Whincorp, Michael --- "Three Positive Theories of International Jurisdiction" [2000] MelbULawRw 14; (2000) 24(2) Melbourne University Law Review 379

Three Positive Theories of International Jurisdiction

MICHAEL WHINCOP[*]

[For what reasons and under what circumstances will a state permit its courts to exercise jurisdiction over cases with international elements? Three positive theories provide contrasting answers. The first is a utilitarian theory, in which the calculus of the utility to be maximised may vary in its scope. The second is a deontological theory, in which courts exercise jurisdiction only in circumstances in which it is fair to the defendant to do so. The third is a political theory. It asserts that politicians will design principles of jurisdiction to maximise their utility, which is likely to involve responding to the rent-seeking demands of well-organised interest groups. Doctrinal evidence of suggestive value is examined. This paper suggests that a sufficiently broad definition of utility substantially equates the utilitarian and deontological theories.]

INTRODUCTION

The law on the international jurisdiction of courts in private law is a little bit like fishing restrictions. They determine what you can catch, and what you have to throw back into the water. Fishing restrictions are passed for important reasons. The most obvious is to ensure a balance between current and future consumption. On the other hand, as we know, some countries would, if they could, permit Australian territorial waters to be fished until they were empty. Related, but not quite the same, is a claim that the rules are passed to protect fish and other marine animals. You can’t catch a dugong or dolphin and carve it up for bait or for a Scouts’ barbecue. The cynics might even say that fishing restrictions are passed to make the fishing industry better off by restricting supply and increasing prices.

I hope that I will not offend private international lawyers by suggesting that these analogies are a useful place to start in thinking about why the law allows you to net some defendants, but forces you to throw back others. There is a utilitarian rationale, designed to increase aggregate welfare of those interested in the waters; a deontological rationale, designed to protect defendants against unfair netting; and a political rationale, designed to advance the interests of a subset of those who ply the waters, often at the expense of the others.

At this point, a lawyer might interject, ‘Your metaphor is wearing a bit thin — tell me why I should be interested in these explanations. I would prefer to content myself with knowing what the High Court says about forum non conveniens, or service out of the jurisdiction.’ Quite apart from the pragmatic value of a positive theory of law, which includes prediction of doctrinal trends, we currently find ourselves caught up in a transition of method in the law on jurisdiction. Australia has outgrown the rigid jurisdictional approach of the earlier 20th century, and has followed, a little diffidently, the bolder steps of the English courts. In doing so, our approach to jurisdiction has changed quite markedly, from narrow rules to broader standards. On the other hand, the Europeans have done the opposite, committing themselves to quite precise rules in conventions on jurisdiction. Australia now faces a choice as to its response to a possible rule-based approach to jurisdiction emanating from the Hague Conference on Private International Law.[1] Positive theory helps us to understand why certain choices might be made, what the content of those choices might be, and the effect of particular changes.

Initially, I describe each of the theories, and then return to examine some suggestive doctrinal evidence in favour of each. The evidence is drawn primarily from United States and Commonwealth law on the establishment and exercise of jurisdiction. I also discuss developments in the recognition of judgments, since that area is inextricably tied to the law on jurisdiction. Although this evidence can be marshalled to test the three theories, that is not my immediate purpose in this article. I seek, first, to point out a path for a more theoretically rich scholarship of international jurisdiction. Second, I show how the apparently Manichean opposition of these theories frequently gives way to mutually supportive explanations of doctrinal phenomena — something which can be an obstacle to empiricism. Third, I seek to argue, in respect of deontological and utilitarian hypotheses, that this is no mere coincidence, by outlining a metathesis that the efficient jurisdictional solution in a multi-state case is also the fair solution.

II A UTILITARIAN THEORY OF JURISDICTION

A The Calculus of Utility

A utilitarian theory of jurisdiction asserts that legal principles that govern the jurisdiction of a court and their exercise are selected to maximise social welfare.[2] What principles make for the ‘greatest good’ of the ‘greatest number’? To answer this question, we must address two issues. The first is how jurisdiction principles affect social costs. The second is who bears those costs, and the means by which they are aggregated to form the relevant social welfare function.

1 Primary Behaviour

Jurisdictional principles can affect social costs by influencing the behaviour of actual or potential litigants. A logical means of dividing up a discussion of behavioural influences is to distinguish between primary behaviour and litigious behaviour. The latter involves all choices persons make in deciding whether to litigate a dispute and the means by which to do so. Primary behaviour includes all forms of behaviour that might result in a litigated dispute. It includes entering and performing contracts, taking safety precautions in the conduct of a business, and so on. The law on jurisdiction may affect primary behaviour if legal liability deters or encourages primary behaviour, and the expected liability associated with primary behaviour varies between forums. Legal actors may respond to expected liability differentials by relocation or by adjusting care or activity levels.

Relocation involves changes to the association between persons and jurisdictions, such as reincorporating a corporate defendant in another state. That tactic may reduce liability if jurisdictional principles provide, for example, that a corporate defendant can only be sued in the forum in which it incorporates. A narrow principle to this effect would give defendants the capacity unilaterally to ‘judgment-proof’ themselves if suit in certain forums was impossible. An obvious problem occurs when a person relocates to forum X but the effects of that person’s actions continue to occur in forum Y. Relocation enables that person to externalise the costs of his or her action that Y’s law would otherwise require that person to internalise. Call this the problem of relocation externalities.

Where liability differs between forums, either in quantum or in the degree of culpability required, this may influence the level of an activity a person engages in and the care the person takes in doing so. However, jurisdictional principles can only have these effects if rational defendants expect to be subject to liability rules having these effects, which is in part a consequence of how widely judgments can be enforced and the forum’s choice of law rules.

Relocation externalities and adjusting care or activity levels have completely different implications for jurisdiction. The former simply counsels some form of mechanism analogous to the effects doctrine of public international law (which recognises legislative competency to address behaviour affecting a state).[3] An obvious example is providing the means to serve defendants who relocate. The latter has no implications. Jurisdiction could only optimise care and activity levels if laws could pre-commit to conferring jurisdiction on a state applying the laws that were efficient to particular cases, and defendants could have reason to expect these applications at the time of choosing care and activity levels. Doing the former would require unanimity about the instrumental effect of legal rules — something tort scholarship is no closer to after decades of analysis, and which is even less likely among forums than among scholars. Mary Keyes and I have rejected the proposition that choice of law principles can normally do the latter.[4] Attempting this at the jurisdiction level, at a further remove from substantive law, seems hopeless.

2 Litigation Costs

The law on jurisdiction influences the social costs of litigation, since it determines the forums in which actions may proceed. The optimal conduct of a case may vary between forums. If litigation costs vary between forums, a utilitarian theory would seek to choose the forum in which costs were minimised.

First, costs are likely to be positively related to the proximity of the parties’ normal place of abode to the forum. Litigants and their attorneys will have higher travel costs when involved in a suit in a distant forum. Additional lawyers may have to be instructed locally and in the forum, which imposes additional agency costs in monitoring and controlling them. Second, costs are likely to be positively related to the proximity of the forum to the cause of action and evidence relating to it. This is obviously linked to travel costs but also to the scope of the coercive power of the court. Litigation costs are lowest when a forum does not depend on cooperation of other states and forums. Call these cost influences party proximity and case proximity.

Third, litigation costs are influenced by settlement transaction costs. The civil procedure system can be conceived as a device for enabling parties to settle their case. Settlement is influenced by information asymmetries and differences in optimism between the parties.[5] Settlement transaction costs are therefore the direct costs of complying with steps in civil procedure, the social costs of delay in congested forums (for example, stale evidence), and the transaction costs of settlement bargaining. The steps that need to be taken to settle are partially endogenous to the civil procedure rules themselves. Some rules can encourage unnecessary litigation; others may unnecessarily protract it. This is because civil procedure rules affect, for example, the distribution of information (for example, discovery rules), the stakes of winning and losing (for example, costs), and the incentives to bargain or litigate strategically.[6]

Fourth, costs may vary between forums given the application of particular substantive law. Proving foreign law, in the absence of full faith and credit, is famously awkward.[7] Particularly where the parties have expressly chosen a state’s law in their contracts, selecting that state as the forum may be efficient, since that country’s law is likely to be given fullest and most accurate effect by its own courts. Call this cost influence foreign law costs.

Fifth, there may be scope economies in litigation. Some parties may be repeat litigators — typically defendants. Businesses may regularly cause injuries that are the subject of frequent suit. The defendant’s place of business is likely to economise on litigation costs across these various suits. Similar economies arise in class actions, or where a defendant seeks contribution from a third party.

Sixth, forums vary in enforcement costs.[8] The scope for the enforcement of foreign judgments in jurisdictions in which the defendant has assets affects the marginal cost of litigating in forums in which he or she does not have assets — the wider the scope, the lower the enforcement costs.

3 Costs to States and State Interaction

States subsidise litigation by funding the justice system, and thus bear costs by exercising jurisdiction. Thus, the law on jurisdiction functions to ration access to that system where the costs to the state of trying a suit are high. Although the fixed cost investments in the justice system complicate estimating these costs, the state’s marginal cost is driven by similar factors to those influencing party costs. Party proximity, case proximity and enforcement costs affect the court’s capacity to enforce its own process. The court also bears foreign law costs, both because of the complex procedure for proving foreign law and because decisions under foreign law do not create binding precedents (precedents being public goods).

The litigation costs of both parties and states depend on how states resolve overlaps of jurisdiction when litigation is commenced in multiple forums. The worst case scenario is one in which both courts persist with the litigation, effectively doubling the litigation burden. The reverse problem is that both courts wash their hands of the litigation. If this result is anticipated — or engineered (for example, by relocation) — by a party, it will affect primary behaviour in a socially costly way. Even if courts share the desire that one, and only one, court resolve the case, there can be co-ordination problems in ascertaining which court that will be. Each of these problems can be fitted into a game theoretic paradigm. The relevance of a utilitarian theory is that it shows that the incidence of these games turns on how courts or law-makers purport to measure social welfare. I demonstrate these things in the following section.

B Measures of Social Welfare

In a positive utilitarian theory of jurisdiction, legal rules on jurisdiction are designed to maximise social welfare. Even if one assumes, as I do, that welfare is measured in terms of wealth, we have not yet answered whose interests count, and how they count, in the aggregation of social welfare. In the subject of interest to us, these definitions can vary in two ways. First, a forum may define welfare unilaterally. If it does, only the welfare of interests associated with the forum counts (for example, those of residents and domiciliaries). Second, states may define welfare in terms of the costs that the state bears, and not those of the parties.[9] These differences alter the pay-offs from exercising or declining jurisdiction, which can change the character of strategic interactions between states — and thus alter the profile of the law.

In explaining these effects, I will use an example, which I vary from time to time. In this example, P has a meritorious case against D, which may be pursued in either forum Y or Z.

matters, assume that if the matter is tried, each forum will grant the same award, A. Assume initially that the grant of the award is socially beneficial, and that the social benefit, B, is equal to A. The plaintiff incurs costs, P, to prosecute the case; the defendant incurs costs, D, to defend; the state incurs costs, S, to try the case. Let these variables take the following values: A = 15; S_Y = S_Z = 1; P_Y = 4; P_Z = 6; D_Y = 6; D_Z = 7.[10]

1 Minimising the Costs of Judicial Administration — A Prisoner’s Dilemma

At the limit, minimising the costs of judicial administration would be inconsistent with the very existence of a justice system. There would then be no principles of jurisdiction to worry about. Nonetheless, minimising the costs of judicial administration may be relevant to international cases, because there is another forum to which the local forum may attempt to off-load the litigation. As I have mentioned, this is particularly so where rights depend on foreign law, since no public good is created by the case’s resolution.

