The love-hate story of arbitral jurisdiction over claims against states in the EU

by Emanuela Matei, Associate Researcher – CELS*

Staging the scene

In October 2013 the European Commission issued a note entitled ‘Platform for Good Tax Governance: Addressing the remaining cases of double taxation in the single market: means to foster arbitration’ in which it exposed the limitations of the EU Arbitration Convention 90/463/EEC. It affirmed that whenever a case covered a situation where no DTC was applicable, the domestic rules have been proven insufficient in regard of the pursuit to eliminate double taxation. The Commission recommended the renegotiation of the existent intra-EU DTCs in order to introduce an arbitration clause. Another option would be the adoption of a directive providing for an arbitration clause. On the other hand, the harmonisation of the rules applicable in the context of DTCs has been regarded as too invasive vis-à-vis the tax autonomy of the Member States.

In the meanwhile, the Commission has been fighting a whole different battle in the field of intra-EU BITs, where actually the elimination of the arbitration clause is one of the main issues. Since the Member States have not acted in order to displace the alleged incompatibilities, the Commission envisages a discussion with the Member States and all interested parties regarding the further improvement of investment protection within the EU.

In the OECD setting, the BEPS project that mainly addresses the risk of double non-taxation would have been incomplete if the elimination of double taxation had been left outside. The question of effective dispute resolution has been addressed in Action 14 of the BEPS. The main topic relates to supplementing the existing provisions of the tax treaties with a mandatory and binding arbitration provision. Exactly as the Commission, the OECD remarks that many DTCs do not contain such a clause and some other DTCs do not guarantee that the access to arbitration is actually recognised.

As it stands now, the EU law does not provide protection against double taxation beyond the scope of the principle of non-discrimination and the four economic freedoms. On the other hand, the issue of double non-taxation is caught under the scope of State aid prohibition[1]. Disfavouring investors is allowed under EU law, as long as the tax measure is not discriminatory. If the tax measure distorts competition, the Commission should react as mentioned above in accordance with the provisions of Article 116 TFEU.

The division of competences in taxation matters reminds you of a quantum system existing in a combination of multiple states that may lead to different outcomes. The multitude of states collapses into a state or another as soon as the system interacts with the external world. In a similar manner, the limits of the national tax autonomy are not predetermined, but only determinable through interpretation, so the matter of jurisdiction to resolve a question entangled with taxation will never be a definitely closed file.

The parallel universe of international investment law

External investment may fall either under the provisions of Articles 206-207 TFEU, if the investment comes from a third country or within the scope of internal market rules, if the investment comes from another Member State. Intra-EU investments fall within an area of shared legislative competence. Taxation measures may attract investments from abroad or put the investor off the idea to move to a different Member State.

The substantive protection derived by investors from the network of BITs is clearly superior, since it includes protection against measures which are not discriminatory in the meaning of Article 18, 49 or 63 TFEU. However, the remarkable difference in favour of the BIT regime can be noticed in the field of procedural rights. The fact that individuals do not have locus standi before the CJEU is the biggest impairment in this context.

The effectiveness of the protection granted by EU law to the intra-EU investor depends on the sincere cooperation of each Member State to take any action that might be necessary in order to achieve the expected results. In Cilfit, the CJEU affirmed that if the correct application of EU law is so obvious as to leave no scope for any reasonable doubt as to the way in which the question raised is to be resolved, the national court may refrain from submitting that question for a preliminary ruling and take upon itself the responsibility for resolving it. Furthermore, it must be established in detail that there is no such doubt[2].

[A] court against whose decisions there is no judicial remedy under national law is required, when a question of EU law is raised before it, to fulfil its obligation to bring the matter before the Court of Justice, unless it has established that the correct application of EU law is so obvious as to leave no scope for any reasonable doubt and that the existence of such a possibility must be assessed in the light of the specific characteristics of EU law, the particular difficulties to which its interpretation gives rise and the risk of divergences in judicial decisions within the European Union[3].

The obligation to refer for a preliminary ruling is strict in the case of courts of last resort, though in practice, national courts may still make use of Cilfit in order to avoid a request for a preliminary ruling. I would also dare to affirm that generally in any domain of science, less you know about a subject, clearer things appear to be. In case that certain national courts still do not dispose of human resources with sufficient knowledge of EU law, the intricacy of a certain question of EU law would not be easily accessible to them. It is very probable that in such a case, the plaintiff will not be able to obtain damages by making use of Köbler doctrine, since the infringement will not be manifest[4]. The Commission as well does not have an outstanding record of starting infringement proceedings in order to discourage the  practice of making use of a broader interpretation of the Cilfit-exception.

