roma 2 simionidis.pdf

7/27/2019 roma 2 simionidis.pdf

1/46

*. Dean and Professor of Law, Willamette University College of Law; LL.B. (Private Law); LL.B.(Public Law) University of Thessaloniki; LL.M., S.J.D. Harvard Law Scho ol.

1

Rome II and Tort Conflicts: A Missed OpportunityBy

SYMEON C. SYMEONIDES*

[To be published in 56 AMERICAN JOURNAL OF COMPARATIVE LAW (2008)All rights reserved]

Abstract

This article reviews the European Unions new Regulation on tort conflicts

(Rome II), which unifies and federalizes the member states laws on this

subject. The review accepts the drafters pragmatic premise that a rule-system built

around the lex loci delicti as the basic rule, rather than American-style ap-proaches, was the only politically viable vehicle for unification. Within this

framework, the review examines whether Rome II provides sufficient and flexible

enough exceptions as to make the lex loci rule less arbitrary and the whole systemmore workable.

The authors answer is negative. For example, the common-domicile

exception is too broad in some respects and too narrow in other respects. Likewise,

the manifestly closer connection escape is phrased in exclusively geographical

terms unrelated to any overarching principle and is worded in an all-or-nothing way

that precludes issue-by-issue deployment and prevents it from being useful in all but

the easiest of cases. The review concludes that, although attaining a proper

equilibrium between legal certainty and flexibility is always difficult, Rome II errs

too much on the side of certainty, which ultimately may prove elusive.

On the whole, Rome II is a missed opportunity to take advantage of the richcodification experience and sophistication of modern European conflicts law.

Nevertheless, Rome II represents a major political accomplishment in unifying and

equalizing the member states laws on this difficult subject. If this first step is

followed by subsequent improvements, Europe would have achieved in a relatively

short time much more than American conflicts law could ever hope for.

Contents

I. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2II.HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

A. Before Rome II. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4B. Rome II .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5III. GENERAL FEATURES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

A. Scope.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7


2/46

SYMEON C.SYMEONIDES ROM E II AN D TORT CONFLICTS

1. The English text of the Regulation can be found in the Official Journal of the European Union,2007 O .J. (L 199) 40. Excerp ts discussed in this essay are reproduced in an Append ix at ???,infra.

2. ROM E II, closing sentence. See also Treaty establishing the Europ ean Comm unity, art. 249(2).

3. ROM E II, art. 32.

4. Rome II wil l not apply to Denmark. See ROM E II, recital (40), art. 1(4).

5. Among the non-contractual obligations arising from acts other than torts are those arising fromunjust enrichment, negotiorum gestio, and culpa in contrahendo. Rome II covers theseobligations in arts. 10-12. This essay does not d iscuss these articles.

56AMERICAN JOURNAL OF COMPARATIVE LAW (2008) PAGE 2 OF 46

B. Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8C. Certainty vs. Flexibility. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8D. Jurisdiction-Selection or Content-Oriented Law-Selection. . . . . . . . . . 9

E. State or National Interests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12F. Issue-by-Issue Analysis andDpeage. . . . . . . . . . . . . . . . . . . . . . . . . 13

IV. THE GENERAL RULE .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15V.THE EXCEPTIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

A. The List. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20B. The Common Habitual-Residence Rule. . . . . . . . . . . . . . . . . . . . . . . . 21C. The General Escape.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

1. The Closer-Connection Escape. . . . . . . . . . . . . . . . . . . . . . . . . 252. The Pre-existing Relationship Exception.. . . . . . . . . . . . . . . . . 31

D. Compensation for Traffic Accidents. . . . . . . . . . . . . . . . . . . . . . . . . . . 32VI. PRODUCT LIABILITY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34VII. ENVIRONMENT AL TORTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37VIII. RULES OF SAFETY AND CONDUCT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39IX.PARTY AUTONOMY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43X.CONCLUDING THOUGHTS.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

I. INTRODUCTION

On July 11, 2007, the European Parliament and the Council of the EuropeanUnion adopted Regulation (EC) No 864/2007 on the Law Applicable toNon-Contractual Obligations known as Rome II. In European Union parlance,1

a regulation is binding . . . and directly applicable in all member countries without2

the need for implementing national legislation in each individual country. TheRegulation is scheduled to go into effect on January 11, 2009. It will preempt the3

national choice-of-law rules of the European Unions member states on non-4

contractual obligations arising from torts or delicts and from other acts or facts. 5

Unlike some other regulations which apply only within the European Union, RomeII will have universal application, in the sense that it will cover torts occurring bothwithin and outside the Union, and it may lead to the application of the law of a non-member state. Rome II is a dramatic step in the federalization or Europeanization


3/46


6. Johan Meeusen, Instrum entalisa tion of Priva te Internationa l Law in the European Un ion:Toward a European Conflicts Revolution? 9 EUR.J. M IGR.& L. 287 (2007).

7. For other discussions of Rome II, including earlier versions,see , e.g ., Janeen M. Carruthers &Elizabeth B. Crawford, Variations on a Theme of Rome II: Reflections on Proposed Ch oice of

Law Rules fo r No n-contrac tua l Ob ligations, 9 EDINBURGH L. REV . 65, 238 (2005;) Peter Hay,Contemporary Approaches to Non-Contractual Obligations in Private International Law

(Conflict of Laws) and the European Com munitys Rome II Regulation, EUROPEAN LEGALFORUM I-137 (4-2007); Cyril Nourissat & Edouard Treppoz, Quelques observations surleavened-projet de proposition de reglement du Conseil sur la loi applicable aux ob ligations

non contractuelles Rome II , 45 J. DR. INT L 130 (2003); Willibald Posch, The Draft

Regulat ion Rome II in 2004: Its Past and Futur e Perspective s, 6 YBK PRIV . INT L L. 129(2004); Ansgar Staudinger, Rome II and Traffic Accidents, E UROPEAN LEGAL FORUM 61 (4-2005); Peter Stone, The Rome II Proposal on the Law Applicable to Non-contractualObligations, EUROPEAN LEGAL FORUM 213 (4-2004); von Gerhard Wagner, Internationales

De liktsrecht, di e Ar beiten an der Rom II Verordnung und der Europische

Deliktsgerich tsstand, IPRAX 372 (2006); Russell J. Weintraub, Rome II and the TensionBetween Predictab ility and Flexib ility, 19 RIV. DIR. INT LE PRIV. E PROCESS 561 (2005).

8. See SYMEON C. SYMEONIDES,THE AMERICAN CHOICE-OF -LAW REVOLUTION :PAS T,PRESENTAN D FUTURE (2006) [hereinafter THE CHOICE-OF -LAW REVOLUTION].

9. See CODE DE DROIT INTERNATIONAL PRIV (Loi du 16 juillet 2004, Moniteur Belge 27 Juillet2004 (1 ), hereafter BELGIAN PILCOD E.

10 . See ACT OF 11 APRIL 2001REGARDING CONFLICT OF LAWS ON TORTS, Staatsblad 2001, 190,

effective 1 June 2001 (hereafter D UTCH PIL ACT). For English translation with anIntroductory Note by P. Vlas, see NETHERLANDS INT L L. REV . 221 (2003-2).

11 . See PRIVATE INT ER NA TIO NA L LAW (M ISCELLANEOUS PROVISIONS) ACT OF 1995 (c 42), 8No vember 1995 , hereaf ter cited as ENG LISH PILACT .

12 . See FEDERAL ACTS OF 1986 AN D 1999 FOR THE REVISION OF PRIVATE INT ER NA TIO NA L LAW(AMENDING THE INT RO DUC TO RY LAW TO THE CIVIL COD E (EGBGB)) (Transl. by Wegen, 27I.L.M. 1, 18 (1988), and Hay, 47 AM .J. COM P.L. 650 (199 9)), hereafter referred to as EGBGB.

13 . See FEDERAL LAW OF 18 DECEMBER,1987 ON PRIVATE INT ER NA TIO NA L LAW (Transl. by Cornu,Hankins & Symeonides, 37 AM .J. COM P.L. 193 (19 89), hereafter cited as SWISS PILACT .


of private international law (PIL) in the EU member states, a step that has been aptlycharacterized as the European conflicts revolution.6

This essay discusses the main provisions of Rome II on tort conflicts. The7

fact that the essays author has spent the last three decades laboring in the Americanconflicts vineyard, inevitably influences his views on the subject. Nevertheless, to theextent this essay criticizes Rome II, it is notat least not consciouslyfor failing toemulate any American models, but rather for failing to take full advantage of therichness, sophistication, and progress of modern European PIL. Indeed, whileAmerican conflicts law was stumbling through a loud revolution and then engagingin endless navel-gazing about how to end it, European PIL was going through a quiet8

evolution, gradually repairing the old system and producing several noteworthy PILcodifications, among them the Belgian, Dutch, English, German, and Swiss,9 10 11 12 13

to mention just a few. Unfortunately, Rome II does not compare favorably with these


4/46


14 . See Kurt H. Nadelmann, The Benelux Uniform Law on Private International Law , 18AM .J.COM P.L. 406 (1970 ). An English translation of the Uniform Law is reproduced id. at 420-425.

