Prüller-Frey: The CJEU on Direct action provided for by national law against the civil-liability insurer
Case-law on Rome II (the law applicable to non-contractual obligations) is only slowly picking up so almost anything coming out of the CJEU is met with excitement. Like Ergo Insurance (so far only the AG’s Opinion), Prüller-Frey concerns insurance contracts. In this case, direct action against an insurer, by the victim of an air traffic accident.
The victim sues in Austria, on the basis of Article 6 or, alternatively, 11 of the Brussels I Regulation (old: Regulation 44/2001). Applicability or not of the Montreal Convention (for the Unification of Certain Rules for International Carriage by Air) and the EU’s implementation of same, is less relevant for this posting. At stake was mostly Article 18 of the Rome II Regulation, which reads
The person having suffered damage may bring his or her claim directly against the insurer of the person liable to provide compensation if the law applicable to the non-contractual obligation or the law applicable to the insurance contract so provides.
The lex contractus is German law. This was so chosen by the insured, Norbert Brodnig, and the insurance company, Axa Versicherung AG. German law does not provide for such direct action. But Spanish law, the lex locus damni (which applies between Prüller-Frey and Brodnig), does. The insurance company calls upon the absence of the action in German law, to reject Prüller-Frey’s action. Szpunar AG and the CJEU itself simply point to the clear language of Article 18: this is not a conflict of laws rule that determines the law applicable between victim and insurer: the insurance company’s obligations will continue to be subject to the lex contractus. Article 18 is simply an alternative connecting factor for the very possibility of direct action against the insurer. Spanish law is the law applicable to the non-contractual obligation and if Spanish law allows for such direct action, then that is enough for there to be one.
Thank you for the team at Dechert to remind us of the potential that BITs may be used to pursue proactive, rather than just reactive environmental litigation. A word of explanation: Bilateral Investment Treaties, in particular their investor-state dispute settlement mechanisms, are currently under a lot of pressure following the public outcry over the TTIP negotiations. Allowing private investors to sue countries that roll out regulation, using vague principles of protection of property, is seen by many as a form of corporate bullying.
Dechert’s briefing however reminds us firstly, specifically vis-a-vis stubborn air pollution in the Indonesia area, that States may carry responsibility in line with Trail Smelter’s nec utere tuo principle. The possibility for individuals (as opposed to neighbouring States) suing on that basis, is of course complicated by the mechanism of (absence of) direct effect of huge chunks of international environmental law. That is where investor-state can come in handily. Such as in Allard v Barbados at the Permanent Court of Arbitration. Dechert’s summary of that case reads ‘the Canadian owner of an eco-tourist facility in Barbados is currently suing the Government of Barbados for an alleged breach of the full protection and security provision (among other provisions) in the Canada- Barbados bilateral investment treaty. Peter Allard argues in his claim that Barbados breached its treaty obligations by failing to enforce its domestic environmental laws, which he alleges led to the environment being spoilt and a loss of tourist revenues at his eco-resort’.
A timely reminder of the good BITs can do, just before I am to speak (again) tomorrow on TTIP and why EU citisens are so suspicious of it.
Of tractors and trailers. Insurance contracts, subrogration, contracts and torts. Sharpston AG on the scope of Rome I and II.
First, a quick heads-up on precedent: the difference between ‘contract’ and tort’ in European private international law is crucial, as regular readers of this blog will have observed. Crucial, yet the concept is left undefined in the Brussels I (and Recast) Regulation (which has a different special jurisdictional rule for both), the Rome I Regulation on applicable law for contracts, and the Rome II Regulation on applicable law for torts. Undefined, for these foundational elements of private law are outside the reach of legal and political compromise in the legislative process. Yet courts of course do have to apply the rules and in doing so, have to distinguish between both.
The CJEU pushes an ‘autonomous’ EU definition of both concepts which in the past has led to the seminal findings in Jakob Handte (C-26/91) and Kalfelis. In Handte the Court held: the phrase ‘matters relating to a contract [ ] is not to be understood as covering a situation in which there is no obligation freely assumed by one party towards another.’ (the double negative exercised scholarship for some time). In Kalfelis the Court had earlier defined ‘tort’ as ‘all actions which seek to establish liability of a defendant and which are not related to a ‘contract’ within the meaning of Article 5(1).’ (5(1) has become 7(1) in the Recast).
Is the relationship between two insurers, having covered liability for a towing vehicle cq a trailer, each subrogated in their insured’s rights and obligations, one of them currently exercising a claim against the other in partial recovery of the compensation due to the victim, non-contractual?
Per Kalfelis, tort as a category is residual. Sharpston AG’s starting point in Joined Cases Ergo Insurance and AAS Gjensidige Baltic, Opinion issued yesterday, therefore is to examine whether the recourse action is essentially contractual in nature. In the negative, the action is non-contractual. The case is evidently made more complex by the underlying relationships between insurer and insured, and the presence of subrogration. In question is not therefore the relationship between the insurer and the victim: this is clearly non-contractual. The question is rather whether the action of one insurer against the other is contractual in nature, given the contractual relationship between insurer and insured, cq the non-contractual relationship between the insured and the victim.
