The ECJ’s Apple Judgement: A Game-Changer in the Architecture of EU Taxation or Just a Story of Procedural Defeat?

Di Giulio Allevato -

Abstract (*) (**)

Through its judgement of 10 September 2024, the European Court of Justice closed the Apple State aid case, that is the largest tax case in history, as it is worth € 13 billion. Contrary to most experts’ expectations, the Court of Justice overturned the General Court’s judgement and ruled in favor of the Commission, thereby upholding the qualification of two advance tax rulings granted by the Irish tax authorities to the Apple group as illegal State aid and condemning Ireland to recover an unprecedented amount of money. At first glance, such outcome seems to be in apparent contrast with the Court of Justice’s own recent case law (in particular, the Fiat and Amazon judgements), which had established that a correct interpretation of Art. 107 TFEU prohibits the use of external standards not incorporated into national law – such as the OECD guidelines on transfer pricing in the case of Luxembourg and Ireland – to determine the reference system for State aid purposes. However, some elements and recent developments may induce to minimize the effects of the Apple judgement on the architecture of EU taxation.

La sentenza Apple della Corte di Giustizia UE: un punto di svolta nell’architettura della tassazione europea o solo una storia di sconfitta procedurale? – Con la sentenza del 10 settembre 2024, la Corte di Giustizia UE ha chiuso il caso sugli aiuti di Stato di Apple, il più grande caso fiscale della storia, dal valore di 13 miliardi di euro. Contrariamente alle aspettative della maggior parte degli esperti, la Corte di Giustizia ha annullato la sentenza del Tribunale e si è pronunciata a favore della Commissione, confermando così la qualificazione di due tax rulings preventivi concessi dalle Autorità fiscali irlandesi al gruppo Apple come aiuti di Stato illegali e condannando l’Irlanda a recuperare una somma di denaro senza precedenti. A prima vista, tale esito sembra essere in evidente contrasto con la recente giurisprudenza della stessa Corte di Giustizia (in particolare, con le sentenze Fiat e Amazon) che aveva stabilito che una corretta interpretazione dell’art. 107 TFUE vieta l’uso di standard esterni non recepiti nel diritto nazionale – come le linee guida dell’OCSE sui prezzi di trasferimento nel caso del Lussemburgo e dell’Irlanda – per determinare il sistema di riferimento ai fini degli aiuti di Stato. Tuttavia, alcuni elementi e sviluppi recenti potrebbero indurre a minimizzare gli effetti della sentenza Apple sull’impianto della tassazione comunitaria.

SUMMARY: 1. The Apple case: a recap. – 2. The backbone of the Commission’s legal reasoning. – 3. The previous ECJ’s judgements and the expectations about Apple. – 4. The Apple judgements and its implications: four potential scenarios. – 5. Conclusions.

1. The recent judgement, by the European Court of Justice (“ECJ”), in the Case C-456/20P (ECJ, Judgement of 10 Sept. 2024, Case 465/20 P, European Commission v. Ireland, Apple Sales International et al, hereinafter simply “Apple”), ends a long-lasting case which has attracted the attention not only of scholars and policymakers, but also of the most informed part of the public opinion, due to the nature of the taxpayer involved (one of the so-called “Tech Giants”) and to the monetary amount at stake. The case originated from an investigation started by the EU Commission (“Commission”) back in 2014, after the so-called LuxLeaks scandal. The results of the investigation led the Commission to issue, in 2016, a decision qualifying two advance tax rulings on intra-group profits allocation granted, in 1990 and 2014, by the Irish tax authorities to the Apple group as illegal State aids under Art. 107 TFEU (see European Commission, Decision (EU) 2017/1283 of 30 Aug. 2016 on State Aid SA.38373 [2014/C] [ex 2014/NN] [ex 2014/CP] implemented by Ireland in respect of Apple, OJ L 187/1, 2017). It may be worth recalling that, according to Art. 107 TFEU a Member State’s measure qualifies as a State aid when it (i) is financed by the State or through State resources (“public funding” requirement); (ii) provides an advantage (“advantage” requirement); (iii) is selective (“selectivity” requirement); and (iv) affects trade between Member States and distorts competition (“trade and competition distortion” requirement).

