We had already seen that choice-of-law in a federal system (interstate choice-of-law) is different from international choice-of-law because of the non-discrimination requirement. We will now consider if interstate choice-of-law is different in other respects, too.
In a federalistic state (union), legislative powers are divided between
the federal government37
and the constituent states38, such as in the US and the EU.39 The federal government
assumes certain tasks that the states have deemed useful, and agreed, to
centralize. Legislative power for non-centralized tasks remains with the
states. The result is vast legislative diversity (of the laws of the different
states). Choice-of-law administers this diversity; it determines which of the
laws of the different states will be applied in a given case.
One can approach the federalistic model from two sides. One side is: Why do states come together to form, and partly abandon legislative powers to, a union in the first place ? If there is advantages from centralization in certain fields, why do only some groups of states enter into federation agreements, while many others stay on their own ? The flipside is: Why do states in a federation not go all the way to full centralization ? After all, maintaining multiple levels of government seems costly and complicated. The answers to these questions are, of course, highly complex, and the actual genesis and shape of federations are often the result of historical coincidences. The purpose of this section is merely to recall some basic, and rather common place, arguments, and to draw some inferences for inter-state choice-of-law from these arguments.40
This section will try to show that the same reasons as for forming a union should make a mutually beneficial cooperative approach to choice-of-law (1.) especially palatable, and almost mandatory, to states in a federal system (2.).
Whether united in a federal union or not, states can increase the effectiveness of their own laws by submitting themselves to a system of neutral choice-of-law rules. Neutral choice-of-law rules are choice-of-law rules that do not favor any particular state's law, or any particular substantive policy. They are of the type "in matters of type X, apply the law of the state, and only of the state, where element Y of the case points to."41 The beneficial effect of such a system has been demonstrated by Lea Brilmayer and Larry Kramer.42
A state could try to give preference to its own laws in every case even loosely related to its legislative domain. This would mean enforcing its laws in many cases where enforcement furthers the policies of the state very little. In other cases, where the state has a strong interest in having its law applied, other states' courts may obtain adjudicative jurisdiction. These other states, following a similar approach as the first state, would follow their respective laws, thus setting aside the first state's law in this case where applying the first state's law would further that first state's policy considerably. All states face this dilemma. The solution lies in cooperation. If the other states reciprocate, each state is better served if it commits itself not to apply its own law where another state's policies are more strongly implicated, or the other state places more importance on the policy implicated. This "increases the pie" of policy realization, or of the realization of important policies, and presumably each state stands to benefit from that. The cumulative application of more than one state's law is not an alternative - every rule is a dividing line, if just between individual freedom and the interests of other parties, and cumulative application of two laws wipes out all except one of them.43
One problem with this approach is, of course, that it might be quite difficult to determine which state's policy is more strongly implicated, or is more important to the state, in a given case. We will get back to this point later. For now let us assume that we can at least for most cases make a satisfactory determination of which state has, or should have, a stronger interest to have its laws applied in a given case.
The other problem with the cooperative approach is that the states are facing a prisoners' dilemma type situation. Each state (or its court) has an incentive in (almost) each case to apply its own law, but overall they are all better off if they refrain from applying their own law if another state has a stronger interest in the case. As William F. Baxter put it: "Baseball's place as the favorite American pastime would not long survive if the responsibilities of the umpire were transferred to the first team member who managed to rule on a disputed event."44 One solution, the one Baxter was primarily advocating, is the neutral umpire that makes Baseball a (more or less) orderly sport. Baxter's proposal was not only about bringing in the federal courts as an umpire, though. He wanted the federal courts to craft choice-of-law rules for the entire American legal system, i.e., rules more specific than the pure principle of maximizing the realization of states' interests. Choice-of-law rules, rather than lose standards, have many advantages in terms of predictability for the parties, and administrability for the courts, and thus, ultimately, for state policy realization.45 But for the cooperative choice-of-law approach, what is most important about more specific rules is that their implementation is better observable than that of a broad standard. To wit, the federal courts in Baxter's proposal would not decide every choice-of-law dispute. Often state courts would rule on choice-of-law issues, and the federal courts would only control that the state courts play by the rules. But controlling that the state courts play by the rules is a very onerous task if the rule is a broad standard, which requires to take into account all the circumstances of the case. Creating more specific rules thus helps solving an agency-problem as between the federal courts (principal) and the state courts (agent). In fact, with more specific rules, one might even be able to get along without a neutral umpire. If rules are clear cut enough so that their (non-)observation by a state's courts is readily observable, fear of retaliation by other states in future cases (= repeat-game situation) might be enough to overcome the prisoners' dilemma situation in each individual case. Put in a positive way, clear cut rules might enable a state to credibly commit to following neutral choice-of-law rules, such that other states may find it worthwhile to follow suit. There are many caveats to this "self-regulating" situation. For present purposes it is sufficient to note that rules, rather than lose standards, are preferable for the cooperative approach to choice-of-law, and that ideally a neutral umpire controls the observation of these rules.