15. BENELUX UNIF OR M LAW , Art. 14 (emphasis added).

16. 191 N.E.2d 279 (N.Y. 1963) .Babcockapplied New York law to a suit by a New York guest-passenger against her New Yo rk host-drive r and his insurer for injury sustained in a tra fficaccident in Ontario, during a short trip there. O ntarios guest-statute prohibited the suit, while

Ne w Y ork law allo wed it.

17. For an English translation of the Draft Convention,see 21 AM . J. COM P.L. 587 (1973). Fordiscussion,see Kurt H. Nadelmann, Impress ionism and Un ificat ion of Law: The EEC DraftConvention on the Law Applicable to Contractual and Non- Contractual Obligations, 24 AM .J. COM P.L. 1 (1976).


codifications, thus underscoring the fact that law reform is far more difficult at themultinational than the national level.

II.HISTORY

A. Before Rome II

In recent European history, the first effort to codify choice of law for torts ata multinational level was the Benelux Uniform Law on Private International Law of1969, which was based on work that began as early as 1951. Although it never14

entered into force, the Uniform Law was surprisingly modern for its day. Its basicrule for tort conflicts was to apply the law of the place of tortious conduct, but subjectto a wisely drafted escape clause. The escape provided that, if the consequences ofa wrongful act belong to the legal sphere of a country other than the one where theact took place, the obligations which result therefrom shall be determined by the lawof that other country. This escape was phrased broadly enough to cover not only15

cross-border torts in which the consequences of conduct in one country are felt inanother country, the country of injury, but also cases such as the one in the famousNew York caseBabcock v. Jackson, namely intra-state torts in which both parties16

are domiciled in another country (common-domicile cases). We shall return to thispoint later.

The next major effort took place in 1967, in the context of the then EuropeanEconomic Community. By 1972, this effort produced the E.E.C. Draft Conventionon Contractual and Noncontractual Obligations. Its main provision on tort conflicts17

also adopted the place of conduct rule, but was accompanied by a more elaborate,two-prong escape:

[I]f, on the one hand, no significant link exists between the situationresulting from the event which caused the damage and the State inwhich the event occurred and if, on the other hand, such situation hasa predominant connection (connexion prpondrante) with another


5/46


18. EECDRAFT CONVENTION ON CONTRACTUAL ANDNONCONTRACTUAL OBLIGATIONS, art. 10(2).

19 . See id. at art. 10(3) ( Such a connection must normally be based o n a connecting factor commonto the victim and the author of the damage.).

20 . See infra at text accompanying note 130.

21 . Id . at art. 10(4). The placement of this statement suggested that it was intended to be used no tonly in applying (or no t applying) the common-dom icile presumption, but also in applying thegeneral, closer connection escape.

22 . See art. 7 of the HAGUE CONVENTION OF 4 MAY 1971 ON THE LAW APPLICABLE TO TRAFFICACCIDENTS & art. 9 of the H AGUE CONVENTION OF 2OCTOBER1973 ON THE LAW APPLICABLETO PRODUCTS LIABILITY .

23 . See EE CDRAFT CONVENTION art. 12. (Irrespective of which law is applicable under Article10, in determining liability account shall be taken of rules of safety and public ord er in force atthe place and time of the event which caused the d amage.).

24 . See ROM E II, art. 17, discussed infra VIII.

25. The consolidated text of the Convention as amended by the various Conventions of Accession,and the declarations and protocols annexed to it, is published in 1998 O .J. (C 27, 26,1) 34 .


State, the law of that State shall apply.18

The escape continued with an explicit common-domicile presumption forBabcock

type cases, and with the sensible notion which Rome II does not repeat that,19 20

in cases of multiple victims, the applicable law shall be determined separately foreach of them. Finally, the draft convention included a special provision, also21

contained in two Hague conventions of the same period, which called for taking22

into account certain rules of safety and public order in force in the conduct state.23

This provision survived in slightly different verbiage in Rome II as discussedbelow.24

With the expansion of the Community to nine members states in 1973,following the accession of the United Kingdom, Ireland, and Denmark, the effortslowed down. It slowed to the point that the decision was made to abandon the tortprovisions of the draft convention and instead to concentrate on contract conflicts.In 1980, the Convention on the Law Applicable to Contractual Obligations, now25

known as the Rome Convention, was opened for signature, and it entered into forceon April 1, 1991.

B. Rome II

The idea of addressing tort conflicts received new impetus with the Treaty ofAmsterdam of October 2, 1997, when the European Commission began solicitingfeedback on measures to implement the provisions of the Treaty regarding judicialcooperation in civil matters with cross-border impact. By 1998, the idea of aconvention on tort conflicts was adopted in principle and, later that year, the first

draft proposal was put forward by the Groupe europen de droit international priv


6/46


26 . See Proposal for a European Convention on the Law Applicable to Non-contractualObligations adopted at the Luxembourg meeting of Sept. 25-28, 1998, available athttp://www.drt.ucl.ac.be /gedip/gedip-documents-8pe.html.

27 . SeePreliminary D raft Proposal for a European Council Regu lat ion on the La w Applica ble toNon-Contractu al Ob ligation s, available athttp://europa.eu.int/comm.justice_home/ unit/civil/consultation/Index_en.html [hereinafterCommission, Preliminary Draft Proposal].

28. For a comprehensive academic proposal put forward during this period by the Hamburg Groupof Private International Law,see Comm ents on the European Commissions Draft Proposal fora Europea n Council Regulation on the Law Applicable to Non-Contractual Obligations (Oct.10, 2002), 67 RABELSZ 1 (2003).

29 . Se eCommission of the European Communities, Proposal for a Regulation of the EuropeanParlia ment a nd the Coun cil on the Law Applica ble to Non-Contractual Ob ligation s, (COM427) (2003) final, 2003/0168(C)D), Brussels, (July 22, 2003) [hereinafter CommissionProposal]. For a critique of this proposal, see Symeon C. Symeonides, Tort Conflicts and Rom e

II; A V iew from A cro ss, in FESTSCHRIFT FRERIK JAYME 935 (H -P. Mansel, et al., eds. 2004).

30. Ms. Wallis is currently the Vice President of the European Parliament and is the former leaderof the Liberal Democrat Europ ean P arliamentary Party (LDEPP). She is an En glish solicitorwho also studied law in Belgium, Germany, and Switzerland and is fluent in French andGerman. For more information on her b ackground and wo rk in the European Parliament, seehttp://www.dianawallismep.org.uk/.

31. In the interest of full disclosure, it should be noted that the undersigned author is one of theacademics consulted by the rapporteur.


(GEDIP). This proposal was an elegant, sophisticated, and flexible document which26

has influenced the general content and coverage of Rome II, although not where itmatters most. In May 2002, the European Commission published a preliminary draft

proposal and invited comments from the public. In addition to holding public27hearings, the Commission received, and posted on the internet, more than eightywritten commentaries from interested parties, including trade and industry groups,professional organizations, governments, practitioners, and academics. Although28

the majority of the commentaries came from industry groups and tended to supportthe most regressive elements of the preliminary draft, the whole process is a goodexample of European democracy at work.

The Commission finalized its proposal on July 22, 2003. The proposal was29

accompanied by a detailed Explanatory Report (hereafter Report) and an article-by-article commentary. This Report is the only complete explanation of the thinking

underlying Rome II and it remains authoritative as to all of the remaining unamendedprovisions.

The process then shifted to the European Parliament where the rapporteur,British MEP Diana Wallis, continued to solicit feedback, especially from30

academics. The rapporteur tried valiantly to inject some flexibility into the31

Commissions proposal and, on July 6, 2005, she succeeded in having Parliament
http://www.drt.ucl.ac.be/http://www.drt.ucl.ac.be/


7/46


32 . See EUR.PAR L., P6_TA(20 05)028 4. (Codecision procedure: first reading..

33 . See EUR.PAR L., P6_TA(2007)0006. (Codecision procedure: second reading..

34 . SeeCommon Position (EC) No 22/2006 adopted by the Council on Sept. 25, 2006, Oy C289E/68 (Nov. 28, 2006).

35 . See supra note 1.

36. Hereafter, the word state is used interchangeably with the word country.

37. The Regulation also does not apply to evidence and procedure, except as provided in Articles21 and 22, which deal with the proof o f juridical acts and presumptions regarding the burdenof proof in torts.

