Article

Interpreting Forum Selection Clauses

I.     Introduction

Forum selection clauses—contractual provisions in which the parties agree to litigate their disputes in a specified forum—are now regularly written into commercial contracts in the United States.1 Although U.S. courts were historically reluctant to enforce such clauses, this is no longer the case.2 Modern courts will generally give effect to these provisions so long as they are not unjust, contrary to public policy, or the product of fraud or overreaching.3 This shift in judicial attitudes has generated extensive commentary relating to the enforceability of forum selection clauses.4  While important, this is not the only legal issue presented by these clauses. The courts are also routinely called upon to interpret certain words and phrases that commonly appear in forum selection clauses.5 To date, the question of how courts should interpret these clauses has attracted far less scholarly attention than the question of whether these clauses are enforceable. Indeed, it has attracted almost no attention at all.6

This Article aspires to fill this gap. Courts called upon to interpret forum selection clauses confront a singular challenge—namely, the words and phrases in these clauses are usually non-negotiated boilerplate.7 While the contracting parties will typically dicker over the identity of the chosen forum—whether litigation must proceed in New York or Texas, for example—they will rarely give much thought to the other words in their forum selection clause. This inattention presents a number of challenges. The goal of contract interpretation, generally speaking, is to give effect to the “intent” of the parties.8 The best evidence of this intent, in turn, is said to be the language of the agreement.9 When the contract language consists of non-negotiated boilerplate, however, an inquiry into the actual intent of the specific parties to a particular agreement will rarely turn up useful.10 The parties are using the same language as have thousands of other parties in thousands of other contracts. In such cases, it is difficult for the courts to credibly maintain that they are giving effect to the intent of these particular parties to this particular contract.11 They are instead assigning a meaning to the language that—they hope—is broadly in line with what most other parties using that same language would want it to mean.

In order to facilitate this interpretive task, the courts have developed a number of canons of construction that assign a presumptive meaning to ambiguous words and phrases that appear in forum selection clauses.  These canons are:

  1. The Canons Relating to Exclusivity. These canons help the court determine whether the parties have agreed to litigate their dispute exclusively in the chosen forum or whether they have merely consented to jurisdiction or venue in that forum. An exclusive forum selection clause stipulates that the parties must litigate in the chosen forum and no other. A non-exclusive forum selection clause merely provides that the parties consent to jurisdiction or venue in the chosen forum. In many cases, one party will argue that a clause is exclusive while the other will argue that the clause is non-exclusive. In these cases, the court must look to the precise words utilized in the clause to resolve the dispute. Over the years, the courts have developed interpretive rules to help distinguish exclusive clauses from non-exclusive clauses.
  2. The Canons Relating to Scope. These canons help the court determine whether the forum selection clause is narrow or broad. Narrow forum selection clauses only select a forum for contract claims arising between the parties; they do not select a forum for tort claims or for statutory claims. Broad forum selection clauses select a forum in which all claims must be resolved regardless of whether they sound in contract, tort, or statute. In distinguishing narrow clauses from broad ones, courts across the United States have developed a dizzying array of interpretive rules, some of which conflict with one another, to determine the scope of the clause.
  3. The Canons Relating to Non-Signatories. These canons help the courts determine whether the forum selection clause covers parties who are affiliated with the contract signatory but who did not actually sign the contract. When a company signs a contract containing an exclusive forum selection clause selecting a particular state, for example, and where the contract counterparty sues both the company and its subsidiary in a different U.S. state, the courts must determine whether the subsidiary—which was not a party to the original agreement—may invoke the forum selection clause. In answering this and similar questions, the courts have developed interpretive rules to determine when these clauses may be invoked by and against non-signatories.
  4. The Canons Relating to Federal Court. These canons help the courts determine whether the parties contemplated litigation in state court to the exclusion of federal court or whether they contemplated litigation in either state or federal court. This issue most commonly arises when one party seeks to remove a case to federal court on the basis of diversity jurisdiction and the other party opposes the motion on the grounds that the parties had previously agreed—via a contractual forum selection clause—to litigate their dispute exclusively in state court. Over the years, the courts have developed several interpretive rules of thumb to assist them in identifying forum selection clauses where each party has waived its right to be in federal court. 

Each of these canons plays an important role in determining the meaning of a forum selection clause and, by extension, whether litigation relating to a particular contract must proceed in a given place.12 To date, however, these canons have attracted surprisingly little attention in the academic literature.13 The Article’s first contribution to the literature, therefore, is to identify these canons, to assign them labels, and to show how they operate in practice to assign meaning to ambiguous language in forum selection clauses. This project is largely descriptive.

The Article’s second contribution to the literature is normative. Here, the Article seeks to evaluate whether the canons identified above produce results that are broadly consistent with the expectations of most contracting parties. This is more challenging than it may appear at first glance. While courts interpreting contract boilerplate generally recognize that they should interpret contract boilerplate to align with the preferences of most contracting parties—rather than the intentions of these contracting parties—they often struggle to determine which of the proffered interpretations best aligns with majoritarian expectations.14 Indeed, courts are often encouraged to adopt idiosyncratic interpretations of contract language by the lawyers who appear before them.15 In situations where a litigant can only win a case by persuading the court to adopt an idiosyncratic interpretation of a piece of contract boilerplate, for example, that litigant’s attorneys will urge the court to select this interpretation even when it is disfavored by a majority of other contract users.

