OPINION OF ADVOCATE GENERAL
ĆAPETA
delivered on 7 April 2022(1)
Joined Cases C‑331/20 P and C‑343/20 P
Volotea, SA (C‑331/20 P)
easyJet Airline Company Ltd (C‑343/20 P)
v
European Commission
(Appeal – Action for annulment – State aid – Imputability – Indirect beneficiaries – Applicability and application of the market economy operator test)
I. Introduction
1. ‘L’inverno aveva rinfrescato anche il colore delle rocce.’ (‘Winter had refreshed even the colour of the rocks.’) (2)
2. The present cases arise many seasons after Grazia Deledda picturesquely described the arrival of spring on the island of Sardinia, Italy. (3) However, at their core, they also relate to the undervalued beauty of Sardinia during the winter months.
3. It was precisely to encourage tourism during that season, that the Regione Autonoma della Sardegna (Autonomous Region of Sardinia, Italy) (‘the Region’) set up a scheme to attract flights to airports located on the island of Sardinia (‘the aid scheme’).
4. The European Commission found that scheme to constitute incompatible aid in favour of several airlines (‘the contested decision’). (4) Among those airlines from which aid should be recovered were Volotea SA and easyJet Airline Company Ltd (collectively, ‘the appellants’). Those parties unsuccessfully challenged the Commission’s decision before the General Court. (5) Both companies appealed against the General Court’s judgments.
5. In the present cases, the Court of Justice is invited to assess a number of findings of the General Court relating to the determination of the Region’s scheme as State aid. However, the ‘elephant in the room’ is the applicability of the market economy operator (‘MEO’) test.
II. Facts and proceedings
A. Background to the dispute
6. The facts and the legal background of the present cases are described in detail in the judgments under appeal. (6) For the purposes of the present Opinion, they can be summarised as follows.
7. By legge regionale n. 10 – Misure per lo sviluppo del trasporto aereo (Regional Law No 10 – Measures for the development of air transport) (7) (‘Law 10/2010’), the Region made funding available to airlines with a view to opening up new routes and developing existing routes to and from Sardinia, including during the winter months.
8. The funding was transferred to the airlines through three Sardinian airports. Only two of those airports are of concern for the present cases: Cagliari-Elmas and Olbia airports. They are operated by companies that are either majority-owned by the chamber of commerce of Cagliari-Elmas or by private shareholders (collectively, ‘the airport operators’).
9. By measures implementing Law 10/2010, the Region defined activities for which funding could be received. Those of interest for the present appeals are the increase in air traffic by airlines (activity 1) and the promotion of Sardinia as a tourist destination by airlines (activity 2).
10. Funding from the Region’s budget was provided in the form of grants. Those grants were awarded to the airport operators, which, in turn, distributed the funds to various airlines. The Region released funding solely after assessment and approval of detailed ‘plans of activities’, which the operators had to submit to it. The Region also monitored their implementation.
11. The appellants are airlines that concluded agreements with the airport operators to carry out activities 1 and 2. Their respective plans of activities were approved by the Region, after which the appellants received compensation from the airport operators, under the scheme described above, for increasing the number of flights and routes to and from the two airports and for promoting Sardinia as a tourist destination.
12. On 30 November 2011, in accordance with Article 108(3) TFEU, the Italian Republic notified the scheme to the Commission. After the initiation of a formal investigation procedure, the Commission adopted the contested decision on 29 July 2016.
13. That decision states that the aid scheme in question does not involve State aid, within the meaning of Article 107(1) TFEU, to the airport operators (Article 1(1) of the contested decision); that it does constitute State aid in favour of the airlines involved, including the appellants, in so far as it relates to the operations of those airlines at Cagliari-Elmas and Olbia airports (Article 1(2) of that decision); that the aid in question was put into effect by Italy in breach of Article 108(3) TFEU and that it is incompatible with the internal market (Article 1(3) and (4) of the contested decision). It, finally, requires that that aid be recovered from the airlines (Article 2(1) of that decision).
B. The actions before the General Court
14. By means of applications lodged on 6 September 2017 and 11 January 2018, the appellants brought actions for annulment of the contested decision.