38. For recent academic commentary on the subject, see Christopher J. Kunke, Rome II an dDefama tion: Will the Tail Wag the Do g? 19 EMORY INT L L. REV . 1733 (2005); AaronWarshaw, Uncertainty from Abroad: Rom e II and the Choice of Law for Defamation Claims ,

(continued...)


adopt, on first reading, several amendments to that effect. However, the Council,32

representing the member states, and the Commission rejected most of the amend-ments. On second reading, Parliament moved closer to the Commissions positions,33

however, the Council and Commission remained unsatisfied. The process then34shifted to a Conciliation Committee, consisting of representatives of the Council,Commission, and Parliament. The ensuing trialogue bridged the differences witha compromise text that was adopted by Parliament on third reading, on July 11,2007. This is the final text and is referred to hereafter as Rome II.35

III. GENERAL FEATURES

A. Scope

Article 1 of Rome II defines the scope of the Regulation. It applies tonon-contractual obligations in civil and commercial matters, in situationsinvolving a conflict of laws, namely, situations having multistate contacts of thekind and pertinence that implicate the laws of more than one state. The Regulation36

does not apply to revenue, customs or administrative matters or to the liability of theState for acts and omissions in the exercise of State authority (acta jure imperii), orto obligations arising out of: marriage, family or similar relationships; matrimonialproperty regimes; wills and succession; bills of exchange, cheques, promissory notesand other negotiable instruments; the law of companies; relations between thesettlors, trustees and beneficiaries of voluntary trusts; nuclear damage; violations ofprivacy and rights relating to personality, including defamation. The last topic was37

initially included in the scope of Rome II, but the proposed article proved mostcontroversial attracting more commentary, especially from the publishing industry,than any other provision. Eventually, the article was dropped, but a CommissionStatement accompanying Rome II pledges to study the subject again and report toParliament by December 2008.38


8/46


3 8. (...co ntinue d)32 BROOKLYN J. INT L L. 269 (2006).

39. For a comparative discussion of the various choices, see SYMEON C.SYMEONIDES,PRIVATEINT ER NA TIO NA L LAW AT THE END OF TH E 20 TH CENTURY : PROGRESS OR REGRESS?, 21-35(1999) [hereinafter Symeonides,PROGRESS ORREGRESS].

40 . See, e.g., the rapporteurs Report to the European P arliament, in Final A6-0211/2005, (June27, 2005).


B. Structure

Rome II begins with forty statements or recitals explaining the drafters

thinking and goals, and continues with thirty-two articles divided into seven chapters.The most important of these chapters, and the focus of this essay, is chapter II, whichcontains the general rule for all torts (Art. 4), followed by special rules on productliability (Art. 5), unfair competition (Art. 6), environmental damage (Art. 7),infringement of intellectual rights (Art. 8), and industrial action (Art. 9). Chapter IIIdeals with unjust enrichment, negotiorum gestio, and culpa in contrahendo, chapterIV deals with party autonomy (Art. 14), and the remaining chapters contain common,general and transitional rules.

The general rule of Article 4 is the lex loci delicti, which is defined as the lawof the place of the injury (lex loci damni). The rule is followed by an exception infavor of the parties common habitual residence, and by a general escape clausebased on the closer connection principle. Articles 5 to 9 are phrased as eitherparticular applications or clarifications of the general rule for certain torts, or asexceptions to the general rule. Before discussing these articles, it is helpful to identifysome of the general methodological features of Rome II, beginning with the balanceit strikes between certainty and flexibility.

C. Certainty vs. Flexibility

Every PIL system encounters the perennial tension between the need forcertainty and predictability on the one hand, and the need for flexibility and equity on

the other. Each system responds differently, striking a different equilibrium betweenthe two needs. For example, the GEDIP proposal, which consisted of a series of39

cascading presumptions rather than rules, was a deliberate and brave choice offlexibility over certainty. Although the GEDIP scheme was nearly perfect, thepotential for uncertainty must have alarmed the Commission, because it movedsharply in the opposite direction of adopting a system of tightly written black-letterrules with relatively few escapes and little room for judicial discretion. The rapporteurand Parliament tried to re-inject flexibility into the draft, but, as the final text40

reflects, the Commission and especially the Council remained unyielding. The onlyprovision that remains from the Parliaments efforts is a recital in the Preamblereferring to the need to do justice in the individual case, juxtaposing it with the


9/46


41 . See ROM E II Preamble, recital (14).

42 . See infra V.2-3.

43 . See ROM E II, recital (6) (The prop er functioning of the internal market creates a need, in orderto improve the predictability of the outcome o f litigation, certainty as to the law applicable andthe free movement of judg ments, for the conflict-of-law rules in the Memb er States to designatethe same national law irrespective of the country of the court in which an action is brought.).

4 4. D avid F. C avers,Resta tem ent of the Law of Co nfl ict o f Laws, 44 YALE L. J. 1478, 1482 (1935).

45. In Professor Weintraubs words, [i]ronically Rome II is more likely to succeed in providingreasonable foreseeability if its rules provide sufficient flexibility. Weintraub, supra note 7, atid.


requirement of legal certainty. However, as explained later in discussing the41

escape clauses of Rome II, the final balance is excessively skewed against flexibility.42

The reasons for the Councils and Commissions political preference forcertainty over flexibility are obvious. The primary motive behind the movement todraft Rome II, as well as the choice of the particular instrument for its implementa-tiona regulation as opposed to a directivewas the need to ensure uniformity ofchoice-of-law decisions within the European Union. These two bodies must have43

concluded that uniformity would be in jeopardy if Rome II were to have too manyflexible rules or escape clauses. This was a plausible, though not necessarily the best,conclusion. The argument that a codification intended for application by the courts ofdifferent countries cannot afford to be flexible, is highly overrated. For example,whatever its other faults, the Rome Convention did not fail for being too flexible.

Moreover, while no one would question the desirability of uniformity andcertainty, one can question the extent to which these values should displace all othervalues of the choice-of-law process, such as the need for sensible, rational, and fairdecisions in individual cases. If the American experience has something to offer, it isa reminder that a system that is too rigidas the traditional American system wasultimately fails to deliver the promised predictability because, in a democratic societyno system can mechanize judgment and, to the extent it attempts to do so, judges44

will ignore it. To be sure, it would be unfair to characterize Rome II as a mechanical45

system. Its drafters were conscious of the need for flexibility and they attempted toprovide for some degree of it. The question is whether the drafters provided enoughflexibility, a question on which reasonable minds can differ. Only time will show

whether the final text of Rome II has found the golden mean between the competingvalues of certainty and flexibility. For what it is worth, and for reasons that willbecome obvious from the following discussion, this authors opinion is that a schemethat is closer to the GEDIP proposal or the rapporteurs ultimately defeated proposalswould have been preferable.

D.Jurisdiction-Selection or Content-Oriented Law-Selection


10/46


46. The term jurisdiction-selection was first coined by Professor Cavers. See David F. Cavers,A Critique of the Choice-of-Law Problem , 47 HAR V.L. REV . 173 (1933).

47. For a full discussion of these concepts, see Symeon C. Symeonides,Am erican C onflic ts L awat the Dawn of the 21st Century, 37 W ILLAMETTE L. REV . 1, 46-60 (2000).

48 . See Symeonides,THE CHOICE-OF -LAW REVOLUTION 399-404.

49. One example of such a rule is the common-domicile rule for loss distribution conflicts. Asexplained infra at text accompanying note 112, this rule produces good results not only whenthe law of the common domicile favors the victim, but also when it favors the tortfeasor.

56AMERICAN JOURNAL OF COMPARATIVE LAW (2008) PAG E 10 OF 46

One difference between traditional and modern PIL is the extent to which eachconsiders the content of the substantive laws of the involved states before choosingthe law that would govern the case at hand. In traditional PIL, the choice was based

exclusively on the physical contacts of the involved states (jurisdiction-selection),without regard to the content of their substantive laws. In contrast, in modern PIL,46

the choice is based not only on physical contacts but also on the content andunderlying policies of the laws of the contact states (content-oriented lawselection). Content-oriented law selection is easier when the court is not bound by47

statutory choice-of-law rules. When such rules exist, however, a content-oriented lawselection is possible only to the extent these rules allow it. The question here is towhat extent the rules of Rome II require jurisdiction-selection and to what extent theypermit content-oriented law selection?

Under Rome II, jurisdiction-selection is the norm and content-oriented law

selection is a limited, often unintended, exception. The majority of the dispositivearticles of Rome II designate the state whose law governs because of that statesphysical contacts rather than because of the content of its law. This alone is not areason to criticize Rome II. After all, in most codified PIL systems, jurisdiction-selecting rules are bound to outnumber content-oriented rules. Secondly, asdocumented elsewhere, carefully crafted, narrowly tailored jurisdiction-selecting48

rules can achieve functionally sound results. For this to happen, however, the draftersmust identify the various typical law-fact patterns and consider whether theapplication of a given states law produces an equally good result regardless of thecontent of that law. If the answer to this question is affirmative, a jurisdiction-selecting rule would be acceptable. If the answer is negative, then one or more49

content-oriented rules should be drafted to accommodate the various patterns. Onequestion explored later in this essay is whether the jurisdiction-selecting rules ofRome II meet this test.