When similar or identical language appears in many thousands of contracts, moreover, the first decision in a particular jurisdiction interpreting that language takes on an outsized importance. If that decision articulates an interpretive default rule that aligns with the expectations of most contracting parties, then all will be well. Conversely, if that first decision interprets a boilerplate provision in a manner that is inconsistent with the expectations of most parties, that decision will impose costs on third parties whose contracts contain similar language.16 Sophisticated third parties must thereafter incur the cost of redrafting their agreements to contract around this interpretation. Unsophisticated third parties, unaware of the court’s decision, must thereafter incur the cost of litigating against the backdrop of an interpretive default rule that is inconsistent with their baseline expectations. When a court adopts an idiosyncratic interpretation of a piece of contract boilerplate, in short, this decision is keenly felt by a range of third parties not involved in the litigation.17

The Article argues that one possible solution to this problem is to conduct surveys and interviews with disinterested lawyers to ask them how they want the courts to interpret ambiguous language in forum selection clauses.18 When these surveys and interviews are conducted outside the context of litigation, the respondents are free to be candid with respect to their preferences; they are not constrained to give answers that advance their client’s immediate interests in an ongoing lawsuit.19 While contract drafters may not always pay close attention to the boilerplate language used in their agreements, they almost always have a preference as to what they want such provisions to mean.20 As things stand, the courts have no easy way to uncover these preferences separate and apart from the self-serving evidence presented to them by the parties in litigation. Surveys and interviews with disinterested parties conducted by scholars and other disinterested parties can help to remedy this information deficit.

To illustrate the utility of such surveys to shed light on the intended meaning of boilerplate contract language, I contacted 86 attorneys to ask them how they generally want courts to interpret ambiguous language in forum selection clauses. Their responses—described in the main body of the Article—suggest that the canons relating to exclusivity generally produce outcomes that are consistent with the expectations of most parties. They also show that the canons relating to non-signatories and the canons relating to federal court frequently produce outcomes that are inconsistent with the expectations of most parties. With respect to the canons relating to scope, the survey responses suggest that the courts should generally err on the side of reading forum selection clauses broadly rather than narrowly. Each of these findings has the potential to reshape the way that courts go about assigning meaning to ambiguous language in forum selection clauses. More importantly, they show how courts might rely on surveys and interviews as aids when interpreting boilerplate contract language.

The Article’s final contribution relates to contract drafting. If a forum selection clause is clear and unambiguous, there will be no need for the courts to invoke any of the canons discussed above. With this in mind, the Article identifies the language—the “magic words”—that the courts have previously identified as signaling a particular meaning in this context. It then urges parties to incorporate these words into their contracts ex ante to avoid incurring the costs associated with litigating their meaning ex post.

The Article proceeds as follows. Part II discusses the canons that the courts use to determine whether forum selection clauses are exclusive or non-exclusive. Part III reviews the canons that determine the scope of generic forum selection clauses. Part IV examines the canons that enable the courts to determine whether a forum selection clause may be invoked by non-signatories. And Part V considers the canons used by the courts to assess whether the parties intended to litigate their dispute in state or federal court. These Parts are largely descriptive. The normative analysis begins in Part VI, which describes the survey methodology. Part VII then applies that methodology to each of the canons in an attempt to determine whether it produces results that are consistent with majoritarian preferences. The Article concludes by offering advice to contract drafters on how to draft forum selection clauses that specifically address each of the foregoing issues.

II.     The Canons Relating to Exclusivity

An exclusive forum selection clause may be usefully conceptualized as a form of waiver.21 By agreeing to litigate a dispute in a specified forum, the contracting parties waive their right to bring a lawsuit elsewhere. When a defendant argues that a contract contains an exclusive forum selection clause that requires the parties to litigate their dispute in a different court, therefore, the court must determine if the plaintiff has waived its right to sue elsewhere. While this task may seem straightforward in theory, it can be surprisingly complex in practice. A waiver of one’s right to sue elsewhere will often resemble a waiver of the right to contest personal jurisdiction in the chosen court.22 It will also often resemble a waiver of the right to object to venue in the chosen court.23 To distinguish among these different types of waivers, the courts have developed interpretive rules to assist them. I refer to these rules as the canons relating to exclusivity.

The canons relating to exclusivity posit that the parties have waived their right to sue elsewhere only when a forum selection clause contains “language of exclusivity.”24 Courts applying this test will look for certain “magic words” that signal the parties’ intent to litigate their disputes in the chosen forum to the exclusion of all other possible venues. The words “exclusive” or “sole” are generally recognized to convey this intent.25 Statements that a claim “must” be brought in a particular forum or that it may “only” be brought in that forum also suffice.26 Each of the following provisions was held to be an exclusive forum selection clause because it contained the requisite language of exclusivity:

Sole and exclusive jurisdiction and venue for any action, suit or litigation arising from or related to this agreement shall be in the state or federal courts located in the State of New Hampshire. . . .27

[T]he Courts of the State of Colorado . . . shall be the exclusive forum for the resolution of any disputes related to or arising out of this Term Agreement.28

In the event that either party brings suit to enforce the terms of this [a]greement both [parties] consent and agree that jurisdiction for such action will lie only in the state and federal courts sitting in Mecklenburg County, North Carolina. . . .29

Charterer further specifically agrees and consents that any causes of action or suits related to this Agreement must be filed in the Second Judicial District Court, Albuquerque, New Mexico, USA. . . .30

Such language, in the view of the courts, clearly signals the parties’ intent to waive their right to bring a lawsuit in any forum other than the chosen forum.