Specifically, the tax rulings at issue involved two Apple group’s Irish-incorporated entities – Apple Sales International (“ASI”) and Apple Operations Europe (“AOE”) – which were part of a cost-sharing agreement with Apple Inc. – i.e., the Apple group’s head-company established and tax resident in the United States. The cost-sharing agreement entitled ASI and AOE to exploit the Apple group’s core technology and marketing intangibles outside the Americas. Both the Apple group and the Irish tax authorities agreed that, according to the relevant legislation in force in Ireland and in the United States at the time, ASI and AOE were not tax resident of either country, their only taxable presence in Ireland being constituted by permanent establishments performing routine functions like manufacturing, procurement, marketing, and support services. Through the advance rulings at issue, the Irish tax authorities validated a profit allocation method which, in light of the routine nature of the activities performed by the Irish PEs, allowed most of the profits of ASI and AOE to stay with their head offices. Thus, considering the stateless status of ASI and AOE, their income substantially escaped any significant taxation, leading to an effective tax rate lower than 1%, as documented in the Commission’s decision.

On the contrary, the Commission came to the conclusion that those rulings had granted a selective advantage to the Apple group vis a vis other entities having a taxable presence in Ireland, and, therefore, they qualified as illegal State aids (considering that other requirements under Art. 107 TFEU – such as the use of State funds and the impairment of competition within the single market – where also met). Indeed, the Commission argued that a correct application of the arm’s length principle (“ALP”) would have led to the attribution of the largest part of the profits to the Irish PEs, since the head offices of ASI and AOE did not perform any relevant activity other than holding board meetings and signing off on documents. Consequently, the Commission’s decision contained an order for Ireland to recover € 13 billion, amounting to the Irish corporate income tax that Ireland would have collected absent those two rulings between 2003 and 2014 (i.e., the years for which the statute of limitation had not run yet). It is important to note that the Commission claimed to ground its determination in the Authorized OECD Approach to the Attribution of Profits to the Permanent Establishment (“AOA”), which the Commission considered as sort of inherent in Irish law, although, as Ireland and Apple vigorously contended during the investigation, had not been incorporated into Irish national law.

On July 2020, the General Court (“GC”), before which Ireland and Apple had filed an action for annulment of the Commission’s decision, had annulled the Commission’s decision based on technicalities related to the implementation of the methods to determine the ALP. In particular, the GC had held that the Commission had wrongly allocated profits to the Irish branches of ASI and AOE using an ‘exclusion’ approach – i.e., without effectively demonstrating that the activities related to the IP licenses had been conducted in Ireland. Furthermore, the GC stated that those Apple entities were in a position to effectively develop and manage the Apple group’s IP and, therefore, to generate profits abroad (see General Court, 15 July 2020, Joined Cases T‑778/16 and T‑892/16, Ireland, Apple Sales International et al. v. European Commission, ECLI:EU:T:2020:338, at para. 166 to 249).

The appeal of the Commission against the GC judgment, therefore, centered on demonstrating that its decision was based on an effective and correct application of the separate entity approach as prescribed by the international and national norms concerning the attribution of profits to a permanent establishment. In its recent judgement, the ECJ accepted the main arguments of the Commission and concluded that the European judges of first instance erroneously applied and interpreted the rules concerning the attribution of profits to a PE. Thus, it overturned the GC’s decision and upheld the Commission’s original decision to qualify the advance tax rulings granted by the Irish tax authorities to the Apple group as illegal State aids, thereby requiring Ireland to recover the € 13 billion.