If Y and Z seek to minimise their costs of judicial administration, neither exercises jurisdiction, since each prefers a pay-off of 0 to -1 (the value of S in each forum). The social benefit from both courts’ declining jurisdiction, W₀, is 0. Let aggregate social welfare, W_F, from the exercise of jurisdiction be defined such that W_F = B – (S + P + D). Thus, if Y exercises jurisdiction, W_Y = 4.[11] If Z exercises jurisdiction, W_Z = 1.[12] Resolving the case by limiting social welfare to the costs of judicial administration is the least desirable result (4 > 1 > 0). If all courts attempted to decline jurisdiction over costly international cases to economise on costs, the residents of all countries would be worse off. Where a strategy is in each individual’s best interest but makes the individual and the other player(s) worse off collectively, game theorists are apt to talk of a prisoner’s dilemma. The game takes its name from a situation involving two suspected criminals, against whom there is sufficient evidence for light convictions to be returned. They are offered, separately and without the opportunity to form a pact, a deal that if one confesses and the other remains silent, the former will get a pardon and the latter a very heavy sentence. If both confess, each will receive a sentence intermediate between the sentence for the light conviction and the very heavy sentence. Although mutual silence maximises collective welfare (by minimising the sum of the sentences), each is better off confessing. If their partner in crime stays silent, they are better off taking the pardon by confession than the light conviction. If their partner confesses, it is better to take the intermediate sentence by confession than the very heavy sentence. We have a prisoner’s dilemma here because a policy of minimising the state’s social costs can only increase a state’s welfare if other states are willing to hear these cases, inconsistently with the assumption of minimising the costs of judicial administration.

However, cost minimisation may not always be a symmetrical objective of states, especially when there is variance in the forum’s connections with the case or simply in the costs to states. There will often be domestic pressure in the more connected forum to exercise jurisdiction whatever the other court does. Thus, the positive implication is that cost-minimising strategies are likely to occur in cases where the forum is minimally connected with the suit, or where enforcement costs are high or case proximity is very low.

2 Unilateralism — A Chicken Game and an Alternative Prisoner’s Dilemma

My first example assumed that A_Y = A_Z. This will often be violated because of differences in substantive law, especially if choice of law rules are not multilateral (that is, the proper law is independent of the forum). In these cases, jurisdictional rules may be selected or applied to maximise the welfare of in-state interests. Examples would include principles that privilege the position of local plaintiffs in bringing suits locally, even if the forum is inappropriate to the litigation.[13] Another example is where pleading to be entitled to the benefit of local law provides a justification for the exercise of jurisdiction. This occurred in the decision of the High Court of Australia in Akai Pty Ltd v People’s Insurance Co Ltd,[14] where an insured was allowed to claim rights under Australian legislation, notwithstanding a choice of English law and forum. Where the law vests discretion in judges to determine whether or not to decline jurisdiction, the factors that influence litigation costs explored in Part II(A) above would not be expected to feature consistently, if judges decide jurisdiction disputes unilaterally. Judges would only rely on them if they favoured a citizen plaintiff who had made an efficient forum choice, or a citizen defendant resisting an inefficient forum choice by an alien. They would reject their relevance in cases where citizen plaintiffs selected the local forum when it did not minimise costs.

Unilateralism encourages forum shopping and creates uncertainty. But there are two more subtle effects. To illustrate the first, I vary the above example by letting A_Y = 15 and A_Z = 0. If P is a resident of Y and D a resident of Z, we have a problem. The Y court will want to assume jurisdiction since, by doing so, P earns a pay-off of 11 relative to a pay-off of 0 in Z (it is 0 because P would choose not to litigate in Z). D will therefore want to litigate in Z in order to obtain a negative declaration that he or she has no liability to P. The Z court will indulge D since his or her pay-off in Z is -7 (–(A_Z + D_Z)) and his or her pay-off in Y is -21 (–(A_Y + D_Y)).[15] If both courts exercise jurisdiction, however, the parties must bear costs in both forums. Unless there is a corresponding social benefit from parallel litigation — which seems highly unlikely — the parties (and the states) could be even worse off than they were if neither court exercised jurisdiction. This interaction is strategically identical to the ‘chicken’ game, in which both players drive their cars at each other, and make simultaneous last instant decisions either to swerve to the right or to drive straight ahead. Each player is better off if he or she drives and the other swerves, but both are worst off if neither swerves. Call this the chicken effect. The second effect — I refrain from calling it the egg effect — lies in the unilateral incentives of states to formulate, before any case comes to be decided, jurisdictional rules which are excessively wide in order to favour locals. This is also a prisoner’s dilemma, since each state has an incentive to respond symmetrically with its own excessively wide rules. Mutual restraint in jurisdictional rules would constrain socially wasteful forms of litigation. This form of prisoner’s dilemma is the paradigm for game theoretic analyses of most international law.[16]

3 Maximising Aggregate Social Welfare — A Co-Ordination Game

Principles of jurisdiction may be formulated to attempt to maximise aggregate social welfare. These principles would constrain forums with high social costs, irrespective of whether they were borne by citizens or aliens. But herein lies a problem. Both legislators and courts may find it difficult to identify the lowest cost forum — legislators, because of the case specific influences on costs; courts, because of difficulties verifying comparative costs. This suggests that the strategic interaction partakes of a co-ordination game. A co-ordination game is one where parties are better off selecting co-ordinated strategies (in our example, Y, the lower cost forum, exercises jurisdiction; Z declines).[17] Since there is more than one combination of co-ordinating strategies, co-ordination may never occur. Co-ordination problems can be knotty, and present an occasion for unilateralism at the margins, where forums are equal in their aggregate merits but one forum’s nationals do well if it exercises jurisdiction, while the reverse is true in the other forum.[18] Co-ordination problems are sometimes overcome because particular heuristic solutions are focal — that is, they are cognitively favoured or preferred.[19] Co-ordination can also be solved if states can reach agreement on solutions, provided the implementation of those solutions can be verified.[20] There is an important trade-off here. These problems of addressing co-ordination impose a certain ambiguity on the form of jurisdiction rules that maximise aggregate social welfare. While it may be efficient ex post to give courts a large measure of discretion to ascertain ex post which forum maximises social welfare, very simple rules may actually be more efficient ex ante because they reduce the need for co-ordination.

III A DEONTOLOGICAL THEORY OF JURISDICTION

The standard critique of utilitarian reasoning, synonymous with Kantianism, is that it treats humans as objects or means and fails to respect their autonomy. The concept of justice addresses this problem by recognising rights. This is described as a deontological approach to the law, because it recognises ethically based duties owed by autonomous human beings to each other. More problematic is how to determine the substance of the rights. Although there are many different approaches to this question, I mention here the two most important — the corrective justice and contractarian approaches to rights.

The concept of corrective justice derives from Aristotle, and has been much used by scholars to explain tort law.[21] A party has a right to corrective justice where that person has been injured by wrongful conduct, irrespective of that person’s or the injurer’s character or social status. It is my opinion that the concept has limited value for a deontological theory of jurisdiction. That is because ‘[t]he definition of wrongs is prior to the duty of corrective justice. ... [S]ociety is always free ... to alter the definition of wrongful conduct.’[22] The problem with conflicts cases is that they raise either disparate definitions of wrongs, or disparate measures of the duty to do corrective justice. Corrective justice concepts perhaps favour application of the proper law which is most closely implicated by the relation between the defendant and the plaintiff.[23] They can never support an argument that a court should accept jurisdiction to enable corrective justice to be done, for that presupposes a concluded view of the merits and the conflict.

This very problem recommends a contractarian approach to justice, since it seeks to derive rights from the contract or agreement of persons who may have conflicting notions of the good. Rawls famously argues that just rights and entitlements may be derived from the heuristic device of determining the agreement of parties deprived of knowledge of their position or endowments in society.[24] Lea Brilmayer has used contractarian methodology in private international law to explain the rights of a person to be free from the exercise of power by states.[25] Her argument is principally a response to unilateral utilitarian approaches to choice of law. Nonetheless, the argument has clear jurisdictional implications.[26] Brilmayer defines these ‘political rights’ in terms of the defendant’s relation and connections to the forum. Certain relations or connections cross the threshold at which the defendant may fairly be subject to the state’s law. Consent to the exercise of the state’s law is an obvious example. Brilmayer also considers that legitimacy may attach to the exercise of power against domiciliaries, to action within the state’s territory, and in a manner that balances the potential benefit and burden to a defendant of being associated with a state. Such rights enable parties to know what suffices to ‘enter’ the state in a politically substantive sense, and what they must do to ‘exit’ it.[27]

To generalise my analysis, a deontological approach to jurisdiction cannot begin with substantive judgments about the merits of the cases. Instead, it must focus on the circumstances in which it is fair for a plaintiff to invoke the power of the state to affect the property rights of the defendant. It may draw findings of fairness from the defendant’s relation to the forum or the defendant’s relation to the plaintiff. As regards the relation to the forum, concepts of transient jurisdiction, supported by mere physical presence, are likely to be unfair because of the trivial relation to the forum. Concepts such as domicile, residence, nationality, and carrying on business are less objectionable, because the stronger connections suggest a more balanced match of benefit and burden from association with a state.

The defendant’s relation to the plaintiff can provide considerations that legitimate the exercise of jurisdiction. This is for two reasons. The first is that it may be very difficult to analyse some defendants’ relations to the forum, when those defendants are purely juristic, as corporations are. As I have explained in detail elsewhere, focusing on transactions — that is, relations between plaintiffs and corporate defendants — is not only more consistent with the character of corporations as nexus of contracts but may limit strategic choices of domicile that would generate what, above, I call relocation externalities.[28] The second is the one meaningful contribution of the concept of corrective justice to theorising jurisdiction. Corrective justice is necessarily a bilateral or relational concept, so those jurisdictions foreign to the plaintiff–defendant relation provide a most unconvincing forum in which any duty of correction might be recognised.

A deontological approach to jurisdiction is inconsistent with the exercise of jurisdiction on the basis of attributes of the law pleaded by the plaintiff. To allow the plaintiff to choose the forum by pleading the application of its law is inconsistent with the focus on the defendant. Even to allow the jurisdictional question to rest on the law selected by a choice of law rule presupposes the absence of bias of the choice of law rule, and the fairness and appropriateness of that law in determining the existence of a wrong requiring correction. The fairness of statutory provisions directing courts to apply forum law in particular cases (not uncommon in Australia)[29] continues to depend on the defendant’s relation to the forum or to the plaintiff.

