OPINION OF ADVOCATE GENERAL

MAZÁK

delivered on 20 October 2011 (1)

Case C‑124/10 P

European Commission

v

Électricité de France (EDF) and Others

(Appeal — State aid — Selective tax exemption linked to an increase in share capital during the recapitalisation of an undertaking — Market economy investor principle — State acting as shareholder and State wielding public power)

1.        The Commission is asking the Court to set aside the judgment (2) by which the Court of First Instance (now ‘the General Court’) annulled Articles 3 and 4 of the Commission Decision on aid measures in favour of EDF and the electricity and gas industries. (3) This appeal is of particular significance for EU State aid law. It raises a question of principle, which concerns the scope (applicability) of a central tenet of EU State aid law — namely, the Market Economy Investor Principle (MEIP) — in a case where the State has exercised its prerogative of wielding public authority. Essentially, the MEIP (or ‘the Private Investor Test’) is the logical expression of the principle, derived from Articles 86 EC and 295 EC, of the equal treatment of public and private enterprises.

I –  Legal framework

2.        Article 38(2) of the French General Tax Code provides that ‘the net profit is the difference between the value of the net assets at the close and at the opening of the period for which the surplus constitutes the taxable income, with the deduction of additional contributions and the addition of amounts withdrawn by the owners or members in that period. Net assets are the amount by which the assets exceed that portion of total liabilities comprising third-party liabilities, verifiable depreciation and accounting provisions.’

3.        Under paragraphs I and II, respectively, of Article 4 of Law No 97‑1026, (4) the structures of the high-voltage electricity transmission network are to be deemed to have been owned by EDF from the time that it was granted the concession for that network and, for the purposes of applying the provisions of paragraph I, as at 1 January 1997, the value of the assets in kind allocated under concession to the high-voltage transmission network appearing as liabilities on EDF’s balance sheet is to be entered, net of the corresponding revaluation differences, under the item ‘Capital injections’.

II –  Background to the dispute

4.        The facts, the administrative procedure and the contested decision are set out in detail in paragraphs 9 to 51 of the judgment under appeal. I shall confine myself to the most essential points. EDF produces, transmits and distributes electricity. At the material time, EDF was wholly owned by the State. It was created by a law which nationalised electricity and gas and which established the principle of transferring to EDF the nationalised electricity concessions. In 1958, the various electricity transmission concessions granted by the State were unified into a single concession known as the ‘high-voltage transmission network’ (réseau d’alimentation générale, ‘the RAG’).

5.        The application to EDF of the General Accounting Plan of 1982, which included accounting rules specific to concessions, led to a change in the accounting treatment of the RAG, in order to take into account certain recommendations made by the National Accountancy Council (Conseil national de la comptabilité, ‘the CNC’). An accounting plan specific to EDF was introduced and was approved by interministerial order. The RAG was entered in the assets on EDF’s balance sheet under the item ‘Fixed assets operated under the concession’. Specific accounting provisions were created for the renewal of the fixed assets operated under the concession in order to enable EDF to return those assets in perfect condition at the end of the concession.

6.        Expenditure incurred by EDF for the purposes of renewal was recorded in the balance sheet item ‘Value in kind of assets operated under concession’ (also called ‘grantor rights’). That item represented a debt which EDF owed to the French State and which was linked to the return without consideration of the replaced assets at the end of the concession. In a 1994 report, the French Cour des comptes (Court of Auditors) noted the irregular nature of the tax relief received by EDF following the irregular creation of the accounting provisions for renewal of the RAG. The French State therefore undertook to restructure EDF’s balance sheet. The ‘State-EDF 1997-2000’ management contract, signed on 8 April 1997, made provision for the normalisation of EDF’s accounts and its financial relationship with the State, with a view to the opening up of the market in electricity as provided for in Directive 96/92/EC of the European Parliament and of the Council of 19 December 1996 concerning common rules for the internal market in electricity (OJ 1997 L 27, p. 20).

7.        Prior to the adoption of Law No 97‑1026, EDF’s balance sheet was presented as follows: (i) on the assets side, an item entitled ‘Fixed assets operated under the concession’ in the amount of FRF 285.7 billion, of which approximately FRF 90 billion was in respect of the RAG; (ii) on the liabilities side, an item entitled ‘Provisions’ of which approximately FRF 38.5 billion was in respect of the RAG, and an item entitled ‘Value in kind of assets operated under concession’, which recorded the expenditure for renewal incurred and amounted to FRF 145.2 billion, of which FRF 18.3 billion was accounted for by the RAG.

8.        Pursuant to Article 4 of Law No 97‑1026, the way in which to restructure the upper part of EDF’s balance sheet was decided and communicated to EDF, on 22 December 1997, by letter of the Minister for Economic Affairs, Finance and Industry, the State Secretary for the Budget and the State Secretary for Industry (see points 25 and 26 below).

9.        In the context of the administrative procedure, the Commission took three related decisions concerning EDF which were published on 16 November 2002 (OJ 2002 C 280, p. 8). The Commission initiated a formal investigation under Article 88(2) EC into the advantage accruing to EDF as a result of the non-payment of the corporation tax due on some of the accounting provisions created, free of tax, for the renewal of the RAG. Article 3 of the contested decision categorises as State aid incompatible with the common market the non-payment by EDF, in 1997, of corporation tax on some of the accounting provisions created free of tax for the renewal of the RAG, corresponding to FRF 14.119 billion in grantor rights reclassified as capital injections and states that the aid in question amounts to EUR 888.89 million. Under Article 4 of the contested decision, France was to recover the aid (with interest) of EUR 1.217 billion, and that sum was in fact repaid to the French State.

III –  The judgment under appeal

10.      In support of its action, EDF essentially raised three pleas in law. The General Court limited its assessment to the first plea and the first three parts of the second plea, rejecting the first plea and the first two parts of the second plea. It upheld the third part of the second plea, annulling Articles 3 and 4 of the contested decision, in respect of which EDF had argued in essence that, by implementing the measure in question, the State had acted like a prudent private investor in a market economy.

11.      In paragraphs 233 to 237 of the judgment under appeal the General Court considered essentially that the fact that it has taken the form of a law is not sufficient to exclude the possibility that the intervention of the State in the capital of an undertaking pursues an economic objective which might also be pursued by a private investor. In paragraphs 243 to 245, it considered in essence that the provisions of Article 4 of Law No 97‑1026 are not tax provisions per se; rather, they are provisions concerning accounting, but carrying tax implications. In paragraphs 247 to 250, the General Court considered essentially that, in the light of the objective pursued, the mere fact that the claim held by the French State against EDF is fiscal in nature and the sole fact that the French State used legislation as its instrument do not mean that the Commission may legitimately refuse to apply the MEIP.

12.      In paragraphs 251 and 252 of the judgment under appeal, the General Court stated that it cannot be ruled out that the form taken by an investment may give rise to differences in terms of the cost of raising the capital and in terms of the return on that capital, from which it could be concluded that a private investor would not have made such an investment in comparable conditions. However, that presupposes an economic analysis in the context of the application of the MEIP, which the Commission had deliberately failed to apply. In paragraph 253, the General Court concluded that the Commission could not examine merely the tax implications of the provisions adopted by the French Republic without at the same time examining — and possibly rejecting after that examination — the merits of the line of argument that the operation satisfied the MEIP.

IV –  Appraisal

13.      The Commission relies on two grounds of appeal: (i) the General Court misinterpreted and distorted the facts of the case and (ii) the General Court made an error of law when it assessed the scope of the MEIP. It is not necessary to reproduce in detail all the parties’ arguments in a separate section. Rather, in order to facilitate the reading of the Opinion, I shall integrate those arguments into my analysis.

A –     The first ground of appeal: distortion of the facts

14.      Essentially, the Commission, Iberdrola and the EFTA Surveillance Authority (‘the Authority’) claim that, contrary to the statements made in the judgment under appeal, the French Republic did not in fact convert a tax claim into capital, but simply granted EDF aid in the form of a corporate tax exemption. EDF and the French Government, on the other hand, contend that the first ground of appeal should be rejected.

15.      First of all, EDF submits that the first ground of appeal is inadmissible, because it is asking the Court to re-examine the assessment made by the General Court in relation to the restructuring of EDF’s balance sheet.

16.      I would point out that, as regards the ground concerning distortion of evidence, whilst it is indeed for the General Court alone to assess the value which should be attached to the items of evidence produced before it, (5) the Court of Justice has none the less held that such a ground is admissible. (6) In the present case, the objection of inadmissibility raised by EDF should be rejected. The Commission is clearly not simply seeking a re-examination of the General Court’s assessment of the restructuring of EDF’s balance sheet. Rather, it emerges from the submissions and documents before the Court that there is a real and present risk that the General Court may have distorted the facts of the case and erred in its characterisation of those facts.

