from the ECT to respect their obligations under the EU Treaties. In that sense, the modernisation effort seeks to address several of these incompatibility issues.25
3. Withdrawal by the EU
3.1. Procedure for withdrawal
Article 218 TFEU governs the procedures for decision-making by the EU institutions in respect of international agreements. This provision lays down the procedure for several key elements concerning the formation and implementation of international agreements by EU institutions. It contains decision-making procedures for the signature, provisional application, conclusion and suspension of international agreements, as well as a decision-making procedure for adopting the positions of the EU within bodies set up by international agreements that produce legal effects. However, an explicit decision-making procedure for termination or withdrawal from an international agreement is absent.
There seem to be two possibilities if no explicit provision is present. The first is that the procedure of Article 218 (9) TFEU is followed. This procedure allows the EU to take positions in bodies set up by international agreements and to (partially) suspend international agreements. Given that the European Parliament is not involved, the choice of this procedure could be characterised as being more in line with the ‘executive prerogative’ used by some states to terminate international agreements.26 , 27 The other possibility is that the same procedure would need to be followed to conclude an international agreement, as provided for by Article 218 (6) TFEU.28 The procedure to terminate would then mirror the procedure for conclusion. This option is based on an ‘actus contrarius’ principle in the literature.29 The second possibility, the procedure based on Article 218 (6) TFEU, would generally ensure greater participation by the European Parliament in decision-making. However, given the involvement of an additional institution, the procedure would generally be more cumbersome. This second possibility appears to be the better option.
First, if one looks at the constitutional traditions of the Member States, practice is so diverse that one cannot speak of a practice ‘common’ to all. Member State practice cannot point to one possibility over the other. On the one hand, fourteen Member States, including France and Germany, favour a system based on ‘executive prerogative’.30 On the other, thirteen Member States favour a system whereby their parliaments are involved in termination to a varying degree.31 Nine of the latter group follow a procedure whereby the powers of their national parliament mirror the powers of the EU Parliament at the ratification stage.32
Second, the practice of the EU institutions suggests a procedure that mirrors the process that would have to be followed for the conclusion of that international agreement rather than the procedure provided by Article 218 (9) TFEU. Kuijper and others identify five agreements concluded by the EU that have since been terminated, the termination occurring in all instances before the entry into force of the Lisbon Treaty.33 In four instances, it is clear from the decision that the procedure for conclusion was followed rather than that of suspension. In the fifth instance (on the data transfer of airline passenger information to the United States), no internal decision was necessary as the decision to conclude the agreement was annulled by the Court. Hence, the Commission issued a notice to the United States.34 Of those four decisions, the European Parliament either consulted or consented to the decision to terminate the international agreement in two instances.35 In two cases involving trade matters, the European Parliament was not involved, as was prescribed for the conclusion of trade agreements at the time. The text of the decision refers to the procedure for the conclusion of an international agreement rather than the procedure for suspending an international agreement.36
Third, the principle of institutional balance would suggest that the institutions involved in making international law should be involved to the same extent when such laws are unmade.37 The Court made clear in Somali Pirates (I) that the procedure of Article 218 TFEU ‘must take account of the specific features which the Treaties lay down in respect of each field of EU activity, particularly as regards the powers of the institutions’.38 Accordingly, Article 218 TFEU ‘establishes symmetry between the procedure for adopting EU measures internally and the procedure for adopting international agreements to guarantee that the Parliament and the Council enjoy the same powers in relation to a given field, in compliance with the institutional balance provided for by the Treaties’.39 By analogy, the termination of an international agreement in a given field should, based on the principle of institutional balance, equally respect those powers in relation to that area internally, as provided for by Article 218 (6) TFEU.