IV A POLITICAL THEORY OF JURISDICTION

A positive political theory of jurisdiction uses public choice theory to explain the exercise of jurisdiction.[30] Public choice dichotomises legal rules into those seeking to maximise wealth and those seeking to redistribute it. Politicians seek to maximise their prospects for election, which they can do by offering their services to ‘rent-seeking’ interest groups which seek wealth transfers and subsidies. The interest groups most likely to prevail in this process are those whose interests are concentrated, rather than diffuse. Competition between these constituencies should limit the social costs of rent-seeking, by limiting the scope of the transfer and reducing its deadweight costs.[31]

Public choice has been used in the analysis of choice of law.[32] The central insight lies in the demand of rent-seeking interest groups, based within the state, for laws which benefit them at the expense of out-of-state persons. The theory is often illustrated by pro-plaintiff products liability awards by state courts in the United States, and the associated use of pro-forum, pro-recovery choice of law methods.[33] The benefits of recovery mostly stay within the state, but the costs are borne by out-of-state manufacturers in proportion to the prevalence of interstate trade. Modern choice of law methods intensify this problem as they may enable a state to keep liability under its own tort rules relatively low, but to use a choice of law rule to invoke the higher awards available under other forum law when out-of-state defendants are involved.

The effectiveness of redistributive legislation depends on the costs to the loser to avoid the effect of the legislation.[34] Laws on jurisdiction interact with choice of law rules to influence these exit costs in various ways. For example, choice of law methods biased in favour of applying local law and jurisdiction rules that predicate the exercise of jurisdiction on attributes of the plaintiff or the applicable law raise exit costs. Laws which are territorial or predicate on defendant attributes lower exit costs. Laws which defer to the parties’ agreement on forum and proper law reduce exit costs further, because physical relocation is unnecessary and because opportunistic forum choice by a plaintiff can be the subject of contract. Federal systems present a special case because they are often structured in a way that minimises discrimination against out-of-state residents (and thus externalities) and that encourages low-cost moving between states. These can encourage beneficial competitive effects.[35]

To develop hypotheses from a positive political theory of jurisdiction, one must specify several parameters. First, are the interest group’s members habitually plaintiffs or defendants? Pro-plaintiff groups (for example, consumer organisations) would prefer to be able to choose the forum unilaterally, but pro-defendant groups (for example, product manufacturers) would like to be able to lock it in contractually. Interest groups associated with these factions may have similar preferences, such as those of ‘plaintiffs’ lawyers’. Lawyers may have unique interests of their own related to their desire for imprecise jurisdictional principles that are resolved litigiously, such as forum non conveniens. In this sense, the complexity of jurisdictional doctrine is a desirable compromise for lawyers, since it balances pro-plaintiff and pro-defendant interests while providing a litigation subsidy for lawyers. Those few states that become jurisdictional ‘capitals’, such as Delaware in corporate law, or Texas in tort, are likely to adopt principles designed to reinforce their leadership. Delaware makes it easy to enter by adopting pro-choice laws; Texas makes it hard to get out by minimising the scope to decline jurisdiction.

Second, preferences for jurisdiction depend on whether the interest group has procured rent-seeking legislation and, if so, in what form. The interest group may seek to eliminate jurisdiction agreements where the law reinforces a cartel arrangement or provides a subsidy in order to protect its rents against competitive dissipation. But that is not true if barriers to entry are naturally high, or if the law favours the interest group through mandatory terms, since it may be beneficial for some members to contract out of the rule.

Third, preferences for jurisdiction differ between the time of seeking redistributive legislation and the time of litigation. Interest groups may not oppose low-cost exit when they lobby for the legislation in order to mitigate the deadweight losses that its victims bear and thus the incentive for them to oppose the legislation. At the time of litigation, they may argue that the legislation intended to limit exit for paternalistic reasons in order to enlarge its scope.[36]

V EXPLAINING THE DOCTRINE

Parts II, III and IV offer analyses of the scope and the exercise or non-exercise of jurisdiction, according to very different positive theories of the law. Ideally, the next step would be to use empirical evidence to discern which theory is supported. Practically, this is not easy to do. First, some relevant phenomena, such as an international convention on jurisdiction and the recognition of judgments, are unique events, which resist aggregation with, say, the law on the interpretation of statutes. Second, small numbers bias statistical tests of significance. Third, the theories I develop are not easily empirically differentiated because, at least some of the time, they may overlap. For instance, the unilateral utilitarian theory offers similar predictions to certain specifications of positive political theory. There are also overlaps between the broadest utilitarian theory and the deontological approach. Thus, my discussion of doctrinal evidence is of suggestive value only.[37] Nonetheless, analysing relevant aspects of the law in light of positive theory improves our understanding of the law and has some normative or critical weight as to how we should respond to recent or proposed changes. My discussion below divides into factors relevant to the establishment of jurisdiction, and factors relevant to its exercise.

A Establishing Jurisdiction

1 The Decline of Narrow Territorialism

In the early 1900s Anglo-American conflicts scholarship and doctrine were dominated by what may be called ‘narrow territorialism’. Courts interpreted their sovereignty in an absolute, but strictly defined manner. They were competent to exercise jurisdiction over any defendant present within the state.[38] This was reinforced by the vested rights choice of law scholarship, in which entitlements were referable solely to the law of the place those rights arose.[39] Such an approach permitted the exercise of transient jurisdiction.

Transient jurisdiction, as I have indicated above,[40] frequently violates fairness requirements — although the violation is substantial only in a society in which defendants are mobile. It may also be inconsistent with utilitarianism since the defendant’s presence is unlikely to be the most reliable indicator of either party or case proximity. Presence is a flawed concept to apply to corporations, because of the metaphysical requirements it imposes.[41] However, ascribing narrow territorialism to a political theory of jurisdiction is difficult, since it imposes a limited scope on legislative bargains and it preserves low-cost exit. The best explanation for narrow territorialism is simply the minimisation of the costs of judicial administration. Presence is a simple, verifiable concept, which limits the complex inquiries a court must make into the comparative advantage of different forums.

Narrow territorialism is decreasing as a result of various factors. Two of these, which I discuss below, are the willingness of courts to stay their proceedings where the forum is inappropriate, and the relaxation of attitudes of courts towards ‘exorbitance’ in the exercise of long-arm jurisdiction. However, some jurisdictions, in particular the United States, have specifically jettisoned narrow territorialism in favour of closer scrutiny of the contacts between the defendant and the forum, and the transaction and the forum.[42]

It is difficult to differentiate between the theories in explaining the decrease of narrow territorialism. Greater attention to forum suitability supports deontological and utilitarian theories, and the demise of transient jurisdiction may limit the persuasive force of the unilateral concept of social welfare, although that effect must be balanced against the broader scope that long-arm statutes afford to the cases that can be commenced in the forum. Likewise, the political implications of the demise of narrow territorialism are ambiguous, although the increasing legal complexity in the establishment of jurisdiction arguably favours lawyers.

2 The Reach of Long-Arm Statutes

Common law jurisdiction, which is based on the narrow territorial concept of presence, allows jurisdiction to be established as of right, subject to the defendant’s entitlement to seek a stay. A case that requires service out of the jurisdiction depends on statutory authority — a so-called ‘long-arm’ statute. In England and Australia the source of that authority has been found in rules of court.[43] Historically, this jurisdiction has been regarded as an extreme, exorbitant one.[44] However, it is important to note that this attitude prevailed because the original rules of court required plaintiffs to obtain prior leave on an ex parte application. That is no longer so in several states, although the plaintiff may be required to justify service at a later stage, which changes the procedural context in which these questions are raised.

There are signs that courts have softened their stance on the exercise of jurisdiction under long-arm statutes. The English courts have toned down their classification of long-arm jurisdiction as ‘exorbitant’.[45] The courts only inquire into whether the grounds for service are clearly satisfied and whether the English forum is appropriate to the resolution of the case. That test is the same as the one used in England for a stay of jurisdiction regularly brought within the forum.[46] The Australian High Court has observed that applications for prior leave or to set aside service out of the jurisdiction must be governed by the same principles as stay applications, simply because ‘the issues raised are the same’.[47] The ultimate question in each case is ‘the appropriateness of the chosen forum’.[48]

This development may recognise a phenomenon that I discuss below — multilateral conventions are leading us away from the distinction between common law and long-arm jurisdiction, and towards the concepts of general and special jurisdiction. Here, the accepted grounds of special jurisdiction over a foreign defendant are similar to some of the conventional grounds for service out of the jurisdiction accepted in Anglo-Australian rules of court, yet the plaintiff may invoke the ground regularly without seeking leave or satisfying a forum non conveniens test.

Explaining the increased reach of long-arm jurisdiction depends on the grounds in the long-arm statute. The concentration on cause of action attributes that is characteristic of Anglo-Australian rules of court supports the (broad) utilitarian and deontological theories. The absence of rules favouring local plaintiffs is inconsistent with the unilateral utilitarian theory, and with relatively extreme versions of the political theory of jurisdiction.

On the other hand, long-arm rules create problems in some categories of tort cases, where they predicate on the place in which damage occurs.[49] In torts in which there is substantial separation in space and time between the impugned action of the defendant and the alleged damage sustained by the plaintiff, rules that predicate on damage can function both unfairly and inefficiently.[50] A good example is a products liability case in which the defendant may be incapable of foreseeing where damage might eventually occur and, thus, which forum (and which law) would be applied. The inefficiency derives from the potential for multiplication of forums, in which evidence in relation to the cause of action is unlikely to be located. The unfairness derives from the defendant’s likely lack of consent and domicile, the weak territorial nexus, and the potential lack of mutuality. The most credible hypothesis explaining this is the unilateral social welfare hypothesis, since the costs are likely to be borne by out-of-state manufacturers, whereas benefits go to in-state plaintiffs and lawyers.

3 Multilateral Pressure towards Pre-Specification

One of private international law’s unresolved tensions is between the case for increased pre-specification, and the opposing case for increased ex post discretion, in the exercise of jurisdiction. As I discuss in Part V(B) below, common law courts have become more willing to review the exercise of jurisdiction in particular cases, through the easing of the principles applicable to stays of jurisdiction on grounds of inappropriateness.[51] However, there is a corresponding pressure, deriving from the advantages of a multilateral convention on jurisdiction and the recognition of judgments, to provide more precise statements of the courts in which cases may be brought and a greatly reduced role for ex post discretion in staying proceedings. This derives in part from the apparent success of the European conventions,[52] in which there is no scope for concepts such as forum non conveniens. England, being obliged to apply the convention to residents of the European Union, is whipsawed between the two movements. At the instigation of the United States, the Hague Conference on Private International Law has adopted a Preliminary Draft Convention quite similar to the European Conventions.[53] Only time will tell how many countries ratify any final convention, and how much influence it will have.

Do our theoretical accounts offer any insight into this cycling between specificity and discretion? The convention’s restrictions on the forums that have general and special jurisdiction embody the deontological theory’s concern with the fairness of subjecting a defendant to suit. More importantly, they typically provide a means by which to pursue broader utilitarian goals, since they enable states to commit credibly to the eschewing of excessive forms of jurisdiction.[54] Although there is a natural concern that pre-specified grounds for jurisdiction may permit plaintiffs to commence litigation in inefficient forums, and may preclude plaintiffs from commencing litigation in efficient forums, only empirical evidence can tell us how substantial the divergence between rules and discretion actually is.