The Commission’s winning of the appeal has surprised many early commentators (see Sheppard L., EU Extortion in Apple, in Tax Notes, 23 Sept. 2024; Daly S., Another take on the [bad] Apple Ruling: is a misapplication of domestic law enough for a finding of State Aid? in Oxford Business Tax Blog, 17 Sept. 2024; Collier R., A Bad Apple Ruling, in Oxford Business Tax Blog, 13 Sept. 2024). First, it surprised that, contrary to the non-binding opinion of the General Attorney Pitruzzella, the ECJ did not refer the case back to the GC for a reassessment of the methodology used by the Commission in determining the allocation of profits between the two foreign Apple companies’ head offices and their Irish branches. Indeed, the General Attorney had recognized that the GC had erroneously established that the Commission had used an ‘exclusion’ approach. Second, and most important for our purposes, in two previous judgments on overlapping State aid cases concerning advance tax rulings where the Commission’s reasoning to demonstrate the existence of a selective advantage was substantially the same as in Apple, the ECJ had torn down the main arguments of the Commission’s legal reasoning and, consequently, had annulled the Commission’s decisions or confirmed the GC’s judgement that had annulled them. This has been the case of the ECJ’s judgement in Fiat and Amazon, and raised the expectation that the Commission would lose also in Apple (see Sheppard L., EU Extortion in Apple, supra).

2. Before entering into the details of the ECJ’s judgment in Fiat and Amazon, it is worth remembering that Apple has not constituted an isolated case. On the contrary, it is part of a wave of investigations originally triggered by the LuxLeaks scandal which led the Commission to issue, between 2016 and 2019, various decisions through which advance price agreements (“APAs”) granted to multinational corporate groups by several Member States – in particular, the Netherlands, Luxembourg, and Ireland – had been qualified as illegal State aids under Articles 107 and 108 TFEU, based on substantially the same legal reasoning (see European Commission: Decision (EU) 2016/2326 of 21 October 2015 on State aid SA.38375 [2014/C ex 2014/NN] which Luxembourg granted to Fiat, [2015] O.J. L 351; Decision (EU) 2017/502 of 21 October 2015 on State Aid SA.38374 [2014/C ex 2014/NN] implemented by the Netherlands to Starbucks [2017] OJ L 83/38; Decision (EU) 2016/1699 of 11 January 2016 on the excess profit exemption State aid scheme SA.37667 [2015/C], O.J. L 260/61; Decision (EU) 2018/859 of 4 October 2017 on State Aid SA.38944 [2014/C] [ex 2014/NN] implemented by Luxembourg to Amazon [2018] OJ L 153/1).

In particular, the backbone of the Commission’s legal reasoning that led to the qualification of all those advance tax rulings as illegal State aids under Art. 107 TFEU was based on the following two main arguments:

  1. That the rulings granted an economic advantage, since, by allowing an intra-group allocation of profits that deviated from an ALP allegedly derived from EU law (specifically, Art. 107 TFEU), the national taxable base and tax burden of the group was lower than it would have been under a “normal” treatment (i.e., absent the advance tax rulings at issue);

  2. That an economic advantage provided to multinational enterprises, and not available to standalone undertakings, would, by nature, also be selective.

