1.

The present Opinion is delivered in Joined Cases C‑790/21 P and C‑791/21 P. It should be read together with my three other Opinions in parallel appeals, ( 2 ) also delivered today, all of which concern the same State aid scheme. By its appeal in Case C‑790/21 P, Covestro Deutschland AG (‘Covestro’), a materials-manufacturing company, seeks to have the judgment of 6 October 2021, Covestro Deutschland v Commission (T‑745/18, EU:T:2021:644) (‘the judgment under appeal’) set aside. That judgment dismissed its action seeking annulment of European Commission Decision of 28 May 2018 on the aid scheme implemented by Germany for baseload consumers under Paragraph 19 StromNEV, ( 3 ) in respect of the years 2012 and 2013 (‘the decision at issue’). By its appeal in case C‑791/21 P, the Federal Republic of Germany (for the sake of simplicity, the Federal Republic of Germany will be referred to as ‘Germany’) seeks to have the judgment under appeal set aside. The Commission has brought a cross-appeal in both of the above cases, whereby it too seeks to have the judgment under appeal set aside.

2.

The background to the dispute is set out in paragraphs 1 to 22 of the judgment under appeal. For the purposes of the present Opinion, it may be summarised as follows.

3.

Paragraph 21 of the Energiewirtschaftsgesetz (Law on the protection of the energy supply), as amended by the Gesetz zur Neuregelung energiewirtschaftlicher Vorschriften (Law on the revision of energy industry regulations) of 26 July 2011, ( 4 ) but prior to the amendments made by the Gesetz zur Weiterentwicklung des Strommarktes (Law on the development of the electricity market) of 26 July 2016, ( 5 ) provides, inter alia, that network charges must be reasonable, non-discriminatory, transparent and based on the costs of an efficient operation of the energy supply system.

4.

Paragraph 17 of the Stromnetzentgeltverordnung (Federal Ordinance on Electricity Network Charges) of 25 July 2005, ( 6 ) defines the calculation method to be used by system operators in order to determine the general charges, which is based, inter alia, on the annual total network costs.

5.

Paragraph 19 of StromNEV 2005 also provides for individual charges for categories of users whose consumption and load profiles are very different from those of other users (so-called ‘atypical users’), which take into account, in accordance with the principle of cost-reflective pricing, the contribution of those users to reducing or preventing an increase in network costs.

6.

In that regard, Paragraph 19(2) of StromNEV 2005 introduces individual charges for the following two categories of atypical users: (i) users whose peak cost contribution may differ significantly from the simultaneous annual peak load of all other users connected to the same network level, that is to say, users who systematically consume electricity outside peak hours (‘anticyclic consumers’); and (ii) users whose annual electricity consumption represents at least 7000 hours of use and more than 10 gigawatts/hour (‘baseload consumers’).

7.

Until the entry into force of StromNEV, as amended by the EnWG 2011 (‘StromNEV 2011’), anti-cyclical and baseload consumers were subject to individual fees calculated according to the ‘physical-route method’ drawn up by the Bundesnetzagentur (Federal Network Agency, Germany; ‘BNetzA’). That method took account of the network costs generated by those consumers, with a minimum charge equivalent to 20% of the published general charges (‘the minimum charge’). The latter guaranteed a fee for the operation of the network to which those consumers were connected if the individual charges calculated according to the physical-route method were lower than or close to zero.

8.

Pursuant to the second and third sentences of Paragraph 19(2) of StromNEV 2011, which entered into force on 4 August 2011, with effect from 1 January 2011 (the date of the retroactive application of that provision), the system of individual network charges for baseload consumers was abolished and replaced by a full exemption from the obligation to pay network charges (‘the exemption at issue’). It was to be granted only once the competent regulatory authority (either the BNetzA or the regulatory authority of the Land concerned) had verified that the legal conditions were met. That exemption was borne by transmission or distribution system operators according to the network level to which the beneficiaries were connected. Individual network charges for non-peak consumers remained in place as well as their obligation to pay at least 20% of the published network charge.

9.

In accordance with the sixth and seventh sentences of Paragraph 19(2) of StromNEV 2011, transmission system operators were obliged to reimburse downstream operators of electricity distribution networks (the distribution system operators) for lost revenues resulting from the exemption at issue and had to compensate each other for the costs entailed by the exemption, by means of financial compensation in accordance with Paragraph 9 of the Kraft-Wärme-Kopplungsgesetz (Law on the promotion of cogeneration of heat and electricity) of 19 March 2002, ( 7 ) with the result that each of them bore the same financial burden calculated according to the quantity of electricity which each supplied to the final consumers connected to its network.

10.

From 2012 onwards, the decision of the BNetzA of 14 December 2011 (BK8-11-024; ‘the BNetzA decision of 2011’) established a financing mechanism. Under that mechanism, the distribution system operators collected from final consumers a surcharge (‘the surcharge at issue’), the amount of which was passed on to the transmission system operators to compensate them for the loss of revenue brought about by the exemption at issue.

11.

The amount of the surcharge was determined each year, in advance, by the transmission system operators, on the basis of a methodology established by the BNetzA. The amount for 2012, the first year in which the system was implemented, was set directly by the BNetzA.

12.

Those provisions did not apply to the costs of the exemption for 2011 and, therefore, each transmission and distribution system operator had to bear the losses relating to the exemption for that year.

13.

During the administrative procedure before the Commission, which resulted in the decision at issue, the exemption at issue was first declared null and void by judicial decisions of the Oberlandesgericht Düsseldorf (Higher Regional Court, Düsseldorf, Germany) of 8 May 2013 and of the Bundesgerichtshof (Federal Court of Justice, Germany) of 6 October 2015. That exemption was subsequently repealed, with effect from 1 January 2014, by the StromNEV, as amended by the Verordnung zur Änderung von Verordnungen auf dem Gebiet des Energiewirtschaftsrechts (Ordinance amending the Energy Regulations) of 14 August 2013. ( 8 ) The latter ordinance reintroduced the individual charges calculated according to the physical route method, with the application – instead of the minimum charge – of standard charges of 10, 15 and 20% of the general charges, based on electricity consumption (7000, 7500 and 8000 hours of annual use of the network) (‘the standard charges’).

