A. Conceptualizing Choice of Law
What is choice of law? What are its goals, and how should we understand what goes on in a court’s choice-of-law analysis? These questions might seem abstract and general, but how they are answered will have important implications for our understanding of particular and concrete legal questions.
I. In General
Choice of law, in general, is typically understood as a method for allocating authority among co-equal sovereigns. Some cases are purely domestic—every relevant contact points to the same state, and that state’s law will govern the case. But some are multistate—different contacts or elements of the case point to different states. Conduct may occur in one state and injury in another, or the parties may reside in different states. In such cases, it is not clear which state’s law should govern the case, or particular issues within it. Choice of law tells us which state has regulatory authority for the case or the issue—it tells us which law will govern.
What kind of a determination is that? Some scholars seem to think of choice of law as a preliminary procedural analysis that is necessary to identify the rights the parties have but independent of that determination.Footnote 1 Before we can figure out what the parties’ rights are, this perspective maintains, we must identify the governing law. Then—and only then—we consult that law to ascertain what rights it gives the parties.
This understanding is sometimes called a jurisdiction-selecting approach, because rather than by considering laws—a process that would include looking at the content of those laws—it considers jurisdictions, or states, without looking at the content of their lawsFootnote 2 . I sometimes describe this approach by analogy to colored glasses. We can imagine a set of several pairs of glasses, with different tints. These are the laws of different states, and the legal relations between the parties will look different depending on which pair we put on. The process of choice of law is the process of picking the right pair, and once we have picked it, we look at the case—or the particular issue—through the lens of the chosen law. And, most adherents of this approach seem to think, once the governing law is identified, the case or issue should be resolved as if it were purely domestic—in the same way as if all relevant contacts were located in the state whose law was chosen.Footnote 3 In the United States, the foremost example of this approach is the territorialism of Joseph Beale and the First Restatement of Conflict of Laws.Footnote 4 Territorialism maintained that law was by its nature limited in scope. A state’s law reached everything that happened within its borders and nothing outside them. One and only one law would govern any legal issue, and the way to identify that law was simply to figure out where the relevant legal event occurred—where a tort was committed, or where a contract was formed, and so on.
A different view of choice of law, which I prefer, maintains that choice of law is in fact substantive and not procedural.Footnote 5 What determines the rights the parties have, on this view, is the laws that create or withhold those rights. Choice of law is a two-step process of first identifying the rights the parties have—not by applying whatever rules the forum likes but by interpreting the laws the parties invoke—and second resolving conflicts, if any, between those rights.Footnote 6 The fundamental choice of law question is not so much “Does state A law or state B law govern this issue?” but more “How should we resolve a conflict between a right under state A law and a contrary right under state B law?”
This is not the place to argue for one of those perspectives rather than the other. I think there is probably no right answer in a strong metaphysical sense. States are free to construct and understand their choice-of-law systems however they want, unless some higher law constrains them. I think that the analysis that works in terms of conflicts between rights tends to be more useful, illuminating, and analytically tractable. In the United States, the Supreme Court has recognized some constitutional limits on what states can do in the choice-of-law process that are easy to understand from that perspective but somewhat more mysterious otherwise.Footnote 7 Again, that does not mean that the resolution of conflicts between rights is what is “really” going on in some metaphysical sense. But it does mean that this perspective provides a more useful way to talk about constitutional limits and, I have argued, choice of law more generally. It is also the perspective adopted by the Third Restatement of Conflict of Laws, currently in progress.
II. In Contract Cases
The perspective that takes choice of law to be about conflicts between rights might seem most plausible with respect to tort law, or noncontractual claims. There, it is easy to see how one party might have a claim under the law of state A—under state A law, the facts alleged support a cause of action—while the other might have a defense under state B law—under that law, the facts alleged do not support a cause of action, or they provide an affirmative defense against liability. It might be harder to see how this perspective works with respect to contracts, where law operates to let parties make enforceable promises, rather than to restrict their behavior with the threat of liability.
But contracts can be understood within the same framework. Generally speaking, what is at issue in a contractual choice-of-law case is whether the contract or a particular term is enforceable. One state would allow enforcement—that is, it gives the plaintiff a right to demand performance or recover for breach. The other state would not allow enforcement—it authorizes the defendant to decline performance without being liable for breach. Conceptually, we have the same issue of a claim under one state’s law conflicting with a defense under the law of another state. And we have the same question of how to resolve that conflict: Which state’s law should be given priority?
How we answer that question depends on what we want a choice-of-law system to do. Generally, I divide choice-of-law desiderata into two different categories, which I call “right answer values” and “systemic values.”Footnote 8 Right answer values are the things that make the answer to a choice-of-law question correct, or at least sensible, rather than arbitrary. People will, of course, disagree on what makes an answer correct, but in the United States there is probably a general consensus that it has something to do with the policies of the relevant states and the expectations of the parties. It is good, that is, to resolve a case in a way that helps states achieve what they are trying to achieve with their laws, or at least does not needlessly frustrate those goals. It is good to resolve a case in a way that does not surprise the parties, subjecting them to a law they did not think would govern their relationship. It is good, maybe, to find a solution that advances the shared policies of the states, or the basic policies underlying the relevant field of law. It would be bad, by contrast—it would be arbitrary—to select the governing law by flipping a coin, or on the basis of which state’s name came first alphabetically.
