I wanted to post this comment since CJEU in October this year handed its perplexing Pinckney C-170/12
ruling. When this Thursday Advocate General Jääskinen published its opinion in Coty Prestige C-360/12
, I felt obliged to go back to its confusing language and arguments and point out why Pinckney
is a bad law.
My regular readers will remember that yet in September
, I predicted that Pinckney
will be most likely rejected for irrelevance of questions, and even if accepted, the decision will most likely follow the line of Pammer
C‑585/08 and Hotel Alpenhof
C‑144/09 and its “targeting” principles. The time has proven my both predictions to be plainly wrong
. CJEU accepted the case, and decided for “mere accessibility” criterion for jurisdiction under Art. 5(3) of Brussels I.
More exactly the court held that:
19 The Court has also held that the justification for a request for a preliminary ruling is not that it enables advisory opinions on general or hypothetical questions to be delivered, but rather that it is necessary for the effective resolution of a dispute (Case C-459/07 Elshani  ECR I-2759, paragraph 42).
20 In the present case, it is common ground that the referring court is dealing with an allegation of copyright infringement resulting from the online offer of a material support reproducing a protected work and that the issue as to whether the French courts have jurisdiction to hear that allegation is the very subject-matter of the dispute in the main proceedings. In the light of all the information before the Court, it does appear that the outcome of that dispute depends on the answer to the questions referred, which, furthermore, require reformulation.
47 In the light of all of the foregoing, the answer to the questions referred is that Article 5(3) of the Regulation must be interpreted as meaning that, in the event of alleged infringement of copyrights protected by the Member State of the court seised, the latter has jurisdiction to hear an action to establish liability brought by the author of a work against a company established in another Member State and which has, in the latter State, reproduced that work on a material support which is subsequently sold by companies established in a third Member State through an internet site also accessible with the jurisdiction of the court seised. That court has jurisdiction only to determine the damage caused in the Member State within which it is situated.
It took me some time and discussions with other people to grasp what decision really tells us. After little bit of thought, the decision, I think, makes two important and at the same time dangerous conclusions:
- that a court can hear a “territory-limited claim” against a defendant, who is not domiciled, and did not act or directly cause an effect by his acts in a state of the court, purely on the basis that that the effect was furthered by somebody else, whose
- website was merely accessible in that state.
I know this sounds complicated. I will try to put it more simple words. The main issue is that even if there is no ground to establish jurisdiction against one defendant under Brussels I., the courts will be still able to extend it by attributing him harm (effects) caused by second person against who the jurisdiction can be established, if they in some unknown way acted together. So if the court has no jurisdiction over defendant “A” under Art. 2 or Art. 5(3) of Brussels I., the plaintiff just needs to point to some second alleged co-tortfeasor, who fulfills the criteria. The person who in this way helps to establish extended jurisdiction does not need to be part of the proceedings.
This is exactly what happened in Pinckney. An Austrian company Mediatech that did not act, or cause any effect by its acts in France, can be still sued in France, only because some unrelated second company from UK furthered Mediatech’s infringement by making the website accessible in France.
My initial reaction was that Pinckney does not mean all of this, but CJEU only did not want to reject the questions. If you, however, look at the ratio decidendi of the ruling, this alternative is not very convincing. The court does not say that French court should examine if accessibility is helpful for it, but without any problem goes and analyzes the behavior of the UK company, which is not defendant in the national proceedings.
Now, you might think that Melzer C‑228/11
rejected this some time ago. The Melzer
, however, only concerned the extended jurisdiction based on attribution of acts
under Art. 5(3) of Brussels I. The effects of acts
, as second alternative of Art. 5(3) since Bier
C-21/76, however, stayed open. And Pinckney
seems to endorse for effects what Melzer
rejected for acts. Sounds weird? Well, its seems that I am not the only one reading the situation this way.