Since 2010 the overall number of infringement procedures has decreased, while the number preliminary rulings under Article 267 TFEU has significantly increased[5]. However, these numbers do not say anything about compliance with the duty to refer for a preliminary ruling and they do not provide any guarantee that the interpretation given by the CJEU has been applied correctly in practice by the national courts. The inability of EU law to effectively defend the implied right of an individual to have a question referred to the CJEU for a preliminary ruling invites the question whether international investment law should provide legal protection via its traditional means i.e. through investor-state arbitral proceedings.

As discussed above, the source of conflict is often the reverse coin of a certain competence. It is not the right to impose taxation, but the decision to abstain from it that may cause problems. It is not the right to allow free movement of capital from and to third countries, but the ability to impose restrictions on it that generated a number of infringement proceedings in EU law. It appears obvious that EU law allows Member States to restrict the protection of the foreign investor in cases where the IIA standard would not. In other cases, the EU law itself contains restrictions that the Member State must apply.

The judicial system of the European Union has nonetheless been created as a complete system of legal remedies and procedures designed to ensure review of the legality of acts of the EU institutions. The Member States are not allowed to confer the jurisdiction to adjudicate disputes between individuals on ‘a court created by an international agreement which would deprive those courts of their task, as ordinary courts within the European Union legal order, to implement European Union law and, thereby, of the power provided for in Article 267 TFEU, or, as the case may be, the obligation, to refer questions for a preliminary ruling in the field concerned[6].

Concerning disputes between Member States in relation to the application of the ECHR within the scope ratione materiae of EU law, the CJEU held that only an explicit exclusion of the ECtHR’s jurisdiction would be compatible with EU law[7]. The tasks attributed to the national courts and the CJEU are deemed indispensable for the preservation of the very nature of the law established by the Treaties[8]. It must be underlined that the Commission has been in favour of establishing a system of external judicial control in many occasions including the adhesion to the ECHR, and agreed with the adoption of an arbitration clause in CETA, TTIP and FTAS[9]. The answer provided by the CJEU until now has been different from the opinion of the Commission.

In a recent request for preliminary ruling the Bundesgerichtshof asked the CJEU whether the application of an ISDS provision in an intra-EU bilateral investment protection agreement would be precluded by Article 344 TFEU or Article 267 TFEU or Article 18 TFEU[10].  Article 344 TFEU merely prohibits Member States from submitting a dispute concerning the interpretation or application of the Treaties to any method of settlement other than those provided for in the Treaties[11]. The theory projected by the Commission in Micula is that by abiding by the rules of the BIT, the Member State in question endorses the submission of disputes concerning the interpretation or application of the Treaties to the arbitral tribunal[12]. The Bundesgerichtshof in Achmea expresses the opposite opinion, namely that only if the plaintiff is a Member State, then the Article 344 TFEU would apply[13].

Article 267 TFEU aims to avoid divergences in the interpretation of European Union law which the national courts have to apply[14]. In Achmea, the arbitral decision has been adopted under the UNCITRAL rules and therefore it shall be enforced in accordance with the New York Convention that allows the domestic court to take into consideration the issue of public policy. The domestic court would have thus the opportunity to refer for a preliminary ruling in this context. The situation is different under the ICSID convention though. In any case, the Bundesgerichtshof finds in its referral no case of incompatibility with Article 267 TFEU.

The Bundesgerichtshof reasoned on the other hand that the application of the ISDS clause might be precluded by Article 18 TFEU. The Bundesgerichtshof proposed in this case to extend the higher level of protection to all investors in a similar position. Apparently, this solution seems to be elegant and certain, however in my view the issue of potential incompatibility is more intricate than it may look at first sight. A higher level of investor protection may potentially counteract the effective application of a provision of Union law protecting a specific interest of the EU[15]. Because the risk of such conflicts exists, the exercise of balancing effectiveness against investor protection must be performed by the competent court or tribunal. The role of the CJEU is to ascertain that the court that performs the assessment does not rely on an erroneous or divergent interpretation of EU law. As I mentioned before, the Cilfit doctrine does not provide a sufficient insurance that national courts apply Union law consistently with the interpretation adopted by CJEU, but at least Member States may be brought to Court for not complying with EU law obligations.

However, the investor has no means to bring a case against the Member State before the CJEU, even if an instance of unfair unequitable treatment under the BIT-rules also implied an infringement of EU law. It could be argued that the arbitral proceedings do not constitute alternative means in relation to proceedings established under EU law. The path paved by Köbler does not constitute a genuine alternative to the system of international arbitration, since it fully relies on domestic remedies and the general principle of state liability for breaches of international law obligations[16].

