I. INTRODUCTION
In Airbus Industrie G.I.E. Respondents v Patel, Lord Goff summarised and outlined two different approaches to the allocation of jurisdiction in international commercial litigation.Footnote 1 The first approach, widely adopted in common law jurisdictions, is to allocate jurisdiction to the place most proximate to the dispute in question. Under this approach, a wide range of factors are relevant in determining whether the court seised is the most suitable for handling the dispute. Another approach, prevalent in continental European countries, is embodied in the Brussels I Recast (Recast).Footnote 2 Being a European Union instrument, it aims to avoid competing assertions of jurisdiction between Member States through well-defined rules. Its general aims can be summarised as follows:
– Unity: to ‘unify the rules of conflict of jurisdiction in civil and commercial matters’;Footnote 3
– Predictability: jurisdictional rules should be highly predictable and foreseeable;Footnote 4
– Avoiding forum actoris: avoiding the claimant having the advantage of suing in the courts of their domicile, thus exposing the defendant to potentially frivolous litigation in unfamiliar jurisdictions.Footnote 5 Unless expressly provided for in the Recast (eg Article 18(1)), the Regulation is generally ‘hostile’Footnote 6 to the idea of conferring jurisdiction to the courts of the claimant's domicile;
– Preventing overlapping jurisdiction: minimising concurrent proceedings and irreconcilable judgments between different Member States.Footnote 7
These aims accord with a strong civil law tradition, which ‘places greater emphasis on certainty and predictability than flexibility’ as compared to common law traditions.Footnote 8
Article 7(1) governs the allocation of jurisdiction in respect of contractual disputes and is one of a number of special jurisdiction rules. Article 7(2) is another special jurisdictional rule, which governs the allocation of jurisdiction in respect of torts, delicts or quasi-delicts, this being to the place where the harmful event occurred. In cases where a dispute concerns a branch, agency or other establishment, Article 7(5) confers jurisdiction on the place where the enterprise is situated. Recital 16 of the Recast provides that one of the aims of the special jurisdiction rules is ‘[to provide] a close connection between the court and the action’.Footnote 9 Special jurisdiction rules under the Recast can therefore be considered as giving effect to an attenuated form of forum conveniens, being exceptions to the general rule in Article 4(1) that the defendants should be sued in their own domicile.
There are two different approaches to determining jurisdiction within Article 7(1). Article 7(1)(a) determines the place of performance of a particular contractual obligation with reference to the lex causae of that obligation. The lex causae is determined by the conflict of law rules of the court seised. Given the drawbacks of this, Article 7(1)(b) sets out a uniform set of rules for determining jurisdiction in respect of contracts for the sale of goods and for the supply of services, irrespective of the substantive law of a particular Member State. The place of performance of such contracts is located at the place of performance (ie, the place where the goods are to be delivered or the services performed), this being determined on the facts.
This article examines the dynamics between these competing approaches to determining jurisdiction, all of which are embodied within this single provision. The interpretations of the Court of Justice of the European Union (CJEU) concerning Article 7(1) have often produced more questions than answers. This article looks principally at two questions in respect of Articles 7(1)(a) and 7(1)(b): (i) how the two subsections determine the place of performance in general, and the respective problems associated with them; and (ii) how they allocate jurisdiction if the contractual obligation is to be performed in more than one Member State.
The first question explores problems associated with determining jurisdiction by reference to the lex causae and explains why a uniform and autonomous interpretation of the place of performance of a contract best accords with the aims of the Recast. It is argued that the approach set out in Article 7(1)(b), namely the place of performance, should be extended to some forms of contract which currently fall within the scope of Article 7(1)(a).
The second question involves examining the fault line between the two different approaches to the allocation of jurisdiction in international commercial litigation, as outlined earlier. When determining jurisdiction in cases involving more than one place of performance, should a rigid, but predictable, or a more flexible approach be preferred? It is argued that, given the Recast's overarching aims, simplicity and certainty should be encouraged. The CJEU should therefore follow its earlier decision in Besix SA v WABAG Footnote 10 and have recourse to Article 4 when a claim involves contractual performances in multiple Member States.
II. DETERMINING THE PLACE OF PERFORMANCE UNDER ARTICLE 7(1) OF THE RECAST
The difficulty of applying Article 7(1)(a) is the main reason why autonomous linking factors have been introduced in relation to the two commonest types of contracts, contracts for the sale of goods and for supply of services. This section argues that more types of contracts should follow the approach of Article 7(1)(b).
A. Determining Jurisdiction under Article 7(1)(a)
The process of allocating jurisdiction under Article 7(1)(a), following Tessili v Dunlop,Footnote 11 involves three steps: (i) identifying the obligation which forms ‘the basis of the legal proceeding’;Footnote 12 (ii) identifying the lex causae of that particular obligation in accordance with the private international law rules of the court seised, and (iii) applying the substantive law rules of the lex causae to determine the place of performance.Footnote 13 Therefore, the place of performance varies according to the particular obligation allegedly in breach. Although the Rome I Regulation can now assist in pinpointing the lex causae,Footnote 14 and thus simplifying step (ii), three problems remain.
First, requiring the court seised to determine the place of performance by reference to the lex causae is unduly complicated.Footnote 15 Most legal systems determine the place of performance of a contractual obligation by reference to the parties’ express or implied intention,Footnote 16 a task which is often best left to be determined during the substantive proceedings. In GIE Groupe Concorde v Master of the Vessel Suhadiwarno Panjan, Advocate General Ruiz-Jarabo noted that the determination of the place of performance under the lex causae could ‘come up against the general antipathy of most legal systems to theoretical definitions’.Footnote 17 The Advocate General described this approach as ‘extremely laborious’,Footnote 18 especially when the court seised is only asked to determine a preliminary question of jurisdiction instead of the substantive merits of the case. Often the reason why national courts appear to have followed Tessili is because the place of performance as determined under the lex fori happens to be the same as that of the law applicable to the obligation.
The uncertainty and delay caused by this approach has prompted some Member State courts to not follow Tessili. Jurisdiction is instead determined according to the lex fori or the factual place of performance.Footnote 19 This rejection of Tessili is most notable in Comptoir Commercial d'Orient S.A. v Medtrafina S.A..Footnote 20 Decided on 11 March 1997, the French Cour de Cassation upheld an application to enforce a Greek judgment in France and found that the Court of Appeal, ‘in investigating the place of performance of the obligation … justifiably defined that place as the place of real and effective performance, by reference to the nature of the obligation and the circumstances of the case’.Footnote 21 However, just seven days later, on 18 March 1997, the same court in Ernesto Stoppani SpA v Stoppani France cited and applied Tessili in full, quashing the Court of Appeal's judgment since it failed to determine the place of performance by reference to the law applicable to the contract.Footnote 22 The Tessili approach has thus been applied ‘very unequally’ and ‘very imperfectly’.Footnote 23 It is therefore apparent why the Court in Car Trim GmbH v KeySafety System Srl Footnote 24 was in favour of adopting autonomous linking factors for Article 7(1)(b) contracts,Footnote 25 leading to inconsistencies in the approaches adopted by the two subsections.Footnote 26
Secondly, the lex causae often determines the places of performance on the basis of principles which differ from those underlying Recital 16. Therefore, whether or not the place of performance determined by the lex causae achieves the aim of ensuring that there is a close connection is very much a matter of chance.Footnote 27 This discrepancy is especially apparent in cases concerning the obligation to deliver and the obligation to pay. As regards the obligation to deliver, consider for example a simple case where goods are sold by English sellers to German buyers, to be resold to German customers. Different substantive law rules can result from there being different places of performance of the same obligation. Under an ‘ex-factory’ contract, the place of delivery is in the seller's country of residence (ie England). For a free-on-board contract, the place of delivery would be the place where the goods are loaded onboard a ship. However, it is Germany that has the closest connection to the dispute. The Advocate General in Custom Made Commercial Ltd v Stawa Metallbau GmbH was right to observe that:
[t]he rules of the lex causae concerning the place for performance of the seller's obligation to supply the goods … contain elements which serve only to share out the risk, in the present case the transport risk, and which give no reliable information on the economic objective of the seller's obligations.Footnote 28
The place of performance of the obligation to pay is another example. Substantive law rules amongst Member States diverge. Under Italian and Dutch law it is for the debtor to seek out the creditor, whilst under German and French law it is the opposite. However, the rationale for adopting these respective rules have almost nothing to do with the procedural question of determining which court has the closest connection to the dispute; they are based on the need to protect the perceived weaker party to the transaction.Footnote 29 Taking these two examples together, determining the place of performance of an obligation on the basis of the lex causae does not in any way guarantee that the place so designated has the closest connection to the dispute.