Systemic values, by contrast, are less about which state’s law is chosen and more about how that law is chosen. Systemic values include simplicity, uniformity, and predictability. It is good, from this perspective, to have a system that is easy to apply—that reduces the costs of litigation. It is good to have a system that gives the same answer regardless of where suit is brought—that reduces forum-shopping. It is good to have a system whose answers can be predicted—that gives greater certainty to parties planning their conduct. Flipping a coin to identify governing law is still bad from a systemic perspective. Picking the state whose name comes first alphabetically is actually pretty good from a systemic perspective—the fact that it is obviously unacceptable shows us that, at some point, the cost to right answer values grows too high. Conversely, a very complicated analytical method may do a very good job of finding the right answer but impose unacceptable systemic costs. The goal of choice of law, then, is to maximize the combination of right answer and systemic values while not straying too far in either direction.Footnote 9
That is the perspective I will be using for the remainder of this Article. Choice of law is a method of determining which state will have the authority to specify the legal consequences of events that touch more than one state. It does this by identifying the rights that the parties have under different states’ laws and, if necessary, resolving conflicts between those rights. And in so doing, it should produce answers that are sensible in terms of the policies of the relevant states and the expectations of the parties, and it should produce those answers in a simple, uniform, and predictable way. That is the general framework. It is time now to turn to the more specific context of choice of law clauses.
B. Why Party Autonomy (and Why Not)
Why should states allow parties to choose the law that governs their relationship?Footnote 10 One answer might be that we value party autonomy more or less for its own sake. Allowing parties to choose the law that governs their relationship gives them the opportunity to experience life under different legal regimes. It enhances their freedom. Choice of law, from this perspective, is not about allocating authority among co-equal sovereigns, but more about empowering individuals.Footnote 11 In a similar vein, one might argue that party autonomy allows parties to avoid inefficient laws. Again, choice of law appears not as a method of allocating authority among states but rather a device for promoting a particular substantive value—in this case, efficiency.Footnote 12
But as I have said, that is not the perspective from which I am working. I think of choice of law as a method of giving regulatory authority over a particular issue to the most appropriate law, where “most appropriate” is understood to contain some combination of what I have called right answer and systemic values. From that perspective, it is easy to see what the justifications for party autonomy are—and also what its limits should be. Thus, this discussion may support my claim that thinking about choice of law from this perspective helps us think about it clearly.
Party autonomy can promote right answer values to at least some degree. Party expectations, most people think, or at least justified expectations, are a relevant factor to consider in choosing law. And this is perhaps particularly true in contract cases, where the point of a contract is to set out the rules for the parties’ relationship. Parties entering into a contract typically place considerable importance on knowing what they are agreeing to, and choice of law matters to that, because it can determine whether particular terms of the contract, or the whole thing, are enforceable. Allowing the parties to choose the law that will govern does help protect their expectations.
But party autonomy can also be threatening to right answer values. In the typical case, a contract will be enforceable under one state’s law—that state, we could say, has a policy in favor of enforcement or broad contracting power. It will be unenforceable under another state’s law—that state believes that certain contractual terms should not be honored, often because they are unfair or oppressive to a party who might not have had a chance to negotiate them, or who lacks equal bargaining power. In the absence of a choice-of-law clause, American courts would usually attempt to identify which of those states has the more significant relationship to the issue.Footnote 13 A choice of law clause might select the law of the state with the lesser interest or the less significant relationship—that is, it might get the question wrong in terms of the policies or relative interests of the relevant states. And so, looking at right answer values on the whole, party autonomy might actually undermine them.
The other justification for party autonomy is that, regardless of its effect on right answer values, it does a very good job of promoting systemic values. “Select the law the parties chose” is a very simple rule, and it is uniform and predictable. So the main justification for party autonomy, within the framework I have set out, is that it always advances systemic values and will only sometimes undermine right answer values.
Thinking about party choice in terms of the tradeoff between right answer values and systemic values helps us understand when it should be allowed and when it should not be. Briefly, the answer is that there should be some limits to ensure that it does not go too far in terms of undermining right answer values. What is the right answer, in the absence of a choice-of-law clause, can be a close and difficult question. In that case using party autonomy as a tiebreaker makes a lot of sense—it advances systemic values with only a slight cost in terms of right answers. But if the right answer is clear, or important state policies are at stake, so that the parties’ contrary choice imposes high costs, then it should not be respected.
This brief answer can be expanded. First, most fundamentally, this perspective explains why parties should be allowed to choose only the law of a state with some connection to the transaction. Connection to the transaction gives a state justification for extending its law to govern the transaction. If a state has no contacts with the parties or the transaction, its law will ordinarily not include the transaction within its scope. It will not create any rights, and thus there is no possibility of giving that law priority over another conflicting law. At least, that is normally the case, and in the United States the rule that a state’s law does not reach events having to contacts with the state is to some degree required by the United States Constitution. Some states have enacted laws authorizing parties to choose their law even for contracts that have no connection to the state, and these are discussed in more detail below, when I talk about what states say about the availability or non-availability of their law.