The protection offered by Köbler jurisprudence is limited to serious infringements of individual rights conferred by EU law, thus its scope is far narrower than under IIA law. The state liability established by CJEU in Köbler aims to fill the gap left behind by the authors of the Treaties and it employs the notion of common legal traditions of Member States that enshrines a set of rules regarding a minimum standard of protection[17]. Moving beyond this minimum standard would require an amendment of the Treaties in the sense of extending the locus standi of individuals under EU law.

The opinion of Bundesgerichtshof that the higher standard imposed by the arbitration clause should be extended to cover all similar situations may be attuned with the principle of equivalence imposed on the domestic system of remedies in matters where the protection of rights derived from EU law is at stake. The CJEU held that ‘in the absence of [EU] rules governing the matter, it is for the domestic legal system of each Member State to designate, with due observance of the requirements stemming from [Article 47 of the Charter] and the principles of effectiveness and equivalence, the courts and tribunals with jurisdiction and to lay down the detailed procedural rules governing actions brought to safeguard rights which individuals derive from [EU] law[18].

However, even if the unfair treatment would in certain instances amount to an infringement of EU law, it is doubtful that the access to arbitral proceedings would constitute the benchmark for determining the equivalent treatment in matters of judicial protection of EU rights. It all comes down to one thing. The scope of a BIT is per definition limited to the natural or legal persons referred to in it. The Commission views this limitation as a reason to conclude that the intra-EU BITs – all of them – are not compatible with EU law, since by acceding to the EU, a Member State is obliged to grant the same advantages related to individual rights comprised within the scope of internal market law to all the other Member States.

In relation to the substantive rights guaranteed by a BIT and their wider application ratione personae adopted in order to cover the investors who are nationals of other Member States than the signatories of the BIT, one can suppose that this extension is – in theory – possible. However, in consideration of the procedural rights and the assessment of compatibility with EU law of a certain obligation derived from the enhanced version of the BIT, the possibility to extend the scope in order to cover natural persons not covered by the BIT is doubtful.

It is doubtful not only for reasons very well-established by the CJEU in its Opinions pursuant to Article 218(11) TFEU, but also because the access to arbitral proceedings is provided under the premise that based on the principle of reciprocity, the counterparty will match this benefit by assuming an identical obligation. The extension of the benefit must be matched by the extension of the obligation to guarantee the access to arbitral proceedings on all other Member States. It would definitely require harmonisation measures.

Moving from reciprocity towards the unconditional duty to carry out tasks flowing from the supranational law entails a big psychological stride in regard of sensitive domains, such as taxation or judicial remedies. It requires that the governments of the Member States refrain from (ab)using blue pills – false impressions of intact autonomy – and start to realise that harmonisation may offer a better way to maintain control over an issue, since this method would allow them to actually define in more concrete terms the limits of their duties under EU law. If the Commission plans to adopt a directive replacing the Arbitration Convention in order to enhance the protection of taxpayers, I see no reason why it could not consider a similar step in order to improve the protection of economic freedoms in general.


[1] Mc Donald’s SA 38945, 2016/C 258/03 and GDF Suez SA 44888, not published yet.

[2] Case C‑379/15 Association France Nature Environnement EU:C:2016:603 paragraph 52.  Under ECHR law, the national courts must state the reasons why they consider it unnecessary to seek a preliminary ruling. A refusal to refer a question for a preliminary ruling may constitute in itself a violation of Article 6 § 1 of the ECHR. However, again if the decision of the national court not to refer is based on an erroneous application of Cilfit, the ECtHR does not have jurisdiction to examine this question. In practice, a simple reference to Cilfit  finding that there was no reasonable doubt may be sufficient in order to avoid a breach of the right to a fair hearing within the meaning of Article 6 § 1. See Ullens de Schooten and Rezabek v. Belgium, nos. 3989/07 and 38353/07, § 62, 20 September 2011, and Dhahbi v. Italy, no. 17120/09, §§ 31-34, 8 April 2014.

[3] Case C‑379/15 Association France Nature Environnement EU:C:2016:603 paragraph 50.

[4] Case C-224/01 Köbler EU:C:2003:513 paragraphs 53 and 124.

[5] Report from the Commission on the Monitoring the application of Union law, 2014 Annual Report. COM/2015/0329 final.

[6] Opinion 1/09 EU:C:2011:123 paragraph 80.

[7] Opinion 2/13 EU:C:2014:2454 paragraph 213.

[8] Opinion 1/09 EU:C:2011:123 paragraph 85.

[9] Opinion 2/15, still pending, date of hearing 13 September 2016.

[10] Case C-284/16, Achmea, pending.

[11] Opinion 1/09 paragraph 63.