15. On 13 May 2020, the General Court delivered the judgments under appeal. It dismissed the appellants’ actions and ordered them to pay the costs of the proceedings.
C. Procedure before the Court of Justice
16. In its appeal before the Court of Justice, lodged on 22 July 2020, Volotea asks the Court to set aside points 1 and 2 of the operative part of the judgment in Case T‑607/17; annul Article 1 of the contested decision and the recovery order on Italy to the extent that it concerns Volotea; and, in the alternative, set aside points 1 and 2 of the operative part of the judgment in Case T‑607/17, refer the case back to the General Court; and order the Commission to pay the costs.
17. In its appeal before the Court, lodged on 23 July 2020, easyJet asks the Court to set aside the judgment in Case T‑8/18 and/or to annul Articles 1, 2, 3 and 4 of the contested decision to the extent that it concerns easyJet; and, in the alternative, to set aside the judgment under appeal, refer the case back to the General Court; and order the Commission to pay the costs.
18. For its part, the Commission contends that the Court should dismiss the appeals and order the appellants to pay the costs.
19. By decision of the President of the Third Chamber of the Court on 22 February 2021, the two cases were joined for the purposes of the Opinion and the judgment.
20. The parties replied to written questions put to them by the Court.
III. Assessment
21. While the Treaties do not contain a definition of what constitutes State aid, the Court has insisted that that concept requires an interpretation on the basis of its objective features. (8) Interpreted in that manner, Article 107(1) TFEU comprises four (main) cumulative conditions: that the measure in question be imputable to the State or financed through State resources; that it concerns an undertaking; to which it grants an advantage; and that that measure is selective. A measure satisfying those criteria is generally considered to be capable of distorting competition and affecting trade in the internal market. (9)
22. Due to the development in how the purpose of State aid control is understood, (10) and as a consequence of Member States’ inexhaustible imagination when it comes to intervening on the market, the Court is faced with the constant need to re-explain the criteria underlying Article 107(1) TFEU and to adjust them for their application to new situations. The present cases are no exception to that rule.
23. Indeed, the scheme at issue in the present cases seems to be an imaginative effort to bring more tourists to Sardinia, including during the winter months, with a view to boosting that region’s economic development.
24. As Sardinia is an island, tourists can obviously arrive by air. Airlines do business with airports. The scheme envisaged by the Region therefore involved airports developing plans and working together with airlines in order to bring more flights to the island. If those plans were approved by the Region, the latter would provide funding to the airport operators, which would distribute those funds to the airlines concerned. The scheme therefore involved an interplay of both public (the Region) and private (the airports and the airlines) actors.
25. For the purposes of the concept of State aid, the design of that scheme raises several questions. For instance, does the channelling of funds through a private intermediary rid those funds of their public provenance? Which of the private parties involved, if not both, must be deemed the beneficiary of that scheme? Finally, does the said scheme provide an advantage which would not have been available on the market?
26. The Court has asked that I focus my analysis on providing answers to those questions. I will, therefore, restrict my opinion to the concepts of imputability (IV); secondary beneficiaries of aid (V); and the methodology for determining or ruling out the existence of an advantage (VI).
IV. Imputability
27. Financing by the State or through State resources is a constitutive element of the concept of State aid. (11) That means that the resources used in a measure must be imputable to the State. (12) In their respective appeals, the appellants challenge precisely that determination: they argue that the funds received from the airport operators could not be imputed to the Region. (13)
28. Their arguments are essentially three-fold. (14) First, they argue that, in the judgments under appeal, the General Court failed to assess appropriately the criteria arising from the judgment in France v Commission. In particular, it is argued that the General Court’s reliance on the existence of a supervisory mechanism to monitor the distribution and use of the funds at issue was insufficient to establish imputability. Second, the appellants allege that the General Court misapplied the judgment in Belgium and Magnetrol International v Commission, (15) in that it did not assess whether the airport operators had a margin of discretion to define essential elements of the aid scheme in question. Third, they argue that the presence of State resources may only be confirmed where the State exercises control over the funds as well as the administrators of those funds (in this case, the airport operators). Such a requirement would follow from the judgment of the Court in Germany v Commission. (16)