Before addressing this question, it is worth noting that Rome II containsseveral provisions that, directly or indirectly, permit consideration of the content ofthe substantive laws of the involved countries. These provisions can be divided intothree categories:

(1) traditional provisions that allow the court to refuse to apply the normally


11/46


50 . See infra text following note 15 4.

51. R OM E II, recital (33).

52 . See infra at text accompanying note 44.


applicable law to the extent it violates the forums mandatory rules (Article16) or its ordre public (Article 26);

(2) provisions that require the court to choose a law that produces a particularresult, such as Article 18, which authorizes a direct action against the insurerif such action is allowed by either the law applicable to the tort or the lawapplicable to the insurance contract; and Article 14(2) and (3) which providethat the contractually chosen law shall be disregarded to the extent it violatescertain mandatory rules; and

(3) provisions that directly allow one litigant to choose between two laws,such as: Article 7, which allows the victim of an environmental tort to choosebetween the law of the place of conduct and the law of the place of injury; andArticle 6(3)(b), which allows the plaintiff to choose the law of the forum incertain cases involving anti-competitive restrictions.

Although some of these content-oriented provisions have long been acceptedby traditional PIL, the presence of the remaining ones suggests a selective receptive-ness to the idea of content-oriented selection on the part of the drafters of Rome II.

Moreover, as in other codifications, certain choice-of-law rules that purport tochoose a state because of its contacts rather than its law, in fact are based on certainassumptions about the likely content of that law and to that extent they disguisespecific substantive choices. One possible example from Rome II is Article 5 onproducts liability, which, as explained below, will lead to the application of the lawof the victims habitual residence in the great majority of cases. When the plaintiff50

is a resident of an EU country with a generous pro-consumer law and the manufactureris based in a state with a pro-manufacturer law, Article 5 will ensure that the EUplaintiff will be protected under EU standards. However, in the converse situation, thesame article will ensure that the EU defendant will be held accountable under thelower standards of the plaintiffs foreign residence.

Similar thinking seems to underlie recital 33 regarding damages for victimsof traffic accidents. The recital states that, when quantifying damages for personalinjury in cases in which the accident takes place in a State other than that of thehabitual residence of the victim, the court should take into account all the relevantactual circumstances of the specific victim, including in particular the actual lossesand costs of after-care and medical attention. The history and meaning of this recital51

are discussed later. Suffice it to say that, despite its precarious placement and52

equivocal wording, this recital does more than recognize the relevance of foreignfacts(i.e., the costs of after-care and medical attention); it also recognizes the relevance of


12/46



54 . Id. at recital (20).

55 . Id . at recital (21).




foreign law as Parliaments unsuccessful amendment expressly provided. Indeed, itis the law of the foreign state that makes these facts relevant, not the other wayaround. Stated another way, the fact that medical costs are high in the victims

residence is not significant unless the victim would be entitled to recover medicalcosts under that states law. To this extent, recital 33 can be seen as an example of acontent-oriented law selection.

E.State or National Interests

A casual look at Rome II confirms the impression that, like most traditionalPIL systems in Europe, Rome II does not subscribe to the notion that ordinaryconflicts disputes at the private-law level implicate the interests of the involvedcountries. Indeed, none of Rome IIs articles refer to state policies, much less interests,

and both the Preamble and the Explanatory Report contain several statementsdescribing Rome IIs goal as one of ensur[ing] a reasonable balance between theinterests of [theparties, i.e.,] the person claimed to be liable and the person who hassustained damage.53

Yet, one who looks below the surface would discover that Rome II is notoblivious to state interests. In several instances, the Preamble refers to broader societalinterests that reach beyond the interests of private litigants. For example, with regardto products liability, recital 20 of the Preamble speaks of the policies of fairlyspreading the risks, . . . protecting consumers health, stimulating innovation, securingundistorted competition and facilitating trade. With regard to unfair competition,54

recital 21 speaks of the need to protect competitors, consumers and the general publicand ensure that the market economy functions properly. Recital 25, with regard to55

environmental torts, states that the need for a high level of [environmental]protection and the principle that the polluter pays justify a choice-of-rule ofdiscriminating in favour of the person sustaining the damage. Finally, recital 3156

recognizes the need to impose certain restrictions on the parties power to choose thegoverning law so as to protect the weaker parties.57

More importantly, some of Rome IIs dispositive articles can only be explainedin terms of public (and thus state) interests, rather than in terms of private interests.Besides Article 26 which codifies the traditional ordre public exception, and Article16 which allows the forum to interpose its own mandatory rules, many otherprovisions of Rome II are designed to be sensitive to certain preferred substantive


13/46


58 . See infra VIII.

59. EUR.PAR L.FINAL A6-0211/2005, (June 27, 2005) p. 19/46.

60. Art. 15 provides that the scope of the applicable law encompasses virtually all issues likely toarise in tort litigation, including the basis and extent of liability, the ground s for exemption fromliability, any limitation of liability and any division o f liability, the existence, the nature and the

(continued...)


polices. The clearest example is Article 7 pertaining to environmental torts, which isdiscussed in some detail later. Other less obvious examples are: Article 14(2)-(3),58

prohibiting choice-of-law clauses to the extent they violate certain mandatory rules of

a non-chosen state or of the Community; Article 14(1)(b), allowing pre-tort choice-of-law clauses only in contracts between commercial parties and subjecting them tocertain limitations; and Article 6(4), prohibiting choice-of-law agreements in cases ofunfair competition and restrictions to competition.

F.Issue-by-Issue Analysis and Dpeage

Modern choice-of-law doctrine and practice have come to recognize that inmany cases the conflict is confined to only some aspects or issues of the case, andthat in other cases the involved states may be interested in different aspects of the

case. Consequently, rather than seeking to choose a law as if all aspects of the casewere in dispute, the modern decisionmaker focuses on the narrow issues with regardto which a conflict exists and proceeds accordingly. Such issue-by-issue analysis iseasier and more likely when the decisionmaker is not bound by statutory choice-of-law rules. When such rules exist, however, such an analysis is possible only to theextent the rules permit it. In turn, this depends on whether these rules are phrased inbroad terms designating the law that would govern the case as a whole, or whetherthey are phrased in narrower terms.

On the whole, Rome II takes a negative stance towards issue-by-issue analysis,apparently because of the drafters desire to avoid as much as possible one of its

byproductsthe phenomenon ofdpeage, namely the application of the laws ofdifferent states to different issues in the same case. Indeed, the Council and theCommission specifically rebuffed the rapporteurs and Parliaments efforts tointroduce issue-by issue analysis. One of Parliaments amendments on first readinghad attached the following concluding sentence to what became the general rule ofArticle 4: In resolving the question of the applicable law, the court seised shall,where necessary, subject each specific issue of the dispute to separate analysis. The59

Council and Commission rejected this amendment. In the end, most of Rome IIsarticles were phrased in broad terms designating the law that would govern the caseas a whole, the tort/delict. Article 15 reaffirms this holistic approach by providinga long list of issues that comprise the scope of the law applicable under these

articles.60


14/46


6 0. (...co ntinue d)assessment of damage or the remedy claimed; injunctive relief, the proper beneficiaries of theright to claim damages, and its transferability or heritability, respondeat superior, and theextinction or prescription of the obligation (statutes of limitation).

61. This provision can lead to the application of community law to some issues and national law toother issues. Art, 27 can also lead to the same p henomenon by recognizing that a case may begoverned partly by the law designated by Rome II and partly by the law designated by otherchoice-of-law rules co ntained in other community instruments.

6 2. A rt. 1 4 c an lea d to dpeage through several different routes. For example, the parties maychoose different laws for different aspects of the case, or they may c hoos e one law for so meissues and none for o thers. Even when they choo se one law for all issues, that law will not be

applied to the extentit violates the mand atory rules described in art. 14(2) or the comm unityrules described in art. 14(3).

63. Arts. 16 and 26 allow courts to displace the applicable law to the extentnecessary to satisfy themanda tory rules of the forum state, orto the extentthat law is incompatible with the forumsordre public.

64. Art. 17 (disc ussed infra VIII) allows the court to take account of the conduct rules of theconduct state when the rest of the case is governed by the law of another state.

65. Under art. 18, the victims right to directly sue the tortfeasors insurer may be governed by thelaw governing the insurance contract, although the tort may be governed by another law.