Some courts have suggested that the word “shall” connotes a similar intent.31 In Slater v. Energy Services Group International, the Eleventh Circuit was asked to construe a clause stating “that all claims or causes of action relating to or arising from this Agreement shall be brought in a court in the City of Richmond, Virginia.”32 That court held this clause was an exclusive forum selection clause because “the term ‘shall’ is one of requirement.”33 There are, however, other cases in which the word “shall” was deemed insufficient. In Hunt Wesson Foods, Inc. v. Supreme Oil Company, the clause at issue stated that “[t]he courts of California . . . shall have jurisdiction over the parties in any action at law relating to the subject matter or the interpretation of this contract.”34 The presence of the word “shall” notwithstanding, the Ninth Circuit reasoned that the clause was not mandatory because it did not prohibit the contract parties from bringing claims in other fora.35

As a form of shorthand, some courts distinguish between “mandatory” and “permissive” forum selection clauses.36 A “mandatory” clause contains language of exclusivity. A “permissive” clause lacks such language.37 When a clause states that a particular court shall “have jurisdiction” over a suit, for example, the parties have not waived their right to sue elsewhere. They have merely waived their right to object to the assertion of personal jurisdiction by the chosen court. This is a permissive clause.38 Similarly, when a clause states that the parties “consent to venue” in a particular court, the parties have not waived their right to sue elsewhere. They have merely waived their right to argue that the chosen venue is improper.39 This is also a permissive clause.40 Consent-to-jurisdiction clauses and consent-to-venue clauses will often look a great deal like exclusive forum selection clauses.41 These clauses are distinguishable, however, because they lack the all-important language of exclusivity. Each of the following provisions was held to be permissive rather than mandatory:

The parties hereto submit and consent to the jurisdiction of the courts present in the state of Texas in any action brought to enforce (or otherwise relating to) this agreement.42

Any suit, action or proceeding arising out of or relating to this Agreement may be commenced and maintained in any court of competent subject matter jurisdiction in Miami-Dade County, Florida and each party waives objection to such jurisdiction and venue.43

Travis County, Texas shall be appropriate jurisdiction and venue for any litigation with respect to the Partnership.44

Each of the foregoing clauses lacks any of the words that are necessary to signal the parties’ intent to select the chosen forum to the exclusion of all others. Accordingly, each was held to constitute a permissive (non-exclusive) forum selection clause rather than a mandatory (exclusive) one.45

III.     The Canons Relating to Scope

A different interpretive question asks whether a forum selection clause applies solely to claims for breach of contract or whether it also applies to tort and statutory claims. The answer to this question will depend on the language in the clause.46 At one end of the spectrum lie clauses that are clearly broad enough to cover non-contractual claims.47 These clauses typically contain the phrase “related to” or “in connection with”:

[A]ll claims or causes of action relating to or arising from this Agreement shall be brought in a court in the City of Richmond, Virginia.48

[The parties] agree that any appropriate state or federal district court located in the Borough of Manhattan, New York City, New York, shall have exclusive jurisdiction over any case or controversy arising under or in connection with this Agreement and shall be a proper forum in which to adjudicate such case or controversy.49

Each of these clauses clearly covers non-contractual claims that relate to the agreement in some way. At the other end of the spectrum, one finds clauses that were clearly not intended to apply to non-contractual claims. These clauses typically contain language that refers specifically to the contract or the obligations set forth in the contract:

Should either party institute legal suit or action for enforcement of any obligation contained herein, it is agreed that the venue of such suit or action shall be in San Diego County, California.50

Any conflict which may arise regarding the interpretation or fulfillment of this contract, shall be submitted expressly to the courts of the City of Ensenada, B.C. 51

There are, however, many clauses that fall between these two poles. These clauses do not contain a clear statement of scope.52 Instead, they contain language such as “arising out of” or “arising hereunder.” Such clauses often look like this:

The sole venue for disputes arising hereunder shall be in Harris County, Texas.53

Any litigation or arbitration between the parties which arises out of this Agreement shall be instituted and prosecuted only in the appropriate state or federal court or other tribunal situated in Miami, Florida. 54

The parties agree to submit to the exclusive jurisdiction over all disputes hereunder to the federal and state courts in the State of New York located in New York County.55

I refer to this third and final group of clauses—which shall occupy the bulk of our attention in this Part—as “generic” forum selection clauses.56 A generic forum selection clause, by definition, is ambiguous when it comes to the question of whether it applies to tort and statutory claims.57

The courts have reached varying conclusions about how to interpret generic forum selection clauses. Some courts have held that these clauses cover tort and statutory claims.58 Other courts have held that these clauses do not cover tort and statutory claims.59 These varying outcomes are attributable, at least in part, to the very different interpretive approaches that different courts utilize to resolve this question. Since these clauses are mostly boilerplate, there is rarely much in the way of extrinsic evidence to shed light on the intent of the contracting parties. To resolve the ambiguity, the courts must make informed guesses as to what most contracting parties probably would have wanted at the time of drafting to assign a meaning to the clause.60 Over time, these judicial guesses have congealed into a set of interpretive default rules that assist courts in determining the scope of a generic forum selection clause. These interpretive rules are the canons relating to scope.