[12] Commission Decision 2015/1470 Micula points 102, 104, 112, 114.

[13] Bundesgerichtshof I ZB 2/15 Achmea ECLI:DE:BGH:2016:030316BIZB2.15.0.

[14] Opinion 1/09 paragraph 83.

[15] By analogy Joined Cases C-402/05 P and C-415/05 P Kadi EU:C:2008:461 paragraphs 353, 371-3. Case C‑105/14 Tarrico EU:C:2015:555 paragraphs 54-55, Case C‑237/15 PPU Lanigan EU:C:2015:474 paragraph 63.

[16] Köbler, supra footnote 4, paragraph 48.

[17] Lenaerts and Gutiérrez-Fons, 2010. The Constitutional Allocation of Powers and General Principles of EU Law, Common Market Law Review 47, pp 1635-6.

[18] Case C-583/11 P Inuit Tapiriit Kanatami and Others v Parliament and Council EU:C:2013:625 paragraphs 100 and 102.


Emanuela Matei,  Associate Researcher at the Centre of European Legal Studies, Bucharest. Juris Master in European Business Law (Lund University, June 2012), Magister legum (Lund University, June 2010), BSc in Economics & Business Administration (Lund University, June 2009).

Just Because State Aid Is The Best Hammer Does Not Mean That All Issues Are Nails (Part II)

by Emanuela Matei, Associate Researcher – CELS*

This article represents Part 2/2 of a larger material regarding the interaction of EU state aid rules and international investment law in the context of recent EC Decisions. Part 1/2 was published earlier this week.

B.  Selectivity

Whether a regulatory measure is selective shall be examined within the context of the particular legal system by verifying whether the measure constitutes an advantage for certain undertakings in comparison with others, which are in a comparable legal and factual situation. Since the present case concerns a regional scheme, the financial autonomy of that region may justify a differentiation[i]. However, the mere fact of acting on the basis of a regional development or social cohesion policy would be insufficient in itself to justify a measure adopted within the framework of that policy[ii]. The disfavoured regions do not enjoy fiscal autonomy, thus the regional character of the measure would be sufficient to prove the selectivity of the aid. The Commission chose nonetheless a different line of argumentation.

‘…compensation for damages will not selectively benefit an individual undertaking only insofar as that compensation follows from the application of a general rule of law for government liability which every individual can invoke, so that it excludes that any compensation granted confers a selective benefit on certain groups in society’.

The Commission affirms that the compensation does not follow from the application of a general rule of law for government liability, since the access to justice is restricted to certain groups of individuals, i.e. foreign investors covered by the BITs. It concludes that to the extent that paying compensation awarded to an investor pursuant to a BIT amounts to granting an advantage, the advantage is selective.

In my view, first and most important, it is not necessary to go so long in order to prove the selectivity of the measure, since the scheme is regional and the award does nothing more than re-establishing the facilities granted by that scheme.

Secondly, if such definition of selectivity were admitted by the CJEU, the scope of the State aid would go beyond ‘wide’. It would potentially cover all situations, where a conflict between a State and an undertaking can be solved by arbitration. It would outlaw investor-State arbitration as such. The current solution for investor protection is based on a network of BIT-agreements including an ISDS-clause together with a worldwide affiliation to the ICSID Convention. The complexity of this structure consists in its apparent bilateralism and inward transnationalism. The Micula dispute may appear to be an issue between Romania and two associated Swedish investors, but in reality, it concerns the reliability of the current system of investor protection.

Thirdly, concerning the ‘selective’ access to justice examined vis-à-vis the matter of State aid control, there is no difference between intra- and extra-EU BITs. There is nothing in the State aid law that stipulates that the prohibition of State aid only applies, if the investor-beneficiary is national of a Member State. If the definition of selectivity is derived from the application of an exceptional rule of government liability, which not every individual can invoke, the extra-EU BITs would also be deemed illegal, unless they could qualify for an exemption as stated by Article 351 TFEU.

C.  Upon an undertaking

An investor can be involved in FDI or could act as a portfolio investor. Micula brothers are direct investors with a long-term strategical approach, so normally, no distinction can be made between the legal situations of a direct investor, who is a natural person and a legal person as vehicle of direct investment[iii]. I agree with the Commission that in the present case no difference can be made depending on whether the compensation collectively awarded to all five claimants by the Tribunal is paid out to the shareholders or to the companies owned by them.

However, I must point out, an important factor. A distinction must be made between the grant upon an undertaking i.e. a single economic unit and the recovery of State aid that obviously must be applied in relation to a person. A shareholder, who is directly involved in the day-by-day management of fully owned companies will be covered by the notion of undertaking[iv]. In the present case, it is the award that states who is entitled to payment and the recovery of aid shall follow the same assessment, as long as the award does nothing more than restoring the fiscal facilities put in place by EGO 24.