Nevertheless, a closer look at some of the other articles reveals that Rome IIdoes notbecause it cannotentirely avoid an issue-by-issue analysis and thus thepossibility ofdpeage. First, some of Rome IIs articles speak of the law applicable

to the obligation arising out of a tort or delict, rather than to the tort or delict as awhole. This is a wise choice of terms, which will enable courts to engage in a moreindividualized evaluation of the multiple obligations that may arise from the sameevents. Moreover, if Rome II uses the term obligation in its original sense of the legalbond between the obligor and the obligee that encompasses both the obligors dutiesand the obligees rights, the court will be able to further differentiate among the rightsof the individual victims and proceed accordingly.

Rome II consciously stops short of the next logical step of encouraging or evenpermitting an issue-by-issue analysis of the various aspects of the obligation. Even so,some of Rome IIs other articles contain the possibility for further splitting the issues.

Among the latter articles are:(a) Article 8(2) on intellectual property rights;61

(b) Article 14 on choice-of-law agreements;62

(c) Articles 16 on mandatory rules and 26 on ordre public;63

(d) Article 17 on rules of safety and conduct;64

(e) Article 18 on direct actions against insurers;65

(f) Articles 19 and 20 regarding subrogation, indemnification, or


15/46


66. Under arts. 19 and 20, the rights of subrogation, indemnification, and contribution between theparties mentioned in the articles may be governed by a law other tha n the law governing thevictims claims against these parties.

67 . See supra text accompanying note 51.68. In fact, the possibil ity ofdpeage is the main reason cited by the Council and Commission for

rejecting Parliaments express rule to that effect. Although the recital does not enjoy the samestatus as an express rule in the Regulations main body, the recital either means what it says, inwhich case the p ossibility ofdpeage remains, or it does not mean what it says, in which caseParliament did not get anything from the purported comp romise.

69 . See infra at text accompanying notes 129-34.

70. R OM E II, art. 4(1).

71 . Id .


contribution; and66

(g) Recital 33, which, as noted earlier, purports to authorize the application67

of the law of the habitual residence of the victim of a traffic accident inquantifying the recoverable damage, even when all the other issues resultingfrom the accident are governed by the law of the accident state.68

The above is a much longer list than the opponents of dpeage wouldordinarily tolerate, but one provision that is missing is the general rule of Article 4,especially the general escape of paragraph 3 in which issue-by-issue analysis (with thepossibility ofdpeage) would have been most useful. The resulting problems arediscussed later.69

IV. THE GENERAL RULE

Rome IIs central provision is Article 4, which contains the general andresidual rules. Paragraph 1 of Article 4 provides that the applicable law shall be thelaw of the country in which the damage occurs (lex loci damni). This law governs70

irrespective of the country in which the event giving rise to the damage occurredand irrespective of the country or countries in which the indirect consequences of thatevent occur.71

The operation of this rule can be illustrated by the following hypotheticalscenario, which is used throughout this essay: Blasting operations by a Swiss miningcompany in the Swiss Alps cause a snow avalanche in the French Alps injuring a

group of English tourists. Although there is some room for contrary argument, itseems that Article 4(1) views Switzerland as the country of the event giving rise tothe damage, France as the country in which the damage occurs, and England as thecountry in which the indirect consequences of that event occur. Translated intosimpler English, Article 4(1) provides that the applicable law is the law of the countryin which the injury occurs, and more precisely the harmful physical impact(France),


16/46


72. AMERICAN LAW INS TIT UT E,RESTATEMENT OF CONFLICT OF LAW S 377 Note (1933).


74 . See ROM E II, recital (15).

75 . See, e.g., AUSTRIAN PILACT of 15 June 1978 48(1); POLISH PILACT of 1965 art. 33(1).

76 . See DUTCH PILACT , art. 3(2); ENG LISH PILACT of 1995 11 (subject to exceptions).

77 . SeePORTUGUESE CIV.COD E, art. 45(1) and (2); SWISS PILACT , art. 133(2).

78 . See SPANISH CIV.COD E art. 9; GREEKCIV.COD E, art.26; CZECHOSLOVAKIAN PILACT of 1963,art 15.

79 . See EGBGB art 40(1); HUNGARIAN PILDECREE of 1979 32(1)(2); ITALIAN PILACT of May31, 1995, art 62(1).



irrespective of the country in which the injurious conductoccurred (Switzerland), andirrespective of the country in which the indirect consequences of the injury are felt(England).

Thus, the general rule of Rome II is nothing but a restatement of the traditionallex loci delicti rule, with its last event sub-rule. It purports to be as categorical as thecorresponding rule of the American First Restatement. In its penchant to avoid anyambiguity, the Restatement provided numerous minute localization sub-rules which,for example, defined the place of injury as the place where the harmful force takeseffect upon the body in personal injury cases, and the place where the deleterioussubstance takes effect in cases of poisoning. The fact that the Restatement never72

attained certainty, despite having attained clarity, is a lesson that subsequent codifiersignore at their peril.

The Rome II codifiers note that [t]he principle of the lex loci delicticommissiis the basic solution for non-contractual obligations in virtually all the MemberStates, which of course is true, except for the fact that in many countries this73

solution is subject to several exceptions. The drafters also correctly note that manycountries disagree in defining the locus delicti. Indeed, some countries opt for the74

place of conduct, others opt for the place of injury, others apply the law of the place75 76

of conduct in some specified cases and the law of the place of injury in other cases,77

others leave the question unanswered, while others allow the victim or the court to78

choose between the two laws. The Rome II drafters decided to resolve these79

differences by unequivocally choosing the law of the place of injury, because such asolution strikes a fair balance between the interests of the person claimed to be liable

and the person sustaining the damage, and also reflects the modern approach to civilliability and the development of systems of strict liability. Neither of these two80

reasons are self-explanatory, and the second reason regarding strict liability iscertainly debatable. As for the first reason, the only balance the lex loci damni rulestrikes between the parties is that it can be equally unfair to the plaintiff in some casesas to the defendant in others.


17/46


81 . See infra IV .

82. Hereafter, the term domicile is used interchangeably with the term habitual residence asused in Rome II.

83. Art. 17 is disc usse d infra VIII.

84. 191 N.E.2d 279 (N.Y. 1963) .

85 . See SYMEONIDES,THE CHOICE-OF -LAW REVOLUTION 123-29 and authorities cited therein.

86. Padula v. Lilarn Props. Corp., 644 N.E.2d 1001, 1002 (N.Y. 1994).


This is not to say that, just because the lex loci damni is an old rule, or justbecause it produces unfair results in some cases, the rule is bad in all cases. Despitethe recent tendency, especially among American academics, to summarily reject this

rule as an outmoded remnant of the past, a dispassionate examination of actual casesindicates that this rule produces good or defensible results in several fact-law patterns,although by no means all. To properly evaluate this rule one should first evaluate theexceptions to which it is subject, and, second, examine the results the rule produces81

in several typical patterns formed by the aggregation or disbursement of the pertinentcontacts (conduct, injury, and parties domiciles ) and the content of the laws of each82

contact state. Specifically, it is helpful to distinguish between cases depending on:

(1) whether the injurious conduct and the resulting injury occurred in the samestate (intrastate torts), or in different states (cross-border torts);

(2) whether the tortfeasor and the victim were domiciled in the same state(common-domicile cases), or in different states (split-domicile cases); and

(3) whether the conflict is between laws that primarily regulate conduct(conduct-regulating rules) or primarily allocate the economic loss resultingfrom the injury (loss-allocation or loss-distribution rules).

The above quoted terms are neither self-explanatory nor universally accepted.Generally speaking, the distinction between conduct-regulating and loss-distributingrules corresponds to the two grand objectives of tort lawdeterrence and reparation.Although these two objectives are often interconnected, conduct-regulating rules serveprimarily the first objective, while loss-distributing rules serve primarily the second

objective. Rome II does not adopt this distinction, except in a oblique and limited waythrough Article 17, which authorizes taking account of the safety and conductrules of the conduct state. In the United States, this distinction, which was first83

articulated by the New York Court of Appeals in the 1963 landmark caseBabcock v.Jackson, has been adopted by many courts, albeit without always using this84

terminology and without a consensus on its precise contours. In the words of the85

New York court, conduct-regulating rules are those that have the prophylactic effectof governing conduct to prevent injuries from occurring. This category includes not86

only rules of the road like speed limits and traffic-light rules, but also rules thatprescribe the civil sanctions for violating traffic rules, including presumptions and


18/46


87. Examples include rules providing that a person involved in a collision while driving in excessof the speed limit, or while b eing intoxicated, is presumed to be negligent, and rules providingthat, in a rear-end vehicular collision, the driver of the rear car is presumed to be at fault.

88 . Padu la,supra note 86, at id.

