Rome II and Tort Conflicts: A Missed Opportunity Abstract Contents

Rome II and Tort Conflicts: A Missed Opportunity 

By 

SYMEON C. SYMEONIDES * 

[To be published in 56 AMERICAN JOURNAL OF COMPARATIVE LAW (2008) 

All rights reserved] 

Abstract 

This article reviews the European Union’s 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 “approaches,” 

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 system 

more workable. 

The author’s 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 rich 

codification 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 ...................................................2 

II. HISTORY .......................................................4 

A. Before Rome II............................................4 

B. Rome II ..................................................5 

III. GENERAL FEATURES..............................................7 

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

*. 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 School. 

1

SYMEON C. SYMEONIDES ROME II AND TORT CONFLICTS 

B. Structure. ................................................8 

C. Certainty vs. Flexibility......................................8 

D. Jurisdiction-Selection or Content-Oriented Law-Selection. . . . . . . . . . 9 

E. State or National Interests...................................12 

F. Issue-by-Issue Analysis and Dépeçage. ........................13 

IV. THE GENERAL RULE.............................................15 

V. THE EXCEPTIONS................................................20 

A. The List.................................................20 

B. The Common Habitual-Residence Rule. . . . . . . . . . . . . . . . . . . . . . . . 21 

C. The General Escape........................................24 

1. The Closer-Connection Escape. . . . . . . . . . . . . . . . . . . . . . . . . 25 

2. The Pre-existing Relationship Exception.. . . . . . . . . . . . . . . . . 31 

D. Compensation for Traffic Accidents.. . . . . . . . . . . . . . . . . . . . . . . . . . 32 

VI. PRODUCT LIABILITY.............................................34 

VII. ENVIRONMENTAL TORTS. .......................................37 

VIII. RULES OF “SAFETY AND CONDUCT”...............................39 

IX. PARTY AUTONOMY. ............................................43 

X. CONCLUDING THOUGHTS..........................................44 

1. The English text of the Regulation can be found in the Official Journal of the European Union, 

2007 O.J. (L 199) 40. Excerpts discussed in this essay are reproduced in an Appendix at ???, 

infra. 

2. ROME II, closing sentence. See also Treaty establishing the European Community, art. 249(2). 

3. ROME II, art. 32. 

I. INTRODUCTION 

On July 11, 2007, the European Parliament and the Council of the European 

Union adopted “Regulation (EC) No 864/2007 on the Law Applicable to 

1 

Non-Contractual Obligations” known as “Rome II.” In European Union parlance, 

2 

a regulation is “binding . . . and directly applicable” in all member countries without 

the need for implementing national legislation in each individual country. The 

3 

Regulation is scheduled to go into effect on January 11, 2009. It will preempt the 

4 

national choice-of-law rules of the European Union’s member states on noncontractual 

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

Unlike some other regulations which apply only within the European Union, Rome 

II will have “universal application,” in the sense that it will cover torts occurring both 

within and outside the Union, and it may lead to the application of the law of a nonmember 

state. Rome II is a dramatic step in the federalization or “Europeanization” 

4. Rome II will not apply to Denmark. See ROME II, recital (40), art. 1(4). 

5. Among the non-contractual obligations arising from acts other than torts are those arising from 

unjust enrichment, negotiorum gestio, and culpa in contrahendo. Rome II covers these 

obligations in arts. 10-12. This essay does not discuss these articles. 

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


of private international law (PIL) in the EU member states, a step that has been aptly 

characterized as the European conflicts revolution. 6 

7 

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

fact that the essay’s author has spent the last three decades laboring in the American 

conflicts vineyard, inevitably influences his views on the subject. Nevertheless, to the 

extent this essay criticizes Rome II, it is not—at least not consciously—for failing to 

emulate any American models, but rather for failing to take full advantage of the 

richness, sophistication, and progress of modern European PIL. Indeed, while 

American conflicts law was stumbling through a loud “revolution” and then engaging 

8 

in endless navel-gazing about how to end it, European PIL was going through a quiet 

evolution, gradually repairing the old system and producing several noteworthy PIL 

9 10 11 12 13 

codifications, among them the Belgian, Dutch, English, German, and Swiss, 

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

6. Johan Meeusen, Instrumentalisation of Private International Law in the European Union: 

Toward a European Conflicts Revolution? 9 EUR. J. MIGR.& 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 Choice of 

Law Rules for Non-contractual Obligations, 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 Community’s “Rome II” Regulation, EUROPEAN LEGAL 

FORUM I-137 (4-2007); Cyril Nourissat & Edouard Treppoz, Quelques observations sur 

leavened-projet de proposition de reglement du Conseil sur la loi applicable aux obligations 

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

Regulation Rome II” in 2004: Its Past and Future Perspectives, 6 YBK PRIV. INT’L L. 129 

(2004); Ansgar Staudinger, Rome II and Traffic Accidents, EUROPEAN LEGAL FORUM 61 (4- 

2005); Peter Stone, The Rome II Proposal on the Law Applicable to Non-contractual 

Obligations, EUROPEAN LEGAL FORUM 213 (4-2004); von Gerhard Wagner, Internationales 

Deliktsrecht, die Arbeiten an der Rom II Verordnung und der Europäische 

Deliktsgerichtsstand, IPRAX 372 (2006); Russell J. Weintraub, Rome II and the Tension 

Between Predictability and Flexibility, 19 RIV. DIR. INT’LE PRIV. E PROCESS 561 (2005). 

8. See SYMEON C. SYMEONIDES, THE AMERICAN CHOICE-OF-LAW REVOLUTION: PAST, PRESENT 

AND FUTURE (2006) [hereinafter THE CHOICE-OF-LAW REVOLUTION]. 

9. See CODE DE DROIT INTERNATIONAL PRIVÉ (Loi du 16 juillet 2004, Moniteur Belge 27 Juillet 

2004 (1), hereafter “BELGIAN PIL CODE.” 

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

effective 1 June 2001 (hereafter “DUTCH PIL ACT”). For English translation with an 

Introductory Note by P. Vlas, see NETHERLANDS INT’L L. REV. 221 (2003-2). 

11. See PRIVATE INTERNATIONAL LAW (MISCELLANEOUS PROVISIONS) ACT OF 1995 (c 42), 8 

November 1995, hereafter cited as “ENGLISH PIL ACT.” 

12. See FEDERAL ACTS OF 1986 AND 1999 FOR THE REVISION OF PRIVATE INTERNATIONAL LAW 

(AMENDING THE INTRODUCTORY LAW TO THE CIVIL CODE (EGBGB)) (Transl. by Wegen, 27 

I.L.M. 1, 18 (1988), and Hay, 47 AM. J. COMP. L. 650 (1999)), hereafter referred to as EGBGB. 

13. See FEDERAL LAW OF 18 DECEMBER, 1987 ON PRIVATE INTERNATIONAL LAW (Transl. by Cornu, 

Hankins & Symeonides, 37 AM. J. COMP. L. 193 (1989), hereafter cited as “SWISS PIL ACT.” 



codifications, thus underscoring the fact that law reform is far more difficult at the 

multinational than the national level. 

A. Before Rome II 

II. HISTORY 

In recent European history, the first effort to codify choice of law for torts at 

a multinational level was the Benelux Uniform Law on Private International Law of 

14 

1969, which was based on work that began as early as 1951. Although it never 

entered into force, the Uniform Law was surprisingly modern for its day. Its basic 

rule for tort conflicts was to apply the law of the place of tortious conduct, but subject 

to a wisely drafted escape clause. The escape provided that, “if the consequences of 

a wrongful act belong to the legal sphere of a country other than the one where the 

act took place, the obligations which result therefrom shall be determined by the law 

15 

of that other country.” This escape was phrased broadly enough to cover not only 

cross-border torts in which the consequences of conduct in one country are felt in 

another country, the country of injury, but also cases such as the one in the famous 

16 

New York case Babcock v. Jackson, namely intra-state torts in which both parties 

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

point later. 

The next major effort took place in 1967, in the context of the then European 

Economic Community. By 1972, this effort produced the E.E.C. Draft Convention 

17 

on Contractual and Noncontractual Obligations. Its main provision on tort conflicts 

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 situation 

resulting from the event which caused the damage and the State in 

which the event occurred and if, on the other hand, such situation has 

a predominant connection (connexion prépondérante) with another 

14. See Kurt H. Nadelmann, The Benelux Uniform Law on Private International Law, 18 

AM.J.COMP.L. 406 (1970). An English translation of the Uniform Law is reproduced id. at 420- 

425. 

15. BENELUX UNIFORM LAW, Art. 14 (emphasis added). 

16. 191 N.E.2d 279 (N.Y. 1963). Babcock applied New York law to a suit by a New York guestpassenger 

against her New York host-driver and his insurer for injury sustained in a traffic 

accident in Ontario, during a short trip there. Ontario’s guest-statute prohibited the suit, while 

New York law allowed it. 

17. For an English translation of the Draft Convention, see 21AM. J. COMP. L. 587 (1973). For 

discussion, see Kurt H. Nadelmann, Impressionism and Unification of Law: The EEC Draft 

Convention on the Law Applicable to Contractual and Non-Contractual Obligations, 24 AM. 

J. COMP. L. 1 (1976). 



State, the law of that State shall apply. 18 

The escape continued with an explicit common-domicile presumption for Babcock 

19 20 

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

in cases of multiple victims, “the applicable law shall be determined separately for 

21 

each of them.” Finally, the draft convention included a special provision, also 

22 

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

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 discussed 

24 below. 

With the expansion of the Community to nine members states in 1973, 

following the accession of the United Kingdom, Ireland, and Denmark, the effort 

slowed down. It slowed to the point that the decision was made to abandon the tort 

provisions of the draft convention and instead to concentrate on contract conflicts. 

25 

In 1980, the Convention on the Law Applicable to Contractual Obligations, now 

known as the “Rome Convention,” was opened for signature, and it entered into force 

on April 1, 1991. 

B. Rome II 

The idea of addressing tort conflicts received new impetus with the Treaty of 

Amsterdam of October 2, 1997, when the European Commission began soliciting 

feedback on measures to implement the provisions of the Treaty regarding judicial 

cooperation in civil matters with cross-border impact. By 1998, the idea of a 

convention on tort conflicts was adopted in principle and, later that year, the first 

draft proposal was put forward by the Groupe européen de droit international privé 

18. EEC DRAFT CONVENTION ON CONTRACTUAL AND NONCONTRACTUAL OBLIGATIONS, art. 10(2). 

19. See id. at art. 10(3) (“Such a connection must normally be based on a connecting factor common 

to 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 not 

only in applying (or not applying) the common-domicile presumption, but also in applying the 

general, “closer connection” escape. 

22. See art. 7 of the HAGUE CONVENTION OF 4 MAY 1971 ON THE LAW APPLICABLE TO TRAFFIC 

ACCIDENTS & art. 9 of the HAGUE CONVENTION OF 2 OCTOBER 1973 ON THE LAW APPLICABLE 

TO PRODUCTS LIABILITY. 

23. See EEC DRAFT CONVENTION art. 12. (“Irrespective of which law is applicable under Article 

10, in determining liability account shall be taken of rules of safety and public order in force at 

the place and time of the event which caused the damage.”). 

24. See ROME 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. 



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

has influenced the general content and coverage of Rome II, although not where it 

matters most. In May 2002, the European Commission published a preliminary draft 

27 proposal and invited comments from the public. In addition to holding public 

hearings, the Commission received, and posted on the internet, more than eighty 

written commentaries from interested parties, including trade and industry groups, 

28 

professional organizations, governments, practitioners, and academics. Although 

the majority of the commentaries came from industry groups and tended to support 

the most regressive elements of the preliminary draft, the whole process is a good 

example of European democracy at work. 

29 

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

accompanied by a detailed Explanatory Report (hereafter “Report”) and an article-byarticle 

commentary. This Report is the only complete explanation of the thinking 

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

provisions. 

The process then shifted to the European Parliament where the rapporteur, 

30 

British MEP Diana Wallis, continued to solicit feedback, especially from 

31 

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

Commission’s proposal and, on July 6, 2005, she succeeded in having Parliament 

26. See Proposal for a European Convention on the Law Applicable to Non-contractual 

Obligations adopted at the Luxembourg meeting of Sept. 25-28, 1998, available at 

http://www.drt.ucl.ac.be/gedip/gedip-documents-8pe.html. 

27. See Preliminary Draft Proposal for a European Council Regulation on the Law Applicable to 

Non-Contractual Obligations, available at http://europa.eu.int/comm.justice_home/ unit/civil/consultation/I 

ndex_en.html [hereinafter Commission, Preliminary Draft Proposal]. 

28. For a comprehensive academic proposal put forward during this period by the Hamburg Group 

of Private International Law, see Comments on the European Commission’s Draft Proposal for 

a European Council Regulation on the Law Applicable to Non-Contractual Obligations (Oct. 

10, 2002), 67 RABELSZ 1 (2003). 

29. See Commission of the European Communities, Proposal for a Regulation of the European 

Parliament and the Council on the Law Applicable to Non-Contractual Obligations, (COM 

427) (2003) final, 2003/0168(C)D), Brussels, (July 22, 2003) [hereinafter Commission 

Proposal]. For a critique of this proposal, see Symeon C. Symeonides, Tort Conflicts and Rome 

II; A View from Across, in FESTSCHRIFT FÜR ERIK 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 leader 

of the Liberal Democrat European Parliamentary Party (LDEPP). She is an English solicitor 

who also studied law in Belgium, Germany, and Switzerland and is fluent in French and 

German. For more information on her background and work in the European Parliament, see 

http://www.dianawallismep.org.uk/. 