3.2. Unilateral or coordinated withdrawal by the EU
A further question is whether the EU can unilaterally withdraw from a mixed agreement without doing so jointly with its Member States. As the Court has made clear, the duty of loyal cooperation (Article 4 (3) TEU) requires that Member States and EU institutions have ‘an obligation of close cooperation in fulfilling the commitments undertaken by them under joint competence when they conclude a mixed agreement’.40 At the very least, this duty involves informing and consulting one another.41
However, as Opinion 1/19 (the Istanbul Convention) has made clear, this duty cannot go as far as preventing the EU from taking decisions within its sphere of competences and cannot require the EU to reach a ‘common accord’ with its Member States in relation to mixed agreements before deciding on Article 218 TFEU.42 The autonomy of EU law means that the EU Treaties govern decision-making by the EU and cannot be made dependent on the actions of its Member States.43 In particular, the duty of close cooperation ‘cannot justify the Council setting itself free from compliance with the procedural rules and voting arrangements laid down in Article 218 TFEU’.44 The EU Treaties, therefore, ‘not only do not require the Council to wait’ for a ‘common accord’ for the withdrawal from a mixed agreement, they even ‘prohibit’ the Council from making the initiating of that procedure contingent on the prior establishment of a ‘common accord’.45
4. Consequences for its Member States of withdrawal by the EU
What are the legal consequences for Member States, as parties to the ECT, when the EU unilaterally withdraws? In Opinion 1/19, the Court draws important conclusions from the fact that, under mixed agreements, the EU and its Member States only act within their respective competences. The Court found that ‘the conclusion of a mixed agreement by the European Union and its Member States in no way implies that its Member States exercise, in that event, competences of the European Union or that the European Union exercises competences of those States; rather, each of those parties acts exclusively within its sphere of competence.’46
As a result, the Court explained that if a Member State does not conclude a mixed agreement, it does not automatically follow that the EU is acting outside its competences by concluding the agreement. On the contrary, according to the Court, the use of the correct legal bases and the use of declarations of competences, as well as other provisions in the mixed agreement itself, should make clear that the EU, in such a case, would stay within its competences when concluding and implementing the agreement. Likewise, in principle it cannot be automatically inferred from the withdrawal by the EU from a mixed agreement that the Member States are exercising their competences over the entirety of the agreement. If that were the case with the ECT, Member States would be acting ultra vires under EU law.
The ECT covers investment protection standards for foreign direct investment in the energy sector, an exclusive competence of the EU. Member States remaining party to the EU agreement cannot assume these obligations under the ECT after the EU withdraws unless the EU explicitly empowers its Member States to do so.47 The EU’s Grandfathering Regulation does not allow such delegation to the Member States.48 First, the Regulation applies to bilateral investment agreements of Member States, not to multilateral investment agreements such as the ECT. Second, the ECT is an agreement the Member States signed before 1 December 2009. As such, it should have been notified to the Commission to benefit from the provisions that allow Member States to maintain their existing bilateral agreements.49
While the Court in Opinion 1/19 considers it cannot be automatically assumed that in cases of incomplete mixity the EU would be acting ultra vires, the Court does indicate how the EU can avoid acting outside its authority.50 These contain a declaration of competences by the EU institutions, explicit provisions in the international agreement itself, and the legal basis for the conclusion of the agreement. However, these instruments are only of limited value for Member States who wish to remain a party to an agreement from which the EU has decided to withdraw. Member States are states, unlike the EU. Therefore, the limited nature of their competences to conclude an international agreement cannot be assumed. Moreover, the current definition of ‘Regional Economic Integration Organisation’ under Article 1 (3) of the ECT indirectly offers support that, in the case of withdrawal by the EU, Member States only remain a party to the ECT to a limited extent. In addition, the current declaration of competences submitted by the EU to the ECT Secretariat offers limited guidance as to the extent of Member State responsibilities under the ECT.51 Finally, while the legal basis for the decision to withdraw from the ECT by the EU may offer some indications as to the extent to which its Member States are exercising their competences, the lack of clarity over the legal basis of EU competence for ISDS agreements and portfolio investments makes this instrument a risky basis to establish the limited nature of Member States’ responsibilities under the ECT.