This beneficial cooperative approach to choice-of-law commends itself in a federal system. To begin with, the federal system has the convenience of having a preexisting neutral umpire, the federal judiciary and legislator. But the cooperative approach is not only more practical in a federal system, it is also inherent to the reasons for forming a union in the first place. The case for establishing a union among formerly independent states can be made, grossly simplifying, from two different perspectives, idealistic and pragmatic, both of which point towards a cooperative approach to choice-of-law.46
From an idealistic perspective, states may feel that they belong to a larger community, be it ethnic or cultural, which they feel the desire to join. This spirit of community should go along with respect for the other states and their laws, making it more palatable for the states to apply the sister states' laws. To be sure, this argument is very fragile, perhaps even speculative, as an empirical claim. A state might be pragmatic about abandoning some powers to a federal structure, while sticking to its chauvinistic approach for all remaining, i.e., not federalized, powers. But hindsight suggests that states do not join a federal structure if they do not have mutual respect for each other, and that encompasses respect for the other states' laws, and the policies embodied therein.
Even more important, and less speculative, is the argument based on the shared value system in a federalistic state. States entering a union usually belong to some cultural or ethnic community with shared values anyway. In addition to this, and perhaps more importantly, the union usually makes some of these shared values mandatory for all states, thus reinforcing (or newly creating) the shared value system. This encompasses human rights bills as well as harmonizing legislation. It reduces the variance , or at least prevents egregious differences, between the laws of the different states. The degree to which a state compromises its own policies in applying a sister state's law should thus be limited, making such application easier.47 In hip game theory jargon, the benefit from shirking in an individual case is smaller than in an average international case, making shirking less likely.
From a pragmatic perspective, states may decide to join a union in order to benefit from economies of scale in legislation and administration, tackle otherwise unmanageable border-transcending problems, and avoid being locked in a race-to-the-bottom.48 This presupposes that in order to further their policy objectives, even if diluted in a union-wide decision-making process, they are willing to give up some of their "sovereignty" to the benefit of the union. It seems that states that were willing to abandon some of their "sovereignty" to a union might have a more pragmatic way of dealing with "sovereignty" issues. One might think that they should then be equally willing to give up some of their "sovereignty" vis-à-vis the sister states in individual cases under the cooperative approach, if this furthers the implementation of their own policy objectives on the whole. Again, the degree to which a state's policy is compromised in a given case should be reduced in a federal system anyway, because of the shared underlying values.
Finally, if one conceives of federalism as a delegation of power from the federal entity to the states, rather than as a delegation of power from the states to the federal entity, the cooperative approach seems to be inescapable: The federal entity would not delegate to the states a power mutually to thwart their respective policies.
In the preceding paragraphs, I advocated a neutral umpire as part of a cooperative choice-of-law approach. I somewhat rashly (and tacitly) jumped to the conclusion that the neutral umpire would not only be the federal courts, but also the federal legislator, i.e., that we need federal choice-of-law rules. It is true that by Baxter's baseball allegory, the point seems to be obvious: Baseball's place as the favorite American pastime would presumably not be preserved if the umpire just controlled that the first team member who managed to rule on a disputed event did so according to some neutral rule defined by that team. Choice-of-law, one could argue, needs common rules for everybody just like baseball does.49 But is this allegory enough to dispose of the general federalistic arguments for legislative diversity as applied to choice-of-law ? If, in a federal system, we encourage legislative diversity in many fields, should not the same reasons support diverse (state) choice-of-law rules ? In fact, the states' power to make their own choice-of-law rules has been considered by some a vital part of their remaining independence.50
The discussion of these questions anticipates a little bit the general discussion of reasons for legislative diversity in the following section B. However, it seems adequate to raise these questions now to round off the discussion of the cooperative approach in general, and of the neutral umpire and the neutral rules in particular.
For the purpose of making the cooperative approach work, it might be sufficient, as noted above, if every state committed to a system of neutral rules, albeit its own system of neutral rules, in order to convince the other states that one is not favoring one's own laws, i.e., that one is playing the cooperative game. Unless some state or the union is more skilled and resourceful to devise optimal choice-of-law rules, no one state (or the union, for that matter) is more likely to come up with the optimal rule than others, so that one might think that the expected degree of policy realization is the same whether one has federal or state choice-of-law rules. Better still, experimentation would be furthered, and an optimal rule would seem to be more likely to be found, if all states devised their own choice-of-law rules.51 State choice-of-law rules might also be better adapted to local preferences in two regards. First, the citizens of different states might have different preferences as to the allocation of specific choice-of-law costs, like information cost (often called "multistate interests").52 Second, I have repeatedly been alluding to the possibility that maximizing state interests might not only have to take into account the degree of potential impairment of different policies, but also the relative importance a state attaches to one or the other of its policies. For example, Germany might put its consumer protection policy above all else, while Spain favors its (anti-)divorce laws, so that both would profit if Germany compromised its divorce laws in exchange for an extension of its consumer protection in relation to Spain. Lastly, if some local circumstances are relevant for choice-of-law decisions, the state legislator has a regulatory (i.e., informational) advantage in making such decisions. These arguments do not stand up to closer scrutiny, though.