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

academics consulted by the rapporteur. 



32 

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

representing the member states, and the Commission rejected most of the amendments. 

On second reading, Parliament moved closer to the Commission’s positions, 33 

34 

however, the Council and Commission remained unsatisfied. The process then 

shifted to a Conciliation Committee, consisting of representatives of the Council, 

Commission, and Parliament. The ensuing “trialogue” bridged the differences with 

a compromise text that was adopted by Parliament on third reading, on July 11, 

35 2007. This is the final text and is referred to hereafter as “Rome II.” 

A. Scope 

32. See EUR. PARL., P6_TA(2005)0284. (Codecision procedure: first reading.. 

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

34. See Common Position (EC) No 22/2006 adopted by the Council on Sept. 25, 2006, Oy C 

289E/68 (Nov. 28, 2006). 

35. See supra note 1. 

III. GENERAL FEATURES 

Article 1 of Rome II defines the scope of the Regulation. It applies to 

non-contractual obligations in “civil and commercial matters,” in “situations 

involving a conflict of laws,” namely, situations having multistate contacts of the 

36 

kind and pertinence that implicate the laws of more than one state. The Regulation 

does not apply to “revenue, customs or administrative matters or to the liability of the 

State for acts and omissions in the exercise of State authority (acta jure imperii),” or 

to obligations arising out of: marriage, family or similar relationships; matrimonial 

property regimes; wills and succession; bills of exchange, cheques, promissory notes 

and other negotiable instruments; the law of companies; relations between the 

settlors, trustees and beneficiaries of voluntary trusts; nuclear damage; violations of 

37 

privacy and rights relating to personality, including defamation. The last topic was 

initially included in the scope of Rome II, but the proposed article proved most 

controversial attracting more commentary, especially from the publishing industry, 

than any other provision. Eventually, the article was dropped, but a Commission 

Statement accompanying Rome II pledges to study the subject again and report to 

38 

Parliament by December 2008. 

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 Articles 

21 and 22, which deal with the proof of juridical acts and presumptions regarding the burden 

of proof in torts. 

38. For recent academic commentary on the subject, see Christopher J. Kunke, Rome II and 

Defamation: Will the Tail Wag the Dog? 19 EMORY INT’L L. REV. 1733 (2005); Aaron 

Warshaw, Uncertainty from Abroad: Rome II and the Choice of Law for Defamation Claims, 

(continued...) 



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, which 

contains the general rule for all torts (Art. 4), followed by special rules on product 

liability (Art. 5), unfair competition (Art. 6), environmental damage (Art. 7), 

infringement of intellectual rights (Art. 8), and industrial action (Art. 9). Chapter III 

deals with unjust enrichment, negotiorum gestio, and culpa in contrahendo, chapter 

IV 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 law 

of the place of the injury (lex loci damni). The rule is followed by an exception in 

favor of the parties’ common habitual residence, and by a general escape clause 

based on the “closer connection” principle. Articles 5 to 9 are phrased as either 

particular applications or clarifications of the general rule for certain torts, or as 

exceptions to the general rule. Before discussing these articles, it is helpful to identify 

some of the general methodological features of Rome II, beginning with the balance 

it strikes between certainty and flexibility. 

C. Certainty vs. Flexibility 

Every PIL system encounters the perennial tension between the need for 

certainty and predictability on the one hand, and the need for flexibility and equity on 

the other. Each system responds differently, striking a different equilibrium between 

39 

the two needs. For example, the GEDIP proposal, which consisted of a series of 

cascading presumptions rather than rules, was a deliberate and brave choice of 

flexibility over certainty. Although the GEDIP scheme was nearly perfect, the 

potential for uncertainty must have alarmed the Commission, because it moved 

sharply in the opposite direction of adopting a system of tightly written black-letter 

rules with relatively few escapes and little room for judicial discretion. The rapporteur 

40 

and Parliament tried to re-inject flexibility into the draft, but, as the final text 

reflects, the Commission and especially the Council remained unyielding. The only 

provision that remains from the Parliament’s efforts is a recital in the Preamble 

referring to the “need to do justice in the individual case,” juxtaposing it with the 

38. (...continued) 

32 BROOKLYN J. INT’L L. 269 (2006). 

39. For a comparative discussion of the various choices, see SYMEON C. SYMEONIDES, PRIVATE 

INTERNATIONAL LAW AT THE END OF THE 20TH CENTURY: PROGRESS OR REGRESS?, 21-35 

(1999) [hereinafter Symeonides, “PROGRESS OR REGRESS”]. 

40. See, e.g., the rapporteur’s Report to the European Parliament, in Final A6-0211/2005, (June 

27, 2005). 



41 

“requirement of legal certainty.” However, as explained later in discussing the 

42 

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

The reasons for the Council’s and Commission’s political preference for 

certainty over flexibility are obvious. The primary motive behind the movement to 

draft Rome II, as well as the choice of the particular instrument for its implementation—a 

regulation as opposed to a directive—was the need to ensure uniformity of 

43 

choice-of-law decisions within the European Union. These two bodies must have 

concluded that uniformity would be in jeopardy if Rome II were to have too many 

flexible 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 of 

different 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 and 

certainty, one can question the extent to which these values should displace all other 

values of the choice-of-law process, such as the need for sensible, rational, and fair 

decisions in individual cases. If the American experience has something to offer, it is 

a reminder that a system that is too rigid—as the traditional American system was— 

ultimately fails to deliver the promised predictability because, in a democratic society 

44 

no system can “mechanize judgment” and, to the extent it attempts to do so, judges 

45 

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

system. Its drafters were conscious of the need for flexibility and they attempted to 

provide for some degree of it. The question is whether the drafters provided enough 

flexibility, 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 competing 

values of certainty and flexibility. For what it is worth, and for reasons that will 

become obvious from the following discussion, this author’s opinion is that a scheme 

that is closer to the GEDIP proposal or the rapporteur’s ultimately defeated proposals 

would have been preferable. 

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

41. See ROME II Preamble, recital (14). 

42. See infra V.2-3. 

43. See ROME II, recital (6) (“The proper functioning of the internal market creates a need, in order 

to improve the predictability of the outcome of litigation, certainty as to the law applicable and 

the free movement of judgments, for the conflict-of-law rules in the Member States to designate 

the same national law irrespective of the country of the court in which an action is brought.”). 

44. David F. Cavers, Restatement of the Law of Conflict of Laws, 44 YALE L. J. 1478, 1482 (1935). 

45. In Professor Weintraub’s words, “[i]ronically Rome II is more likely to succeed in providing 

reasonable foreseeability if its rules provide sufficient flexibility.” Weintraub, supra note 7, at 

id. 



One difference between traditional and modern PIL is the extent to which each 

considers the content of the substantive laws of the involved states before choosing 

the 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”), 

46 

without regard to the content of their substantive laws. In contrast, in modern PIL, 

the choice is based not only on physical contacts but also on the content and 

underlying policies of the laws of the contact states (“content-oriented law 

47 

selection”). Content-oriented law selection is easier when the court is not bound by 

statutory choice-of-law rules. When such rules exist, however, a content-oriented law 

selection is possible only to the extent these rules allow it. The question here is to 

what extent the rules of Rome II require jurisdiction-selection and to what extent they 

permit 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 dispositive 

articles of Rome II designate the state whose law governs because of that state’s 

physical contacts rather than because of the content of its law. This alone is not a 

reason to criticize Rome II. After all, in most codified PIL systems, jurisdictionselecting 

rules are bound to outnumber content-oriented rules. Secondly, as 

48 

documented elsewhere, carefully crafted, narrowly tailored jurisdiction-selecting 

rules can achieve functionally sound results. For this to happen, however, the drafters 

must identify the various typical law-fact patterns and consider whether the 

application of a given state’s law produces an equally good result regardless of the 

content of that law. If the answer to this question is affirmative, a jurisdiction- 

49 

selecting rule would be acceptable. If the answer is negative, then one or more 

content-oriented rules should be drafted to accommodate the various patterns. One 

question explored later in this essay is whether the jurisdiction-selecting rules of 

Rome II meet this test. 

Before addressing this question, it is worth noting that Rome II contains 

several provisions that, directly or indirectly, permit consideration of the content of 

the substantive laws of the involved countries. These provisions can be divided into 

three categories: 

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

46. The term “jurisdiction-selection” was first coined by Professor Cavers. See David F. Cavers, 

A Critique of the Choice-of-Law Problem, 47 HARV. L. REV. 173 (1933). 

47. For a full discussion of these concepts, see Symeon C. Symeonides, American Conflicts Law 

at the Dawn of the 21st Century, 37 WILLAMETTE 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. As 

explained infra at text accompanying note 112, this rule produces good results not only when 

the law of the common domicile favors the victim, but also when it favors the tortfeasor. 



applicable law to the extent it violates the forum’s mandatory rules (Article 

16) or its ordre public (Article 26); 

(2) provisions that require the court to choose a law that produces a particular 

result, such as Article 18, which authorizes a direct action against the insurer 

if such action is allowed by either the law applicable to the tort or the law 

applicable to the insurance contract; and Article 14(2) and (3) which provide 

that the contractually chosen law shall be disregarded to the extent it violates 

certain 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 choose 

between the law of the place of conduct and the law of the place of injury; and 

Article 6(3)(b), which allows the plaintiff to choose the law of the forum in 

certain cases involving anti-competitive restrictions. 

Although some of these content-oriented provisions have long been accepted 

by traditional PIL, the presence of the remaining ones suggests a selective receptiveness 

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 to 

choose a state because of its contacts rather than its law, in fact are based on certain 

assumptions about the likely content of that law and to that extent they disguise 

specific substantive choices. One possible example from Rome II is Article 5 on 

products liability, which, as explained below, will lead to the application of the law 

50 

of the victim’s habitual residence in the great majority of cases. When the plaintiff 

is a resident of an EU country with a generous pro-consumer law and the manufacturer 

is based in a state with a pro-manufacturer law, Article 5 will ensure that the EU 

plaintiff will be protected under EU standards. However, in the converse situation, the 

same article will ensure that the EU defendant will be held accountable under the 

lower standards of the plaintiff’s foreign residence. 

Similar thinking seems to underlie recital 33 regarding damages for victims 

of traffic accidents. The recital states that, “when quantifying damages for personal 

injury in cases in which the accident takes place in a State other than that of the 

habitual residence of the victim,” the court should “take into account all the relevant 

actual circumstances of the specific victim, including in particular the actual losses 

51 

and costs of after-care and medical attention.” The history and meaning of this recital 

52 

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

equivocal wording, this recital does more than recognize the relevance of foreign facts 

(i.e., the costs of after-care and medical attention); it also recognizes the relevance of 

50. See infra text following note 154. 

51. ROME II, recital (33). 




foreign law as Parliament’s unsuccessful amendment expressly provided. Indeed, it 

is the law of the foreign state that makes these facts relevant, not the other way 

around. Stated another way, the fact that medical costs are high in the victim’s 

residence is not significant unless the victim would be entitled to recover medical 

costs under that state’s law. To this extent, recital 33 can be seen as an example of a 

content-oriented law selection. 

E. State or National Interests 

A casual look at Rome II confirms the impression that, like most traditional 

PIL systems in Europe, Rome II does not subscribe to the notion that ordinary 

conflicts disputes at the private-law level implicate the interests of the involved 

countries. Indeed, none of Rome II’s articles refer to state policies, much less interests, 

and both the Preamble and the Explanatory Report contain several statements 

describing Rome II’s goal as one of “ensur[ing] a reasonable balance between the 

interests of [the parties, i.e.,] the person claimed to be liable and the person who has 

sustained damage.” 53 

Yet, one who looks below the surface would discover that Rome II is not 

oblivious to state interests. In several instances, the Preamble refers to broader societal 

interests that reach beyond the interests of private litigants. For example, with regard 

to products liability, recital 20 of the Preamble speaks of the policies of “fairly 

spreading the risks, . . . protecting consumers’ health, stimulating innovation, securing 

54 

undistorted competition and facilitating trade.” With regard to unfair competition, 

recital 21 speaks of the need to “protect competitors, consumers and the general public 

55 

and ensure that the market economy functions properly.” Recital 25, with regard to 

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 of 

56 

“discriminating in favour of the person sustaining the damage.” Finally, recital 31 

recognizes the need to impose certain restrictions on the parties’ power to choose the 

governing law so as to protect the weaker parties. 57 

More importantly, some of Rome II’s dispositive articles can only be explained 

in 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 Article 

16 which allows the forum to interpose its own mandatory rules, many other 

provisions of Rome II are designed to be sensitive to certain preferred substantive 


54. Id. at recital (20). 






polices. The clearest example is Article 7 pertaining to environmental torts, which is 

58 

discussed in some detail later. Other less obvious examples are: Article 14(2)-(3), 

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-oflaw 

clauses only in contracts between commercial parties and subjecting them to 

certain limitations; and Article 6(4), prohibiting choice-of-law agreements in cases of 

unfair competition and restrictions to competition. 