14.

StromNEV 2013 introduced a transitional scheme, in force from 22 August 2013 and applicable retroactively to baseload consumers who had not yet received the exemption at issue for the years 2012 and 2013 (‘the transitional scheme’). Instead of the individual fees calculated in accordance with the physical route method and the minimum fee, that scheme provided exclusively for the application of the standard fees.

15.

In its two identical cross-appeals in Cases C‑790/21 P and C‑791/21 P, the Commission puts forward three grounds of appeal. The first two are, in essence, identical to the two grounds raised in the cross-appeals in Cases C‑792/21 P to C‑796/21 P and C‑800/21 P. In Case C‑790/21 P, Covestro relies on four grounds of appeal. In Case C‑791/21 P, Germany puts forward a single ground of appeal. However, as requested by the Court of Justice, the present Opinion will address only the first ground of the Commission’s cross-appeals (admissibility of an action for annulment) and Covestro’s third ground of appeal, which chimes with Germany’s single ground of appeal (in so far as these concern the condition relating to the existence of an intervention by means of ‘State resources’).

16.

The Commission submits that the General Court erred in law, in paragraphs 37 to 44 of the judgment under appeal, by adopting a broad interpretation of the concept of ‘publication’ within the meaning of the sixth paragraph of Article 263 TFEU. In the first place, according to the Commission, the General Court’s interpretation is contrary to the case-law of the Court of Justice, in which the latter established a parallel between the sixth paragraph of Article 263 TFEU and Article 297 TFEU. In the Commission’s submission, it is clear from that case-law that publication is the starting point of the period prescribed for instituting proceedings only if it is a precondition for the entry into force of the measure in question, and if it is provided for by the Treaty itself. In the second place, as regards, more specifically, the publication in the Official Journal of the European Union (‘Official Journal’) of a Commission decision closing a formal investigation procedure, that does not constitute a ‘publication’ within the meaning of the second subparagraph of Article 297(2) TFEU. It does not therefore constitute the starting point of the period prescribed for instituting proceedings. In the third place, the Commission puts forward a series of arguments which, in its view, support its interpretation of the sixth paragraph of Article 263 TFEU, such as the scheme of that provision, the principle of equality of arms or the mandatory nature of the time limits for bringing proceedings.

17.

Germany and Covestro submit, in essence, that the General Court’s interpretation of the concept of ‘publication’, within the meaning of the sixth paragraph of Article 263 TFEU, is not vitiated by any error of law.

18.

It is true that, in practice and as has been the case in this instance, State aid beneficiaries are often informed of the content of a Commission decision declaring the aid scheme incompatible with the internal market and unlawful, and ordering the recovery of the aid before that decision is published in the Official Journal. The beneficiaries suffer the legal effects of that decision in the context of the recovery procedure, which must take place immediately after the decision was adopted and notified to the Member State concerned.

19.

The Commission argues that, in those conditions and contrary to what the General Court ruled in the judgment under appeal, it is the date on which the decision at issue came to the knowledge of the State aid beneficiary which must be taken into account for the purposes of calculating the time limit for bringing an action for annulment, and not the date of its publication in the Official Journal. The Commission therefore submits that the General Court should have declared Covestro’s action inadmissible, since it was out of time.

20.

I propose that the Commission’s interpretation should be rejected, as it is supported neither by the wording of the sixth paragraph of Article 263 TFEU, nor by the case-law of the Court, nor by the purpose of that provision.

21.

The wording of the sixth paragraph of Article 263 TFEU provides that ‘the proceedings provided for in this Article shall be instituted within two months of the publication of the measure, or of its notification to the plaintiff, or, in the absence thereof, of the day on which it came to the knowledge of the latter, as the case may be’.

22.

As has been pointed out by Germany, and as follows from the Court’s case-law, it can be inferred from the wording of that provision that the criterion of the date on which a measure came to the knowledge of an applicant, as the starting point of the period prescribed for instituting proceedings, is subsidiary to the criteria of publication and notification.

23.

As regards notification, it is important to bear in mind that the procedure for reviewing State aid is a strictly bilateral procedure between the Commission and the Member State concerned. Aid beneficiaries have only the status of ‘interested parties’, within the meaning of Article 1(h) of Regulation (EU) 2015/1589. ( 9 ) It follows that that Member State is the sole addressee of a negative decision of the Commission and the entry into force of that decision results from this bilateral procedure, that is, the Commission’s notification of the decision to the Member State. Since such a decision has already entered into force upon notification, the determination of the starting point for unnotified parties to institute proceedings should be unrelated to the conditions for a measure to enter into force and should, instead, be derived from an interpretation of the wording and purpose of the sixth paragraph of Article 263 TFEU.

24.

According to the wording of the sixth paragraph of Article 263 TFEU, as regards persons who are not notified, the publication of the measure in the Official Journal becomes the main element in determining the starting point for calculating the time limit for instituting proceedings. It follows that the moment when the decision came to the knowledge of the aid beneficiary, in so far as it is subsidiary to the other criteria, would only become relevant in a situation where no publication has taken place (even though it was required).

25.

In my view, the Commission cannot rely on the judgment in Portugal v Commission. ( 10 ) That judgment concerned a different situation to that in the present case. Indeed, in the former case, the decision at issue was specifically addressed to the Portuguese Republic, whereas Covestro, as the aid beneficiary, is only an interested party; it is not an addressee of the decision at issue.

26.

In addition, the Commission relies on a misreading of that judgment as regards the alleged parallel between the sixth paragraph of Article 263 TFEU and the second and third subparagraphs of Article 297(2) TFEU. In that judgment, the Court did not interpret the concept of ‘Bekanntgabe’ (‘notification’/‘publication’ in English) as meaning ‘actual awareness’. The Court ruled only on the relationship between the two alternatives provided for in the sixth paragraph of Article 263 TFEU, in order to hold that the decision addressed to the applicant takes effect upon its notification and at the same time that notification constitutes the starting point of the period prescribed for instituting proceedings.

27.

