In an era of globalisation and digitalisation, complex issues may arise as regards the laws applicable to situations that either involve cross-border elements or occur in cyberspace. The disconnect between the borderless nature of modern trade and the essentially domestic character of economic laws may give rise to complex issues of jurisdiction.
Rules and principles of private international law often determine which national laws are applicable in any given situation. However, no equivalent system of binding rules exists as far as the public enforcement of antitrust rules is concerned. As a result, the same business conduct may potentially fall within the jurisdiction of a number of States, each having its own rules. Clearly, problems may not only arise when those rules are dissimilar (or interpreted and applied differently), but also when various jurisdictions with identical rules are applied cumulatively to the same conduct.
This piece, available here, reviews European jurisprudence concerning the reach and scope of the EU’s competition laws. It does so as follows:
Section 2 briefly looks at earlier ECJ’s case law on the jurisdictional reach of EU’s competition law.
According to settled case-law, EU law must be interpreted in conformity with public international law, which requires that, to establish jurisdiction, a State must be able to identify a sufficient nexus (or genuine connection) between itself and the conduct on which it asserts regulatory and/or enforcement powers. On this basis, the EU Courts have consistently held that the application of EU law presupposes an adequate link to the EU territory.
Nevertheless, until recently, only a few decisions of the EU Courts specifically concerned the scope of application of EU competition rules. In past cases, the Court of Justice resorted to the single economic entity doctrine (ICI), or referred to the implementation of anticompetitive conduct within the European Union (Woodpulp I). In addition, the General Court had also endorsed a ‘qualified effects’ test with regard to the applicability of the EU merger regulation (Gencor).
Section 3 looks at recent developments.
The situation has known two important recent developments. The first concerned the calculation of fines imposed by the European Commission in the LCD cartel, which was discussed in the piece above. The Commission identified three categories of sales of the cartelised product: ‘direct EEA sales’ (sales of panels to a third party within the EEA), ‘direct EEA sales through transformed products’ (sales of panels, incorporated outside the EEA into a finished product by a vertically integrated company, to a third party within the EEA); and ‘indirect sales’ (sales of panels to a third party outside the EEA, which incorporates them into finished products sold within the EEA). The Commission took the view that, to calculate the amount of the fines, it needed to examine only the first two categories.
Inno-Lux appealed this on the grounds that ‘direct EEA sales through transformed products’ should not be used as the basis to calculate fines in the EU, but the General Court found that the Commission was entitled to do so. Inno-Lux made a further appeal to the Court of Justice. In his Opinion, AG Wathelet found that taking into account direct EEA sales through transformed products entailed the taking into account of (internal) sales of the product concerned that were made entirely outside the EEA. In his view, this did not amount to ‘implementation’ in the EEA within the meaning of Woodpulp I. On the other hand, it could be said that the sale in the EEA of products manufactured and sold outside the EEA could create ‘qualified effects’ in the EEA, and hence trigger the Commission’s jurisdictions. However, the AG concluded that no specific, credible and adequate evidence had been provided on this issue, and thus he considered that jurisdiction was not triggered in this respect. The ECJ, however, found Inno-Lux’s arguments to be irrelevant –it was undisputed that the Commission had jurisdiction. The question concerned the calculation of the fine, and, in that respect, jurisdictional matters were not relevant. The Commission was entitled – in order to ensure that the amount of the fine reflected the economic importance of the infringement and the relative weight of the undertaking concerned in the infringement—to take into account the sales of finished products by InnoLux in the EEA to independent third parties, even if the internal sales of the goods concerned were made outside the EEA.
The second development was the Intel judgment (which I discussed extensively elsewhere, even if the focus then was on the appropriate treatment of rebates and the relevance of the as-effective-competitor test). In short, the Commission found that Intel had abused its dominant position by implementing a strategy aimed at foreclosing AMD from the market for a specific type of microprocessors (x86 CPUs). Intel challenged that decision before the General Court, inter alia on the ground that the Commission lacked jurisdiction to apply EU competition rules to certain parts of its conduct. The court considered that two alternative criteria may be applied to assert jurisdiction: implementation and qualified effects. In the court’s view, the application of those criteria led, in the case at hand, to the same conclusion: the Commission had jurisdiction with respect to Intel’s conduct in its entirety.