The previously distinct line between the grant of aid to an undertaking as element of State aid definition, which is an abstract concept and the matter of recovery, which is a concrete device, a legal remedy, has been blurred by a recent CJEU ruling[v]. This new theory of a maintained separate legal personality of the State aid beneficiary in relation to its controlling shareholder – even if it appears to be inconsistent with the usual understanding of the doctrine of single economic unit – could support the argumentation claiming that the shareholders’ interests as natural persons would depart from the interests of the three corporate claimants[vi].

D.  Imputability

The question of imputability is theoretically the most interesting. The initial advantage is granted by Romania to investors established in a certain disfavoured region, but the enforcement of the award may be ordered by any of the ICSID-members, inside and outside the EU. Would such a payment still be imputable to Romania?

According to the Commission, the answer is affirmative, since the voluntary agreement of Romania to enter into the BIT created the conditions for the selective advantage resulting from the award. Is the act of signing a BIT five years before the accession to the EU illegal under EU law? Is there a direct causal relation between this act and the grant of State aid? If the act were illegal would the culpability be attributed to Romania alone? If Romania had chosen to terminate the BIT in January 2007, the BIT sunset clause would have nonetheless maintained the protection for investments already in place until 2027. It would not have changed anything with regard to the Micula dispute.

The adoption of EGO 24 is definitely an act of State, but the enforced payment of the award by means of seizing assets abroad ordered by foreign courts or bailiffs cannot be attributed to Romania based on the simple observation that Romania did not terminate its BIT in 2007 or because it entered into a BIT agreement in 2002. The question of imputability is extremely complex and the State aid instrument is in my opinion too blunt to be able to cope with this complexity. While a payment ordered by a Romanian court or bailiff is imputable to the Member State in line with the obligations assumed according to the Treaties, a payment ordered by a Belgian or an U.S. federal court cannot be imputable to Romania, unless it can be substantiated that the act of engaging in BIT agreements is illegal per se under EU law.

III.          Conclusions

According to the available information, the adoption of EGO 24 established a derogation from the regime of ‘normal taxation’ implying an economic advantage. This advantage is selective due to its regional character and as any other regulatory measure is imputable to the state. It is supported by state resources, as a negative advantage that consists in a derogation from generally applicable fiscal obligations. The measure has not been notified to the Commission and it has been found illegal by the Romanian Council of Competition in May 2000. According to Atzeni a compensation that restores an illegal State aid cannot be allowed under EU law, therefore the Asteris exemption is not applicable.

The Commission tries instead to prove that the enforcement of the award issued in 2013 would in itself constitute unlawful State aid. I disagree with this line of argumentation. First, it would be redundant to prove a distinct aid entailed by the enforcement of the award and secondly, it would lead to unreasonable implications. The fact that Romania signed the BIT five years before its accession to the EU and three years before even becoming an acceding country, supports the idea that Romania cannot be held culpable under EU law for the decision to engage in such agreements. If the signing of a BIT had been forbidden under EU law, the attention of the Commission should have been directed towards the other party of the agreement, Sweden, a Member State with full rights and obligations at the relevant time.

Concerning the allegeable obligation to terminate the BIT, it must be said, first that the measure would have no effect on present investments and secondly, that Romania cannot be held as solely responsible for the maintaining of a parallel system of protection that potentially could threaten the autonomy of the Union legal order. The legality of the BIT in question should not be examined in isolation, since the practicability of the present system of investor protection relies on a network structure and the privileged access to arbitration of foreign investors. By disconnecting one node from the network, the problems indicated by the Commission in its decision at point 66, namely, the fact that the current system of State liability is not applicable to any investor, will not be solved.

The applicants in the present case personify the global community of foreign investors and represent the interests pleading in favour of maintaining the system of protection that pre-existed the EU law. State aid control is the appropriate tool for treating bi-dimensional relations between States and undertakings, but it does not seem to be adequate for dealing with triangular relations between a State and the community of foreign investors represented by the web of transnational institutions established under international law. Such matters of incompatibility between the pre- and post-Lisbon system of investor protection or between pre- and post-accession State aid measures should have been addressed by making use of other more appropriate instruments.


[i] Case C-88/03, Portugal/Commission [2006] ECR I-07115 [67].

[ii] Idem [82].

[iii] Case C-222/04 Cassa di Risparmio di Firenze SpA and Others [2006] ECR I-00289 [112].

[iv] Case C-170/83 Hydrotherm [1984] ECR I-2999 [11].

[v] Case C‑357/14 P Dunamenti v Commission, not yet reported [115].