F. Issue-by-Issue Analysis and Dépeçage 

Modern choice-of-law doctrine and practice have come to recognize that in 

many cases the conflict is confined to only some aspects or “issues” of the case, and 

that 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 case 

were in dispute, the modern decisionmaker focuses on the narrow issues with regard 

to which a conflict exists and proceeds accordingly. Such issue-by-issue analysis is 

easier and more likely when the decisionmaker is not bound by statutory choice-oflaw 

rules. When such rules exist, however, such an analysis is possible only to the 

extent the rules permit it. In turn, this depends on whether these rules are phrased in 

broad terms designating the law that would govern the case as a whole, or whether 

they 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 

byproducts—the phenomenon of dépeçage, namely the application of the laws of 

different states to different issues in the same case. Indeed, the Council and the 

Commission specifically rebuffed the rapporteur’s and Parliament’s efforts to 

introduce issue-by issue analysis. One of Parliament’s amendments on first reading 

had attached the following concluding sentence to what became the general rule of 

Article 4: “In resolving the question of the applicable law, the court seised shall, 

59 

where necessary, subject each specific issue of the dispute to separate analysis.” The 

Council and Commission rejected this amendment. In the end, most of Rome II’s 

articles were phrased in broad terms designating the law that would govern the case 

as a whole, the “tort/delict.” Article 15 reaffirms this holistic approach by providing 

a long list of issues that comprise the scope of the law applicable under these 

articles. 60 

58. See infra VIII. 

59. EUR. PARL. 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 to 

arise in tort litigation, including the basis and extent of liability, the grounds for exemption from 

liability, any limitation of liability and any division of liability, the existence, the nature and the 




Nevertheless, a closer look at some of the other articles reveals that Rome II 

does not—because it cannot—entirely avoid an issue-by-issue analysis and thus the 

possibility of dépeçage. First, some of Rome II’s articles speak of the law applicable 

to the “obligation” arising out of a tort or delict, rather than to the tort or delict as a 

whole. This is a wise choice of terms, which will enable courts to engage in a more 

individualized evaluation of the multiple obligations that may arise from the same 

events. Moreover, if Rome II uses the term obligation in its original sense of the legal 

bond between the obligor and the obligee that encompasses both the obligor’s duties 

and the obligee’s rights, the court will be able to further differentiate among the rights 

of the individual victims and proceed accordingly. 

Rome II consciously stops short of the next logical step of encouraging or even 

permitting an issue-by-issue analysis of the various aspects of the obligation. Even so, 

some of Rome II’s other articles contain the possibility for further splitting the issues. 

Among the latter articles are: 

61 

(a) Article 8(2) on intellectual property rights; 

62 

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

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

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

65 

(e) Article 18 on direct actions against insurers; 

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


assessment of damage or the remedy claimed; injunctive relief, the proper beneficiaries of the 

right to claim damages, and its transferability or heritability, respondeat superior, and the 

extinction 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 to 

other issues. Art, 27 can also lead to the same phenomenon by recognizing that a case may be 

governed partly by the law designated by Rome II and partly by the law designated by other 

choice-of-law rules contained in other community instruments. 

62. Art. 14 can lead to dépeçage through several different routes. For example, the parties may 

choose different laws for different aspects of the case, or they may choose one law for some 

issues and none for others. Even when they choose one law for all issues, that law will not be 

applied to the extent it violates the mandatory rules described in art. 14(2) or the community 

rules described in art. 14(3). 

63. Arts. 16 and 26 allow courts to displace the applicable law to the extent necessary to satisfy the 

mandatory rules of the forum state, or to the extent that law is incompatible with the forum’s 

ordre public. 

64. Art. 17 (discussed infra VIII) allows the court to “take account” of the conduct rules of the 

conduct state when the rest of the case is governed by the law of another state. 

65. Under art. 18, the victim’s right to directly sue the tortfeasor’s insurer may be governed by the 

law governing the insurance contract, although the tort may be governed by another law. 



66. Under arts. 19 and 20, the rights of subrogation, indemnification, and contribution between the 

parties mentioned in the articles may be governed by a law other than the law governing the 

victim’s claims against these parties. 

67. See supra text accompanying note 51. 

68. In fact, the possibility of dépeçage is the main reason cited by the Council and Commission for 

rejecting Parliament’s express rule to that effect. Although the recital does not enjoy the same 

status as an express rule in the Regulation’s main body, the recital either means what it says, in 

which case the possibility of dépeçage remains, or it does not mean what it says, in which case 

Parliament did not get anything from the purported compromise. 

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

70. ROME II, art. 4(1). 

71. Id. 

66 

contribution; and 

67 

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

of the law of the habitual residence of the victim of a traffic accident in 

quantifying the recoverable damage, even when all the other issues resulting 

from the accident are governed by the law of the accident state. 68 

The above is a much longer list than the opponents of dépeçage would 

ordinarily 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 the 

possibility of dépeçage) would have been most useful. The resulting problems are 

discussed later. 69 

IV. THE GENERAL RULE 

Rome II’s central provision is Article 4, which contains the general and 

residual rules. Paragraph 1 of Article 4 provides that the applicable law shall be the 

70 

law of the country in which “the damage occurs” (lex loci damni). This law governs 

“irrespective of the country in which the event giving rise to the damage occurred” 

and “irrespective of the country or countries in which the indirect consequences of that 

event occur.” 71 

The operation of this rule can be illustrated by the following hypothetical 

scenario, which is used throughout this essay: Blasting operations by a Swiss mining 

company 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, it 

seems that Article 4(1) views Switzerland as the country of the “event giving rise to 

the damage,” France as the country in which “the damage occurs,” and England as the 

country in which “the indirect consequences of that event occur.” Translated into 

simpler English, Article 4(1) provides that the applicable law is the law of the country 

in which the injury occurs, and more precisely the harmful physical impact (France), 



irrespective of the country in which the injurious conduct occurred (Switzerland), and 

irrespective 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 traditional 

lex loci delicti rule, with its “last event” sub-rule. It purports to be as categorical as the 

corresponding rule of the American First Restatement. In its penchant to avoid any 

ambiguity, the Restatement provided numerous minute localization sub-rules which, 

for example, defined the place of injury as the place where “the harmful force takes 

effect upon the body” in personal injury cases, and the place where “the deleterious 

72 

substance takes effect” in cases of poisoning. The fact that the Restatement never 

attained certainty, despite having attained clarity, is a lesson that subsequent codifiers 

ignore at their peril. 

The Rome II codifiers note that “[t]he principle of the lex loci delicti commissi 

is the basic solution for non-contractual obligations in virtually all the Member 

73 

States,” which of course is true, except for the fact that in many countries this 

solution is subject to several exceptions. The drafters also correctly note that many 

74 

countries disagree in defining the locus delicti. Indeed, some countries opt for the 

75 76 

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

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

78 

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

79 

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

differences by unequivocally choosing the law of the place of injury, because such a 

solution “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 civil 

80 

liability and the development of systems of strict liability.” Neither of these two 

reasons are self-explanatory, and the second reason regarding strict liability is 

certainly debatable. As for the first reason, the only balance the lex loci damni rule 

strikes between the parties is that it can be equally unfair to the plaintiff in some cases 

as to the defendant in others. 

72. AMERICAN LAW INSTITUTE, RESTATEMENT OF CONFLICT OF LAWS § 377 Note (1933). 


74. See ROME II, recital (15). 

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

76. See DUTCH PIL ACT, art. 3(2); ENGLISH PIL ACT of 1995 § 11 (subject to exceptions). 

77. See PORTUGUESE CIV. CODE, art. 45(1) and (2); SWISS PIL ACT, art. 133(2). 

78. See SPANISH CIV. CODE art. 9; GREEK CIV. CODE, art.26; CZECHOSLOVAKIAN PIL ACT of 1963, 

art 15. 

79. See EGBGB art 40(1); HUNGARIAN PIL DECREE of 1979 § 32(1)(2); ITALIAN PIL ACT of May 

31, 1995, art 62(1). 




This is not to say that, just because the lex loci damni is an old rule, or just 

because it produces unfair results in some cases, the rule is bad in all cases. Despite 

the recent tendency, especially among American academics, to summarily reject this 

rule as an outmoded remnant of the past, a dispassionate examination of actual cases 

indicates 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 the 

81 

exceptions to which it is subject, and, second, examine the results the rule produces 

in several typical patterns formed by the aggregation or disbursement of the pertinent 

82 

contacts (conduct, injury, and parties’ domiciles ) and the content of the laws of each 

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

(1) whether the injurious conduct and the resulting injury occurred in the same 

state (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 resulting 

from 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-distributing 

rules corresponds to the two grand objectives of tort law—deterrence and reparation. 

Although these two objectives are often interconnected, conduct-regulating rules serve 

primarily 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 way 

through Article 17, which authorizes “taking account” of the “safety and conduct” 

83 

rules of the conduct state. In the United States, this distinction, which was first 

articulated by the New York Court of Appeals in the 1963 landmark case Babcock v. 

84 

Jackson, has been adopted by many courts, albeit without always using this 

85 

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

New York court, conduct-regulating rules are those that “have the prophylactic effect 

86 

of governing conduct to prevent injuries from occurring.” This category includes not 

only “rules of the road” like speed limits and traffic-light rules, but also rules that 

prescribe the civil sanctions for violating traffic rules, including presumptions and 

81. See infra IV. 

82. Hereafter, the term “domicile” is used interchangeably with the term “habitual residence” as 

used in Rome II. 

83. Art. 17 is discussed 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). 



87 

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

sites, buildings, and other premises; rules that impose punitive damages; and rules 

defining as tortious conduct such as “alienation of affections,” “interference with 

marriage,” or “interference with contract,” and granting an action against the actor. In 

contrast, loss-distributing rules are those that “prohibit, assign, or limit liability after 

88 

the tort occurs.” They include not only guest statutes, which now are virtually 

extinct, but also rules that define the amount of compensatory damages, and rules of 

interspousal immunity, parent-child immunity, worker’s compensation immunity, and 

loss of consortium. 

Admittedly, the line between the two categories is not always as bright as one 

would like. While some tort rules are clearly conduct-regulating and others are clearly 

loss-distributing, there are many tort rules that do not easily fit in either category, and 

some 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, this 

distinction provides a useful starting point for resolving or analyzing many tort 

conflicts, although in many such conflicts the distinction will not make a difference. 

The starting point is a presumption that conduct-regulating rules are territorially 

oriented and loss-distribution rules are not necessarily territorially oriented. 

Consequently, territorial contacts (namely the places of conduct and injury) remain 

relevant in conduct-regulation conflicts, while both territorial and personal contacts 

(i.e., the parties’ domiciles) are relevant in loss-distribution conflicts. 

89 

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

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, regardless 

of where the parties are domiciled; 

(2) intrastate torts in which the issue is one of loss distribution and either the 

tortfeasor or the victim is domiciled in the state of conduct and injury; and 

(3) cross-border torts involving either conduct-regulation or loss distribution 

conflicts, in which: (a) the state of injury prescribes a higher standard of 

conduct for the tortfeasor or of financial protection for the victim than the state 

of conduct; and (b) in which the occurrence of the injury in the former state 

was objectively foreseeable. 

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

87. Examples include rules providing that a person involved in a collision while driving in excess 

of the speed limit, or while being intoxicated, is presumed to be negligent, and rules providing 

that, in a rear-end vehicular collision, the driver of the rear car is presumed to be at fault. 

88. Padula, supra note 86, at id. 

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



injury passes the scrutiny of a functional choice-of-law analysis, regardless of whether 

that law favors the tortfeasor or the victim. Consequently, a jurisdiction-selecting rule 

like 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 the 

majority of cases. In the first pattern, the state of both the conduct and injury clearly 

has the strongest claim to apply its law to conduct within its borders, even if the 

90 

parties are both domiciled in the same foreign state. The lex loci rule honors this 

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

present the direct or true conflict paradigm and two the inverse or “no interest” 

paradigm. On balance, the application of the law of the state that has the two territorial 

contacts (conduct and injury) and one personal contact (the domicile of one party) is 

justified 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 interest 

in protecting its domiciliary victim injured there, while the state of conduct has an 

interest in protecting a tortfeasor acting (and often domiciled) there. In these cases, the 

application of the law of the state of injury is justified if the occurrence of the injury 

93 

in that state was objectively foreseeable. The lex loci damni rule of Rome II produces 

this result, but its failure to include a foreseeability defense may make it unfair to the 

defendant, depending on the other circumstances of the case. 94 

The lex loci damni rule does not produce good results in cases of the converse 

pattern, namely cross-border torts in which the state of conduct prescribes higher 

95 

standards of conduct for the tortfeasor than the state of injury. Suppose for example 

that, in the above snow avalanche hypothetical, Switzerland imposes a negligence per 

se 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 ordinary 

negligence standard. In such a case, Article 4(1) calls for the application of French law 

96 

and specifically excludes Swiss law. In contrast, if this were an environmental tort, 

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 lex 

loci rule in these cases, Rome II commits a serious error, which can only be corrected through 

the back door of art. 17. See infra text accompanying notes 113 & 172. 

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

REVOLUTION 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-LAW 

REVOLUTION 200-02, 223-28. 