29. I disagree with all three arguments.
30. As to the first argument, in the judgment in France v Commission, the Court laid down a number of indicators which collectively may show that measures taken by certain undertakings are actually imputable to the State. (17) Through that assessment, the Commission can then establish whether the circumstances of a case point towards the presence of State direction. (18) That is to say, it may conclude that the State was involved in the adoption of a measure, or at least that it is unlikely that it was not involved in influencing its adoption. (19)
31. However, the judgment in France v Commission concerned public undertakings. When faced with the question of the evidentiary threshold in relation to private undertakings, the Court in the judgment in Commission v Italy and Others explained that the Commission must furnish more than circumstantial evidence: what is required is that it establishes a complete chain of events and thus show decisive proof of imputability. (20)
32. Given the date of delivery of the judgment in Commission v Italy and Others, it is unfortunate that the parties to the present appeals were not able to take a position on whether that move away from the aphorism ‘circumstances cannot lie’ should have brought about changes in the General Court’s analysis.
33. However, I read the judgments under appeal as indicating that the General Court appeared to be satisfied that the Commission’s investigation revealed decisive proof of imputability.
34. Thus, in paragraphs 70 and 94 of the respective judgments under appeal, the General Court observed that ‘it is clear … from its implementation in practice that the funds transferred by that region to the airport operators were those used by the airport operators to remunerate co-contracting companies.’ Similarly, in paragraphs 86 and 118 thereof, the General Court explained that ‘the Commission found, in the contested decision, that the funds made available to the airport operators by the Autonomous Region had to be and were actually used in accordance with the instructions prescribed by that region, in this case as remuneration for services provided by the airlines.’ (21) Finally, in several paragraphs of the judgments under appeal, the General Court recalled that the airport operators themselves admitted to acting under the control and direction of the Region, and that they only transferred the funds at issue with the expectation that those sums would be repaid by that public authority. (22)
35. It follows, to my mind, that the General Court found that the evidence available to it showed, to the requisite standard of proof, evidence of the factors set out by the Court of Justice in the judgment in France v Commission, as applied to the case of private undertakings in the judgment in Commission v Italy and Others.
36. As to the second argument, I am equally unconvinced that the General Court misapplied the judgment in Belgium and Magnetrol International v Commission. (23)
37. That case concerned, inter alia, the question of whether systematic administrative practices to approve requests for excess profit exemption by multinational companies were capable of constituting a relevant factor in the assessment of the de facto existence of an aid scheme. (24)
38. First, as the Commission rightly points out, the appeals are silent on how that judgment would be relevant in the context of the present cases. In the scheme at issue in the present cases, we are not faced with a systemic and unwritten administrative practice. Quite the opposite, in fact, the General Court found that there were express objectives and criteria laid down by the Region in Law 10/2010, which the operators were implementing. (25)
39. Second, even if that judgment were relevant, it should be observed that, in the judgments under appeal, the General Court did assess the remit of discretion of the airport operators to define essential elements of the aid scheme at issue. (26) It found that there was no indication that the airport operators could go beyond the margins set by Law 10/2010. Indeed, according to the General Court, the operators’ discretion was limited by the Region to the extent that the latter was able to control both the amount of the aid to the airport operators as well as the redirection of those sums (in full) to the appellants. (27)
40. Turning finally to the third argument, it does not, in my view, follow from the judgment in Germany v Commission that State resources can only be deemed present when the administrator of State funds is subject to constant State control. What that case confirms is that the State ensures a power of disposal over the funds as well as their administration. (28) That requirement is far from novel. (29)
41. The converse would, to my mind, permit the rules on State aid to be circumvented through the creation of institutions charged with administering aid, but which, as in the case of the airport operators, are not themselves controlled by the State. (30)
42. Indeed, it is for precisely that reason that the Court has consistently dismissed the need to distinguish between whether aid is granted directly by the State or by public or private bodies that have been established or appointed to implement the State’s scheme. (31)
43. In view of the above, I propose that the Court should dismiss the appellants’ challenge to the finding of imputability as unfounded. (32)
V. Beneficiaries
44. In the contested decision, the Commission found that a number of airlines, including the appellants, were beneficiaries of State aid (Article 1(2) thereof). At the same time, the Commission found that the airport operators did not receive such aid and thus could not be deemed beneficiaries of the aid scheme (Article 1(1) of the contested decision).