4 The Consolidation of Forum Contractibility

In general, most jurisdictions permit parties to select the forum for litigation in their contracts.[55] However, this has not always been the case, and the position was only definitively established in United States federal law in the 1970s.[56] An entitlement to contract over forum is a substantial concession to a utilitarian theory of jurisdiction, since, if transaction costs are low, parties should make efficient, cost-minimising forum choices. The exercise of agreed jurisdiction states the strongest case for the legitimacy of that forum’s exercise of sovereignty over the defendant. Contractibility of forum is inconsistent with the claim that interest groups prefer to restrict exit from the forum in order to increase the scope of an interest group bargain.

5 The Interpretation of Statutes

The interpretation of statutes is perhaps the single most important context for a political theory argument, since interest group behaviour is much more likely to be reflected in legislation than in common law.[57] Two matters are important — first, how legislation alters private international law rules and, second, how courts interpret the legislation. As regards the first, we find something of a puzzle. Legislation relating to contracts often directs courts to ignore choices of foreign law when determining the proper law of the contract.[58] However, the legislation does not explicitly make equivalent provision in relation to contractual choice of jurisdiction by directing courts to ignore forum clauses and to exercise jurisdiction. I am not aware of Australian legislation that specifically mandates the exercise of jurisdiction, irrespective of the existence of a jurisdiction agreement, even though it is presumably no harder to draft than the equivalent provision in a choice of law agreement. This therefore raises the second question — what should courts do in the absence of these provisions?

In the High Court decision in Akai, a majority exercised jurisdiction to enforce Australian law, in spite of a choice of foreign law clause (to which the legislation did not permit effect to be given) and a foreign jurisdiction agreement (in respect of which the legislation was in fact silent).[59] By contrast, foreign courts to which the parties have submitted contractually rarely have any similar concern to enforce foreign mandatory rules.[60] The English litigation in Akai Pty Ltd v People’s Insurance Co Ltd enjoined further Australian proceedings, illustrating the dichotomy between local mandatory rules and foreign mandatory rules.[61]

Such a result seems to enhance the political theory of jurisdiction, since it provides as much scope to the interest group bargain as local sovereignty permits. However, it is less clear why courts should choose to further these bargains, while the parliament, which is more susceptible to interest group influence than courts, does nothing. My only answer is this — there do not seem to be very many cases in which parties are actually willing to litigate out of the country, if the decided cases are any guide. If that is the case, the costs of exit must be high, so that interest groups have no very strong incentive to include explicit provisions on the exercise of jurisdiction. The reaction of courts probably has nothing to do with the furtherance of interest group behaviour, but simply represents judicial biases in favour of taking expansive views of the apparent policy behind the legislation in question.[62] For instance, the interest group provenance of the Insurance Contracts Act 1984 (Cth) in Akai[63] is not very strong; furthering its policy appears to be a public-regarding result, even though its principal effect is to assist a party to avoid its contract. We see similar effects in several recent cases involving applications for stays in which relief was sought under s 52 of the Trade Practices Act 1974 (Cth).[64] These cases could also be rationalised as furthering unilateral social welfare.

6 Foreign Land

Historically, courts have refused to exercise jurisdiction over a case involving a question as to the title to, or torts infringing, land situated outside the jurisdiction.[65] The court does not, however, deny jurisdiction in cases which involve a contract in relation to foreign land or restitutionary suits.[66] Although such a rule could simply be dismissed as an accident of history, it is more useful to regard the rule as an example of a narrow utilitarian approach to the exercise of jurisdiction. A rule denying jurisdiction over foreign land minimises the cost of judicial administration by excluding cases which are tangibly foreign. Distinguishing torts from contracts also flows from a unilateral definition of social welfare. Locals are likely to benefit from a wider jurisdiction to enforce contracts, as it reduces the transaction costs of dealing in land. The benefits associated with trade presumably accrue to parties within the jurisdiction. Torts, on the other hand, do not involve exchanges. Torts protect property entitlements; however, the benefit of that protection must primarily remain in the situs.[67]

The point can be made in a similar way as a political hypothesis. If one examines some of the recent cases in which foreign land has been in issue, the denial of jurisdiction has primarily enabled the externalisation of costs by foreign companies to local residents.[68] The rule therefore favours in-state interests at the expense of out-of-state interests, even though the exercise of jurisdiction may have had utilitarian justifications in decreasing pollution or more efficient use of environmental entitlements. The statutory repeal of the common law rule in New South Wales, although little litigated, is arguably consistent with a broader definition of social welfare — although it also serves the interests of local litigators.[69]

B Exercising Jurisdiction

1 The Stay

I have mentioned that most common law jurisdictions have eschewed the highly restricted grounds for declining jurisdiction asserted in the early part of the century. The English law principle, articulated in St Pierre v South American Stores (Gath & Chaves) Ltd,[70] was that a stay would only be granted if the continuance of the action was oppressive, vexatious, or an abuse of process, and the stay would not cause injustice to the plaintiff. Courts now recognise, to a greater or lesser extent, a latitude to stay cases when the forum in which proceedings have been brought is unsuitable or inappropriate.[71] England, Canada and the United States have adopted a forum non conveniens principle. It permits the grant of a stay where another forum, having jurisdiction over the parties, is appropriate for the trial of the action, based on the interests of the parties and the ends of justice. Australia has not been as progressive but has softened the meanings of ‘oppressive’ and ‘vexatious’. It now suffices that the applicant shows that ‘there is an available and appropriate tribunal in some other country and that the local court is a clearly inappropriate forum in all the circumstances.’[72]

Which theory best explains this sea change in the law? In this article, I will answer this question by examining the declared doctrine. However, it is appropriate to acknowledge that the question can only be comprehensively answered by empirical examination of the way in which cases are actually decided. In particular, are factors such as the nationality of the person seeking and the person resisting the stay significantly related to outcomes?

Jaffey on the Conflict of Laws synthesises the factors that English courts consider in determining the forum conveniens in stay cases.[73] These factors substantially accord with broadly defined utilitarianism. First, the parties’ territorial connections with the forum and the locus of the evidence are simply party and case proximity. Second, the existence of another forum in which proceedings are significantly further advanced also has an economic basis in marginal costs, provided one looks at future costs in both forums, not at costs already incurred.[74] Third, courts will lean towards a stay where a suit is part of complex multi-party litigation which is most expeditiously resolved in another forum. This is the scope economies in litigation factor. Wealth maximisation seems to be a guiding principle in these cases. This is apparent since the grant or refusal of a stay can be justified, even if that order makes the parties to the suit before it worse off, if parties to other suits were made sufficiently better off to be able, in theory, to compensate the former litigants. It therefore trades off fairness and cost minimisation.[75] The fourth factor also recognises the economies of scope from litigating cases in a single forum, which are related not in terms of parties but in terms of subject matter and legal representatives. The fifth is the desirability of staying proceedings in a forum in which the language used is not the same as that of a majority of the relevant documents. It is an aspect of case proximity that reflects judicial error and litigation costs.

The final factor in Jaffey on the Conflict of Laws refers to choice of law issues, and thus invokes foreign law costs. The authors state that it may be preferable for the legal system whose law applies to provide the forum.[76] It is true that proving foreign law has positive costs, including expected judicial errors. However, the legitimacy of choice of law depends on how that law is selected. Multilateral choice of law rules, which are as likely to select forum law as foreign law, legitimate the relevance of substantive law. However, using methods biased towards the lex fori would be more consistent with a unilateral definition of welfare and, where statutes have interest group provenance, of interest group behaviour. In general, however, courts outside the United States have generally been more restrained in applying forum law, although it does figure in the rule in Phillips v Eyre.[77] I note below that some Australian cases apply local statutes in a questionable pre-emptive manner.

Choice of law is related to the ‘second stage’ of forum non conveniens in English law, in which a case in ‘justice’ may be made for permitting the plaintiff to continue in an inappropriate forum.[78] I do not regard this reference to justice as consistent with the deontological theory. If justice is equated with equal treatment, it is unjust to favour plaintiffs by refusing a stay when the local forum is non conveniens but not to grant a stay to a defendant when the local forum is conveniens. As Part III above shows, the defendant is in need of procedural protection, not the plaintiff, who chooses the forum. The courts have, however, regarded these juridical advantages as being less important than connecting factors.[79]

In Australia, the Voth test for the grant of stays requires that the Australian forum be clearly inappropriate. In adopting the test, the High Court thought that it would only occasionally lead to different results from Spiliada.[80] Richard Garnett, however, identifies several categories of Australian cases in which he considers Spiliada would favour a stay but in which jurisdiction was exercised. The first category involves a local plaintiff and a foreign defendant.[81] Two of the three cases cited[82] fit the paradigm I described in which there is a spatiotemporal separation between action and damage. This separation, as I outlined above, is likely to be problematic.[83] These cases present ready opportunities for externalising costs. However, the results are mitigated by the fact that evidence was located locally, which suggests some case proximity.[84] The second category involves a claim to a remedy for breach of s 52 of the Trade Practices Act 1974 (Cth), which has sufficed for the court to refuse a stay.[85] Jurisdiction depends on characteristics of the law the plaintiff pleads. This is unfair, because of the difficulty for the defendant to avoid the statute. The third category involves the subjection of the defendant to jurisdiction because it suited the plaintiff to join that suit with another in which the forum had jurisdiction. This is not strictly inconsistent with wealth maximisation, but it represents a quite controversial furtherance of it.[86] These cases drive a wedge between the English approach, with its emphasis on fairness to the defendant and the minimisation of litigation costs, and the effect of the Voth test. The divergence often seems to be most easily explained in terms of unilateralism and the favouring of local plaintiffs. I do not see any consistent interest group effect explaining these cases (unless it is the interests of Australian plaintiffs’ lawyers).

To conclude the discussion of stays, we may note that in general, a strong vein of wealth maximising utilitarianism underpins the approach to the award of stays. Yet at the same time, by enabling objections by the defendant, it reflects deontological concerns as well. The Australian cases deviate to some extent in a way that seems more strongly unilateral. It is, however, difficult to measure the overall significance of the deviation.

2 The Anti-Suit Injunction

The anti-suit injunction is an exercise of power by a local forum over a person subject to its jurisdiction that restrains the continuance of foreign litigation.[87] The jurisdiction has long been a sensitive one, because of its interference with the process of another court. The virtue of comity has often been cited as a reason for caution in granting these orders. There are instances where comity is at stake, such as the Laker Airways litigation.[88] That case involved a significant question of competition law in an inter-continental market. Other cases are unlikely to be quite so political, and it seems unreasonable to assume comity will always be impaired.[89]

Anti-suit injunction cases break down into several identifiable categories.[90] The clearest case for the grant of an anti-suit injunction is to enforce a contract, such as a jurisdiction or arbitration agreement, or a covenant not to sue agreed in a settlement. The English courts are coming to regard this as a straightforward case for the grant of an injunction to which comity is irrelevant.[91] This is consistent with broad utilitarianism, and, unless there is something unfair about the contract, enforcement of party agreement is fair because consent legitimates it.

In the second category, the relief the plaintiff seeks is only available in one forum (‘single forum cases’). These can be differentiated according to whether the foreign forum’s jurisdiction is sound in terms of the private international law of the court asked to enjoin it. Where the jurisdiction is sound, an anti-suit injunction is inappropriate since that forum has a prima facie claim to be the efficient forum. The opposite is true where the plaintiff has invoked an exorbitant or extraterritorial jurisdiction. The usual example of the latter is the extraterritorial aspect of American antitrust law.[92] Courts have been more willing to enjoin suits where the behaviour on which the plaintiff relies is not connected with that jurisdiction, and the forum is non conveniens.