Thirdly, the problem of forum actoris frequently arises since the substantive law rules allow unpaid claimants to sue in their own courts. Defendants, who very often rely on the quality of the counter-performance as a defence, will be obliged to defend themselves before a foreign court. In such cases, the court of the place where the goods are to be delivered generally has a closer connection to the subject matter of the dispute than the place of payment. The adoption of an autonomous linking factor in Article 7(1)(b) is intended to prevent an unpaid seller from routinely being able to obtain ‘home-court advantage’.Footnote 30
B. The Uniform Interpretation of the Place of Performance under Article 7(1)(b)
Article 7(1)(b) is an improvement upon Article 5(1) of the Brussels Convention 1968Footnote 31 in two regards. First, it is the places of delivery and provision of services which assume jurisdiction, regardless of the specific obligation in question. Secondly, the place of performance is autonomously defined on the basis of a ‘purely factual criterion’,Footnote 32 by reference either to the terms of the contract or to the place where the goods were physically transferred. Therefore, Article 7(1)(b) precludes the application of the rules of private international law of the Member State and the applicable substantive law rules. However, there are five problems with this essentially factual approach.
The first problem is that the court having jurisdiction can still not be the most proximate to the dispute. There are two points to note here. First, it is essential to distinguish between the factual and legal connections between the court and the dispute. In Car Trim, the place of performance of a sale of goods contract was the place where the goods were ‘physically transferred or should have been physically transferred to the purchaser at their final destination’.Footnote 33 This ensures that there is a close factual connection, which is especially important in cases where goods delivered are allegedly defective, since goods are most likely to be examined by the buyer at the final destination. However, the legal connection between the court and the dispute may be weakened if the buyer's forum has to apply the law of the seller's habitual residence pursuant to Article 4(1)(a) of the Rome I Regulation. This issue is discussed further discussed in section II.D below.
The second problem is that disputes unrelated to the contract's principal obligation may have no connection with the court designated under this rule. Kareda v Benkö, in which the CJEU ruled that a claim between debtors should be brought at the place of the lender's registered office, illustrates this point.Footnote 34 While the weakness of the connection between the place having jurisdiction and the claim was apparent, the CJEU was clearly persuaded by the fact that the place of performance was more foreseeable, and it avoided there being the multiplicity of jurisdictions that would otherwise occur if the places of performance were different for bank–debtor and debtor–debtor claims.Footnote 35
The third problem is that the court first seised must determine whether the place of delivery is apparent from the provisions ‘under the contract’. The reference to provisions ‘under the contract’ in Article 7(1)(b) refers to terms that ‘do not identify directly and explicitly the place of performance’,Footnote 36 and thus include, for example, international terms of trade. Electrosteel Europe SA v Edil Centro SpA involved a dispute concerning a contract of sale between an Italian seller and a French buyer on ‘ex-works’ terms. The CJEU upheld the determination of the place of performance by reference to the terms of the contract, provided such reference did not involve the application of the rules of private international law and the substantive law applicable to the contract. Therefore, the Italian court had jurisdiction pursuant to the terms of the contract, even though the ultimate destination of the goods was France, the headquarters of the buyer. Utilising international terms of trade to determine the place of performance may not necessarily ensure that the forum so designated has the closest connection with the dispute, for reasons outlined in section II.A above. However, this is consistent with the overarching approach adopted in Car Trim, according to which jurisdiction is determined by a fact-oriented approach to the ‘place of performance’, without resorting to substantive law rules.
A fourth problem is that the words ‘unless otherwise agreed’ in Article 7(1)(b) allow the parties to expressly designate a place of performance. While the place designated by the parties may not be the place where the obligation is to be performed, this appears to reflect the autonomy of the contracting parties, a principle reinforced by Recital 14 of the Recast. This autonomy is, however, limited by Mainschiffahrts-Genossenschaft eG (MSG) v Les Gravieres Rhenanes Sarl,Footnote 37 which ruled that ‘a place of performance which has no actual connection with the real subject-matter of the contract becomes fictitious’, and thus must satisfy the formal requirements of a jurisdiction clause under Article 25.Footnote 38 The threshold for establishing an ‘actual connection’ is, however, unclear. The Advocate General seemed to support the need for an objective connection, requiring ‘a direct and, above all, objective connection between the dispute and the courts having jurisdiction to entertain it’.Footnote 39
Consider a case in which an Italian seller sells defective goods on free-on-board terms to British buyers, to be resold in the British market, the goods to be loaded onto a ship at a port in the Netherlands. A free-on-board contract is influenced by ‘technical modalities of transport and may overemphasise the importance of the port of loading …’.Footnote 40 If an ‘objective connection’ is needed, the Netherlands arguably has no real connection with the subject matter of the dispute. On the other hand, the MSG case seemingly formulates a subjective test, with the place of performance only becoming fictitious if the designation is made with ‘the sole aim of specifying the courts having jurisdiction’.Footnote 41 In the above example, as long as the place of performance in the Netherlands is not fictitious, it is unlikely that the designation would fail the subjective test. It is suggested that only if the subjective and objective test are both satisfied should the expressly designated place of performance be displaced. This respects the principle of party autonomy, even though it could potentially weaken the ‘close connection’ between the dispute and the claim.
The fifth problem is that Article 7(1)(b) is based on the idea that the contractual terms and/or the actual performance of the contract can provide the ‘factual criterion’ necessary for determining the jurisdiction of the court to be seised. It is true that the place of performance of a contract can in most cases be determined by either of the methods under consideration. However, it is not difficult to envisage cases in which the complaint concerns the non-delivery of goods or non-provision of service, but the place of performance is not provided for in the contract.Footnote 42 Rogerson has suggested two possible solutions: (i) that Article 7(1)(a), including Tessili, fills the gap, pursuant to Article 7(1)(c),Footnote 43 or (ii) that Article 7 is simply inapplicable.Footnote 44 The latter is the better option.