45. The appellants contest the latter finding. They argue essentially that the airport operators received an advantage under that scheme.
46. Even though the challenge to Article 1(1) of the decision is not the object of these appeals, those arguments are also relevant in relation to the challenge to Article 1(2) thereof, and possibly also as regards the effects arising from Article 2 of the contested decision, given that the resulting finding could have an impact on the amount of the aid to be repaid by the airlines. (33)
47. I will, therefore, address the two arguments on the basis of which the General Court concluded that the airport operators are not (secondary) beneficiaries under the aid scheme.
48. The first (and main) argument relied on by the General Court is that the airport operators passed on to the airlines the entirety of the funds received from the Region.
49. Is that the relevant argument? To my mind, it is not. (34)
50. An advantage can be conferred on undertakings other than those to which State resources are directly transferred (that is what is also referred to as an ‘indirect advantage’). (35) Such an advantage must be distinguished from mere secondary economic effects that arise and are inherent in most State aid measures, and which are exempt from State aid review. (36)
51. That raises the question of how to distinguish between economic effects that arise incidentally, as a side-effect of an aid scheme, from ‘actual’ indirect advantages geared towards certain ‘hidden’ beneficiaries.
52. In its Notice on the notion of State aid, the Commission explains that ‘the foreseeable effects of the measure should be examined from an ex ante point of view’. (37) Thus, an indirect advantage would be present only ‘if the measure is designed in such a way as to channel its secondary effects towards identifiable undertakings or groups of undertakings’. (38)
53. Accordingly, it seems to me, that in order to verify that an indirect advantage is not ‘dressed up’ as the ancillary effect of a scheme, in the present cases, the Commission was under an obligation to examine whether the design of the scheme envisaged ex ante certain advantages for the airport operators that were more than inherent features of the aid scheme in question. (39) After all, the airport operators had an active role to play in that design and so had an economic interest in that scheme operating at a certain level of capacity (and that is, despite their plans having to be approved by the Region and despite the fact that the latter oversaw the implementation of those plans).
54. Indeed, I can envision a situation whereby, for instance, in the design of the flight schedule for a particular airport, the addition of one more flight than what would be necessary to cover its operating costs might have secured an advantage to that airport (and thus its operator). (40)
55. It is precisely for that reason, that I would argue that it is only logical that, where, as in the present cases, a party makes a sufficiently substantiated claim that there exists, in fact, a ‘hidden’ or ‘additional’ beneficiary, the General Court is required to assess, on the basis of all of the circumstances available to it, whether the Commission carried out a sufficiently complete assessment to identify also the presence of any indirect beneficiaries under that aid scheme.
56. As the General Court did not do so in the present cases, I take the position that it erred in finding that the Commission was right to hold as irrelevant the ex ante analysis of the economic profitability of the service agreements concluded between the airport operators and the airlines, as accepted by the Region. (41)
57. The second argument relied on by the General Court is that the airport operators would not have entered into the agreements with the airlines had it not been for the Region’s funding under the scheme. (42)
58. I do not see how that fact (if that is indeed a finding of fact) contributes to concluding in itself that participation in the aid scheme at issue did not create an advantage also for the airport operators.
59. If anything, it points to the contrary conclusion: that the airport operators opted to participate in the scheme solely because they were the beneficiaries of an advantage that would not have accrued to them under normal market conditions.
60. However, that element of advantage is precisely what was not investigated further. To my mind, a more thorough analysis as to the benefits potentially received by the airport operators would have been necessary, in order to remove all doubt.
61. On that basis, I propose that the Court of Justice find that the General Court erred when it deemed the ex ante analysis of the economic profitability of the service agreements between the airport operators and the airlines irrelevant for determining the beneficiaries under the aid scheme at issue.