Although each of these canons is distinct, they all seek to answer the same basic question: whether a plaintiff’s non-contractual claims are so closely intertwined with the parties’ contractual relationship that the parties would have wanted them to be covered by the forum selection clause. Unlike the canons relating to exclusivity, the canons relating to scope are not always closely tied to specific words and phrases.61 While the courts frequently utilize the canons to construe specific phrases in the agreement—typically “brought hereunder” or “arising under”—they also deploy these canons where these phrases are absent.62 The canons may therefore be usefully conceptualized as a set of interpretive default rules that are applied to ambiguous clauses as a whole rather than to specific words and phrases within each clause.63

The canons relating to scope encompass five distinct interpretive methodologies that assign a scope to a generic forum selection clause. These are:

  1. Origination. A generic forum selection clause never covers non-contractual claims because these claims do not originate in the contract.
  2. Same Operative Facts. A generic forum selection clause covers non-contractual claims only when these claims arise out of the same operative facts as a parallel contract claim.
  3. Contract Analysis. A generic forum selection clause covers non-contractual claims only when (a) the claims relate in some way to the interpretation of the contract, (b) the court must construe the contract to resolve the claims, i.e., read implied terms into it, or (c) the claims cannot be adjudicated without determining whether the defendant is in compliance with the contract.
  4. Existence of a Contractual Relationship. A generic forum selection clause covers non-contractual claims only when (a) the claims would not have arisen but for the contractual relationship between the parties, or (b) the claims have a direct relationship to the contract.
  5. Hybrid. A generic forum selection clause covers non-contractual claims only when (a) they arise out of the same operative facts as a parallel contract claim, (b) they require contractual analysis, or (c) they depend upon the existence of a contractual relationship.64

At the outset of this inquiry, it should be noted that—at least in this area of law—the courts are magpies. They routinely pick and choose among interpretive methodologies without regard to precedent or hierarchy of authority.65 While some methods are more likely to be applied by certain courts, to search for perfect consistency across the many hundreds of cases in this area is to invite madness.66

A.     Origination

Some courts have held that tort and statutory claims are never covered by a generic forum selection clause. Courts that follow this approach reason that non-contractual claims do not “originate” in the contract—they originate in the common law of tort or in a statute—and therefore do not arise out of the contract.67 This approach represents the most parsimonious approach to clause construction.

In Phillips v. Active Audio Ltd, for example, the Second Circuit was asked to determine whether the plaintiff’s claim for copyright infringement arose out of a recording contract between a U.S. musician and an English recording studio.68 The court concluded that this claim did not arise out of the contract—and was therefore not covered by the forum selection clause naming England as the exclusive forum—because it did not originate in the contract.69 The court held that the claim originated in the federal copyright statute.70 Accordingly, it concluded that while the plaintiff’s breach of contract claim had to be litigated in England—per the exclusive forum selection clause—his copyright claim could be litigated in New York.71

This narrow interpretive approach has been rejected by most other courts. These courts have observed that the approach encourages plaintiffs to recast contract claims as tort and statutory claims to evade an otherwise valid forum selection clause.72 In the words of the New Hampshire Supreme Court:

The Second Circuit’s test, by asking whether the rights of the claims originate in the contract, may allow savvy plaintiffs to disguise what could be contract-based claims as statutory or common-law tort claims, even if the parties would have expected to litigate such disputes in the selected forum.73

This approach is also disfavored because it makes it more likely that litigation arising out of a single contractual relationship will proceed in multiple fora. As the Second Circuit observed at the conclusion of its opinion in Phillips:

We are aware that the commencement of separate proceedings in two countries is a likely inconvenience to the parties and that they . . . may have intended to bundle all claims constituting any proceeding to avoid fractured litigation. . . . [B]ut our twin commitments to upholding forum selection clauses where these are found to apply and deferring to a plaintiff’s proper choice of forum constrain us in the present context to treat [Plaintiff’s] claims separately.74

In summary, while there may be a certain linguistic elegance in the notion that non-contractual claims do not “arise out of” a contract because they do not “originate” in that same contract, most courts have shied away from this approach due to concerns about artful pleading and fragmented litigation proceedings.75

B.     Same Operative Facts

Other courts have held that tort and statutory claims are covered by generic forum selection clauses only when such claims arise out of the “same operative facts” as a parallel contract claim.76 In essence, this analytical approach directs courts to consider whether the facts underlying the plaintiff’s tort or statutory claim could also support a breach of contract action.77 If so, these non-contractual claims are covered by the forum selection clause.

The “same operative facts” test was utilized by the New Hampshire Supreme Court in Hansa Consult of North America v. Hansaconsult Ingenieurgesellschaft.78 The plaintiff in Hansa was a U.S. company “in the business of detecting fuel leaks at airports.”79 The defendant was a German company.80 In 2001, these two companies entered into a distribution agreement in which the U.S. company was named the exclusive distributor of the German company’s products and services in the United States and Canada.81 In 2009, the U.S. company sued the German company in New Hampshire state court.82 The plaintiff alleged that the defendant had—prior to the expiration of the distribution agreement—hired one of its former employees and that this employee had stolen trade secrets from the plaintiff’s computer server.83 The plaintiff further alleged that the defendant had—after the expiration of the distribution agreement—wrongfully and incorrectly told customers and potential customers that the plaintiff lacked authority to utilize a certain type of software to detect fuel leaks.84

Based on these factual allegations, the plaintiff asserted five non-contractual claims: (1) misappropriation of trade secrets, (2) conversion, (3) violation of the New Hampshire Consumer Protection Act, (4) common law unfair competition, and (5) tortious interference with contractual relations.85 The question before the court was whether these claims fell within the scope of the forum selection clause in the distribution agreement, which stipulated that “[p]lace of [j]urisdiction is only Hamburg.”86 The court decided to apply the “same operative facts” test to determine the scope of the clause because in the court’s view this test “best effectuates the legitimate expectations of contracting parties.”87 Under this approach, the court must “examine the claims and facts side-by-side and . . . determine which disputes can fairly be considered to ‘arise under’ the contract.”88

In applying this test to the facts before it, the court held that two of the plaintiff’s claims—misappropriation of trade secrets and conversion—were covered by the forum selection clause because the operative facts underlying these claims would have supported a parallel cause of action for breach of contract.89 In reaching this conclusion, the court relied heavily on the fact that the contract contained a provision obliging the parties to protect each other’s trade secrets.90 The court ordered these two claims to be dismissed in favor of a German forum.91 The court then held that the three remaining claims—violation of the consumer protection act, unfair competition, and tortious interference—were not covered by the forum selection clause because there was “no parallel breach of contract claim involving the same operative facts.”92 Accordingly, the court held that these three claims could be brought in New Hampshire state court because they did not grow out of the “same operative facts” as the contract claim.