96. Art. 17 allows the court to “take account,” but not necessarily to apply, the Swiss negligence 

per se rule on the assumption that it qualifies as a rule of “safety and conduct.” See infra VIII. 



97 

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

extend this option (which the Report characterizes as “the principle of favouring the 

98 victim”) to other torts , because “this solution would go beyond the victim’s 

99 

legitimate expectations.” 

This, however, is the wrong dilemma. The dilemma is not whether one should 

favor victims over defendants, but rather whether—in a subject called “conflict of 

laws”—one 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 to 

benefit victims as such, but rather out of deference to the policies of the state of 

conduct, which is the only state that has something to lose from the non-application 

of its law. In this case, Switzerland’s negligence per se rule is intended to deter people 

from engaging in inherently dangerous operations like blasting. Because the defendant 

acted within Swiss territory, Switzerland has every reason to insist in determining the 

legal consequences of that activity, even if in this case, the injury occurred across the 

border. The effectiveness of this policy of deterrence would be seriously impaired if 

exceptions to it were made for out-of-state injuries. Moreover, in terms of fairness and 

party expectations, there is nothing unfair in subjecting a tortfeasor to the law of the 

state in which he acted. Having violated the standards of that state, the tortfeasor 

should bear the consequences of such violation and not be allowed to invoke the lower 

standards of another state. Conversely, there is little reason to apply French law. Its 

ordinary negligence rule was intended to protect mining defendants who operate 

100 

within French territory, not foreign operators operating elsewhere. In conclusion, 

there is a good deal of wisdom in the rules that allow the victim or the court to choose 

between the laws of the state of conduct and the laws of the state of injury in cases of 

101 

cross-border torts. It is regrettable that the drafters of Rome II have chosen not to 

adopt a similar rule as they did with regard to environmental torts. 102 

A. The List 

97. Art. 7 is discussed infra VII. 

98. Report, art. 3, 11. 

99. Id. at 11-12. 

V. THE EXCEPTIONS 

As noted earlier, one cannot properly evaluate Rome II without also analyzing 

the exceptions to its basic rule of lex loci damni. A careful perusal of the whole 

100. Similar arguments could be made if France, but not Switzerland, limited the amount of damages 

that could be recovered from mining operators. Again, art. 4(1) would mandate the application 

of French law, albeit subject to the “closer-connection” escape of art. 4(3). However, for 

reasons explained later, this escape will not necessarily lead to Swiss law. 

101. See, e.g., EGBGB arts. 40.1, 41; SWISS PIL ACT Arts. 133(2), 136-39; ITALIAN PIL ACT art. 62; 

VENEZUELAN PIL ACT art. 32(2); HUNGARIAN PIL ACT art. 32(2). 

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



document reveals that this rule is subject to many more exceptions than are readily 

apparent. They can be divided into general exceptions, namely those that apply to all 

cases, 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 connection,” 

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 and 

conduct” rules of the state of conduct, under Article 17; 

(5) The application of the law chosen by the parties before or after the 

occurrence of the tort, under Article 14; 

(6) The non-application of the lex loci (or any other law) when it is manifestly 

incompatible 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 common 

domicile of the tortfeasor and the victim, the domicile of the victim, the state 

of the product’s acquisition, or the state of the “manifestly closer connection,” 

under Article 5; 

(8) The application of the law of the forum in certain cases involving 

restrictions to competition under Article 6(3)(b); 

(9) The application of the law of the state of conduct at the victim’s behest in 

environmental torts, under Article 7; and 

(10) The possible application of the law of the victim’s habitual residence for 

quantifying recoverable damages in traffic accident cases, under recital 33. 

Because of space limitations, this essay discuses only some of the 

exceptions. 103 

B. The Common Habitual-Residence Rule 

The first “official” exception to the lex loci rule is found in paragraph 2 of 

Article 4, which provides that, if at the time of the injury, the tortfeasor and the victim 

have their habitual residence in the same country (hereafter “common domicile”), then 

103. Additional exceptions can be found in arts. 10-12 in cases of unjust enrichment, negotiorum 

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



104 

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

repeated in Article 5 on product liability, Article 6 on unfair competition cases in 

which the competition affects “exclusively” the interests of a specific competitor, 105 

106 

and Article 9 on industrial action. In contrast, this exception does not apply to other 

unfair competition cases, cases falling within Article 7 on environmental torts, or 

Article 8 on infringement of intellectual property rights. One can infer that the reason 

for this exclusion is an implicit recognition that cases involving the latter categories 

of cases implicate broader societal interests that go beyond the interests of the 

litigants. 

In adopting the common-domicile exception, Rome II joins the majority of 

recent PIL codifications and international conventions in accepting the premise that 

when 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 in 

another state. This notion is implemented either through a common-domicile rule (as 

in the codifications of Louisiana, Puerto Rico, Switzerland, Quebec, Belgium, and the 

107 

Hague Convention on Products Liability), or through an exception from the lex loci 

rule. The exception is phrased either in common-domicile or common-habitual 

108 

residence language (as in the Dutch, German, and Hungarian codifications), or in 

common-nationality language (as in the Italian, Polish, and Portuguese 

109 

codifications). 

Article 4(2) also mirrors parallel developments in the United States. As 

110 

documented elsewhere, 32 of the 42 cases decided since the 1960s in which an 

104. Art. 13 provides that the habitual residence of companies and other corporate or unincorporated 

bodies is the place of their central administration and, for injuries caused or sustained in the 

course of operation of a branch, the place where the branch is located. Art. 13 also provides that 

the habitual residence of a natural person acting in the course of his or her business activity is 

his or her principal place of business. 


106. This escape is also repeated in the art. 10-12, dealing with unjust enrichment, negotiorum 

gestio, and culpa in contrahendo, which are not discussed here. However, the escape becomes 

operable only if the primarily applicable law “cannot be determined.” 

107. See LA. CIV. CODE ANN. art. 3544(1); PUERTO RICO DRAFT CODE art. 47(a); SWISS PIL ACT, 

art. 133; QUEBEC CIV. CODE, art. 3126; BELGIAN PIL CODE, art. 99 (1); HAGUE PRODUCTS 

LIABILITY CONVENTION, art. 5. 

108. See DUTCH PIL ACT, art. 3(3); EGBGB art. 40(2); HUNGARIAN PIL DECREE § 32(3). 

109. See ITALIAN PIL ACT, art. 62; POLISH PIL ACT, art. 31(2); PORTUGUESE CIV. CODE, art. 45. For 

an exception that displaces the lex loci when the parties have either a common nationality or a 

common habitual residence, see CHINA’S MODEL DRAFT LAW OF PIL, art. 114 (6th Draft 2002). 

110. For citations and discussion, see SYMEONIDES, THE CHOICE-OF-LAW REVOLUTION 145-59; 

EUGENE SCOLES, PETER HAY, PATRICK BORCHERS & SYMEON C. SYMEONIDES, CONFLICT OF 




American court of last resort abandoned the lex loci rule involved the commondomicile 

pattern. Subsequently, an additional 18 common-domicile cases have 

reached 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 state 

supreme courts in the post-lex loci era. Of these 50 cases, 44 cases (or 88%) have 

applied the law of the common domicile, regardless of the particular choice-of-law 

111 

methodology the court followed. The majority of these cases (35 out of 50) involved 

the Babcock v. Jackson pattern in which the law of the common domicile favors 

recovery more than the law of the state of conduct and injury. These cases present the 

classic false conflict paradigm in which only the state of the common domicile has an 

interest in applying its law. The remaining cases involved the converse-Babcock 

pattern in which the law of the common-domicile prohibits or limits recovery more 

than the law of the state of conduct and injury. These cases are not as clear false 

conflicts as Babcock was because the accident state arguably has an interest in 

applying its law to compensate those injured in its territory and to facilitate recovery 

112 

of local medical costs. On balance, however, the application of the law of the 

common domicile in both the Babcock pattern and its converse is entirely justified. 

In this sense, a common-domicile rule that does not depend on the content of the law 

of the common domicile (i.e., a jurisdiction-selecting rule) is generally preferable to 

any other iteration. 

It is important to stress, however, that all of the above American cases 

involved conflicts between “loss-allocation” or “loss-distribution” rules, as opposed 

to “conduct-regulation” rules. In contrast, the common-domicile rule of Rome II is 

much 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 has 

an interest in enforcing its conduct-regulating rules even if neither the violator nor the 

victim is domiciled in that state, and even if both parties are domiciled in the same 

foreign state. For example, an Austrian motorist involved in a French accident may 

not claim exemption from French traffic rules, and if injured by conduct that violates 

these rules, France may not deny her the benefit of their protection. Even if both 

parties are domiciled in Austria, France has the exclusive claim to apply its law to the 

conduct-regulating aspects of the case. These aspects are not limited to rules of public 

law or pure traffic rules such as speed limits and red lights; they also extend to rules 

that impose civil liability to violations of traffic rules or attach presumptions and 

inferences of fault that arise from certain violations, such as not maintaining sufficient 

distance from the preceding car. Because the common-domicile rule of Rome II is 


LAWS, 799-806 (4th ed. 2004). 

111. The six cases that applied another law are distinguishable, overruled, or discredited. See 

SYMEONIDES, THE CHOICE-OF-LAW REVOLUTION, 149 n.17, 152-54. 

112. See id. at 154. 



phrased in global terms, it would mandate the application of Austrian law to all 

aspects of the case. This problem is made worse when, as in the Italian, Quebec, and 

113 

Polish codifications, the rule is not subject to any escapes. Rome II’s rule is subject 

to two escapes: the “closer connection” clause of Article 4(3), and the possibility that 

the court may “take account” of the “safety and conduct” rules of the conduct state 

under Article 17. Unfortunately, as explained below, these escapes are not easily 

deployable. 

From a different perspective, the common-domicile rule of Rome II is too 

narrow in that it applies only when the parties are domiciled in the same state but not 

when they are domiciled in different countries that have the same laws. The better 

view is that the latter cases are functionally analogous to common-residence cases and 

114 

should be treated accordingly. Suppose, for example, that while hunting in Kenya, 

a French hunter injures a Belgian hunter with whom he has no pre-existing 

relationship. Suppose that French and Belgian law provide the same amount of 

compensation, which is much higher than that provided by Kenya. This is the classic 

false conflict in which Kenya has no interest in applying its low recovery law. In such 

a case, there is no reason to apply Kenyan law and every good reason to apply either 

Belgian or French law. Yet, Article 4(1) of Rome II mandates the application of 

Kenyan law, and, unfortunately, none of Rome II’s exceptions to the lex loci rule 

would be operable in this case. 

C. The General Escape 

Paragraph 3 of Article 4 provides an escape from both the lex loci rule of 

paragraph 1 and the common-residence rule of paragraph 2. Echoing similar escapes 

113. See ITALIAN PIL ACT, art. 62(2); QUEBEC CIV. CODE, art. 3126(2); POLISH PIL ACT, art. 31(2). 

114. The Louisiana codification provides that “[p]ersons domiciled in states whose law on the 

particular issue is substantially identical shall be treated as if domiciled in the same state.” LA. 

CIV. CODE ANN. Art. 3544(1). This legal fiction, which is particularly useful in cases with 

multiple victims or defendants, enables a court to resolve these false conflicts by applying the 

law of the domicile of either party, unless the general escape clause of the codification dictates 

a different result. For pertinent discussion, see Symeon C. Symeonides, Louisiana’s New Law 

of Choice of Law for Tort Conflicts: An Exegesis, 66 TUL. L. REV. 677, 759-63 (1992). The 

American Law Institute has recommended a similar rule for mass tort cases. See AMERICAN 

LAW INSTITUTE, COMPLEX LITIGATION: STATUTORY RECOMMENDATIONS AND ANALYSIS § 

6.01(c)(2) & (3) (1994). In addition, in certain cases involving corporate tortfeasors, the 

Louisiana common-domicile rule is subject to further expansion, or contraction, through art. 

3548, which provides that a juridical person that is domiciled outside the forum state but 

transacts business in that state and incurs a delictual obligation arising from such activity may 

be treated as a domiciliary of that state, if such treatment is appropriate under the principles of 

art. 3542. 



115 

found in recent European codifications and international conventions, the escape 

authorizes the court to apply the law of another country if “it is clear from all the 

circumstances of the case that the tort/delict is manifestly more closely connected with 

116 

[that other] country.” Paragraph 3 provides an example by stating that a manifestly 

closer connection “might” be based on “a pre-existing relationship between the 

parties, 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 cases 

covered by the general rule, the final text repeats the escape in the articles dealing with 

products liability (Art. 5(2)), unfair competition cases in which the competition affects 

“exclusively” the interests of a specific competitor (Art. 6(2)), and choice-of-law 

agreements (Art 14(2)). 118 

Despite serious reservations about the scope and wording of this particular 

escape, 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 statutory 

scheme. Because perfection is not for this world and more and more modern 

legislatures have begun to recognize their fallibility, escapes have become a common 

119 

feature of almost all recent codifications. As Aristotle recognized many centuries 

ago, any pre-formulated rule, no matter how carefully or wisely drafted, may, “due to 

120 

its generality,” or because of its specificity, produce results that are contrary to the 

purpose for which it was designed. This “is a natural consequence of the difference 

121 

between law making and law application.” The question here is to what extent this 

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

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


117. Id. 

1. The Closer Connection Exception 

The final phrasing of the escape clause is a significant improvement over that 

of the Commission’s preliminary draft proposal, which was based on the failed EEC 

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 CLAUSES 

D’EXCEPTION EN MATIÈRE DE CONFLITS DE LOIS ET DE CONFLITS DE JURIDICTIONS – OU LE 

PRINCIPE DE PROXIMITÉ (1994). 