VI. Advantage
62. In their appeals, the appellants take issue with the fact that the General Court endorsed the Commission’s finding that the MEO test was not applicable to the aid scheme at issue in order to exclude the existence of an advantage. (43)
63. The MEO test is an analytical tool used by the Commission to exclude the existence of an advantage to an undertaking obtained by virtue of the State’s intervention on a given market. That test requires a complex economic analysis to assess whether a hypothetical private operator on a given market would engage in a comparable intervention and on the same terms as the State.
64. As rightly referred to by the appellants, the Court has held that the Commission has an obligation to conduct the MEO test in all situations in which it is applicable. (44)
65. It is precisely that question of the applicability of the MEO test which has, so far, not been sufficiently illuminated by the case-law. In particular, it remains unclear in which situations the intervention of the State in the market is reasonably comparable to the potential behaviour of a market operator.
66. In order to offer some ideas about how to resolve that question, it is useful to reflect on the purpose for which the MEO test was developed and how that purpose evolved with the case-law.
A. The purpose and evolution of the MEO test
67. Introduced in parallel to the 1980 Transparency Directive, (45) the MEO test was initially intended to assess the behaviour of public undertakings (that is to say, those in which the State had ownership) when competing on the same market as private undertakings. (46)
68. That is also where the idea of the State acting as a ‘shareholder’ (which, in the language of the time, was known as the ‘acquisition of public authorities’ holdings in company capital’) originates from. (47).
69. The rationale of the MEO test was thus rooted in the need to find a way to enable the State to be a market participant without undermining the effectiveness of the State aid discipline. (48) More specifically, that test served to exclude from the concept of State aid those actions of the State whereby the latter ‘descended onto the market’ to compete with other undertakings on like terms. Such like terms were deemed present where the State’s investment in its own undertaking, or by its own undertakings, was of a type that could be contemplated by a private investor. (49)
70. It is precisely in the light of that rationale that the MEO test is frequently described in relation to the possibility for Member States to opt for a mixed economy, understood as one in which the State is also present on the market alongside private undertakings. (50) Hence, the often-repeated reference to the property ownership neutrality of Article 345 TFEU and the principle of equal treatment linked thereto. (51)
71. However, the more recent case-law of the Court appears to indicate that the MEO test should also find application outside those traditional confines of State ownership and shareholding. (52)
72. That gives rise to two ways in which the applicability of the MEO test could be conceived.
73. On the one hand, when viewed against its historical application, the applicability of the MEO test could be restricted solely to those situations in which the State acts as a shareholder on a particular market. One the other hand, given that the State can also intervene on the market through means other than ownership or shareholding, confining the MEO test to those situations appears too narrow an interpretation of the usefulness of that test.
74. Indeed, to my mind, the MEO test can prove useful, and therefore applicable, in all those situations in which a State’s intervention on the market can be assimilated to the action of a private market participant. That assessment of ‘equivalency’ goes beyond the traditional confines of shareholding and ownership. That is, if a private market operator can be seen to act in the same way as the State, the MEO test can reveal whether the State’s action was market-conform. That would then, in turn, assist in ruling out the existence of an advantage (and thereby a benefit constituting State aid).
B. The dichotomy of ‘public authority’ versus ‘economic actor’
75. Seemingly detached from its original rationale, the applicability of the MEO test appears to have evolved to be governed by a confusing (and, quite frankly, misleading) dichotomy.
76. Consider, for instance, the judgment in Commission v EDF. (53) There, the Court held that ‘the applicability of the private investor test ultimately depends … on the Member State concerned having conferred, in its capacity as shareholder, and not in its capacity as public authority, an economic advantage on an undertaking belonging to it’. (54) In the same vein, in Ryanair v Commission, the General Court noted that ‘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 private investor principle, application of that principle must be excluded in the event that the State acts as a public authority.’ (55) Finally, similar wording is also reproduced in the Commission’s Notice on the notion of State aid. (56)
77. Judging by those examples and the language used therein, the question of the applicability of the MEO test is linked to the State acting as an economic operator (by means of shareholding). Those activities would have to be distinguished from the activities of the State acting as a public authority. In the former scenario, according to the logic of the case-law, the MEO test applies. In the latter scenario, that test would not apply.