89 . See SYMEONIDES,THE CHOICE-OF -LAW REVOLUTION 141-263.


inferences attached to the violation; rules that prescribe safety standards for work87

sites, buildings, and other premises; rules that impose punitive damages; and rulesdefining as tortious conduct such as alienation of affections, interference with

marriage, or interference with contract, and granting an action against the actor. Incontrast, loss-distributing rules are those that prohibit, assign, or limit liability afterthe tort occurs. They include not only guest statutes, which now are virtually88

extinct, but also rules that define the amount of compensatory damages, and rules ofinterspousal immunity, parent-child immunity, workers compensation immunity, andloss of consortium.

Admittedly, the line between the two categories is not always as bright as onewould like. While some tort rules are clearly conduct-regulating and others are clearlyloss-distributing, there are many tort rules that do not easily fit in either category, andsome rules that appear to fit in both, namely they both regulate conduct and effect or

affect loss distribution. Nevertheless, despite the difficulties in its application, thisdistinction provides a useful starting point for resolving or analyzing many tortconflicts, although in many such conflicts the distinction will not make a difference.The starting point is a presumption that conduct-regulating rules are territoriallyoriented and loss-distribution rules are not necessarily territorially oriented.Consequently, territorial contacts (namely the places of conduct and injury) remainrelevant in conduct-regulation conflicts, while both territorial and personal contacts(i.e., the parties domiciles) are relevant in loss-distribution conflicts.

For reasons explained in detail elsewhere, as well as later in discussing the89

exceptions to the lex loci damni rule, the view of this author is that this rule produces

functionally defensible results in the following patterns of cases:(1) intrastate torts involving conflicting conduct-regulation rules, regardlessof where the parties are domiciled;

(2) intrastate torts in which the issue is one of loss distribution and either thetortfeasor or the victim is domiciled in the state of conduct and injury; and

(3) cross-border torts involving either conduct-regulation or loss distributionconflicts, in which: (a) the state of injury prescribes a higher standard ofconduct for the tortfeasor or of financial protection for the victim than the stateof conduct; and(b) in which the occurrence of the injury in the former statewas objectively foreseeable.

In the first two patterns, the application of the law of the state of conduct and


19/46


90. For documentation and defense of this thesis,see id. at 213-20.

91. However, by adopting a common-domicile rule that is phrased so broadly as to trump the lexloci rule in these cases, Rome II com mits a serious error, which can only be corrected throughthe back door of art. 17. See infra textaccompanying notes 113 & 172.

92. For documentation and defense of this thesis, see SYMEONIDES, THE CHOICE-OF -LAWREVOLUTION 162-91.

93. For documentation and defense of this thesis,see id. at 192-200, 228-36.

94 . See infra text accompanying note 164.

95. For documentation and defense of this thesis, see SYMEONIDES, THE CHOICE-OF -LAWREVOLUTION 200-02, 223-28.

96. Art. 17 allows the court to take account, but not necessarily to apply, the Swiss negligenceper se rule on the assumption that i t qualifies as a rule o f safety a nd conduct. See infra VIII.


injury passes the scrutiny of a functional choice-of-law analysis, regardless of whetherthat law favors the tortfeasor or the victim. Consequently, a jurisdiction-selecting rulelike that of Article 4(1) which is indifferent on its surface to the content of the

involved substantive laws is acceptable, besides being relatively easy to apply in themajority of cases. In the first pattern, the state of both the conduct and injury clearlyhas the strongest claim to apply its law to conduct within its borders, even if theparties are both domiciled in the same foreign state. The lex loci rule honors this90

claim. The second pattern consists of four different sub-patterns, two of which91

present the direct or true conflict paradigm and two the inverse or no interestparadigm. On balance, the application of the law of the state that has the two territorialcontacts (conduct and injury) and one personal contact (the domicile of one party) isjustified in all four sub-patterns, whether that law favors the tortfeasor or the victim.92

The third pattern presents the true conflict paradigm: the state of injury has an interestin protecting its domiciliary victim injured there, while the state of conduct has aninterest in protecting a tortfeasor acting (and often domiciled) there. In these cases, theapplication of the law of the state of injury is justified if the occurrence of the injuryin that state was objectively foreseeable. The lex loci damni rule of Rome II produces93

this result, but its failure to include a foreseeability defense may make it unfair to thedefendant, depending on the other circumstances of the case.94

The lex loci damni rule does notproduce good results in cases of the conversepattern, namely cross-border torts in which the state of conduct prescribes higherstandards of conduct for the tortfeasor than the state of injury. Suppose for example95

that, in the above snow avalanche hypothetical, Switzerland imposes a negligenceperse rule on mining operators who engage in blasting activities in certain areas or time

periods, while France, in order to protect its mining industry, follows an ordinarynegligence standard. In such a case, Article 4(1) calls for the application of French lawand specifically excludes Swiss law. In contrast, if this were an environmental tort,96


20/46


97. Art. 7 is disc ussed infra VII.

98 . Repo rt, art. 3, 11.99 . Id. at 11-12.

100. Similar arguments could be made if France, but not Switzerland, limited the amount of damagesthat could be recovered from mining op erators. Again, art. 4(1) would mandate the applicationof French law, albeit subject to the closer-connection escape of art. 4(3). However, forreasons explained later, this escape will not necessarily lead to Swiss law.

101. See, e.g., EGB GB arts. 40.1, 41; SWISS PILACT A rts. 133(2), 136-39; ITALIAN PILACT art. 62;VENEZUELAN PILACT art. 32(2); HUNGARIAN PILACT art. 32(2).

102. For contrary views, see Hay,supra note 7, at I-145; Wagner, supra note 7, at 379.


Article 7 would allow the victim to opt for Swiss law. The drafters decided not to97

extend this option (which the Report characterizes as the principle of favouring thevictim) to other torts , because this solution would go beyond the victims98

legitimate expectations.99

This, however, is the wrong dilemma. The dilemma is not whether one shouldfavor victims over defendants, but rather whetherin a subject called conflict oflawsone should seek to first determine whether the involved laws actually conflict.As in the case of environmental torts, the reason for giving victims a choice is not tobenefit victims as such, but rather out of deference to the policies of the state ofconduct, which is the only state that has something to lose from the non-applicationof its law. In this case, Switzerlands negligenceper se rule is intended to deter peoplefrom engaging in inherently dangerous operations like blasting. Because the defendantacted within Swiss territory, Switzerland has every reason to insist in determining the

legal consequences of that activity, even ifin this case, the injury occurred across theborder. The effectiveness of this policy of deterrence would be seriously impaired ifexceptions to it were made for out-of-state injuries. Moreover, in terms of fairness andparty expectations, there is nothing unfair in subjecting a tortfeasor to the law of thestate in which he acted. Having violated the standards of that state, the tortfeasorshould bear the consequences of such violation and not be allowed to invoke the lowerstandards of another state. Conversely, there is little reason to apply French law. Itsordinary negligence rule was intended to protect mining defendants who operatewithin French territory, not foreign operators operating elsewhere. In conclusion,100

there is a good deal of wisdom in the rules that allow the victim or the court to choosebetween the laws of the state of conduct and the laws of the state of injury in cases of

cross-border torts. It is regrettable that the drafters of Rome II have chosen not to101adopt a similar rule as they did with regard to environmental torts. 102

V.THE EXCEPTIONS

A. The List

As noted earlier, one cannot properly evaluate Rome II without also analyzingthe exceptions to its basic rule of lex loci damni. A careful perusal of the whole


21/46


103. Additional exceptions can be found in arts. 10-12 in cases of unjust enrichment, negotiorumgestio , and culpa in contrahendo, which are not discussed here.


document reveals that this rule is subject to many more exceptions than are readilyapparent. They can be divided into general exceptions, namely those that apply to allcases, and specific exceptions that apply to particular torts. The general exceptions

include the following:

(1) The application of the law of the parties common habitual residence,under Article 4(2);

(2) The application of the law of a state that has a manifestly closer connec-tion, under the escape clause of Article 4(3);

(3) The application of the mandatory rules of the forum state, under Article 16;

(4) The taking into account (and possible application) of the safety andconduct rules of the state of conduct, under Article 17;

(5) The application of the law chosen by the parties before or after theoccurrence of the tort, under Article 14;

(6) The non-application of the lex loci (or any other law) when it is manifestlyincompatible with the ordre public of the forum, under Article 26.

The specific exceptions include the following:

(7) In product liability cases, the application of the law of the commondomicile of the tortfeasor and the victim, the domicile of the victim, the stateof the products acquisition, or the state of the manifestly closer connection,under Article 5;

(8) The application of the law of the forum in certain cases involvingrestrictions to competition under Article 6(3)(b);

(9) The application of the law of the state of conduct at the victims behest inenvironmental torts, under Article 7; and

(10) The possible application of the law of the victims habitual residence forquantifying recoverable damages in traffic accident cases, under recital 33.