[vi] Commission Decision (EU) 2015/1470 of 30 March 2015 on State aid SA.38517 (2014/C) (ex 2014/NN) implemented by Romania, OJ L 232, 4.9.2015, p. 43–70 [59].


Emanuela Matei,  Associate Researcher at the Centre of European Legal Studies, Bucharest. Juris Master in European Business Law (Lund University, June 2012), Magister legum (Lund University, June 2010), BSc in Economics & Business Administration (Lund University, June 2009).

Just Because State Aid Is The Best Hammer Does Not Mean That All Issues Are Nails (Part I)

by Emanuela Matei, Associate Researcher – CELS*

This article represents Part 1/2 of a larger material regarding the interaction of EU state aid rules and international investment law in the context of recent EC Decisions. Part 2/2 will be published later this week.

 I.          Introduction

In May 2014, Obama defended a more relaxed foreign policy that entailed less military interventions, by stating, I cite: ‘Just because we have the best hammer does not mean that every problem is a nail[i]. The same observation can be made in relation to the Commission’s all-encompassing use of the versatile tool of State aid control. It would most probably not nail all forms of state liability. In particular with regard to regulatory measures adopted by a Member State before its accession to the EU, the application of State aid rules must be more precisely calibrated.

The Micula arbitral award established in December 2013 that by annulling an investment incentive scheme four years prior to its scheduled expiry in 2009, Romania failed to comply with its obligations assumed via the Romania-Sweden BIT, which had come into force in 2003.

In its decision of 30 March 2015, the Commission found that the compensation paid by Romania according of the named arbitral award breached the State aid prohibition. State aid is forbidden unless notified and approved by the European Commission. An extra intricacy of the present case relates to the fact that at the moment, when the notification had to be made, Romania was not yet a Member State of the Union. The standstill obligation existed according to the acquis, though the prerogatives of control were attributed to the Romanian Council of Competition.

II.      State aid Definition

According to the Commission, the revoked investment incentive scheme selectively favoured certain investors and was therefore deemed to be incompatible with the state aid rules. In order to classify a national measure as State aid, the following criteria must be examined and cumulatively fulfilled:

  • The measure must confer a selective economic advantage upon an undertaking;
  • The measure must be imputable to the state and financed through state resources;
  • The measure must distort or threaten to distort competition;
  • The measure must have the potential to affect trade between Member States.

Four conditions will be analysed in this post: the presence of an advantage, its selectivity, its imputability and the definition of an undertaking which may benefit from the aid.

A.  An economic advantage

A.1. A derogation from ‘normal taxation’ is an advantage

The concept of State aid is broader than that of a subsidy, since it comprises not merely positive benefits, such as subsidies themselves, but also interferences which, in diverse forms, mitigate the charges that are regularly included in the budget of an undertaking and which, without therefore being subsidies in the strict meaning of the word, are comparable in character and have the same effect[ii].

The concept of aid has constantly been interpreted by the CJEU as not covering measures that distinguish between undertakings in relation to charges, where that differentiation is the result of the nature and general scheme of the fiscal system. The very existence of an advantage may be established only when compared with ‘normal’ taxation’[iii].

The line of argumentation followed by the Commission asserts that ‘by repealing the EGO 24 scheme, Romania re-established normal conditions of competition on the market on which the claimants operate, and any attempt to compensate the claimants for the consequences of the revocation of the EGO 24 incentives grants an advantage not available under those normal market conditions’[iv]. I see no valid explanation for choosing the test of ‘normal market conditions’ instead of the benchmarking of ‘normal taxation’ in the present case.

In its recent negative decision in case SA.38375[v], the Commission repeats this choice while applying the ‘normal market conditions’ test (one form of it, the Market Economy Investor Principle) to a case concerning tax rulings, despite the fact that the measures concerned were administrative i.e. non-economic in nature[vi].

The Romanian legislation in Micula granted a derogation from the general regime of taxes and customs duties and the award re-established the initial economic advantage by ordering the compensation of those previously abolished fiscal facilities.

The AG Colomer[vii] has noticed in his Opinion in the Atzeni case that even if the entitlement to compensation is recognised, the amount prescribed cannot be equal to the sum that must be recovered according to the standstill prohibition enshrined in Article 108(3) TFEU and the article 14 of the Regulation No 659/1999[viii]. The Commission affirms that the Award was based on an amount corresponding to the fiscal facilities provided by EGO24 including lost profits plus interest.

The compensation provided for by the Award is based on an amount corresponding to the customs duties charged on raw materials, lost profits and interest on the total sum of damages awarded[ix].