17.      Indeed, according to settled case‑law, where an appellant contends that the General Court has made findings which the documents in the file show to be substantially incorrect, or that it has distorted the clear sense of the evidence before it, objections based on findings of fact and their assessment in the judgment under appeal are admissible. (7) The evaluation of evidence is clearly a matter for the General Court and cannot be reviewed by the Court of Justice, but that is not the case where the General Court has distorted evidence by inferring from it things which it clearly did not contain — and that, in my view, is precisely the case here. (8) Distortion of evidence (9) (dénaturation) occurs when the court dealing with the substance of a case exceeds its jurisdiction by interpreting a clearly and unambiguously worded document (such as an agreement, a will, a report, a judgment, or a foreign law) in a way incompatible with that wording. (10)

18.      It follows from all the foregoing considerations that the first ground of appeal is admissible.

19.      Turning to the substance of the first ground of appeal, it would appear that the General Court erred indeed in the characterisation of the facts — in particular, insofar as it held that the French Republic had converted a tax claim into capital (see, inter alia, paragraph 258 of the judgment under appeal), whereas what that State did, in fact, was to grant EDF a selective corporate tax exemption. (11)

20.      Contrary to the arguments put forward by EDF, and as will be made apparent below, the distortion of facts is obvious from the documents before the Court and, accordingly, it is not necessary for the Court to proceed with a new detailed assessment of facts.

21.      The French Government seeks to justify the General Court’s approach by insisting that the conversion into capital of a tax claim was undertaken as part of the restructuring of EDF’s balance sheet, one of the steps in that process being a recapitalisation.

22.      No one disputes the fact that a recapitalisation of EDF was under way at the same time, which was not considered in itself to constitute State aid. However, it is important to point out that it was the tax implications — which are dissociable from the recapitalisation — which were categorised as State aid (see paragraph 241 of the judgment under appeal). I consider the French Government’s argument futile, because to my mind the renunciation of a tax debt and its conversion into a capital injection must manifestly be regarded as constituting two distinct and consecutive operations. In fact, in paragraph 97 of its response, the French Government would seem to concede this.

23.      Manifestly, a reading of the case-file confirms that it is impossible to find the taxable amount in any of those documents. There is no tax assessment notice, no trace of a payable tax claim and no decision on the allocation of the taxable amount to the recapitalisation of EDF. In fact, if the General Court’s interpretation of the documents and facts submitted to it were to be followed, the allocation of the taxable amount would fall to be deduced simply from the silence of Law No 97‑1026 and/or the letter of 22 December 1997 on this point.

24.      In fact, the General Court states in paragraph 242 of the judgment under appeal that ‘all the parties agree that tax would have been payable on the amount of FRF 14.119 billion before it was recorded under the item entitled “Capital injection”’.

25.      It follows from paragraphs 239 to 242 of the judgment under appeal that all the measures for restructuring EDF’s balance sheet were recorded on its balance sheet, with the two-fold aim of strengthening its net assets and stabilising the financial relationship between the State and EDF on a basis similar to ordinary law (see paragraph 31 of the judgment under appeal).

26.      The following points, in particular, should be noted: (i) the assets constituting the RAG were reclassified, in the amount of FRF 90.325 billion, as ‘own assets’; (ii) the unused accounting provisions for renewal of the RAG were posted as retained earnings and reclassified as accumulated losses, that account being thereby reconciled and the balance allocated to the reserves (I would add that that transaction was taxed); and (iii) the ‘grantor rights’ were allocated directly to the capital injections item in the amount of FRF 14.119 billion.

27.      Importantly, however, the conversion into capital of the taxable amount, from which EDF was exempted, was in no way whatsoever recorded in the undertaking’s accounts.

28.      In fact, ‘… the French authorities themselves acknowledge that the transaction was illegal. In a memorandum dated 9 April 2002 addressed by the Directorate-General for Taxation [of the French Ministry of Economic Affairs, Finance and Industry] to the Commission, the French authorities stated that “the grantor rights in respect of the RAG represent an unowed debt which was unjustifiably exempted from tax by being incorporated into the capital” and that “before this reserve was incorporated into the capital, it should have been transferred from the enterprise’s liabilities, where it was incorrectly posted, to a net assets account, thereby resulting in a positive variation in net worth that was taxable under Article 38(2)” of the General Tax Code’ (see recital 89 of the contested decision).

29.      Despite the unambiguous opinion expressed by the Directorate-General for Taxation, no tax claim was duly recorded before the amount covered by the ‘claim’ was injected as capital.

30.      Indeed, it would appear that, in the present case, the ‘claim’ never existed and there was never any debt to waive. As Iberdrola pointed out, there was no conversion of the claim to capital nor, apparently, was there any desire on the part of the authorities to carry out such a conversion. It would even appear that the General Court relied on the attempt made by the French Government to justify its actions a posteriori on the basis of the MEIP. To my mind, such allegations — in particular, as they are presented in tempore suspecto — cannot retrospectively transform a tax exemption into a course of action which can be regarded as ‘comparable’ to the conversion of a claim into capital.

31.      In paragraph 282 of the judgment under appeal, the General Court recognised that there was never any claim or debt to waive: ‘… before Law No 97‑1026 was adopted and EDF was deemed to own the RAG, no tax was owed. Furthermore, since the French Republic waived the right to levy the tax, that tax was no longer owed and therefore could not be posted in the balance sheet as a debt owed by the undertaking’.

32.      I agree with the Commission that, if the General Court’s reconstruction of the facts were to be accepted, it would result in a capital increase which is implicit and non-transparent, undertaken without the express agreement of the legislature and against the advice of the tax authorities.

33.      In that connection, paragraph 243 of the judgment under appeal is central to the General Court’s reasoning. There, the General Court appears to use the argument that Article 4 of Law No 97‑1026 was intended to restructure EDF’s balance sheet and increase own funds in order to conclude that ‘consequently, the provisions of that article are not tax provisions per se, but rather accounting provisions having tax implications, as is shown by the letter sent to EDF … on 22 December 1997’.

34.      In fact, the remainder of the General Court’s reasoning is based on the finding that the French State had carried out an investment and, in particular, that it had ‘increase[d] the capital of [EDF] by waiving that tax claim’ (see paragraph 246 of the judgment under appeal). A reading of the relevant passages of the judgment under appeal confirms that, in fact, this element is the key to the General Court’s reasoning and that, for the General Court, it represented the essential point that had to be resolved in order to rule on the third part of EDF’s second plea.

35.      Indeed, that idea can be traced throughout the whole judgment. It can be found in more or less explicit terms in paragraphs 248, 249, 250, 252, 253, 258 and 259. In addition, the General Court refers to the objectives purportedly pursued in paragraphs 229, 233, 234, 235, 236, 237, 247 and 259 of the judgment under appeal.

36.      Clearly, the General Court based its assessment on the terms of Article 4 of Law No 97‑1026. Essentially, all the argumentation of the French Government and EDF relies heavily on the terms of that provision. They argue that it is clear from those terms that the General Court was right to consider that the assistance given in 1997 was a capital injection. However, contrary to those arguments, I consider that the terms of Article 4 of Law No 97‑1026 do not lead to the inevitable conclusion that that assistance was a capital injection. Moreover, the French Government has not demonstrated in its submissions that it flows necessarily from the nature of the financial operations that they constitute a capital injection. It is clear from the foregoing considerations that the French Government’s argument that the alleged capital injection was carried out in the ‘most transparent’ way (because it takes the form of a law) must also be rejected.

37.      At most, it was transparent in form: it was certainly not transparent in content.

38.      Manifestly, Article 4 of Law No 97‑1026 merely sets out the accounting consequences of the financial operations covered. The fact remains that it does not set out the tax consequences or implications of the operations in question. In fact, EDF would seem to admit that in its defence. As the Commission noted, during the legislative process, a proposed amendment to the text of the law was rejected because it sought to define the accounting process by which the accounting provisions relating to the structures in question would be converted into own funds. (12)

39.      First, suffice it to point out that there is nothing in Law No 97‑1026 to suggest that the French State carried out an operation which consisted, in essence, in the conversion of a tax claim into EDF’s capital. Secondly, it emerges from the report of the finance committee of the National Assembly, drawn up in the course of the consideration of the text, that the legislature did not decide on the tax process by which the accounting provisions relating to the structures of the electricity transmission covered by the concession to the high-voltage transmission network would be converted into own funds. Importantly, that report confirms that the law contains no indication of the tax consequences of the accounting reclassification to be carried out pursuant to its provisions.

40.      As regards the letter of 22 December 1997 from the French Government to EDF, it would appear that the French authorities themselves made a clear distinction between the measures relating to the restructuring of the balance sheet (see Annex 1 to that letter) and the tax consequences (see Annex 3 to that letter). As was pointed out by the Commission, Annex 1 simply shows the direct allocation of the grantor rights to the capital injections item in the amount of FRF 14.119 billion, whereas Annex 3 removes the tax consequences of this measure for EDF. Again, there is nothing in that document to suggest that a tax claim was converted to capital.

41.      With regard to the letter of 9 April 2002 from the French authorities to the Commission, the situation is the same: again, there is nothing to suggest that the French State carried out a conversion of a tax claim to capital.

42.      It follows from the documents before the Court that it was late in the administrative procedure — not until 9 December 2002, 18 months from the start of the investigation — that, in the observations sent to the Commission, the French authorities sought to present the tax exemption granted to EDF as an operation ‘comparable’ to a ‘further capital injection’, and, importantly, in doing so, the French authorities provided no objective evidence to support those claims.