Because of space limitations, this essay discuses only some of theexceptions.103

B. The Common Habitual-Residence Rule

The first official exception to the lex loci rule is found in paragraph 2 ofArticle 4, which provides that, if at the time of the injury, the tortfeasor and the victimhave their habitual residence in the same country (hereafter common domicile), then


22/46


104. Art. 13 provides that the habitual residence of companies and other corporate or unincorporatedbodies is the place of the ir cen tra l administratio n and, fo r inju ries caused or sustained in thecourse of operation of a b ranch, the place where the branch is located. Art. 13 also provides thatthe habitual residence of a natural person acting in the course of his o r her business activity ishis or her principal place of business.

105. R OM E II, art. 6(2).

106. This escape is also repeated in the art. 10-12, dealing with unjust enrichment, negotiorumgestio , and culpa in contrahendo, which are not discussed here. However, the escape beco mesoperable only if the primarily applicable law cannot be determined.

107. Se e LA.CIV.COD E ANN . art. 3544(1); PUERTO RICO DRAFT COD E art. 47(a); SWISS PILACT ,

art. 133; QUEBEC CIV.COD E, art. 3126; BELGIAN PILCOD E, art. 99 (1);HAGUE PRODUCTSLIABILITY CONVENTION, art. 5.

108. See DUTCH PILACT , art. 3(3); EGBG B art. 40(2); HUNGARIAN PILDECREE 32(3).

109. See ITALIAN PILACT , art. 62; POLISH PILACT , art. 31(2);PORTUGUESE CIV.COD E, art. 45. Foran exception that displaces the lex loci when the parties have eithera common n ationality or acommon habitual residence, see CHINAS MODEL DRAFT LAW OF PIL, art. 114 (6th D raft 2002).

110. For citations and discussion, see SYMEONIDES, THE CHOICE-OF -LAW REVOLUTION 145-59;EUGENE SCOLES ,PETERHAY ,PATRICKBORCHERS & SYMEON C. SYMEONIDES,CONFLICT OF

(continued...)


the law of that country applies to the exclusion of the lex loci. This exception is104

repeated in Article 5 on product liability, Article 6 on unfair competition cases inwhich the competition affects exclusively the interests of a specific competitor,105

and Article 9 on industrial action. In contrast, this exception does not apply to other106unfair competition cases, cases falling within Article 7 on environmental torts, orArticle 8 on infringement of intellectual property rights. One can infer that the reasonfor this exclusion is an implicit recognition that cases involving the latter categoriesof cases implicate broader societal interests that go beyond the interests of thelitigants.

In adopting the common-domicile exception, Rome II joins the majority ofrecent PIL codifications and international conventions in accepting the premise thatwhen both the tortfeasor and the victim are affiliated with the same state or state(through nationality, domicile, or habitual residence) that state has the best claim to

determine their respective rights and obligations, even if the tort occurred entirely inanother state. This notion is implemented either through a common-domicile rule (asin the codifications of Louisiana, Puerto Rico, Switzerland, Quebec, Belgium, and theHague Convention on Products Liability), or through an exception from the lex loci107

rule. The exception is phrased either in common-domicile or common-habitualresidence language (as in the Dutch, German, and Hungarian codifications), or in108

common-nationality language (as in the Italian, Polish, and Portuguesecodifications).109

Article 4(2) also mirrors parallel developments in the United States. Asdocumented elsewhere, 32 of the 42 cases decided since the 1960s in which an110


23/46


110. (.. .continued)LAW S, 799-806 (4th ed. 2004).

111. The six cases that applied another law are distinguishable, overruled, or discredited. SeeSYMEONIDES,THE CHOICE-OF -LAW REVOLUTION , 149 n.17, 152-54.

112. See id. at 154.


American court of last resort abandoned the lex loci rule involved the common-domicile pattern. Subsequently, an additional 18 common-domicile cases havereached the highest courts of the states that had previously abandoned the lex loci rule,

thus raising to 50 the total number of common-domicile cases that have reached statesupreme courts in the post-lex loci era. Of these 50 cases, 44 cases (or 88%) haveapplied the law of the common domicile, regardless of the particular choice-of-lawmethodology the court followed. The majority of these cases (35 out of 50) involved111

the Babcock v. Jackson pattern in which the law of the common domicile favorsrecovery more than the law of the state of conduct and injury. These cases present theclassic false conflict paradigm in which only the state of the common domicile has aninterest in applying its law. The remaining cases involved the converse-Babcockpattern in which the law of the common-domicile prohibits or limits recovery morethan the law of the state of conduct and injury. These cases are not as clear falseconflicts as Babcock was because the accident state arguably has an interest inapplying its law to compensate those injured in its territory and to facilitate recoveryof local medical costs. On balance, however, the application of the law of the112

common domicile in both theBabcockpattern and its converse is entirely justified.In this sense, a common-domicile rule that does not depend on the content of the lawof the common domicile (i.e., a jurisdiction-selecting rule) is generally preferable toany other iteration.

It is important to stress, however, that all of the above American casesinvolved conflicts between loss-allocation or loss-distribution rules, as opposedto conduct-regulation rules. In contrast, the common-domicile rule of Rome II ismuch broader in that it encompasses not only loss-distribution issues, but also

conduct-regulation ones. This is a serious defect. As a general proposition, a state hasan interest in enforcing its conduct-regulating rules even if neither the violator nor thevictim is domiciled in that state, and even if both parties are domiciled in the sameforeign state. For example, an Austrian motorist involved in a French accident maynot claim exemption from French traffic rules, and if injured by conduct that violatesthese rules, France may not deny her the benefit of their protection. Even if bothparties are domiciled in Austria, France has the exclusive claim to apply its law to theconduct-regulating aspects of the case. These aspects are notlimited to rules of publiclaw or pure traffic rules such as speed limits and red lights; they also extend to rulesthat impose civil liability to violations of traffic rules or attach presumptions andinferences of fault that arise from certain violations, such as not maintaining sufficientdistance from the preceding car. Because the common-domicile rule of Rome II is


24/46


25/46


115. For a comparative discussion, see SYMEONIDES,PROGRESS ORREGRESS, 31-35.

116. R OM E II, art. 3(2).

117. Id .118. This escape is also repeated in the Articles 10-12, dealing with unjust enrichment, negotiorum

gestio , and culpa in contrahendo, which are not discussed here.

119. For comprehensive discussions of this subject, see D. KOKKINI-IATRIDOU, LES CLAUSESDEXCEPTION EN MATIRE DE CONFLITS DE LOIS ET DE CONFLITS DE JURIDICTIONS OU LEPRINCIPE DE PROXIMIT (1994).

120. Aris totle ,NICOMACHEAN ETHICS , V. x 7.

121 . Peter Hay ,Fle xibility Versus Predicta bility and U niformity in Choice of Law , 226 RECUEIL DESCOURS 281, 291 (1991-I).


found in recent European codifications and international conventions, the escape115

authorizes the court to apply the law of another country if it is clear from all thecircumstances of the case that the tort/delict is manifestly more closely connected with

[that other] country. Paragraph 3 provides an example by stating that a manifestly116closer connection might be based on a pre-existing relationship between theparties, such as a contract, that is closely connected with the tort/delict in question.117

In contrast to the preliminary draft, which limited the scope of the escape to casescovered by the general rule, the final text repeats the escape in the articles dealing withproducts liability (Art. 5(2)), unfair competition cases in which the competition affectsexclusively the interests of a specific competitor (Art. 6(2)), and choice-of-lawagreements (Art 14(2)).118

Despite serious reservations about the scope and wording of this particularescape, this author applauds the drafters for including an escape in the final version

of Rome II. Indeed, escapes clause are necessary in any less than perfect statutoryscheme. Because perfection is not for this world and more and more modernlegislatures have begun to recognize their fallibility, escapes have become a commonfeature of almost all recent codifications. As Aristotle recognized many centuries119

ago, any pre-formulated rule, no matter how carefully or wisely drafted, may, due toits generality, or because of its specificity, produce results that are contrary to the120

purpose for which it was designed. This is a natural consequence of the differencebetween law makingand law application. The question here is to what extent this121

escape will help cure the deficiencies of the general rules of Rome II.

1. The Closer Connection Exception

The final phrasing of the escape clause is a significant improvement over thatof the Commissions preliminary draft proposal, which was based on the failed EEC


26/46


122. See supra II.1.

123. See, e.g., BELGIAN PILACT , art. 19(1); SWISS PILACT , art.15(1); QUEBEC CIV.COD E art. 3082.For a softer phrasing, see ENG LISH PILACT . 12.