In the absence of legislative indication that a state’s law is supposed to reach contracts having no connection with the state, then, it should be impossible for parties to select that state’s law.Footnote 14 They cannot choose to be governed by a law that does not reach them. They can still, if they so desire, attempt to incorporate the content of that law into their contract—so they can attempt to provide by contract for the same rights that they would have under the law if it reached them. But, importantly, their ability to do so is limited by whatever law does govern the contract, and if the incorporation is successful, the rights of the parties are contractual and not statutory in nature. They would not, for example, be subject to criminal or administrative penalties for violating an incorporated rule, even if such penalties would be imposed for violating the relevant state law, and they would not be entitled to bring suit in specialized tribunals to enforce contractual rights, even if they would be entitled to do so for statutory rights. These points are considered in more detail below.
Selecting the law of a state that has no contacts with the transaction is the extreme example of parties choosing the wrong answer to the choice-of-law question presented by the contract. But even less extreme examples may be wrong enough that they should not be allowed. At some point, the cost in terms of right answer values is unacceptably high, and party autonomy should be restrained. As the next Part discusses, most choice of law systems follow these principles.
C. State Policies (or Interests)
When should the parties’ choice of law not be respected? I’ve said that the answer to this is when it generates an answer that is sufficiently wrong to outweigh the gains that party autonomy brings in terms of systemic factors.Footnote 15 And an answer is wrong, I’ve said, if it selects the law of a state with no contacts to the parties or the transaction, or if it selects a law that would not otherwise govern and use of that law is sufficiently offensive to the policies of the state whose law would otherwise govern.Footnote 16 All of this can be explored at greater length, and the following sections do that. But it might be useful, as a preliminary, to say something about state policies and interests.
I have been told, although I am not sure how accurate it is, that Europeans often react with bemusement to American talk of state interests. When a case involves two private parties, Europeans supposedly ask, how can it affect the interests of a state? The importance that Americans place on state interests, then, is often supposed to mark a difference between American and European approaches to choice of law.
I find this a little hard to believe, because European codifications do seem to reflect the same kind of policy analysis that in the United States is carried out in terms of state interests.Footnote 17 But it might be worth spending just a moment to clarify what it means to say that a state has an interest in a case or an issue. Generally speaking, the point is that states enact laws for particular purposes. State lawmakers might think that certain behavior is socially undesirable and should be deterred, and they might try to do that by making people who engage in that behavior pay for injuries they cause as a result. They might think that certain contractual terms are oppressive or unfair and should not be enforced. When Americans talk about a state having an interest, all they mean is that using the state’s law will advance the purposes or policies behind that law.
Tort law, for instance, is generally understood to have both a deterrent and a compensatory purpose. It is supposed to deter behavior that is considered socially undesirable—meaning, roughly, that its costs exceed its benefits. And it is supposed to compensate people injured by such behavior, on the grounds that its costs should be borne by tortfeasors, rather than victims.
When are these policies advanced? American policy analysis will generally say that the relevant contacts for the deterrent purpose are territorial. States intend to regulate conduct that occurs within their borders, or that causes injury within their borders. The deterrent purpose of New York tort law, for instance, would be implicated if there is conduct within New York, or conduct anywhere that causes injury in New York. The relevant contacts for the compensatory purpose are typically personal—that is, they relate not to where conduct or injury occur but to who is injured. The compensatory purpose of New York tort law would be implicated if a New Yorker is injured, no matter where the injury occurs. New York is interested in the application of its tort law, an interest analyst would say, in cases in which conduct or injury occurs in New York, or in which the injured party is from New York. “From” here means whatever kind of connection the state thinks is relevant—typically domicile or habitual residence.
It may well be true that this personification of states is metaphysically inaccurate. States are not people, and they do not have interests or desires in the same way people do. But the point of the discussion above is not to make philosophical claims about state identities; it is to identify situations in which it makes sense to select a state’s law to govern an issue, or at least to consider that law as a possible candidate for selection. The personification of states is simply a means to that end, and it should be assessed not in terms of whether it is consistent with whatever metaphysical notions of state identity we might have but rather whether it is a useful way of talking for that purpose. The practice of choice of law in American courts using policy analysis suggests to me that it is—policy analysis typically avoids the arbitrary results that sometimes occur under jurisdiction-selecting approaches. Thus, in the following sections, I will go on to talk about state interests and policies as a convenient shorthand. Everyone agrees, I think, that states enact laws for particular purposes, and some sets of facts implicate those purposes more than others. That is all that policy analysis claims.
I. The Forum
The policies of the forum, or at least some of them, are always relevant to a choice-of-law analysis. For one thing, because litigation occurs in the courts of the forum, the forum has an interest in the use of its procedural rules. Those are the rules it has adopted to govern litigation in its courts, attempting, presumably, to increase the accuracy and efficiency of adjudication, understanding that sometimes these values, like systemic and right answer values, must be traded off against each other.
Procedural rules are not really relevant in our analysis of choice-of-law clauses, though. It is normally assumed that parties intend to select the substantive law of a state, rather than its procedural law, so the issue of whether they could select procedural law if they wanted to does not arise. Instead, the relevant forum policies tend to be those embodied in its substantive law.
This is not usually because forum law would otherwise govern the issue—it is not because the forum has the closest connection or most significant relationship or whatever connecting factor forum choice-of-law rules deem controlling. I deal here with what is sometimes called the “forum qua forum”—that is, the interests that the forum has merely by virtue of being the forum and not because of any other contacts. There are two distinct, though related, sets of interests that the forum has simply because it is the forum. The first is its mandatory overriding rules; the second is its public policy.