43.      Indeed, as the Commission rightly pointed out, in its observations submitted before the General Court on 3 February 2005, EDF presented a study entitled ‘A hypothetical market investor perspective’, but even that study was carried out only ex post in an attempt to justify retrospectively the economic logic of the transaction in question. That is to say, there was an attempt to present that transaction as the conversion of a claim into capital, an interpretation of events which, on the facts, would appear not to have matched reality, or ever to have been analysed as such by the French Republic, at the time of the restructuring in question.

44.      Next, the fact that the General Court departed from the facts of the case by distorting them to such an extent that it ultimately ruled on a case other than the case before it is evident from paragraph 242 of the judgment under appeal. In that paragraph, the General Court emphasised that all the parties agreed that tax would have been payable on the transaction in question. By contrast, in paragraphs 266 to 269, the General Court goes on to express a doubt that it harbours as to whether the tax is actually payable in such circumstances. It would appear from the documents before the Court of Justice, however, that the transaction in question — in the form in which it was carried out — should indeed have given rise to the collection of tax.

45.      To my way of thinking, it was not open to the General Court to find as a fact that a capital injection had taken place as a result of the abandonment of a tax claim, in the absence of any objective evidence attesting to the actual existence of such a transaction.

46.      Accordingly, since such objective evidence was manifestly lacking, the General Court’s conclusion that there was a capital injection is a finding based on a distortion of the documents before it.

47.      Alternatively, as the Commission noted, if the General Court’s finding were interpreted, not as a finding of an objective fact, but as a legal characterisation of the facts, based on the fiction of an implicit and non-transparent conversion of a tax into capital, or as a transaction corresponding to such a conversion, then that legal characterisation of the facts would be manifestly erroneous, because it would invalidate the whole of the General Court’s reasoning.

48.      Lastly, paragraph 225 of the judgment under appeal rightly points out that, according to the case‑law of the Court of Justice, the MEIP is not material where the intervention by the State has no economic character. That is the position where the public authorities pay a subsidy directly to an undertaking, or grant an exemption from tax, or agree to a reduction in social security contributions. I agree with the Commission, therefore, that it would even appear as if, by re-characterising the transaction in question, the General Court sought in fact to circumvent the case‑law of the Court of Justice regarding the scope of the MEIP.

49.      It follows from all the foregoing considerations that the first ground of appeal should be upheld.

B –     The second ground of appeal: interpretation of Article 87 EC

1.      First part of the second ground of appeal: criteria for distinguishing between the State as shareholder and the State as wielding public power

50.      In essence, the Commission, Iberdrola and the Authority contest the General Court’s finding that the distinction between the State as shareholder and the State wielding public power depends primarily on the objective pursued by the State (in the present case, the recapitalisation of EDF) and not on elements which are objective and verifiable, as required by the case‑law. EDF and the French Government contend that the first part of the second ground of appeal should be rejected.

51.      First of all, EDF and the French Government argue essentially that the second ground of appeal is based on the alleged distortion of facts by the General Court, which forms the subject-matter of the first ground of appeal. They consider that the two grounds of appeal are complementary and that the first constitutes, in fact, an indispensable premiss for the second.

52.      However, in my view, the second ground of appeal is manifestly separate and independent from the first. This is true, notwithstanding the fact that the appeal necessarily concerns Articles 3 and 4 of the contested decision, concerning the non-payment by EDF in 1997 of certain amounts of corporation tax. Indeed, as the Authority and Iberdrola rightly pointed out, Articles 3 and 4 of the contested decision can stand or fall whether the aid measure in question is characterised as a capital injection or as a non-payment of tax, if no rational and prudent private investor would make those funds available to a publicly owned undertaking on ‘terms which a private investor would find acceptable in providing funds to a comparable private undertaking when the private investor is operating under normal market economy conditions’. (13)

53.      Clearly, the material finding made by the Commission in that part of the contested decision was that the financial assistance given by the French State in 1997 constituted illegal State aid.

54.      I consider that, in any event, the fact remains that the second ground of appeal is well founded. Indeed, it is common ground that, in the present case, the French State exercised its prerogatives as a public authority.

55.      Turning to the substance of the first part of the second ground of appeal, I believe that one of the most important issues in the present appeal is the paramount importance which the General Court attaches to the objective of the measure under assessment, coupled with the fact that that approach finds no basis in the case‑law of the Court of Justice.

56.      As consistently interpreted by the Court, Article 87(1) EC ‘does not distinguish between measures of State intervention by reference to their causes or their aims but defines them in relation to their effects’. (14) Indeed, the nature of the objectives underlying a State measure is not sufficient ipso facto to take it outside the ambit of State aid. Otherwise, the Member State would only have to invoke the legitimacy of the aim pursued by the public intervention to avoid the application of the Treaty rules on State aid. (15)

57.      Here EDF counters with the argument that the General Court did not assess the measure in question exclusively on the basis of its objective. According to EDF, the General Court assessed that measure in the light of a variety of criteria, such as its nature, object and objectives, and by reference to all its aspects and its overall context.

58.      While it may be accepted that, prima facie, the General Court did not base its approach exclusively on the objective pursued, the fact remains that a closer reading of the whole judgment under appeal shows that the General Court did, in fact, base its findings principally — if not exclusively — on the objective pursued, in order to determine whether the State acted qua shareholder or qua public authority. As will be apparent, it cannot be disputed that, in its analysis, the General Court attributed predominant significance to the objective pursued.

59.      In paragraph 229 of the judgment under appeal, the General Court held that ‘[i]n order to determine whether measures taken by the State represent the exercise of State authority or whether they are the consequence of obligations that the State must assume as shareholder, it is important to look not at the form of those measures, but at their nature, their subject-matter and the rules to which they are subject, while taking into account the objective pursued’ (emphasis added).

60.      Next, in paragraph 233, the General Court again underlines the need to determine whether the State pursued an ‘economic objective’ which might also be pursued by a private investor, in order to establish whether or not the MEIP is applicable. Paragraphs 234 and 235 also mention that the objective pursued is important in this context. In paragraph 236, the General Court concludes that ‘the measure must be examined not solely according to its form, but on the basis of its nature, its object and its objectives, which presupposes that all aspects of it are examined and that its context is taken into consideration’ (emphasis added). Lastly, paragraph 237 also makes a reference to the objective of the measure under consideration.

61.      Then, it is in paragraph 247 that the General Court once again states — decisively — that ‘in the light of the objective of recapitalising EDF pursued by Law No 97‑1026, the mere fact that the claim … is fiscal in nature and the sole fact that the French State used legislation do not allow the Commission to refuse to ascertain whether, in similar circumstances, a private investor could have been persuaded to inject the same amount of capital and, therefore, whether the capital was provided by the State in circumstances corresponding to normal market conditions’ (emphasis added).

62.      Thus, in order to rule on the applicability of the MEIP, the General Court manifestly took into account, and attached decisive importance to, the objective pursued and, in consequence, that ruling is based on an error of law.

63.      In addition, it is clear that the only case‑law on the basis of which the General Court seeks to justify taking into account, in the context of the applicability of the MEIP, the objective allegedly pursued by the Member State is the judgment of the Court of Justice in SAT Fluggesellschaft. (16)

64.      However, in the relevant paragraph of that judgment (paragraph 30), the Court of Justice in no way refers to the objective pursued by the measure in question.

65.      First of all, SAT Fluggesellschaft is not a case which concerns State aid. That judgment sought to determine whether a certain international organisation constituted an ‘undertaking’ within the meaning of Articles 82 EC and 86 EC. In particular, the Court was asked whether or not the European Organisation for the Safety of Air Navigation (Eurocontrol) constituted an ‘undertaking’.

66.      In order to rule on that question, the Court examined: (i) the nature of Eurocontrol’s activities (paragraph 19), (ii) the aim of those activities as defined in the convention which established it (paragraph 21), (iii) Eurocontrol’s tasks as defined in the amended convention (paragraph 22), (iv) its competences, in particular, to establish and collect route charges (paragraphs 23, 28 and 29), (v) Eurocontrol’s operational activity (paragraph 24), and, lastly, (vi) its method of financing as an international organisation (paragraph 26).

67.      In paragraphs 30 and 31 of SAT Fluggesellschaft, therefore, the Court came to the conclusion that ‘taken as a whole, Eurocontrol’s activities … are connected with the exercise of powers relating to the control and supervision of air space which are typically those of a public authority. They are not of an economic nature justifying the application of the Treaty rules of competition’. Thus Eurocontrol did not constitute an undertaking under Articles 82 EC and 86 EC.

68.      It should be noted that, in the present case, it was the General Court itself which recognised in paragraph 246 of the judgment under appeal that ‘the State used its prerogatives as a public authority when waiving [the tax claim in question]’. It is not disputed that the measure in question is connected with the exercise of the prerogatives of public authority. Furthermore, as the Commission rightly pointed out, the test applied by the Court of Justice in SAT Fluggesellschaft concerned the categorisation of a given organisation as an undertaking and not the categorisation of a certain operation carried out by the State as State aid, as is the case here.

69.      The only conclusion which can be drawn from the foregoing considerations is that, contrary to the French Government’s arguments, SAT Fluggesellschaft was and remains not relevant for the purposes of deciding the present case and, in particular, that judgment is irrelevant for the purposes of establishing whether or not, in a given case pertaining to Article 87 EC, the MEIP is applicable.