The Commission discusses the ‘notification’, claiming that the undertakings were notified by the German authorities in the present case. However, when those authorities – which are not the institution that adopted the act – send a copy of the decision at issue to the aid beneficiaries, that cannot be classified as a formal notification of a Commission decision within the meaning of Article 297 TFEU. Such a communication of a decision to the beneficiaries therefore does not have the same effects as a notification within the meaning of that article. In any event, ‘notification’ (‘Zustellung’ in German) of a Commission decision is not a live issue in the present case.

28.

The other precedents cited by the Commission do not support the hypothesis defended by the latter either. Thus, the order in Fryč v Commission is not relevant in so far as it concerned the action for annulment of regulations published in the Official Journal. Similarly, the order in Iordăchescu v Parliament and Others ( 11 ) concerned a directive also published in the Official Journal.

29.

Finally, the purpose of the time limits for instituting proceedings is to create objective criteria in order to guarantee legal certainty. It follows that such a time limit must be clear to and easily identified by everyone and seeks to avoid any discrimination or arbitrary treatment in the administration of justice. ( 12 )

30.

Thus, as Germany observes, it is precisely in order to maintain legal certainty that the sixth paragraph of Article 263 TFEU identifies the publication of the measure as the main event for determining the time limit for bringing proceedings. Indeed, it is always possible to determine that date with certainty, unlike the date on which the measure came to the beneficiary’s knowledge, proof of which is much more difficult to provide. The alternative put forward by the Commission would go against this legally certain calculation. In that alternative, any time limit would have to be determined individually, and the General Court would be compelled to carry out, with significant effort, an in-depth investigation into the actual date on which the applicant became aware, perhaps by chance, of the measure in question.

31.

Furthermore, the interpretation advocated by the Commission would have the effect of limiting the effective legal protection of the aid beneficiary, in so far as, for instance, the latter could receive from the Member State in question an incomplete or redacted version of the Commission’s decision. Such notification is not governed by EU law and the Member State could omit, arbitrarily or through negligence, to notify some or all of the aid beneficiaries, which would pose obvious problems as regards the principle of equal access to justice.

32.

The Commission’s argument that, in the present case, Covestro – which received the decision at issue in the context of the national procedure for the recovery of the aid – brought its action for annulment before the publication of the decision at issue in the Official Journal, does not alter my conclusion.

33.

Indeed, the Court has recalled in previous case-law that ‘the third paragraph of Article 33 of the ECSC Treaty, which prescribed notification and publication as the formalities from which the time limit for bringing an action for the annulment of a decision was to run, did not prevent an applicant from lodging an application with the Court as soon as the contested decision had been adopted, without waiting for it to be notified or published. One of the actions at the origin of that judgment was therefore admissible, notwithstanding the fact that it had been lodged at the Court Registry prior to publication of the contested decision’. Therefore, the Court ruled in the judgment cited that ‘nothing in the sixth paragraph of Article 263 TFEU, which corresponds to the third paragraph of Article 33 of the ECSC Treaty, prevents that case‑law law from being transposed to the present case’ ( 13 ) and held, essentially, that the fact that an action for annulment was brought prior to the publication of the decision in question did not constitute a reason for declaring the action inadmissible.

34.

It follows from the above assessment that the fact of bringing an action for annulment early (that is, prior to the actual publication of the act at issue) does not have any bearing on the starting point for instituting proceedings and for calculating the time limit for bringing such proceedings, as defined by the sixth paragraph of Article 263 TFEU.

35.

Finally, contrary to the Commission’s argument, the Opinion of Advocate General Campos Sánchez-Bordona in Georgsmarienhütte and Others (C‑135/16, EU:C:2018:120) is irrelevant for the purposes of the present cases, in that the Court ruled in that case on the admissibility of a request for a preliminary ruling and did not rule on the starting point of the period for bringing an action before the General Court for annulment of the Commission’s decision.

36.

It follows, in my view, that the General Court rightly held that the inadmissibility raised by the Commission at first instance had to be rejected and that the Commission’s cross-appeals should be dismissed as unfounded.

37.

Covestro’s third ground of appeal (Case C‑790/21 P) relating to the concept ‘State resources’ is based on three arguments: (i) it argues that the existence of tax and State control are not alternative criteria; (ii) it contests the classification of the surcharge at issue as a tax; and (iii) it submits that there was no State control over the resources generated by the surcharge at issue. Germany’s single ground of appeal (Case C‑791/21 P) is divided into three parts. The first part, contesting that ‘the existence of a tax and State control are two alternative criteria’, is based on the following arguments: (a) there are contradictions in the judgment under appeal; (b) it does not follow from the case-law of the Court of Justice that it is sufficient to conclude that the measure involves State resources solely on the basis that the surcharge can be classified as a tax; (c) the scheme of the Treaties is contrary to the General Court’s interpretation; (d) that interpretation has consequences which are ‘not foreseen by the Treaties’; and (e) even where there is a tax, it is necessary to assess whether there is State control. The second part, ‘the surcharge imposed by the BNetzA is not a tax within the meaning of the [case-law of the Court of Justice]’ is based on the following arguments: (a) there is no ‘tax’ if the tax debtor is not the final consumer; and (b) the General Court should have examined not only the obligation to collect the tax, but also the obligation to pay it. By the third part, Germany argues that the General Court erred in law by holding that the exclusive use of the network charges collected did not rule out that the State could dispose of those funds.

38.

The Commission contends that Covestro’s third ground of appeal and Germany’s single ground of appeal must be rejected as unfounded.

39.

Given that the various arguments concerning State resources raised by Covestro in the third ground of appeal (Case C‑790/21 P) and by Germany in the single ground of appeal (Case C‑791/21 P) significantly overlap, it is appropriate to address them together, and in the order and following the structure of the judgment under appeal. As a result, it is first necessary to deal with the alternative nature of the criteria; next, the classification of the surcharge at issue as a ‘tax’; and, finally, the existence of State control over the resources generated by the surcharge at issue.

40.

Covestro and Germany argue, in essence, that the General Court erred in law when it considered that the existence of a tax and State control over the funds generated by that tax (or over the administrators of those funds) were ‘two factors which together form an alternative’. The mere existence of such a tax is not sufficient to conclude that the present case involved ‘State resources’.

41.