1 . Overriding Mandatory Rules
An overriding mandatory rule is a rule that cannot be varied by agreement—mandatory—and that overrides ordinary choice-of-law analysis, including a choice-of-law clause—overriding.Footnote 18 It directs courts to select a particular law, usually forum law, to govern particular issues when particular contacts are present. In the United States, for instance, several states have statutes providing that insurance contracts payable to state residents or insuring property within the state shall be governed by the state’s law, regardless of whether the parties select a different law via a choice-of-law clause.Footnote 19 These statutes bind the courts of the forum, but not other courts. The forum’s contacts with the case matter to the extent that such statutes specify particular contacts that trigger the required choice of law: If the insured party is a forum resident, if the insured property is located within the forum state, and so on. Somewhat less often, a state’s choice of law rules may override party choice regardless of contacts with the forum. An example is the Rome I rule—Article 6, Section 2— providing that party choice may not be used in a consumer contract to deprive the consumer of nonwaivable protections of his home law.Footnote 20
2. Public Policy
More often, the relevant policy is the general public policy exception. This traditional rule allows a forum to refuse to recognize or enforce foreign law when doing so would sufficiently offend the forum’s public policy.Footnote 21 Typically, states require a degree of offense that goes well beyond mere disagreement—if disagreement were the test, a state would never select foreign law that differed from its own, undoing the whole enterprise of choice of law. In the United States, a commonly-quoted formulation is that of Judge and future Justice Benjamin Cardozo in Loucks v. Standard Oil Co.: Courts “do not close their doors unless help would violate some fundamental principle of justice, some prevalent conception of good morals, some deep-rooted tradition of the common weal.”Footnote 22
As the phrase “close their doors” indicates, invocation of the traditional public policy exception led courts to dismiss the suit rather than to select a different law. In modern cases, courts are more likely to decide the case or issue under a different law.Footnote 23 In those cases, contacts tend to play a larger role. In courts that still follow a territorialist approach, the public policy exception sometimes operates simply as a means to reach the outcomes that modern policy analysis would prescribe, such as selecting the law of shared domicile for loss-allocating issues.Footnote 24 In other cases, the likelihood of a court’s invoking the public policy exception seems to be affected not merely be the repugnance of the foreign law but by repugnance considered in light of the contacts between the forum and the case.Footnote 25
II. The Non-Selected State Whose Law Would Otherwise Govern
The policies of the forum qua forum will place limits on the parties’ ability to select governing law for three reasons. First, the forum has an interest in the procedures that its courts follow and will apply its own procedural law.Footnote 26 Second, there may be relevant forum choice-of-law statutes, which will bind the court to select a particular law.Footnote 27 Third, the general public policy exception may operate to prevent the court from enforcing the parties’ choice.Footnote 28
All of these limits are intelligible within the framework I have developed. Procedural issues typically come within the scope of forum law and not foreign law—that is, I think, the best understanding of what it means to call an issue procedural.Footnote 29 Choice-of-law statutes direct courts to select a particular law; that is, they identify that law as the right answer, in the view of the legislature. They also usually do so in a clear and simple way, although results may not be uniform if other states prescribe different results. And the public policy exception represents the forum’s view that an answer is so egregiously wrong—because the forum policy is so strong, in light of the forum’s contacts with the case—that it cannot be allowed.
The framework also explains the role played by the policies of the non-selected state. Suppose that we were trying to identify situations in which party choice strayed too far from the right answer. First, of course, the answer the parties chose would have to be different from the one our ordinary choice-of-law analysis directed. So we would test the parties’ choice against the law that would otherwise govern—what ordinary choice-of-law analysis identifies as the right answer.
Second, we would want to identify not just a wrong answer but an egregiously wrong one. There are two ways of thinking about what makes an answer egregiously wrong, which we can call interests and policy significance. From the interests perspective, an answer is wrong to the extent that it picks a state with a lesser interest, or a less significant relationship. It is egregiously wrong if that state’s interest is lesser by a lot. From a policy significance perspective, an answer is wrong to the extent that it undermines the policy of the state with the greater interest. It is egregiously wrong if that state’s policy is very important. While the assessment of interests will, in the main, be the same for all states—certain contacts will generally implicate certain interests in the same way for all states, as described in the discussion of tort law above—policy significance can, at least for this purpose, be thought of as subjective: One state might decide that its policy about protecting consumers is fundamental and essential, while another state with the same policy might consider it relatively unimportant. So to articulate limits on party autonomy, we would want to say something about interests and policy that captured the two senses in which an answer might be wrong.
The Second Restatement did this, although not necessarily in a perfect way or one that is easy to apply.Footnote 30 The non-selected state can effectively veto the parties’ choice of law, under the Second Restatement, if two criteria are met. First, the non-selected state must have “a materially greater interest”—so that the parties’ choice is egregiously wrong in terms of interests.Footnote 31 Second, the policy at stake must be fundamental, so that the outcome is egregiously wrong in terms of policy.Footnote 32 Neither “materially greater” nor “fundamental” is easy to articulate in an objective way. Generally, however, the inquiry should be understood as similar to the one the legislature undertakes in deciding whether to write an overriding mandatory rule: Given certain policies and certain contacts, is it sufficiently important to the state that its law govern a particular issue?