Local preferences, and local circumstances, certainly might matter for choice-of-law (it is another question if they matter much, i.e., if the local differences are significant). But since a choice-of-law necessarily involves the local preferences and circumstances of at least two states, it is not particularly helpful to leave the assessment of these preferences and circumstances to one of them alone.53 Only a bilateral treaty would be superior to a federal rule - if only two states are concerned. If there is more states' policies involved in a case, the treaty would have to include all these states. It is not hard to imagine that following through on this line (of treaties between individual states) would result in an extremely complex, and presumably totally impracticable system. Rarely, if ever, would one expect local differences to be at the same time so important as to justify the hassle of negotiating and implementing such tailor-made rules, and complementary such that an agreement (like the fictional agreement between Germany and Spain mentioned above) departing from neutral rules would significantly increase state interest maximization.
As to the experimentation with different choice-of-law solutions, the experiments will be flawed if, and to the extent that, plaintiffs can avoid even the best choice-of-law rule by choosing another forum which follows another choice-of-law rule.
Most importantly, states cannot be assured of other states' cooperation in choice-of-law just because the other states adopt nominally neutral choice-of-law rules. States have ample opportunity to discriminate in fact with nominally neutral choice-of-law rules. An example: Imagine, not completely unrealistically, that Germany's major contacts with Spain are German tourists going to Spain, and Spanish workers coming to Germany. Then a set of nominally neutral choice-of-law rules that both divorce and consumer contracts are governed by the law of residence of the divorcing couple and the consumer, respectively, assures application of German law, and realization of German policies, in all cases in relation to Spain. It seems therefore not enough that the union control that all states adopt nominally neutral rules. The union should set neutral rules itself.
In sum, the case for diversity in choice-of-law rules is untenable. At most, one could argue for federal or treaty choice-of-law rules which, while not trying to advantage one state or the other, are not all nominally neutral, in order to accord for local conditions. This should be a rare exception, though. Generally, choice-of-law in a federal system requires neutral choice-of-law rules set by the federal legislator (or the federal courts, for that matter). After all, choice-of-law describes the boundaries of states' regulatory authority, and as with other boundary conflicts between states, the union seems to be the only adequate arbiter.54
I argued that states in a federal system would, and should be ready to, maximize their policy interests implicated in choice-of-law by submitting themselves to neutral choice-of-law rules, ideally rules devised and controlled on the federal level. But which rules would achieve this goal ? I said that state policies might be more or less strongly implicated in a given case, and that a state might attach more or less importance to one policy or the other. Concerning the latter point, no abstract indications seem to flow from federalism. Concerning the former point, the arguments for forming a union seemed to have nothing to say, i.e., these arguments were agnostic to the question which law should be designated as applicable. The arguments for preserving the states might fill that gap.55 The reasons why we have different laws in the first place might guide us in determining which of them to apply. This section will therefore be dedicated to scrutinizing the different reasons for federalist plurality for their possible implications for choice of law.
The most obvious explanation for decentralized government and decentralized law-making is based on fundamental principles of democracy.56 The voice of a single citizen is the stronger the smaller the entity within which it is exercised. At the same time, many questions handled by governments are exclusively or overwhelmingly local in their implications, so that participation by others than local citizens in the rule-making process would dilute the votes of the locals for no reason. Democratic participation and self-rule thus seem to be enhanced by delegation of power to local entities. This way, the local citizenry may enact its preferences, which may differ from those of the larger population, for local questions. Of course, there is limits to how small the local entities can be, or, rather, to which questions they should handle. The smaller the entity, the bigger the probability that any given question will have relatively large implications outside of the local realm, or that addressing the question will necessitate the use of resources outside of the state. But the basic argument remains: Local questions should be submitted to local government in the name of democracy. Which questions are considered "local" instead of "federal" is a question of vertical distribution of power.
For the horizontal distribution of power, i.e., choice of law, the question to be asked from the perspective just presented seems to be simply : To which local entity does the question belong? In other words, the local governance argument for federalism seems to call for localizing the legal dispute - unfortunately, the answers to be given to this question have been disputed ever since Savigny discovered it in the 19th century. Savigny was the first to frame choice of law as the quest "[t]o discover for every legal relation (case) that legal territory to which, in its proper nature, it belongs or is subject (in which it has its seat)."57 Similarly, the dominant school in contemporary European choice of law scholarship looks for "the closest relationship" of the question with one state or another.58 The failed First Restatement, although much more formalistic than modern European choice of law, might also be counted in this corner. One look at the widely differing answers to specific choice of law questions that the different legal systems have come up with in different times should be enough to show that the localization question does not seem to have a definite answer. That is not to say that framing choice of law as a localization problem is meaningless. At the very least, it might be a different focus than interest analysis, and it definitely points towards neutral choice of law rules.
