Before we can turn to choice-of-law questions in the federal system, a little prologue is necessary. This prologue concerns the jungle of choice-of-law methodology. "Jungle" seems to be an appropriate word. Theories on choice-of-law methodology are as diverse as species in a tropical rain forest.13 This paper is an inappropriate instrument for thinning out this jungle. But some general remarks seem necessary on the methodology used, or rather the assumptions made, in this paper. The brief remarks may not convince someone who does not agree anyway, but they will at least make clear between which trees of the jungle this paper stands.
One assumption made throughout this paper is that states are not, and should not be, indifferent to whether their (mandatory) laws are applied in a particular case or not. States can have a more or less strong desire to have their laws applied depending on the case. This puts me on the American side of what is largely an Atlantic divide in choice-of-law theory. Being European, I guess I owe a brief justification in the form of answers to the questions: Why would a state care about the application of its laws ? Would it care at all ?
I cannot help thinking that the answer to the last question is a clear yes: If a state enacts a law, it does so for a reason - it wants to pursue some kind of policy, or, if we do not want to call any state decision (or any judge-made case law, for that matter) a policy, mediate between opposing private interests in conformity with some conception of "justice", this state's conception of "justice."14 This policy gets thwarted to a greater or less extent (this depends on the content of the other state's law, and on how many cases "escape" the state's law) if the state's law is not applied. A state would and should therefore not be indifferent to the application of its laws.15 The (mainly European) counter-voices contend that what matters in choice-of-law with regard to private law (= private international law, as opposed to choice-of-law in administrative law, social security law, and so on) is the interests of the private parties involved.16 Such interests surely exist, and we will discuss them later in section VI., in the context of the burden-test.17 But the argument that they are the only interests that matter is founded on a mistakenly sharp distinction of private law and public law, a distinction that we will encounter again, and try to refute, when discussing the burden-test18. This paper acknowledges that both state and private interests exist and matter.
The answer to the question if states care at all about the application of their laws was also already the answer to why they would care: in order to realize their respective substantive policies. Now it is a common argument that states' policies are designed to further the interests not of any citizens, but only of their own citizens, and that consequently states have an interest in having their laws applied only if this is advantageous for one of their own citizens in the given case.19
Factually, it is of course conceivable, although by no means obvious20, that states want to further only the interests of their own citizens.21 Be it submitted here that even if that were the case, a state would be well advised not to withhold the beneficial application of its laws from out-of-state individuals in any given case. This is because the out-of-state individuals will in anticipation of the state's partiality change their behavior, with potentially highly detrimental consequences for the state's citizens. For example, a state might decide to apply all its consumer protection legislation, including anti-fraud statutes and the like, only if the consumer is a state citizen. At first sight, this might benefit business in the state, since the local businesses would not bump up against protective legislation anymore when they sue out-of-state consumers in the state's courts. But the out-of-state consumers might eventually notice this, and stop shopping in the state altogether. For a more drastic example, imagine a state which does not enforce its criminal laws if the victim is an out-of-state individual - presumably, not many tourists from other states would come to visit. Even an "egoistic" state might therefore want to apply its laws to the benefit of out-of-state individuals in many instances.
But quite apart from these factual issues, states in a federal system are legally barred from so limiting the application of their laws. The reason is the non-discrimination requirement. This would be abundantly clear if a state enacted a law protecting a person in a certain situation, or with certain attributes, with the addition "... and if the person is a citizen of this state." (The criminal law example in the last paragraph is another obvious demonstration.) It should not be different in the choice-of-law context.22 Attempts have been made to save the narrow (i.e., strongly egoistic) definition of state interests, and the resulting choice-of-law system, from the verdict of discrimination by clever reformulation of the problem, but these defenses are ultimately unconvincing. 23
The preceding assertion calls for some clarifying remarks. First, following the conceptual line of this paper, one might want to ask if a non-discrimination requirement is a necessary part of a federal system. This is an interesting question, but not one that will be asked in this paper, for the simple reason that almost all federal systems have a non-discrimination requirement, at least the ones we are mainly concerned with here, the US and the EU. Second, and most importantly, one might want to object to the short treatment of non-discrimination that the issue is not so simple after all. Non-discrimination poses important problems in a federal system, especially in connection with the provision of government subsidies (like welfare payments).24 Choice-of-law in particular faces a dilemma: Is it discriminatory to apply a different law to a person in function of her different residence/nationality, because non-discrimination requires that people in otherwise (i.e., apart from residence/nationality) identical situations are submitted to the same rules? Or is it discriminatory not to apply a different law in function of her different residence/nationality, because equal treatment requires that everybody be submitted to their "own" law25 ?26 We will take up some aspects of this question in pragmatic terms in later sections - the dilemma seems to be inevitable (one way or another, there is some kind of discrimination), and so one should look pragmatically at which solution is better for the individuals concerned.27