78. I do not think that that distinction holds water. (57)
79. There are two possible understandings of the expression of the ‘State acting as a public authority’. For one, that expression might denote a State which acts in the public interest (for example, to enhance economic development, save jobs, or protect the environment). For another, that expression can be understood as the State acting through instruments of public policy (that is to say, legislation, regulation or, for that matter, taxes). However, neither interpretation really supports the conclusions reached in the case-law (or by the Commission) as to why the MEO test was or was not applicable.
80. As to the first understanding of the concept of a ‘State acting as a public authority’, the case-law considers that the reasons which motivate a State to intervene on the market are irrelevant. (58) Indeed, whether a State acts in order to preserve employment, to protect the environment, to enhance tourism or simply to manage its own company is wholly unrelated to the question of whether the beneficiary undertaking acquired a benefit arising from that intervention which could have been provided by the market under the same conditions. The applicability of the MEO test cannot, therefore, be excluded for those reasons alone.
81. As to the second understanding, it must be observed that the MEO test was applied in various instances in which the State acted through the use of instruments of public power; that is to say, instruments which are not available to a private market operator. For instance, in a number of cases, the State intervened on the market to defer or forgive the payment of tax debt and other public financial obligations owed to the State by private undertakings. (59) Similarly, in Commission v EDF, (60) the State intervened through fiscal and regulatory powers, and in Ryanair v Commission (61) the intervention occurred through legislative and regulatory powers. In all those cases, however, some form of the MEO test was applied. That means that its applicability cannot be excluded solely because the State is acting through instruments of public power.
82. In untying that dichotomy, it certainly does not help that the case-law sometimes draws on the distinction between the State acting as an ‘economic actor’ and the State acting as a ‘public authority’ not in order to decide on the applicability of the MEO test, but rather in the course of its application. (62) That is so particularly in situations when the Court wishes to stress that the costs and benefits of the State as a public authority, relating, for instance, to the rate of employment or increase in budgetary revenues, are not to be taken into account when applying that test. Conceptually, that distinction is logical, as a hypothetical market operator, who, after all, serves as comparator, would not take such concerns into consideration when deciding on its market moves. (63) However, those considerations only enter the picture once the applicability of the MEO test is confirmed and not at any point prior to that.
83. In short, I take the position that it is incorrect to draw the dividing line of applicability of the MEO test on the basis of the dichotomy between ‘public authority’ and ‘economic operator’. The MEO test seeks to determine or discard the existence of an advantage created for an undertaking by reference to the market-economic conformity of certain behaviour by the State. As such, it is realistically applicable in all situations in which the State performs activities comparable to ‘market activities’, including when ‘the State as a public authority’ acts in a way comparable to a market operator. That implies that the applicability of the MEO test should be decided by an objective enquiry into how the State engages with undertakings on a given market, irrespective of the form and reasons for that engagement.
C. How to decide whether the MEO test is applicable?
84. On the basis of the foregoing, I am of the opinion that the applicability of the MEO test is governed by the nature of the activities of the State. That is to say that the applicability of the MEO test presupposes that the State intervenes through an activity that is comparable to what could be envisaged on the market.
85. But how does one assess whether an equivalent activity is possible on the market?
86. To my mind, the Commission must place itself in the position of the beneficiary undertaking and assess, from its point of view, whether the benefit received (for example, the grant of a credit guarantee, capital injection, or debt forgiveness) could also have been offered by other undertakings on that market. If the answer to that question is in the affirmative, the MEO test is applicable.
87. A few examples might clarify my proposition. On the market, an undertaking can receive a loan deferral from a creditor. That activity can also be performed by the State: it could, for example, approve a tax deferral to an undertaking. As such, from the point of view of the recipient undertaking, the State is granting a benefit of the same nature as a (private) creditor. As State action of that kind, consequently, finds equivalency on the market, notwithstanding the fact that the State may have acted as a public authority, it must be possible to assess its intervention by means of the MEO test. (64)
88. To take a second example: if the State purchases marketing services on the market, this benefits the undertaking from which it purchases those services. The acquisition of such services is not an activity that only a State can perform. It is a simple purchasing transaction. In other words, given that the State performs an activity comparable to that of any other private undertaking, the MEO test should be applicable to verify the market-conform nature of that (service) acquisition.