But the democracy explanation for local government might have a slightly different implication for choice of law. As said above, local government is (partly) about enacting local preferences into local policies for local questions. A federal system should make sure that, as a general rule, local policies are followed where a case has implications for the locality, and that other localities do not interfere and frustrate these policies. So, where only one state's policies are affected by the decision in a given case (this resembles the "false conflict" in Currie's theory and terminology), this state's law should govern.59 The problem with this is that choice of law problems more often than not implicate more than one state's policies, if these policies are properly understood60 (this resembles the "real conflicts").61 The solution might be to apply the law of the state whose policies are the most strongly affected by a given case (and, in extension, the incentives following from deciding the case one way or another), or, inversely, whose policies would be most impaired by applying another law. This methodology to choice of law in a federal system has been proposed as early as in 1963 by Baxter, and has since been known as "comparative impairment" analysis.62 The theory seems plausible, but it should be clear that it is very hard to implement in practice. How do we know which state's policy is more impaired ? Attempts have been made to formulate precise rules under this approach, and the results look strikingly similar to contemporary European choice of law rules which are based on the "closest relationship" approach.63
A related reason for decentralized law-making is to adapt rules to local conditions.64 It is related to the democracy argument in that it, too, aims at providing a closer fit of rules with local circumstances. Unlike the democracy argument, however, the local circumstances it aims at are not the preferences of the local citizens, but rather the local empirical data that go into formulating appropriate rules. Differing climates, landscape, economic situation, and so on, all call for differing rules fine-tuned to the respective circumstances. Theoretically, this could be done by a centralized rule-maker with an open eye for local differences. But it seems that local governments and voters are better positioned to assess local facts than even the most diligent central government could ever be. The idea is old; Richard Posner has given it the catchy name "comparative regulatory advantage."65 For choice of law, it means that a case should be governed by the law of the state with the regulatory advantage, i.e., who is in a better position to assess a case of the given kind.
Again, the approach seems plausible, but it is by no means clear cut. In an automobile accident case, for example, one might think that the state where the accident appears knows its roads better66 - but the state where the driver is from might know its drivers better, and the state where the car is registered probably knows the condition of its cars better. And that is not even all the states with a potential regulatory advantage in a case like this. Recovery for tortious injury might be seen as a protection of the victim against being left a cripple with no resources to support oneself (after the accident occurred, i.e., not from a deterrence perspective). It seems that the state best positioned to judge whether and to what extent such support from monetary damages is necessary is the state of residence of the victim, which would also be the state to pay disablement pensions and the like. The overall assessment of which state enjoys the comparative regulatory advantage would have to be made depending on what facts go into forging rules in a particular area. In conclusion, whereas regulatory advantage can be an element in choice of law theory, it provides no abstract answers.
Another reason for decentralized lawmaking could be that it has important advantages for regulatory innovation.67
Legislatures, ideally, strive to develop the best possible solutions to the problems of their state. A multitude of state legislatures is more likely, it is assumed, to come up with, and, most importantly, test a good idea on a smaller state scale than a single central legislature could for the nation as a whole. If the idea is successful, other states or the federal legislature can then copy it. Thus, the states function as regulatory laboratories.68
As to why the states legislatures would strive to develop and implement good solutions in the first place, much discussion has been devoted in more recent times to the argument of regulatory competition. The argument is highly controversial, and many commentators think that "regulatory competition" is but a pseudonym for a race to the bottom.69 States, the argument assumes, compete for citizens and businesses (most importantly, but not only, in their capacities as tax-payers) like commercial enterprises compete for clients. They do this by trying to offer the best possible package of state activity and regulation.70
This involves both finding the best solution as judged by shared standards, and catering for the tastes of different groups of society. Competition through offering "objectively" superior solutions furnishes an alternative explanation (other than benevolence of legislators) why states would strive to find good solutions to problems of their state in the first place, and why they would copy better solutions of other states. Trying to attract a particular "client group" is to some extent related to the point made above that federalism enables legislation to be adapted to the preferences of the local population. The local population is not static71 - people may move to states whose policies are aligned with their preferences, and states may actively seek to attract such immigration, with different states attracting different groups.
As with all competition, it seems that regulatory competition can only work if externalities are kept to a minimum. To the extent a state can reap the benefits of its policy while shifting the costs to a "competitor", its success will not be the result of it offering a better "product", and "competition" will not bring about the best "product". The test of new regulation in a state "laboratory" might be thwarted by such positive or negative externalities. In many cases, a race to the bottom might result - the target of the critique against the idea of regulatory competition. Of course, some areas of legislation are more likely to give rise to significant externalities than others. In a federal system, many or most areas likely to produce strong externality problems are confined to the federal legislature. Where state legislatures remain competent, it seems that the authority of the states vis-à-vis each other should be allotted such as to minimize externality problems on the margins. This would be the task of choice-of-law: horizontal allocation of rule-making authority such as to minimize externalities from state rule-making.72
The application of this yard-stick for choice-of-law rules will often be extremely difficult, and may even lead to an irresolvable tie. Consider the case of a resident of state A entering into a contract in state B, when this person is considered a minor under state A's law, and of legal age under the law of B. Applying state A's law to the question of legal capacity would add the benefit of comprehensiveness in protection of minors to state A's product. The cost of "producing" this product, however, would most likely be borne mostly by the residents of state B, leaving state A's product unaltered, and making B's less attractive. The cost results from the increase in the risk for people entering into transactions of finding themselves with empty hands against a minor, or the increase in monitoring required by the addition of a new group of minors. People entering into transactions in state B will accordingly have to marginally increase their prices. Some customers might decide to rather enter into a transaction in state C (where there is less state A residents). Assuming that most people paying those higher prices will be state B residents, and most people losing customers as well, state B residents bear the cost, and state B cannot deliver as good a product in this regard as it could without minors from A entering into state B transactions. Applying state B's law, on the other hand, would give B the benefit of being able to deliver the "product" of safer, and therefore cheaper, transactions. But it would shift the cost of burdening a person needy of protection with legal obligations to a resident of another state.