70.      What is more, I consider that, in fact, the General Court did not even apply the criteria set out by the Court of Justice in paragraph 30 of SAT Fluggesellschaft. The General Court treated the objective pursued by the State — a criterion extraneous to that judgment of the Court — as by far the most determinative factor and, at the same time, failed to take into account the nature of the State’s action in waiving the claim in question, which, in the General Court’s own words, was an instance of the State using its prerogatives as a public authority.

71.      First, a criterion based on the intention of the State would be particularly inappropriate for assessing the existence of State aid insofar as such a criterion is, by its very nature, subjective and open to interpretation. Secondly, as was pointed out by Iberdrola, the taking into account of objectives allegedly pursued by the Member State lends itself to manipulation, affecting the competition between the private and public sectors. There is a real risk that, in order to avoid the application of State aid law, Member States would feel encouraged merely to plead (ex post) profit-making preoccupations.

72.      Lastly, the determinant role accorded by the General Court to the objective pursued is difficult to reconcile with the fact that ‘the concept of aid is an objective one’. (17) Indeed, it must not be forgotten that, according to the case‑law of the Court of Justice, ‘State aid, as defined in the Treaty, is a legal concept which must be interpreted on the basis of objective factors’. (18) Finally, the presence of an element of aid is assessed with reference solely to available objective and verifiable elements. Otherwise, the Commission’s duty of supervision would be affected and the Community courts would be unable to ‘carry out a comprehensive review as to whether a measure falls within the scope of Article 87(1) EC’, (19) and indeed that is precisely the risk should one follow the approach suggested in the judgment under appeal.

73.      Next, EDF argues that what is relevant in the framework of applying the MEIP is not only the objective pursued but also, in particular, the context. It relies on some cases in which the Commission was required to assess the context of various operations. In particular, EDF refers to Chronopost I, (20)P&O European Ferries (Vizcaya) and Diputación Foral de Vizcaya v Commission (‘P&O European Ferries’), (21)Bundesverband deutscher Banken v Commission, (22) and Ryanair v Commission (‘Ryanair’). (23)

74.      Suffice it to point out, however, that in all those cases what was at issue were transactions of a commercial nature. In Chronopost I, it was the transfer of a client base; in P&O European Ferries, the sale of ferry tickets; in Bundesverband deutscher Banken v Commission, the making available of assets to a bank which were then used to underpin its competitive business; and, lastly, in Ryanair, the fixing of airport charges, that is to say, an activity directly connected with the management of airport infrastructure, which is an economic activity. In particular, the General Court made it clear in Ryanair that it was appropriate to regard the charges in question as ‘fees’ and not as ‘taxes’ — indeed, contrary to the situation here, a scheme reducing those charges could have been put in place by a private operator.

75.      In particular, it would appear that the General Court took inspiration from its judgment in Ryanair, by which it also annulled a State aid decision of the Commission. Specifically, the General Court held in Ryanair that the Commission’s refusal to examine together the advantages granted to Ryanair by the Walloon Region and by Charleroi Airport, and to determine whether, taken together, those two entities acted as rational operators in a market economy, was vitiated by an error in law. However, it is clear that the present case must be distinguished from Ryanair. Indeed, the case‑law had already recognised the economic nature of the activities in question in Ryanair. (24) Even though the charges at issue in that case were fixed by regulatory means, the General Court considered that there was economic activity because ‘the setting of airport charges is closely connected with the use and operation of Charleroi airport, which must be described as an economic activity’. The General Court essentially considered that not recognising the fact that, by fixing the charges in question, the Walloon Region could act as a private investor would amount to discrimination between private and public airports. It should perhaps be noted that the Commission did not appeal the Ryanair judgment.

76.      However, unlike the fixing of such charges, no private operator may (ever) relinquish a tax claim. Indeed, I consider that EDF is manifestly wrong when it tries to argue that the fact that the instrument used by the State is inaccessible to the private investor cannot mean that the MEIP must be rejected. There is clearly good reason for the legal commentators’ view that the MEIP cannot apply in tax matters, that is, in cases where the advantage at issue is achieved by means of tax law. (25)

77.      To my mind, the General Court’s approach in the judgment under appeal must be rejected a fortiori because, if the State wishes to act potentially as a private investor, it still can: all it needs to do is to proceed to inject capital in an undertaking after it has exercised its fiscal powers and thus used its prerogatives of public authority. Accordingly, contrary to the arguments of EDF that the scope of the MEIP would be excessively restricted and would amount to discrimination against public undertakings, the approach I advocate here would not cause any discrimination to the detriment of public undertakings. It is designed simply to limit the risk of discrimination to the detriment of private enterprises.

78.      Next, as the Authority noted, while there are circumstances in which a State can enter into contractual and commercial relations with companies, the debt owed by EDF to the French State in the present case was not at all of a contractual or commercial nature. It was a tax debt. Taxes are imposed by the State and are not owed as a consequence of voluntary contractual or commercial relations. In any event, the Court has distinguished operations carried out by the State acting as investor from those in which it acts as creditor. (26)

79.      Moreover, according to the MEIP, whenever a company receives assistance through public funds, such assistance will be regarded as State aid if a private investor would not, under normal circumstances, have made such an investment. If, on the other hand, the State acted in the same way as any normal private shareholder acting under normal market conditions, the financial assistance would not be considered to be State aid. As the Authority pointed out, it is clear that the State imposes taxes in the exercise of its public authority. It can hardly be claimed that the State has the power to tax in its capacity as a private investor. The corollary of imposing taxes — the waiving of tax debts — is also an activity which the State undertakes as a public authority. Consequently, fiscal activities of the State — the imposition, collection, refund or remitting of taxes — are undoubtedly undertaken in the exercise of its public authority and cannot by definition be undertaken as a private investor.

80.      Indeed, that is why an important distinction has been drawn by the Court. The Court held in Spain v Commission (‘Hytasa’) (27) that ‘a distinction must be drawn between the obligations which the State must assume as owner of the share capital of a company and its obligations as a public authority. Since the three companies in question were constituted as limited companies, the Patrimonio del Estado, as owner of the share capital, would only have been liable for their debts up to the liquidation value of their assets. That means … that the obligations arising from the cost of redundancies, payment of unemployment benefits and aid for the restructuring of the industrial infrastructure must not be taken into consideration for the purpose of applying the [MEIP]’.

81.      In Germany v Commission (‘Gröditzer’), (28) the Court held that ‘in order to establish whether the privatisation of Gröditzer Stahlwerke (GS) for a negative selling price of DEM 340 million involves elements of State aid, it is necessary to assess whether, in similar circumstances, a private investor of a dimension comparable to that of the bodies managing the public sector could have been prevailed upon to make capital contributions of the same size in connection with the sale of that undertaking or whether it would instead have chosen to wind it up’. (29) Again, the Court recalled that a distinction must be drawn between the obligations which the State must assume as owner of the share capital of a company and its obligations as a public authority.

82.      Next, in Altmark, (30) Advocate General Léger noted that the Court applies the MEIP only in situations where the intervention of the State is of an economic nature. He stated that the MEIP is material because the conduct of the State is capable of being adopted, at least in principle, by a private operator acting with a view to profit (an investor, a bank, a surety, an undertaking or a creditor). However, the MEIP is not material where the intervention by the State has no economic character. In such situations, the intervention by the State cannot be adopted by a private operator acting with a view to profit but falls within the exercise of public powers of the State, such as tax policy or social policy. The MEIP is therefore not material, since, by definition, there cannot be any breach of equal treatment as between the public and private sectors. He concluded that, in consequence, the MEIP does not apply to interventions by the State which fall within the exercise of public powers.

83.      Indeed, as the General Court rightly held in Ryanair, (31) ‘while it is clearly necessary, when the State acts as an undertaking operating as a private investor, to analyse its conduct by reference to the [MEIP], application of that principle must be excluded in the event that the State acts as a public authority. In the latter event, the conduct of the State can never be compared to that of an operator or private investor in a market economy’.

84.      In Westdeutsche Landesbank Girozentrale and Land Nordrhein-Westfalen v Commission, (32) the General Court correctly noted that ‘as regards the Land’s argument based on increased tax revenue, the Court observes that the Land’s position as a public body and its position as the owner of a business must not be confused. The line of argument based on an increase in tax revenue would be wholly irrelevant for a private investor’.

85.      Lastly, it should be pointed out that the Court has already held that an entity which acts as a tax collector does not act as an economic operator. (33) In addition, according to the case‑law of the Court, the Treaty rules on competition do not apply to an activity which, by its nature, its aim and the rules to which it is subject, does not belong to the sphere of economic activity, or which is connected with the exercise of the powers of a public authority. (34)

86.      Here EDF and the French Government counter with the argument that there have been other cases in which the Court adopted a less stringent approach in situations where use had been made of State power. The French Government submits that the means are irrelevant when different forms of investment can be envisaged by the State. It maintains that the Court has already accepted that measures which were not available to private investors could be compared to the behaviour of private operators. (35)

87.      However, I consider it important not to treat in the same way two notions which are quite different. A distinction should be made between the private creditor (in a market economy) principle and the private investor (in a market economy) principle (the MEIP). Indeed, contrary to the argument of the French Republic, it is not possible to compare the position of a private investor with the position of a private creditor. Whereas an investor seeks to generate a profit by making representations to undertakings, a creditor seeks to obtain payment of sums owed to it by a debtor in financial difficulties. (36)

88.      Accordingly, it is significant that here neither the French Government nor EDF claimed that this was a case of an undertaking in financial difficulties. This is clearly not a case of a creditor seeking to recover sums owed to it by a debtor in financial difficulties. Moreover, the private creditor test does not depend on the State’s role as a shareholder: under that test, undertakings public and private are treated in exactly the same way. By contrast with the MEIP and the situation which concerns us here, in the case of the private creditor test there is no risk of unequal treatment.