89. Consider, lastly, a third example: if a State awards an otherwise fee-incurring operating licence to an undertaking free of charge, the State intervenes on a market through the performance of an activity that it alone holds as gatekeeper to that market. A private undertaking could never replicate that activity. Therefore, in the assessment of whether that free operating licence constitutes State aid, there is no place for the MEO test. The yardstick of comparison could never be a private economic operator.
90. The point I wish to make with those examples is the following: the MEO test is applicable whenever a State performs an activity that could be replicated on the market by private operators. Obviously, that can occur not only when the State acts as a shareholder, but also when it acts as a public authority (whether it pursues economic, social, environmental or other policy aims). The determinative question is whether the type of economic activity in question is of a nature that could be envisaged on the market. If the answer to that question is in the affirmative, the MEO test is applicable.
91. I would like to point out that the applicability of the MEO test does not, in any way, prejudice the result of its application. Whether or not an activity could be replicated on the market (so that the MEO test is applicable) is not decisive for the question of whether a particular benefit constitutes an advantage. That latter question depends on the application of the MEO test, which is based on an economic analysis examining whether a (hypothetical) market operator would grant the same benefit under similar conditions as those granted by the State. (65)
D. Application to the present cases
92. In the judgments under appeal, the General Court upheld the Commission’s dismissal of the applicability of the MEO test on the basis of two arguments. First, it considered that the Region did not act as a shareholder, as it was not the owner of the airport operators involved. (66) Second, it relied on the fact that, in adopting the aid scheme, the Region acted ‘in the framework of a general economic policy’, therefore ‘exclusively as a public authority’. (67)
93. In the light of the foregoing explanations, I am of the opinion that the arguments put forward by the General Court are insufficient to exclude the applicability of the MEO test.
94. As I have explained in points 71, 74 and 83 of this Opinion, the more recent case-law of the Court as well as the entire rationale underlying the MEO test do not limit its applicability solely to situations where a State becomes part of the market by means of ownership in an undertaking. The fact that the Region does not own the airport operators through which it has implemented its scheme is, therefore, in itself insufficient to exclude the application of the MEO test.
95. Equally, it is not decisive for an assessment relating to the applicability of the MEO test whether the Region acted ‘in the framework of a general economic policy’ or as a ‘public authority’. As I have explained in point 80 of this Opinion, the intention behind the State’s intervention is simply not a factor to take into account at this stage of the analysis.
96. Indeed, as I have explained in the previous section of this Opinion, when reviewing whether the Commission appropriately excluded the applicability of the MEO test, the General Court should have verified whether that institution assessed, from the point of view of the appellants, whether the funds received from the Region were granted by means of an activity which could have been replicated on the market.
97. I consider, therefore, that the General Court erred in the way it has justified the non-applicability of the MEO test in the present cases.
98. Next, after concluding that the MEO test was not applicable, the General Court assessed the existence of an advantage to the appellants by having regard to the activities of the Region from the perspective of an acquirer of services.
99. Even though I am doubtful that the award of grants for the creation of new flights may constitute an acquisition of services, I agree that, in principle, the ‘service acquirer’ test could indeed act as a gateway to determining the existence or exclusion of an advantage. However, given that that approach presupposes that the State acted as an economic operator in order to purchase services on a market, that assessment constitutes, by its nature, nothing more than one of the iterations of the MEO test. That is, the very test which the General Court rejected as inapplicable because it considered that the Region had not acted as an economic operator.
100. Be that as it may, there are two points relating to the General Court’s alternative assessment (relying on the ‘service acquirer’ test) that deserve further analysis. The first point relates to the appellants’ argument that the General Court exceeded its jurisdiction when upholding the contested decision on grounds of alternative reasoning. The second point relates to the steps of the analysis within the service purchaser test, as proposed by the General Court.