120. Aristotle, NICOMACHEAN ETHICS, V. x 7. 

121. Peter Hay, Flexibility Versus Predictability and Uniformity in Choice of Law, 226 RECUEIL DES 

COURS 281, 291 (1991-I). 



122 123 

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

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 the 

draft’s rules; and (b) that there was a substantially “closer” connection with another 

country. The problem with this scheme was that, if taken literally, the first prong 

would 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 a 

comparison of the two connections, but rather an independent determination that the 

first connection was not significant. Only when that determination confirmed the 

insignificance of the first connection did the escape allow examination of the 

closeness of the other connection. The final text resolved much of the problem by 

eliminating the first prong and by encouraging a comparison between the two 

connections. Under the final text, a party that invokes the escape need not show that 

the connection of the country whose law governs under the rule is “insignificant.” All 

one needs to show is that the connection with another country is “manifestly closer” 

and this of course can only be determined after comparing the two connections. This 

is a significant improvement for which the drafters deserve praise. 

Even with these changes, however, the escape remains problematic because, 

124 

like its European counterparts: (a) it is phrased in exclusively geographical or 

quantitative terms that are not correlated to an overarching principle; and (b) it does 

not 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 system 

because most of its dispositive rules depend on the place in which a single critical 

event occurred, or in which one or both parties reside. Very few non-geographical 

factors affect the choice, and the content of the conflicting laws is a factor that appears 

125 

only in some narrow exceptions. Having relied on geography in erecting this 

system, the drafters may have felt bound to also rely on geography to handle the 

exceptional cases and overcome the inevitable impasses. Thus, if the geographically 

chosen place of injury or of the parties’ common residence turn out to have a nonsignificant 

connection, one must search for a place that has a “closer” connection. This 

logic, however, will rarely overcome the impasses. Escapes are designed to cure the 

rule’s 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 

122. See supra II.1. 

123. See, e.g., BELGIAN PIL ACT, art. 19(1); SWISS PIL ACT, art.15(1); QUEBEC CIV. CODE art. 3082. 

For a softer phrasing, see ENGLISH PIL ACT.§ 12. 

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

125. See supra III.4. 



and the values and goals the rule seeks to promote. To simply say that one should look 

for a “closer” connection gives courts little meaningful guidance and entails the risk 

of 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 of 

the Restatement (Second) and the Louisiana codification. The Restatement provides 

in Section 6 that the goal of the choice-of-law process is to identify the state that has 

the “most significant relationship.” Although literally the quoted phrase appears to 

contemplate a determination based on geography, the content of Section 6 negates any 

such inference because it lists a series of substantive policies intended to guide this 

determination. 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 significant 

126 

relationship “under the principles stated in § 6.” Similarly, Article 3542 of the 

Louisiana codification enunciates the general goal of the choice-of-law process for tort 

conflicts as one of identifying the state whose policies would be most seriously 

impaired if its law were not applied. After establishing specific rules based on that 

goal, the codification also provides an escape clause in Article 3547 which authorizes 

the 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 not 

applied to the particular issue.” 127 

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

of Rome II—issue-by-issue evaluation and correlation to non-geographical 

overarching principles. The only hope comes from a statement in recital 14, which 

appears to articulate the philosophy of Rome II as a whole: 

The requirement of legal certainty and the need to do justice in 

individual cases are essential elements of an area of justice. This 

Regulation provides for the connecting factors which are the most 

appropriate to achieve these objectives. Therefore, this Regulation 

provides for a general rule but also for specific rules and, in certain 

provisions, for an ‘escape clause’ which allows a departure from these 

rules where it is clear from all the circumstances of the case that the 

tort/delict is manifestly more closely connected with another country. 

This set of rules thus creates a flexible framework of conflict-of-law 

rules. Equally, it enables the court seised to treat individual cases in 

126. AMERICAN LAW INSTITUTE, RESTATEMENT (SECOND) OF CONFLICT OF LAWS, § 146 (emphasis 

added). 

127. LA. CIV. CODE ANN. Art. 3547 (emphasis added). 



an appropriate manner. 128 

This recital is what survives from the rapporteur’s and Parliament’s efforts to 

inject more flexibility into the text of the general rule of Rome II. As noted earlier, the 

Council and Commission rebuffed these efforts and recital 14 is the resulting 

compromise. While much of this recital is self-congratulatory and merely describes 

the scheme of Rome II, the italicized phrases can be viewed as providing instruction 

to courts on when and how to use the escape: to “treat individual cases in an 

appropriate manner” and to “do justice in individual cases.” Thus, a court should 

resort to the escape when the law designated as applicable by the general rule leads 

to a result that is incompatible with “the need to do justice in individual cases.” 

To be sure, one can object—and many will—to such a loose reading of the 

escape as giving a licence for loose, ad hoc, subjective judging. While this danger is 

real, 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 quantitative 

employment of the escape will not resolve the aforementioned Kenyan hunting 

accident case involving a French and a Belgian hunter, whereas an employment of the 

escape 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-byissue 

deployment and evaluation. As it is, the escape contemplates situations in which 

the 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 law 

129 

applicable to “a non-contractual obligation arising out of a tort/delict.” Thus, the 

escape does not even permit the court to look separately to the possibly multiple 

obligations that may arise from the same facts, such as when (but not only) the case 

130 

involves multiple tortfeasors or victims. Instead, the phrasing of the escape forces 

the court to only look at the tort as a whole. If the court finds that the tort as a whole 

has a closer connection with another country, then the court is authorized to displace 

the otherwise applicable law in its entirety and replace it with the law of that other 

country. Thus, the escape is an “all or nothing” proposition; and therein lies its most 

serious flaw. 

This flaw will make the escape unavailable in all but the obvious cases. For 

example, 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 negligence per se or 

damages limitations) is “manifestly” more closely connected with Switzerland. Even 

if the victims’ domicile, England, also allowed unlimited damages, the court could not 

128. ROME II, recital 14 (emphasis added). 

129. ROME II, art. 4(1) (emphasis added). 

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



avoid the application of French law because the connection with England (which 

Article 4(1) a priori condemns to a state of insignificance) will probably not qualify 

as “manifestly” closer than the connection with France. The same would be true with 

other issues affecting the victim’s recovery. Suppose, for example, that the avalanche 

caused the death of one of the English tourists, and one of the issues in the case is who 

is entitled to compensation for his wrongful death. Suppose that French law provides 

that compensation is due to the victim’s surviving spouse and children together, 

whereas English law provides that compensation is due to the surviving spouse to the 

exclusion of the children. Under Article 4(1), the applicable law shall be the law of 

the country in which “the damage occurs” (France) “irrespective of the country or 

countries in which the indirect consequences of that event occur” (England). Article 

15 reiterates that in such a case, French law will apply to virtually all issues likely to 

arise in tort litigation, including “[which] persons [are] entitled to compensation” and 

“whether a right to claim damages or a remedy may be transferred, including by 

131 

inheritance.” Yet England is the country most intimately involved and has the best 

claim to apply its law to this issue. Rules designating the beneficiaries of a wrongful 

death action reflect a society’s assumptions regarding how a person’s death impacts 

his survivors and which of his survivors are likely to have the highest need for 

compensation. These assumptions and value judgments belong to the society in which 

the victim lived, not to the society in whose territory the injury occurred. In Rome II’s 

terminology, England is “manifestly more closely connected” with regard to the issue 

of wrongful death beneficiaries, even if its connections with regard to other issues may 

not be the closest. Yet, the phrasing of the escape clause does not permit this focus on 

the specific issue, and therefore does not allow a court to cure the rule’s deficiency. 

Similar problems are encountered in employing the “manifestly closer 

connection” 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 argue 

that, with regard to issues of conduct and safety, France has a “manifestly closer 

connection” than Austria, and thus French law should govern. The problem with this 

otherwise sound argument is that it runs against the restrictive and holistic wording 

of the escape, which does not allow an issue-by-issue analysis and instead speaks of 

the whole “tort/delict,” as opposed to certain aspects of it, as being more closely 

connected with another country. This wording makes it difficult to argue that the 

entire tort is more closely connected with France, while also being governed by the 

law of Austria. The same problem exists in the case of the Kenyan hunting accident 

involving the French and Belgian hunters. Because the escape is worded in 

geographical terms, and the hunters are not domiciled in the same country, geography 

would work in favor of, not against, the lex loci. One provision that can help in the 

French, but not the Kenyan, accident case is Article 17, which allows a court to “take 

131. ROME II, art. 15 (f) & (e). 



account” of the “safety and conduct” rules of the state of conduct. However, as 

discussed later, this provision offers only limited relief. 

A more nuanced escape clause would soften the common-domicile rule and 

produce more rational results, not only in the above cases, but also in other 

132 

exceptional cases, or in cases in which that rule adversely affects the rights of third 

parties. The latter problem is illustrated by two hypotheticals. The first involves a 

single-car accident in Portugal caused in part by faulty road conditions and resulting 

in injury to the car’s Spanish passenger. The passenger sues the local Portuguese 

authority responsible for maintaining the road, and the authority then impleads the 

car’s Spanish driver for contribution and indemnification for his fault in the accident. 

The driver and passenger are married to each other and Spanish, but not Portuguese, 

law prohibits interspousal suits. In such a case, one could argue that the local 

authority’s claim for contribution falls outside the scope of the common-domicile 

133 

rule, but if the argument fails, the only way to avoid Spanish law would be by 

stretching the wording of the “closer connection” escape. In the second hypothetical 

the local authority is not involved, the car was rented in Portugal, and the laws are 

reversed so that Portuguese, but not Spanish, law prohibits interspousal lawsuits. The 

injured passenger sues the driver and the car’s insurer invoking Spanish law, while the 

insurer denies coverage invoking Portuguese law. Again, if the insurer’s obligation for 

134 

coverage falls within the scope of the common-domicile rule, then the only way to 

avoid Spanish law would be through the difficult road of employing the “closer 

connection” escape. 

In conclusion, while both the adoption of the common-domicile rule and the 

inclusion of an escape clause in Article 4 are significant steps in the right direction, 

both the rule and the escape would have benefitted from more nuanced drafting. In the 

132. A case like Schultz v. Boy Scouts of America, 480 N.E.2d 679 (N.Y. 1985), is arguably such an 

exceptional case. Schultz involved a suit by New Jersey parents against a New Jersey charitable 

corporation for injury resulting from a sexual molestation of the plaintiffs’ child by one of 

defendant’s employees in New York. Invoking the common-domicile rule, the New York Court 

of Appeals applied New Jersey charitable immunity law, denying recovery. 

133. The argument would be based in part on art. 20 of Rome II, which provides that, if a creditor 

has a claim against several debtors who are “liable for the same claim,” the question of that 

debtor’s right to demand compensation from the other debtors “shall be governed by the law 

applicable to that debtor’s non-contractual obligation towards the creditor.” 

134. The pertinent Rome II articles for answering this question are art. 18 & 19. Art. 18 is not 

particularly helpful, except perhaps indirectly, because it only applies to the victim’s right to 

“directly” sue the tortfeasor’s insurer (i.e., without joining the tortfeasor) and not to the merits 

of the suit. The article allows the victim to chose between the law governing the insurance 

contract and the law governing the tort. Art. 19 provides that, when a creditor has a noncontractual 

claim against a debtor, and a third person has a duty to satisfy the creditor, “the law 

which governs the third person’s duty to satisfy the creditor shall determine whether . . . the 

third person is entitled to exercise against the debtor the rights which the creditor had against 

the debtor under the law governing their relationship.” 



final analysis, a good escape clause can bring substantive improvements to even a bad 

rule system, in addition to helping attain the proper equilibrium between the two 

perpetually competing goals of certainty and flexibility. However, to perform this role 

the escape must possess some degree of built-in flexibility. The drafters’ preference 

for a tight escape that does not swallow the rules is understandable. However, an 

escape that is so tight as to be rarely utilized, or one that is phrased in broad all-ornothing 

terms, is only slightly better than no escape at all. 

2. The Pre-existing Relationship Exception 

As noted earlier, the second sentence of Paragraph 3 of Article 4, attempts to 

provide an example of a manifestly closer connection by stating that such a connection 

might be based on a “pre-existing relationship between the parties, such as a contract, 

135 

that is closely connected with the tort/delict in question.” As with the general 

escape, this exception is also conceived of in all-or-nothing terms rather than in terms 

of specific issues. Once again, the drafters’ excessive concern with avoiding the 

possibility of dépeçage renders this exception far less useful than it might have been. 

In fact, rather than allowing a splitting of the various tort issues, this exception 

contemplates grouping them together with the issues arising from the pre-existing 

relationship. 