89.      The line of authority to the effect that the MEIP has to be applied ‘leaving aside all social, regional-policy and sectoral considerations’, so that there can be no question of taking account of considerations that obviously relate to a Member State’s role as a public authority, flows from the earliest cases. (37)

90.      The first case‑law on the MEIP is Meura and Boch II. (38) However, the subject had already been touched upon by Advocate General VerLoren van Themaat in Intermills v Commission. (39) He stated that ‘the present case clearly relates to an injection of capital into an undertaking in difficulties, which has also already been regarded by the Commission as aid [under] Article 92 … Furthermore … the present capital participation in the undertakings in the Intermills group is mainly intended to cover losses. It seems to me beyond doubt that such a very extensive covering of losses, which artificially keeps an undertaking going, must also be regarded as aid [under] Article 92, especially as it must be inferred from the permanent nature of the injection of capital and from the fact that the factories still in production are still not profitable even four years after the aid was granted that capital could not have been obtained in the private capital market’.

91.      I consider it clear — as I try to show in the present Opinion — why it is justified that both the Commission and the Court have always been firm about maintaining a clear distinction between the State’s exercise of its public powers and its role as an investor — so that, even where the measure adopted pursuant to public powers has much the same effect as one that could perhaps have been adopted by the State in its capacity as an investor, it must nevertheless be disregarded when assessing claims that the State acted as a market investor. (40)

92.      It is my belief that, in the judgment under appeal, the General Court breached a distinction meticulously marked by the Court between acta iure gestionis and acta iure imperii. (41) In essence, to my mind, the General Court’s approach might lead to legal uncertainty and a lack of transparency, as well as (tax) privileges for public enterprises. I would add that the last possibility could have a particularly adverse impact on the numerous sectors which were recently liberalised or which are in the process of being liberalised.

93.      Specifically, the approach of the General Court in the judgment under appeal runs counter to the requirement of transparency prescribed by EU law. The Court of Justice has held, for instance, that ‘a system of undistorted competition, as laid down in the Treaty, can be guaranteed only if equality of opportunity is secured as between the various economic operators’. The Court concluded from this that ‘maintaining effective competition and securing transparency require[s] the drawing-up of technical specifications, monitoring their application and granting type-approval to be carried out by a body independent of public or private undertakings offering competing goods and/or services in the telecommunications sector’. (42) There is no doubt in my mind that transparency has a very important role to play in EU State aid law, too.

94.      It is relevant to refer in this respect to Altmark. (43) If the principles upheld in that judgment are applied, the position is that, in order for compensation for public service obligations not to constitute State aid, the recipient undertaking must, in particular, actually have public service obligations to discharge and these must be ‘clearly defined’. In addition, ‘the parameters on the basis of which the compensation is calculated must be established in advance in an objective and transparent manner’ (emphasis added). In the context of the fourth condition set out by Altmark, it is also relevant that the choice of the undertaking which is to discharge public service obligations must be decided by a public procurement procedure.

95.      To my mind, it is clear that the judgment under appeal in fact challenges certain elements of the Altmark case‑law. In Altmark, the Court, taking a prescriptive approach, sought to eliminate any possibility of manipulation on the part of the Member States (in this respect, see also point 71 above) and to bring transparency and clarity into the Member States’ activities on the market.

96.      In my view, the present case calls for the same approach. The Commission was right, therefore, to take a principled line in the contested decision, insofar as there should be a visible separation of the role of the State qua public authority from the role of the State qua shareholder. I consider that the MEIP test should not be applicable until there is a level playing field for the various economic operators, and equality before tax. It must not be forgotten that the ratio of the MEIP is precisely to prevent any discrimination as between public enterprises and private enterprises, with a view to ensuring the correct application of the Treaty provisions on State aid. As we have seen, however, in the judgment under appeal, the General Court strayed away from this equality and, accordingly, from the very raison d’être for the MEIP.

97.      It follows that the first part of the second ground of appeal should be upheld.

2.      The second part of the second ground of appeal: the General Court should have assessed whether the behaviour of the French State was open to a private operator without any privileges

98.      In essence, the Commission, Iberdrola and the Authority criticise the General Court for not having based its assessment on a comparative study of the behaviour that a prudent private operator without privileges would have adopted in similar circumstances, on the one hand, and the actual behaviour in the present case of the French State, with its prerogatives of public authority, on the other. EDF and the French Government contend that the second part of the second ground of appeal should be rejected.

99.      However, contrary to EDF’s arguments that the General Court assessed the behaviour effectively adopted by the State, I consider that its analysis is in fact based on the conduct which the French State ‘could have’ adopted — had it acted differently. The General Court carried out a study of the course of action that the State could have adopted under the ‘long plan’. This consisted in, first of all, allocating to EDF’s capital a net amount after corporation tax, then requesting EDF to pay a tax corresponding to the variation in net worth and, lastly, making a further capital injection of an amount equal to the tax paid.

100. In this connection, the Commission is right when it points out that, under the ‘long plan’, the State budget, which is closely controlled, would have guaranteed transparency — the lack of which is precisely the crux of the present case. Under the plan actually applied in this case, the resources escaped all budgetary discipline, the principle of equality before tax was disregarded and EDF enjoyed special treatment without any transparency whatsoever.

101. As Iberdrola argues, the assertion that it makes no difference whether a taxable amount is directly allocated to the capital of a public undertaking — in a wholly non-transparent way — or whether that amount funds the State budget is clearly misconceived. It is not at all a foregone conclusion that the capital injection to which the General Court assimilated the tax exemption in question would have actually come about if the French State had first collected the tax in order to contribute the amount to its budget and then, by means of the applicable procedures, sought to invest in EDF an amount corresponding to the amount of the tax collected.

102. In any event, the case‑law makes it clear that the categorisation of a measure as State aid does not depend on the measures which ‘could have’ otherwise been adopted. It depends on the objective characteristics of the measure which was, in actual fact, implemented by the Member State. (44)

103. In addition, in BNP Paribas and BNL v Commission (45) the General Court rightly held that ‘it is not for the Commission, in examining a scheme in the light of the rules on State aid, to envisage the subjective choices that might have been made by the beneficiaries of that scheme in the absence of such a scheme but to examine the scheme in order to determine whether it entails, from an objective standpoint, an economic advantage by reference to the tax provisions from which it derogates which would normally have been applicable in the absence of the scheme … The fact that, in the absence of the … scheme at issue, the undertakings concerned would allegedly not have disposed of their assets is, in the context of such an objective assessment, irrelevant’.

104. It should be noted that, in the present case, the General Court’s approach requires the Commission to undertake an ‘overall’ analysis, that is to say, an analysis covering the loss of tax revenue flowing from the tax exemption accorded to EDF, on the one hand, and the alleged further capital injection by the State-shareholder, on the other. Suffice it to say that such an approach presupposes that the revenue from the tax was directly hypothecated to the alleged capital injection. However, it is now clear that, had the tax been collected, it would have flowed to the general budget of the State, without any concrete hypothecation.

105. According to the case‑law, a joint assessment of a tax as a mode of financing a measure is only then called for when it is indissociable from the aid, which is the case when certain revenue is specifically ‘allocated’ to the use in question. Indeed, ‘for a tax to be regarded as forming an integral part of an aid measure, it must be hypothecated to the aid measure under the relevant national rules, in the sense that the revenue from the tax is necessarily allocated for the financing of the aid and has a direct impact on the amount of that aid’. (46)

106. Next, the question whether or not it was a rational move on the part of the French State to adopt the measure at issue here is simply irrelevant. That is not the purpose of the MEIP. Rather, it is clear from the case‑law that its purpose is to compare the situation of the public undertaking with that of the private undertaking and not to compare the costs for the State under two different operations.

107. In that connection, paragraph 262 of the judgment under appeal states that ‘like any creditor who is the owner of a company, the State can waive a claim by converting that claim into capital of an equivalent amount. That operation, whereby the owner of a company increases the company’s capital by waiving a claim which it holds against that company, constitutes a form of compensation which a prudent private investor is also capable of effecting under normal market conditions’. At the hearing, the Court of Justice therefore asked: supposing that a private undertaking which created accounting provisions in order to reimburse a claim held by its sole shareholder and the latter decides to relinquish its claim so as to increase the capital of the undertaking, does that operation give rise to taxation under Article 38(2) of the French General Tax Code? The Commission replied that it is common ground that the amount of the capital increase would never be the same, because the amount of the tax could never be recovered by a private operator (who would have to pay tax and would be unable to convert it into capital), whereas here the French State was able to re-inject that same amount.

108. It is thus apparent that, on one side of the comparison, the General Court considered the conduct that the French State ‘could have adopted’ — that is to say, the long plan (as opposed to the actual conduct as adopted by the State) — whereas, on the other side, it failed to define a reference private operator which could exist on the market.