The General Court examined the case-law on the criterion of ‘State resources’ and considered that ‘in essence, the case-law of the Court of Justice … relies on two main factors in order to assess whether resources are State resources: first, the existence of a compulsory levy on consumers or customers, normally classified as a “charge”, or more specifically as a “parafiscal charge”, and, secondly, State control over the administration of the scheme, in particular through State control over the funds or the (third-party) administrators of those funds. These are, in essence, two factors which together form an alternative’ (judgment under appeal, paragraph 95, emphasis added).

42.

According to the General Court, ‘it is therefore necessary to ascertain whether the [financing] mechanism at issue meets the conditions laid down by the relevant case-law as regards the use of State resources … and therefore whether the surcharge at issue is in fact a compulsory burden and, therefore, comparable to a parafiscal charge or, if not, whether the State has, at the very least, control over the funds collected or over the bodies responsible for managing those funds’ (judgment under appeal, paragraph 109).

43.

Covestro and Germany submit that the criteria of a ‘tax’ and of State control are cumulative and rely on the judgments in FVE Holýšov I and Others v Commission, ( 14 )Germany v Commission ( 15 ) and PreussenElektra. ( 16 ) However, these and other (more recent) judgments of the Court argue instead in favour of the criteria being alternative, in line with what the General Court held in paragraphs 90 and 95 to 97 of the judgment under appeal. ( 17 )

44.

I consider that the arguments put forward by Covestro and Germany are unfounded in so far as they may be rejected on the basis of the Court’s most recent case-law, which dispels any possible doubts as to whether the criteria are alternative or cumulative.

45.

In the judgment in DOBELES HES, ( 18 ) the Grand Chamber of the Court dealt with two issues that are relevant to the present appeals: (i) the definition of the concept of a tax and (ii) the relationship between the criterion of a tax and the criterion of State control over the financing mechanism.

46.

In paragraph 39 of that judgment, the Court made clear that the relevant criteria in Article 107(1) TFEU form part of an alternative: ‘the criterion mentioned in the preceding paragraph of the present judgment [that is, the existence of a charge,] is not the only criterion for identifying “State resources” within the meaning of that provision. The fact that sums constantly remain under public control, and therefore available to the competent national authorities, is sufficient for them to be categorised as “State resources”’.

47.

It follows that it is sufficient to satisfy one or other of the two criteria, which is confirmed by paragraph 42 of the judgment in DOBELES HES, where the Court makes direct reference to the existence of ‘two alternative criteria’.

48.

The General Court therefore did not err in law in the judgment under appeal when interpreting Article 107(1) TFEU, as it followed pre-existing case-law of the Court of Justice when it stated in paragraph 95 that the above criteria ‘together form an alternative’, and that approach has since been again confirmed in the judgment in DOBELES HES. As a result, it was sufficient for the General Court to identify, in the judgment under appeal, either the existence of a tax or State control over the financing mechanism, whereas each of those criteria would have been sufficient on its own for the purposes of confirming that State resources were involved. It follows that the arguments put forward by Covestro and Germany that those criteria are cumulative should, in my view, be rejected as manifestly unfounded.

49.

It is true that the General Court did not state this expressly, but I take the view that it is clear from a reading of the judgment under appeal (in particular, paragraph 109) that it examined the second criterion merely for the sake of completeness. Given that the alternative character of the criteria was the subject of much disagreement between the parties before the General Court, the latter was guided by the principle of the sound administration of justice and examined both factors of the alternative.

50.

Next, in support of its argument that the criteria of a ‘tax’ and State control should be cumulative, Germany also submits that the interpretation of Article 107(1) TFEU in the judgment under appeal (which is confirmed by the Court’s case-law) is, in any case, contrary to the scheme of the Treaties and would result in ‘consequences which are not foreseen in the Treaties’. It states that any tax or additional charge imposed by the State would necessarily come under the notion of ‘State resources’, which would run counter to the fiscal sovereignty of Member States. It argues further that the prohibition of illegal State aids seeks to safeguard against State infringements of free competition, which implies that only infringements of competition with a link with the State budget or involving resources subject to the State’s power of disposal are concerned.

51.

However, as has been pointed out by the Commission, those arguments are based on an incorrect reading of the Treaties ( 19 ) as well as on an amalgamation of, on the one hand, the case-law relating to Articles 30 and 110 TFEU and, on the other hand, that relating to Article 34 TFEU. The prohibition of discrimination under the former two articles is explained by the fact that both customs duties and taxes are, by nature, fiscal (they constitute taxes used by the State to achieve political objectives). Moreover, suffice it to point out that the main deciding factor in relation to State resources is that the distortion of competition is the work of the State (and not of an undertaking). However, significantly, contrary to what Germany appears to suggest, the scheme of the Treaties does not contain any indication that only aid ‘financed by the State budget’ is to be regarded as State aid. Indeed, the Court’s case-law has always made clear that the simple fact that the flow of resources passes exclusively through private-law entities is not enough to refute the existence of the State’s effective control within the meaning of the notion of aid. ( 20 )

52.

Contrary to the arguments advanced by Germany, the systematic difference in position in the FEU Treaty between Articles 30 and 110 and Article 107 has no material impact for the purposes of the present cases. Those three provisions have instead something in common, that is to say, the fact that they relate to the fiscal action of the State. Indeed, the Court has adopted, in relation to the two categories of provision, a broad interpretation of ‘State control’, in particular, so as to prevent circumvention of the rules contained in those provisions. ( 21 )

53.

Furthermore, it is incorrect to claim that ‘any [transfer] of assets between private [legal persons], ordered by law, [would be assimilated] to a use of State resources’. For instance, a minimum price set by law is not a tax. It is sufficient to refer to the judgments in PreussenElektra and in EEG 2012. The judgment in PreussenElektra concerned a legal minimum price applied between two private-law entities. The private-law entity which was required to apply the minimum price could not pass the costs on to its customers. Therefore, there was no tax.

54.

In the judgment in EEG 2012, the Court clarified that the mere possibility of an impact on customers was not sufficient to consider that the resources used were State resources. The use of State resources – in the form of a tax – can only be presumed if the private-law entity on which the obligation weighs also has the obligation to pass on the additional cost to a third party by way of collecting from it a tax.