Sharpston AG first gets two issues out of the way. Lithuania (both referred cases are pending in Lithuanian courts) is a signatory State to the Hague Convention on the law applicable to traffic accidents, which is left unaffected by Rome II by virtue of Article 28. However the Convention itself holds that it does not apply to recourse action and subrogation involving insurance companies. Further, a suggestion that Directive 2009/103 (relating to insurance against civil liability in respect of the use of motor vehicles, and the enforcement of the obligation to insure against such liability) includes a conflict of laws (applicable law) rule which is lex specialis vis-a-vis the Rome Regulation, was quickly dismissed. Indeed the Directive’s provisions do not indicate whatsoever that they can be stretched.
Then comes the core of the issue, the nature of the relationship underlying the claim. This, the AG suggests, is contractual. Relevant precedent referred to includes Brogsitter and OFAB. Essentially the AG puts forward an ancestry test: what is the ancestry of the action, without which the parties concerned would not be finding themselves pleading in a court of law?: she uses ‘centre of gravity’ (‘the centre of gravity of the obligation to indemnify is in the contractual obligation’); ‘rooted in’ (‘the recourse action by one insurer against the other…is rooted in the contracts of insurance’); and ‘intimately bound up’ (‘[the action] is intimately bound up with the two insurers’ contractual obligation‘). (at 62).
Incidentally, in para 20 of her Opinion the AG refers, in giving context, to the difference between Lithuanian and German law (the accidents both occurred in Germany) as regards the limitation periods for bringing a recourse action. In Rome II, limitation periods are included in Article 15 as being covered by the lex causae; ditto in Article 12 of Rome I. This pre-empts discussion on the matter for whether limitation periods are covered by lex fori (as a procedural issue) or the lex causae is otherwise not necessarily the same in all Member States.
If the CJEU confirms, preferably using the terminology of its AG, the tort /contract discussion in my view will have been helpfully clarified.
Postscript 24 September 2015: the incoterm ‘ex works’ was at issue in Cimtrode The Electrode Company GmbH v Carbide BV at Gerechtshof ‘s-Hertogenbosch. Judgment (on appeal) was issued 1 September 2015. The court held inter alia that whether the incoterm was actually part of the agreement between parties, could only be judged in accordance with the lex causae. The agreement was a verbal agreement, and any choice of court which one of the parties claimed had been made, had not been confirmed in writing. Reference to relevant standard terms and conditions on the invoices sent later, following execution of the agreement, could not, the court held, be regarded as confirmation of the choice of court.
In Roonse Recycling & Service BV v BSS Heavy Machinery GmbH, the Court at Rotterdam first of all discussed the factual circumstance of a possible choice of court agreement between parties, in favour of the courts at Eberswalde (Germany). Such choice of court is made in the general terms and conditions of seller, BSS. Whether parties had actually agreed to these, was in dispute. Roonse suggests the reference on the front page of the order form to the general terms and conditions on the backside (‘umseitiger‘) was without subject for that back page was blank. The court therefore suggests that agreement depends on whether, as was suggested, the standard terms and conditions were attached (stapled, presumably) to the order form. Whether this was the case is a factual consideration which Rotterdam does not further entertain for even if the choice of court agreement is invalid, the court found it would not have jurisdiction under the only other alternative: Article 7(1) special jurisdictional rule for ‘contracts’.
Roonse suggest that the parties had agreed that the contract, a delivery of good, is performed in Rotterdam for that, it argues, is where delivery took place per the Incoterm CPT (carriage paid to). The CJEU has flagged the inconclusive effect of the mere use of Incoterms for the purposes of finding an agreement between parties under Article 7, in Electrosteel Case C-87/10 (in that case with respect to the use of ‘ex works’) and has generally insisted, per Car Trim Case C-381/08 that the courts need to make reference to all relevant terms and conditions in the agreement so as to determine the place of delivery.
Rotterdam in casu held the Incoterm CPT Rotterdam as being mostly a reference to costs, not place of delivery. Where it is impossible to determine the place of delivery on that basis, without reference to the substantive law applicable to the contract, that place at least for the sale of goods, the CJEU held, is the place where the physical transfer of the goods took place, as a result of which the purchaser obtained, or should have obtained, actual power of disposal over those goods at the final destination of the sales transaction. In casu, this was found to be in the geographical jurisdiction of the courts at Den Haag. Given that Article 7(1) does not merely identify the courts of a Member State but rather a specific court within a Member State, Rotterdam has no jurisdiction.
The case is a good reminder of the limited power of Incoterms to determine jurisdiction. Better have a specific choice of court clause (which here may or may not have presented itself here in the general terms and conditions of seller).