109. As is now clear, however, it is settled law that a capital injection by the State or public authorities cannot be regarded as State aid where a private investor operating under normal market conditions would have made the investment. (47) In that connection, it is necessary to consider whether, in similar circumstances, a private investor of a size comparable to the public body in question would have provided capital of such an amount. (48)

110. And, as follows from the considerations discussed above, it is clear in the present case that a private undertaking would not have been able to convert the funds set aside under the accounting provisions into capital without first incurring corporation tax. As a consequence, the private investor would no longer have had that taxable amount available to him and would thus not have had at his disposal the amount supposedly used for the further capital injection, which the French State — exercising its prerogatives — was able to ‘invest’.

111. In that connection, it is relevant to refer here to the Court’s judgment in Chronopost I, (49) insofar as, in my view, what the Court essentially did there is what is called for in the present case. In Chronopost I, the Court considered that it was not appropriate to compare the situation of La Poste with that of the private investor.

112. In particular, it was also in that case that the General Court upheld the application of the MEIP, while subsequently the Court rejected it. This was necessary because La Poste was in a very particular situation: it had to maintain a network of post offices in order to provide Services of General Economic Interest, and a private investor — not operating in a reserved sector — would never find himself in the same position.

113. It accordingly follows from Chronopost I that, when a public operator is placed in a situation with which, by definition, a private operator could not be faced, there is nothing to compare. In my view, the issue in Chronopost I was to determine the amount of compensation for a public service and it was not a case which directly concerned the MEIP as such. However, that judgment is relevant in the present case and should be applied by analogy.

114. I therefore consider that the General Court misapplied that case‑law when, in paragraph 278 of the judgment under appeal, basing its findings on paragraph 38 of Chronopost I, it held that ‘the absence of a reference private investor does not preclude the need to examine the operation in the light of “normal market conditions”, which are necessarily hypothetical and must be assessed by reference to the objective and verifiable elements which are available’.

115. It is helpful first to read the Opinion of Advocate General Tizzano in Chronopost I, which was followed by the Court. In point 47, the Advocate General explains that ‘“under normal market conditions” a private undertaking that was not obliged to maintain a public postal network comparable to that commanded by La Poste in order to guarantee the provision of a universal postal service … would not have such a postal network and could not therefore provide one of its subsidiaries with logistical assistance of the kind at issue ... In asking the Commission to find out what charge would have been made for such assistance by a hypothetical private holding company … which was not required to provide a universal postal service and so did not enjoy the benefit of a reserved sector, the [General Court] erred in its interpretation of Article 87 EC because it took as a benchmark for the purpose of determining whether there was State aid a private operator which would not in fact exist “under normal market conditions”.’ He concludes in point 45 that ‘the [General Court] … required the Commission to apply a test that was clearly unrealistic and consequently unsuitable for the purposes of determining in a similar case whether there was State aid within the meaning of Article 87 EC’.

116. It is clear that the MEIP was not pertinent in Chronopost I. The Court agreed with that analysis and that is precisely why the Court held that the General Court was wrong and why the Court rejected the application of the MEIP in such a case. Indeed, the Court ruled in paragraph 33 of that judgment that the General Court’s ‘assessment [based on the application of the MEIP], which fails to take account of the fact that an undertaking such as La Poste is in a situation which is very different from that of a private undertaking acting under normal market conditions, is flawed in law’.

117. That is why, in a certain sense, the approach adopted by the Court in that case addresses the same concern as lies at the heart of the present appeal: the need to exclude from the scope of the MEIP situations where there is no real private operator with whom the conduct of the State could be compared.

118. That is why the test laid down in the judgment under appeal cannot be upheld insofar as it is based on an accumulation of fictions affecting the real behaviour of the State and the standard private investor. Both elements of the comparison — the conduct of the State, on the one hand, and the conduct of the private investor, on the other — are altered to the extent that what is at issue is no longer a comparison of the actual conduct of the State with that of a real private operator in the market economy.

119. It follows that the MEIP test as adumbrated by the General Court is rendered meaningless. Under the General Court’s version of the test, the ‘normal’ market conditions are no longer assessed on the basis of real elements, which would characterise a reference private investor corresponding to a standard investor on the market (which the General Court failed to define); nor are they assessed in the light of the means used to provide capital, but on the basis of purely hypothetical considerations. By contrast, the MEIP manifestly requires verification as to whether a private investor would have made the investment in question under the same conditions. As the Commission noted, different investments may imply different costs and different perspectives of profitability.

120. Lastly, according to paragraph 249 of the judgment under appeal, ‘it was for the Commission to determine whether a private investor would have invested a comparable amount in similar circumstances, irrespective of the form of the intervention by the State to increase EDF’s capital and the possible use of tax resources to that end, with a view to ascertaining the economic rationale for that investment and to comparing it to the actions such an investor would have taken with respect to the same undertaking in the same circumstances’.

121. Indeed, concurring with the analysis of the General Court, EDF argues essentially that all behaviour of the State should be assessed through the prism of profitability. (50)

122. However, as Iberdrola noted, if it was solely the economic rationale for the State’s investment which counted, public undertakings would be allowed to enjoy advantages on the basis of the status of their owner, because the State investor is in an intrinsically different situation from that of the private investor. On that analysis, public undertakings could ultimately be completely exempt from taxes.

123. In that connection, as the General Court rightly held in Westdeutsche Landesbank, (51) when applying the MEIP, it was not sufficient to compare the return which the Land obtained through the transaction at issue with the return which it obtained on the Wohnbauförderungsanstalt des Landes Nordrhein-Westfalen’s assets before the transaction, as those assets were not subject to the logic of a private investor. It was necessary to compare the return obtained by the Land under the transaction at issue with the return which a notional private investor would have required for that transaction. (52) The fact that the transaction at issue was reasonable for the Land did not preclude the application of EU law on State aid. It did not obviate the need to ascertain whether that transaction strengthened Westdeutsche Landesbank’s position by giving it an advantage which it would not have obtained under normal market conditions.

124. Indeed, if the General Court’s approach were to be followed, it would be possible for Member States to proceed on the basis of an overall calculation of economic rationale, in the context of which they would be allowed to make use of the prerogatives of public authority. For instance, in the present case, the French Republic would be able to exercise its fiscal powers for the benefit of EDF if — viewed as a whole — the operation could be construed as rational conduct for an operator acting with a view to profit.

125. However, it is this sort of overall calculation of rationale that was censured by the Court in Hytasa and Gröditzer. (53) As Iberdrola pointed out, those cases concerned the possibility for Germany and Spain to be allowed to take advantage of the rationale of the measures in question (essentially, capital increases), in view of the obligations incumbent on them as public authorities in the event of the liquidation of the recipient undertaking, obligations relating for instance to the payment of certain unemployment benefits. The Court censured such an overall calculation of rationale because it brought into play the acts of public authority. The Court emphasised the incompatibility between, on the one hand, the calculation of rationale which is specific to the MEIP and, on the other, the taking into account of advantages and charges resulting from the powers and obligations specific to the State.

126. In that connection, contrary to the statements made in paragraph 256 et seq. of the judgment under appeal and to the arguments put forward by EDF, the fact that Hytasa and Gröditzer concerned ‘obligations’ on the State alone, and not — as in the present case — the powers existing for its own benefit, does not affect the fact that the obligations — but also the rights — which the State may have as a public authority must never be considered in conjunction with those which the State has as an investor.

127. Accordingly, as is clear from paragraph 258 of the judgment under appeal, the General Court is able to commit such an error only by once again making the purported objective of the measure in question prevail over the insurmountable exclusivity of the rights necessary to implement it. However, neither in Hytasa nor in Gröditzer did the Court take into account the objectives of overall profitability claimed by the two Member States, insofar as the Court rejected their relevance in the logic of an investor: the nature of the measures was enough in itself to exclude them from the calculation of the rationale claimed by the Member States.

128. Lastly, as the Commission pointed out, the General Court’s approach is paradoxical notably because it is introduced in the context of liberalisation of the market concerned. Since the tax exemption represents a considerable advantage for the public undertaking in question, it is clear that such a measure tends to perpetuate its dominant position — in spite of the liberalisation of the market.

129. It follows from all the foregoing considerations that the second part of the second ground of appeal should be upheld.

3.      The third part of the second ground of appeal: the principle of equal treatment

130. Essentially the Commission and Iberdrola submit that the judgment under appeal fails to apply the principle of equal treatment as between State enterprises and private undertakings, thus making it possible for more favourable tax treatment to be accorded to the State, including those undertakings in which the State is not the only shareholder. EDF and the French Government contend that the third part of the second ground of appeal should be rejected.

131. I would recall that the MEIP is a reflection of the principle of equal treatment and, therefore, as the Commission noted, by allowing the State to grant tax exemptions to undertakings of which it is the 100% owner — as long as their profitability is sufficient — the General Court is, in fact, conceding those undertakings a tax favour.

