Otis: Another Great Judgment on Private Enforcement from the CJEU… But It Could be Better

By Miguel Sousa Ferro1 & Guilherme Oliveira e Costa2



2019 was coming to its end when the Court of Justice of the European Union (“CJEU”) issued its judgment in Otis3 – another judgment in the escalators and lifts cartel saga, making it the third landmark case on private enforcement during that year, alongside Skanska4 and Cogeco5 (plus the less memorable Tibor Trans case6).

In Otis, the CJEU, agreeing with AG Kokott’s Opinion7 (which is essential reading for a complete understanding of what is at stake), delivered another great judgment reinforcing private enforcement of EU competition law and protecting the effectiveness of Articles 101 and 102 TFEU. It is increasingly clear that, among the EU institutions, the CJEU is the greatest paladin of private enforcement, using the referral mechanism to protect, develop and foster this type of competition law enforcement and to partly Europeanize private law applicable in this context. However, we argue the Luxembourg judges should ponder reviewing their way of replying to these preliminary rulings, in particular by giving clearer and more precise guidelines to the national judges.

The Otis Judgment

As is well-known, in 2007, the European Commission applied fines to several undertakings in the market for the installation and maintenance of escalators and lifts in Belgium, Germany, Luxembourg and the Netherlands. In 2008, many of those undertakings were also fined in Austria for a similar cartel in that domestic market. Numerous entities lodged follow-on claims in order to be compensated for the loss caused to them by the cartel. One such entity was the Province of Upper Austria, a public entity which granted legally mandated promotional loans for the financing of building projects, allowing the beneficiary of those loans to obtain funding at better conditions than market rates. The Province of Upper Austria claimed that the costs connected with the installation of lifts, included in the overall building costs paid by those beneficiaries, were increased as a result of the cartel at issue. This resulted in that entity being obliged to grant loans in higher amounts. If the cartel at issue had not existed, the Province of Upper Austria would have granted smaller loans and it could have invested the difference at the average interest rate of federal loans.

The nature of the damage claimed, together with the fact that the claimant was active neither on the relevant market, nor on any upstream or downstream market, led to contrary opinions in the Austrian national courts8 and, consequently, to this preliminary ruling where, in essence, the Austrian Supreme Court asked the CJEU if, under Article 101 TFEU, a person such as the Province of Upper Austria – with its specificities and absence from markets directly or indirectly affected by the cartel – could seek damages from the cartel members.9

The CJEU concluded that if the losses are causally linked to an infringement of Article 101 TFEU and if those were suffered “as a result of the fact that, since the amount of those subsidies was higher than what it would have been without that cartel, those persons were unable to use that difference more profitably,”10 compensation could be sought.

This means, in other words, that there is a right to damages, independently of the market on which the injured person is active, and even if the injured person is not active on any market related to the one affected by the cartel, as long as there is a causal link between the damage and the infringement of Article 101 TFEU.

What Is Otis Really About and Why It Could Be Better

A superficial interpretation of this judgment could lead to the idea that Otis’s main implication is related to the right to damages in general, and, in particular, to the right to compensation for indirect damages suffered in upstream or downstream markets (even though this was not the kind of damage which was at stake). Of course, the judgment is important in that regard, deepening the line of case law initiated in Kone,11 and continuing to widen the scope of the right to damages based on infringements of Articles 101 and 102 TFEU, deepening their deterrent effect. But this is just the onion’s first layer.

Otis has deeper and more significant implications for the private enforcement system as a whole. Firstly, it reaffirmed that the requisites of the right to compensation based on Article 101 (and 102) TFEU are dictated by EU law, while the detailed rules governing the exercise of that right are a matter of national law, which must ensure the effectiveness of said right.

Secondly, and similarly to what happened in Skanska, the CJEU gave another example of something which is a requisite for the existence of the right to damages, rather than a detailed rule governing the exercise of that right. As highlighted by AG Kokott, in this case, the touchstone was who has legitimacy, under Article 101 TFEU, to claim compensation, and for which injury.12 These are requisites of the right to damages, since its very existence depends on their verification. In this regard, it is also worth noting the examples given by AG Kokott of what should be considered to be the detailed rules that ought to govern the existence of this causal link.13

It follows that, contrary to what is likely to be the assumption of many jurists in antitrust private-enforcement cases, it is already becoming settled in EU case law that the “who,” “what” and “why” of the right to damages arising from Articles 101 and 102 TFEU is determined by EU – rather than national – law. National law determines the “how,” within the limits of equivalence and effectiveness.

Thirdly, because the right to damages is determined by EU law according to its own logic, not all the requisites found in the tort laws of some MS will be applicable. In this case, the Court specifically ruled out the need for the damage to fall within the scope of protection of Article 101 TFEU.14 This scope of protection issue is a requisite found, namely, in Member States with German-style tort laws, but it has no place – as such – in the requisites for a right to damages deriving from Article 101 or 102 TFEU. EU law is concerned simply with identifying sufficient causality between the damage and the anticompetitive practice. Alternatively, at least for practical purposes, this may be perceived by Germanophile lawyers as the Court’s implicit statement that compensation of any and all damages that have a sufficient causal link with the anticompetitive practice falls within the scope of protection of Articles 101 and 102 TFEU.