Due to the problems with the Tessili approach identified in section II.A above, it is best not to apply Article 7(1)(a) at all. Applying the default rules under the lex causae to fill the gap (eg by adopting Article 31 of the United Nations Convention on Contracts for the International Sale of Goods) goes against the factual approach on which Article 7(1)(b) is built. More fundamentally, the place of delivery under the lex causae may not be the place having the closest connection with the dispute. Rather, it is the reason for non-performance that is a more useful pointer for determining the issue. For example, if it is because the defendant challenges the existence and/or the validity of the contract that they refuse to perform, the place where the contract was allegedly concluded would be most proximate to the dispute.Footnote 45 Jonathan Hill's discussionFootnote 46 of Tesam Distribution Ltd v Schuh Mode Team GmbH Footnote 47 and Hanbridge Services Ltd v Aerospace Communications Ltd,Footnote 48 both cases concerning non-performance of contractual obligations, has shown how focusing on the specific obligation alleged to have been breached without sufficiently inquiring into the reasons for the non-performance has meant that jurisdiction has not been allocated to places with the closest factual and legal connection to the dispute. Therefore, abandoning the special jurisdictional rules in this scenario best limits the scope of Article 7(1) to cases where proximity between the court and the dispute can be more readily established.
C. Autonomous Linking Factors for Article 7(1)(a) Contracts—Following in the Footsteps of Article 7(1)(b)?
Having evaluated the merits of the approaches under Articles 7(1)(a) and (b) in isolation, it is necessary to consider whether more types of contracts should follow the approach of vesting jurisdiction in the courts where the relevant obligations are to be performed. Article 7(1)(a) continues to govern a large number of contracts, such as third-party guarantees, licensing agreements, sale of intangibles, joint venture agreements, etc. It is suggested that more categories of contracts should adopt the ‘place of performance’ approach.
First, the CJEU has long preferred adopting autonomous interpretations, which also fulfil one of the most important purposes of the Recast: unity (Recital 4). As early as its 1993 decision in Mulox IBC Limited v Hendrick Geels, the CJEU recognised the importance of unifying the rules governing jurisdiction, reasoning that
[a]utonomous interpretation alone can ensure uniform application of the Convention, the objectives of which are, inter alia, to harmonize the jurisdiction rules of the courts of the Contracting States by avoiding as far as possible the multiplication of bases of jurisdiction with regard to the same legal relationship … .Footnote 49
Clarifications of Article 7(1) by the CJEU have consistently favoured autonomous community-wide interpretations of the terms in the Recast, such as the meaning of ‘contract’Footnote 50 and the distinction between ‘sale of goods’ and ‘supply of services’.Footnote 51 It is only in comparatively rare cases, where the Recast specifically provides for recourse to domestic law, that the CJEU reverts back to the approach under Article 7(1)(a). One such example is the meaning of ‘domicile’.Footnote 52
Secondly, Member State's courts are already familiar with determining the place of perfomance, under Article 4(2) of the Rome I Regulation.Footnote 53 The list of contracts in Article 4(1) of the Rome I Regulation, and the linking factors adopted therein, can equally be applied to Article 7(1)(a) contracts. For licensing agreements, the court of the place where the licence is granted can have jurisdiction, whilst for third-party guarantees it could be the place where the guarantor is obliged to pay.Footnote 54 This eliminates the uncertainties of allocating jurisdiction on an obligation-by-obligation basis, focusing instead on the performance that characterises the contract as a whole. As pointed out in the Giuliano and Lagarde Report, the characteristic performance of a contract generally constitutes ‘the centre of gravity and the socio-economic function of the contractual transaction’.Footnote 55 Therefore, granting jurisdiction to the place where the characteristic obligation is performed is a step towards achieving the aim of ensuring a close connection between the court and the claim. All secondary obligations (eg claim for damages or declarations) in those contracts would then be litigated at the place of performance of the characteristic obligation.
If a contract does not have a characteristic obligation (eg barter),Footnote 56 it is arguable that Article 7 should not be applied at all. Given that there is no distinct linking factor in these types of contracts that can ensure a close connection between the court and the dispute, there is no reason why the special jurisdiction rule should be triggered in the first place.
The characteristic obligation solution is not perfect. In fact, no solution within the framework of the Recast can ensure there is a close connection in all cases. This would be best achieved by the forum conveniens doctrine. However, in Custom Made, Advocate General Lenz was steadfast in not adopting ‘an interpretation which takes account solely of the court's proximity to the dispute, but which might undermine the concept of place of performance and turn Article 5(1) into a vague forum conveniens rule’.Footnote 57 It is therefore difficult to reconcile the need for a simple and predictable jurisdictional rule, as embodied by the Recast, with the more fluid approach of forum conveniens. Even for sale of goods cases under Article 7(1)(b), claims for non-payment may sometimes have no significant connection with the place of delivery, such as when the condition of the goods is not an issue in dispute.Footnote 58 However, this situation is comparatively rare.Footnote 59 Therefore, if a connecting factor has to be chosen as a default rule, and applied with certainty, the place of performance best ensures a close connection. As rightly observed by Advocate General Lenz,
[t]he objective cause of the dispute, whether it takes the form of a claim for payment of a price or a claim for damages, is often at the place of performance of the non-pecuniary obligation, so that the designation of the corresponding court is likely to favour the criterion of the close connection between the dispute and the competent court.Footnote 60
Adopting this approach maximises the chances of there being a close connection between the place assuming jurisdiction and the crux of the dispute.
D. Further Harmonisation between the Recast and the Rome I Regulation?
Viewed macroscopically, centring jurisdiction under Article 7(1) of the Recast on the place of performance would also play a valuable role by further harmonising jurisdiction and choice of law rules, the disjuncture between which has already been seen in section II.B above. In a sale of goods contract, if goods are to be delivered to a buyer's residence, Article 4(1)(a) of the Rome I Regulation means that courts in the country in which the buyer is domiciled may have to apply the law of the country in which the seller habitually resides when determining the dispute. This is suboptimal.
Currently, a solution to this problem can be found in Article 4(3) of the Rome I Regulation. This provides that, if it is clear from all the circumstances of the case that the contract is manifestly more closely connected with a country other than that indicated in Articles 4(1) and (2), the law of that other country shall be applied instead. Recital 16 of the Rome I Regulation provides that whilst legal certainty and a high degree of foreseeability are desirable, courts retain a degree of discretion when determining which body of law is most closely connected to the dispute. Article 4(3) thus acts as an ‘escape clause’ in this regard.
Case law suggests that where the place of performance is different from that of the performers’ habitual residence, the court would more readily displace the law of the latter in favour of the law of the former. For example, in Definitely Maybe (Touring) Ltd v Marek Lieberberg Konzertagentur GmbH (No. 2) Footnote 61 there was a contract between the English claimants, who acted for the pop group Oasis, and the German defendants. The claimants had agreed to procure Oasis for concerts to be held in Germany. Morison J held that the presumptions under what is now Articles 4(1) and (2) were more readily displaced when ‘the place of performance is different from the place of the performer's business’.Footnote 62 Therefore, although the claimant's habitual residence was in England, the commercial centre of gravity was in Germany, the place of performance. With the promulgation of the Rome I Regulation, case lawFootnote 63 and commentatorsFootnote 64 have both suggested that the rules in Articles 4(1) and (2) are more than presumptions, and only in exceptional cases would the rules be displaced. However, it is safe to suggest that the escape clause would still be highly relevant in cases where the place of performance differs from that of the performer's habitual residence.Footnote 65
The incongruence between the linking factors in the Recast and the Rome I Regulation regarding contractual disputes has long been noted. The United Kingdom, citing Definitely Maybe, argued, in a position paper sent to the Committee on Civil Law Matters (Rome I) in September 2006,Footnote 66 that the place of performance can often differ from the habitual residence of the performer. In particular, it is very possible that in service contracts the service provider has no ‘branch, agency or any other establishment’ in the actual place of performance. The United Kingdom therefore proposed a new draft Article 4(3), which provided that:
Notwithstanding paragraphs 1 and 2 of this Article, where it is clear from all the circumstances of the case that the contract is manifestly more closely connected with a country other than that indicated by those paragraphs, the law of that other country shall apply. In particular, a manifestly closer connection with another country might exist where:
(a) the characteristic performance is to be effected in that other country; or
(b) the contract is closely linked to another contract which constitutes part of the same transaction and the transaction as a whole is most closely connected with that other country.