132. That, in my view, amounts to a breach of the principle of equal treatment and cannot be reconciled with the objective of the MEIP. It is clear that the test as conceived by the General Court would inevitably engender serious distortion of competition between public and private undertakings and, accordingly, the fact that the instrument used must also be accessible to the private investor is not merely ornamental: that question goes to the very heart of the MEIP. As a consequence, I consider that the test as conceived by the General Court is incompatible with Article 295 EC, read in conjunction with Article 87 EC. In particular, application of that test would subvert the logic intrinsic to the principle of neutrality enshrined in Article 295 EC. (54)

133. Indeed, according to the General Court, since the nature of the claim is irrelevant, all financial advantage awarded by the State through the exercise of its public authority should be assessed in the light of that principle — an approach which would become a means of excluding from the concept of State aid measures which a private investor is unable to adopt.

134. As was pointed out by Iberdrola, the principle that the State’s activities as shareholder should be kept strictly separate from activities undertaken by the State as a public authority is essentially the same underlying idea as that which led the Court to consider that the fact that a dominant undertaking had acted at the same time both as a holder of special rights and as an economic operator created a risk of conflict of interest, in breach of Articles 82 EC and 86 EC. A system of undistorted competition, such as that provided for by the Treaty, can be guaranteed only if equality of opportunity is secured as between the various economic operators. (55) Similarly, the Court held that a State measure constitutes an infringement of Articles 86(1) EC and 82 EC where it gives rise to inequality of chances between economic operators and, in consequence, to distorted competition. (56) To my mind, therefore, the MEIP should not allow public enterprises to carry out operations in conditions which are more favourable than those applying to private enterprises or to carry out operations which private enterprises may never be able to undertake. Indeed, it should not be forgotten that, in the present case, the State could just as easily have adopted an attitude comparable to that of a private investor (see point 77 above).

135. As is evident from all the foregoing considerations, I consider that the MEIP should clearly not be applicable in the case of a tax exemption.

136. Lastly, there is reason to believe that, even though prima facie the General Court would appear to seek to limit the scope of its judgment to the hypothesis of an undertaking in which the State is sole shareholder, I do not believe it possible to discern in the logic underlying the judgment the reasons for that limitation. On the contrary, as the Commission noted, once there are private shareholders on the side of the State, the application of the MEIP becomes even more plausible, because then the conduct of the other shareholders may often serve as a valid reference point.

137. It follows that the third part of the second ground of appeal should be upheld.

4.      The fourth part of the second ground of appeal: the burden of proof

138. The Commission, Iberdrola and the Authority claim that the General Court disregarded the rules governing the apportioning of the burden of proof in relation to the applicability of the MEIP. EDF and the French Government contend that the fourth part of the second ground of appeal should be rejected.

139. First, I would point out that the General Court indicates in paragraphs 248, 249 and 250 of the judgment under appeal that it is the Commission who bears the burden of proving that the operation in question satisfied the MEIP. In fact, in paragraph 278, the General Court even reproaches the Commission for ‘fail[ing] to establish that there is no reference private investor with whom to compare the public investor’. While I agree with the EDF that the Commission is required to undertake a diligent and impartial examination of the file, I consider that that most certainly does not mean that the Commission should be required to carry out an assessment of a tax exemption in the light of the MEIP merely because the Member State ‘mentioned’ that principle 18 months into the investigation. This is manifestly not a question of the MEIP depending on the cooperation of the Member State.

140. It is clear that it is for the Commission to establish that financial assistance granted by the State objectively meets all the conditions necessary to be characterised as State aid. However, the fact remains that, once the Commission has done so, it is then incumbent on the State to show that it acted in the same way as a rational private investor. As the Authority pointed out, the application of the MEIP serves to rebut a finding that the assistance granted was State aid. Accordingly, it is for the State in question to adduce evidence in rebuttal.

141. Indeed, in Freistaat Thüringen v Commission, (57) the General Court rightly stated in essence that, once all the conditions for the existence of State aid are met, it is for the Member State to ‘put forward … evidence to show that the measure in question had been in keeping with the conduct of a private investor in a market economy’.

142. After all it is logical that the Member State which is relying on this argument must bear the burden of proving that the conditions for the MEIP are fulfilled. Also, it is obvious that the Member State and the recipient undertaking are the only parties who have at their disposal the economic and accounting data showing the quality and characteristics of a given investment.

143. In any event, this is a fortiori the case if the conduct which the Member State claims satisfies the MEIP appears at the outset to reflect the exercise of State power.

144. It is apparent from the documents before the Court that the French authorities failed to produce any proper and specific evidence during the administrative procedure to substantiate their claim that the French State had acted as a rational private investor. Indeed, the French authorities merely alleged that the capital injection was justified on the basis of prospects of profitability, without any supporting evidence. Contrary to EDF’s arguments, therefore, in view of the circumstances of the present case and given that, while the burden of proof was incumbent on the Member State, the French authorities failed to produce the slightest specific evidence, the Commission was entitled not to apply the MEIP, in particular in the circumstances of the present case, where the State has exercised its public authority.

145. It follows that the General Court erred in law insofar as it failed to comply with the rules on the allocation of the burden of proof.

146. In fact, as was pointed out by the Authority, the reversal of the burden of proof effected by the General Court places the Commission in an invidious position. As a rule, the Commission cannot know the precise circumstances of the State support given to an undertaking unless it is informed of those circumstances by the State in question. All the relevant arguments — and, in particular, the underlying facts and evidence — must be submitted by the State. If the Commission is not in possession of all the necessary facts, it must be able to adopt a decision in the light of the information which it does have in its possession.

147. In addition, I agree with the Commission that the absence of evidence to show that the tax exemption in question should have been regarded as an investment, in relation to which the State had appraised the prospects of profitable return before committing itself, as any private investor would have done, reveals another error of law.

148. Indeed, as the Court recalled in Commission v Scott, (58) ‘the lawfulness of a decision concerning State aid falls to be assessed by the [EU] judicature in the light of the information available to the Commission at the time when the decision was adopted’.

149. However, at the time of the adoption of the contested decision in the present case, the Commission had at its disposal no evidence to show that the MEIP was relevant or that that principle would have been satisfied. Thus, reproaching the Commission for not having assessed the tax exemption in the light of the MEIP is tantamount to criticising it for not taking account of information which the Commission did not have at its disposal.

150. Lastly, it should be pointed out that, during the administrative procedure, the French authorities provided no objective evidence to support their claim that the MEIP was applicable. They provided no report, expertise or internal study to show that there had been an assessment of profitability. Indeed, even when the General Court requested production of such a report, the report subsequently submitted to it was also carried out a posteriori: in other words, the French authorities had no report dating from the time of the alleged investment. It follows that the French State would appear to have invested EUR 888.89 million in an undertaking without undertaking any prior analysis or drawing up an ex ante business plan, and yet it still claims that its conduct should be regarded as comparable to that of a prudent private investor operating in a market economy.

151. It follows that the fourth part of the second ground of appeal should be upheld also and, as a consequence, that the second ground of appeal should be upheld in its entirety.

5.      The consequences of annulment

152. In accordance with the first paragraph of Article 61 of the Statute of the Court of Justice, the latter may, where the decision of the General Court has been set aside, either itself give final judgment in the matter, where the state of the proceedings so permits, or refer the case back to the General Court for judgment.

153. In the present case, the General Court did not assess all the pleas raised by EDF. It is appropriate, therefore, to refer the case back to the General Court for reconsideration and to reserve the costs of the appeal.

154. The fact remains, however, that the Court may itself give final judgment on an issue, where the state of the proceedings so permits. In my view, the Court now has at its disposal all the information necessary for it to rule on the merits of the third part of the second plea raised by EDF at first instance, by which EDF claimed essentially that the measures at issue should have been categorised as a capital injection and analysed in the overall context of clarifying the financial relationship between the State and EDF, thus claiming that the State acted like a prudent private investor in a market economy. As is apparent from all the foregoing considerations in the present Opinion, however, that third part of the second plea should be rejected.

V –  Conclusion

155. In the light of all of the foregoing, I propose that the Court should:

–        set aside the judgment of the General Court of the European Union (Third Chamber) of 15 December 2009 in Case T‑156/04 EDF v Commission, insofar as that judgment annulled Articles 3 and 4 of the Commission Decision on aid measures in favour of EDF and the electricity and gas industries (C 68/2002, N 504/2003 and C 25/2003) adopted on 16 December 2003;

–        refer the case back to the General Court for reconsideration;

–        reserve the costs of the appeal proceedings.

1 – Original language: English.

2 – Case T‑156/04 EDF v Commission [2009] ECR II‑4503 (‘the judgment under appeal’).

3 – Decision (C 68/2002, N 504/2003 and C 25/2003) adopted on 16 December 2003 (‘the contested decision’).

4 – Law of 10 November 1997 concerning urgent fiscal and financial measures (JORF, 11 November 1997, p. 16387).

5 – Case C‑136/92 P Commission v Brazzelli Lualdi and Others [1994] ECR I‑1981, paragraph 66.

6 – See, inter alia, Case C‑53/92 P Hilti v Commission [1994] ECR I‑667, paragraph 42.

7 – See Case C‑237/98 P Dorsch Consult v Council and Commission [2000] ECR I‑4549, paragraphs 35 and 36.

8 – See the Opinion of Advocate General Tesauro in Case C‑362/95 P BlackspurDIY and Others v Council and Commission [1997] ECR I‑4775, point 32.

9 – See also Hilti v Commission, cited in footnote 6, paragraph 42.