Fourthly, the CJEU stated that the burden of proof of causality lies on the injured party, which must demonstrate before the national court that the damage was indeed caused by the anticompetitive practice. This could be the case if the claimant had the possibility of making more profitable investments with the capital which was unavailable to it as a result of the infringement.15

Finally, Otis also provides a clue for quantification: the damage will consist in the remuneration of the capital which could have been obtained, had it been free to be invested otherwise. This is in line with the right to full compensation, laid down in Article 3 of the Damages Directive.16

Notwithstanding the important clarifications deriving from Otis, the brevity and content of the judgment, especially when compared to AG Kokott’s Opinion, is still a little disappointing. We believe the CJEU could have gone further, without overstepping the limits of the referral procedure.

While there are also important differences, there is an obvious parallelism between these recent judgments of the CJEU on liability for infringements of EU Competition Law, and the role played by the CJEU in the 1990s when it set out the principles for liability of Member States for infringements of EU Law.17 In that regard, the lessons of the past should not be forgotten. The relative vagueness with which the Court set out the criteria for Member States’ liability left much room for legal uncertainty and for heterogenous solutions throughout the European Union. The Court is again showing the same tendency for vagueness.

While in each referral the CJEU is merely providing clarifications meant to assist the national judge in that specific case, it is obviously providing clarifications which will be useful (and indirectly binding, through the principle of sincere cooperation) in future cases. A concrete outcome to a specific case is often more helpful than a vague abstract rule. The latest judgments in antitrust private enforcement, while extremely exciting to specialized EU competition practitioners, have done less to provide legal certainty than they could have, because the vagueness of some of their phrasing and the incompleteness of the clarifications they provided left room for national courts to fill in the blanks with their own views as to the legal solutions.

Tibor Trans was, mostly, a good example of a clear statement of the law in a simple factual context, but this very simplicity meant that the Court ended up leaving many doubts for future, less straightforward, cases.18 Skanska19 is a great example of how a clear statement of law can be well received and applied by national courts,20 even going beyond the most immediate implications of the judgment.21

On the other side of the spectrum is Cogeco.22 The Court avoided providing clarifications which were superfluous to that case (but extremely important for future cases). As for the main issue it did answer, after the CJEU declared that Portuguese rules on time-barring infringed EU law, the 1st instance Portuguese court said that the CJEU had failed to take into account one characteristic of Portuguese Law, and that it did not infringe EU law.23 As this case highlights, one of the main problems of referrals – which is continuing to come up in private enforcement cases24 – is that national courts need to know what to ask, and need to provide the relevant information to the CJEU, as the latter’s proactivity in rephrasing questions and filling gaps in information varies from case to case.

The CJEU could choose to engage in more of a dialogue, in a spirit of cooperation with national courts. And it should keep in mind who its audience is, and seek to adopt phrasing which ensures its message is correctly understood by the largest possible part of that audience. Otherwise, it is partly depriving its own judgments, and EU Law and the rights deriving from it, of their effectiveness.

Otis is a good example of this. The judgment is, arguably, so vague that there is really no way of knowing whether the Province of Upper Austria will, in the abstract, be entitled to damages or not. It is entirely possible, in light of the wording of the judgment, that two courts deciding the same case would arrive at different outcomes, not because of different assessments of the facts and the evidence, but because of different understandings of the legal test.

The Court had a chance to provide a clear answer about whether a person in the situation of the claimant was entitled to damages under EU Law. It could have said: “if requisites X, Y and Z are met, such a person is entitled to damages as a matter of EU law,” it being up to the national court to determine if those requisites were met in the specific case.

It did not do that. The formulation of the ruling itself (specifically, the operative part of the judgment) – which, unfortunately, is often the only focus of attention of national courts – merely says that a person in the position of the claimant “may seek an order” for compensation. And even if we read – as we should – the full judgment, the answer is not clear. The Court says that such a person “must be able to request compensation.”25 Notwithstanding the phrasing of the question referred to the CJEU, this is a very surprising, procedural-sounding way of responding to a question which was, ultimately, about the existence of a subjective right. Some could interpret this, not as the declaration that there is a right to compensation deriving from EU Law in such a situation, but only that there is a right to seek compensation. This is a subtle but fundamental difference. We believe the reasoning followed by the Court, and the interpretation of the case law as a whole, shows that the CJEU was affirming that there is a subjective right to damages deriving from EU law in cases such as this one, as long as causality can be established. But it will be far from surprising if many national judges and lawyers disagree with that assessment. And that is precisely the point we are trying to make. Furthermore, what degree of causality is required?

The Court tells us that if there is a causal link, there is a right to damages.26 And it tells us one does not have to be active on the market affected by the cartel to have a right to damages.27 But it did not tell us whether this situation showed a sufficient causal link. Paragraph 33 started off with very helpful phrasing: “It is therefore for the referring court to determine whether, in the present case, the Province of Upper Austria actually suffered such loss, by verifying, in particular, whether that authority had the possibility of making more profitable investments…”. But it then took a sharp turn into murky waters by adding: “… and, if that is the case, whether that authority adduces the evidence necessary of the existence of a causal connection between that loss and the cartel at issue.”