While subsection (b) of the proposal has now been reflected in Recital 20 of the Rome I Regulation, the proposed amendment in subsection (a) has not been properly reflected in any of the Recitals.
It is therefore suggested that Article 7(1) of the Recast and Article 4 of the Rome I Regulation would be better aligned if both provisions were based on the concept of characteristic performance—if no other law had been chosen, the law governing the contract would be the law of the place of performance, rather than that of the performer's habitual residence. This would lessen the need to rely on the Article 4(3) ‘escape clause’, since the default choice of law is more likely to be closer to the ‘commercial centre of gravity of a contract’.Footnote 67 If Article 7(1) of the Recast was also modified along the lines suggested in section II.C above, this would then achieve the desirable outcome of the court having jurisdiction in most cases applying its own law instead of foreign law when determining the dispute. This enhances the conceptual unity underlying choice-of-law and jurisdiction rules regarding contractual disputes.
III. PERFORMANCE OF CONTRACTUAL OBLIGATIONS IN MULTIPLE PLACES
A second controversy that the CJEU's jurisprudence has created concerns cases where there is a multiplicity of places of performance. Article 7(1)(a) allocates jurisdiction to the place where the principal obligation of a contract is being performed. Article 7(1)(b) allocates jurisdiction to the principal place of delivery or the principal place where the service is to be performed. These two subsections offer in total three divergent solutions in cases where the principal obligation or place cannot be determined: dépeçage, jurisdiction being allocated to the place the claimant so chooses, and jurisdiction of the place where the service provider is domiciled. Inconsistency aside, none of these three potential solutions is attractive.
This entire area of law could be significantly simplified if the scope of Article 7(1) did not extend to cases where the aim of ensuring a ‘close connection’ cannot generally be guaranteed. It is suggested that Article 7(1) should not be applied at all in cases concerning contractual performance to be performed in multiple Member States. Such cases should fall under Article 4, with jurisdiction centred on the place of the defendant's domicile.
A. Article 7(1)(a): The Problem of Multiplicity of Jurisdictions
The approach of Article 7(1)(a) to claims arising from contractual performance in multiple places has been summarised in Leathertex v Bodetex: Footnote 68 it is the court where the principal obligation is to be performed that has jurisdiction.Footnote 69 If the principal obligation cannot be determined, the claimant would have to commence proceedings separately at each place of performance. There are two problems with this approach.
First, what constitutes the ‘principal obligation’ can be difficult for the court seised of the case to answer. As has been frankly admitted by the CJEU in GIE Groupe, the question ‘could hardly be resolved without reference to the applicable law’.Footnote 70 If English procedural law applies, the principal obligation refers to the obligation that is the ‘real ground of complaint’.Footnote 71 This is a formulation that does no more than rehearse the phrase ‘principal obligation’ in another way. It is a far cry from achieving the aim of predictability set out in Recitals 15 and 16 of the Recast. In Leathertex, the CJEU reasoned that the claims on outstanding commission and compensation in lieu of notice were obligations of equal rank. However, the alternative interpretation advanced by the United Kingdom Government is at least arguable, in that the failure to perform the obligation to pay commission was the principal obligation, because ‘the only reason why Bodetex alleges that Leathertex terminated the agreement without notice and therefore claims compensation in lieu is that the latter had not paid the commissions’.Footnote 72 Further, determining the principal obligation is complicated if the court seised is asked to identify which is the principal obligation when there are several, each being governed by a different substantive body of law. There is no mechanism under the Recast by which these obligations could be systematically ranked.Footnote 73
Secondly, dépeçage is not an ideal solution when it is not possible to identify the principal obligation. ‘Dépeçage’ in this context refers to cases where the claims arising from the same contract have to be split up, and to be heard in more than one jurisdiction. Dépeçage can occur in two types of cases: (i) where two distinct obligations breached are of equal rank, or (ii) where the contract provides for the same obligation to be performed in different jurisdictions.Footnote 74 In the former scenario, the substantive law governing the contract could possibly point towards two jurisdictions for the two different breaches.Footnote 75 In the latter scenario, the claimant can sue in each place of performance, but only to the extent that the breach has occurred within it. The CJEU has emphasised that it is procedurally undesirable for a court to only have jurisdiction over one part of a dispute where the various parts are closely related.Footnote 76 The CJEU in Leathertex was clearly aware of this problem but brushed aside this concern by arguing that the claimant could bring the entire claim in the defendant's domicile if they wished.Footnote 77 This is questionable. If Article 4 is the preferred course, why offer the claimant the opportunity of using Article 7(1)(a) in the first place? Perhaps the availability of a dépeçage option could possibly encourage the claimant to initiate proceedings in multiple jurisdictions, thus pressurising the defendant into making a more favourable settlement.Footnote 78
It is true that Article 30(2) of the Recast provides for the consolidation of claims in a single jurisdiction. However, Article 30(2) is of no help in here, since it only applies when ‘the court first seised has jurisdiction over the actions in question’. Therefore, in a case where the specific obligation approach in Tessili results in there being two different places of performance for two distinct obligations of equal rank, the court first seised will not have jurisdiction over the other obligation, unless its jurisdiction is based on other Articles of the Recast. As rightly pointed out in Leathertex, Article 30(2) is not jurisdiction-conferring.Footnote 79
B. Article 7(1)(b): Conflicting Principles, Conflicting Solutions
1. Color Drack v Lexx International VerbtriebsFootnote 80
The starting point of this discussion is Color Drack v Lexx International Verbtriebs, a case concerning deliveries to multiple places within Austria. It was held that the court of the principal place of delivery, determined by ‘economic criteria’, had sole jurisdiction. If a ‘principal’ place could not be determined, the claimant could pursue the entire claim in any place where the goods were delivered.
First, the ambiguity of the ‘economic criteria’ test is evident. For example, in a sale of goods case adopting a different benchmark (eg volume, market value, or net profit) can potentially lead to a different principal place of performance.Footnote 81 Choosing the relevant criteria, as well as apportioning weight to each of them, are both delicate exercises, since establishing a close connection between the claim and the court having jurisdiction (Recital 16) is highly fact-sensitive. For a contract lasting for several years, what should be the time frame within which a selected criterion is examined?Footnote 82 The CJEU left the heavy lifting to the particular Member State's court seised.Footnote 83 Such an intense focus on the facts inevitably requires examination of the terms of the contract, the interpretation of which may necessitate considerations of the lex contractus yet again.Footnote 84 This is an undesirable outcome, and something which Article 7(1)(b) aimed to avoid. Without a complex assessment from the court first seised, both parties are left in limbo as regards the place where the claim should be brought.Footnote 85 Given that jurisdiction can seldomFootnote 86 be challenged once it has been assumed by a Member State's court,Footnote 87 the criterion chosen to determine whether or not a court has jurisdiction can be left unscrutinised or unchallenged. Even if a Member State's court declines jurisdiction, ruling on the appropriate criteria can create an issue estoppel.Footnote 88 The Advocate General in Color Drack explicitly rejected the ‘economic criteria’ test, since it had:
… the result of reintroducing into the system of optional jurisdiction in matters relating to a contract provided for in Regulation No 44/2001 complex criteria that the Community legislature manifestly wished to abandon. It would be very difficult for the parties to a contract to determine clearly when a delivery was a principal delivery. Such a classification would again depend on detailed rules that could be provided only by case-law.Footnote 89
It was unfortunate that the CJEU refused to accept this opinion.