Advocate General in Coty Prestige C-360/12
struggles with Pinckney
in the following way (Google translate version as English version is not yet online):
68 If extended to other types of tort, the logic [of Pinckney] in my opinion, could lead to the conclusion that a court according to the place where the damage occurred, would have jurisdiction if, as in the main proceedings, this damage is caused by the fact that “the danger” that the suspect goods are made available to consumers in the Member State in which that court is located, and that such an act leads to civil liability under the lex fori. In the light of the judgment Pinckney, the fact seems to be irrelevant that the alleged tort has, but produced effects in the district of the court because of the actions of an independent third party, that has acted outside the Member State both of the court and the plaintiff and was not sued.
69 Because the facts of this case underlying similarities with that in the Pinckney case has and me the differences between these cases do not appear critical, it is difficult for me to find arguments, which made it possible, given the way for the determination of the place where the damage has occurred to depart within the meaning of Article 5 No. 3 of the Brussels I Regulation.
The opinion shows how important Advocate Generals are for the Union law and work of the CJEU. They provide for self-reflection, that is unfortunately not always heard by the judges. AG in his Opinion rightly points out (§§ 57-64) how dangerous this approach can be. Not only it would lead to extension of Art. 5(3) of the Brussels I, less legal certainty and new possibilities of forum shopping, but would also require creating the quantitative or qualitative criteria for such attribution of effects. Purely procedural jurisdictional attribution, that would probably somehow try to mirror joint and several liability on the substantial law level – a very difficult construct.
However appealing this might sound to some of you, imagine the consequences. In order to establish jurisdiction against any on-line user-committed tort, it would suffice that you find an ISP who furthered the infringement and might be liable for his omission. This would suffice to establish the jurisdiction against the user
. And it would of course work both directions. The limits of the attribution would depend only on the CJEU.
If you think that requirement of directness of the damage (Dumez C-220/88, Marinari C-364/93) solves this, try to find any reference to it in the Pinckney. In fact, this kind of attribution is only a back-door solution circumventing CJEU’s own requirement of directness of the damage.
Now, when you combine this first part of Pinckney
with the mere accessibility
criterion, the worrisome picture is well painted. Of course, the mere accessibility is limited by mosaic doctrine of Shevill C-68/93
, but this represents only a very weak remedy against possible jurisdictional abuse (e.g. consider costs of defending yourself in all of those countries, even against groundless or de minimis claims). It would make more sense if CJEU would:
- either reconsider mosaic theory entirely and substitute it with some other solution solving CJEU’s fear of forum shopping (e.g. targeting theory),
- or at least supplement the mosaic theory with some additional corrective for online environment, such as “targeting” requirement of Pammer C‑585/08 and Hotel Alpenhof C‑144/09.
If you read the Pinckney
, you realize that CJEU has no real arguments
against such corrective. The fact that no such condition is literally written in the Art. 5(3) of Brussels sounds like a bad joke for those who know what kind of “acrobatic” argumentation CJEU usually uses to push European integration. It is especially disappointing that CJEU did not take advantage of any existing academic proposals, like the CLIP project
The CLIP recognizes the existence of problem of pure “accessibility theory” by providing for an escape clause to prevent jurisdiction to be assumed over de minis claims under mosaic theory (which it continues to use).
Article 2:202: Infringement
In disputes concerned with infringement of an intellectual property right, a person may be sued in the courts of the State where the alleged infringement occurs or may occur, unless the alleged infringer has not acted in that State to initiate or further the infringement and her/his activity cannot reasonably be seen as having been directed to that State.
CLIP has advantage that it puts the burden to prove the circumstances of the escape clause upon the defendant. This makes sense because sometimes targeting might be not so obvious from the outside, which would make it too difficult to prove for the plaintiff (e.g. consider a Russian website that is in Russian with no references to UK territory, but has substantial amount of visitors from the UK – the fact that only operator of that website can prove).
Luckily, CJEU has two upcoming decisions, where both problems can be appropriately solved. The first issue can be reconsidered in Coty Prestige C-360/12
, and the second in Hejduk C-441/13
. Let’s hope that judges hear to what Advocate Generals say – at least on the first issue. Any thoughts?
Some comments by others
- Graham Smith (great in-depth analysis you should definitely read)
- Ge Chen (little too empathetic reading of the case)