However, the calculation of State aid cannot be based on the Award, but it must be assessed independently taking into consideration the difference between the ordinary expenses and the subsidised expenses under EGO24 with reference to the situation prior to payment of the aid[x]. In Dunamenti it has been established that even if the Article 107 TFEU became applicable on the accession date, the analysis of the measure must be done in the context of the period in which it had been granted.

The relevant factual circumstances of the grant cannot be disregarded solely because they have preceded the accession[xi]. In case, the Award exceeded the amount of aid granted under EGO24, which is the regulatory measure examined by the Romanian Council of Competition in its decision of May 2000, this excess cannot be deemed to constitute indirect State aid as established by the Atzeni Opinion.

A.2.   Compensation for damages is not an advantage

The principle of State responsibility for loss and damage caused to individuals because of breaches of European Union law for which the State can be held liable is enshrined by the system of the treaties on which the European Union is based[xii]. In Asteris, the CJEU established that State aid is fundamentally different in its legal nature from the amounts paid to individuals as compensation for the damage caused by public authorities and that, in consequence, such damages do not constitute aid for the purposes of Articles 107 and 108 TFEU[xiii].

Acknowledging the fact that EU law does not allow an unconditional permeation of obligations derived from international law, the main difference between a Micula entitlement and the Asteris right to bring an action for payment is the source of the obligations on which the claim for payment is based.[xiv] Greece was expected to implement EU law obligations, while Romania is responsible for implementing an obligation arisen from a BIT and affirmed by an ICSID-award.

The CJEU recognised a distinction between an action for damages under Article 340 TFEU and an action for payment concerning the liability of the national authorities responsible for implementing Union law, which individuals are seeking to establish before national courts[xv]. Hence, it can concluded that the Asteris liability is a distinct case, since it does not entail a sufficiently serious breach of an EU law obligation borne by the State.

On the other hand, the entitlement to payment in Micula ought to be interpreted in the light of the relevant rules of customary international law, which is part of the Union legal order and is compulsory for its institutions[xvi]. The BITs obligations go beyond the field of customary international law, ensuring a higher level of protection for the investor that can be recognised under EU law, only if it concurs with the specific characteristics and the autonomy of the Union legal order[xvii]. The Commission considered that the BIT obligations were not conform to EU law and subsequently, the action for payment could not be supported by an Asteris claim.


[i] Read more at: http://www.nationalreview.com/corner/378955/obama-west-point-because-we-have-best-hammer-does-not-mean-every-problem-nail-andrew.

[ii] Case 30/59 De Gezamenlijke Steenkolenmijnen in Limburg v High Authority [1961] ECR 1.

[iii] Case 173/73 Italy/Commission [1974] ECR I- 00709 [15].

[iv] Commission Decision (EU) 2015/1470 of 30 March 2015 on State aid SA.38517 (2014/C) (ex 2014/NN) implemented by Romania, OJ L 232, 4.9.2015, p. 43–70 [92].

[v] Commission Decision, State aid SA. 38375 (2014/NN) (ex 2014/CP) Brussels, 11.06.2014. The final decision of 21 October 2015 not yet published. Read also my commentary (click here).

[vi] Read more at: http://www.kluwertaxlawblog.com/blog/2015/10/28/the-interplay-between-the-state-aid-rules-and-other-beps-preventing-tools-sa-38375/.

[vii] AG Opinion, AG Colomer, Joined Cases C-346/03 and C-529/03 Atzeni [2006] ECR I-01875 [198].

[viii] ‘It should be noted that, if an entitlement to compensation is recognised, the damage cannot be regarded as being equal to the sum of the amounts to be repaid, since this would constitute an indirect grant of the aid found to be illegal and incompatible with the common market’.

[ix] Commission Decision (EU) 2015/1470 of 30 March 2015 on State aid SA.38517 (2014/C) (ex 2014/NN) implemented by Romania, OJ L 232, 4.9.2015, p. 43–70 [123].

[x] Case T‑473/12 Aer Lingus, not yet reported [83]. See, to that effect, C‑277/00 Germany /Commission [2004] ECR I-03925 [74-5].

[xi] Opinion AG Wathelet, Case C‑357/14 P Dunamenti Erőmű, not yet reported [121].

[xii] Joined Cases C-6/90 and C-9/90 Francovich and Others [1991] ECR I-5357 [35]; Joined Cases C-46/93 and C-48/93 Brasserie du Pêcheur and Factortame [1996] ECR I-1029 [31]; and Case C‑445/06 Danske Slagterier [2009] ECR I‑0000 [19].

[xiii] Joined cases 106 to 120/87 Asteris/Greece (Asteris III), [1988] ECR I-05515.