This second part of the sentence concerns the standard of proof, which is a procedural aspect left up to national law (the “how” to prove the right to damages), but it also concerns the determination of a causal connection, and the legal test for this must derive from EU Law itself. The same injured party cannot be entitled to damages or not, based on Article 101 or 102 TFEU, depending on the Member State in which it files its action (and that Member State’s own legal test for causality in tort actions).

Was there a sufficient causal connection in Otis or not? What legal standard should the national court use to assess if there is a sufficient causal connection in this case? Is it enough for the claimant to prove that it was obliged to pay out more than it would have paid in the absence of the cartel, and that it could have used that capital to make more profitable investments? It sounds to us like this is what the Court meant. But then why not just say that?!


Otis is another great step-forward in EU Law on antitrust private-enforcement. It confirmed several important matters and contributed to the crystallization of the CJEU’s core ideas in its previous rulings, guaranteeing them a place as settled case law. The Court has picked up where the EU legislator left off and is leading the private enforcement revolution in Europe.

This case concerned the requisite of causality, which is of particular importance, namely because it was one of the matters which the Damages Directive only marginally addressed, and where case law has a large void to fill.

Asking hundreds of national courts to apply extremely complex law to extremely complex sets of facts is bound to lead to heterogenous solutions. It is a very big challenge to get national judges to, first, recognize that several of the solutions imposed by EU Law run counter to their national solutions in civil and procedural law and to their legal instinct developed on the basis of those national rules, and then to be able to properly identify those solutions.

The CJEU is proving to be a friend to EU antitrust private enforcement, in its time of need. But it is a friend who is, to some extent, speaking in riddles, and being far less clear than it could be. And that makes its help somewhat less helpful. If a question on the interpretation of EU Law comes before the Court and, after its reply, national judges still disagree about the answer provided by the Court, something has gone wrong. And while the Court may not always be responsible, it cannot shirk all responsibility. In Otis, for example, it could have been clearer. It could have expressed itself in such a way that all but the most unwavering defense lawyers would know whether a claimant in the circumstances of this case would have a right to damages, and making papers, such as this one, discussing the meaning of the judgment, unnecessary.

The greatest challenges for private enforcement in the next decade are connected to the interpretation and application of the law in the various Member States by the national courts. This is a decentralized system, with the CJEU as the only real guardian of the uniformity and effectiveness of Articles 101 and 102 TFEU. The clearer its stand on the specific cases that come before it, the greater will be its contribution, and the sooner legal certainty will be achieved.

Click here for a PDF version of the article

1 Professor at University of Lisbon Law School and European University (Lisbon). Partner at Sousa Ferro & Associados. In compliance with ASCOLA’s Transparency and Disclosure Declaration, the author discloses that he represents the plaintiff in the pending Cogeco case, mentioned in this paper.

2 Trainee Lawyer at Sérvulo & Associados. Master in International and European Law (Nova Law School).

3 Case C-435/18 Otis EU:C:2019:1069.

4 Case C-724/17 Skanska EU:C:2019:204.

5 Case C-637/17 Cogeco EU:C:2019:263.

6 Case C-451/18 Tibor-Trans EU:C:2019:635.

7 Opinion of AG Kokott in Case C-435/18 Otis EU:C:2019:651 (“AG Kokott’s Opinion”).

8 Otis, paras. 11-18.

9 Otis, paras. 19-20.

10 Otis, para. 34.

11 Case C-557/12 Kone EU:C:2014:1317.

12 AG Kokott’s Opinion, para. 60.

13 AG Kokott’s Opinion, paras. 57-59.

14 Otis, para. 31.

15 Otis, para. 33.

16 Directive 2014/104/EU of the European Parliament and of the Council of 26 November 2014 on certain rules governing actions for damages under national law for infringements of the competition law provisions of the Member States and of the European Union (OJ L 349/1, 05/12/2014).

17 See, e.g. Cases C-6/90 and C-9/90 Francovich EU:C:1991:428; and Cases C-46/93 and C-48/93 Brasserie du Pêcheur EU:C:1996:79;

18 Case C-451/18 Tibor-Trans EU:C:2019:635.

19 Case C-724/17 Skanska EU:C:2019:204.

20 See, e.g. Finnish Supreme Court judgment of October 22, 2019, Skanska (KKO:2019:90), described here.

21 See, e.g. judgment of the Court of Appeal of Arnhem of November 26, 2019, TenneT (NL:GHARL:2019:10165), described here.

22 Case C-637/17 Cogeco EU:C:2019:263.

23 Lisbon Judicial Court Order of June 11, 2019, Cogeco v. Sport TV et al (case no. 5754/15.7T8LSB). An appeal is pending.

24 See the recent referral in Case C-716/19 Repsol.

25 Otis, para. 32.

26 Otis, para. 30.

27 Otis, para. 32.