C.     Contract Analysis

Other courts have held that a non-contractual claim is covered by a generic forum selection clause only when the claim requires the court to analyze the contract in some way. Some courts utilizing this approach have held that non-contractual claims are covered when they relate to the “interpretation” of the contract.93 Other courts applying this test routinely ask whether the adjudication of the non-contractual claims will require the court to read “implied terms” into the contract.94 Still others ask whether the non-contractual claims can be adjudicated without assessing whether the parties are in “compliance” with the contract.95 Each of these approaches is briefly set forth below.

1.     Interpretation

In one iteration of the contract-analysis inquiry, the court considers whether the non-contractual claims relate in some way to the “interpretation” of the contract.96 The U.S. District Court for the Northern District of California adopted this approach in Graham Technology Solutions v. Thinking Pictures.97 The plaintiff was a computer programmer who brought a copyright infringement suit against a technology company.98 The essence of the complaint was that the company was illegally using code written by the programmer.99 In its defense, the company argued that (1) the code was produced pursuant to detailed hardware and software specifications that the company delivered to the programmer, and (2) the parties had executed a professional services agreement (“PSA”) stipulating that the company “would own all right, title and interest to the [programmer’s] completed work product.”100 This agreement also contained a forum selection clause stating that “[t]he parties agree to submit to the exclusive jurisdiction over all disputes hereunder to the federal and state courts in the State of New York located in New York County.”101 When the plaintiff filed suit in federal court in California, the company moved to dismiss the case in favor of a New York forum, citing the forum selection clause.102 In response, the plaintiff argued that the copyright claims were not covered by the forum selection clause and that the case should remain in California.103

In considering these competing arguments, the court first observed that “forum selection clauses can be equally applied to contractual and tort causes of action where resolution of the tort claims relates to the interpretation of the contract.”104 It then went on to observe that

[f]or either party to prevail in this action, it must prove that its interpretation of the PSA and the work to be completed by [the plaintiff] under the terms and conditions of the PSA is correct. The causes of action alleged by [the plaintiff] relate to the central conflict over the interpretation of the PSA, and therefore fall within the scope of the forum selection clause.105

Since it was necessary for the court to interpret the PSA in order to evaluate the merits of the plaintiff’s non-contractual claims, in other words, the court concluded that these claims must be covered by the contractual forum selection clause.106

2.     Construction and Implied Terms

A slightly different iteration of this inquiry posits that a generic forum selection clause covers all non-contractual claims whose resolution “arguably depend[s] on the construction of an agreement.”107 This version of the test was utilized by the Seventh Circuit in Omron Healthcare, Inc. v. Maclaren Exports, Ltd.108 The plaintiff in Omron was a U.S. company that sold baby strollers.109 Omron entered into a distribution agreement with a British firm whereby Omron was named the exclusive distributor of its strollers in the United States.110 After this agreement was terminated, the British firm entered into a new distribution agreement with a different U.S. company.111 The first shipment of goods delivered to this new U.S. company contained strollers that still bore the “Omron” trademark because they had been produced for Omron prior to the termination of the old distribution agreement.112 Omron then sued the British firm for trademark infringement in the Northern District of Illinois.113 The British firm moved to dismiss on the basis of the following forum selection clause in the distribution agreement: “The parties hereto agree that all disputes arising out of this Agreement which cannot be resolved amicably between the parties shall be referred to the High Court of Justice in England which will have exclusive jurisdiction to determine such disputes.”114 The plaintiff argued that the trademark claims did not “arise out of” the distribution agreement and were therefore not covered by it.115 The defendants argued in response that the trademark claims “arose out of” the distribution agreement because it would be necessary for the court to construe the contract to resolve those claims.116

The Seventh Circuit sided with the defendants. It held that the plaintiff’s trademark claims fell within the forum selection clause because the resolution of these claims would require the court to read implied terms into the distribution agreement.117 Unlike the Graham court, the Omron court did not focus on the need to interpret words that were actually in the contract.118 Instead, the Omron court focused on the fact that it would be necessary to read certain default rules into the contract in order to resolve the plaintiff’s claims. Reading implied terms into an agreement is not, strictly speaking, an act of interpretation—it is an act of construction. Nevertheless, the Omron court held that the need to perform this task operated to bring the plaintiff’s non-contractual claims within the scope of the forum selection clause.