In the third category, the plaintiff can pursue its cause of action in the foreign forum and the forum in which the injunction is sought. The defendant will allege that litigation should proceed in the latter forum. English and Australian cases state that demonstrating that the local forum is conveniens is necessary but not sufficient. The defendant must demonstrate that litigation in the foreign forum is oppressive or vexatious (analogously to the St Pierre test for stays).[93] It is difficult to apply the deontological theory to these cases. They depend on whether the plaintiff in the foreign proceedings will also be the plaintiff in the local proceedings (for instance, the plaintiff may seek a negative declaration in the foreign forum, or there may be major cross claims). Where the cases have the same ‘polarity’, the concept of fairness to the defendant has some sort of identifiable meaning. The defendant is seeking the relocation of the litigation, so the fairness of the foreign forum must be an important guide. Where the polarity reverses, the concept of fairness becomes highly complex.[94] The applicant seeks not only a forum change but the advantages that flow from firing the first shot in that forum; the respondent seeks to resist the forum change to keep the tactical advantage in the chosen forum. Efficiency questions are less complicated than these ones of fairness, and should be expected to have a stronger influence on outcomes.

In practice, courts have resolved multi-forum anti-suit injunctions in a way that treats fairness concerns as being of somewhat secondary importance — but in which efficiency is not consistently furthered, either. In particular, courts do little to restrain quite plain instances of forum shopping. Two examples are the House of Lords decision in Castanho v Brown & Root (UK) Ltd,[95] and the High Court decision in CSR Ltd v Cigna Insurance Australia Ltd.[96] In the former, litigation in Texas was not enjoined, despite the absence of any real connection with that forum and the imminence of settlement in England. In CSR the High Court restrained proceedings in Australia and refused to enjoin American proceedings commenced by an insured, despite the very strong Australian nexus, and the fact that in another very similar suit, the insured had commenced in Australia. By contrast, the Privy Council decision in Société Nationale Industrielle Aerospatiale v Lee Kui Jak,[97] enjoining a plaintiff from litigation in Texas, in favour of Brunei, furthered efficiency, since the defendant’s claim for contribution against a third party could not have been litigated in Texas. So, at the highest level, the cases are ambivalent.

In the final category of case, the court in which the injunction is sought is not proposed as a forum under any circumstances. In Airbus Industrie GIE v Patel,[98] litigation in relation to an air crash commenced in India. However, the plaintiffs (including some British citizens) began proceedings in Texas. The House of Lords refused to restrain those proceedings, holding that it lacked a sufficient interest in the matter.[99] The case is best explained in terms of narrow utilitarianism. Refusing the injunction favours British plaintiffs and avoids committing the court to the expenditure of judicial resources in resolving foreign disputes. Sometimes, social welfare will require disinterested, ‘non-combatant’ courts to exercise jurisdiction, since the inefficiency of a forum may be palpable in some cases,[100] and the plaintiff in these proceedings may be able to avoid the jurisdiction of the natural forum.

The anti-suit injunction is a hard subject to theorise. The contract-enforcing cases further both fairness and broad utilitarian interests. The non-combatant cases further unilateral interests and minimise the costs of judicial administration. The other cases lack obvious explanations apart from the aberrant concept of fairness to the plaintiff, a subject I consider in Part VI.

C Conclusion

The above sections of Part V do not unanimously support any single theory, in part because of overlapping predictions, but utilitarian factors seem to drive much of the law on jurisdiction. At the same time, much of the evidence supporting utilitarian concerns is also consistent with a hypothesis that the law on jurisdiction protects defendants against the unfair exercise of state power. The forces produced by multilateral negotiation of conventions and treaties have intensified the power of these hypotheses. The narrower utilitarian motivations seem a less powerful explanation, except for cases which only minimally concern the forum, such as non-combatant anti-suit injunctions and torts to foreign land. Recall that, in Part II(B)(1) above, I observed that these were the cases where courts would be expected to economise on judicial resources. However, there are sufficient Australian cases in which jurisdiction has been exercised, despite minimal connections, that one should accept unilateralism as influencing outcomes if not the formal legal principle. The political theory has no straightforward representation, in part because its predictions depend on various parameters, except to the extent that courts rarely decline jurisdiction when confronted with apparently mandatory legislation. Yet, even here, paternalism often competes with political theory as an explanation for such events. We should not of course expect the law to be completely consistent, nor should we be surprised by different explanations holding in different circumstances. However, given that international law is often thought to be in the grip of a strictly competitive, non-cooperative prisoner’s dilemma,[101] it is surprising that courts further broad utilitarian and fairness considerations as consistently as they do.

VI THE UTILITARIAN–DEONTOLOGICAL CONVERGENCE

A The Theoretical Case for Convergence

At the end of the last section, I made a claim that many doctrinal phenomena reflect both utilitarian and fairness considerations. This must come as something of a surprise, since utilitarianism and fairness are often thought to be Manichean opponents. Some theorists, of whom lawyer-economist Richard Posner is illustrative, have long dismissed the preoccupation with justice as simply a codeword for efficiency,[102] but that claim has always been controversial, as has Posner’s relentless advocacy of wealth maximisation as a norm of social choice. It is my present purpose to argue a tentative case for convergence between the outcomes favoured by both a broadly defined measure of social welfare and the concept of fairness in resolving many conflictual situations in private international law. The general argument is this: conflicts cases involve dissonance between jurisdictions, and between substantive policies. Incapable of using the furtherance of policy as the basis for decision — since there is no longer a single policy — courts have intuitively found principles to resolve these uncertainties in more abstract norms of social choice. At that level of abstraction, fairness and efficiency inevitably yield similar prescriptions.

Let us first examine why wealth-maximising jurisdictional principles will be fair ones. If information was complete, perfect, costless, and distributed equally between the litigants and the courts, jurisdictional principles would maximise wealth by conferring jurisdiction on the forum with the lowest costs. Information, however, is not costless. In particular, a party will find it difficult to observe particular phenomena associated with the other party (such as their costs in alternative forums), and will find it difficult to verify cost phenomena to the court. This makes it impossible for courts to predicate jurisdiction on cost minimisation.

In these circumstances, is there a second-best solution? Plaintiffs will prefer to litigate in the forum that maximises their ‘net award’, given the probabilities that forums will accept jurisdiction and apply their own law. Where law differs between forums, a plaintiff may be better off litigating in a forum in which their costs are higher than in some other forum. This is problematic because the litigation costs are social costs, but the award is often only a private benefit matched by the defendant’s private cost. Sometimes, the forum the plaintiff prefers will be particularly unsuitable for the defendant. The plaintiff’s forum choice can externalise costs to the defendant. The second-best solution is therefore to limit the plaintiff’s choice of forum to those in which the defendant’s costs to litigate can be expected to be low across a range of cases, such as the court of the defendant’s domicile. This achieves two things. First, it limits the cost externality the plaintiff can impose on the defendant. Second, the defendant’s own ties to the forum decrease a pro-plaintiff bias that might be expected to increase the scale of the private benefit to the plaintiff, and the costs the plaintiff incurs to pursue these. In other words, fairness to the defendant encourages the plaintiff to make forum choices that minimise the social costs of litigation. This can be reinforced by denying recognition of judgments of courts which are unfair to the defendant, which English law does to some extent.[103]

This claim is related to my argument (with Keyes) that the lex loci delicti (the law of the place where the tort was committed) simultaneously satisfies both fairness and efficiency criteria in choice of law in non-contractual torts.[104] By providing for a multilateral rule that will produce the same outcomes across forums, the lex loci delicti encourages the plaintiff to choose a cost-minimising forum, in much the same way as defining available jurisdictions in terms of their connections with the defendant, or with the cause of action, limits the cost externality the plaintiff can impose on the defendant. The rule’s predictability, its clarity as to what a defendant must do to avoid subjection under any particular tort law, and its neutrality between the parties, make for a strong argument that the lex loci delicti would be selected by parties in an original position — and thus that the rule is just.[105]

There is a case for seeing in all of the principal theorisations of private international law aspects of this simultaneous efficiency–fairness concern. The first modern conflicts scholarship was vested rights theory, which held that choice of law rules simply gave effect to legal rights and obligations that had vested elsewhere.[106] Vested rights theory was formalistic, in the sense that it claimed that the resolution of conflicts was a matter of pure logic. This claim to apodictic certainty was shown to be false by later realist scholarship. Nonetheless, the choice of law rules offered by vested rights theory aspired to forum-neutral consistency in choice of law decisions. That consistency provides in general the sort of economic and fairness benefits that can be ascribed to the lex loci delicti in tort cases. The realists and their epigones, including the proponents of interest analysis, also criticised the proposition that choice of law could ever be forum neutral.[107] Given that premise, their preference for forum law simplified (and reduced the cost of) adjudication, and the concept of ‘interest’ provided something of a control device to limit excessive — and thus inefficient and unfair — application of forum law. The critics of interest analysis, such as Brilmayer and other economists, repudiate the capacity of the ‘interest’ concept to function in this way. They have sought stronger bulwarks against unfairness or inefficiency, which have emphasised concepts of territoriality and consent in a way that is not very different in essential respects.

B Is There a Case against the Convergence Conjecture?

If my convergence conjecture is correct, has it any relevant limits? What circumstances might drive a wedge between the furtherance of efficiency and the furtherance of fairness? I limit my discussion to matters of jurisdiction, and briefly identify four case types — the multi-party case; the paternalist case; the deserving plaintiff case; and the guillotine case.

1 The Multi-Party Case

I have already sketched the logic of these cases above.[108] One influence on the cost of litigation is scope economies in litigation. Average litigation costs decline over a number of related claims heard together. There are three basic fact templates in these cases — the class action, joined defendants, and third party actions. The class action involves multiple plaintiffs. It drives no wedge between efficiency and fairness, since the reduction of costs should greatly benefit the defendant. Although some plaintiffs may be worse off, our theory does not recognise plaintiff political rights. The other two cases are more complex.