The second problem is whether there is, in fact, a need for the CJEU to deliver a judgment in the first place. Color Drack concerned a case where deliveries, whilst in different locations, were all located within Austria. Nevertheless, the CJEU held that Article 7(1) was apt to determine both international and local jurisdiction, and therefore proceeded to determine which court had jurisdiction, without reference to Austrian domestic procedural rules.Footnote 90 While the adoption of autonomous linking factors under Article 7(1)(b) has rendered reference to the substantive law of a Member State largely obsolete, it is questionable whether a Member State's procedural law should equally be disregarded. One justification for this interventionist approach by the CJEU stems from the wording of Article 7(1)(b) itself, which provides that jurisdiction shall be allocated to ‘the place in’ a Member State where the characteristic obligation is to be performed. However, this overtly literal approach does not follow from the Recast's Recitals. It has never been the intention to regulate questions of jurisdiction within a Member State. One of the major aims of the Recast was to unify Member States’ differing national rules on the question of jurisdiction (see Recital 4). Determining which court within a Member State should have jurisdiction has almost nothing to do with the sound operation of the internal market and is a matter best left to the civil procedure rules of the Member State in question.Footnote 91
The third objection is that the claimant has an unfettered choice when the principal place cannot be determined, arguably an even more expansive approach than the dépeçage solution found in Leathertex. It is unclear why Article 7(1)(b) should be expanded in this way without clear textual support. This is inconsistent with the CJEU's general approach, which is that special jurisdiction rules should be interpreted narrowly.Footnote 92 Of course, the provision of a choice in Color Drack could have been defended on the ground that, since the deliveries were all made to places in Austria, the problem of foreseeability was less acute.Footnote 93 Despite Harris strongly disagreeing,Footnote 94 it was argued in the Jenard Report that it is more difficult in general ‘to defend oneself in the courts of a foreign country than in those of another town in the country where one is domiciled’.Footnote 95 In Color Drack, neither the Advocate General's opinion nor the actual judgment identified any particular problems for the defendants as a result of their being sued in St Johann-im-Pongau rather than in other places in Austria.
However, and as seen in subsequent case law, the reasoning in this case has become a convenient means by the CJEU to apply the ‘economic criteria’ test to contractual obligation performed in multiple Member States. This has the effect of further undermining the certainty and predictability of this area of law.
2. Rehder v Air Baltic CorporationFootnote 96
In Rehder, the CJEU extended the reasoning of Color Drack to services provided in different Member States. Being an airline contract, the claimant was afforded a choice between suing in the places of take-off or of landing. The CJEU reasoned that both places had a ‘sufficiently close link of proximity to the material elements of the dispute’.Footnote 97 Foreseeability was not compromised since ‘the applicant's choice is limited to two possible judicial fora’.Footnote 98 There are three problems with the CJEU's reasoning.
First, the element of choice in Color Drack arose in the context of the places of performance all being within a single Member State, a point repeatedly emphasised by the CJEU in that case.Footnote 99 In Rehder the CJEU extended the ratio of Color Drack (i) from the supply of goods to the provision of services, and (ii) from places of performance concentrated in one Member State to those in multiple Member States. While the former development may be less contentious, the same cannot be said of the latter. The only justification given for doing so was in paragraph 37 of the judgment, which did no more than recite the various objectives of the Recast and claim that a differentiated approach would run contra to them.
Secondly, and more importantly, the nature of the particular contract in Rehder—an air-transport contract with two distinct boarding and landing locations—made the identification of the places of performance relatively easy. However, in paragraph 38 the CJEU made a fairly sweeping statement, stating that ‘where there are several places at which services are provided in different Member States, it is also necessary to identify the place with the closest linking factor between the contract in question and the court having jurisdiction’. This exposes an underlying lack of attention as to what foreseeability actually means. The aim is insufficiently protected if the defendant can only foresee the number of places in which they might be sued, no matter how large that number potentially could be. It is the ability to foresee the actual forum which facilitates the aim of promoting the ‘sound administration of justice’ under Recital 16. The claimant having different jurisdiction choices results in disparate pretrial preparations and cost regimes, thus affecting the defendant's litigation or settlement strategies. Foreseeability is unduly stifled when the claimant is treated so favourably, having a free-standing choice, which is close to forum shopping, in cases where there are in multi-State contractual arrangements.
Thirdly, consider the following example. The German Philharmonic Orchestra agrees to perform 30 concerts, 10 in each of London, Paris and Brussels. Litigation is brought only in relation to the concerts in Paris. Would the claimant nevertheless be free to choose between the three jurisdictions following Rehder? This example differs from Rehder in one essential aspect: the obligation in Rehder was indivisible, since ‘air transport consists, by its very nature, of services provided in an indivisible and identical manner from the place of departure to that of arrival of the aircraft’.Footnote 100 There is currently no case law on whether the French court, and the French court alone, would have jurisdiction in the above scenario.Footnote 101 For this to be the case, the concerts in Paris would have to be treated as a severable ‘provision of service’ under Article 7(1)(b). However, there are at least two problems with this interpretation. First, this may not accord with the express wording of the contract or the intention of the parties, since treating the provision of service at each location as severable may influence the application of rules on equitable set-off. Secondly, a separate supplementary rule under Article 7(1)(b) would also have to be developed in order to allocate jurisdiction to the Member State where that severable part of the service is rendered. Also, what would happen if a claim was brought in relation to the concerts in both Paris and Brussels, but not London? Would the claimant be allowed a free-standing choice, or would dépeçage be the only solution? These difficult questions remain unanswered.
3. Wood Floor Solutions v Silver TradeFootnote 102
The rationes decidendi of Color Drack and Rehder were applied in Wood Floor. In this case, the claimant sought damages for the termination of a multi-State agency contract. The CJEU held that the national courts should identify the place of the main provision of service.Footnote 103 Broad factors, such as ‘the time spent in those places and the importance of the activities carried out’, were provided as guidelines.Footnote 104 If the principal place could not be determined, the court where the agent (claimant) was domiciled would have jurisdiction. There are three problems with the reasoning of the CJEU.
First, the CJEU departed from Rehder and Color Drack regarding the appropriate response if the principal place of performance cannot be determined. Advocate General Trstenjak in Wood Floor was unenthusiastic about the option of allowing the claimant a free choice, since this was inconsistent with the aim of predictability and created the danger of forum shopping.Footnote 105 Therefore, the CJEU departed from Rehder and decided that jurisdiction lay with the domicile of the service provider. While this solution avoids the problem of forum shopping and offers predictability as compared to Rehder, the CJEU's judgment did not explain why it departed from its earlier decision. The only substantive distinction between the two cases can be found in Advocate General Trstenjak's opinion, where he pointed out that in Rehder, and unlike in Wood Floor, ‘the applicant had only two places at which it could bring proceedings’.Footnote 106 However, this cannot be a complete answer as to when and why a choice is offered to the claimant. It begs the question of the role played by the number of possible fora, which is itself contingent on the particular contract.