Rome I also operates in roughly this way.Footnote 33 It does have a general public policy exception in favor of the forum—Article 21—and also an exception for the overriding mandatory rules of the forum—Article 9.Footnote 34 But rather than a general exception in favor of the non-selected state, it sets out particular circumstances under which, effectively, there is an overriding mandatory rule directing selection of a particular state’s law—individuals in employment contracts, for instance, and consumers, are both entitled to nonwaivable protections of particular laws, as evident in Article 6(2) and Article 8(1). Rome I may not think of these as overriding mandatory rules, because it defines such rules in terms of the importance of policy: “[O]verriding mandatory provisions are provisions the respect for which is regarded as crucial by a country for safeguarding its public interests, such as its political, social, or economic organisations.”Footnote 35 But as a functional matter, any nonwaivable rule that directs courts to select a particular law works as an overriding mandatory rule, and it seems both difficult and pointless to speculate about which statutes are motivated by a sense that the policy is crucial in this sense. The legislature has decided which law shall govern; that is all courts need to know. Thus, I understand Rome I’s set of rules requiring the application of a particular state’s law as an attempt at a uniform expression of the circumstances under which a state has a materially greater interest and a fundamental policy.
III. The Selected State
Thus far, we can see that American and European approaches to party autonomy are broadly similar. Indeed, some degree of respect for party autonomy, and some degree of convergence with respect to the limits on that autonomy, are universal—or at least worldwide—features of choice-of-law regimes.Footnote 36 The vast majority of systems give parties some freedom and limit that freedom by reference to the law of the forum and the law of the non-selected state.
But what about the selected state? Here, I think, the United States may be ahead of the field. Or it may be that other countries have found the issues I will discuss easy enough to resolve that there has not been the kind of controversy that has occurred in the U.S. I have been surprised by the amount of disagreement over these issues. But whatever the explanation, I think that American scholars talk more about the law of the selected state.
When we think about party autonomy in choice of law, we can imagine states saying three basic things—again, personification is for the purposes of simplicity and clarity. First, a state might say that its law should control even if the parties do not want it to. That is the import of overriding mandatory rules and public policy exceptions. Second, a state might say that its law is available for the parties to select. We presume, I think, that states that have relevant contacts to the parties or the transaction make their law available in this way—to put it in the terms the Third Restatement uses, states extend the scope of their contract law to transactions that have relevant contacts with the state, which means that party choice can, at least sometimes, be used to give priority to that law.Footnote 37 And it might say further that its courts will honor the parties’ choice in some circumstances, even if the courts of other states would not. This, as I will discuss below, is the import of an inbound choice-of-law statute. Third and last, it might say that its law is not available for selection in some circumstances. This is the issue that has received the most discussion in America recently, and which will be my focus.
1. Making State Law Available: Inbound Choice-of-Law Statutes
We have already considered attempts by states to ensure that their law will govern an issue even if the parties do not choose it, and even if ordinary choice-of-law analysis would not choose it either. Overriding mandatory rules and public policy exceptions all function this way. Attempts by states to make their law available to parties tend to receive less attention, but the issue is important. If we think that there are some limits on which laws the parties may choose, then we should consider the extent to which states may lift those restrictions.
The main such limit is typically a requirement that the state whose law is chosen have some connection to the parties or the transaction. This requirement seems very sensible to me, as I have noted already, and we can explain it in a few different ways, consistent with the framework developed earlier. First, thinking in terms of rights under state law, I have said that party choice can be used as a method to decide which of a set of conflicting rights will be given priority. Party choice cannot, however, give priority to rights that do not exist, so if a state’s law gives a party no rights, party choice cannot change that. Second, thinking in terms of state policies or interests, I have said that party choice is an appealing tiebreaker between competing state claims of regulatory authority. But this framing likewise depends on the idea that there is a tie to break. If only one state’s law reaches the transaction, there is no tie.
Despite this, many jurisdictions go farther in the direction of party autonomy. The Second Restatement allows party choice based on either a “substantial relationship” to the parties or the transaction, or some other “reasonable basis” other than a substantial relationship.Footnote 38 The black letter of the Second Restatement does not restrict this analysis to multistate cases, so it apparently recognizes at least some ability of parties to choose governing law in a purely domestic case.Footnote 39 Rome I does not expand contracting power in purely domestic cases: If all relevant elements are located in a country other than the one whose law was chosen, mandatory provisions of that other country’s law may not be altered. But in multistate cases, it does not require a connection to the state whose law is chosen—at least, not for most contracts.Footnote 40
This expansive view of party autonomy strikes me as questionable. I think that for parties to choose a law to govern their contract, rather than merely to incorporate the terms of the law into their contract, the law must reach them of its own force. I do not think—and this point will be developed further below—that parties can expand the scope of state law and make it reach persons or events the state says it does not. In the absence of any indication to the contrary, American courts assume that state laws do not reach events that have no connection with the state. Such indications can be given: Some states have enacted laws, so-called inbound choice-of-law statutes, that provide that parties may choose the state’s law to govern even transactions with no connection to the state, as long as they meet some minimum dollar-value requirement. With an inbound choice-of-law statute, the choice of law is supported not just by party choice but by the active cooperation of the state. How much of a difference that makes will be explored below.