Both these arguments have been largely discussed and criticized by scholars and other experts and commentators. The first argument has been criticized for using the ALP as a de facto counterfactual against which the existence of an advantage granted to a specific undertaking should be assessed. Indeed, according to the established interpretation of Art. 107(1) TFEU resulting from both ECJ case law and Commission’s practice, an advantage should be assessed based on a reference system constituted by the effective treatment under national law, and not on the basis of approximations resulting from the application of external and abstract standards and criteria (see ECJ, Judgment of 6 October 2021, World Duty Free Group and Spain v Commission, Joint Cases C-51/19 P and C-64/19 P, EU:C:2021:793, para. 62. See also Schön W., Tax Legislation and the Notion of Fiscal Aid – A Review of Five Years of European Jurisprudence, Max Planck Institute for Tax Law and Public Finance Working Paper, 2015). Instead, the Commission concluded that the tax rulings gave rise to an advantage because the application of the transfer pricing methodology endorsed by the Member States’ tax authorities departed from an ALP stemming from the principle of equal treatment which is supposedly immanent to Article 107(1) TFEU (hereinafter “EU ALP”). However, for the practical implementation of this EU ALP, the Commission substantially referred to the OECD standards. And the fact that, at the times the advance rulings were issued, the Member States involved had not incorporated in their domestic tax system the OECD standards aimed at concretely determining the arm’s length value of intragroup transactions and to attribute profits to PEs did not constitute, for the Commission, an impediment to the application of the above-mentioned ALP. Therefore, the Commission’s assessment of the advantage requirement seems to be based on the application of an abstract and external criterion, rather than on national legislation and administrative practice (see Kyriazis D., From Soft Law to Soft Law Through Hard Law: The Commission’s Approach to the State Aid Assessment of Tax Rulings, 15 Eur. St. Aid L. Q. 3, 428-439, 2016; Peeters C., Critical Analysis of the General Court’s EU Arm’s Length Tool: Beware of the Reflexivity of Transfer Pricing Law! in EC Tax Review 1, 30, 2022; Allevato G., The Commission’s State Aid Decisions on Advance Tax Rulings: Criticisms and Potential Impact on the Future of Direct Taxation within the European Union, in Almudí J.M. – Ferreras Gutiérrez J.A & Hernández González-Barreda P. (eds.), Combating Tax Avoidance in the European Union: Harmonization and Cooperation in Direct Taxation, Amsterdam, 2018, at 490-92; Judicial Review of the State Aid Decisions on Advance Tax Rulings: A Final Opportunity to Safeguard the Rule of Law, in 62 European Taxation 2, 86-94, 2022; Rossi-Maccanico P., Fiscal State Aids, Tax Base Erosion and Profit Shifting, in 24 EC Tax Rev. 2, 73-77, 2015; Lyal R., Transfer Pricing Rules and State Aid, in 38 Fordham Intl. Law J., 4, 1041, 2015; Traversa E. & Flamini A., Fighting Harmful Tax Competition through EU State Aid Law: Will the Hardening of Soft Law Suffice? 14 Eur. St. Aid L. Q. 3, 330, 2015; Moreno González V.S., State Aid and Tax Competition: Comments on the European Commission’s Decisions on Transfer Pricing Rulings, in 15 Eur. St. Aid L. Q. 4, 556-574, 2016). The second pillar of the Commission’s reasoning has been criticized for conflating the advantage and the selectivity requirements, and therefore circumventing the need to separately assess the ‘selective’ nature of the tax measure under investigation, as prescribed by the established interpretation – by the ECJ and the Commission itself – of Art. 107 TFEU (see Lovdahl Gormsen L. & Mifsud-Bonnici C., Legitimate Expectation of Consistent Interpretation of EU State Aid Law: Recovery in State Aid Cases Involving Advanced Pricing Agreements on Tax, in 8 Journal of European Competition Law & Practice 7, 424, 2017; Kyriazis D., supra, at 433 and Allevato G., Judicial Review of the State Aid Decisions on Advance Tax Rulings, supra, at 91).

The Member States and the taxpayers involved filed requests for annulment of the Commission’s decisions at issue. Despite strong arguments raised by the affected corporate taxpayers and Member States, the GC has never disavowed the legal reasoning of the Commission, including in those judgements where it annulled the Commission’s decisions based on certain technicalities related to the implementation of the transfer pricing methodologies (such as Starbucks and, as already illustrated, Apple itself). Subsequently, the GC decisions were the subject of appeal and cross-appeal before the ECJ. The ultimate endorsement, by the ECJ, of the Commission’s legal reasoning would have an unprecedented impact on the evolution of corporate income taxation in the EU. Indeed, in such case, we would have a sort of “harmonization through the back door”, since any national tax measure would be potentially subject to review of its consistency with an EU law-sourced ALP by the Commission, which would, therefore, convert into a super-national tax authority (see Daly S., The Power to Get It Wrong, in 137(2) Law Q. Rev. 280, 2021; Allevato G., Judicial Review of the State Aid Decisions on Advance Tax Rulings, supra at 87). Instead, in the case of ultimate rejection, by the ECJ, of the Commission’s legal reasoning, the effectiveness of the fight, at the EU level, against harmful tax competition and BEPS would be weakened and left to bilateral or multilateral initiatives at the EU, OECD and G20 level, whose development and implementation require long time and negotiations and whose success and effectiveness is not guaranteed, as the recent events related to the International Framework’s Two-Pillar Solution are demonstrating (Allevato G., Judicial Review of the State Aid Decisions on Advance Tax Rulings, supra at 87).