Moreover, the ECT as it currently stands remains incompatible with EU law. Therefore, the remaining Member States do not have the option to remain a party to the agreement. This means that the Member States concerned must either renegotiate the agreement or withdraw. Given that the ECT is a multilateral agreement, amendments to the agreement’s text are complex, as demonstrated by the failed effort to modernise the ECT. If renegotiation is pursued, however, Member States would need to convince third states that the ECT would only bind them in areas of Member State competence. It is highly unlikely that this would be an acceptable outcome for those other parties since an ISDS is very much at the heart of the agreement. Alternatively, the EU could empower those Member States to remain a party to a renegotiated ECT, either through an amended Grandfathering Regulation or through a similar EU legal instrument. This would allow Member States to assume all responsibilities under an international agreement, even if that agreement partially covers an area of EU-exclusive competence.
5. Member State withdrawal
Member States can withdraw from mixed agreements, such as the ECT, provided that they respect their loyalty obligations towards the EU under Article 4 (3) TEU. Furthermore, as the Court considers that the EU and its Member States act ‘exclusively within [their] sphere of competence’, under mixed agreements Member States can choose to no longer exercise their competences by withdrawing from a mixed agreement.52
When the EU and its Member States are parties to a mixed agreement, they have ‘an obligation of close cooperation’ to fulfil the commitments undertaken by them under joint competence when they conclude a mixed agreement.53 As Delgado and Larik (initially) pointed out, the Court ‘routinely characterised the duty of co-operation through general statements in favour of inter-institutional co-operation without specifying the concrete results required to achieve that co-operation’.54 However, more recent case law suggests that, at the very least, this duty amounts to an obligation to ‘inform and consult’ the Commission and, in certain instances, to refrain from certain actions on the international stage.
This so-called ‘duty to remain silent’ follows from the PFOS (perfluorooctane sulfonate) ruling.55 In PFOS, the Commission had started infringement proceedings against Sweden for unilaterally proposing to the Conference of the Parties under the Stockholm Convention to add a substance, PFOS, to Annex (A) to eliminate that substance. The matter fell within shared competence, which allows the EU and its Member States to go beyond the levels of environmental protection sought by the EU under Article 193 TFEU. However, the Court found that Sweden had violated its loyalty obligations because it had ‘disassociated’ itself from a common strategy within the EU not to list the substance, and the proposal would have had consequences for the EU. In particular, following PFOS, a duty to ‘abstain’ from international action exists where (1) there is an existence of a ‘common strategy’ of the EU; (2) Member State action would amount to a ‘disassociation’ from that common strategy; and (3) that disassociation has ‘consequences’ for the Union.
The PFOS ruling concerned the implementation of international agreements where unilateral actions by Member States would have impacted EU legislation. However, as Heliskoski pointed out, ‘one should probably exercise caution drawing general conclusions from the PFOS judgment insofar as [it] concerns the constraints imposed by Union law on Member States exercising their shared competence as parties to mixed agreements’.56
The withdrawal by Member States, particularly in the context of an EU-led effort to renegotiate an agreement, the ECT, raises several issues that would qualify any ‘duty to remain silent’.57 First, the duty of loyalty finds its logical limits with the division of powers between the EU and its Member States. Contrary to the situation in PFOS, a renegotiation of the ECT would require ratification by the parties to the ECT. It would not result in an obligation to refrain from an action (a duty to ‘remain silent’); it requires taking action. Member States would be forced to ratify an international agreement. In most cases, this would result in putting such an agreement up for a vote in national parliaments and informing the parliamentarians that they have only one choice under EU law, notwithstanding the Member States’ competences. While the duty of loyalty can require Member States and the EU institutions to cooperate closely, it cannot go so far as to encroach upon the division of powers between the EU and its Member States. In other words, the Member States, acting exclusively within their sphere of competence, can reject the outcome of negotiations. A national parliament cannot be required under EU law to ratify an agreement partially within the exclusive sphere of competence of that Member State because of loyalty obligations. If it were otherwise, the duty of loyalty would upend the very logic of mixed agreements. If this is the political choice of the Member State in question, the only logical option left for the Member State would be to withdraw, given that the current ECT is incompatible with EU law.
