Chillin'Competition

Yesterday the General Court handed down the first and much awaited Judgments regarding the assessment of individual tax rulings under State aid. The Court roasted the Starbucks decision, but the Commission managed to drive home the Fiat one.

These Judgments raise complex issues and fit within a complex line of case law. Those interested in a full overview of these should attend a forthcoming presentation at the ULB’s Mardis de la Concurrence (more information here; incidentally, I’ll be the speaker).

But here is an appetizer. My colleague (and main specialist on the issue) José Luis Buendía, Pablo and myself all immediately devoured the Judgments yesterday and spoke to several media outlets about them. The text below contains a collage of our views:

General comments

The Judgments are carefully crafted in their content, outcome and intended message. Not much of a sexy headline, but the takeaway is that tax rulings may, or may not, amount to State aid: the Commission needs to do a thorough job, and the Court has demonstrated its willingness to exercise a strict control. In Pablo’s words “proceed – but with caution, because the court is watching.”

As I told Politico yesterday, “the Court has not challenged the Commission’s policy as a matter of principle, but has sent a clear message that the devil lies in the details of each specific ruling”. This complicates predictions (also as to the outcome of the Apple case, which moreover raises different issues) and arguably legal certainty.

The Judgments are nuanced on the law, avoiding some of the thorniest legal issues (notably the questions related to selectivity, which are effectively bypassed by focusing virtually all of the debate on the requirement of “advantage”). The Judgments are also fact-heavy, which might minimize the chances of success of possible appeals on either side.

Politically, the credibility of the Court and of the EU system may be reinforced by the fact that – fortuitously –  it was only the US company (and the Netherlands) getting out of the hook, not the European one (and Luxembourg).

In our view, the key to understand the opposite outcomes of the Judgments is the Court’s insistence in taking the Commission’s burden of proof seriously. The Commission needs to establish the existence of an advantage, without shortcuts, however tempting.

Advantage 

The GC confirms that the Commission was entitled to rely on the arm’s length principle (“ALP”) as a tool for screening whether a given tax measure is in line with market conditions. The Court seems cautious (just like the Commission itself) not to extract a general ALP principle directly from EU Law, but it accepts that it may be relied upon as a tool. The Judgments seem to be premised on the “nationalization” of the ALP; i.e., on the understanding that both the Dutch and Luxemburguese legal systems had – even if implicitly – incorporated this principle.

Given their recognition at the national level, the Commission was entitled to rely on this principle tool to check whether the specific tax rulings had accepted transfer pricing arrangements departing from normal market conditions (in the sense of Art. 107) and, therefore, departing from what had been the normal taxation of these situations without the tax rulings. Our view is that the Court did not endorse the far-reaching Commission theory that the ALP originates directly from EU law and would apply regardless of its recognition at the Member State level. One could argue that the “nationalization” of the ALP may have required an interpretation of the decision and of national laws that goes beyond the obvious.

Even if the Commission was entitled to apply the ALP in both cases, the judgement underlines that the ALP admits different methodologies, as the OECD guidelines show. Member States thus have a certain choice between them, and the Commission’s assessment ALP must therefore recognize the existence of a certain margin of possible inaccuracies. Crucially, the GC notes that the choice of a particular methodology, in and of itself, is beyond reproach. Thus, the mere fact that the Member State has relied on a method different from the one preferred by the EC does not automatically create a presumption of advantage.

Very much on the contrary, both judgements underline that it is the Commission that has the burden of proving the presence of an “advantage”, that this exercise may be demanding (as exemplified in Starbucks), but nevertheless feasible (as shown in Fiat). It is therefore not for the Member State or the beneficiaries to prove the absence of an advantage, but up to the Commission to prove its existence.

In fact, a too proactive defense by the Member State might even backfire in some cases, by providing the Commission with an analysis that was originally absent. In Fiat the Court validated the reasons offered by the Commission to question the methodology used by Dutch authorities. In Starbucks, it has ruled that non-compliance with methodological requirements cannot be used as a shortcut to find an advantage. Paradoxically, this may require the Commission to do a greater job when Member States do worse. In other words, the Commission cannot simply assume that because the Member State acted in a seemingly arbitrary manner the outcome was wrong.  As Vice President van der Woude noted at the Apple hearing last week, one cannot fully exclude that there may have been a happy coincidence. The message, again, is that nothing is evident and that the Commission cannot take things for granted.

Selectivity

The General Court clearly avoided discussing “selectivity” to the extent possible. The Starbucks Judgment did not need to go there due to the finding that there was no advantage. The Fiat Judgment did require a selectivity assessment, but there the Court found its own shortcut.

Indeed, the Fiat Judgment endorses a  “presumption” of selectivity for cases concerning individual aid (not in cases related to aid schemes). The tax rulings at stake look very much as individual aid. The GC considers that they are not the product of an “scheme” with a predetermined content but rather that they are tailor made by the administration applying discretionary powers. Therefore, if the Commission can prove that the individual tax ruling is advantageous (like in Fiat), it does not need to prove in addition that it is also selective.  This is automatically presumed. The existence of such presumption in EU Law is, in our view, not totally clear. This could therefore be a point of law that the applicants could raise in a possible appeal.

Perhaps because the existence of the said “presumption” of selectivity is far from obvious, the General Court also applied – just in case – the “3 steps test” in order to confirm the selective character of the Fiat tax ruling. The Judgment claims that  ruling would constitute a “derogation”, irrespective of how one were to define the reference system (only the treatment of companies belonging to a group vs those and also stand-alone companies). However, our impression is that the judgement does not entirely justify that claim. Instead, it equals the already established finding of an “advantage” with the finding of a “derogation” for selectivity purposes. This equivalence may be legally questionable and may also offer fertile ground for debate, perhaps on appeal.

Analysis

At the end of the day, the General Court has confirmed that the Commission may indeed use State aid rules against certain tax rulings, but it has also carefully framed its power to do so. It can rely on the ALP (except in the unlikely, and arguably purely theoretical, case that the national legislation explicitly excludes it), but has to recognize a certain margin of “error” to the Member States. Above all, the Commission must prove clearly that the tax ruling confers a real advantage.

The impact on these judgements on the pending tax rulings cases is hard to predict. It would largely depend upon the investigative evidence provided by the Commission in the different decisions. The impact on these judgements on other fiscal cases, dealing with general schemes, might be none.

The Commission may not even file an appeal against the Starbucks Judgment, since the judgement is eminently based on a factual assessment and avoids a taking position on important legal principles. Fiat and Luxembourg may have more incentives to try an appeal, perhaps arguing that the GC has denaturalized national law and raising points of law concerning selectivity. In any case, the chances of success may not be very high.

These judgements will allow the Commission to claim a partial victory and to vindicate its crusade against tax rulings. At the same time, however, they also underscore the need for detailed and thorough investigations. The Commission will no doubt be particularly careful in picking its cases from now on.