124. See supra note 115; AUSTRIAN PILACT , 48(1); EGBGB art. 41(1).

125. See supra III.4.


draft convention of 1972 and later emulated by several national codifications.122 123

That escape was very problematic because it consisted of two independent prongs,both of which had to be satisfied for the escape to apply. One had to show (a) that

there was no significant connection to the country whose law governed under thedrafts rules; and (b) that there was a substantially closer connection with anothercountry. The problem with this scheme was that, if taken literally, the first prongwould rarely be satisfied, thus making the second prong nothing more than cosmetic.The problem was confounded by the fact that the escape did not contemplate acomparison of the two connections, but rather an independent determination that thefirst connection was not significant. Only when that determination confirmed theinsignificance of the first connection did the escape allow examination of thecloseness of the other connection. The final text resolved much of the problem byeliminating the first prong and by encouraging a comparison between the twoconnections. Under the final text, a party that invokes the escape need not show thatthe connection of the country whose law governs under the rule is insignificant. Allone needs to show is that the connection with another country is manifestly closerand this of course can only be determined after comparing the two connections. Thisis a significant improvement for which the drafters deserve praise.

Even with these changes, however, the escape remains problematic because,like its European counterparts: (a) it is phrased in exclusively geographical or124

quantitative terms that are not correlated to an overarching principle; and (b) it doesnot permit an issue-by-issue evaluation.

In one sense, it is logical that a system of geographically-based rules also relies

on geography when formulating escapes from those rules. Rome II is such a systembecause most of its dispositive rules depend on the place in which a single criticalevent occurred, or in which one or both parties reside. Very few non-geographicalfactors affect the choice, and the content of the conflicting laws is a factor that appearsonly in some narrow exceptions. Having relied on geography in erecting this125

system, the drafters may have felt bound to also rely on geography to handle theexceptional cases and overcome the inevitable impasses. Thus, if the geographicallychosen place of injury or of the parties common residence turn out to have a non-significant connection, one must search for a place that has a closer connection. Thislogic, however, will rarely overcome the impasses. Escapes are designed to cure therules deficiencies, not to reproduce them. To intelligently employ the escape, one

must know the reasons for which the drafter made the choices embodied in the rule


27/46


126. AMERICAN LAW INS TIT UT E,RESTATEMENT (SECOND) OF CONFLICT OF LAW S, 146 (emphasisadded).

127. LA.CIV.COD E ANN . Art. 3547 (emphasis added).


and the values and goals the rule seeks to promote. To simply say that one should lookfor a closer connection gives courts little meaningful guidance and entails the riskof degenerating into a mechanical counting of physical contacts. This risk is reduced

when the escape is correlated to the overarching principles that permeate the rules,and/or when the escape allows an issue-by-issue evaluation.

For purposes of illustration, not emulation, one can consider the schemes ofthe Restatement (Second) and the Louisiana codification. The Restatement providesin Section 6 that the goal of the choice-of-law process is to identify the state that hasthe most significant relationship. Although literally the quoted phrase appears tocontemplate a determination based on geography, the content of Section 6 negates anysuch inference because it lists a series of substantive policies intended to guide thisdetermination. The subsequent sections of the Restatement provide specific rules,most of which contain an escape authorizing the judge to apply the law of another

state if with respect to the particular issue that state has a more significantrelationship under the principles stated in 6. Similarly, Article 3542 of the126

Louisiana codification enunciates the general goal of the choice-of-law process for tortconflicts as one of identifying the state whose policies would be most seriouslyimpaired if its law were not applied. After establishing specific rules based on thatgoal, the codification also provides an escape clause in Article 3547 which authorizesthe judge to apply the law of another state if, under the principles of Article 3542,the policies of that other state would be more seriously impaired if its law were notapplied to the particular issue.127

The italicized phrases signify what is missing from the escape of Article 4(3)

of Rome IIissue-by-issue evaluation and correlation to non-geographicaloverarching principles. The only hope comes from a statement in recital 14, whichappears to articulate the philosophy of Rome II as a whole:

The requirement of legal certainty and the need to do justice inindividual cases are essential elements of an area of justice. ThisRegulation provides for the connecting factors which are the mostappropriate to achieve these objectives. Therefore, this Regulationprovides for a general rule but also for specific rules and, in certainprovisions, for an escape clause which allows a departure from theserules where it is clear from all the circumstances of the case that thetort/delict is manifestly more closely connected with another country.This set of rules thus creates a flexible framework of conflict-of-lawrules. Equally, it enables the court seised to treat individual cases in


28/46


128. R OM E II, recital 14 (emphasis added).

129. R OM E II, art. 4(1) (emphasis added).

130. See the EEC Draft Convention, supra at 20.


an appropriate manner.128

This recital is what survives from the rapporteurs and Parliaments efforts to

inject more flexibility into the text of the general rule of Rome II. As noted earlier, theCouncil and Commission rebuffed these efforts and recital 14 is the resultingcompromise. While much of this recital is self-congratulatory and merely describesthe scheme of Rome II, the italicized phrases can be viewed as providing instructionto courts on when and how to use the escape: to treat individual cases in anappropriate manner and to do justice in individual cases. Thus, a court shouldresort to the escape when the law designated as applicable by the general rule leadsto a result that is incompatible with the need to do justice in individual cases.

To be sure, one can objectand many willto such a loose reading of theescape as giving a licence for loose, ad hoc, subjective judging. While this danger isreal, it is worth the price. The alternative is a quantitative employment of the escape,which will resolve only the easiest of conflicts. For example, a quantitativeemployment of the escape will not resolve the aforementioned Kenyan huntingaccident case involving a French and a Belgian hunter, whereas an employment of theescape with a view toward doing justice in the individual case would.

The second major problem with the escape is its failure to allow an issue-by-issue deployment and evaluation. As it is, the escape contemplates situations in whichthe entire tort/delict is manifestly more closely connected with another country.Paragraph 3 of Article 4 not only avoids using the dirty word issue but also avoids(perhaps unintentionally) the phrase used in paragraph 1, which speaks of the lawapplicable to a non-contractual obligation arising out of a tort/delict. Thus, the129

escape does not even permit the court to look separately to the possibly multipleobligations that may arise from the same facts, such as when (but not only) the caseinvolves multiple tortfeasors or victims. Instead, the phrasing of the escape forces130

the court to only look at the tort as a whole. If the court finds that the tort as a wholehas a closer connection with another country, then the court is authorized to displacethe otherwise applicable law in its entirety and replace it with the law of that othercountry. Thus, the escape is an all or nothing proposition; and therein lies its mostserious flaw.

This flaw will make the escape unavailable in all but the obvious cases. Forexample, in the above snow avalanche hypothetical, it would be difficult to convince

a court that the tort/delict (as opposed to the specific issues of negligenceper se ordamages limitations) is manifestly more closely connected with Switzerland. Evenif the victims domicile, England, also allowed unlimited damages, the court could not


29/46


131. R OM E II, art. 15 (f) & (e).


avoid the application of French law because the connection with England (whichArticle 4(1) a priori condemns to a state of insignificance) will probably not qualifyas manifestly closer than the connection with France. The same would be true with

other issues affecting the victims recovery. Suppose, for example, that the avalanchecaused the death of one of the English tourists, and one of the issues in the case is whois entitled to compensation for his wrongful death. Suppose that French law providesthat compensation is due to the victims surviving spouse and children together,whereas English law provides that compensation is due to the surviving spouse to theexclusion of the children. Under Article 4(1), the applicable law shall be the law ofthe country in which the damage occurs (France) irrespective of the country orcountries in which the indirect consequences of that event occur (England). Article15 reiterates that in such a case, French law will apply to virtually all issues likely toarise in tort litigation, including [which] persons [are] entitled to compensation andwhether a right to claim damages or a remedy may be transferred, including byinheritance. Yet England is the country most intimately involved and has the best131

claim to apply its law to this issue. Rules designating the beneficiaries of a wrongfuldeath action reflect a societys assumptions regarding how a persons death impactshis survivors and which of his survivors are likely to have the highest need forcompensation. These assumptions and value judgments belong to the society in whichthe victim lived, not to the society in whose territory the injury occurred. In Rome IIsterminology, England is manifestly more closely connected with regard to the issueof wrongful death beneficiaries, even if its connections with regard to other issues maynot be the closest. Yet, the phrasing of the escape clause does not permit this focus onthe specific issue, and therefore does not allow a court to cure the rules deficiency.

Similar problems are encountered in employing the manifestly closerconnection escape to cases falling within the scope of the common-domicile rule,especially those in which, as noted earlier, that rule is either too broad or too narrow.In the case of the French traffic accident involving two Austrians, one could arguethat, with regard to issues of conduct and safety, France has a manifestly closerconnection than Austria, and thus French law should govern. The problem with thisotherwise sound argument is that it runs against the restrictive and holistic wordingof the escape, which does not allow an issue-by-issue an

Documents

roma 2 simionidis.pdf