Where a common strategy by the EU to renegotiate can be discerned, Member States should facilitate the EU’s ability to secure the renegotiated text’s entry into force. Interestingly, by remaining a party to the ECT, Member States may do the opposite. In other words, withdrawal would facilitate the EU’s tasks. This is, perhaps, a counterintuitive conclusion that follows from how amendments to the ECT text can be made. Under the ECT, amendments need to be ratified by three-quarters of the contracting parties.58 If there is no support within a Member State to ratify amendments, acting within its sphere of competences, but does not withdraw, reaching this threshold would be more difficult.
In any event, for there to be any loyalty obligations based on PFOS, one would need to establish the existence of a ‘common strategy’ of the EU to do something that would not allow for Member States to withdraw. As pointed out above, the fact that the EU seeks to renegotiate the current ECT text cannot force Member States, based on loyalty, to accept such a result. Regardless, such a strategy might not exist in the first place. While the Commission obtained authorisation to negotiate amendments to the ECT, and the Commission had proposed to (passively) endorse the negotiated result at the Energy Charter Conference by participating in the vote and raising no objection to the initialling of the modernised text,59 the proposal failed to obtain a qualified majority in the Council.60
Following PFOS, the threshold for engineering a new ‘common strategy’ is low. In PFOS, the Court maintained that a common strategy could be discerned from discussions within a working party of the Council.61 In those discussions, the substance Sweden wanted to list was discussed but was not part of the subsequent proposal of the Commission. In any event, a new proposal from the Commission would amount to the initiation of such a strategy, or as the Court states are ‘the point of departure for concerted [EU] action’.62
Summarising the above, where the EU has a common strategy to renegotiate the ECT, loyalty obligations result in an obligation to consider the negotiated result and inform and consult the Commission of the political choices made by its Member States in accepting the negotiated result (or not). Where the Council adopts a position to be taken at the Energy Charter Conference that would allow for the initialling of the negotiated results, Member States, following Article 36 (7) ECT, cannot exercise their voting rights and block such a decision. However, such obligations do not prevent Member States from withdrawing, even before a decision of the Energy Charter Conference is made, as withdrawing has no negative consequences for the decision to initial the negotiated text.
6. Consequences for the EU of Member State withdrawal
Under EU law, can the EU remain a party to a mixed agreement such as the ECT when one or more Member States withdraw? As elaborated above, Opinion 1/19 suggests this is possible, as the Court considers that the EU and its Member States each act exclusively within their spheres of competence.63 Therefore, one cannot automatically assume that the EU would be acting ultra vires under the agreement if not all Member States are party to a mixed agreement, as that would incur international liability. The Court bases this argument on the fact that the third parties to the Istanbul Convention are ‘aware of the limited nature’ of the EU’s competences.64 In that sense, the EU ‘gives indications’ as to the extent of its limited powers under the international agreement through the legal bases used in the conclusion of that agreement and through a possible declaration of competences.65
While a number of academics have criticised this approach, it appears that the ECT does not currently meet the low threshold set by the Court.66 Under Article 1(2)–(3), the ECT indirectly suggests that the EU’s membership is based on limited competences.67 However, neither the original legal bases nor the EU and the Member States’ ‘Joint Statement pursuant Article 26 (3) (b) (ii) ECT’ allow for a meaningful understanding of the extent of the EU’s competences.68 On the contrary, the Joint Statement suggests that the EU is exercising competence over the ECT’s ISDS, something it cannot do without its Member States’ consent.69 This may lead to a situation where an investor of a third state submits a dispute for arbitration against the EU for actions taken by a Member State that it considers as falling under EU competence (for instance, unlawful expropriation of foreign direct investment). Given that an increasing number of Member States are withdrawing from the ECT, this situation is likely to occur often. To no longer act ultra vires under EU law, the EU institutions should, at the very least, remedy this situation by submitting a declaration of competences declaring that the EU is not exercising powers over Article 26 ECT and does not have the power to act as a respondent in any cases brought under that article even if the ECT does not allow parties to make reservations.70 It should be emphasised, however, that such an approach would not likely avoid EU liability under international law.