In the ordinary case, the rule that a state’s law does not reach cases with which the state has no contacts has at least some constitutional force. The United States Supreme Court has explained in Allstate Insurance Co. v. Hague that for a state’s law to be “selected in a constitutionally permissible manner, that State must have a significant contact or significant aggregation of contacts, creating state interests, such that choice of its law is neither arbitrary nor fundamentally unfair.”Footnote 41 It seems, then, that use of a state’s law to govern a contract with which the state has no contacts is constitutionally suspect.
Unpacking the Allstate test allows us to explore the question in more detail. Allstate consolidated two different constitutional provisions, the Due Process Clause of the Fourteenth Amendment and the Full Faith and Credit Clause of Article IV.Footnote 42 These two clauses serve different functions and protect different values, so it makes more sense to think about them separately.
Due Process is about the relationship between states and individuals. It places a limit on the coercive power of states—a limit on the scope of legislative or prescriptive jurisdiction. A state’s law may not reach persons or events that have no contacts with the state. Similar restrictions on legislative jurisdiction exist under international law. The primary reason the Supreme Court has given for this rule is unfair surprise: Parties who have no contact with a state cannot be assumed to have notice of its law, and it is unfairly surprising to subject them to a law whose application they could not have foreseen.Footnote 43 Allstate captures the Due Process requirement by saying that the state whose law is chosen “must have a significant contact or significant aggregation of contacts….”Footnote 44
Full Faith and Credit is different. The Full Faith and Credit Clause provides that “[f]ull faith and credit shall be given in each state to the public acts, records, and judicial proceedings of every other state.”Footnote 45 This clause is about the relationship between sister states of the American union, and it requires states to respect the laws of other states. In the choice-of-law context, the Supreme Court has explained that it means that a state cannot select its law to govern an issue unless the state has some interest. Allstate captures this requirement by saying that the contacts must “create[e] state interests.”Footnote 46
Is it constitutionally permissible to allow the parties to choose the law of a state with no relation to the parties or the transaction? If we think of Due Process in terms of unfair surprise, the requirement is easily satisfied. Neither party can claim to be unfairly surprised by the application of a law they selected. So a choice-of-law clause seems adequate to overcome Due Process objections.
If we think of Due Process in terms of the permissible scope of state law, the issue is murkier. It violates Due Process, from this perspective, to determine the parties’ rights by reference to a law that does not reach them.Footnote 47 Party consent does not change that. If the state does not intend its law to reach the contract—and if there are no contacts with the state, I think that lack of intent should be presumed—then the law does not reach it and cannot be selected.
An inbound choice-of-law statute, however, can change this. Such a statute indicates that the state intends to extend its law to reach the transaction despite the lack of contacts. Ordinarily, the state lacks the power to do so, but I believe party consent can overcome that barrier. Legislative jurisdiction is not limited, as people used to think, by the borders of the state.Footnote 48 Nor is it limited, in any metaphysical sense, by the requirement of contacts: If parties consent and the state so desires, I think that legislative jurisdiction can include transactions with no connection to the state.
But what about Full Faith and Credit? The interest being protected there is not an interest of the parties, but of sister states. It is not waivable by the parties. If a contract is entirely internal to New York, for instance, New York has an interest in regulating that contract. New York might have a minimum wage or maximum hour policy, or it might want to protect consumers, or some other weaker party. No other state should be able to override New York’s policy. Similarly, if the contract has contacts with only New York and New Jersey, Massachusetts should not be able to override their regulatory schemes.
In constitutional terms, the question would be whether, in this second hypothetical, Massachusetts has an interest sufficient to meet the Full Faith and Credit test. If Massachusetts has no contacts and no inbound choice-of-law statute, then the answer is relatively clearly no: Massachusetts is not trying to interfere with the regulatory authority of New York and New Jersey, and the parties should not be able to bring its law in against its wishes. If Massachusetts does have an inbound choice-of-law statute, then it has expressed an intent to reach the transaction. But does that give it an interest?
This requires us to decide what counts as a state interest for Full Faith and Credit purposes, an issue the Supreme Court has never explained. It is always possible to come up with reasons that deciding a case under a state’s law benefits the state—it might create employment for local lawyers, or generate precedents that help develop the law. But clearly some interests cannot count, or else the requirement of an interest means nothing. In Phillips Petroleum Co. v. Shutts, for instance, the Supreme Court suggested that procedural interests—the convenience of deciding many claims in a class action format—could not justify the choice of Kansas substantive law to govern claims that had no connection to Kansas.Footnote 49 My tentative guess is that an interest must be related to the purposes of the substantive law the court seeks to apply. The question, then, would be whether it advances the purposes of Massachusetts contract law, or whatever other law, such as franchise law, that might be selected by the choice of law clause, to apply it to a contract with no contacts to Massachusetts. My view—which again is tentative—is that it does not, and that therefore Massachusetts law should not be allowed to displace mandatory terms of the otherwise-governing law.
The argument seems even stronger if the otherwise-governing law embodies a fundamental policy of the enacting state, so that an ordinary choice-of-law clause would be invalid under the test of the Second Restatement.Footnote 50 It is not clear that all inbound choice-of-law statutes are intended to displace the “contrary to a fundamental policy of the state with a materially greater interest whose law would otherwise govern” element of the Second Restatement’s test.Footnote 51 Some explicitly so provide—North Carolina law, for instance, states that the parties to a “business contract” may agree that North Carolina law governs “whether or not” there is a reasonable relation to the state or the contract is contrary to the fundamental policy of the jurisdiction whose law would otherwise govern.Footnote 52 New York’s statute, by contrast, explicitly lifts the “reasonable relation” requirement but does not mention the fundamental policy exception.Footnote 53 At the least, I think, statutes that are silent about the fundamental policy exception should not be read to override it, in light of the constitutional questions such an interpretation would raise.