3.     Compliance with the Contract

The third and final iteration of the contract-analysis inquiry holds that non-contractual claims are subject to a forum selection clause when these claims cannot be properly adjudicated without determining whether the parties were “in compliance” with the contract.119 In essence, this approach asks whether it is possible to resolve the non-contractual claims without inquiring into whether the defendant was in breach of the contract. If so, then these claims are not covered by the forum selection clause. If not, then these claims fall within the scope of the clause. This approach was utilized by the Second Circuit in Magi XXI, Inc. v. Stato della Città del Vaticano, a case involving a sublicense to reproduce images of various items in the Vatican library collection.120 The plaintiff asserted claims for fraud, negligence, breach of contract, unjust enrichment, and conversion.121 The defendant sought to dismiss on the basis of a forum selection clause requiring all disputes to be resolved in Vatican City.122 While the plaintiff argued that its tort claims were not covered by this clause, the Second Circuit disagreed. It stated that

[e]ach of [the plaintiff’s] claims relates to the rights and duties set out in the [license agreements]. These claims cannot be properly adjudicated without determining whether the parties were in compliance with the Sublicense Agreements and the Master License Agreement. Because the resolution of these claims requires interpretation of the contracts at issue, they fall within the scope of the forum selection clauses in the [Sublicense] Agreements.123

In light of this analysis, the court ordered the plaintiff’s claims to be dismissed in favor of a forum in Vatican City.124

To sum up, some courts have—in determining the scope of ambiguous forum selection clauses—asked whether the adjudication of non-contractual claims requires interpretation of the contract. Others have asked whether the adjudication of these claims requires the courts to read implied terms into the contract. Still others have asked whether the adjudication of these claims necessitates an inquiry into whether the defendant is in compliance with the contract. If the answer to any of these questions is yes, the non-contractual claim will be deemed to fall within the scope of the forum selection clause and the plaintiff must sue in the chosen forum. If the answer is no, then the plaintiff may ignore the clause and sue elsewhere.

D.     Existence of a Contractual Relationship

Still other courts have held that non-contractual claims are covered by generic forum selection clauses where these claims “ultimately depend on the existence of a contractual relationship” between the parties.125 In practice, the courts have adopted two different approaches to resolving this question. Some courts have held that any claims that would not have arisen “but for” the contract are covered by the forum selection clause.126 This approach will typically result in a clause being given a broad scope. Other courts, by contrast, have held that a clause will only cover non-contractual claims when these claims have a “direct relationship” to the contract at issue or where the existence of the contract is the “proximate caus[e]” of the claims.127 This approach will typically result in the clause being given a narrower scope.128

1.     “But For” Causation

A good example of the “but for” line of reasoning can be found in Doe v. Seacamp Association.129 In this case, the U.S. District Court for the District of Massachusetts was called upon to determine whether a minor’s tort claims against a summer camp were covered by a forum selection clause in the enrollment contract requiring all suits to be brought in Florida. In concluding that the plaintiff’s tort claims were covered by the clause, the court reasoned that “[a]bsent the contract enrolling [plaintiff] at Seacamp, no duty would have been owed to him by this defendant.”130 In essence, this analytical approach imagines a hypothetical world in which the contractual relationship never came into existence and asks whether the non-contractual claims would still exist. In Doe, the court reasoned that if the plaintiff had never signed the enrollment contract, he would never have come to camp; if he had never come to the camp, he never would have been assaulted by a counselor; and if he had never been assaulted, the tort claims never would have been brought.131 Accordingly, the court concluded that the contract was a “but-for” cause of the tort claims and that these claims were therefore covered by the forum selection clause.132

This sweeping approach has attracted criticism.133 The Seventh Circuit has argued that “[b]ut-for causation is an unsatisfactory understanding of language referring to ‘disputes arising out of’ an agreement.”134 Similarly, the Missouri Court of Appeals has held that “[a] forum selection clause in a contract does not control the site for litigation of a tort claim simply because the dispute that produced the tort claim would not have arisen absent the existence of the contract.”135 In these courts’ view, there must be some limiting principle on the scope of a generic forum selection clause when the issue is whether the non-contractual claim depends upon the existence of a contractual relationship. The limiting principle, for these courts, is the concept of proximate cause.

2.     Proximate Causation

The Eleventh Circuit utilized a proximate cause analysis in Bailey v. ERG Enterprises, LP to determine the scope of a forum selection clause.136 In that case, the trial court—relying on notions of “but for” causation—had held that the plaintiffs’ non-contractual claims fell within the scope of the clause because their claims “ar[o]se out of a relationship that was established by the . . . contracts” and because they “would not have any claims had they not entered into the . . . contracts.”137 The Eleventh Circuit reversed. For the plaintiffs’ claims to fall within the scope of the contracts’ forum-selection clauses, the court held, the claims must either have a “direct relationship” to the contract at issue or be the “fairly direct result of the performance of contractual duties.”138 Neither condition was met on these facts:

[A]lthough the claims would not exist but for the [plaintiffs] purchasing the [property], this but-for relationship does not mean that the claims relate to the lot purchase contracts. The claims must result from the performance of duties under the lot purchase contracts; the [plaintiff’s] claims do not.139

Although the plaintiffs’ non-contractual claims would not have arisen “but for” their entering into the contract in the first place, the Eleventh Circuit held that these claims were not covered by the forum selection clause because they did not have a “direct relationship” to that contract.140 The existence of the contractual relationship, in short, was not the proximate cause of the plaintiff’s non-contractual claims. Accordingly, the court held that the claims were not covered by the clause.141

E.     Hybrid Approaches

While some courts have applied the various interpretive approaches set forth above in isolation, other courts have combined the previous three approaches into a multi-part test. The seminal case adopting such an approach is Terra International, Inc. v. Mississippi Chemical Company.142 The court observed that the resolution of the plaintiff’s claims did not “seem to relate to or require the interpretation of the license agreement.”143 It went on to conclude, however, that the plaintiff’s tort claims “ultimately depend on the existence of a contractual relationship” and “involve the same operative facts as would a parallel claim for breach of contract.”144 Accordingly, the court held that the plaintiff’s tort claims fell within the scope of the forum selection clause and ordered the case dismissed in favor of the jurisdiction selected in the exclusive forum selection clause (Mississippi).145

In the years since Terra was decided, its hybrid approach has attracted many adherents.146 There are a significant number of cases in which courts consider not one, not two, but all three of the approaches discussed immediately above.147 In cases where the courts have adopted a hybrid approach, the defendant typically need only prove that the criteria in one of these tests has been satisfied in order to prove that the plaintiff’s non-contractual claims are subject to the forum selection clause.148 Even if a particular tort claim does not arise out of the same operative facts as a contract claim or require the interpretation of the contract, for example, it may still be covered if the claim would not exist absent a contractual relationship between the parties. In practice, therefore, courts applying a hybrid approach give the litigant seeking to enforce the clause multiple bites at the apple, thereby increasing the chances that tort and statutory claims will be covered by the clause.