7. The competence to neutralise the sunset clause
The ECT contains a so-called sunset clause in Article 47(3) ECT. The sunset clause extends the applicability of the entire agreement in relation to investments made in the region of one party by investors from another for an additional twenty years. An important question is who, under EU law, would be competent to conclude any subsequent inter se agreement between (former) parties to neutralise or terminate this sunset clause.71
Previously, the Member States and not the EU have – through an international agreement – terminated bilateral investment agreements between themselves.72 That international agreement contained provisions neutralising any sunset clauses that might be present in such bilateral investment agreements.73 This solution makes sense from an international law perspective because the EU was not a party to these agreements. However, in the context of the ECT, the situation is comparable as it concerns the termination of the sunset clause between EU Member States. Such intra-EU disputes are between Member State investors and Member States and do not concern the EU as a party to any such agreement. Moreover, termination does not involve third states. Therefore, from an EU law perspective, such agreements do not concern ‘foreign direct investment’, an exclusive EU competence. Consequently, it seems evident that an inter se agreement between Member States neutralising the ECT’s sunset clause is not an exclusive EU competence and can be exercised by Member States.
However, in Opinion 2/15, the Court takes the view that terminating international agreements with third states follows the substantive division of powers, as discussed in Section 1. Under the Court’s ‘theory of succession’, where the EU obtains exclusive competence over a particular area, the EU succeeds its Member States in their international commitments and thus has the power to modify or terminate such obligations.74 Accordingly, an international agreement that would neutralise the sunset clause in the ECT with third states would follow the division of powers between the EU and its Member States. In other words, the question of who is competent to conclude such an agreement would need to be answered by the scope of the inter se agreement with the third state itself. This would mean, for instance, that if the agreement results in the termination of or the continuation of an ISDS, this would remain a Member State competence or, at the very least, any EU competence to do so that can only be exercised with the Member States’ consent. On the other hand, if the agreement removes all direct investment from the scope of the sunset clause, it would be possible, under EU law, for the EU to conclude this agreement.
By concluding the ECT, the EU and its Member States have created a genie that is difficult to put back in the bottle. Withdrawal by the EU or its Member States from the ECT creates such complicated international and EU law questions that one is not surprised by the reluctance of Commission officials to engage with this issue. This contribution has sought to answer two main EU law questions: who has the power under EU law to decide on withdrawal from the ECT, and can this power be exercised unilaterally? In other words, is there an EU legal obligation for a ‘coordinated withdrawal’ by the EU and its Member States?
This contribution argues that unilateral withdrawal by the EU and its Member States is possible, subject to loyalty obligations to inform and consult one another. This concerns the view taken by the Court in Opinion 1/19 on mixed agreements. Under mixed agreements, the EU and its Member States act ‘exclusively within [their] sphere of competence’, thus allowing them to exercise these powers in a disjointed fashion. Loyalty, moreover, may require Member States to refrain from acting in certain instances. However, given that the ECT is incompatible with EU law, refraining from withdrawal is not an option in the case of the ECT if a Member State rejects a renegotiated text. Member States cannot remain a party to an international agreement that violates EU law. Thus, while the Member States, within their sphere of competences, remain empowered to reject any negotiated result, even if that result is part of a ‘common strategy’, they are essentially left with one option – to withdraw from the agreement.
This is not to say that unilateral withdrawal is without legal problems. Most of those are of an international law nature, but those problems may also touch upon the division of power issues between the EU and its Member States. Member States that have withdrawn may still face claims through the backdoor should the EU decide to act as a respondent in ISDS cases, triggering complicated questions of what to do with such apparent ultra vires actions by EU institutions. Likewise, the withdrawal of the EU puts Member States in the position of being party to an agreement that falls only partially within their competence, which risks trespassing on exclusive EU powers. Coordinated withdrawal would remedy many of these problems. Given that eight Member States, representing over 70 per cent of the EU population, have withdrawn or are in the process of withdrawing, and there is no support for a renegotiated text, it is much more politically expedient for the EU to commit to a joint effort rather than remain stuck with an agreement that is incompatible with EU law.