I conclude, then, that while a state may extend its law to contracts that have no contacts with the state, allowing the parties to choose that law, it should still respect the fundamental policies of the state whose law would otherwise govern. In the United States, as far as relations between sister states are concerned, I believe there is a constitutional requirement that operates at some point.Footnote 54
1. Making State Law Unavailable: Territorial Limits
In the absence of specification, I have said, states should generally be presumed to make their law broadly available for selection in a multistate case. A state’s tort law, for example, should probably be presumed to be available for all cases involving conduct or injury in the state, and for cases involving conduct and injury outside the state if a resident is injured. A state’s contract law should be available for all contracts having a reasonable relation to the state. These are default rules about the scope of state law, but of course states can override them with explicit statements.
The preceding section has considered one way in which states may specify the scope of their law. Inbound choice-of-law statutes make clear that the state’s law is intended to reach certain transactions even in the absence of a reasonable relation to the state. States sometimes enact statutes that make clear that the state’s law is not intended to reach certain events: They limit the scope of their law.
Most often, these limits take the form of an “in this state” restriction. A wrongful death statute might create a cause of action for deaths caused “in this state,” a franchise statute might set out rules for franchises “in this state,” or a wage and hour law might set terms for work performed “in this state.” Courts confronting such limits in tort cases have reached what seems the obviously correct answer: If the statute says that a particular case does not fall within its scope, the statute is not relevant and cannot be selected through a choice-of-law analysis. As the Wisconsin Supreme Court put it in just such a case, Waranka v. Wadena, because the Wisconsin wrongful death statute is limited to deaths caused in Wisconsin, “we need not undertake a conflict of laws analysis” with respect to a death caused in Michigan.Footnote 55 Only Michigan law is relevant.
But can a choice-of-law clause change this? According to the framework I have developed, the answer should be no. A choice-of-law clause can determine which set of rights gets priority, if rights conflict. It can break a tie between dueling state claims of regulatory authority. But it cannot alter the scope of state law. It cannot create rights that a state intends to withhold. In the Waranka case mentioned above, the parties might have a contractual relationship, and they might provide that all claims arising from the relationship are to be governed by Wisconsin law.Footnote 56 But that will not extend the Wisconsin wrongful death statute to a death caused in Michigan.
Again, American cases are essentially uniform on this point. Cotter v. Lyft, Inc. is illustrative.Footnote 57 In that case, drivers for Lyft attempted to bring a nationwide class action under California wage and hour laws.Footnote 58 As interpreted by California courts, those laws do not reach work performed exclusively outside California.Footnote 59 The plaintiffs argued that because all the relevant contracts selected California law, the choice-of-law clause allowed workers in every state to brings claims under the wage and hour statutes.Footnote 60
The Cotter court rejected that argument:
A court conducts a conflict of laws analysis only where the laws of multiple states could conceivably apply to the same claim. Where only one state’s law applies, no such analysis is necessary. And…the California wage and hour laws asserted here simply do not apply to employees who work exclusively in another state.Footnote 61
From a policy perspective, this might seem strange. Analysis of choice-of-law clauses, as I’ve said, tends to focus on the policies of the state whose law is not chosen. That is natural—a court that declines to apply a state’s law runs a risk of thwarting that state’s policies. And it also seems natural to suppose that applying a state’s law does not run such a risk.
That is true, I think, if the law is applied as written. We could imagine “applying” California wage and hour law to work performed in Arizona in this sense. We would ask what rights the law gave the parties, and the answer would be none—California wage and hour law simply does not reach work performed in Arizona, so it gives the plaintiff no claim and the defendant no defense. Applying California law to conduct outside its scope is the same as applying no law at all.
But that was not what the plaintiffs wanted. They wanted to erase the territorial restriction and apply a version of the statute that was not what the legislature wrote. That is changing the meaning of state law, and doing that certainly can thwart state policies.
Brief reflection reveals several reasons that states might want to limit the scope of their law. First, most generally, they might do so out of deference to other states. The California legislature might believe that some things that happen outside the borders of California are simply not California’s business. Some other law should govern; California law has nothing to say about the matter.
But, one might object, how does it affect California’s interest if its law is given greater scope? Using California law to govern work performed in Arizona might not advance any California policy, if the California legislature wants to restrict rights under California law. But how can it harm California policy if Arizona courts want to use California law? California can defer to Arizona in its own courts, but can’t Arizona courts decide that Arizona does not want that deference?
One answer—which I think is sufficient—is that this is a decision for the legislature. They can limit the scope of their law if they want, and if that’s what they have done, no other state has the power to contradict them on it. Whether a state statute gives rights to a particular person in particular circumstances is a question of the content and meaning of that statute. Legislatures get to make state law; they have power over its content and meaning. Courts of other states do not. Neither do parties using a choice-of-law clause, and those parties, notably, may have interests distinct from both California and Arizona. Even if we trust Arizona courts to decide when Arizona law should be displaced by California law, that is, we have no similar reason to trust private parties.