55.

In my opinion, it follows from all the foregoing considerations that the General Court rightly considered that those criteria form part of an alternative. Therefore, the judgment under appeal is not vitiated by any error of law in that regard and the first series of arguments raised by Covestro and Germany should be dismissed as unfounded.

56.

Covestro and Germany criticise the General Court in relation to the first of the alternative criteria discussed above, in that it wrongly classified the surcharge at issue as a ‘tax’. They argue that that classification is based on an incorrect interpretation of the concept of a ‘tax’ and on a distortion of national law.

57.

In the first group of arguments, Covestro and Germany submit, in essence, that the General Court misinterpreted the concept of ‘final consumer’ within the meaning of national law, since neither national law nor the German national authorities impose a payment obligation on consumers or system operators. Such an obligation was possible only under a network use contract drawn up by the regulatory authorities, which did not exist for the years 2012 and 2013. According to Covestro, it was for the electricity supplier, as the network user, to pass on the surcharge at issue to the final consumer. Even if all suppliers had systematically passed that surcharge on to the final consumer, such a practice would not have been sufficient for it to be found that there was a legal obligation.

58.

I consider that it is necessary to examine whether the question of who the ultimate debtors of the surcharge at issue were is a decisive consideration.

59.

Covestro argued at first instance that the surcharge was only collected from network users and that it was not compulsorily passed on to final consumers of electricity, so that it should have been understood as a network fee and not as a tax. However, in paragraph 119 of the judgment under appeal, the General Court ruled that that argument was irrelevant, in view of the fact that the ultimate debtors of that surcharge were network users (the providers themselves and the final consumers directly connected to the network), and not other final consumers.

60.

Also in paragraph 120 of the judgment under appeal, the General Court found that the decision at issue identified an obligation to collect and pass on the surcharge at issue to ‘final consumers’. To my mind, it is relevant that this interpretation was corroborated by point 20 of the judgment of the Bundesgerichtshof (Federal Court of Justice) of 6 October 2015, which rejects Covestro’s argument that the surcharge at issue was a network fee. That court confirmed that the surcharge was not a contractual compensation, but an external collection of charges, which was imposed on the operators.

61.

Moreover, it follows from the Court’s case-law that, in order for a measure to constitute a ‘charge’ within the meaning of Articles 30 and 110 TFEU, it is sufficient that that measure was levied on intermediate products or services, without it necessarily being passed on to the final consumers of the downstream products or services. There is no support, in the same case-law of the Court, ( 22 ) for the view that a tax implies the existence of State resources only if that tax is applied to final consumers. Indeed, I consider that the concept of tax, which is referred to in Articles 30 and 110 TFEU, must be construed broadly on the basis of a teleological interpretation.

62.

The General Court rightly relied on the same case-law in paragraph 121 of the judgment under appeal.

63.

I consider that the General Court was correct to base its approach on the question whether private or public bodies, which are tasked with implementing the scheme, must assume the expenses through their own financial resources (a situation involving no State aid) or whether they can collect those resources from third parties (a situation in which there is State aid). It matters little whether those third parties are final consumers or whether they are an intermediate link in the value chain. Indeed, the only decisive factor is the redistributive effect of the tax, namely that a State or a private-law entity collects that tax from undertakings and uses it to finance an advantage granted to another category of undertakings.

64.

It follows from a teleological interpretation and the Court’s case-law on the concept of ‘tax’, within the meaning of Articles 30 and 110 TFEU, that the identity of the person liable to pay the tax in each case is irrelevant. The decisive factor is whether the tax relates to the product concerned or to a necessary activity in connection with that product. ( 23 )

65.

In my opinion, it follows that the first group of arguments should be rejected as unfounded.

66.

Covestro contends that the General Court relied solely and incorrectly on the BNetzA decision of 2011, whereas that decision was declared unlawful, retroactively, by the German courts.

67.

The question arises whether the General Court was entitled to conclude that the surcharge at issue was mandatory by way of relying on that BNetzA decision, despite the fact that it was subsequently annulled and declared null and void by the Oberlandesgericht Düsseldorf (Higher Regional Court, Düsseldorf) and, on appeal, by the Bundesgerichtshof (Federal Court of Justice).

68.

I would recall that it follows from the judgments in Commission v Aer Lingus and Ryanair Designated Activity ( 24 ) and in Heiser, ( 25 ) that the possible unlawfulness of a national scheme is not capable of removing its State aid nature.

69.

According to the judgment in Heiser, ‘even if the legislation providing for the adjustment of deductions … is unlawful, the fact nonetheless remains that that legislation is liable to have an impact as long as it is not repealed or, at the very least, as long as its unlawfulness is not established’ (emphasis added).

70.

Similarly, according to the judgment in Commission v Aer Lingus and Ryanair Designated Activity, ‘the fact that a tax measure is contrary to provisions of EU law other than Articles 107 and 108 TFEU does not mean that the exemption from that measure enjoyed by certain taxpayers cannot be classified as State aid, as long as the measure in question produces effects vis-à-vis other taxpayers and has not been either repealed or declared unlawful and, therefore, inapplicable’ (emphasis added).

71.

As the Commission stated at the hearing, the German legislator dealt with the situation in a manner that ensured that the collection of the surcharge at issue, following the annulment ordered by the Bundesgerichtshof (Federal Court of Justice), was retroactively corrected and continued to have legal effects and, therefore, maintained the obligation of the surcharge at issue.

72.

To my mind, in EU law on State aid, a measure must be considered from the viewpoint of its effects.

73.

In my opinion, it follows from the aforementioned case-law that the General Court was entitled to hold (in paragraphs 107 and 125 of the judgment under appeal) that for the purposes of reviewing State aid, the fact that the BNetzA decision of 2011 was actually applied and produced legal effects during the relevant period was a decisive factor.

74.

That decision was legally binding during the relevant period and required network operators to collect the surcharge at issue from network users.

75.

The wording of paragraph 3 of the BNetzA decision of 2011 imposed the obligation to collect the surcharge at issue on distribution system operators, which were therefore required to collect that surcharge from their customers. Given that that decision formed part of the scheme in force during the relevant period and produced binding effects – a fortiori since those effects were actually not withdrawn following the annulment ordered by the national courts (by way of provisions which successively repealed that scheme) – the General Court correctly ruled that the scheme based on the surcharge at issue produced legally binding effects (judgment under appeal, paragraph 125).