Secondly, the claimant can effectively sue in their own domicile if the principal place cannot be determined, thus offering the claimant a benefit the CJEU has repeatedly cautioned against—forum actoris. As observed by the Italian Supreme Court in Socetà Kretschmer v Muratori, claimants tend to pursue cases before his own court, not only because of economic convenience but also for ‘compulsive psychological reasons’.Footnote 107 Therefore, while forum actoris is not necessarily objectionable per se, Advocate General Lenz in Custom Made argued that ‘where an interpretation of Article 5 et seq. has the result that the court having jurisdiction is that of the plaintiff's domicile, particular care should be taken to see whether that interpretation accords with the aim of the provision in question’.Footnote 108
This leads to the third point, which is that awarding jurisdiction to the court where the agent is domiciled does not accord with the aim of achieving a close connection. The distinctive feature of Article 7(1)(b), applied consistently from Car Trim to Color Drack, is to determine the place of performance by factual criteria—ie, the place where the services were actually provided. Therefore, establishing jurisdiction on the basis of domicile presumes that at least one of the places where the services are provided happens to be the place of domicile. The CJEU is, of course, right in that ‘the agent will in all likelihood provide a substantial part of his services there’.Footnote 109 However, it is clearly possible for an agent only to carry out preparatory work in the place of domicile, without any ‘end-product’ taking place there. This weakens the connection between the place of the agent's domicile and the places of performance of the service.
Magnus and Mankowski have thus argued that even ‘organising, re-arranging and in particular administrating contracts’ can be treated as a ‘provision of service’ under Article 7(1)(b).Footnote 110 This interpretation is not without problems. Following Falco, the characteristic obligation of a service contract is that the party ‘carries out a particular activity in return for remuneration’.Footnote 111 While it is therefore possible that preparatory work forms part of the consideration triggering a reciprocal right to remuneration, whether this is always so depends on the construction of the particular service contract in question. This necessitates a determination by the court seised, according to the lex contractus, of whether the preparatory work is merely an auxiliary element of the contract. This re-invites the consideration of the substantive law of a Member State by the backdoor.
C. Back to Simplicity—The Article 4 Solution
For contractual claims concerning obligations to be performed in multiple places, Article 7(1) should not be applied at all. Article 4 should be applied instead. According to Besix SA v WABAG, when interpreting the phrase ‘place of performance’ in what is now Article 7(1)(a), ‘a single place’ must be identified.Footnote 112 As argued by Advocate General Siegbert Alber in Besix, according to all the applicable language versions of the Recast, the forum of the place of performance refers to a single place, and not to several places.Footnote 113 While Besix concerns the interpretation of what is now Article 7(1)(a), there is no reason why the same words in Article 7(1)(b) warrant a different interpretation. Therefore, in a case involving contractual obligations to be performed in multiple Member States, Article 7(1) should not be applied at all. Instead, jurisdiction should be centred at the domicile of the defendant under Article 4.
Unfortunately, neither Besix nor the argument in favour of resorting to Article 4 was considered in the main judgments of Color Drack, Rehder and Wood Floor. The most extensive discussion of the potential applicability of Besix in multi-State contractual disputes was by Advocate General Trstenjak in Wood Floor, where he gave three reasons for not applying it in cases involving multiple places of performance in different Member States. Each of these arguments against doing so will be considered in turn.
1. Adopting Besix would render the special jurisdiction rule nugatory
The first argument advanced by the Advocate General was that adopting Article 4 would negate the application of Article 7(1)(b) to many contracts where goods or services are provided in several Member States.Footnote 114 This would be contrary to the purpose of Article 7(1)(b), which is to determine the place of performance of sale of goods and supply of services contracts autonomously. Similarly, some commentators contend that disapplying Article 7(1) in cases involving multi-State obligations deprives the claimant of an alternative choice of forum when the claim is clearly contractual.Footnote 115 Since the Recast provides that the place of contractual performance is a ground of special jurisdiction, it would be consistent with its aims to at least try to apply Article 7(1), before falling back on Article 4.
There are four responses to this argument. First, it is common ground that the state of the CJEU jurisprudence as it stands, and as outlined in section III.B above, is unsatisfactory. However, it is questionable whether any amendments by the European regulator, or further clarifications by the CJEU, could actually provide clarity to the economic criterion test. Various suggestions as to the factors that could be relevant in determining the principal place of performance have been made: Advocate General Trstenjak in Wood Floor listed at least nine.Footnote 116 However, the fluidity of the economic criteria test became apparent when the Advocate General was forced to add numerous caveats to the relevant factors, most notably to the weight to be given to ‘turnover’ as an economic criterion.Footnote 117 Magnus and Mankowski suggested that the contract price of the goods or services provided is the best economic criterion, since it is readily ascertainable and predictable for both parties.Footnote 118 While this may be a plausible starting point, this would open a Pandora's box of preliminary rulings, since parties would be encouraged to challenge the criterion adopted by the court seised. It is difficult to see how further clarifications by the CJEU could introduce certainty to such a fact-sensitive exercise.
Secondly, in Wood Floor the CJEU did not consider whether the phrase ‘the place in a member state’ (emphasis added) in Article 7(1)(b) supports the proposition that the Article is only applicable to contracts where the place of performance is within a single Member State. The CJEU's failure to give proper attention to the plain wording of the phrase was in marked contrast to its adoption of a literal interpretation of the same phrase in Color Drack. In Color Drack, the CJEU's principal justification for allocating jurisdiction to a particular place within Austria rested on a literal interpretation of the word ‘place’. When the same phrase presents difficulties in Wood Floor, it seems that the Court conveniently sidestepped the problem.
Thirdly, in Wood Floor Advocate General Trstenjak argued that a literal interpretation of Article 7(1)(b) was contrary to the purpose of Article 7(1) and that a ‘teleological and schematic interpretation’ should be adopted instead.Footnote 119 However, it is highly questionable whether a purposive interpretation supports the adoption of the ‘economic criteria’ approach. Under this approach, the principal place of performance identified may not be the place that has the closest connection to the dispute. For example, in a multi-State sale of goods contract involving defective deliveries to five Member States, where the percentages of good received in each were 23 per cent, 22 per cent, 21 per cent, 18 per cent and 16 per cent, over 70 per cent of the contract may have little connection (eg proximity of evidence and witnesses) with the Member State assuming jurisdiction.Footnote 120 The place where 23 per cent of the goods are delivered would have jurisdiction under the current approach, but it can hardly be considered to be the centre of the dispute. Viewed in the light of this simple example, extending Article 7(1)(b) to cases involving multiple places of performance neither accords with the text of the provision nor the purpose of the special jurisdiction rule.
Fourthly, it is pertinent to recognise the proper relationship between Articles 4 and 7(1). Magnus and Mankowski describe Article 7(1) as an ‘advantage’ offered to the claimantFootnote 121 and that such an option is ‘in principle guaranteed’ by the relevant provision.Footnote 122 Some commentators have refused to interpret Article 7 restrictively.Footnote 123 However, this goes against the repeated dicta of the CJEU,Footnote 124 and Recital 15 expressly provides that ‘jurisdiction is generally based on the defendant's domicile … save in a few well-defined situations’. While it is true that Article 7(1) aims to provide an alternative forum centred at the place of performance of the contractual obligation, it does not follow that the CJEU should stretch Article 7(1) and give the claimant the benefit of choosing between Articles 4 and 7(1), when giving the claimant such an ‘advantage’ does not accord with the aim of ensuring a close connection between the court and the claim under Recital 16.