But if that is not enough, there are in fact some clear policies that could be thwarted by extending the scope of a state’s law. First, there is the problem presented by Cotter itself. If territorial limits in a state’s wage and hour or consumer protection laws can be overridden by a choice-of-law clause, then local corporations will apparently subject themselves to nationwide class actions merely by selecting their home law to govern contracts with employees or consumers. This prospect has been considered by courts, and almost uniformly rejected. As the New York Court of Appeals put it in Goshen v. Mutual Life Ins. Co., it would subject corporations to “a tidal wave of litigation…not intended by the Legislature.”Footnote 62
Second, there is the problem that parties with rights under a state’s law are often entitled to litigate in specialized tribunals or to pursue administrative remedies. California might, for instance, have a Workers Compensation Commission that investigates and adjudicates claims under its wage and hour laws. Should workers in Arizona be entitled to call for an investigation and bring their claims before the Commission, against the wishes of the California legislature, simply because they have a choice-of-law clause selecting California law? Presumably not.
The clearest illustration of this point involves the limited subject-matter jurisdiction of federal courts in the United States. Federal courts can hear cases in which the parties invoke rights under federal law, what is known as federal question jurisdiction. What if parties outside the United States, and hence outside the scope of a federal employment statute, include a choice-of-law clause selecting federal law? Do they now have rights under the statute that support federal question jurisdiction? That question has been raised in federal court, and the answer is no.Footnote 63 It is clear that parties cannot create federal question jurisdiction by agreement.
American cases, as I’ve said, are essentially uniform on this question, and they are correct according to the framework I have developed. It should be understood, however, that while the parties cannot extend the scope of a state’s law, they can incorporate the terms of that law, if doing so is within their contracting power. Because the state whose law is being incorporated does not intend to expand their contracting power, I believe it should be the same as it would be in a purely domestic case. That is, parties can incorporate the terms of a state law that excludes them from its scope to the same extent as they could write those terms explicitly into their contract. And what that means is that they cannot displace mandatory provisions of the law that would otherwise govern.
The distinction between extension of scope and incorporation explains most of the cases, and most of the cases explicitly recognize the distinction. So do at least some commenters.Footnote 64 Keeping it in mind allows us to handle the puzzles presented by choice-of-law clauses that select scope-limited laws. Consistent with this distinction, Section 8.04 of the draft Third Restatement provides as follows:
If the parties choose a state’s law to govern their contract or relationship and some parts of the chosen state’s law contain limits on scope that exclude some or all of the relevant issues, the choice is ineffective as to those issues. Issues outside the scope of the selected law will be governed by the law that would govern in the absence of an effective choice of law. Courts may interpret a reference to a scope-limited law as an attempt by the parties to incorporate the substance of that law into their agreement.Footnote 65
This section is designed to recognize the basic fact that parties cannot use a choice-of-law clause to extend the scope of a state law, for the reasons noted above. Technically, as also noted above, it follows that if the parties have selected a scope-limited law to govern an issue outside its scope—if they have selected California wage and hour law to govern work performed in Arizona, for instance—they have chosen a legal regime that gives neither party any rights. But that makes very little sense—it is unlikely that the parties chose the California wage and hour law in an attempt to place themselves outside all laws. More likely, they selected “California law” without considering the question of what wage and hour law would govern work performed in Arizona. In any case, California does not intend to allow them to avoid the Arizona wage and hour law, if that would otherwise govern, so their contracting power should be no greater than usual. Consequently, “no law” is the wrong answer. The governing law should be the law of Arizona, except that if the parties intended to displace it by incorporating the terms of California law, they should be able to do so with respect to non-mandatory provisions of Arizona law.
D. Conclusion
I have tried to say something about the various ways in which party choice and state legislative authority might interact. My general point is that we should think about the interests and policies of the state whose law is selected, as well as the forum and the state whose law is not selected. Sometimes states limit the scope of their laws, and we should pay attention to those limits.
If we think about the range of possibilities with those points in mind, we can see how several different scenarios should be resolved. These resolutions generally track the cases, which is an indication that courts have been able thus far to reach sensible results without a fully theorized account of contractual choice of law.
First, if a state has no contacts with the parties or the transaction and has not adopted an inbound choice-of-law statute, the parties should not be able to select its law to govern their relationship. They may be able to incorporate the terms of that law, to the extent the otherwise-governing law permits, but they will not have access to specialized tribunals or other state enforcement mechanisms.
Second, if a state has no contacts with the parties or the transaction but it has adopted an inbound choice-of-law statute, parties whose transaction meets the statutory requirements can select that law to govern their relationship. They will be entitled to specialized tribunals and state enforcement mechanisms—subject, in the United States, to possible limits under the Full Faith and Credit Clause.
Third, if a state places limits on the scope of its law, parties outside those limits cannot select the law even if they do have contacts with the state. Giving literal effect to a choice of law that does not reach the parties would mean deciding that no law governs on that issue, but this is unlikely to be the parties’ intent and is not a result encouraged by the state that restricts its law. Consequently, it makes more sense to simply submit those issues to the otherwise-governing law, understanding that the parties may have intended to incorporate the terms of the scope-limited law and that they have the same power to do so as if they had explicitly written those terms into their contract.
Acknowledgements
The author declares none.
Competing Interests
The author declares none.
Funding Statement
No specific funding has been declared in relation to this Article.
Open access