[xiv] Opinion 2/13 of 18 December 2014, not yet reported [201].

[xv] Asteris III [25-6].

[xvi] Case C‑179/13, Evans, not yet reported [35]

[xvii] Opinion 2/13 of 18 December 2014, not yet reported [174].


Emanuela Matei,  Associate Researcher at the Centre of European Legal Studies, Bucharest. Juris Master in European Business Law (Lund University, June 2012), Magister legum (Lund University, June 2010), BSc in Economics & Business Administration (Lund University, June 2009).

Defining International Investment Law for the 21st Century (A Reply)

by Emanuela Matei, Of Counsel – Mircea and Partners*

This post represents a reply to Horia Ciurtin’s material “The Future of Investment Treaties: Metamorphosis or Deconstruction?”, published on the EFILA Blog on 8th September. Another reply will follow from Horia Ciurtin in the following weeks.

Of Two Evils Choose Neither

We are living in a hologram designed by a very confused mind. Witnessing the 21st century we all experience a degree of restlessness and fuzziness. In this context, the choice between two evils may be no more than a false dilemma. The misconception of the limits of international law is part of this holographic picture.

In his post “The Future of Investment Treaties: Metamorphosis or Deconstruction?“, Horia Ciurtin revisits the challenging task of defining – in our not-so brave new world – the concept of international law, in general, and of investment treaties law, in particular. I both agree and disagree with the author’s concerns. I fully agree with him that international law and legal institutions can provide effective means to solve human problems. I disagree with the either-or equation though and I will describe it as a deceiving choice between two evils.

The First Evil

In a world where the interactions are multiple and ubiquitous, it is very often not possible to determine which event occurs first and define it as the cause of a subsequent event, called effect. State interests do not exist outside the social sphere and the actions of states are therefore influenced by the attitudes of non-state constituencies. In other words, the border between state and non-state has been blurred.

It is up to the observer to judge. If the observer believes that coercion is the source of order and well-being in the world, he will naturally think that international law cannot have an influence on actual state behaviour. Such an observer sees international law as a source of democratic concern, arguing against the implementation of international law norms domestically. In my view, this hostile approach is the first evil and – so far – Horia Ciurtin and I agree with each other.

The Second Evil

The affirmation that the sovereign entities are “no longer needed as ‘procedural proxies’ for aggrieved investors, being able themselves to directly involve in international litigation and be compensated for their losses” is on the other hand not immune to criticism. Having a right and being able to exercise it effectively should be seen as two sides of the same coin. A right, which is not enforceable has no legal significance. It has only a symbolic value. States comply with international law as long as the social sphere – in which their interests are continuously defined – requires them to do so.

Moreover, the author pleads for the de-politicisation of the disputes by unconditionally escaping the domestic remedies. My counterargument is that such disputes are nonetheless political in nature, so their de-politicisation would provide no more than an empty gesture.

For a legal pragmatist as I am, the ICSID-convention is a tool designed to serve a set of functions. It is nothing unexpected in the fact that this tool has been designed at a certain moment in time and that time is gone. The question that must be answered is what kind of functional design shall be chosen for the 21st century FTAs? Attention, the designer may be somebody else than before! Again, the political configuration which is part of the social sphere is different now compared with 1950!

Furthermore, the situation of intra-EU BITs is a special case. I believe that the comparison between the South America and the Central-Eastern Europe is a bit misplaced. A conflict between supra-state constitutional law and international law obligations on one side, and between individual rights derived from international law and the obligation of the state to implement supranational law, on the other, constitutes an extra-complication that must be faced by countries like Romania, Hungary or Slovak Republic.

The either-or dilemma is often projected by the advocates of arbitration as a support for the affirmation that without an ISDS-system the protection of the investor will be severely depreciated. It can be true that some strategic contrivances will no longer be available. However, it must be recognised that the accession to the EU of the Central-Eastern European countries had a positive impact on their legal systems and the socio-economic environment is now more stable than in the nineties and early noughties.

More than so, the capital is the most mobile of all factors of production. If some jurisdictions became hostile to investors, the capital would vote with its feet as it does in all other cases, where the regulatory choices of the state or supra-state give an incentive to corporations to move, stay or entry. Thus, my contemplation of the post-Westphalian field of battle is much more optimistic in this particular sense. The second evil – no protection for the investor in the 21st century – is nothing else than a false alarm!

The discussion starts to sound irresistibly interesting to me when we begin to imagine deterritorialised ideas of governance … but this is a different kind of story. This is the true and exciting post-Westphalian realm left unexplored by the mainstream despotique!


* Emanuela Matei, Jurismaster; Of Counsel – Mircea and Partners; Associate Researcher – Centre for European Legal Studies.