No doubt there are easy cases in which the ‘economic criterion’ test produces straightforward answers. Consider a sale of goods contract involving delivery of 100 identical goods of the same contract price, market price and resale price to two different Member States, 99 to one country and 1 to another. The apparent simplicity of this example shows just how many assumptions are required before the parties to the contract can intuitively point to a definitive principal place of performance. Given the complexities of international trade, decisions in most cases will depend on an intricate examination of the weight attached to various criteria. Parties would have to engage in mental gymnastics and go to the court they consider to be the principal place of performance in order to test out their predictions. As early as 1979, when Professor Schlosser published his report on revising the Brussels Convention to accommodate the accession of Denmark, the United Kingdom and Northern Ireland to the Convention, it had been argued that the claimant ‘should not have to waste his time and money risking that the court concerned may consider itself less competent than another’.Footnote 125 The principle of certainty is ill-served by ex post facto assessments.
The restrictive interpretation of Article 7(1) suggested here would, no doubt, create some problems of its own. Consider the example of the German Philharmonic Orchestra in section III.B above. As already argued, in a case where there are multiple places of performance, but the claimant has only complained about defective performance in one place, there is currently no guidance from the case law on whether jurisdiction would be allocated to that place alone, or whether it would still be allocated to the principal place of performance. It is suggested that even if the restrictive interpretation of Article 7(1) proposed by this article were to be adopted, Article 7(1) should still remain applicable when the claimant only claims against defective performance in one Member State. This is so even if the contractual obligation was in fact to be performed in multiple Member States.
Of course, this suggestion could invite abusive litigation tactics by the claimant. For example, if the defendant has supplied defective goods to England, France and Germany, the claimant might decide to bring three separate sets of proceedings under Article 7(1) in each of the three States where deliveries due in relation to each. One solution to this problem could be Article 30(1) of the Recast, which provides that ‘where related actions are pending in the courts of different Member States, any court other than the court first seised may stay its proceedings’. In The Tatry, the CJEU held that the meaning of ‘related actions’ must be interpreted broadly to avoid the risk of conflicting decisions.Footnote 126 In the scenario outlined above, the proceedings in England, France and Germany are highly likely to be related actions, and the Member State courts other than the court first seised can exercise their discretion to stay the proceedings pending the resolution in the court first seised. This avoids the defendant having to simultaneously defend proceedings in multiple jurisdictions. Of course, this solution does not compel the claimant to consoldiate all proceedings in one Member State. However, on balance, dépeçage is still a more desirable solution than reverting to the test of determining the principal place of performance by ‘economic criterion’. Even though one of the aims of the Recast is to consolidate proceedings in one Member State, unprincipled consolidation is an approach that the CJEU should actively avoid.
2. Article 7(1)(a) should have been applied first before resorting to Article 4
The second argument offered by the Advocate General in rejecting the Article 4 solution is that even if Article 7(1)(b) is not to be applied for contracts with multiple places of performance, Article 7(1)(a) should be the fall-back provision in light of Article 7(1)(c).Footnote 127 Only if it is impossible to determine the place of performance under Article 7(1)(a) should Article 4 be applied.
The complexities of the jurisprudence under Article 7(1)(a) have been outlined above. The difficulties of determining the place of performance with reference to the lex causae in Article 7(1)(a) (section II.A) as well as the danger of fragmentation of jurisdictions if determining the principal place of performance is impossible (section III.A) are both strong reasons for not applying Article 7(1)(a), even if Article 7(1)(b) is not to be applied. The default rule under Article 4(1) provides a more certain and reliable route for determining jurisdiction.
English domestic jurisprudence offers mixed responses as to whether the Article 4 solution in Besix should be applied to contractual cases involving multiple places of performance. In Canyon Offshore Ltd v GDF Suez E&P Nederland BV, a Queen's Bench Division case, Canyon was the subcontractor of GDF. When the principal contractor failed to pay Canyon, GDF assured Canyon that it would pay for the work done. Canyon initiated proceedings against GDF to enforce this assurance.Footnote 128 There are two points to note about the facts of this case. First, this contract was classified as falling under Article 5(1)(a) (now Article 7(1)(a)), since the characteristic obligation of the contract was ‘GDF's obligation to pay Canyon, not the provision of trenching services by Canyon’, and GDF did not step into the shoes of the principal contractor.Footnote 129 Secondly, the place of performance of the payment obligation was left floating, to be performed either in England or Scotland. Taken together, this is a good test case to evaluate the solution offered by Article 7(1)(a) when there is more than one places of performance.
Unfortunately, the solution proposed in the case is far from satisfactory. In essence, Judge Mackie refused to apply Besix, and instead relied on ‘an evolution’ from CJEU jurisprudence (eg Wood Floor; Color Drack and Rehder) which permitted ‘claimants to have a choice, within art. 5, of where to sue provided that there is sufficient proximity and predictability’.Footnote 130 This reasoning is problematic for two reasons.
First, it is questionable whether and how this evolution can be deduced, since the decision and reasoning of Besix were not considered in the main judgment in Wood Floor and Rehder at all. Secondly, Judge Mackie considered the aims of ‘proximity’ and ‘predictability’ together, without sufficiently distinguishing between them. The meaning of the two concepts was not specifically discussed, despite Judge Mackie acknowledging that ‘[t]he scope of the words … has not yet been much worked out’.Footnote 131 In this case, considerations of proximity and predictability arguably pull in different directions. As regards proximity, Judge Mackie was right that the essence of what GDF was required to do under the contract, akin to a guarantee, was to make a payment to Canyon.Footnote 132 Therefore, the places of performance (Scotland or England) could both be considered as the centre of gravity of the contract.
The problem concerns the reasoning regarding foreseeability. Judge Mackie argued that ‘a normally well-informed defendant would be reasonably able to foresee that, apart from at his domicile, he might be sued in either of the places where payment was required to be made under the alleged contract’.Footnote 133 First, the fact that there were only two possible places of performance weighs heavily in favour of the decision. However, this does not provide a principled solution as to when the number of possible fora becomes so large that it is no longer foreseeable by the defendant. This creates a sorites paradox. There is still no ‘single place of performance’ under the contract, and whether the number of possible fora is two or more than two is purely a question of degree.Footnote 134 Secondly, it fails to address the importance of foreseeability as to the precise forum to which the defendant can be sued. Even as between England and Scotland, there are differences concerning rules of civil procedure (eg disclosure, settlement offers and caveats) which affect a party's pretrial litigation strategy and actions.
Therefore, resorting to Article 7(1)(a) would not help clarify the CJEU's current jurisprudence. Instead, the problems concerning Article 7(1)(b) would continue, and the solution in Canyon is a long way from achieving the aims of certainty and predictability set out in the Recast.
By contrast, in the earlier English Court of Appeal decision of Mora Shipping Inc v Axa Corporate Solutions Assurance Sa,Footnote 135 insurers domiciled in Member States undertook, in the form of a guarantee, to pay general average contributions to a shipowner, payment to be made either to the shipowner outside the European Union or to the shipowner's agents in London. The Court of Appeal ruled that as a matter of construction, there was no obligation to make payment in England. In particular Clarke LJ cited Besix when establishing more generally that ‘if the obligation was required to be performed in more than one jurisdiction, or could be performed in more than one jurisdiction, no single jurisdiction would be established under Art.5.1’.Footnote 136 There can hardly be a more succinct summary of Besix, and this offers a predictable solution to this area of law.