So if the interests of a state in the application of its laws are not defined by the wish to favor the state's own citizens, what can we say about these interests in order to determine which law to apply in a given case ? The answer was given above: The substantive policies of a state determine its interest in having its law applied in a given case. The interest of the state will depend on how strongly its policies are implicated (i.e., to what extent they would be furthered by application of its law, or thwarted by non-application), and, arguably, on how much importance the state attaches to the policy that is implicated. One of this paper's purposes is to develop some guidelines as to the limits of these policies, or as to how conflicts between two states policies should be resolved.28
If state policies matter for choice-of-law, predictability should matter for choice-of-law, too. The simple reason is that many (and some, including myself, would say: most) rules' policies are implemented primarily through anticipation of positive or negative sanctions provided by the rule, i.e., through (dis-)incentives. Incentives only work where the sanctions can be foreseen, and sanctions can only be foreseen where the applicable rule is clear. If it is not clear whether a rule is applicable, the incentive mechanism will not work properly, because private actors will have to expect too much or too little sanctions with some degree of likelihood.29
Even if one does not buy into last paragraph's argument, it seems clear that unpredictability as to the applicable law will not only protract existing litigation, but also lead to too much or too little litigation (or both, in different places), since potential litigants have more difficulty predicting the outcome of the suit.30
Predictability as to the applicable law can only be achieved by some degree of specificity of choice-of-law rules. Of course, it is not possible to design choice-of-rules which perfectly allocate legislative spheres for all individual cases. Even if it were, the rules would have to be so complex that they could only be understood by a handful of specialists, and would be ignored in practice. At the same time, rules do not need to be inflexible and metaphysical as the First Restatement, always cited as the negative example of choice-of-law rules by those who dislike them. Substantive law, too, follows and embodies policies, and still it is, and must be (to make the task of judging feasible) at least partially laid down in rules, not broad standards - the judge who cannot understand a complex (choice-of-law) rule is unlikely to make a better (choice-of-law) decision free-handedly. Generally, the (theoretical) benefits (in terms of substantive policy realization) of refining choice-of-law through more complex rules should be weighed against the added administrative cost and risk of non-observance.31 Some set of choice-of-law rules is thus needed.32 This paper will proceed on that basis, and we will give some additional reasons for the superiority of rules vs. standards in choice-of-law below.
Lastly, the criteria relevant for choice-of-law are overwhelmingly territorial, at least in interstate choice-of-law in a federal system.
"Territoriality" has been somewhat of a bad word in choice-of-law theory of most of the last century. If territoriality of a choice-of-law approach meant that the approach attaches primary, or exclusive, importance to the place where a (or: the last) physical act was committed, territoriality would indeed deserve to be discarded. The standard negative example is Beale's First Restatement.33 But if territoriality means that choice-of-law is concerned with regulatory authority of states, and that regulatory authority of states is (at least primarily) defined as the regulatory authority for some territory of this planet's surface, the concept seems almost self-evident. The alternative would be a system that vests a legislature or administration with the power to regulate the affairs, or certain affairs, of a certain group of people. But a US State is not an entity responsible for, say, all Scottish immigrants and their descendents; it is an entity responsible for a certain part of the US territory. And a Member State of the EU is not, or at least not only, an entity responsible for a certain people, e.g., all Francophones, or all French; it is an entity (at least also) responsible for a certain part of the territory of the EU. This is not to say that US States and EU Member States are responsible only for their territory. EU Member States may have certain responsibilities for, or authority on, their nationals, US States may have the same for their residents.34 Determining the respective realms of territorial and personal repartition of state responsibilities and powers is in fact quite a complex matter which we will briefly discuss below.35 But given the (at least dominantly) territorial definition of states, the general idea of using territorial criteria in choice-of-law should be unquestionable.36
13 See generally, e.g., Lea Brilmayer, Conflict of Laws (2nd ed. 1995); Axel Flessner, Interessenjurisprudenz im IPR (1990); Gerhard Kegel, Paternal Home and Dream Home: Traditional Conflict of Laws and the American Reformers, 27 Am. J. Comp. L. 615 (1979); Larry Kramer, Rethinking Choice-of-Law, 90 Colum. L. Rev. 277 (1990) [hereinafter: Rethinking], and More Notes on Methods and Objectives in the Conflict of Laws, 24 Cornell Int'l L. J. 245 (1991) [hereinafter: More Notes]; Paul Lagarde, Le Principe de Proximité dans le Droit International Privé Contemporain, 196 R.C.A.D.I. 9 (1986-I); Eugene F. Scoles, Peter hay, Patrick J. Borchers & Symeon C. Symeonides, Conflict of Laws, ch. 2 (3rd ed. 2000).
14 One might want to object that legislators do not pursue any ideal of justice (this is the reason why I put the word in quotation marks), and that all the law expresses is the result of bargaining between interest groups. For present purposes, this does not seem to matter - for the interest groups would presumably not like to see their compromise undermined by the application of another state's law.