However, this is only one of the problems with this provision. The major 

problem is determining which of two equally plausible objectives the drafters 

intended, namely: (a) to apply the same law as that which governs the pre-existing 

relationship, or (b) to apply the law of the same state in which the pre-existing 

relationship is primarily centered. The Explanatory Report suggests that the drafters 

136 

intended the first objective. However, unlike some European codifications which 

137 

expressly provide to that effect, Article 4(3) of Rome II does not do so. This makes 

viable the other possibility, namely, applying the law of the state in which the preexisting 

relationship is centered. 138 


136. See Explanatory Report, p. 13 ("By having the same law apply to all their relationships, this 

solution respects the parties' legitimate expectations and meets the need for sound administration 

of justice. On a more technical level, it means that the consequences of the fact that one and 

the same relationship may be covered by the law of contract in one Member State and the law 

of tort/delict in another can be mitigated.”). 

137. See, e.g., SWISS PIL ACT, art. 133(3) (“. . . when the tortious act constitutes a violation of a 

pre-existing legal relationship between the tortfeasor and the injured party, claims founded on 

this act are governed by the law applicable to that legal relationship.” BELGIAN PIL CODE, art. 

100 (“an obligation resulting from an injurious event having a close connection with a preexisting 

legal relationship between the parties is governed by the law applicable to that 

relationship.”). 

138. This possibility is consistent not only with the previous sentence of paragraph 3 of Article 4, but 

also with the whole geographical orientation of Rome II. 



To be sure, in some cases, the two objectives will lead to the same law. For 

example, if the pre-existing relationship is a family relationship centered in state X, 

then the law of that state will govern that relationship and, under the above quoted 

139 

provision, the court may apply the same law to a related delictual obligation. If, 

140 

however, the relationship is contractual, then there is no guarantee that the state in 

which the relationship is centered will also be the state whose law will govern the 

contract. For example, the contract may contain a choice-of-law clause stipulating for 

the law of state Z, even if that state has a relatively tenuous but otherwise sufficient 

141 

connection with the relationship. In such a case, the question is which, if any, of the 

two states, X or Z, will be the candidate for the closer connection exception? Z cannot 

be because, in this scenario, it does not have a close enough factual connection. On 

the other hand, X has the factual connection, but the application of its law will defeat 

the apparent purpose of this exception, which is to apply the same law to both the tort 

and contract aspects of the case. 142 

Finally, the main advantages of applying the same law to both the tort and 

contract aspects of a dispute are practicality and simplicity. Obviously, these 

advantages are not present when the particular dispute involves only tort issues. In 

such a case, the rationale for this exception must be sought elsewhere, such as in the 

ostensible expectations of the parties. 

4. Compensation for Traffic Accident Victims 

139. The same result can be obtained more directly through the common-residence rule of Paragraph 

2. This illustrates that the above quoted sentence of Paragraph 3 is superfluous in most cases 

in which the parties to the relationship are residents of the same state. 

140. If the relationship is merely social rather than legal, as in Babcock v. Jackson in which the 

parties where neighbors who drove together from New York to Ontario, it makes little sense to 

say that the tort will be governed by the same law that governs the relationship because the 

social relationship may not, as such, be governed by any law. However, it does make sense to 

say that the tort will be governed by the law of the state in which the relationship was centered. 

141. Under the Rome Convention, the choice of state Z law will be upheld unless “all the other 

elements relevant to the situation at the time of the choice are connected with [another] 

country.” Even then, the choice will be disregarded only to the extent it “prejudices the 

application of rules of the law of that [other] country which cannot be derogated from by 

contract.” ROME CONVENTION, art. 3(3) (emphasis added). 

142. Another variation of this problem is when the contract contains a choice-of-law clause that is 

partially ineffective under the Rome Convention. For example, under art. 6, a choice-of-law 

clause in an employment contract will be disregarded to the extent it would deprive the 

employee of the protection afforded by the mandatory rules of the state in which the employee 

“habitually carries out his work.” However, the rest of the contract will be governed by the 

chosen law. Thus, if a German company hires in Germany a German employee for work in 

Belgium and the clause stipulates for German law, the clause will be ineffective to the extent 

it violates Belgian mandatory rules but effective with regard to the rest of the contract. One of 

the questions in such a case is which of the two countries (and for what issues) would qualify 

as the “contract state” for purposes of applying the exception of art. 4(3) of Rome II. 



During Parliament’s first reading of Rome II, the rapporteur proposed, and 

Parliament approved, the insertion of the following exception to the general lex loci 

rule of what later became Article 4: 

In the case of personal injuries arising out of traffic accidents, . . . the 

court seised . . . should, for the purposes of determining the type of 

claim for damages and calculating the quantum of the claim, apply the 

rules of the individual victim’s place of habitual residence unless it 

would be inequitable to the victim to do so. 143 

The Council and Commission rejected this amendment, and the resulting 

compromise was the insertion of the following statement in recital 33 of the final text 

of the Preamble: 

According to the current national rules on compensation awarded to 

victims of road traffic accidents, when quantifying damages for 

personal injury in cases in which the accident takes place in a State 

other than that of the habitual residence of the victim, the court seised 

should take into account all the relevant actual circumstances of the 

specific victim, including in particular the actual losses and costs of 

after-care and medical attention. 144 

Obviously, this recital narrows down considerably the scope and import of 

Parliament’s amendment. Besides the fact that a recital does not have the same status 

and cogency as a rule in the main body of the Regulation, the recital does not 

authorize the application of the law of the victim’s habitual residence. It merely 

authorizes “taking into account” that law in “quantifying” damages, rather than in also 

“determining the type of claim for damages,” as Parliament intended. Thus, the recital 

appears to be no more than an invitation (à la Article 17) for the court to take account 

145 

of facts, such as the cost of medical care in the victim’s residence, in fixing the final 

amount of recoverable compensation. Nevertheless, unless one assumes that the 

Parliament got nothing in return for this “compromise,” the recital must mean 

143. EUR. PARL. FINAL A6-0211/2005 (June 27, 2005). The accompanying justification stated that 

in traffic accident cases, “applying the law of the state of the victim's place of habitual residence 

is more equitable (for the victim, e.g., where he or she needs lifelong care) and more practicable 

for insurers and the courts,” and that “not only the level of compensation is important, i.e., the 

question of how much; rather, what FORM of compensation to be received by an injured party 

is also important (in particular whether there is an entitlement to damages for pain and suffering, 

a nursing and attendance allowance or certain pensions). Accordingly, the type of compensation 

should be governed by the law applicable at the injured party's place of habitual residence.” Id. 

20/46. 

144. ROME II, recital 33. In addition, the Commission issued a Statement, reproduced as an appendix 

to the final text of Rome II, promising to study the matter and report to Parliament before the 

end of 2008. 

145. But see supra text accompanying notes 51-52. 



something. What it may mean is that—consistently with the admonition to “do justice 

in individual cases” in recital 14—a court should keep in mind (for example, in 

applying the closer connection escape) the need to adequately compensate victims of 

traffic accidents. Unfortunately, even if this argument is accepted, it cannot help the 

Belgian hunting accident victim or the injured tourist in the snow avalanche 

hypothetical because they were not victims of a traffic accident, although they were 

equally deserving of protection. 

146. Art. 5 makes no distinction between situations in which the product was acquired by the victim 

and situations in which the product was acquired by a third party, such as a previous purchaser 

or a transportation carrier. For the significance of this distinction, see SYMEONIDES, THE 

CHOICE-OF-LAW REVOLUTION 268-70, 351-52. 

147. Art. 5, and Rome II in general, offer little guidance for cases involving peripatetic injuries, such 

as cases involving pharmaceutical products used by their eventual victims over long periods of 

time while residing in different countries. For the American experience on this matter, see 

SYMEONIDES, THE CHOICE-OF-LAW REVOLUTION 268-69. 


149. Id. 

VI. PRODUCTS LIABILITY 

Article 5 provides a special rule for non-contractual obligations arising from 

an injury caused by a product. Paragraph 1 of the article designates, in successive 

order, three countries whose law may govern: (a) the country of the victim’s habitual 

146 

residence; (b) the country in which the product was acquired; and (c) the country in 

147 

which the injury occurred. The application of each country’s law depends on 

148 

whether the product was “marketed in that country.” For example, if a German 

plaintiff is injured in India by a product acquired in Egypt, the applicable law will be 

that of Germany, if the product was marketed there; or, if not, Egypt, if the product 

was marketed there; or, if not, India, if the product was marketed there. 

It appears that the burden of proving that the product was marketed in the 

particular country would rest with the plaintiff, although the defendant may also have 

an incentive, and should be allowed, to either disprove or prove that fact. Moreover, 

the last sentence of paragraph 1 expressly gives defendants a defense—they can avoid 

the application of the law of each of the above three countries by demonstrating that 

they “could not reasonably foresee the marketing of the product, or a product of the 

149 

same type” in that country. If taken literally, this could mean that, even if the 

plaintiff proves (and the defendant does not disprove) that the product was actually 

marketed in the particular country, the defendant can still get a second line of defense 

by showing that, despite the actual marketing, “he or she could not reasonably foresee 

the marketing.” This provision is unduly generous to the defendant. Fortunately, as 

noted below, contemporary marketing patterns suggest that in most cases this defense 



150 

is unlikely to succeed. In any event, if the defense does succeed, the applicable law 

will not be that of the country next in line under paragraph 1 (e.g., Egypt after 

Germany, or India after Egypt), but rather the law of the defendant’s habitual 

151 residence. Thus, if the product was manufactured by a Japanese defendant, Japanese 

law will govern the case, unless of course Japanese law is more favorable to the 

plaintiff than say Egyptian or Indian law, in which case the defendant will not invoke 

this defense to begin with. 

Paragraph 1 of Article 5 applies “[w]ithout prejudice to Article 4(2),” which 

contains the common-residence rule. This means that, if the parties have their habitual 

residence in the same country, its law applies to the exclusion of all others, even if the 

product was not marketed in that country. Thus, if in the above scenario the product 

was manufactured by a German defendant, German law would govern, even if the 

product was not marketed in Germany. 

Finally, all of paragraph 1 (including the cross-reference to the commonresidence 

rule) is subject to the “manifestly closer connection” escape contained in 

152 

paragraph 2 of Article 5. This escape authorizes a court to either: (a) deviate from 

the order established in paragraph 1 and apply the law of one of the countries listed 

there; or (b) apply the law of a country not listed in paragraph 1, such as the country 

153 

of the product’s manufacture, upon showing that the country has a manifestly closer 

connection than the country whose law would normally govern under paragraph 1. 

Although Article 5 appears complex in its wording, its actual operation in 

practice may be much simpler, depending on how easy it will be to satisfy the basic 

condition of a product’s marketing in a particular country. One could surmise that, in 

today’s global market, this condition will be more and more easily satisfied in the 

great majority of cases without much further inquiry or counter-proof. At least this is 

what the American experience suggests. A recent study of products liability cases 

decided in the United States between 1990 and 2005 shows that, in none of these cases 


151. If the defendant is a juridical person, the place of its central administration is deemed to be its 

habitual residence. See ROME II, art. 23(1). Even so, the defendant’s “residence” (at least when 

the defendant is the manufacturer rather than the local importer or distributor ) would seem to 

be the least relevant contact in today’s world of corporate mobility. In most cases, the 

manufacturer is likely to be a corporate entity whose “residence” or cental administration may 

be located in a country that has little relationship with the case, the product, or its manufacture. 

152. ROME II, art. 5(2). The escape also repeats the “pre-existing relationship” exception with all the 

problems described earlier. In addition, in all cases in which the victim was also the acquirer 

of the product, either side can claim a “pre-existing relationship” between the victim and the 

defendant manufacturer, distributor, or retail seller. 

153. It is somewhat surprising that art. 5 does not consider the place of manufacture as a pertinent 

connecting factor. For the role of this factor in American product liability conflicts, see 

SYMEONIDES, THE CHOICE-OF-LAW REVOLUTION 270, 327, 351. 



154 

did the foreign defendant invoke a similar defense. Although most of these cases 

involved American manufacturers, several cases involved foreign manufacturers. 

Thus, unless marketing patterns are much different in Europe, the marketing defense 

will probably be unsuccessful in all but the rarest instances. In turn, this suggests that 

Article 5 will lead to the application of the law of: (a) the parties’ common habitual 

residence, in the few cases when such common residence exists; and (b) the victim’s 

habitual residence in most of the remaining instances. Few cases would trickle down 

to the law of the country of acquisition, and even fewer, if any, to the law of the 

country of injury. 

If these assumptions are correct, the next question is whether these results are 

acceptable. In answering that question, it helps to remember that actual cases are often 

far less complex than classroom hypotheticals. For example, although in the abstract 

there may be good reasons to criticize the application of the law of the country of the 

victim’s residence as such, it is helpful to know that, in the majority of cases that 

country is likely to have at least one or more additional pertinent contacts. This was 

so in seventy-two percent of the disputes in the aforementioned American study. 155 

Nevertheless, although the presence of these additional contacts make the application 

of the law of the victim’s domicile more defensible in practice, Article 5 itself must 

also be defensible in those cases in which these other contacts are lacking. Moreover, 

the fact that Article 5 does not differentiate between cases in which the law of the 

victim’s domicile favors and those in which it disfavors the victim raises additional 

questions. One such question is whether Article 5 favors residents of developed 

156 

countries and disfavors residents of lesser developed countries. In the above 

hypothetical, the German plaintiff who was injured in India by a Japanese product 

acquired in Egypt will get the benefit of German law. However, an Indian plaintiff 

who is injured in Austria by a German product acquired in Germany will be confined 

to the remedies provided by Indian law. Was this deliberate? If yes, it is one more 

example of a territorialist choice-of-law rule which, though seemingly value-neutral, 

disguises specific policy choices. 