In Canyon, it is unfortunate that Judge Mackie refused to refer the case to the CJEU for determination.Footnote 137 Currently, there is no direct CJEU authority in which Article 4(1) has been applied rather than Article 7(1) in a case concerning obligations to be performed in multiple Member States. The only support for this approach can be found in Advocate General Bot's opinion in Color Drack, which deserves to be quoted in full:
The solution adopted in the Besix judgment could be transposed, in my view, if the places of delivery were in different Member States. In such a case, I consider that the optional jurisdictional rule laid down in Art.5(1)(b) of Regulation 44/2001 would not be applicable, because the objective of foreseeability could not be achieved since the courts that potentially had jurisdiction under this provision were in several Member States. I am also of the opinion that, in such a scenario, taking account of the objective of foreseeability of the rules on jurisdiction pursued by Regulation 44/2001 and the difficulties of applying Art.5(1)(a) of that regulation, the latter provision would not be applicable either. As the latter provision is purely subsidiary, its application should be confined to the situation envisaged by the Commission in its draft regulation of 1999, in other words where the place of delivery of goods or provision of services is in a third state. If the goods are delivered or the services provided in several Member States, the court with jurisdiction could not, in my opinion, be other than that of the domicile of the defendant, in accordance with the principle set out in Art.2 of Regulation 44/2001.Footnote 138
3. Positive versus negative obligations
The last argument of Advocate General in Wood Floor for not adopting Besix was that it could be distinguished.Footnote 139 Besix concerned the breach of a negative obligation. The parties had agreed to ‘act exclusively and not … commit themselves to other partners’, so theoretically there was no territorial limit to their obligation.Footnote 140 It was therefore argued that Besix had no bearing on cases involving positive obligations to be performed in multiple Member States. It is unfortunate that neither the Advocate General nor the CJEU in Wood Floor appreciated that the reasoning of Besix can in fact be applied to both positive and negative obligations. First, in Besix the CJEU did not limit its reasoning to cases involving negative obligations. It said that Article 4(1) should be applied whenever the factual pattern results in a ‘multiplicity of competent courts’.Footnote 141 In particular, the CJEU noted that ‘in a case such as that now referred, characterised by a multiplicity of places of performance of the contractual obligation in question, a single place of performance has to be identified’.Footnote 142
Secondly, the CJEU rehearsed the need for jurisdictional rules to be foreseeable, allowing ‘a normally well-informed defendant reasonably to foresee before which courts, other than those of the State in which he is domiciled, he may be sued’.Footnote 143 In particular, the CJEU noted that:
where parties have agreed a contractual obligation not to do something, applicable without any geographical limit, [Article 7(1)] does not avoid a multiplicity of competent courts, since it leads to the result that the places of performance of the obligation in question are in all the contracting states. It also involves the risk that the claimant will be able to choose the place of performance which he judges to be most favourable to his interests.Footnote 144
Consider an agency contract in which the defendant is required to act as the exclusive marketing agent of the claimant's products in all Member States of the European Union. A distinction between the positive and negative obligations in this contract would mean that a claim concerning the defective provision of service would engage Wood Floor and the complicated assessment it entails, whereas the breach of the exclusivity clause could potentially engage Besix instead. This cannot be right as a matter of principle. Both obligations involve performance, either positive or negative, in multiple Member States. The two concerns noted by the CJEU, namely the lack of foreseeability and the danger of forum-shopping, are equally applicable to both obligations.
Thirdly, in Besix the CJEU was fully aware that one of the overarching aims of the special jurisdiction rules was to ensure a close connection between the court and the claim. It therefore noted that ‘Art. 5(1) of the Brussels Convention is not apt to apply… where it is not possible to determine the court having the closest connection with the case by making jurisdiction coincide with the actual place for performance of the obligation…’.Footnote 145 If one of the reasons why Article 7(1) should not be applied in cases of negative obligations is because it is not possible to identify the court that has the closest connection with the case, the line of authority from Color Drack to Wood Floor could only serve to fortify the conclusion that this concern equally applies to positive obligations performed in multiple Member States.
The above arguments show that the attempted differentiation by the Advocate General between positive and negative obligation is a distinction without a difference.
IV. CONCLUSION
The arguments presented in this article ultimately raise a question concerning the priority of the aims of the Recast: in cases where the general aim of certainty and predictability conflict with the specific aim of a close connection between the court and the claim, which aims should Article 7(1) prioritise? Answering this question helps determine the ambit of Article 7(1).
The first suggestion made in section II, which involves amending Article 7(1), favours an autonomous linking factor approach when determining jurisdiction. During the review of the Brussels Convention, there were at least three different views concerning the then Article 5(1): complete abolition of the contract jurisdiction rules, maintaining the status quo, and concentrating jurisdiction at the place where the characteristic obligation is performed.Footnote 146 The current Article 7(1)(b) represents a last-minute compromise between the three camps.Footnote 147 This is an incomplete reform. This article therefore suggests a further reform, whereby autonomous linking factors are adopted based on the characteristic obligation of each type of contract. No doubt there would still be cases where a breach of the non-characteristic obligation results in jurisdiction being allocated to a court that is not most proximate to the dispute. It would then have to be asked if this default rule, developed in the context of Recast, does enough to achieve the aim of ensuring a close connection. The Recast seeks to allocate jurisdiction to an appropriate forum, even if it is not always the most suitable forum.Footnote 148 Centralising jurisdiction at the place where the characteristic obligation of the contract is performed strikes a delicate balance between the close-connection aim under Recital 16, and the overarching aims of certainty and predictability under Recital 15.
The second suggestion is made in section III, concerning contractual obligations performed in multiple Member States, and prioritises certainty of application and predictability over a close connection between the forum and the dispute. Commentators have generally been reluctant to abandon Article 7(1) in favour of Article 4. Magnus and Mankowski believe that this is no different from ‘committing suicide for fear of death’.Footnote 149 However, this article has demonstrated why attempts by the CJEU to clarify the test have been futile. It has proven extremely difficult to reach a result that ensures a close connection between the forum and the dispute, without at the same time sacrificing the aims of certainty and predictability that the Recast prides itself on. This explains why, if a choice has to be made, it is better to favour a solution that accords with the general aims of the Recast. One only has to look at the first stage of the Spiliada test to realise the fluidity of a multi-factor approach (eg factual connection, legal connection, rules on lis pendens), an approach which deliberately leaves the court seised with discretion as to the weight to be accorded to each factor.
The result is that the enormous case law on forum conveniens remains no more than illustrative of how the doctrine works in practice. This is why this article takes a dim view of possible clarifications of the meaning of ‘principal obligation’ or ‘principal place of business’. An uncertain test gives rise to the temptation of speculative litigation in order to gain procedural advantages.Footnote 150 Therefore, while Article 4 may not always result in the allocation of jurisdiction to the most appropriate forum in a contractual dispute, it cannot be denied that allocating jurisdiction to the place where the defendant is domiciled is at least always an appropriate solution. Macroscopically, the uncertainties associated with determining the principal place of delivery or service by ‘economic criteria’ under Article 7(1)(b) has been cited as one of the reasons why Article 7(1)(a) should be retained. Resorting to Article 4 in these complex cases thus removes a reason for not harmonising Articles 7(1)(a) and 7(1)(b) in accordance with the latter's autonomous linking factor approach.
Unless and until the CJEU realises that Article 7(1) need not be expanded beyond what Recital 16 originally intended, uncertainties will continue to plague Member State's courts, an unfortunate state of affairs in an area where certainty is the cardinal principle.