However, there are areas where it seems that avoiding externalities could be a useful guiding principle for choice of law. One of them is product liability. Whatever theory of tort law one may adhere to, it is undeniable that stricter products liability comes at a price - the higher demands on making a product safe will be factored into the retail price of the product. Each state may experiment with which combination of safety, compensation for injury, and prices it finds suitable, and, following the argument of regulatory competition, citizens may then choose between the different options offered by different states. State competition on this level is distorted if a state can offer its citizens the benefits of high safety and compensation for injury under one state's strict products liability rule, while making them pay the low prices resulting from another state's lax products liability rule. Stated the other way around, state competition is distorted if one state's citizens have to pay the high prices resulting from a high protection rule, while getting only low product safety and low compensation under another state's low protection rule. This kind of distortion would arise if products liability were governed by the law of the state of the place of manufacturing. While theoretically every buyer could inquire into the origin of the product, and the products liability regime hence applicable, so that the products liability regime would be factored into the price, transaction costs would be prohibitive. In fact, the high transaction cost of explicit contracting on questions concerning damages caused by faulty products, and the ensuing potential for a market for lemons, are the reason for products liability regimes in the first place. Since in practice an individual buyer cannot obtain, lest process, the necessary information on product liability to factor it into his buying decision, products from a state with a lax products liability regime would fetch the same high price as other states' products, while offering only low safety and potential compensation. Manufacturers, on the other hand, are repeating the same transactions over and over and are therefore much more likely to be in a position to ascertain products liability cost. If the law of the market where the sale was made governs product liability, prices will then automatically reflect the applicable products liability regime. Competition works. The capacity of the manufacturer to factor in products liability into its prices is what sets products liability apart from the problem of legal capacity discussed above. A manufacturer of state A held to the stricter products liability standard of state B where the product was sold, is not suffering a set back (in ex ante perspective), since it was able to be compensated for the risk of liability through higher prices. There is an important caveat, however: If arbitrage is possible, and if we do not want to, or cannot, link the applicable law to the place where the arbitrageur bought the merchandise (because this would recreate the information problem for the consumer), a producer cannot sustain higher prices in the state with higher products liability.73 In this case, strong harmonization of products liability law, or restricting the flow of the product from one state to the other, are the only solutions - the latter being precluded in an internal market.
The idea of state competition at first glance seems to be at loggerheads with the use of an immutable citizenship (like the national citizenship in the EU), or a semi-immutable domicile, as relevant factor for choice of law. States can compete for citizens only in so far as choice of law allows them to apply their laws to their residents (or in other cases where citizens seek to benefit from the state's laws, for example by going there for shopping). If immigration to another state does not change the applicable law on some point, because the applicable law is fixed in function of an immutable citizenship, the states cannot compete on this point.74
At closer look, the use of an immutable nationality criterion might encourage competition, rather than prevent it. If all applicable law, like family law, inheritance law, child support law, and so on, changes with a change of residence, then citizens can only get a full-package-deal. If, on the other hand, some questions remain attached to citizenship, they can choose to enjoy the commercial law, tax law, and so on, of another state, while preserving the more personal legal environment that they are attached to.
Full competition in the sense of optimal "consumer" satisfaction would require that citizens be able to choose whether they want to preserve certain aspects of their old home state's law, and which, or not. If such freedom of choice would be socially acceptable would depend on the interrelationship of different areas of law75, and on the normative basis of family law and others76. This demonstrates not only the problematic of the regulatory competition model. It also points to a central dilemma of choice-of-law in general, and of legislative diversity in a federal system in particular, which is that distinct, but related questions are subjected to different states' rules, despite the fact that within one legal system, all these rules are, or should be, adapted to one another. Finally, it drives home the point made earlier that non-discrimination can be a very tricky issue in choice-of-law in a federal system. These are all fascinating questions. But since we are here mainly concerned with the impact of the internal market on choice-of-law in a federal system, and the areas where the citizenship criterion is used are almost exclusively areas only loosely related to market activities, we will not try to develop the answers to these questions.
Exemplifying, clarifying, and expanding on, what we have learned so far, we can now quickly turn to some elements of traditional choice-of-law theory that should be banished from choice-of-law in a federal system. All rules with a substantive bias have no place in a federal system.77 The basic reason is that diversity as such is not a bad in a federal system; rather, it is precisely what federalism is all about. The preceding section canvassed the reasons why such diversity is desired.