If the only objective of products liability law is to ensure the “right” amount 

of compensation for victims, then the application of German law to the German 

plaintiff and Indian law to the Indian plaintiff may be defensible. However, to the 

154. See id. at 337. 

155. See id. at 320-22. American courts applied the law of the victim’s domicile in sixty-five percent 

of all cases, but in all but twelve percent of those cases that state had one or two additional 

contacts and in more than half of them, that state had a pro-defendant law. See id. at 325. 

156. For purposes of illustration, it is worth noting that American courts, which are often accused of 

favoring plaintiffs, applied a plaintiff-favoring law in only fifty-two percent of the cases. See 

id. at 332-33. 



157 

extent that product liability law is also designed to serve other objectives, such as 

deterring the manufacture and proliferation of unsafe products, there are good reasons 

to disagree with the application of Indian law in the Indian plaintiff’s case. One can 

only hope that a court would avoid such a result by invoking the closer connection 

escape of Article 5. 

In fairness to the Rome II drafters, it should be noted that product liability 

conflicts are inherently complex and thus far no one has the perfect formula for 

158 

resolving them. This includes the present author who has drafted two statutory rules 

159 160 

for such conflicts and has proposed two other rules for the same purpose. The fact 

that each of those rules differs from the others is this author’s admission that the 

search for the perfect formula must continue. Article 5 of Rome II is far from the 

perfect formula, but the real question is whether it is good enough. 

157. ROME II, recital (2), states that “[t]he conflict-of-law rule in matters of product liability should 

meet the objectives of fairly spreading the risks inherent in a modern high-technology society, 

protecting consumers’ health, stimulating innovation, securing undistorted competition and 

facilitating trade.” 

158. For a comparative discussion of enacted and proposed rules for product liability conflicts, see 

E. Scoles et al., CONFLICT OF LAWS, 934-41 (4th ed. 2004). 

159. See LA. CIV. CODE ANN. Art. 3545, discussed in Symeonides, Exegesis, supra note 114, at 749- 

59; PUERTO RICO DRAFT CODE, art. 48. 

160. See Symeon C. Symeonides, The Need for a Third Conflicts Restatement (And a Proposal for 

Tort Conflicts), 75 INDIANA L. J. 437, 450-51, 472-74 (2000); Symeonides, THE CHOICE-OF- 

LAW REVOLUTION, 346-64. 

161. Explanatory Report, art. 7, at 19. 

VII. ENVIRONMENTAL TORTS 

Article 7 of Rome II deals with environmental torts. Through a cross reference 

to the general rule of Article 4(1), Article 7 provides that the applicable law is the law 

of the place of injury, unless the victim of the tort opts for the law of the place of 

conduct. As noted earlier, Article 7 is not subject to either the common-domicile or 

the closer connection exceptions. Because this victim’s choice is meaningful only in 

cases of cross-border torts in which the country of conduct and the country of injury 

prescribe different standards of conduct for the tortfeasor or financial protection for 

the victim, the discussion below is limited to these cases. 

It is worth noting that, although at some point the Explanatory Report speaks 

of “a victim in a low-protection country . . . enjoy[ing] the higher level of protection 

161 

available in neighbouring countries,” the reason for which Article 7 gives the victim 

a choice is not to benefit the victim as such. Rather, the reason is to promote the 

interests of the respective countries and of the Union as a whole in deterring pollution. 

Applying whichever of the two laws subjects the polluter to a higher standard 



promotes this interest. Giving the victim a choice is simply the vehicle for ensuring 

this result. 

When the law of the state of conduct has a higher standard of conduct for the 

tortfeasor or of financial protection for the victim than the state of injury, the victim 

will opt for the law of the conduct state, thus producing the deterring effect the 

drafters intended. As the Report notes, if the victim did not have this choice, the 

operator would have “an incentive to establish his facilities at the border so as to 

discharge toxic substances into a river and enjoy the benefit of the neighbouring 

country’s laxer rules. This solution would be contrary to the underlying philosophy 

of the European substantive law of the environment and the ‘polluter pays’ 

162 

principle.” Indeed, the application of the higher standards of the conduct state in this 

scenario is fully justified, whether one thinks in terms of party reliance or expectations 

or, more aptly, state interests. After all, the operator should not complain for being 

subjected to the standards of the state in which he acted. Having violated those 

standards, he should bear the consequences of the violation and not be allowed to 

invoke the lower standards of another state. In terms of state interests, the application 

of the higher standards of the conduct state promotes that state’s policy in policing 

conduct within its borders, without subordinating the policies of the state of injury 

because the latter state does not have an interest in applying its lower-standards to 

protect conduct occurring, or tortfeasors acting, beyond its borders. In other words, 

this is a case that is described in the American conflicts lexicon as a classic “false 

conflict.” As noted earlier, Rome II would have been a better system if the drafters had 

adopted the same logic when drafting the general rule of Article 4(1). 

When the law of the state of injury prescribes a higher standard than the state 

of conduct, the victim will not opt for the law of the conduct state, thus again 

producing the deterrent effect the drafters intended. As the Report points out, this 

application is 

[C]onducive to a policy of prevention, obliging operators established in 

countries with a low level of protection to abide by the higher levels of 

protection in neighbouring countries, which removes the incentive for an 

operator to opt for low-protection countries. The rule thus contributes to 

raising the general level of environmental protection. 163 

Significantly, because Article 7 is not subject to any exceptions, the polluter 

cannot avoid the application of the law of the state of injury by arguing, for example, 

that the occurrence of the injury in that state was not foreseeable. Apparently the 

article assumes that, regardless of distance, foreseeability is always present, or 

alternatively, that it should make no difference. One avenue for providing some 

protection for the polluter in this case is Article 17, and the Report alludes to this 

162. Id. 19-20. 

163. Explanatory Report, art. 7, at19. 



164 

possibility. This author’s view is that an objective foreseeability proviso is a better 

vehicle for ensuring fairness. If the polluter should have foreseen that his conduct in 

one state would cause injury in another state with higher standards, the polluter should 

not be able to seek refuge behind the lower standards of the state of conduct. 

Conversely, when foreseeability is lacking, the court should be allowed to take 

account of this factor, without necessarily exonerating the actor. For, although the 

application of the law of the state of injury would be unfair to the polluter, this result 

would be defensible because the non-application of that law would be equally unfair 

to the victim who has no control over the situation. 

164. See id. at20. 

165. ROME II, art. 17. 


VIII. RULES OF “SAFETY AND CONDUCT” 

Buried towards the end of the Rome II regulation is the familiar yet somewhat 

enigmatic provision of Article 17. It provides that, regardless of which law governs 

the non-contractual obligation under the other articles of Rome II, in “assessing the 

conduct of the person claimed to be liable, account shall be taken, as a matter of fact 

and in so far as is appropriate, of the rules of safety and conduct which were in force 

165 

at the place and time of the event giving rise to the liability.” As noted earlier, 

provisions similar to Article 17, albeit with slightly different wording, are found in the 

166 

1972 EEC draft convention on contractual and non-contractual obligations, the 

167 

Hague conventions on traffic accidents and products liability, as well as the 

168 169 170 

Belgian, Dutch, and Swiss codifications enacted in the interim. The Preamble 

of Rome II states that the term ‘rules of safety and conduct’ should be interpreted as 

referring to “all regulations having any relation to safety and conduct, including, for 

167. See art. 7 of the Hague Convention of 4 May 1971 on the Law Applicable to Traffic Accidents 

(“rules relating to the control and safety of the traffic”), and art. 9 of the Hague Convention of 

2 October 1973 on the Law Applicable to Products Liability (“rules of conduct and safety”). 

168. See BELGIAN PIL CODE, art. 102 (“consideration must be given to the safety and conduct rules” 

of the conduct state). 

169. See DUTCH PIL ACT, art. 8 (authorizing the “taking into account of the traffic and safety 

regulations and other comparable regulations for the protection of persons or property in force 

at the place of the tort.”). 

170. See SWISS PIL ACT, art. 142(2) (“Rules of safety and conduct in force at the place of the act are 

taken into consideration”). Cf. also PORTUGUESE CIV. CODE, art. 45(3) (“without prejudice to 

provisions of local state laws which must be applied to all persons without differentiation”); 

Hungarian PIL Decree, § 33.1 (“The law of the place of the tortious conduct shall determine 

whether the tortious conduct was realized by the violation of traffic or other security 

regulations.”). 



171 

example, road safety rules in the case of an accident.” The very fact that Rome II 

contains a separate article dealing with these rules, is an oblique and grudging 

recognition of the distinction noted earlier between conduct-regulating rules and loss 

distributing tort rules. 

However, it is unclear whether this article is a true rule of choice-of-law rather 

than a mere evidentiary instruction about which facts are relevant in determining the 

172 

degree of the defendant’s culpability. Although the article’s wording clearly 

suggests the latter possibility, it is worth exploring whether the article is capable of 

being used as a choice-of-law rule, which can lead to applying, rather than merely 

“taking account” of the law of the conduct state in appropriate cases. If the answer is 

affirmative, then Article 17 can provide a flexible exception to all of Rome II’s 

articles that lead to a law other than that of the state of conduct, particularly: (a) the 

lex loci damni rule of Article 4(1) in conflicts arising from cross-border torts; and (b) 

the common-domicile rule of Article 4(2), in conflicts arising from intrastate torts. 

Unfortunately, there are several obstacles to transforming Article 17 into a true 

rule of choice-of-law. The first obstacle, which is not insurmountable, is that, despite 

173 

using the imperative “shall,” the article is entirely discretionary, as it should be. The 

second obstacle is that, under the article, the court need not apply the rules of safety 

and conduct, but may simply “take them into account.” Although the article itself does 

not preclude the application of these rules, the Explanatory Report suggests that this 

is precisely what was intended. It states that “[t]aking account of foreign law is not the 

same thing as applying it: the court will apply only the law that is applicable under the 

conflict rule, but it must take account of another law as a point of fact, for example 

when assessing the seriousness of the fault or the author’s good or bad faith for the 

174 

purposes of the measure of damages.” This statement, especially the italicized word 

“only,” seems to preclude the application of the safety and conduct rules of the 

175 

conduct state. The third obstacle is that these rules are to be taken into account “as 

a matter of fact” in “assessing the conduct” of the tortfeasor, a notion that echoes 

171. ROME II, recital (34) (emphasis added). 

172. For the concept of foreign law as “datum,” see Erik Jayme, Ausländische Rechtsregeln und 

Tatbestand inländischer Sachnormen—Betrachtungen zu Ehrenzweigs Datum-Theorie, in 

GEDÄCHTNISSCHRIFT FÜR EHRENZWEIG 35 (1976); Herma H. Kay, Conflict of Laws: Foreign 

Law as Datum, 53 CALIF. L. REV. 47 (1965). 

173. This is clear from both the legislative history and the phrase “in so far as appropriate” in art. 17, 

although that phrase also has a quantitative component. 

174. Explanatory Report, art. 13, p. 25 (emphasis added). 

175. The Report acknowledges the origin of Article 17 in, among others, Article 9 of the Hague 

Products Liability Convention, which does “not preclude consideration being given” to the rules 

of conduct and safety of the state of the product’s marketing. Despite this equivocal wording, 

however, the Explanatory Report of the Hague Convention states that a court could well apply 

the conduct rules of that state. See Willis Reese, Explanatory Report to the Hague Products 

Liability Convention, ACTS AND DOCUMENTS OF THE TWELFTH SESSION, v. III, 251, 268 (1972). 



176 

Ehrenzweig’s concept of foreign law as a datum. This phraseology is considerably 

narrower than the Commission’s proposal, which provided that these rules were to be 

177 

taken into account “in determining liability.” This restrictive formulation reduces 

Article 17 to a mere evidentiary instruction of how to assess the tortfeasor’s 

culpability. Unfortunately, this minimalist conception prevents Article 17 from 

becoming the useful corrective tool that Rome II sorely needs. 

Moreover, the Rome II drafters seem to envision an even narrower, one-sided 

role for Article 17. They describe it as tool for helping the tortfeasor, but not 

necessarily the victim. Rome II’s preamble states that Article 17 is necessary “[i]n 

178 

order to strike a reasonable balance between the parties,” while the Report seems 

to be preoccupied with the plight of “the perpetrator [who] must abide by the rules of 

179 

safety and conduct in force in the country in which he operates.” In rejecting 

Parliament’s attempt at first reading to make Article 17 inapplicable to cases of 

defamation and unfair competition, the Commission saw “no reason for depriving the 

perpetrators of these two categories of liability of the protection which this rule gives 

180 

them.” Elsewhere, the Report describes a situation falling within Article 7 in which 

the perpetrator complied with the lower environmental standards of the state of 

conduct but not with the higher standards of the state of injury. The Report concludes 

that, in such a situation, “[u]nder Article 17, the court must then be able to have regard 

to the fact that the perpetrator has complied with the rules in force in the country in 

which he is in business.” 181 

The view of this author is that, in cases such as the one described above, the 

concern for the perpetrator is excessive, if not misplaced. The key question in such 

cases should be whether, under these facts, a reasonable person should have foreseen 

that his conduct in the one state would produce injury in the other state. For example, 

one who operates a chemical factory in close proximity to the border in state A should 

have foreseen that the wind will likely carry the factory’s emissions across the border 

into state B. Under these circumstances, the operator should not be allowed to seek 

refuge in the lower standards of state A. Using Article 17 as such a refuge would be 

contrary to the spirit of Article 7 and the “polluter pays” principle that article 


177. Commission Proposal, art. 13 (2003). Also, the phrase “as a matter of fact and in so far as is 

appropriate” was not part of the Commission’s proposal. 

