A rather curious judgment was handed down last month by the CJEU, concerning the 2001 Brussels Regulation. At first glance, the decision is not surprising. It merely confirms that the EU has moved with the times and, outside the business-to-consumer context at least, has permitted jurisdiction agreements to be communicated by “click wrapping”. This means allowing them to be seen by clicking on a hyperlink rather than appearing automatically on a web page or being sent to a party in hard copy or by fax or email. But incidentally, the judgment also highlights the inconsistency of different language versions of EU legislation, and serves as a warning not to rely exclusively on English language ones.
First of all, some background to the CJEU’s decision. When the 2001 Regulation was updated recently, becoming the Recast Brussels Regulation, its general architecture and key provisions remained largely intact, but a few significant changes were made to the details of its rules and the way they interact. Where jurisdiction agreements are concerned, the most significant change was to give exclusive agreements priority over the “court first seised” rule (new Article 31(2)). This means that the chosen court can proceed to try a case, regardless of whether the dispute is already being litigated elsewhere in the EU. Before the change it had to wait until the “court first seised” had declined jurisdiction, which was an open invitation to litigants to engage in delaying tactics by issuing proceedings pre-emptively in a jurisdiction that had nothing to do with the dispute, but where the wheels of justice turn very slowly (hence the popular term for this tactic, the “Italian torpedo”). Another welcome change was to recognise jurisdiction agreements in favour of EU courts, regardless of the domicile of the parties (new Article 25(1)). Previously, at least one of them had to be based in the EU. Less welcome, perhaps, is the new wording in Article 25(1), which complicates matters by making the validity of a jurisdiction agreement dependent in part on the law of the member state where the chosen court is located. However, the EU Commission generally chose to leave well alone, and much of old Article 23 was carried forward unchanged into new Article 25.
One aspect of old Article 23 that survived the “recasting” unscathed was the wording (now in Article 25(1) and (2)) setting out the formal requirements for a valid jurisdiction agreement. On the face of it, these seem straightforward enough.The agreement must either be in writing (or evidenced in writing) or be in a form that accords with practices which the parties have established between themselves. A concession is also made in certain circumstances to relevant usages in international trade or commerce (see Article 25(1)(c)). In addition, “any communication by electronic means which provides a durable record of the agreement shall be equivalent to ‘writing'” (Article 25(2)).
But simple though these provisions sound, they are not necessarily straightforward to apply in practice. For example, what has to be in writing is not so much the terms of the jurisdiction agreement itself as the fact that they have been agreed. Now the CJEU has had to deal with another aspect of the formal requirements that seems straightforward, but can be interpreted in different ways. In El Majdoub v CarsOnTheWeb.Deutschland, a dealer who purchased a car claimed that he was not bound by a jurisdiction agreement, partly because it was “click wrapped” rather than opening automatically on his computer screen. This, he argued, did not satisfy Article 23(2) of the original 2001 Brussels Regulation, now Article 25(2).
Oddly enough, if the buyer had been British or Irish and relied on the English language version of Article 25(2), quoted above, his argument might have had some force. It is quite plausible to maintain that being given access to terms and conditions on a website is not the same as being “provided” with a durable record of them, for example by being sent them in hard or soft copy. So when I first read the CJEU’s judgment it seemed questionable, and I could not understand why the court referred to the communication being “capable” of providing a durable record (paragraph 20 of the judgment) or “the possibility” of it doing so (paragraphs 32 and 33). No word with either meaning appears in the English language version of Article 25. But, as it happens, the parties and national court in the case were German, so presumably consulted the German language version of the Regulation. That version, like the French one, expresses Article 25(2) in rather different terms, providing only that the relevant terms should be transmitted in a way that allows (“ermöglichen” in German, “permet” in French) the creation of a durable record. There is no suggestion that a durable record should in fact be provided.
Seen in that light, the CJEU’s decision is not surprising. The court pointed out that the buyer was able not only to view the relevant terms on his computer, but to save them and print them out if he wanted, and that was enough to satisfy the Regulation’s formal requirements (paragraph 32 of the judgment onwards). But the decision raises an important question. It may be sufficient if the communication allows for the creation of a “durable record”, but what does that expression really mean? I see that Briggs and Rees, in the 5th edition of their Civil Jurisdiction and Judgments (about to be superseded by a 6th edition), believe it excludes text and voicemail messages, since these are “ephemeral” (paragraph 2.115). But isn’t that view a little old-fashioned now? After all, text messages are saved indefinitely on many smart phones, and voicemail messages are routinely converted into sound files and attached to emails in the office environment. It’s true that standard terms and conditions are unlikely to be communicated in either of these ways, but a party’s agreement to them might well be.
More broadly, the decision is unsettling because it highlights (unintentionally) how inconsistent the different language versions of EU legislation can be. It is a general principle that all official language versions are equally authentic, and that differences have to be resolved by giving the provision in question a purposive interpretation. This is in fact alluded to by the CJEU in paragraph 30 of its judgment, where it says that one of the aims of old Article 23/new Article 25 is “ensuring the real consent of the parties” to the jurisdiction of the agreed court. But it would certainly be nice if litigants didn’t have to deal with these inconsistencies in the first place.