Consequently, the forum law approach, i.e., simply applying the forum's law to any legal question presented to the forum, is not an option in a federal system. Whereas an independent state might take the stand that it would be ideal if all people and states in the world lived by its laws, which it considers to be the best for everybody, and whereas such an independent state might therefore apply its own laws to all cases it gets its hands on, such a stance seems to be inconsistent with a state's integration into a federal system. This is not to say that the application of the forum's law as, and because it is, forum law is excluded. A neutral rule submitting certain questions to the law of the adjudicating state can be the optimal rule from a state interest maximizing standpoint. Naturally, most procedural questions will be governed by forum law, as tradition has it too. Also, the cost of applying foreign law in form of search costs, and perhaps an imprecise application of the foreign law, can be high, and the benefit from applying foreign law in form of state policy realization marginal. In this case, the alternative of applying forum law might sometimes be the optimal solution for all states - but mind you that if the benefit from applying the foreign law would be high, declining jurisdiction will usually be the better solution to the problem. My guess is that in situations where actors can anticipate the application of forum law by the courts, the negative impact on states' policies can be expected to be substantial. After the event, a race to the courthouse will result if courts apply forum law, and multiple forums are available. In most cases then, the balance will be against the forum law rule.78
Similarly, it is not admissible for a court to select the rule which it considers superior, as suggested by Leflar's "better-law approach." This approach is dubious (mainly under separation-of-powers considerations), to say the least, in the habitual choice-of-law setting. In the federal setting, it is impermissible as a matter of power allocation between the different states and the union - a court of one state, or of the union, cannot rule on the merits of another state's law.79
The more recent history of (European and international) choice-of-law has seen a proliferation of rules providing for the application of whichever rule out of a set of states is more favorable to a certain substantive result.80 These rules favor a substantive result. But in a case where certain states' substantive rules would provide for this substantive result, and others would not, applying the former states' rule potentially thwarts the latter states' policy. As usual, the question should be which state's policy is more strongly implicated in the case, and which state attaches most importance to its policy that is implicated in the case. There is no reason why one or the other policy should be favored. Of course, one can observe certain trends in substantive law (perhaps of all the states) to favor certain parties, for example children. But these trends have limits (otherwise, all the laws would rule for the same result), and a choice-of-law rule which favors the result conforming with the trend overlooks these limits. To say, as Louise Weinberg does, that we generally want to provide injured parties with a remedy against the injurer, and that we should therefore allow the injured party to select the law which is most favorable to recovery, is simply forgetting that we grant a remedy only under certain circumstances, and for a reason.81 Choice-of-law rules with the kind of substantive bias described should thus be eliminated. This is, in my view, recommended for general choice-of-law; it is mandatory for choice-of-law in a federal system.
Lastly, and most importantly, the public policy exception ("ordre public") has hardly any place in a federal system. Under the public policy exception, a court can eschew or modify the application of the otherwise applicable foreign law, if the substantive result of applying that law unmodifiedly is considered too shocking by the standards of the forum. One can argue that the limits of what can be considered shocking are fixed by the federal constitution, which sets limits to all state legislation, and that a state is therefore preempted of adding to these limits of acceptability.82 This argument is dubious, though, because a federal rule setting minimum standards for the internal legislation of a state does not necessarily preempt questions relating to the enforcement of such legislation in other states. Accepting that another state do something is not the same as being obliged to participate in it. But if a state may refuse to participate in enforcing another state's policy, it certainly must not interfere with it either, if the federal choice-of-law system allocates regulatory authority in a case to that other state. All the state invoking the public policy exception may do at the adjudication stage, if at all, is refuse to decide the case - declare itself a substantive forum non conveniens, so to speak. However, the public policy exception may maintain a (very limited) role at the enforcement stage.83
In sum, although federalism theory, not surprisingly, does not present
clear cut solutions to complex conflict of law issues, it can give important
general indications. First of all, choice of law rules in a federal system
should be neutral and without substantive bias. In addition, some
recommendations as to the content of these rules can be derived. Depending on
the federalist argument used, one might recommend a comparative impairment
approach, look for regulatory advantages, or try to avoid regulatory
externalities in so far as possible. The comparative impairment approach and
the desire to avoid externalities seem to be largely overlapping, or even
identical. A regulatory externality necessarily impairs the effectiveness of
one of the implied states' policies. Minimizing externalities should therefore
serve to minimize impairment of state policies. Regulatory advantage could
theoretically point to a different solution in any single case. In general,
however, it would seem that where one state has a clear regulatory advantage,
it is unlikely that strong externalities would result - otherwise, the
regulatory advantage would not be clear.
The preceding discussion also suggests that, and why, choice of law does matter.84 Does it get the attention it deserves, and the content it needs, in the EU and the US ?
37 Throughout the text, "federal government" will refer to the European institutions and/or the US federal government, respectively.
38 Throughout the text, "states" will refer to the entities making up the union: Member States in the EU, and States in the US.
39 The USA is without a doubt a federalistic state. The EU is perhaps not properly called a federalistic state, but in view of its distribution of powers it can be called a federalistic structure, or analyzed as such, see, e.g., Hans-W. Micklitz & Stephen Weatherhill, Federalism and Responsibility, in Federalism and Responsibility 1-2 (Hans-W. Micklitz, Thomas Roethe & Stephen Weatherhill eds. 1994); Bernd Martenczuk, Die differenzierte Integration und die föderale Struktur der Europäischen Union, Europarecht [EuR] 2000, 351, 355-356 (with further references); cf., even before the creation of the EU, and at the time of the adoption of the Single European Act (1986), Daniel J. Elazar & Ilan Greilsammer, Federal Democracy: The U.S.A. and Europe Compared - A Political Science Perspective, in Integration Through Law, supra note 8, vol. 1 bk. 1, 71, especially 92-112; Joseph H.H. Weiler, The Future of the European Community in the Light of the American Federal Experience (1986).