76.

My assessment in the present Opinion is also supported by the Court’s judgment in Deutsche Post v Commission, ( 26 ) in which it was held that an annulled decision is still capable of producing legal effects. Since the action for annulment in that case had been brought prior to the General Court giving judgment annulling the decision at issue (and thus retroactively excluding it from the legal order), the admissibility of that action had to be assessed by reference to the time when it was brought since the negative decision of 2002 was still in force and continued to form part of the EU legal order.

77.

Finally, it should be pointed out in that regard that the production of legal effects was, moreover, the stated objective of the German legislator. ( 27 )

78.

In my opinion, the arguments seeking to rely on the fact that the BNetzA decision of 2011 was declared unlawful by the national courts should therefore be rejected as unfounded.

79.

Covestro and Germany submit that in paragraph 121 of the judgment under appeal, the General Court misinterpreted the concept of ‘tax’, within the meaning of Articles 30 and 110 TFEU, so as to conclude that the surcharge at issue corresponded to that concept. It is true that, in order to justify the fact that an obligation to levy a surcharge linked to the use of the network can also be a tax, the General Court refers to the concept of tax for the purposes of Articles 30 and 110 TFEU.

80.

As explained in point 61 of the present Opinion, the concept of tax, which is referred to in Articles 30 and 110 TFEU, must be construed broadly.

81.

In the judgment in Essent Netwerk Noord (paragraph 40), the Court held that ‘Articles [30 and 110 TFEU] … complement each other in pursuing the objective of prohibiting any national fiscal measure that is liable to discriminate against products coming from or destined for other Member States by constituting a restriction on their free movement within the [European Union] in normal conditions of competition’ (emphasis added).

82.

As has been observed in point 64 of the present Opinion, it follows from the concept of ‘tax’, within the meaning of Articles 30 and 110 TFEU, that the identity of the person liable to pay the tax in each case is irrelevant. The decisive factor is whether such a tax relates to the product concerned or to a necessary activity in connection with that product.

83.

In the present cases, the surcharge at issue is collected in respect of the use of the network. In the General Court’s correct assessment, the fact that the surcharge does not constitute consideration for the use of the network, but rather a mandatory charge, is decisive for its classification as a tax.

84.

Covestro seeks to rely on the argument that, in previous cases dealt with by the Court, the relevant product was not the use of the network, but the electricity supplied. Consequently, the only question that mattered was whether, in the electricity supply relationship, a tariff supplement was levied, which had to be classified as a tax.

85.

However, Covestro’s argument is futile, in so far as, in the present cases, the use of the network must be regarded as an intermediate product or, more specifically, as an intermediate service for the purposes of Articles 30 and 110 TFEU, in accordance with the case-law of the Court of Justice discussed above. The General Court therefore did not err in law, in my view, by relying on Articles 30 and 110 TFEU in defining tax as a charge on a product or service provided.

86.

Furthermore, the resolution of the question regarding the mandatory nature of the charge as a condition for determining the existence of State resources can be derived from the existing case-law of the Court, in so far as the present cases are similar to the factual situation which gave rise to the judgment in Essent Netwerk Noord.

87.

The judgment in Essent Netwerk Noord concerned a tax in the Netherlands. A network operator (wholly controlled by public authorities) collected from domestic purchasers of electricity a tax (price surcharge) on the use of its electricity network. Similar to the present cases, in the judgment in Essent Netwerk Noord, that tax was foreseen by national law, and the network operator collected money from consumers and in such a manner financed the State aid.

88.

The General Court correctly referred to the judgment in that case and pointed out in paragraph 91 of the judgment under appeal that ‘the Court concluded that a surcharge on electricity transmitted, fixed according to objective criteria, imposed by legislation on electricity consumers and collected by the system operators, constituted a “charge”, the source of which was a State resource’.

89.

In the judgment in Essent Netwerk Noord, the Court of Justice makes several important observations for the purposes of the present appeals.

90.

First, the Court notes that it matters little whether the tax affects the product as such, for instance, electricity, or whether the tax will affect a necessary activity in relation to the product, for instance, the transport of electricity (judgment in Essent Netwerk Noord, paragraph 44). In the present cases, the tax affects the transport of electricity.

91.

Secondly, the Court points out that the fact that the surcharge is a charge imposed unilaterally is decisive for the existence of a tax (judgment in Essent Netwerk Noord, paragraph 45). In the present cases, the BNetzA decision of 2011 imposed on distribution system operators, in a legally binding way, the obligation to collect the surcharge at issue from final consumers, as network users (paragraph 132 of the judgment under appeal).

92.

Thirdly, the Court rules that ‘it is of little account that the financial charge is not levied by the State’ (judgment in Essent Netwerk Noord, paragraph 46). In the present cases, distribution system operators were responsible for collecting the surcharge at issue.

93.

Finally, in the present cases, similar to the situation which gave rise to the judgment in Essent Netwerk Noord, the surcharge at issue was intended to cover network usage costs which German baseload consumers would normally have had to bear themselves.

94.

The exemption from network charges for baseload consumers has therefore improved their competitive position. Due to the tax status of the surcharge at issue in the present cases, it would be appropriate to recognise those resources as being State resources within the meaning of Article 107(1) TFEU. Further, that would be consistent with the case-law, according to which the financing of an advantage through the proceeds of a tax implies that the advantage originates from a State resource. ( 28 )

95.

I consider that not only the judgment in Essent Netwerk Noord, but also that in FVE Holýšov I are particularly relevant. Both of those cases have a similar factual background to that of the present cases and show that the Court has, in fact, already addressed the questions raised in the present appeals.

96.

In the judgment in FVE Holýšov I (paragraph 46), the Court explained that it does not matter whether or not the financing mechanism in question is classified, under national law, as a tax or a parafiscal charge. EU State aid law is only concerned with knowing that such a charge is levied, and that levy is ordered unilaterally and through an act of the State.

97.