A third party action would involve A suing B, who seeks indemnity from C. These cases create problems where B and C have substantially different preferences as to forum. B might be content with the forum, F, that A has chosen, but that forum may not minimise C’s costs. It should be obvious that simply registering the party’s ‘votes’ for forum is likely to produce neither fair nor efficient outcomes. C’s marginal cost litigating in F may be high enough to offset the marginal benefit to A and B litigating there. The voting solution is also problematic to the point of impossibility if there are three possible forums, F, G and H. Consider a case in which the parties’ preferences over forum were ranked as follows:

Table 1 Party Rankings of Preferences for Forums

Voting cannot produce a single outcome in this case. If parties must express preferences between F and G, F wins, because A and C prefer it, although B does not. If parties must express preferences between G and H, G wins, because A and B prefer it, although C does not. Despite the fact that F beats G, and G beats H, F will not beat H if parties must express preferences between the two. B and C prefer H, although A does not. This paradox — first attributed to Condorcet in the eighteenth century[109] — is a simple illustration of a very general theorem in economics, the Arrow impossibility theorem.[110] It demonstrates the inevitable cycling and indeterminacy in any unbiased attempt to aggregate voted preferences, once there are three options and three voters. It therefore can never be sensible for a judge to resolve a third party case by saying that the plaintiff and the defendant’s preference outweigh the third party’s objection. Is there a uniquely fair solution in these cases? There may be one if there is a single forum in which B and C’s political rights are not infringed, but that may not be the case. We are therefore faced with a trade-off — grant a stay to C, and require B to conduct litigation in two forums, or refuse a stay and violate C’s rights to fairness. Fairness is therefore only adventitiously of assistance. Utilitarianism provides a solution based on the comparison of costs, if one accepts wealth maximisation as a criterion. One simply compares total litigation costs under the stay. However, such an order inevitably drives a wedge between fairness and utilitarianism, unless the party who gains from it is forced to share some of those gains with the loser. For example, if the stay was refused, B’s costs might be $100 and C’s costs $200. If the stay was granted, B’s costs in the suit against A might be $90, and in the separate suit against C, $80, and C’s costs $150. Although refusing a stay is wealth maximising, it makes C worse off by $50. The unfairness might be rectified by ordering B to compensate C by at least that amount, which B can do, since B is better off by $70.[111]

The case of joined defendants raises similar problems. It is the reverse of the class action, since it involves a joinder of suits against defendants. The plaintiff may often find this an advantage, and indeed it may often be wealth maximising for the suits to occur together. However, as in the third party action, this may not be fair to one (or possibly both) of the defendants. Unlike the third party action, however, the plaintiff is dictating the unfairness. If it is wealth maximising to allow the defendant to be joined, the plaintiff may be able to infringe the political right of the defendant to whom he or she would otherwise be obliged to submit. Again, however, the gap between fairness and efficiency can be bridged if the plaintiff must compensate the defendant for marginal litigation costs in the forum.

2 The Paternalist Case

This case involves situations in which the court trumps an express contractual submission to a forum which a party then seeks to avoid. Although courts generally enforce jurisdiction agreements, they do not do so in all cases.[112] The jurisdiction to stay litigation in the agreed forum implicates standard paternalistic arguments against freedom of contract, such as inequality of bargaining power and contracts of adhesion, many of which lack economic substance. The information costs associated with the presence of the clause, and the nature of the cost differential associated with forums is a more serious objection. Although non-enforcement is rarely an optimal response to information costs problems, a court in a conflicts case is often incapable of pursuing more efficient solutions unilaterally, such as disclosure obligations.[113] A more complex argument is that jurisdiction agreements are insufficiently state contingent — meaning that the parties have not optimally differentiated the forums to which they submit to correspond with different types of dispute. Stays may allow the inefficiencies of such clauses to be mitigated, although whether judicial activism is a good solution to contracting problems of this sort is hotly disputed territory.[114] A good example of a paternalist case is Carnival Cruise Lines Inc v Shute,[115] with its echoes of the classic ticket cases, although in that case enforcement (and presumably efficiency) was favoured. The tendency to exercise jurisdiction when a local plaintiff pleads rights under consumer protection legislation could also be seen as the paternalistic pre-emption of efficiency or fairness.[116]

3 The Deserving Plaintiff Case

The deserving plaintiff case is similar to the paternalist case, except that the focus is not on the time of contracting, but on the plaintiff’s claims to indulgence at the time of the suit. The best example is Castanho,[117] in which a severely injured plaintiff was permitted to discontinue in a jurisdiction in which a settlement had almost been concluded in order to pursue a much larger damages claim in Texas. Clearly, an anti-suit injunction would have minimised litigation costs, but the court’s compassion seemingly got the better of it. Garnett discusses a similar case, Teare v British Nuclear Fuels plc,[118] in which a plaintiff was allowed to continue in Australia, despite England’s overwhelming connections to the case, because he was unfit to travel. This is an instance of how efficiency and fairness are most likely to divide when the court moves away from abstraction to the facts of particular cases. Fairness becomes conceived in personal, result-oriented terms, rather than institutional or procedural terms.

4 The Guillotine Case

All litigation is costly, and we have observed that efficiency is judged in terms of minimising costs. If that is so, how might an efficiency-regarding judge resolve a jurisdiction problem between forums when the litigation is impossible in one of the forums, for instance because its statute of limitations applies, or because of differences in law on survivorship? I describe this as a ‘guillotine case’, since a judge who stays litigation in favour of the forum in which action is barred, necessarily decapitates the plaintiff’s suit. Examples occur in practice where the defendant cannot show that a foreign forum would assume jurisdiction over the case if the matter was stayed.[119] This is not simple. If the suit was inherently a nuisance suit, or would have never won on the merits, foreclosing the source of litigious prolongation is clearly efficient — but making such a classification of the case often requires far more knowledge than a court has available to it at the interlocutory stage. If the case is inherently meritorious, but in one of the forums some non-discriminatory law sterilises it at some point in time after the cause of action arose, we face a very difficult choice.[120] What is the appropriate efficiency comparison? Is the court to compare the costs as if the cause of action had not been sterilised, on the basis that it is not its place to prejudge the merits? Or does it make the comparison in favour of an outcome which minimises the social costs of litigation, which will be zero if the suit cannot go forward, and in so doing leave the plaintiff without any remedy? As with deserving plaintiff cases, the court’s view of the merits is likely to govern decisions at the jurisdictional level, even though, as I have argued, no coherent view of corrective justice compels this.

5 Conclusion

Even though doctrine will reflect the utilitarian–deontological convergence, results in practice may be more variable, as courts manipulate the discretion the law reposes in them to do what they see as justice in the case, even though the system is not unjust or unfair. Without wishing to paint this as improper, we need to recognise such result-oriented judging, and to design legal principles to make its existence transparent, and therefore the legitimate subject of scrutiny, rather than shrouding it in secretive obscurity.

VII CONCLUSION

I started this essay with a dialogue between a doctrinal lawyer and myself. The former was not much interested in positive theory, but in legal principles. In this essay, I hope to have shown two things. The first is that studying cases can be a different enterprise from studying what judges say in those cases. As we have seen, private international law is often formulated in a consciously abstracted manner, in which facts are often hidden behind a veil of ignorance. However, the facts of particular cases, above and beyond those on which the legal principles formally predicate, can introduce a tension between the systemic and the specific which practising lawyers need to understand.

The second point is aimed at the values of legal scholarship. Many academics distrust economic analysis, because it appears to affirm norms they have been conditioned to believe the legal system does not partake of. This paper demonstrates that economic-utilitarian principles not only explain the law on jurisdiction, but have much in common with those traditional virtues of the lawyer, justice and fairness.

The next step in the research program that I have begun in this article is to complete the positive theorisation of the law by examining empirical regularities in the results of cases in which the establishment and exercise of jurisdiction has been in issue. That will provide an important extension of, or qualification on, the doctrinal evidence. Such an understanding of the concept of sovereignty in private law is as important to those who practise the law, as it is to those who purport to theorise it.

[*] BCom (Hons), LLB (Hons), MFM (Qld), PhD (Griffith); Associate Professor, Faculty of Law, Griffith University; Leader, Business Regulation Program, Key Centre for Ethics, Law, Justice and Governance. This article is part of a larger project, entitled Policy and Pragmatism in the Conflict of Laws, which is funded by an Australian Research Council Small Grant. I would like to thank Tanya Pridannikoff for research assistance and the anonymous referees for their helpful comments.

[1] See text accompanying below n 53.

[2] It is in this sense of selecting forums on the basis of welfare maximisation, rather than with reference to some given set of rights, that I describe this theory as ‘utilitarian’. The theory could, without loss of generality, be conceived as one based on the efficiency criterion, wealth maximisation, familiar from the economic analysis of law. This is because I only purport to measure costs in monetary equivalents, rather than in terms of personal utility unsupported by willingness to pay. I therefore do not encounter the intractable problems associated with interpersonal utility comparisons: Richard Posner, The Problems of Jurisprudence (1990) 356–8. My approach is at one with most of the literature: eg, Alan Schwartz, ‘Proposals for Products Liability Reform: A Theoretical Synthesis’ (1988) 97 Yale Law Journal 353, 357–62.

[3] The Case of the SS Lotus (France v Turkey) [1927] PCIJ (ser A), No 10.

[4] Michael Whincop and Mary Keyes, ‘Economic Analysis of Conflict of Laws in Torts Cases: Discrete and Relational Torts’ [1998] MelbULawRw 16; (1998) 22 Melbourne University Law Review 370.

[5] See, eg, William Landes, ‘An Economic Analysis of the Courts’ (1971) 14 Journal of Law and Economics 61 (optimism); Lucian Bebchuk, ‘Litigation and Settlement under Imperfect Information’ (1984) 15 Rand Journal of Economics 404 (private information).

[6] See, eg, Steven Shavell, ‘Suit, Settlement, and Trial: A Theoretical Analysis under Alternative Methods for the Allocation of Legal Costs’ (1982) 11 Journal of Legal Studies 55.

[7] Richard Fentiman, ‘Foreign Law in English Courts’ (1992) 108 Law Quarterly Review 142.

[8] See generally Michael Whincop, ‘The Recognition Scene: Game Theoretic Issues in the Recognition of Foreign Judgments’ [1999] MelbULawRw 17; (1999) 23 Melbourne University Law Review 416.

[9] The first bias produces the second in a suit where there are no local interests.

[10] The value of a variable in forums Y and Z respectively is denoted by the subscript Y and Z.

[11] 15 – (1 + 4 + 6) = 4.

[12] 15 – (1 + 6 + 7) = 1.

[13] Local courts may also attempt to protect local defendants from suit. However, unilateral social welfare may allow local defendants to be sued locally, since it may make them worse off if they have to face the process of a foreign court. The question is therefore somewhat equivocal. It may depend on the extent to which the defendant is actually liable to suit in the foreign forum.

[14] (1996) 188 CLR 418 (‘Akai’). In this case an insurance contract between a Singaporean insurer and an Australian supplier contained both a choice of English law clause and an English choice of forum clause. The High Court of Australia held that both clauses were invalid, as they defeated the purpose of a Commonwealth Act, which was directed at preventing insurers from circumventing the jurisdiction of Australian courts. Consequently, the insured was allowed to claim rights in Australia under Australian legislation.

[15] This example is based on a ‘US rule’ for the allocation of party costs, under which each party bears its own costs. However, the analysis does not lose its generality if the British rule is used in each forum. Under the British rule the defendant bears the litigation cost, p_F(A_Y + A_Z), and the plaintiff bears the litigation cost, (1 – p_F)(A_Y + A_Z), where p_F is the probability of the plaintiff’s victory in forum F. Assume that the pay-offs, A_Y and A_Z, are certain in amount and probability (p_F = 1). The plaintiff’s pay-off litigating in Y becomes 15 (rather than 11), and the plaintiff will continue to choose not to litigate in Z, because the pay-off there is negative. Under the British rule, the defendant’s pay-off litigating in Z is 0 (rather than -7), and in Y is -25 (rather than -21). The effects of different costs regimes on forum choices and jurisdictional objections are studied in Michael Whincop, ‘Choice of Forum, the Stay Application, and Jurisdictional Trade: An Economic Model of International Litigation’ (Unpublished paper, May 2000) (copy on file with author).

[16] Kenneth Abbott, ‘Modern International Relations Theory: A Prospectus for International Lawyers’ (1989) 14 Yale Journal of International Law 335.

[17] In the present example, the cost of litigating in Y would be (1 + 4 + 6) = 11, while the cost of litigating in Z would be (1 + 6 + 7) = 14.