101. First, as to the competence of the General Court, it is true that, by virtue of the administrative nature of the State aid procedure, the General Court enjoys only limited jurisdiction. As such, it may not substitute its own economic assessment for that of the Commission. (68)
102. In essence, that requires the General Court to assess whether the statement of reasons in the contested decision contains all the relevant information to allow it to substantiate the conclusions drawn from it. (69) That is particularly so, given that, in the realm of State aid, the Commission closes its investigative procedure by way of a decision. (70)
103. However, the Court of Justice has clarified that, except where there is no material factor to justify that course of action, the General Court may be led, in proceedings for annulment, to interpret the reasoning of the contested measure in a manner which differs from that of its author, and even, in certain circumstances, to reject the latter’s formal statement of reasons. (71)
104. It is precisely that leeway which the General Court made use of in the present cases. Thus, in the judgments under appeal, that court complemented the (admittedly short and, by itself, insufficient) reasoning already contained in recital 386 of the contested decision (72) by arguments exchanged between the parties as well as information obtained during the written and oral procedures before it. (73)
105. Such a situation does not, therefore, fall within the category of substituted reasoning in the context of the General Court’s review jurisdiction since the General Court relied on the existing reasoning in the decision, without changing its content.
106. Second, as to the constituent elements of the ‘service acquirer’ test, the General Court considered that the existence of an advantage could only be excluded if, first, the acquisition of the services in question occurred by means of an open, transparent and non-discriminatory tender procedure and, second, the Region had an actual need for the acquired services.
107. As to the requirement for an open and competitive selection procedure, I can be brief. The economic rationality of purchases of goods and services can indeed be demonstrated by the very fact that they were made through such a procedure. That is why such purchases by public authorities for their own operating purposes are, as a rule, subject to the public procurement directives. (74) However, I am not convinced that, in a situation such as that in the present cases, in which the Region acted through the intermediary of private market operators, insisting on a procedure similar to that required by the public procurement directives would be the only way to ensure that the services in question are priced at market terms.
108. As far as the genuine need requirement is concerned, the apparent logic underlying that assessment seems to be the following: even if purchases by a public authority were made at market terms, there is the additional requirement that those purchases are proportionate to the ‘genuine needs’ of the authority in question. (75)
109. It is true that an economically rational market operator is presumed to acquire only those services (or goods) that it actually requires for the performance of its activities (or to achieve the aim it pursues). (76) The General Court considered that the Region could have a genuine need to purchase the services in question (related to flights and marketing) in order to promote the island of Sardinia as a tourist destination.
110. However, how is one to quantify that need?
111. In the present cases, the Region did not ‘acquire’ flights and marketing services to fulfil its own needs (as, for example, when a State buys office chairs for government offices). Accordingly, there was no internal benchmark against which the number of acquired services could be assessed. What the Region sought by purchasing those services (if indeed both of them can be considered services) was to create, at least in part, new supply (through flights) and new demand (through marketing services) for the Sardinian market. For that type of activity, a market operator would have conducted a cost-benefit analysis taking account of the funds required and the expected return. It is not clear from the contested decision whether the Region carried out such an assessment or whether it operated according to set expectations relating to, for instance, the number of tourists the scheme was intended to attract. Nor does it appear from the contested decision that the Commission tried to investigate whether such an analysis had been carried out.
112. In such circumstances, I am at a loss as to the basis on which the General Court could legitimately claim that the acquisition of the marketing services in question was in excess of ‘genuine needs’. (77) Consequently, I consider that the General Court also erred in the application of the elements constituting its ‘service acquirer’ test.
113. I therefore propose that the Court of Justice should uphold the appellants’ arguments that the General Court erred in the way in which it conducted the ‘service acquirer’ test.
VII. Conclusion
114. On the basis of the foregoing conclusions, I propose that the Court of Justice should uphold the appellants’ arguments relating to the way in which the General Court:
– confirmed the European Commission’s finding that the ex ante analysis of the economic profitability of the service agreements concluded between the airport operators and the airlines was irrelevant for determining the beneficiaries under the aid scheme at issue;
– justified the non-applicability of the market economy operator test; and
– conducted the ‘service acquirer’ test.