In both of those judgments, the Court observes that this is an obligation to collect money unilaterally and that these are therefore State resources. This issue brings us to the recent judgment in DOBELES HES (paragraph 34), which provides further clarification of the judgment in Essent Netwerk Noord. I consider that this is important for the purposes of the outcome of the present appeals.

98.

In that paragraph 34, the Court states that ‘as regards, in the second place, the condition that the advantage be granted “through State resources”, which the referring court specifically raises, the Court has held that amounts resulting from the price surcharge imposed by the State on purchasers of electricity are similar to a charge which is levied on electricity and have their origin in “State resources” within the meaning of Article 107(1) TFEU’. ( 29 )

99.

I would point out that the judgment in DOBELES HES goes on, in paragraphs 36 and 37, to clarify other existing case-law in addition to that in the judgment in Essent Netwerk Noord (in particular, the judgment in EEG 2012, paragraphs 70 and 71).

100.

The Court highlights differences between the judgments in EEG 2012 and DOBELES HES. It explains that, in the former case, there was no compulsory contribution, since the collection of the EEG charge was not legally binding and was on a voluntary basis only. The Court therefore makes clear that it is the mandatory nature which is decisive in determining whether the charge in question is a compulsory contribution. The mere fact that money is transferred or can be transferred is not enough in itself. However, in the present cases, there was precisely an obligation to collect the tax (or charge), whereas the judgment in EEG 2012 did not relate to a tax.

101.

Contrary to the situation in the judgment in EEG 2012, the collection of the surcharge at issue in the present cases is not a commercial decision of the network operators, but is instead carried out on the basis of legal provisions (in particular, the decision of the BNetzA).

102.

Therefore, it is necessary to make a distinction between the financing mechanism at issue in the present cases and the mechanism concerned in the judgment in EEG 2012. In the present cases, the existence of the surcharge per se is sufficient for it to be classified as constituting State resources.

103.

As far as concerns the other of the two criteria from the judgment in EEG 2012 – that is, State control (power of disposal) over the funds themselves or over the entity which administers those funds – as we have seen above, the two criteria are alternative and not cumulative.

104.

Had there been a mandatory surcharge in the judgment in EEG 2012, then it would not have been necessary for the Court to enter into the discussion in paragraph 72 of that judgment.

105.

Indeed, it is precisely because the conditions in paragraph 71 were not met that the Court found it necessary to continue its line of reasoning in paragraph 72.

106.

In paragraph 71 of that judgment, the Court ruled that ‘the fact, noted by the General Court …, that “in practice”, the financial burden resulting from the EEG surcharge was passed on to the final customers and, consequently, could “be assimilated, from the point of view of its effects, to a levy on electricity consumption” is not sufficient for it to be concluded that the EEG surcharge had the same characteristics as the electricity price supplement examined by the Court … in the judgment [in Essent Netwerk Noord]’.

107.

The Court therefore begins paragraph 72 with the word ‘consequently’: ‘Consequently, it is necessary to determine whether the other two factors, … recalled … in paragraph 62 of the present judgment, allowed [the General Court] nonetheless to conclude that the funds generated by the EEG surcharge constituted State resources, since they constantly remained under public control and were therefore available to the public authorities … In that situation, it is irrelevant whether or not the EEG surcharge may be classified as a “levy”’.

108.

In other words, the situation in the judgment in EEG 2012 was entirely different to the present cases.

109.

In the judgment in EEG 2012, the German legislature allowed electricity undertakings to collect certain charges, but this was merely a possibility and not an obligation. The State gave the operators a choice: if they availed themselves of that possibility, then the money collected had to be affected to a specific use determined by the State (that is to say, it was intended to compensate supplementary costs linked to renewable energies). Therefore, in that case, the Court came to the conclusion that, as there was no obligation to collect those charges, the State did not control those funds.

110.

It follows from the foregoing considerations and case-law that the surcharge at issue was, as the Commission found in the decision at issue, a legally mandatory collection of a tax, as also established by the Bundesgerichtshof (Federal Court of Justice), which, as confirmed most recently in the judgment in DOBELES HES, is sufficient to find that State resources are involved.

111.

I therefore take the view that the judgment under appeal is not vitiated by an error of law in this regard.

112.

In my opinion, the arguments based on Articles 30 and 110 TFEU (concept of ‘tax’) and on the judgment in Essent Netwerk Noord and subsequent case-law (mandatory nature of the charge) should be rejected as unfounded.

113.

Covestro and Germany argue, in essence, that the General Court distorted national law in so far there was no obligation to collect the surcharge at issue.

114.

I consider that the classification of the surcharge at issue as a tax is not based on any distortion of national law.

115.

In paragraphs 120 and 122 to 124 of the judgment under appeal, the General Court found, on the basis of national law (in particular, the decision of the BNetzA of 2011), that the charge at issue was levied on final consumers. Thus, the argument that there is a ‘charge’, as defined in the judgment in Essent Netwerk Noord, only where the person liable for payment of the charge is the final consumer, is ineffective. Indeed, Germany cannot dispute that finding of the General Court relating to national law, which is a finding of fact, on appeal before the Court of Justice, and that Member State has in fact refrained from doing so directly.

116.

As explained in points 59 to 64 of the present Opinion, it follows from the case-law that the identity of the person liable to pay the tax in each case is irrelevant. Instead, the decisive factor is whether the tax relates to the product concerned or to a necessary activity in connection with that product.

117.

Moreover, I agree with the Commission that the collection obligation and the payment obligation are two sides of the same coin, ( 30 ) with the result that any determination as to whether the payment obligation imposed on network users stemmed from the national legislation or the BNetzA decision of 2011 is also irrelevant.

118.

In that connection, contrary to the arguments put forward by Covestro, the fact that the BNetzA decision of 2011 made no provision for penalties for the non-collection of the surcharge at issue is not decisive. The BNetzA exercised normal supervisory powers over the transmission system operators, and could address binding decisions to those operators, in the event of failure to comply with their obligations. ( 31 )

119.

It follows that the arguments based on the distortion of national law and on the lack of obligation to collect the surcharge at issue and the lack of obligation to pay it should, in my opinion, be rejected as unfounded.

120.