3. As anticipated, in its relevant judgements on the tax ruling cases rendered before Apple, the ECJ had taken an opposite position than the GC, by tearing down the backbone of the Commission’s legal reasoning (see ECJ: Judgement of 8 Nov. 2022, Joined Cases C-885/19 P and C-898/19 P, Fiat Chrysler Finance Europe v Commission, EU:C:2022:859, hereinafter simply “Fiat”; Judgement of 14 Dec. 2023, Case C-457121 P, European Commission v. Gran Dutchy of Luxembourg, Amazon et al, OJ C, C/2024/1066, 5.2.2024,:C:2024, hereinafter simply “Amazon”). After recalling that, as recognized by established case law, “the determination of the reference framework for the purposes of determining whether a selective advantage has been granted by a Member State’s measure must follow from an objective examination of the content, the structure and the specific effects of the applicable rules under the national law of that State” (Fiat, para. 72; Amazon, para. 38), and that compliance with this established principle is even more important in areas falling outside the spheres in which EU tax law has been harmonized, such as income taxation (see Fiat, para 73; Amazon, para 38), the ECJ determined that «parameters and rules external to the national tax system at issue cannot therefore be taken into account in the examination of the existence of a selective tax advantage within the meaning of Article 107(1) TFEU and for the purposes of establishing the tax burden that should normally be borne by an undertaking, unless that national tax system makes explicit reference to them» (Fiat, para 96; Amazon, para 44). Therefore, after noting that Luxembourg had enacted and implemented specific rules on the calculation of transfer prices in the case of group financing companies, such as FFT – specifically, as pointed out by Luxembourg during the investigation leading to the Commission’s decisions, Art. 164(3) of the national Tax Code and the Circular No 164/2 – the Court concluded that «such errors in determining the rules actually applicable under the relevant national law and, therefore, in identifying the ‘normal’ taxation in the light of which the tax ruling at issue had to be assessed necessarily invalidate the entirety of the reasoning relating to the existence of a selective advantage» (Fiat, para 118; Amazon, para 57).

As a result, in Fiat and Amazon, the ECJ established that the only acceptable reference system for assessing whether a tax ruling constitutes a selective advantage under Art. 107 TFEU is what is ultimately stipulated in national law, and not parameters and rules external to it such as an ALP allegedly deriving from Article 107 TFEU itself or from OECD guidelines, unless the latter have been incorporated into national law itself. This is particularly relevant in the cases involving Member States such as Ireland and Luxembourg which, at the time the advance rulings were granted, had not incorporated the OECD standards for the determination of the arm’s length value into their national legislation (see Dourado A.P., The FIAT Case and the Hidden Consequences, in 51 Intertax 1, 4, 2023; Doleman R., In Principle, [Im] possible Harmonizing on EU Arm’s Length Principle, in EC Tax Review, 3, 93, 2023).