On appeal to the ECJ, Intel argued that two of its agreements (with Lenovo, a Chinese OEM) had no effect in Europe. The appeal raised two important issues: what was the correct standard to establish jurisdiction, and how to apply it. According to AG Wahl, EU competition provisions are concerned with conduct that has anticompetitive effects within the internal market. Accordingly, ‘implementation’ is a valid jurisdictional criterion for EU competition rules: conduct implemented in the Union may be assumed to generally produce effects within the internal market. A conduct can be considered to be ‘implemented in the Union’ when one of the essential constituent elements of the anticompetitive conduct takes place within the EU territory. Whether that is so depends mainly on the nature, form and scope of the conduct in question. In addition, the AG suggested that a ‘qualified effects’ criterion can also be used to establish jurisdiction, even if this criterion should be used with a certain amount of restraint to reflect the multiplicity of competition regimes in the world. The application of EU competition rules to some specific conduct should only occur when that conduct has foreseeable, immediate and substantial effects in the Union. Accordingly, the criterion of ‘qualified’ effects is not satisfied where, for example, the effect in the Union is merely hypothetical or, in any event, of minor significance.
The Court of Justice endorsed this analytical framework and held that the qualified effects criterion pursues the same objective of the implementation criterion: to identify conduct which, while not adopted within the Union, has anticompetitive effects liable to have an impact on the EU market. The court emphasised, however, that in order to determine whether the Commission has the necessary jurisdiction to apply EU competition law in a given case, it is necessary to examine the conduct of the undertaking in question ‘viewed as a whole’. A different conclusion would lead, according to the Court, to an artificial fragmentation of anticompetitive conduct capable of affecting the market structure in Europe into a collection of separate forms of conduct which might escape the EU’s jurisdiction. In the case at hand, the Court found that Intel’s conduct vis-à-vis Lenovo formed part of an overall strategy intended to ensure that no Lenovo notebook computer equipped with an AMD CPU would be available on the market, including in the Union.
Section 4 provides a comment on these developments.
In Intel, the Court of Justice expressly dealt with the jurisdictional boundaries of EU competition rules. On the one hand, the judgment is clear and unambiguous in finally embracing a qualified effect test, aligning the EU with the US. On the other hand, however, the judgment is somewhat less clear regarding the application of the implementation and qualified effects tests. In essence, the Court limits itself to bundle together different forms of conduct by Intel and to find that some of those had qualified effects in the Union. It followed from this that Intel’s arguments concerning an alleged lack of jurisdiction were dismissed. More importantly, it is yet to be clarified how strictly the qualified effects criterion is to be applied in other cases, since the court’s approach was very fact-specific and involved identifying an overall strategy which clearly encompassed the EU. It is unclear how jurisdiction will be established if the sanctioned conduct had a less expansive scope.
It is also unclear how the implementation and qualified-effects tests relate to one another after Intel. Is, as suggested by the Advocate General, the implementation test a subspecies of the qualified effects’ test; or, as decided by the General Court, is it an alternative to that test? Would there be any need to refer to the implementation criterion in the future, given that the qualified effects criterion seems broader and easier to meet? These doubts seem all the more justified since, admittedly, the distinction between effects and implementation is not always obvious, and the implementation criterion seems to be unknown in other jurisdictions.
Section 5 concludes.
The criteria traditionally employed to determine a State’s jurisdiction—such as personal and, most of all, territorial criteria—appear increasingly inadequate when dealing with today’s global phenomena and modern business dynamics. The Permanent Court of International Justice considered some form of extraterritorial application of laws permissible as early as 1927. In 2019, it is unrealistic to expect States to adopt a strictly territorial approach when it comes to protecting their citizens’ welfare against companies’ anticompetitive practices. This is apparent in how several countries nowadays follow an effect-based approach to determine the application of their antitrust laws. Furthermore, thanks also to bilateral channels of dialogue and international fora, current cross-border antitrust enforcement reflects increased cooperation.
Yet, international cooperation (both bilateral and multilateral) has its limits, in particular because of the absence of binding rules and strong international institutions endowed with powers of enforcement. If some degree of extraterritoriality in the application of antitrust rules is by now largely accepted (or at least tolerated) by the international community, there is no general consensus about the conditions under which, and limits within which, that application is permissible. It is not even clear, for example, that all countries that adopt an effects-based approach to jurisdiction interpret and apply that concept similarly. A balanced approach on the issue of jurisdiction is thus, for both legal and political reasons, as appropriate today as it was in the past. The judgment in Intel may be regarded as a step—and a significant one—in the Court of Justice’s quest for such a balanced approach. However, because of its brevity on the point, the EU Courts will most likely be called upon, in future cases, to clarify and refine their case-law on such a key and sensitive issue.
This is a crisp and succinct analysis of where we stand on the jurisdictional reach of EU competition law – perhaps unexpectedly, given that the author has worked both as a referendaire at the ECJ and for the Commission’s legal service. It does not go much further than that – e.g. the piece does not tell us what the author thinks should be the courts’ approach to jurisdiction in greater detail, whether there are problems with the approach which has been adopted, or how the European approach compares to others – but, as someone who is limited in what I can say because I am an international civil servant, I am very sympathetic to the author’s predicament.