The summaries of the pleas of the actions for annulment lodged by Luxembourg and the Netherlands in the Fiat and Starbucks cases respectively were recently published (see here and here). Moreover, the pleas of Fiat in its own action for annulment against the Commission’s decision were also published (see here).
These pleas are important because they provide a glimpse into the Commission’s reasoning in its final decisions, which were adopted in October 2015 but have yet to be published. Naturally, there is not much that can be said with certainty, since the applicants are arguing their case and are not necessarily presenting an undistorted account of the Commission’s findings. Still, the comparative analysis of the three aforementioned actions does provide some interesting insights.
The first takeaway is that all applicants are focusing on the Commission’s assessment of the two key State aid conditions, i.e. advantage and selectivity. There are hints in the Dutch and Fiat actions that the Commission has stuck to its approach in the opening decisions where it conflated the advantage and selectivity conditions and instead only examined the existence of a “selective advantage”. More specifically, the Netherlands is arguing that ‘the Commission did not adequately — and separately — demonstrate that the selectivity criterion was fulfilled’, while Fiat is arguing that the Commission misapplied the concept of “selective advantage”.
These assertions, if true, could cause the Commission severe trouble before the General Court, especially given that the ECJ recently reaffirmed that ‘the requirement as to selectivity under Article 107(1) TFEU must be clearly distinguished from the concomitant detection of an economic advantage, in that, where the Commission has identified an advantage, understood in a broad sense, as arising directly or indirectly from a particular measure, it is also required to establish that that advantage specifically benefits one or more undertakings’ (Case C-15/14P, Commission v MOL para 59). The requirement for a separate analysis of these two conditions is not undermined by the ECJ’s subsequent dictum in paragraph 60 of the same judgment that in cases of individual aid ‘the identification of the economic advantage is, in principle, sufficient to support the presumption that it is selective’ since this does not mean that the two conditions become one, but merely that in principle, i.e. not always, the fulfilment of the first one creates a (rebuttable) presumption that the (separate) selectivity condition is also fulfilled. This is thus one of the points to keep in mind when going through the long-awaited final decisions.
As regards selectivity per se, Luxembourg reproaches the Commission for not adducing adequate proof in relation to the selective character of Fiat’s tax ruling, while the Netherlands disputes the Commission’s choice of reference framework. More specifically, the Dutch position is that the relevant reference framework should not be the general Netherlands corporation system but a combination of the “Law on Corporation Tax” and the “Settlement Price Decree”, from which the Starbucks’ APA does not depart and is thus not selective.
Objections to the Commission’s selectivity analysis are unsurprising, especially if the Commission’s analysis is as terse as it was in its opening decisions, where it claimed – for instance in the Fiat opening decision – that ‘to the extent the Luxembourgish authorities have deviated from the arm’s length principle as regards the FFT APA, the measure should also be considered selective’ (para 82).
Moving on, in the pleas of the Netherlands and Luxembourg there is no explicit mention of any OECD document, which could mean that the Commission has ceased to refer to the OECD Transfer Pricing Guidelines as “appropriate guidance”, thus introducing its own definition of the Arm’s Length Principle (ALP). This assumption seems to be confirmed by Fiat’s fourth plea in law, where it alleges that the Commission’s decision ‘breaches the principle of legitimate expectations since the Commission has created a legitimate expectation that for state aid purposes it assesses transfer pricing arrangements on the basis of the OECD Guidelines and its sudden departure from this has breached the principle of legitimate expectations.’
Although this departure, if true, might shield the Commission from criticism that it is turning soft law (OECD documents) into hard law (Article 107 TFEU), it certainly does not come without its drawbacks, since the Commission will have to show that its very own “EU ALP” is part of primary EU law, even though it is doubtful whether the ECJ has ever endorsed it (especially to the extent claimed by the Commission when it tries to generalise the findings of Forum 187). In the same vein, Fiat touches the heart of the matter when it accuses the Commission of failing to show ‘how it derives the arm’s length principle from Union law, or even what the principle is’. These are harsh words, and a similar argument is put forward by the Netherlands in an even more unequivocal manner, where it argues that ‘there is no arm’s length principle in EU law and that that principle is not part of a State aid assessment.’
Surprisingly enough, Luxembourg is also seeking the annulment of the Fiat Decision because the Commission did not adduce proof of a “restriction of competition”. Luxembourg is obviously referring to the “distortion of competition” condition, a condition that is notoriously easily fulfilled. Still, if the Commission was too complacent and relied on very general or vague reasoning, then Luxembourg could possibly rely on the Court’s case law (most notably Case C-494/06P, Italy and Wam) to challenge the Commission’s findings. The fact that Fiat also argues that the Commission failed to show that its APA was liable to distort competition is, perhaps, a further indication of possible complacency in this regard.
I am afraid nothing more can be said on the Fiat and Starbucks cases until the two Commission decisions are published – only then will we be able to evaluate the merits of the three actions for annulment. The ball is now in the General Court’s court (pun intended), which, apart from clarifying the European Courts’ case law, could also be required to grapple with the nitty-gritty of transfer pricing.
In any case, the judgments of the General Court in these cases will have far-reaching consequences. Should the Court reject one of the Commission’s main arguments, most notably its assertion that a deviation from the ALP confers a “selective advantage” on the recipient, then most probably all its final decisions will be annulled, since they are based on the same doctrinal “pillars”. If, moreover, the ECJ does not vindicate its approach on appeal, the Commission’s use of the State aid apparatus to crack down on tax avoidance will have failed dramatically.