It is worth noting that such ECJ’s position with respect to the determination of the reference system was confirmed in another judgement concerning the qualification, by the Commission, of a tax ruling as an illegal state aid in a context other than transfer pricing (ECJ, Judgement of 5 Dec. 2023, Joined Cases C-451/21 and C-454/21 P, Grand Duchy of Luxembourg, Engie Global LNG Holding Sàrl et al v. European Commission, OJEU 29.1.2024, hereinafter “Engie”). Specifically, the case concerned a tax ruling through which the Luxembourg’s tax authorities had substantially allowed a hybrid mismatch scheme between four entities of the Engie group. Indeed, in the tax ruling the Luxembourg’s tax authorities had agreed on the deductibility of interest payments on a convertible loan issued by two Engie group’s entities which was not offset by the taxation of the income in the hands of other two Engie group’s companies that were holding the convertible instruments. In its decision, the Commission assumed that Luxembourg’s national law aimed to tax all the resident entities and not to grant any room for double non-taxation, and that the deductibility of a payment by the payor is subject to the condition that the same item is taxable in the hands of its recipient. Therefore, the ruling, which in fact granted double non-taxation – through deductibility of the interest payments from the taxable base of the debtor and exemption of the capital gain in the hands of the holders of the convertible instruments – constituted an advantage selectively granted to the Engie group. The GC had upheld the Commission’s decision. When the case reached the ECJ, the latter came, instead, to the conclusion that both the Commission and the GC wrongly determined the reference system. Indeed, the ECJ firstly explained that, in tax matters which are not harmonized at EU level, only the national law applicable in the EU Member State concerned must be taken into account in order to identify the reference system; secondly, it highlighted that neither the wording of Luxembourg national law nor the administrative practice of the local tax authorities conditioned the deductibility of the interest payments to the taxation of that amount in the hands of the recipient, as demonstrated by Luxembourg. Furthermore, the ECJ reiterated that the Commission is not allowed to ground its assessment on the assumption of a general objective of taxation of all resident companies.

4. The judgements illustrated above had led to the expectation that, in deciding the Apple case, the ECJ would have confirmed the annulment of the Commission’s decision on an even stronger basis – i.e., the rejection of the key part of the Commission’s legal reasoning concerning the determination of the reference system accordingly to external standards not incorporated into the national tax system. Nevertheless, in Apple, the ECJ substantially limited its analysis to determining whether the Commission had used an “exclusion” approach for the purposes of PE profit attribution, as stated by the GC, or not. Indeed, contrary to what it did in the Fiat and Amazon judgements, the ECJ did not pronounce itself on the use of the ALP as a criterion to assess the existence of a selective advantage. This is due, according to the Court itself, to the fact that neither Apple group nor Ireland had raised that argument before the ECJ. Indeed, the ECJ highlighted that, since the complaints in relation to the reference framework have not been the subject of a cross-appeal, the rejection of such arguments in the GC’s judgement under appeal has the force of res judicata (see Apple, para. 276 and 303). This is the main peculiarity of the ECJ judgement in Apple, because it leads to a schizophrenic result: the ECJ validated a measure whose legitimacy is ultimately based on a reference system that the ECJ itself had considered as not acceptable in previously decided analogous cases. Indeed, to most early commentators, the outcome of the Apple case appears inconsistent. Such concern is well expressed by Stephen Daly’s statement that «we’re a bit lost why Luxembourg won in Fiat and Ireland lost in Apple» (as reported by Peter A., CJEU Reinstates €13 Billion State Aid Decision Against Apple, in Tax Notes, 11 Sept. 2024).

Indeed, besides the recovery of an unprecedented amount of money it ultimately imposed, the ECJ’s judgement in Apple raises the question of whether it will have an impact on the architecture of corporate income taxation in the EU or not. In such regard, four main scenarios seem to emerge. The first and most radical scenario is the one portrayed by Ruth Mason, who stated that «the Court of Justice effectively overruled Fiat and reinstated the commission’s ability to use OECD guidance to second-guess member states’ rulings, regardless of whether the member state adopted that guidance into domestic law» (as reported by Peter A., CJEU Reinstates €13 Billion State Aid Decision Against Apple, supra). According to this position, in Apple the ECJ has substantially changed its established case law by accepting the use of an EU ALP and of supranational standards to determine the reference system, regardless of whether such criteria are effectively incorporated into a national tax system. As a result, the ECJ has ratified the harmonization through the backdoor of national corporate income tax systems and the transformation of the Commission into a supranational tax authority entitled to assess national tax measures, such as advance rulings, and strike them down in case of contrast with the above-mentioned reference system. This reading of the significance of the ECJ’s judgement of Apple is rooted in the argument that the Irish domestic law “corresponded in essence” to the functional and factual analysis conducted under the Authorized OECD Approach to the attribution of profits to the PE (see Apple, para. 123). Obviously, this scenario seems to be in open violation of fundamental principles such as legal certainty and the principle of legitimate expectations: first, because of the contrast with the previously established EU case law; second, because international consensus on the OECD Authorized Approach was reached only in 2010, when the OECD Guidelines on the Attribution of Profits to the Permanent Established were published – that is, almost two decades after the issuance of the first of the two rulings at issue.