F.     Summing Up

The canons relating to scope constitute a set of judicial guesses as to what most parties to a particular contract probably would have wanted the scope of their generic forum selection clause to be. It is useful, therefore, to recast the discussion above as a set of guesses as to party intent:

  1. Origination. Parties never want their generic forum selection clauses to cover non-contractual claims.
  2. Same Operative Facts. Parties want their clauses to cover non-contractual claims only when the claims arise out of the same operative facts as a parallel contract claim.
  3. Contract Analysis. Parties want their clauses to cover their non-contractual claims only when (a) the claims relate in some way to the interpretation of the contract, (b) the court must construe the contract to resolve the claims, i.e., read implied terms into it, or (c) the claims cannot be adjudicated without determining whether the defendant is in compliance with the contract.
  4. Existence of a Contractual Relationship. Parties want their clauses to cover non-contractual claims only when (a) the claims would not have arisen but for the contractual relationship between the parties, or (b) the claims have a direct relationship to the contract.
  5. Hybrid. Parties want their clauses to cover non-contractual claims only when (a) they arise out of the same operative facts as a parallel contract claim, (b) they require contractual analysis, or (c) they depend upon the existence of a contractual relationship.

 

Each of these judicial guesses as to party intent is empirically testable. It should be possible, in other words, to determine whether each guess tracks the actual preferences of most parties who enter into agreements that contain generic forum selection clauses. One need only ask these parties—or more realistically, their lawyers—about their baseline expectations when it comes to the scope of their forum selection clauses. This Article undertakes just such an inquiry in Parts VI and VII. Before proceeding to that analysis, however, it is necessary to review the remaining two categories of canons.

IV.     The Canons Relating to Non-Signatories

Still another issue that sometimes arises is whether an exclusive forum selection clause covers non-signatories to the agreement. At first glance, the answer to this question may seem obvious. It is hornbook law that one cannot be bound by a contract that one has not signed.149 On a number of occasions, however, courts have held that a litigant may invoke a forum selection clause in an agreement to which it was not a party if that litigant is affiliated in some way with one of the contracting parties.150 In reaching these decisions, the courts have frequently invoked the presumed intent of the parties.151 The result is a set of interpretive rules that I call the canons relating to non-signatories.152

The canons relating to non-signatories may be distinguished from the canons discussed above in several respects. First, they are generally not tied to any particular words or phrases that commonly appear in forum selection clauses. Instead, they are applied to interpret the clause as a whole. Second, while other provisions in a contract may be relevant to the analysis, it is exceedingly rare for the parties to address the issue of non-signatories in the forum selection clause itself. Since many attorneys assume that contracts do not bind non-signatories, they do not think to address the question of whether such persons are covered by a forum selection clause.

The issue of whether a forum selection clause applies to non-signatories frequently arises when a plaintiff brings a lawsuit against companies that are affiliated in some way with the contracting party but that are not actually parties to the contract.153 In Lu v. Dryclean-U.S.A. of California, for example, the plaintiffs were California residents who had entered into a franchise agreement with a Florida-based dry cleaning company.154 The franchise agreement contained a forum selection clause requiring all litigation to be brought in Florida.155 The plaintiffs sued their counterparty, its parent company, and its grandparent company in California state court.156 The parent and grandparent companies moved to dismiss, citing the forum selection clause.157 The plaintiff opposed the motion on the grounds that these companies were not parties to the contract and hence not entitled to invoke the forum selection clause.158 The California Court of Appeal sided with the defendants and ordered the suit dismissed.159 It held that because these companies were “closely related” to the counterparty and that it was “foreseeable” to the plaintiff that the clause would apply to them, the clause applied to the parent and the grandparent company.160 In a subsequent case presenting similar issues, this same court explained that “[g]iving standing to all closely related entities honors general principles of judicial economy by making all parties closely allied to the contractual relationship accountable in the same forum, thereby abating a proliferation of actions and inconsistent rulings.”161

Many courts have similarly held that a defendant non-signatory to a contract will be covered by a forum selection clause where that person is “closely related” to a signatory and where it is “foreseeable” that the non-signatory would be bound.162 As the Second Circuit has explained:

[A] non-signatory to a contract containing a forum selection clause may enforce the forum selection clause against a signatory when the non-signatory is “closely related” to another signatory. In such instances, the relationship between the non-signatory and that (latter) signatory must be sufficiently close that the non-signatory’s enforcement of the forum selection clause is “foreseeable” to the signatory against whom the non-signatory wishes to enforce the forum selection clause.163 

This outcome, in the view of the Second Circuit, comports with the “legitimate expectations of the parties, manifested in their freely negotiated agreement.”164