The exam season is over, otherwise Goldhar v Haaretz would have made a great case for comparative analysis. Instead this can now feed into class materials. This is an interlocutory judgment on the basis of lack of jurisdiction and /or abuse of process. Plaintiff lives in Toronto. He is a billionaire who owns i.a. Maccabi Tel Aviv. (Chelsea’s first opponent in the Champions League. But that’s obviously an aside). Mr Goldhar visits Israel about five or six times per year. Defendant is Haaretz Daily Newspaper Ltd. which publishes Haaretz, Israel’s oldest daily newspaper (market share about 7%). It also publishes an English language print edition. Haaretz is published online in both English and Hebrew.
Haaretz published a very critical article on Mr Goldhar in November 2011. The print version was not published in Canada, in either English or Hebrew. However, Haaretz was made available internationally on its website in Israel in both Hebrew and English – the judgment does not say so specifically however I assume this was both on the .co.il site – even if currently Haaretz’ EN site is available via a .com site.
Information provided by the defendants reveals that there were 216 unique visits to the Article in its online form in Canada. Testimony further showed that indeed a number of people in Canada read the article – this was sufficient for Faieta J to hold that a tort was committed in Ontario and thus a presumptive connecting factor exists. Presumably this means that the court (and /or Canadian /Ontario law with which I am not au fait) view the locus delicti commissi (‘a tort was committed’) as Canada – a conclusion not all that obvious to me (I would have assumed Canada is locus damni only). Per precedent, the absence of a substantial publication of the defamatory material in Canada was not found to be enough to rebut the finding of jurisdiction.
Forum non conveniens was dismissed on a variety of grounds, including applicable law being the law of Ontario (again Ontario is identified as the locus delicti commissi: at 48). Plaintiff will have to cover costs for the appearance, in Canada, of defendants’ witnesses. Importantly, plaintiff will also only be able to seek damages for reputational harm suffered within Canada.
I can see this case (and the follow-up in substance) doing the rounds of conflicts classes.
In particular of course, a contract for employment needs to be distinguished from a contract for the provision of services. ‘Contract of employment’ was addressed in the abstract by the CJEU in Shenavai, Case 266/85, where the Court identified a double requirement for it referred to the need for a contract to be qualified as a contract of employment: there must be durable relation between individual and company: a lasting bond, which brings the worker to some extent within the organisational framework of the business; and a link between the contract and the place where the activities are pursued, which determines the application of mandatory rules and collective agreements. However precedent value of Shenavai for the Brussels I and recast Regulation is necessarily incomplete, for a the time employees as a protected category did not yet exist in the Regulation and the Court’s findings on contracts of employment took place within the need to identify a ‘place of performance’ under the Brussels Convention’s special jurisdictional rule on contracts.
The Jenard and Möller report to the 1988 Lugano Convention suggested the relationship of subordination of the employee to the employer.
In Holterman the Court throws into the mix reference to its interpretation of secondary EU law on health and safety at work as well as European labour law, holding that ‘the essential feature of an employment relationship is that for a certain period of time one person performs services for and under the direction of another in return for which he receives remuneration’ (at 41).
Consequently the national courts now have quite a number of criteria which need to apply in practice: it is not for the CJEU to do so in an individual case. In Holterman the Court does seem to suggest that once a worker finds himself qualified as an employee, for the purposes of the application of the Jurisdiction Regulation, that qualification will trump any other roles which that individual may play in the organisation (at 49: ‘the provisions of Chapter II, Section 5 (Articles 18 to 21) of Regulation No 44/2001 must be interpreted as meaning that they preclude the application of Article 5(1) and (3) of that regulation, provided that that person, in his capacity as director and manager, for a certain period of time performed services for and under the direction of that company in return for which he received remuneration, that being a matter for the referring court to determine.’).
In light of the deference to the factual assessment of the national court, the CJEU does complete the analysis with respect to (now) Article 7(1): if the contract is not one of employment, then the special jurisdictional rule of Article 7(1) needs to be applied. The director of a company, the Court holds, provides a service to the company within the meaning of Article 7(1)b. In the absence of any derogating stipulation in the articles of association of the company, or in any other document, it is for the referring court to determine the place where Mr Spies in fact, for the most part, carried out his activities in the performance of the contract, provided that the provision of services in that place is not contrary to the parties’ intentions as indicated by what was agreed. For that purpose, it is possible to take into consideration, in particular, the time spent in those places and the importance of the activities carried out there, it being a matter for the national court to determine whether it has jurisdiction in the light of the evidence submitted to it (at 64).
Finally, should national law also allow for an action in tort against the director of a company, the locus delicti commissi is the place where the director carries out his duties for the company (at 76). The locus damni is the place where the damage alleged by the company actually manifests itself; it cannot be construed so extensively as to encompass any place where the adverse consequences can be felt of an event which has already caused damage actually taking place elsewhere (at 77-78).
All in all, a useful completion of the Shenavai criterion, and in the main a referral to the national court for factual analysis.