The second scenario, instead, would lead to a distinction between cases where the national tax system of the Member State features clear and specific rules arising from legislative and administrative acts – such as in Luxembourg, as demonstrated by its government during the Commission’s investigation and during the litigation before the GC – and other cases where, instead, the national legal framework regulating a matter is rather vague – such as in Ireland at the time the rulings were granted. Only in the latter cases, supranational external standards which are broadly accepted may be exceptionally used to define the reference system. Therefore, Fiat and Amazon would continue to be landmark case law in proceedings involving Member States where the national tax treatment is clearly identifiable and precisely defined, while the relevance of Apple would be limited only to those cases where domestic law is rather vague.

The third scenario would also be based on a case distinction. In such scenario, however, the distinction would be between those rulings concerning “traditional” intra-group transfer pricing – i.e., where the ALP is used to allocate profits between different group entities – and those concerning, instead, the attribution of profits within the same entity (i.e., PE profit attribution). Thus, Apple would be relevant only in the latter cases, while Fiat and Amazon would continue to be landmark case law for all the other cases. This scenario finds its potential rationale in the Commission’s and ECJ’s peculiar and debatable interpretation of the AOA, according to which the allocation of profits must be entirely based on the functions carried out within the same individual entity, with no considerations for significant people functions performed by other group entities (see Collier R., supra).

Under the fourth scenario, Fiat and Amazon would continue to be the only relevant case law. Such scenario assumes that the Apple group and Ireland lost in Apple simply due to a procedural issue. By downgrading Apple to an isolated decision arising from a procedural “accident”, this scenario would grant more stability and coherence to the system. Indeed, under this scenario, the ECJ’s judgement in Apple would not lead to the “harmonization through the back door” that several scholars had originally imputed to the Commission’s decision.

5. Although it is extremely difficult to make reliable predictions on which scenario is more likely to prevail, some elements and recent developments lean towards the fourth scenario described above, according to which the Apple judgment should not have substantial implications on the architecture of EU taxation. Indeed, the ECJ expressly stated that it could not pronounce itself on the legitimacy and appropriateness of the reference system adopted by the Commission and confirmed by the GC, due to the fact that such issue had not been the subject of cross-appeal by either Ireland or Apple and therefore – regardless of whether the Court agreed with that or not – it had to be considered res judicata. Therefore, due to procedural constraints. the ECJ simply did not have the possibility to reject the reference system applied by the Commission and replace it with a reference system consistent with the criteria established in Fiat and Amazon (and Engie). The idea that the Court did not intend to take any novel position on the criteria to determine the reference system for the purposes of assessing the existence of a selective advantage under Art. 107 TFEU appears to be further strengthen by the ECJ’s judgement, issued on September 19 (i.e., after the Apple judgement), in the UK CFC State aid case (see ECJ, Decision of 19 Sept. 2024, Joined Cases C-555/22 P, C-556/22 P and C-564/22, United Kingdom et al v. P, LSEGH et al, hereinafter simply “UK CFC”). Although this case did not deal with intra-group transfer pricing, it provided the opportunity to the ECJ to stress again that, also in tax cases, the determination of the reference framework for the purpose of applying Art. 107 TFEU should be based only on national law as interpreted and implemented by local authorities. Furthermore, in developing this argument, the ECJ made extensive reference to its judgements in Fiat and Engie, thereby demonstrating that it continues to consider them the relevant case law with respect to the determination of the reference system for State aid purposes in non-harmonized tax matters (see UK CFC, para. 97-98).

Those two elements – the ECJ’s qualification of the reference system in Apple as res judicata and the subsequent UK CFC judgement –go, in fact, in the same direction of confining the Apple judgement to a case of mistaken litigation strategy (or unlucky litigation strategy, considering that the judgements in Fiat, Amazon and Engie were issued after the deadline to submit the appeals and cross-appeals in Apple had passed), rather than attributing to it the status of landmark case law in tax ruling State aid cases. Of course, the Apple judgement will continue to have a significant relevance due to the assessed amount of aid and to the taxpayer involved, but the moral of the story seems to be “cross-appeal everything”, rather than “harmonization through the backdoor”.