15 See, e.g., William A. Baxter, Choice of Law and the Federal System, 16 Stan. L. Rev. 1-25 (1963). Cf. also Kramer, More Notes, supra note 13, 261-269. Incidentally, at least the judges have a strong interest, or assume a strong state interest, in applying their law - even where the foreign law is well known, courts around the world favor their own law which they, unsurprisingly, perceive to be more just (forum bias).
16 See, e.g., Albert A. Ehrenzweig, Conflicts in a Nutshell 18 (3rd ed. 1974) (Ehrenzweig taught at Berkeley, but he was one of the German immigrants there); Jan Kropholler, Internationales Privatrecht 18 (4th ed. 2001).
17 See infra VI.D.2.a), p. 39. These private interests are, of course, also state interests, if one assumes, as one should, that the states generally want to further the interests of their citizens.
18 See infra V.B.4, p. 29. Cf. for the implications of a more modern conception of private and state interests and the nature of "private" law for choice of law also Joel P. Trachtman, Conflict of Laws and Accuracy in the Allocation of Government Responsibility, 26 Vanderbilt J. of Transn'l L. 975, 997-998, 1009-1010.
19 This was the position of, above all, the father of interest analysis, Brainerd Currie. See, e.g., Brainerd Currie, Selected Essays on the Conflict of Laws 77, 81-86 (1963) (reprinted from Brainerd Currie, Married Women's Contracts: A Study in Conflicts-of-Law Method, 25 U. Chi. L. Rev. 227 (1958)).
20 "Humanitarian interventions", development aid, and many other (at least possibly) altruistic actions by states show that states do not care exclusively about their own citizens.
21 For a thorough discussion of the problematic definition and determination of state interests generally, see Brilmayer, supra note 13, ch. 2.
22 Cf. the references given for the US constitutional debate, infra note 96.
23 Douglas A. Laycock, Equal Citizens of Equal and Territorial States: The Constitutional Foundations of Choice of Law, 92 Colum. L. Rev. 249, 283-285 (1992).
24 Cf., e.g., Mark P. Gergen, The Selfish State and the Market, 66 Texas L. Rev. 1097 (1988); Richard Posner, Economic Analysis of Law § 25.7 (5th ed. 1998); and Camps Newfound/Owatonna, Inc. v. Town of Harrison, 520 U.S. 564 (1997).
25 The latter idea becomes more understandable if one thinks of family law, inheritance, and so on.
26 See, e.g., G. Fischer, Gemeinschaftsrecht und kollisionsrechtliches Staatsangehörigkeitsprinzip, in Gemeinschaftsrecht und IPR 157 (Christian von Bar ed. 1991); Mark P. Gergen, Equality and the Conflict of Laws, 73 Iowa L. Rev. 893 (1988); ECJ Case C-430/97, Johannes ./. Johannes, 1999 E.C.R. I-3475 (holding that submitting the obligations resulting from a divorce to the national law of the divorcees does not violate Art. 12 EC); and Lemmon v. People, 20 N.Y. 562, 608 (1860) (same for the use of domicile and the Privileges and Immunities Clause of the US Constitution).
27 See infra III.B.3, p. 14, and VI.D, p. 36.
28 There is a subtle difference between the first and the second way of phrasing the problem, as noted by many scholars. My personal opinion is that from a pragmatic perspective, the difference does not seem to matter.
29 Cf., e.g., William H. Allen & Erin A. O'Hara, Second Generation Law and Economics of Conflict of Laws: Baxter's Comparative Impairment and beyond, 51 Stan. L. Rev. 1011, 1041-1047 (1999); Larry Kramer, On the Need for a Uniform Choice of Law Code [hereinafter Uniform Code], 89 Mich. L. Rev. 2134, 2137 (1991).
30 E.g., Michael H. Gottesmann, Draining the Dismal Swamp: The Case for Federal Choice of Law Statutes, 80 Geo. L. J. 1, 12-13 (1991).
31 See generally Louis Kaplow, The Value of Accuracy in Adjudication: An Economic Analysis, 23 J. Legal Stud. 307 (1994).
32 Cf. Scoles et al., supra note 13, § 2.26 with numerous further references.
33 A typical example of the critique of this kind of "territorialism" is Kermit Roosevelt III, The Myth of Choice of Law: Rethinking Conflicts, 97 Mich. L. Rev. 2448, 2455-2458, 2472-2474 (1999).
34 See Donald H. Regan, Siamese Essays: (I) CTS Corp. v. Dynamics Corp. of America and Dormant Commerce Clause Doctrine; (II) Extraterritorial State Legislation, 85 Mich. L. Rev. 1865, 1908-1909 (1987) - even then, residence, and to a lesser extent nationality, is linked to the territory where a person lives.
35 Cf. infra III.B.3.c), p. 16.
36 Cf. generally Laycock, supra note 23, 315-320, 323; Regan, supra note 34, 1885-1891; also James A. Martin, Constitutional Limitations on Choice of Law, 61 Cornell L. Rev. 185, 230 (1976); Breavington v Godleman (1988) 168 CLR 41, 128-129 (Deane, J., dissenting) (Austl.).