10 – See the Opinion of Advocate General Sharpston in Case C‑141/08 P Foshan Shunde Yongjian Housewares & Hardware v Council [2009] ECR I‑9147, point 72.

11 – Case C‑387/92 Banco Exterior de España [1994] ECR I‑877, paragraph 14. See also Case C‑6/97 Italy v Commission [1999] ECR I‑2981, paragraph 16.

12 – See report ‘Migaud’ carried out in the name of the Commission des finances, de l’économie générale et du plan of the National Assembly, attached to the Commission’s application, pp. 79 to 81, recounting the discussion of the amendment ‘De Courson’. In fact, the discussion on this amendment shows that the principal members of the committee (and, in particular, its President and the General Rapporteur) considered that the provision to be adopted should entail the payment by EDF of corporation tax on the amount of provisions transformed to capital.

13 – Cf. Commission communication to the Member States — Application of Articles 92 and 93 [EEC] and of Article 5 of Commission Directive 80/723/EEC to public undertakings in the manufacturing sector (OJ 1993 C 307, p. 3), paragraph 11.

14 – Joined Cases C‑71/09 P, C‑73/09 P and C‑76/09 P Comitato ‘Venezia vuole vivere’ v Commission [2011] ECR I‑4727, paragraph 94 and the case‑law cited.

15 – See Opinion of Advocate General Tizzano in Case C‑172/03 Heiser [2005] ECR I‑1627, point 45.

16 – Case C‑364/92 [1994] ECR I‑43.

17 – See the Opinion of Advocate General Léger delivered on 19 March 2002 in Case C‑280/00 Altmark Trans and Regierungspräsidium Magdeburg (‘Altmark’) [2003] ECR I‑7747, point 77.

18 – Case C‑487/06 British Aggregates v Commission [2008] ECR I‑10515, paragraph 111.

19 – See, inter alia, Case C‑83/98 P France v Ladbroke Racing and Commission [2000] ECR I‑3271, paragraph 25.

20 – Joined Cases C‑83/01 P, C‑93/01 P and C‑94/01 P Chronopost and Others v Ufexand Others (‘Chronopost I’) [2003] ECR I‑6993, paragraph 128.

21 – EDF refers to the Opinion of Advocate General Tizzano in Joined Cases C‑442/03 P and C‑471/03 P [2006] ECR I‑4845, point 87.

22 – Case T‑163/05 [2010] ECR II‑387.

23 – Case T‑196/04 [2008] ECR II‑3643.

24 – See, to that effect, Case T‑128/98 Aéroports de Paris v Commission [2000] ECR II‑3929, paragraphs 107 to 109, 121, 122 and 125.

25 – And by regulatory means in general. See inter alia Hancher, L., Ottervanger, T., and Slot, P.J., EC State Aids, Sweet & Maxwell, 2006, p. 74; Jaeger, Th., Beihilfen durch Steuern und parafiskalische Abgaben, NWV, 2006, paragraph 195; Mamut, M.-A., ‘Privatinvestorgrundsatz und Steuerbeihilfen’, in Jaeger (Ed.), Jahrbuch Beihilferecht 09, p. 341; and Haslehner, W., ‘Die Anwendbarkeit des Privatinvestorentests bei Steuerbeihilfen’, in Jaeger (Ed.), Jahrbuch Beihilferecht 2011, NWV, 2011, p. 273.

26 – Case C‑342/96 Spain v Commission [1999] ECR I‑2459, paragraph 46.

27 – Joined Cases C‑278/92 to C‑280/92 [1994] ECR I‑4103, paragraph 22. See also Joined Cases T‑129/95, T‑2/96 and T‑97/96 Neue Maxhütte Stahlwerke and Lech-Stahlwerke v Commission [1999] ECR II‑17, paragraph 119, and Commission Decision 2008/722/EC of 10 May 2007 on State aid C 2/06 (ex N 405/05) which Greece is planning to implement for the early voluntary retirement scheme of OTE (OJ 2008 L 243, p. 7), paragraphs 85 and 86.

28 – Case C‑334/99 [2003] ECR I‑1139, paragraphs 133 and 134.

29 – See Case C‑482/99 France v Commission (‘StardustMarine’) [2002] ECR I‑4397, paragraph 70.

30 – Opinion of 14 January 2003 in Altmark, cited in footnote 17, points 20 to 27.

31 – Cited in footnote 23, paragraph 85 (emphasis added).

32 – Joined Cases T‑228/99 and T‑233/99 [2003] ECR II‑435, (‘WestdeutscheLandesbank’).

33 – Case C‑207/01 Altair Chimica [2003] ECR I‑8875, paragraph 35.

34 – Case C‑309/99 Wouters and Others [2002] ECR I‑1577, paragraph 57.

35 – It refers to Case C‑342/96 Spain v Commission, cited in footnote 26, paragraph 46, and Case C‑256/97 DM Transport [1999] ECR I‑3913.

36 – See, to that effect, Case C‑342/96 Spain v Commission, cited in footnote 26, paragraph 46, and DMTransport, cited in footnote 35, paragraph 24. See also the Opinion of Advocate General Poiares Maduro in Case C‑276/02 Spain v Commission [2004] ECR I‑8091, points 24 and 25, and Case T‑152/99 HAMSA v Commission [2002] ECR II‑3049, where the capitalisation of the undertaking’s debt to the State was assessed under the MEIP, but the General Court pointed out that the correct comparator was the private creditor.

37 – See Commission Decision 2006/900/EC of 20 October 2005 on the State Aid implemented by Finland for investment aid to Componenta Corporation, OJ 2006 L 353, p. 36, paragraph 26 (one must not mix the roles of the city as a public authority and as commercial co-owner of the company), and Commission Decision C 56/2006 of 30 April 2008 on State aid C 56/06 (ex NN 77/06) implemented by Austria for the privatisation of Bank Burgenland, OJ 2008 L 239, p. 32: The Commission rejected an argument used by Austria, pointing out that liabilities under public law should not be mixed up with the analysis of the State’s acts in its capacity as an investor, but noted that the position would have been different if the State had given the same type of guarantee through a private law guarantee contract.

38 – Respectively, Case 234/84 Belgium v Commission [1986] ECR 2263, and Case 40/85 Belgium v Commission [1986] ECR 2321.

39 – Case 323/82 [1984] ECR 3809 (see 3842).

40 – See Khan, N., and Borchardt, K.-D., ‘The private market investor principle: reality check or distorting mirror?’, EC State aid law — Le droit des aides d’État dans le CE: liber amicorum Francisco Santaolalla Gadea, Kluwer, 2008, p. 115.

41 – Acta iure gestionis, insofar as the measures at issue sought to ‘normalise’ an undertaking’s accounts, in this case, EDF’s balance sheet (see point 6 above) and acta iure imperii, because the restructuring of EDF’s balance sheet was ‘decided and communicated to EDF’ by the French authorities on the basis of a law adopted for this purpose (see points 6 to 8 above).

42 – Emphasis added. See Case C‑91/94 Tranchant [1995] ECR I‑3911, paragraph 18 and the case‑law cited.

43 – Cited in footnote 17, paragraphs 89, 90 and 93.

44 – Case C‑148/04 Unicredito Italiano [2005] ECR I‑11137, paragraph 114.

45 – Case T‑335/08 [2010] ECR II‑3323, paragraph 169.

46 – Case C‑206/06 Essent Netwerk Noord and Others [2008] ECR I‑5497, paragraphs 89 and 90 and the case‑law cited.

47 – See, inter alia, Case C‑303/88 Italy v Commission (‘ENI-Lanerossi’) [1991] ECR I‑1433, paragraphs 20 to 24; Case C‑305/89 Italy v Commission (‘Alfa Romeo’) [1991] ECR I‑1603, paragraphs 19 to 20; and Joined Cases C‑328/99 and C‑399/00 Italy and SIM 2 Multimedia v Commission [2003] ECR I‑4035, paragraph 38.

48 – See Alfa Romeo, cited in footnote 47, paragraph 19; Case T‑234/95 DSG v Commission [2000] ECR II‑2603, paragraph 119; Case T‑296/97 Alitalia v Commission [2000] ECR II‑3871, paragraph 96; and Westdeutsche Landesbank, cited in footnote 32, paragraph 245.

49 – Cited in footnote 20.

50 – It refers to Case 234/84 Belgium v Commission, cited in footnote 38, paragraph 14, and Italy and SIM 2 Multimedia v Commission, cited in footnote 47, paragraph 38.

51 – Cited in footnote 32, paragraphs 313 to 315.

52 – See, to that effect, DM Transport, cited in footnote 35, paragraph 25.

53 – Cited, respectively, in footnote 27, paragraph 22, and footnote 28, paragraph 134.

54 – See Joined Cases 188/80 to 190/80 France and Others v Commission [1982] ECR 2545, paragraph 21, and Westdeutsche Landesbank, cited in footnote 32, paragraphs 192 to 194.

55 – See Case C‑49/07 MOTOE [2008] ECR I‑4863, paragraph 48 et seq. and paragraph 51.

56 – See Case C‑462/99 Connect Austria [2003] ECR I‑5197, paragraph 84.

57 – Case T‑318/00 [2005] ECR II‑4179, paragraph 180.

58 – Case C‑290/07 P [2010] ECR I‑7763, paragraph 91 and the case‑law cited.