[18] This is a ‘battle of the sexes’ paradigm, in which husband and wife must choose independently to view a particular film, in circumstances where each selects a different film to watch, and each prefers to be with the other at the screening of the film they less prefer, rather than to be alone at the individually preferred film. However, each maximises utility, selfishly, if they are together at that person’s preferred film.

[19] Thomas Schelling, The Strategy of Conflict (1960).

[20] Conventions on jurisdiction are an example: Whincop, ‘The Recognition Scene’, above n 8.

[21] See, eg, Ernest Weinrib, ‘Toward a Moral Theory of Negligence Law’ (1983) 2 Law and Philosophy 37. For a powerful critique of the concept, see Posner, The Problems of Jurisprudence, above n 2, 313–34.

[22] Posner, The Problems of Jurisprudence, above n 2, 322.

[23] Michael Whincop and Mary Keyes, Policy and Pragmatism in the Conflict of Laws (2000) 102–3 (forthcoming).

[24] See, eg, John Rawls, A Theory of Justice (1971).

[25] Lea Brilmayer, ‘Rights, Fairness, and Choice of Law’ (1989) 98 Yale Law Journal 1277.

[26] Lea Brilmayer, ‘Consent, Contract, and Territory’ (1989) 74 Minnesota Law Review 1.

[27] Brilmayer, ‘Rights, Fairness, and Choice of Law’, above n 25, 1297–313.

[28] Michael Whincop and Mary Keyes, ‘The Nexus of Forums: A Contractual Theory of Jurisdiction over the Corporation’ [1998] UNSWLawJl 61; (1998) 21 University of New South Wales Law Journal 681.

[29] See, eg, Carriage of Goods by Sea Act 1991 (Cth) ss 11, 16; Civil Aviation (Carriers’ Liability) Act 1959 (Cth) s 41H; Health Insurance Act 1973 (Cth) s 126(3); Insurance Contracts Act 1984 (Cth) s 8(2); Trade Practices Act 1974 (Cth) s 67.

[30] As to public choice generally, see Gary Becker, ‘A Theory of Competition among Pressure Groups for Political Influence’ (1983) 98 Quarterly Journal of Economics 371; Sam Peltzman, ‘Toward a More General Theory of Regulation’ (1976) 19 Journal of Law and Economics 211. As to a positive political theory of choice of law, see Bruce Kobayashi and Larry Ribstein, ‘Contract and Jurisdictional Freedom’ in Francis Buckley (ed), The Fall and Rise of Freedom of Contract (1999) 325.

[31] See Becker, above n 30.

[32] See, eg, Michael McConnell, ‘A Choice-of-Law Approach to Products-Liability Reform’ in Walter Olson (ed), New Directions in Liability Law (1988); Kobayashi and Ribstein, above n 30.

[33] McConnell, above n 32.

[34] Kobayashi and Ribstein, above n 30.

[35] Charles Tiebout, ‘A Pure Theory of Local Expenditures’ (1956) 64 Journal of Political Economy 416; Frank Easterbrook, ‘Antitrust and the Economics of Federalism’ (1983) 26 Journal of Law and Economics 23; Barry Weingast, ‘Constitutions as Governance Structures: The Political Foundations of Secure Markets’ (1993) 149 Journal of Institutional and Theoretical Economics 286.

[36] Erin O’Hara, ‘Opting Out of Regulation: A Public Choice Analysis of Contractual Choice of Law’ (2000) 53 Vanderbilt Law Review (forthcoming). Consider an example. Assume that real estate agents are lobbying for legislation which regulates agents’ commissions to minimise price competition. The incentive for other interest groups to oppose such legislation depends on the cost the legislation imposes on them. The agents may be able to minimise opposition from other interest groups by keeping open the option for parties contracting with agents to opt out of the legislation by a contractual choice of foreign law. That may provide the political coalition necessary for the legislation to be passed. By contrast, should a contract opting out of local law ever be litigated, the agent will try to allege that the choice of law clause should not be enforced and will try to assert rights under the commission-regulating legislation.

[37] To minimise selection bias, I often rely on other assessments of these trends. For Australian evidence, I gratefully adopt Richard Garnett’s survey on stays of jurisdiction: Richard Garnett, ‘Stay of Proceedings in Australia: A “Clearly Inappropriate” Test?’ [1999] MelbULawRw 2; (1999) 23 Melbourne University Law Review 30.

[38] See, eg, Laurie v Carroll [1958] HCA 4; (1958) 98 CLR 310; Pennoyer v Neff, [1877] USSC 166; 95 US 714 (1877).

[39] See, eg, Joseph Beale, A Treatise on the Conflict of Laws (1935).

[40] See above p 389.

[41] Whincop and Keyes, ‘The Nexus of Forums’, above n 28.

[42] See, eg, International Shoe Co v Washington, [1945] USSC 158; 326 US 310 (1945).

[43] See, eg, High Court Rules (Cth) Order 10; Federal Court Rules (Cth) Order 8; Supreme Court Rules 1970 (NSW) Part 10; Supreme Court (General Civil Procedure) Rules (Vic) Order 7; Rules of the Supreme Court (Qld) Order 11; Supreme Court Rules 1987 (SA) Rule 18; Rules of the Supreme Court 1971 (WA) Order 10; Rules of the Supreme Court 1965 (Tas) Order 11; Supreme Court Rules (NT) Order 7; Supreme Court Rules (ACT) Order 12. For the English equivalent, see RSC Order 11.

[44] See, eg, The Hagen [1908] UKLawRpPro 6; [1908] P 189, 201.

[45] To be fair, the historical approach was not always radically different from a simple balancing approach: see, eg, Strauss & Co v Goldschmid (1892) 8 TLR 512, 513; Rosler v Hilbery [1925] Ch 250, 259.

[46] Compare Spiliada Maritime Corp v Cansulex Ltd [1986] UKHL 10; [1987] 1 AC 460, 481 (Lord Goff, Lords Keith, Templeman, Griffiths and Mackay concurring) (‘Spiliada’) (describing ‘exorbitant’ as an ‘old-fashioned word which perhaps carries unfortunate overtones’, and noting that the grounds under RSC Order 11 ‘are of great variety, ranging from cases where, one would have thought, the discretion would normally be exercised in favour of granting leave ... to cases where the grant of leave is far more problematical’) and Société Générale de Paris v Dreyfus Brothers [1885] UKLawRpCh 106; (1885) 29 Ch D 239, 242–3 (Pearson J). In Australia, see Sydbank Soenderjylland A/S v Bannerton Holdings Pty Ltd [1996] FCA 1224; (1996) 68 FCR 539.

[47] Voth v Manildra Flour Mills Pty Ltd [1990] HCA 55; (1990) 171 CLR 538, 563 (‘Voth’). The case concerned a negligence action commenced in NSW by companies in the Manildra Group against a United States solicitor. The High Court held that NSW was an inappropriate forum in which to proceed, as the alleged tort had been committed in Missouri and the solicitor would have been liable under Missouri law. Consequently, the action was stayed.

[48] Ibid 563–4.

[49] Distillers Co (Biochemicals) Ltd v Thompson [1971] AC 458; Castree v ER Squibb & Sons Ltd [1980] 2 All ER 589; Metall und Rohstoff AG v Donaldson Lufkin & Jenrette Inc [1990] 1 QB 391.

[50] McConnell, above n 32; C G J Morse, ‘Products Liability in the Conflict of Laws’ (1989) 42 Current Legal Problems 167. Garnett reports several Australian cases which involve separation in space and time between the defendant’s actions and the plaintiff’s damage: Garnett, above n 37, 43–4. In these cases, jurisdiction was exercised despite the seeming inefficiency of Australia as a forum.

[51] Piper Aircraft Co v Reyno, [1982] USSC 28; 454 US 235 (1981); Voth [1990] HCA 55; (1990) 171 CLR 538; Spiliada [1986] UKHL 10; [1987] 1 AC 460; Amchem Products Inc v British Columbia (Workers Compensation Board) [1993] 1 SCR 897; (1993) 102 DLR (4th) 96.

[52] Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters, opened for signature 26 May 1990, OJ No C 189, 2, 29 ILM 1413 (entered into force 28 July 1990) (‘Brussels Convention’), consolidated and updated version of an original convention by the same name, opened for signature 27 September 1968, OJ No L 299, 32, 8 ILM 229 (entered into force 31 December 1972); Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters, opened for signature 16 September 1988, OJ No L 319, 9, 28 ILM 620 (entered into force 25 November 1988) (‘Lugano Convention’).

[53] Preliminary Draft Convention on Jurisdiction and Foreign Judgments in Civil and Commercial Matters, adopted by the Special Commission of the Hague Conference on Private International Law, 30 October 1999, Hague Conference on Private International Law <http://www.hcch.net/e/

conventions/draft36e.html> (copy on file with author).

[54] Whincop, ‘The Recognition Scene’, above n 8.

[55] In English law, see The Eleftheria [1970] P 94. See also Huddart Parker Ltd v The Ship Mill Hill [1950] HCA 43; (1950) 81 CLR 502.

[56] The Bremen v Zapata Off-Shore Co, [1972] USSC 129; 407 US 1 (1972), affirmed in Carnival Cruise Lines Inc v Shute, [1991] USSC 56; 499 US 585 (1991).

[57] Jonathan Macey, ‘Promoting Public-Regarding Legislation through Statutory Interpretation: An Interest Group Model’ (1986) 86 Columbia Law Review 223.

[58] See, eg, Carriage of Goods by Sea Act 1991 (Cth) ss 11, 16; Civil Aviation (Carriers’ Liability) Act 1959 (Cth) s 41H; Health Insurance Act 1973 (Cth) s 126(3); Insurance Contracts Act 1984 (Cth) s 8(2); Trade Practices Act 1974 (Cth) s 67.

[59] (1996) 188 CLR 418. It is true that the High Court’s decision was based on other provisions of the statute but none of them explicitly required the result in that case. The outcome turned on questionable purposive interpretation: Michael Whincop and Mary Keyes, ‘Statutes’ Domains in Private International Law: An Economic Theory of the Limits of Mandatory Rules’ [1998] SydLawRw 18; (1998) 20 Sydney Law Review 435.

[60] Vita Foods Products Inc v Unus Shipping Co Ltd [1939] AC 277.

[61] [1998] 1 Lloyd’s Rep 90.

[62] Cf Whincop and Keyes, ‘Statutes’ Domains in Private International Law’, above n 59.

[63] (1996) 188 CLR 418.

[64] DA Technology Australia Pty Ltd v Discreet Logic Inc (Unreported, Federal Court of Australia, Gummow J, 10 March 1994) <http://www.austlii.edu.au/au/cases/cth/federal_ct/unrep6652

.html> at 30 August 2000 (copy on file with author); CE Heath Underwriting & Insurance (Australia) Pty Ltd v Barden (Unreported, Supreme Court of NSW, Rolfe J, 19 October 1994). Cf Leigh-Mardon Pty Ltd v PRC Inc [1993] FCA 324; (1993) 44 FCR 88, 99–100 (Beazley J).

[65] British South Africa Co v Companhia de Moçambique [1893] UKLawRpAC 53; [1893] AC 602. In the United States, see Massie v Watts, 10 US [1810] USSC 22; (6 Cranch) 148 (1810).

