On December 21st the European Court of Justice (ECJ) handed down its judgment in the so-called Spanish goodwill cases (Joined Cases C‑20/15P Commission v World Duty Free Group and C‑21/15P Commission v Banco Santander and Santusa).
It is a dense judgment that State aid and tax lawyers will need time to digest and properly “pigeonhole” in the selectivity mosaic. While the decision is a landmark ruling, its pronouncements on selectivity should not, as some have suggested, have any direct impact on the contentious Apple/Fiat/Starbucks tax ruling State aid cases.
These joined cases are the culmination, but not the end, of the Spanish Goodwill saga that started when Spain introduced a tax provision stipulating that, when a company which is taxable in Spain acquires a holding in a foreign company of at least 5% and holds it without interruption for at least one year, the goodwill resulting from that holding can be deducted through amortization. By contrast, when the “target” company was a Spanish company, this provision did not apply.
The Commission found that this amounted to State aid and ordered its recovery in its 2009 (non-Spanish EU shareholdings) and 2011 (non-EU shareholdings) decisions. In 2014, the General Court (GC) annulled both Commission Decisions, finding that the Commission had failed to establish the selective nature of the alleged aid measure.
The Commission appealed the GC’s judgment before the ECJ, and the ECJ has now sided with the Commission in relation to the selectivity issues that had been examined by the GC.
The ECJ’s selectivity standard
The Commission relied on a single ground of appeal made up of two parts. Only the first part will be analysed in this post, since it concerns selectivity and is the most interesting for State aid doctrine in general.
More specifically, the Commission argued that the GC had erred in law by asserting that the fact that the measure derogated from the reference system was not sufficient to establish its selective nature.
In the GC’s view, for the selectivity condition to be satisfied, it is always necessary that a particular category of undertakings be identified that are exclusively favoured by the measure concerned and that can be distinguished by reason of specific properties common to them and characteristic of them. If that is not possible, then the measure is effectively open to all undertakings and thus not selective.
The ECJ refused to accept the GC’s reasoning and found that it had erred in law by inferring such a ‘supplementary requirement’ from the case law; in doing so, the ECJ also sided with the non-binding opinion (recommendation) of Advocate General Wathelet earlier this year.
In the ECJ’s view, all that matters in the context of the selectivity exercise is that the measure, ‘irrespective of its form or the legislative means used, should have the effect of placing the recipient undertakings in a position that is more favourable than that of other undertakings, although all those undertakings are in a comparable factual and legal situation in the light of the objective pursued by the tax system concerned’ (para 79). The fact that the number of potential beneficiaries is large and the criteria for the measure’s application are lax is not relevant according to the ECJ: the measure is still selective.
From the above it is obvious that the ECJ endorsed the “expansive” interpretation of selectivity that the Commission had advanced, at the same time rejecting the GC’s more nuanced and “restrictive” interpretation.
It is certain that many articles will be written about how persuasive the ECJ’s interpretation of its previous case law was, most notably the Gibraltar and Adria Wien rulings. For the moment, suffice it to ponder whether this interpretation of selectivity, the key condition in fiscal State aid cases, is overly broad and counterintuitive, given that it will by definition capture measures from the application of which no undertaking is de facto excluded.
Even though every undertaking could in practice take advantage of the Spanish goodwill measure, since every undertaking could acquire 5% of a non-Spanish company and hold it for a year, this measure has now been found to be selective because Spain chose not to include Spanish companies as target companies. This formulation of the selectivity test and the de facto rejection of the concept of “derogations/exceptions of general application” means that the chasm between reality and State aid reality deepens.
The importance of drawing sensible limitations to the concept of selectivity was eloquently expressed by Advocate General Kokott in 2015 in Case C-66/14 Linz. AG Kokott had made it clear that if a fiscal “provision concerns neither one or more individually identifiable sectors capable of being defined by reference to their economic activity, nor individually identifiable undertakings, as the wording of Article 107(1) TFEU requires, then the provision in question cannot in principle be assumed to be selective.”
AG Kokott’s conclusion was derived not only from her interpretation of ECJ case law, but also from her concern that “too broad an understanding of the selectivity of national provisions […] harbours the risk of adversely affecting the division of competences between the Member States and the European Union.”’ It seems that arguments of this sort hold little sway over the Court of Justice.
Still, the Spanish Goodwill saga is not over yet. The ECJ decided to refer the case back to the GC, for the latter to examine the remaining three pleas of the alleged aid beneficiaries. Even though the (doctrinal) selectivity question has now been settled, the applicants can therefore still prevail.
Apple/Fiat/Starbucks tax ruling cases
A final question that seems to have emerged is whether the ECJ’s judgment in the Spanish Goodwill cases is bad news for Apple, Fiat, Starbucks etc.
In my view, there is no direct link between these two sets of cases, and the ECJ’s ruling in no way prejudges the outcome of the actions for annulment in the tax ruling cases. Not only are the facts very different, with the former concerning a tax scheme and the latter allegedly individual aid, but the Spanish Goodwill ruling did not examine, since it was not relevant, the most contentious legal argument advanced by the Commission in the Apple/Fiat/Starbucks cases, namely that EU State aid rules require the compliance of Member States’ tax rulings with an independent, EU Law, arm’s length principle.
Moreover, as regards selectivity, in the tax ruling cases the Commission is largely relying on the “MOL presumption,” namely, on the ECJ’s assertion in Case C-15/14P MOL 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 the Spanish Goodwill measure was an aid scheme and not an individual aid measure, the ECJ did not get a chance to clarify the limits of the MOL presumption.
The ECJ’s Spanish Goodwill judgment will no doubt be studied for a long time to come and eventually enter the “pantheon” of State aid judgments together with PreussenElektra, Altmark, Azores, Gibraltar et alia. As with all landmark rulings, its legacy is already proving controversial.