40 This section will thus try to answer in policy terms the question that Justice Robert H. Jackson had left open in 1945 in terms of positive law, citing lack of case law: "The ultimate answer [to the choice-of-law problem], it seems to me, will have to be based on considerations of state relations to each other and to the federal system. What is the basis of power in a constituent state of our federation to govern a controversy, when is it exclusive of a like power asserted for another state on the same or different basis, and when is it entitled to prevail even in the forums of another state?", Full Faith and Credit - The Lawyer's Clause of the Constitution, 45 Colum. L. Rev. 1, 28-29 (1945). The answer given here is incomplete, however, since it does not take into account the specific arrangements of a particular federal system. Such arrangements might include, for example, a constitutional recognition of the idea of state competition, or a clearer definition of the personal attachment of a person to her home state, and the resulting personal jurisdiction and responsibility of that state.
41 The notion of neutral choice-of-law norms corresponds to the Continental European notions of "allseitige Kollisionsnorm" or "règle de conflit multilatérale."
42 Brilmayer, supra note 13, ch. 4; Kramer, More Notes, supra note 13, 272-275, and Uniform Code, supra note 29, 2140-2146; cf. Trachtman, supra note 18, 1036-1040, 1047-1048. The following discussion summarizes their arguments. The nucleus of the idea can already be found in Baxter, supra note 15, 24-25, 42. Even Currie acknowledged that the broadest possible application of forum law would "be rational ... in the sense that, in the short run, without considering how other states and higher authority might react, the state would in this manner be doing all it could to maximize its own interests." (Currie, supra note 19, 89) (emphasis added).
43 Cf. Kozyris, supra note 12, 514-515; Trachtman, supra note 18, 1045-1046, 1052. Louise Weinberg's polemic (Choice of Law and Minimal Scrutiny, 49 U. Chi. L. Rev. 440, (464: "[T]he American tort plaintiff ... will not be defeated by territorial limits on state power.") (1982); Against Comity, 80 Geo. L. J. 53 (1991)) in support of unilateralism in choice of law overlooks that a general principle like "A party sustaining an injury is entitled to compensation from the injurer." is not the equivalent of a principle that anybody claiming to have sustained an injury is entitled to compensation regardless of the circumstances under which the injury was sustained. Different states draw the lines between compensating the victim, and limiting the freedom of potential injurers, differently, and there is no reason why the state drawing the line most in favor of the victim should impose its policy on other states. Cf. Jan von Hein, Das Günstigkeitsprinzip im Internationalen Deliktsrecht (1999).
44 Baxter, supra note 15, 23. Cf. Henry M. Hart, Jr., The Relations Between State and Federal Law, 54 Colum. L. Rev. 489, 515 (1954).
45 See supra II.C, p. 6.
46 Cf. Peter Hay, Full Faith and Credit and Federalism in Choice of Law, 34 Mercer L. Rev. 709, 722-727 (1983).
47 Cf., e.g., Laycock, supra note 23, 259-261; and for the Australian Commonwealth Breavington v Godleman (1988) 168 CLR 41, 77-78 (Mason, C.J., dissenting).
48 Cf. David Shapiro, Federalism - A Dialogue 44-50(1995).
49 And Baxter did so argue, of course.
50 Cf. Aaron D. Twerski, On Territoriality and Sovereignty: System Shock and Constitutional Choice of Law, 10 Hofstra L. Rev. 149, 159 (1981); and for Australia: Breavington v. Godleman (1988) 168 CLR 1, 111-116 (per Brennan, J.); McKain v R W Miller and Co (South Australia) Pty Ltd (1991) 174 CLR 1, 36 (per Brennan, Dawson, Toohey, McHugh, JJ.).
51 Cf. Linda Silberman, Can the State of Minnesota Bind the Nation?: Federal Choice-of-Law Constraints After Allstate Insurance Co. v. Hague, 10 Hofstra L. Rev. 103 (1981).
52 Cf. infra VI.D.2.a), p. 39.
53 Cf., e.g., Hart, supra note 44, 514; Laycock, supra note 23, 259.
54 Baxter, supra note 15, 23; Hart, supra note 44, 515; Hay, supra note 46, 727; Harold W. Horowitz, Toward a Federal Common Law of Choice of Law, 14 U.C.L.A. L. Rev. 1191, 1199-1207 (1967); Laycock, supra note 23, 333-334; cf. Jackson, supra note 40, 26; Roosevelt, supra note 33, 2504; even Currie, supra note 19, 179 and 183 (reprinted from Notes on Methods and Objectives in the Conflict of Laws,  Duke L. J. 171) but see Friedrich K. Juenger, Tort Choice of Law in a Federal System, 19 Sydney L. Rev. 529, 536-537 (1997); Russell Weintraub, Who's Afraid of Constitutional Limitations on Choice of Law?, 10 Hofstra L. Rev. 17, 25 (1981).