(*) Il saggio è stato sottoposto a double blind peer review con valutazione positiva. Esso confluirà nel fascicolo n. 2/2024 (semestrale) della Rivista telematica di diritto tributario.

(**) This paper was partially funded by MCIU/AEI/10.13039/501100011033/FEDER, EU Grant No. PID2023-149184OB-C43.

REFERENCES

Allevato G., Judicial Review of the State Aid Decisions on Advance Tax Rulings: A Final Opportunity to Safeguard the Rule of Law, in 62 European Taxation 2, 86-94, 2022

Allevato G., The Commission’s State Aid Decisions on Advance Tax Rulings: Criticisms and Potential Impact on the Future of Direct Taxation within the European Union, in Almudí J.M. – Ferreras Gutiérrez J.A & Hernández González-Barreda P. (eds.), Combating Tax Avoidance in the European Union: Harmonization and Cooperation in Direct Taxation, Amsterdam, 2018

Collier R., A Bad Apple Ruling, in Oxford Business Tax Blog, 13 Sept. 2024

Daly S., Another take on the (bad) Apple Ruling: is a misapplication of domestic law enough for a finding of State Aid? in Oxford Business Tax Blog, 17 Sept. 2024

Daly S., The Power to Get It Wrong, in 137(2) Law Q. Rev. 280, 2021

Doleman R., In Principle, (Im) possible Harmonizing on EU Arm’s Length Principle, in EC Tax Review 3, 93, 2023

Dourado A.P., The FIAT Case and the Hidden Consequences, in 51 Intertax 1, 4, 2023

Kyriazis D., From Soft Law to Soft Law Through Hard Law: The Commission’s Approach to the State Aid Assessment of Tax Rulings, 15 Eur. St. Aid L. Q. 3, 428 -439, 2016

Lovdahl Gormsen L. & Mifsud-Bonnici C., Legitimate Expectation of Consistent Interpretation of EU State Aid Law: Recovery in State Aid Cases Involving Advanced Pricing Agreements on Tax, in 8 Journal of European Competition Law & Practice 7, 424, 2017

Lyal R., Transfer Pricing Rules and State Aid, in 38 Fordham Intl. Law J. 4, 1041, 2015

Miceli R., Il divieto di aiusti di Stato nel contrasto ai rulings fiscali. Limiti e opportunità, in Riv. dir. trib. int., 1, 71, 2021

Moreno González V.S., State Aid and Tax Competition: Comments on the European Commission’s Decisions on Transfer Pricing Rulings, in 15 Eur. St. Aid L. Q. 4, 556-574, 2016

Parada L., Between Apple and Oranges: The EU General Court’s Decision in the ‘Apple Case’, in EC Tax Review, 2, 55, 2022

Peeters C., Critical Analysis of the General Court’s EU Arm’s Length Tool: Beware of the Reflexivity of Transfer Pricing Law! in EC Tax Review 1, 30, 2022

Peter A., CJEU Reinstates €13 Billion State Aid Decision Against Apple, in Tax Notes, 11 Sept. 2024

Rossi-Maccanico P., Fiscal State Aids, Tax Base Erosion and Profit Shifting, in 24 EC Tax Rev. 2, 73-77, 2015

Schön W., Tax Legislation and the Notion of Fiscal Aid – A Review of Five Years of European Jurisprudence, Max Planck Institute for Tax Law and Public Finance Working Paper, 2015

Sheppard L., EU Extortion in Apple, in Tax Notes, 23 Sept. 2024

Traversa E. & Flamini A., Fighting Harmful Tax Competition through EU State Aid Law: Will the Hardening of Soft Law Suffice? 14 Eur. St. Aid L. Q. 3, 330, 2015

Scarica il commento in formato pdf

Tag:, , , , , , , ,