[66] Stephen Lee, ‘Jurisdiction over Foreign Land: A Reappraisal’ (1997) 26 Anglo-American Law Review 273.

[67] Michael Whincop, ‘Conflicts in the Cathedral: Towards a Theory of Property Rights in Private International Law’ (2000) 50 University of Toronto Law Journal 41.

[68] See, eg, Dagi v Broken Hill Proprietary Co Ltd [No 2] [1997] 1 VR 428. This case was filed in the Supreme Court of Victoria and included causes of action in trespass and nuisance to land located in Papua New Guinea. The plaintiffs claimed that this land had been affected by the discharge of by-products from the defendant’s copper mine. Byrne J dismissed these claims, applying the ‘Moçambique principle’ which provides that claims concerning possessory or proprietary rights to foreign land are non-justiciable. Other non-proprietary claims were not, however, struck out.

[69] Jurisdiction of Courts (Foreign Land) Act 1989 (NSW); Family Provision Act 1982 (NSW) s 11. The relevant question, however, is how often courts refuse to exercise the jurisdiction on forum non conveniens grounds. This question remains unanswered. The cross-vesting legislation also enables courts to adjudicate with respect to land in other states: Jurisdiction of Courts (Cross-Vesting) Act 1987 (Cth) s 5.

[70] [1936] 1 KB 382, 398 (‘St Pierre’). The principle applied to suits commenced regularly in the forum, not those relying on long-arm jurisdiction. As to former United States law, see Restatement (2nd) of the Law: Conflict of Laws (1971) §84 (stating a test alike Australia’s current standard).

[71] See Piper Aircraft Co v Reyno, [1982] USSC 28; 454 US 235 (1981); Voth [1990] HCA 55; (1990) 171 CLR 538; Spiliada [1986] UKHL 10; [1987] 1 AC 460; Amchem Products Inc v British Columbia (Workers Compensation Board) [1993]

1 SCR 897; (1993) 102 DLR (4th) 96.

[72] Oceanic Sun Line Special Shipping Co Inc v Fay [1988] HCA 32; (1988) 165 CLR 197, 251 (Deane J).

[73] Christopher Clarkson and Jonathan Hill, Jaffey on the Conflict of Laws (1997) 116–19.

[74] Australian law also seems to apply a similar doctrine of lis pendens: Henry v Henry [1996] HCA 51; (1996) 185 CLR 571.

[75] Garnett, above n 37, 48–9, makes an important point when he notes that the significance of reducing multiple proceedings differs between a case in which a defendant seeks a stay in order to pursue a claim for contribution against a third party, and a case in which the plaintiff joins local and foreign defendants and relies on this factor in resisting a stay. To recast this argument in the terms of the analysis in this paper, the first case raises an issue in which a balance must be struck between fairness to the original defendant and to the third party. In these cases, it is not really surprising that wealth maximisation provides a useful tie breaker. On the other hand, refusing a stay, as occurred in several Australian cases studied by Garnett, may be seen to favour the local defendant, and thus represents a form of unilateralism. In the second case, wealth maximisation is much more controversial, since it condones the plaintiff taking, as it were, the defendant’s political right to fairness because of the saving to the plaintiff in litigating its case in conjunction with another suit in which that defendant may have little or no role. I share Garnett’s reservations: see below Part VI(B)(1).

[76] Clarkson and Hill, above n 73, 118.

[77] (1870) 6 LR QB 1, 28–9.

[78] Spiliada [1986] UKHL 10; [1987] 1 AC 460.

[79] Ibid 477–8; Voth [1990] HCA 55; (1990) 171 CLR 538, 564–5 (Mason CJ, Deane, Dawson and Gaudron JJ).

[80] [1990] HCA 55; (1990) 171 CLR 538, 558 (Mason CJ, Deane, Dawson and Gaudron JJ).

[81] Garnett, above n 37, 43–4.

[82] Al-Ru Farm Pty Ltd v Hedleys Humpers Ltd (Unreported, Supreme Court of SA, Matheson J, 10 January 1991); Phosphate Co-operative Company of Australia Ltd v SGS Supervision Services Inc (Unreported, Federal Court of Australia, Gray J, 7 April 1993); WFM Motors Pty Ltd v Maydwell (Unreported, Supreme Court of NSW, Bryson J, 23 April 1993). The last two cases fit the paradigm.

[83] See above p 394.

[84] I am less concerned by the case Garnett identifies (Garnett, above n 37, 44–5) in which contacts were evenly split and a stay was refused: Shepherd Woolskins Pty Ltd v Bighorn Sheepskin Co (Unreported, Supreme Court of NSW, James J, 9 December 1993). Where costs are not likely to differ in total, and the plaintiff’s selected forum is not unfair, courts should deny stays because applications represent socially costly forms of strategic behaviour.

[85] See DA Technology Australia Pty Ltd v Discreet Logic Inc (Unreported, Federal Court of Australia, Gummow J, 10 March 1994); CE Heath Underwriting & Insurance (Australia) Pty Ltd v Barden (Unreported, Supreme Court of NSW, Rolfe J, 19 October 1994). Cf Leigh-Mardon Pty Ltd v PRC Inc [1993] FCA 324; (1993) 44 FCR 88.

[86] See the discussion in above n 75.

[87] Anti-suit injunctions are effectively irrelevant to interstate cases: Service and Execution of Process Act 1992 (Cth) s 21.

[88] British Airways Board v Laker Airways Ltd [1984] UKHL 7; [1985] 1 AC 58.

[89] Allendale Mutual Insurance Co v Bull Data Systems Inc, [1993] USCA7 1229; 10 F 3d 425 (7th Cir, 1993); China Trade & Development v MV Choong Yong, [1987] USCA2 1067; 837 F 2d 33, 40 (2nd Cir, 1987) (Senior Circuit Judge Bright (dissenting)):

[I]n this day of exceedingly high costs of litigation, where no comity principles between nations are at stake in resolving a piece of commercial litigation, courts have an affirmative duty to prevent a litigant from hopping halfway around the world to a foreign court as a means of confusing, obfuscating and complicating litigation already pending ... in this country.

[90] I adopt the classification schemata in Julian Wilson, ‘Anti-Suit Injunctions’ [1997] Journal of Business Law 424, although I do not analyse cases arising under the Brussels Convention or the Lugano Convention. Not every case fits the categories below. An example is National Mutual Holdings Pty Ltd v Sentry Corporation (1989) 22 FCR 209, in which the foreign proceedings were enjoined because they would interfere with local litigation by compelling a party to change its solicitor.

[91] Aggeliki Charis Compania Maritima SA v Pagnan SpA (The Angelic Grace) [1995] 1 Lloyd’s Rep 87, 96; Akai Pty Ltd v People’s Insurance Co Ltd [1998] 1 Lloyd’s Rep 90; Apple Corps Ltd v Apple Computer Inc [1992] RPC 70; Continental Bank v Aeakos SA [1994] 1 WLR 588.

[92] See, eg, Midland Bank plc v Laker Airways Ltd [1986] QB 689; CSR Ltd v Cigna Insurance Australia Ltd (1997) 189 CLR 345.

[93] Société Nationale Industrielle Aerospatiale v Lee Kui Jak [1987] UKPC 12; [1987] AC 871; Bank of Tokyo Ltd v Karoon [1987] AC 45; CSR Ltd v Cigna Insurance Australia Ltd (1997) 189 CLR 345.

[94] The incidence of these polarised cases might be reduced if judges took a more adverse approach to the entitlement of a party to seek a negative declaration: cf News Corporation v Lenfest Communications Inc (1996) 21 ACSR 553. Some polarised cases would remain, however, because of the possibility of cross-claims.

[95] [1981] AC 557 (‘Castanho’).

[96] (1997) 189 CLR 345.

[97] [1987] UKPC 12; [1987] AC 871.

[98] [1998] UKHL 12; [1999] 1 AC 119.

[99] Ibid 140–1.

[100] Adrian Briggs, ‘The Unrestrained Reach of an Anti-Suit Injunction: A Pause for Thought’ [1997] Lloyds Maritime and Commercial Law Quarterly 90.

[101] Anne-Marie Slaughter, Andrew Tulumello and Stepan Wood, ‘International Law and International Relations Theory: A New Generation of Interdisciplinary Scholarship’ (1998) 92 American Journal of International Law 367; Larry Kramer, ‘Rethinking Choice of Law’ (1990) 90 Columbia Law Review 277.

[102] Richard Posner, Economic Analysis of Law (5th ed, 1998) 30.

[103] Emanuel v Symon [1907] UKLawRpKQB 174; [1908] 1 KB 302. It does this by limiting recognition principally to judgments entered in forums in which the defendant resides and to which the defendant has consented or submitted: see Whincop, ‘The Recognition Scene’, above n 8.

[104] Whincop and Keyes, Policy and Pragmatism in the Conflict of Laws, above n 23, 101–5.

[105] The sorts of cases in which the lex loci delicti is least attractive are torts in which there is potential separation between the defendant’s act or negligence and the plaintiff’s injury, such as products liability. Here, the lex loci delicti is potentially manipulable and uncertain, which simultaneously decreases its efficiency and its fairness. I have already made a similar argument above: see text accompanying above n 50.

[106] See, eg, Joseph Beale, A Treatise on the Conflict of Laws (1935).

[107] See, eg, Walter Wheeler Cook, The Logical and Legal Basis of the Conflict of Laws (1942).

[108] See above n 75 and accompanying text.

[109] See, eg, Iain McLean and Arnold Urken, ‘Did Jefferson or Madison Understand Condorcet’s Theory of Social Choice?’ (1992) 73 Public Choice 445.

[110] Kenneth Arrow, Social Choice and Individual Values (2nd ed, 1963).

[111] ($80 + $90) – $100 = $70.

[112] See The Eleftheria [1970] P 94.

[113] Schwartz, ‘Proposals for Products Liability Reform’, above n 2.

[114] Alan Schwartz, ‘Relational Contracts in the Courts: An Analysis of Incomplete Agreements and Judicial Strategies’ (1992) 21 Journal of Legal Studies 271.

[115] [1991] USSC 56; 499 US 585 (1991). There, the United States Supreme Court enforced a standard clause by which the parties submitted to the jurisdiction of courts in Florida. The clause was contained in tickets purchased by consumers resident in Washington State for a cruise off the Mexican Coast ex-Los Angeles. The Shutes were incapable of litigating in Florida because of the substantial physical and financial burdens that would have been placed on them had they been forced to pursue suit in that forum.

[116] See Akai (1996) 188 CLR 418; DA Technology Australia Pty Ltd v Discreet Logic Inc (Unreported, Federal Court of Australia, Gummow J, 10 March 1994); CE Heath Underwriting & Insurance (Australia) Pty Ltd v Barden (Unreported, Supreme Court of NSW, Rolfe J, 19 October 1994). Cf Leigh-Mardon Pty Ltd v PRC Inc [1993] FCA 324; (1993) 44 FCR 88.

[117] [1981] AC 557.

[118] (Unreported, Supreme Court of Victoria, Harper J, 19 October 1993).

[119] See Garnett, above n 37, 39.

[120] See, eg, The Pioneer Container [1994] UKPC 5; [1994] 2 AC 324. There the Privy Council enforced an exclusive jurisdiction clause despite the fact that the suit was time-barred in the agreed jurisdiction.