55 And they might also reinforce my (and others') earlier claim that state interests matter in choice-of-law.
56 Cf. Shapiro, supra note 48, 91-94; Advisory Commission on Intergovernmental Relations, The Condition of Contemporary Federalism: Conflicting Theories and Collapsing Constraints ch. 1-II (1981).
57 Friedrich von Savigny, A Treatise on the Conflict of Laws § 360 (p. 89) (William Guthrie trans., 2nd ed. 1869) (original: System des heutigen römischen Rechts, vol. 8, 1849).
58 The standard reference is Lagarde, supra note 13.
59 Cf. James R. Pielemeier, Why We Should Worry About Full Faith and Credit to Laws, 60 S. Cal. L. Rev. 1299, 1330-1334 (1987).
60 Cf. supra note 43, and accompanying text.
61 Cf. Pielemeier, supra note 59, 1332, 1335.
62 See Baxter, supra note 15, 1-23.
63 Compare the rules suggested by Laycock, supra note 23, 322-331, with current European practice.
64 Cf. Sproles v. Binford, 286 U.S. 374 (390) (per Hughes, C.J.).
65 Posner, supra note 24, § 21.15.
66 For the lex loci delicti therefore Posner, id., and Allen & O'Hara, supra note 29, 1043-1044.
67 See generally Shapiro, supra note 48, 76-91.
68 But see Susan Rose-Ackerman, Risk Taking and Reelection: Does Federalism Promote Innovation?, 9 J. Legal Stud. 593 (1980).
69 In its
(highly controversial) Centros decision of 1999, the ECJ quite
explicitly approved of the idea - but the discussion is certainly far from over
in Europe and elsewhere. See ECJ Case C-212/97, Centros Ltd ./.
Erhvervs-og Selskabsstyrelsen, 1999 E.C.R. I-1459, para. 27; and from
the numerous comments Eva-Maria Kieninger, Note, Zeitschrift fuer Unternehmens-
und Gesellschaftsrecht [ZGR] 1999, 724; Wulf-Henning Roth, "Centros": Viel
Lärm um Nichts?, ZGR 2000, 31. The decision is set in the area of
corporate law, and, although one would not think so when reading the decision,
this is the area where the debate on the concept of regulatory competition
seems to be most advanced, due to the Delaware phenomenon. See,
e.g., Roberta Romano, The State Competition Debate in Corporate
Law, 8 Cardozo L. Rev. 709 (1987); Lucian A. Bebchuk & Allen
Ferrell, A New Approach to Takeover Law and Regulatory Competition, 87
Virg. L. Rev. 111; both with numerous further references.
This is, by the way, a nice example that choice of law rules do matter - be it a race to the bottom or a race to the top, in Europe the race was called off before the start by choice of law rules in most Member States submitting corporations to the law of the place of their main establishment.
70 See, e.g., Richard Greve, Real Federalism, ch. 1 (1999), and the extensive discussion by numerous contributors in Competition among States and Local Governments (Daphne A. Kenyon & John Kincaid eds. 1991).
71 Provided, and to the extent that, the free movement of persons, including the right to take residence and employment, is guaranteed, and migration is not otherwise impeded, for example by language barriers.
72 Cf. Posner, supra note 24, § 25.8; Allen & O'Hara, supra note 29, 1024-1027; Trachtman, supra note 18, 985-988, 1032.
73 Cf. Bruce L. Hay, Conflicts of Law and State Competition in the Products Liability System, 80 Geo. L. J. 617 (1992); Posner, supra note 24, § 25.8.
74 Of course, one can derive from this that use of citizenship as a criterion for choice-of-law is or should be impermissible in a federal system only if, and to the extent that, one subscribes to the idea of state competition. The idea of state competition is, as said above, far from being universally accepted, least of all for areas like family law where most uses of the citizenship criterion for choice of law can be found. Cf. also Trachtman, supra note 18, 1030-1031.
75 E.g., can one sensibly separate social security law and family law?
76 E.g., is family law a set of default rules for the benefit of the individuals concerned, designed to prevent some family members like parents from putting burdens on third parties by, e.g., abandoning their children to society, or a sort of legitimate moralizing by society ?
77 Another important set of rules that are unacceptable in a federal system are of course rules that discriminate on the grounds of nationality, citizenship, or residence. See supra II.B, p. 4.
78 Cf. Allen & O'Hara, supra note 29, 1023; and generally Kropholler, supra note 16, § 7, discussing, inter alia, the possibility of severely limiting adjudicative jurisdiction, such that applying only the forum law would be more palatable (lex fori in foro proprio).
79 Cf. Jackson, supra note 40, 27-28; Kramer, Uniform Code, supra note 29, 2139-2141..
80 See generally, e.g., Kropholler, supra note 16, § 20 II.
81 See supra note 43.
82 Larry Kramer, Same-Sex Marriage, Conflict of Laws, and the Unconstitutional Public Policy Exception [hereinafter: Public Policy Exception], 106 Yale L. J. 1965, 1986-1987 (1997).
83 Provisions like Art. 27(1) of the Brussels Convention (soon to be replaced by Art. 34(1) of Council Regulation 44/2001/EC, 2001 O.J. (L 12)) are thus acceptable.
84 Also cf. supra note 69.