Last week, AG Saugsmandsgaard Øe rendered his Opinion on Case C-59/19 Wikingerhof, which we first reported in this post by Krzysztof Pacula. The following post has been written by Michiel Poesen, PhD Candidate at KU Leuven, who has been so kind as to share with us some further thoughts on the underlying problem of characterisation.
Characterisation is not just a bee that has been buzzing in conflicts scholars’ bonnets, as Forsyth observed in his 1998 LQR article. Given its central role in how we have been thinking about conflicts for over a century, it has pride of place in jurisprudence and literature. The Wikingerhof v Booking.com case (C-59/19) is the latest addition to a long string of European cases concerning the characterisation of actions as ‘matters relating to a contract’ under Article 7(1) of the Brussels Ia Regulation n° 1215/2012.
Earlier this week, Krzysztof Pacula surveyed Advocate General Saugsmandsgaard Øe’s opinion in the Wikingerhof case on this blog (Geert Van Calster also wrote about the opinion on his blog). Readers can rely on their excellent analyses of the facts and the AG’s legal analysis. This post has a different focus, though. The Wikingerhof case is indicative of a broader struggle with characterising claims that are in the grey area surrounding a contract. In this post, I would like to map briefly the meandering approaches to characterisation under the contract jurisdiction. Then I would like to sketch a conceptual framework that captures the key elements of characterisation.
1. Not All ‘Matters Relating to a Contract’ Are Created Equal
There are around 30 CJEU decisions concerning the phrase ‘matters relating to a contract’. Three tests for characterisation are discernible in those decisions. In the first approach, characterisation depends on the nature of the legal basis relied on by the claimant. If a claim is based on an obligation freely assumed, then the claim is a matter relating to a contract to which the contract jurisdiction applies. Statutory, fiduciary, or tortious obligations arising due to the conclusion of a contract are also contractual obligations for private international law purposes. I will call this approach the ‘cause of action test’, because it centres on the nature of the cause of action pleaded by the claimant. In recent decisions, for example, the cause of action test has been used to characterise claims between third parties as contractual matters (C-337/17 Feniks, blogged here; C-772/17 Reitbauer, blogged here; joined cases C-274/16, C-447/16 and C-448/16 flightright).
The second approach to characterisation is to focus on the relationship between the litigants. From this standpoint, only claims between litigants who are bound by a contract can be characterised as ‘matters relating to a contract’. This approach has for example been used in the Handte and Réunion européenne decisions. We will call it the ‘privity test’. Sometimes scholars relied on this test to argue that all claims between contracting parties are to be characterised as matters relating to a contract.
The third and final approach emphasises the nature of the facts underlying the claim brought by the claimant. This approach was first developed in the Brogsitter decision (C-548/12). However, it is predated by AG Jacob’s opinions in the Kalfelis (C-189/87) and Shearson Lehmann Hutton (C-89/91) cases (which since have been eagerly picked up by the Bundesgerichtshof of Germany). The Brogsitter decision provided that a claim is a contractual matter if the defendant’s allegedly wrongful behaviour can reasonably be regarded to be a breach of contract, which will be the case if the interpretation of the contract is indispensable to judge. I will dub this approach the ‘factual breach test’, since it directs attention to factual elements such as the defendant’s behaviour and the indispensability to interpret the contract. It is plain to see that this is by far the most complicated of the three approaches to characterisation we discussed here (among other things because of the unclear relation between the different layers of which the test is composed, an issue that AG Saugsmandsgaard Øe entertained in Wikingerhof, –, and C-603/17 Bosworth v Arcadia).
The use in practice and literature of the three approaches laid out above demonstrates a tale of casuistry. Similar claims have been subjected to different approaches, and approaches developed in a specific setting have been applied to entirely different contexts. For example, a few CJEU decisions characterised claims between litigants who are not privy to consensual obligations as non-contractual in nature under the privity test. Other decisions characterised such claims as contractual in nature, applying the cause of action test. A similar dichotomy underlies the characterisation of claims between contracting parties. Initially, the CJEU jurisprudence applied the cause of action test, focussing on the nature of the legal basis relied on (see C-9/87 Arcado v Haviland). Later, the Brogsitter decision adopted the factual breach test, which shifted the focus to the nature of the facts underlying the claim.
It is difficult to understand why these divergences have occurred. How can they be explained?
2. The Theories Underlying Characterisation
A good way to start is to conceptualise characterisation further along the lines of this scheme:
Seen from the perspective of this scheme, the previous section described three ‘tests for characterisation’. A ‘test for characterisation’ refers to the interpretational exercise that lays down the conditions under which a claim can be characterised as a matter relating to a contract. Each test elevates different elements of a ‘claim’ as relevant for the purpose of characterisation and disregards others. Those elements are the identity of the litigants, the claim’s legal basis, or the dispute underlying the claim. As such, it concretises an idea about the broader purpose the contract jurisdiction should serve, which is called a ‘theory’. The divergences among the tests for characterisation outlined above is explained by the reliance on different theories.
The AG’s considerations about Brogsitter in the Wikingerhof opinion illustrate the scheme. The AG observed that the factual breach test is informed by what I will dub the ‘natural forum theory’. According to that theory, the contract jurisdiction offers the most appropriate and hence natural forum for all claims that are remotely linked to a contract (for the sake of proximity and avoiding multiple jurisdictional openings over claims relating to the same contract). This theory explains why the factual breach test provides such a broad, hypothetical test for characterisation that captures all claims that could have been pleaded as a breach of contract. Opining against the use of the factual breach test and underlying natural forum theory, the AG suggested that the cause of action test be applied. He then integrated the indispensability to interpret the contract (originally a part of the factual breach test) into the cause of action test as a tool for determining whether a claim is based on contract ( et seq). Essentially, his approach was informed by what I will call the ‘ring-fencing theory’. In contrast to the natural forum theory, this theory presumes that the contract jurisdiction should be delineated strictly for two reasons. First, the contract jurisdiction is a special jurisdiction regime that cannot fulfil a broad role as a natural forum contractus (–). Second, a strict delineation promotes legal certainty and efficiency, since it does not require judges to engage in a broad, hypothetical analysis to determine whether a claim is contractual or not (–). The scheme was applied succinctly here, but the analysis could be fleshed out for example by integrating the role of the parallelism between the Brussels Ia and Rome I/II Regulations.
The scheme can be used to understand and evaluate the CJEU’s eventual judgment in Wikingerhof. I hope that the decision will be a treasure trove that furthers our understanding of the mechanics of characterisation in EU private international law.