In Covestro’s submission, contrary to what the General Court states in paragraphs 126 to 130 of the judgment under appeal, the national authorities granted the network operators no full compensation for loss of revenue and costs generated by the exemption from charges. It wrongly inferred, in paragraph 127 of the judgment under appeal, the existence of a tax.

121.

Contrary to what Covestro maintains, the General Court rightly found that the disputed surcharge made it possible fully to compensate for the loss suffered by the network operators due to the exemptions from network charges. This resulted from the sixth and seventh sentences of Paragraph 19(2) of StromNEV 2011 and the BNetzA decision of 2011. What was important for the assessment of the surcharge under EU law on State aid was the fact that the exemption of baseload consumers from network charges and the resulting advantage were entirely financed by the surcharge.

122.

However, in paragraph 130 of the judgment under appeal, the General Court rightly rejected the argument that there was no legal mechanism to ensure full compensation for loss suffered (in particular on the ground that it was impossible to pass on the costs of the surcharge at issue in the event of irrecoverable debts). Indeed, the classification of the surcharge at issue as a parafiscal charge is sufficient for the revenue from that charge to be regarded as State resources, without it being necessary for the State to undertake to offset the losses generated by the non-payment of that surcharge, in particular in the case of irrecoverable debts.

123.

It follows, in my view, that the arguments based on the lack of full reimbursement of the loss of revenue and costs should be rejected as unfounded.

124.

In my opinion, it follows from all the foregoing considerations that the second series of arguments raised by Covestro and Germany should be rejected as either ineffective or unfounded.

125.

Germany submits essentially that paragraphs 95 to 97 and 109 of the judgment under appeal contradict paragraphs 133 to 134 of that judgment. As regards the case-law of the Court of Justice, the decisive factor is the control and power of disposal over the funds concerned. It also argues that the General Court erred in finding that the relationship between the supplier and the final consumer of electricity was not decisive in concluding in casu that there was a mandatory charge, on the ground that the surcharge at issue is not levied on electricity consumption but on the use of the network.

126.

Covestro submits, in essence, that the judgment under appeal is vitiated by an error of law in that the General Court considered, on the basis of an incorrect description of national law, that there is State control over the funds generated by the surcharge at issue. It also relies on these arguments: (i) the finding that the revenue generated by the surcharge at issue was allocated to the scheme at issue argues against the State nature of the resources; (ii) even if the collection of the surcharge at issue were to have a legislative basis, that would not be sufficient to conclude that the resources were State resources; (iii) factors such as those relating to the mandate of network operators, or to the monitoring by the public authorities of the proper implementation of the system, do not establish the existence of State control over the funds; and (iv) the financing mechanism at issue does not meet the criteria identified by the Court for the assessment of State control, in particular those resulting from the judgment in EEG 2012.

127.

Given the alternative nature of the two criteria referred to in my assessment above, the existence of a tax is already sufficient for the State nature of the resources to be recognised. Given that the ‘tax’ criterion is fulfilled in the present cases and the General Court’s assessment in that regard is correct, the third series of arguments in the present appeals is ineffective and it is no longer necessary to examine it.

128.

It is therefore only for the sake of completeness that I will make the following remarks.

129.

Despite the fact that the ‘tax’ criterion is met and the assessment of the General Court – and, now on appeal, that of the Court of Justice can stop at this point – the fact remains that the General Court’s assessment of the ‘State control’ criterion would also have to be upheld in the present case.

130.

The additional costs resulting from the exemption of network charges for certain consumers were passed on to final consumers in accordance with the binding provisions of the law of the Member State concerned. ( 32 ) In addition, the mechanism of the surcharge at issue ensured that network operators were fully compensated for the loss of revenue, since the amount of that surcharge was adapted to that of the resources required on account of the exemption at issue. ( 33 )

131.

As regards Covestro’s argument that the amount of that surcharge was not prescribed by the State, suffice it to point out that the transmission system operators had no leeway in determining that charge and, as the General Court rightly held in paragraph 134 of the judgment under appeal, there was State control over the funds, that is to say, over the entire mechanism for collecting and allocating the surcharge at issue.

132.

Covestro and Germany, seeking to rely on the judgment in EEG 2012, submit that the exclusive allocation of the resources generated by the surcharge at issue precludes State control. However, as the General Court rightly explains in paragraphs 144 and 145 of the judgment under appeal, the Court of Justice in that judgment did not reverse its settled case-law – which was moreover, confirmed by more recent case-law ( 34 ) – but the Court confined itself to stating that, in the absence of other factors, that factor was not per se decisive in demonstrating the existence of such control. Indeed, in the present cases, the General Court did identify another such factor: the existence of a tax.

133.

The argument that there is no State control over the full mechanism for collecting the surcharge at issue is contradicted by the fact that there is a mandatory hypothecation between, on the one hand, the surcharge at issue (as a parafiscal charge) and, on the other hand, the aid granted in the form of an exemption from network charges. In accordance with the case-law of the Court, where there is such a link between the aid measure and its financing, it automatically follows from the existence of a tax, which serves to finance the aid in well-defined proportion, that the aid is granted by means of State resources (that is, the proceeds of the tax). ( 35 )

134.

Finally, Covestro’s arguments alleging that the rules relating to the levy at issue in the case which gave rise to the judgment in EEG 2012 are stricter than those governing the surcharge at issue in the present cases are unfounded. On the contrary, I consider that the rules governing the surcharge at issue in the present cases appear to be stricter than the levy at issue in EEG 2012, since the present surcharge is based on binding legal provisions as opposed to a commercial/voluntary decision of the network operators in that case.

135.

It follows that the third series of arguments raised by Covestro and Germany should be rejected, in my view, as ineffective and, in any case, unfounded.

136.

Therefore, Covestro’s third ground of appeal and the single ground of appeal raised by Germany, in so far as these concern the condition relating to the existence of an intervention by means of State resources, should be rejected as unfounded.

137.

In the light of the foregoing, I propose that the Court of Justice should: (i) reject the first ground of the Commission’s cross-appeals; and (ii) reject Covestro’s third ground of appeal and the single ground of appeal raised by the Federal Republic of Germany, in so far as these concern the condition relating to the existence of an intervention by means of State resources.