178. ROME II, recital (34) (“In order to strike a reasonable balance between the parties, account must 

be taken of the rules of safety and conduct in operation in the country in which the harmful act 

was committed, even where the non-contractual obligations is governed by the law of another 

country.”). 

179. Explanatory Report, art. 13, at 25. 

180. COD/2003/0168 (Feb. 21, 2006) Commission modified legislative proposal (emphasis added). 

181. Explanatory Report under art. 7. 



182 embodies. Likewise, returning to the snow avalanche hypothetical, the Swiss 

operator who engaged in blasting operations near the Swiss-French border in the Alps 

should have anticipated that his operations may cause avalanches, some of which 

would occur across the border in France. If French law imposes higher standards of 

conduct than Swiss law, why should the operator be allowed to seek refuge in Swiss 

law? 

In any event, if one is to use Article 17 to protect the tortfeasor in this scenario, 

there is little justification for not using the same article in the converse scenario in 

which Swiss law imposes higher standards of conduct for blasting operations than 

French law. Whether one thinks in terms of state interest or simply in terms of 

evenhandedness toward litigants, there is good reason to “take account” and indeed 

to apply Swiss law in this case. The fact that the operator violated the standards of 

Swiss law implicates Switzerland’s policies in policing conduct within its borders, 

even though the consequences of that conduct in this case materialized in France. 

Conversely, this case does not implicate the policies underlying the French lowerstandard 

rule because that rule is designed to protect or encourage conduct within and 

not beyond French borders. In other words, this is what is known in the American 

conflicts lexicon as a classic “false conflict” in which only Switzerland has an interest 

in applying its law, or at least to have it “taken into account.” 

To be sure, the quoted term may be anathema to Rome II given its limited 

recognition of the role of state interests, despite, for example, embracing the concept 

of mandatory rules (Art. 16). Instead, Rome II places a premium on the need to 

“ensure a reasonable balance between the interests of the person claimed to be liable 

183 

and the person who has sustained damage,” and the Explanatory Report purports to 

explain most of Rome II’s rules in terms of the parties’ expectations. As noted earlier, 

the reason for the drafters’s decision to reject a general rule of allowing the victim to 

choose between the laws of the places of conduct and injury in cross-border torts other 

than environmental torts is because such a solution “would go beyond the victim’s 

184 

legitimate expectations.” However, one can also turn that question around. Does the 

application of the law of the state of conduct violate the legitimate expectations of the 

tortfeasor who violated the conduct standards of that state, just because the injury 

occurs across the border? The only Rome II provision that can provide a negative and, 

in this author’s view, justified answer to this question is Article 17. If the drafters’ 

intent was to use this article only when it helps the tortfeasor but not when it helps the 

victim, then it would have been preferable to suppress the article from the final text. 


183. ROME II, recital (16), et passim. 

184. Explanatory Report, art. 3, at 11-12. 



185. In cases of unfair competition, restriction to competition, and infringement of intellectual 

property rights, Rome II does not allow choice-of-law agreements, either before or after the tort. 

See ROME II, arts. 6(4), 8(3). 

186. ROME II, art. 14(1)(b). 

187. Id. The fact that Article 14 does not expressly impose the same requirement in the case of posttort 

agreements may give rise to an a contrario argument that free negotiation is not necessary 

in such cases. Of course, the argument should be rejected. Apparently, the drafters intended to 

ensure higher scrutiny of pre-tort agreements and not to free post-tort agreements from any 

scrutiny. 

188. Id. art. 14(1). 

189. Id. 

IX. PARTY AUTONOMY 

Under the heading “freedom of choice,” Article 14 of Rome II introduces the 

notion that the parties to a tort may agree on which law will govern their rights and 

185 

obligations resulting from the tort. The article properly distinguishes between 

choice-of-law agreements made before and after the tort. It allows post-tort 

agreements between all parties and allows pre-tort agreements only if all the parties 

186 

are “pursuing a commercial activity.” In the latter case, the agreement must be 

187 

“freely negotiated.” For the remainder, Article 14 treats both pre-tort and post-tort 

agreements alike: (1) both must be “expressed or demonstrated with reasonable 

188 

certainty by the circumstances of the case;” and (2) neither may prejudice the rights 

189 

of third parties, or derogate from the mandatory rules of a state in which “all the 

190 

elements relevant to the situation . . . are located,” or, in certain cases, from the 

mandatory rules of Community law. 191 

Obviously, post-tort agreements are far less problematical because, after the 

occurrence of the tort, the parties are in a position to know of their rights and 

obligations and have the opportunity to weigh the pros and cons of a choice-of-law 

agreement. For this reason, these agreements need little policing by the legal system. 

In fact, the system benefits from these agreements insofar as they promote judicial 

economy. Precisely the opposite is true of pre-tort agreements. The parties do not (and 

should not) contemplate a future tort, they do not know who will injure whom, or 

what will be the nature or severity of the injury. Moreover, a weak or unsophisticated 

party may uncritically sign such an agreement, even when the odds of him being the 

victim are much higher than the odds of him being the tortfeasor. For these and other 

190. ROME II, art. 14(2) (“Where all the elements relevant to the situation . . . are located in a country 

other than the country whose law has been chosen, the choice of the parties shall not prejudice 

the application of provisions of the law of that other country which cannot be derogated from 

by agreement.”). 

191. See ROME II, art.(3) (“Where all the elements relevant to the situation . . . are located in one or 

more of the Member States, the parties’ choice of the law applicable other than that of a 

Member State shall not prejudice the application of provisions of Community law . . . which 

cannot be derogated from by agreement.”). 



192 

reasons, pre-tort agreements should either be prohibited or, if permitted, they should 

be closely policed. 

Regrettably, Article 14 requires only minimal scrutiny. The only restrictions 

it imposes on pre-tort agreements (that it does not impose on post-tort agreements) is 

that it must be “freely negotiated” and that the parties must be “pursuing a commercial 

activity.” This is neither sufficient nor free of problems. Even if the term “commercial 

activity” was clearly defined or uniformly understood throughout the EU, it would still 

include within its scope, relationships that are one-sided, such as those arising from 

franchise, licensing, or insurance contracts. For example, a franchise contract is clearly 

commercial, yet the franchisee is usually in a very weak bargaining position (which 

is why so many states have enacted consumer-protection type statutes to protect 

193 

franchisees). By allowing pre-tort choice-of-law agreements in these contracts, 

Article 14 does not live up to the statement in recital 32 that “[p]rotection should be 

194 

given to weaker parties by imposing certain conditions on the choice.” As with 

some other freedom-laden ideas, Article 14 may well become the vehicle for taking 

advantage of weak parties, many of whom are parties to “commercial” relationships. 

The argument that the “mandatory rules” of paragraphs 2 and 3 of Article 14, or the 

ordre public exception of Article 26 will protect the weak parties is overly optimistic 

because of the high threshold these provisions require before they become operable. 

192. See, e.g., EGBGB, art. 42 “After the event giving rise to a non-contractual obligation has 

occurred, the parties may choose the law that shall apply to the obligation.”) (emphasis added); 

BELGIAN PIL CODE, accord. But see DUTCH PIL ACT, art. 6 (“Where the parties have chosen 

the law applicable to any matter relating to tort, . . . that law shall apply between them . . .”). For 

the position of American cases law on this issue, see SYMEON C. SYMEONIDES, AMERICAN 

PRIVATE INTERNATIONAL LAW _???__ (forthcoming 2008). 

193. See SYMEONIDES, AMERICAN PRIVATE INTERNATIONAL LAW, supra note 192 at _???_. 


X. CONCLUDING THOUGHTS 

Considering its starting point—the Commission’s preliminary draft— Rome 

II could have been much worse. Considering the rapporteur’s valiant efforts and 

Parliament’s amendments, Rome II could also have been much better. The amendments 

injected more flexibility, introduced issue-by-issue analysis, and differentiated 

between issues of conduct regulation and loss distribution. Had any of these 

amendments survived, they would have considerably improved Rome II. Unfortunately, 

the Council and Commission rejected these and other amendments. In the end, 

Rome II is what it is. In this author’s view, it is a missed opportunity to do much 

better. This view is reinforced by the fact that the European continent, the birthplace 

of PIL, has never had a shortage of talent, sophistication, and experience in drafting 

PIL legislation. Recent national codifications are a solemn testimony of that 



experience and wisdom, as is the GEDIP proposal which was drafted specifically for 

the pan-European level. 

Be that as it may, the final text of Rome II illustrates the difficulties of the 

process of lawmaking in a multinational democratic society. More often than not, the 

need to obtain a consensus, which is so important and healthy, necessitates compromises 

that disturb the coherence and consistency of the final scheme. This author can 

empathize with the drafters’ unenviable position, having found himself in a similar 

situation in drafting two PIL codifications and being currently involved in a third. 195 

During the six-year process of drafting the Louisiana codification, he had to make 

196 

compromises which, at the time, he strongly resisted. In retrospect, however, 

besides recognizing that with regard to some of them the critics may have had the 

197 

better argument, he views the rest of the compromises as the political price that had 

to be paid in return for getting the bill passed. It was either that, or nothing at all. The 

bill became law and, for the last fifteen years, it has worked well. 198 

Based on this experience, the author assumes that, as an outside observer, he 

sees no more than half of the picture and acknowledges the objective political 

difficulties the Rome II drafters faced in obtaining consensus. These difficulties at 

least partially explain the differences between Rome II and, for example, the GEDIP 

proposal. The reason the GEDIP proposal was so close to perfection had more to do 

with the homogeneity of that academic group than with the fact that it encompasses 

la creme de la creme of European conflicts thought. The proposal was a sophisticated 

document drafted for sophisticated judges capable of, and entrusted with the discretion 

necessary for making the fine balancing the proposal envisioned. In contrast, Rome 

II is a pragmatic document that aims for simplicity and uniformity and primarily seeks 

to preserve the status quo rather than to dramatically alter it. Whether this was the 

right choice in the long run is a matter of opinion, and this author’s opinion deserves 

no more deference than that of the erudite men and women who drafted Rome II or 

195. See Symeon C. Symeonides, Private International Law Codification in a Mixed Jurisdiction: 

The Louisiana Experience, 57 RABELSZ 460 (1993); Symeonides, Exegesis, supra note 193; 

Symeon C. Symeonides, Revising Puerto Rico’s Conflicts Law: A Preview, 28 COLUM. J. 

TRANSNAT’L L., 601 (1990); Symeon C. Symeonides, Codifying Choice of Law for Contracts: 

The Puerto Rico Project, in LAW AND JUSTICE IN A MULTISTATE WORLD: ESSAYS IN HONOR OF 

ARTHUR T. VON MEHREN, 419-37 (John Nafziger & Symeon Symeonides eds., 2002); Symeon 

C. Symeonides, Oregon’s Choice-of-Law Codification for Contract Conflicts: An Exegesis, 44 

WILLAMETTE L. REV. (forthcoming Dec. 2007); Symeon C. Symeonides, Codifying Choice of 

Law for Contracts: The Oregon Experience, 67 RABELSZ 726 (2003). 

196. See Symeonides, Exegesis, supra note 193, at 712-14, 742-48, 758-59. 

197. See Symeonides, THE CHOICE-OF-LAW REVOLUTION 263; Symeon C. Symeonides, In Search 

of New Choice-of-Law Solutions to Some Marital Property Problems of Migrant Spouses: A 

Response to the Critics, 13 (3) COMM. PROP. J. 11 (1986). 

198. See Patrick J. Borchers, Louisiana’s Conflicts Codification: Some Empirical Observations 

Regarding Decisional Predictability, 60 LA. L. REV. 1061 (2000). 



the many commentators who applaud it. 

In any event, although European PIL would have been better off with a 

“better” Rome II, the more realistic question is whether it would have been better off 

without Rome II at all. On balance, this author’s answer is in the negative. If nothing 

else, and despite its flaws, Rome II will unify and thus equalize, the private 

international law of the member-states of the European Union. Although for some of 

these states this equalization will amount to regress, for many more states it will be 

clear progress. From a transatlantic perspective, whatever its flaws, this unification or 

Europeanization cannot be worse than the complete lack common direction, much less 

unity, that characterizes modern American conflicts law. Moreover, one hopes that the 

present text of Rome II is not the last step in the arduous process of modernizing and 

unifying European private international law in tort conflicts. Indeed a review clause 

in Rome II leaves the door open for future changes and adaptations. 199 

199. See Rome II, art. 30 (“Not later than 20 August 2011, the Commission shall submit to the 

European Parliament, the Council and the European Economic and Social Committee a report 

on the application of this Regulation. If necessary, the report shall be accompanied by proposals 

to adapt this Regulation.”).

Rome II and Tort Conflicts: A Missed Opportunity Abstract Contents

Create successful ePaper yourself

Delete template?

Save as template?