While the Lu case addressed the ability of a non-signatory defendant to enforce a forum selection clause against a signatory plaintiff, this is not the only setting in which this particular issue arises. In other cases, a signatory defendant will attempt to enforce a clause against a non-signatory plaintiff when the latter sues the former in a forum other than the one named in the clause.165 In Marano Enterprises of Kansas v. Z-Teca Restaurants, L.P., a lawsuit was brought in federal district court in Missouri by three plaintiffs.166 Two of these plaintiffs were parties to several franchise and development agreements with the named defendants.167 The third plaintiff was not.168 The defendants moved to dismiss the suit based on a forum selection clause requiring suit to be brought in Colorado.169 The non-signatory plaintiff argued that he should not be bound by the clause because he was not a party to the contracts.170 The Eighth Circuit rejected that argument:

[T]he plaintiffs jointly represented to the District Court that [the third plaintiff] is a shareholder, officer, and director of Marano Enterprises, which was a party to the agreements. As such, he is, without question, “closely related” to the disputes arising out of the agreements and properly bound by the forum-selection provisions. [The third plaintiff] joined with [the other two plaintiffs] in bringing suit against [the defendant] under the agreements, arguably acquiescing in the forum-selection clauses within those agreements.171

The basic judicial inquiry is thus the same regardless of whether the clause is being applied to a non-signatory defendant or a non-signatory plaintiff.172

The “closely related” test is not without its critics. One court has described the test as “so vague as to be unworkable.”173 Other courts have turned instead to traditional contract principles—specifically, the law of third-party beneficiaries—to determine whether a non-signatory is covered by a forum selection clause.174 This doctrinal move to third-party-beneficiary law should, in theory, result in fewer non-signatories being bound to forum selection clauses because the “closely related” test is more expansive than third-party-beneficiary law.175 In practice, however, few courts strictly apply the law of third-party beneficiaries in this context.176 Some courts have even stated—offering little in the way of analysis or support—that the “closely related” test and the “third-party beneficiary” test are essentially interchangeable.177 Although using third-party-beneficiary law to resolve the thorny question of whether non-signatories are bound by forum selection clauses may be more intellectually defensible than using the “closely related” test, this doctrine consistently plays second fiddle to the “closely related” test in contemporary practice.178

One thing that the third-party-beneficiary doctrine and the “closely related” test have in common is that both direct the courts to consider the intentions of the parties in making their decisions. One court has observed that “third-party beneficiaries of agreements containing . . . forum-selection clauses may be bound to the clause, depending on the parties’ intentions at the time the contract was executed.”179 Another court has stated that the “closely related” test “comports with the ‘legitimate expectations of the parties, manifested in their freely negotiated agreement.’”180 Since both tests are keyed to party intent, it follows that the parties have the power to alter each test in their agreements should they wish to do so.

In a few instances, they have done exactly that. In Pinto Technology Ventures, L.P. v. Sheldon, for example, the Texas Supreme Court held that the following contract provision clearly expressed an intent to prevent non-signatory defendants from invoking a forum selection clause:

This Agreement . . . shall inure to the benefit of and be binding upon, the successors, permitted assigns, legatees, distributees, legal representatives and heirs of each party and is not intended to confer upon any person, other than the parties and their permitted successors and assigns, any rights or remedies hereunder.181

The court concluded that this language “disclaims any intent to extend the contract’s benefits to nonparties” and that to allow non-signatories to enforce the forum selection clause in the contract “would contravene the parties’ expressed intent and is, thus, impermissible.”182 In Casville Invs., Ltd. v. Kates, the U.S. District Court for the Southern District of New York was presented with a very similar clause:

[T]he provisions of this Agreement shall be binding upon and shall inure to the benefit of the parties hereto and their respective successors and permitted assigns. . . . Neither this Agreement nor any provision hereof is intended to confer upon any person other than the parties hereto any rights or remedies hereunder.183

In light of this language, this court also concluded that the clause could not be invoked by non-signatory “[p]laintiffs because they [were] barred from doing so under the express terms of the Agreement itself.”184

On the one hand, the courts in each of these cases are clearly correct that the plain language of the contract’s no-third-party-beneficiary clause stated that it did not confer any benefit on any non-signatory.185 On the other, it is not altogether clear that the parties foresaw this issue at the time of contracting and consciously drafted this clause to limit the ability of non-signatories to partake of the contract’s forum selection clause.186

V.     The Canons Relating to Federal Court

When two contracting parties agree to an exclusive forum selection clause, it is not always clear whether they wanted to litigate their disputes in state court or whether they wanted to preserve the option of going to federal court.187 This issue sometimes arises when a defendant seeks to remove a case to federal court and the plaintiff objects on the ground that the defendant had previously waived its right to remove in a forum selection clause.188 It also arises when a plaintiff files a suit in federal court and the defendant moves to dismiss on the ground that the parties had agreed to litigate their disputes exclusively in state court.189 When presented with these arguments, the courts have developed three canons of construction that seek to resolve the interpretive issue. These are (1) the “of” canon, (2) the “in” canon, and (3) the “county” canon. Collectively, I refer to these canons as the canons relating to federal court.

In many cases, there will be no need to apply any these canons because the forum selection clause clearly states that claims may only be brought in state court or, alternatively, that claims may be brought in either state or federal court.190 Where the clause is ambiguous with respect to this issue, however, the courts will apply the canons relating to federal court to ascertain whether the clause permits suits to be brought in—or removed to—federal court.191

A.     The “Of” Canon

The first canon posits that the phrase “the courts of State A” signals an intent to select the state courts to the exclusion of the federal courts. This canon was memorably deployed by the Fifth Circuit in the case of Dixon v. TSE International.192  In that case, the court was asked to construe a forum selection clause stating that “[t]he Courts of Texas . . . shall have jurisdiction over all controversies with respect to the execution, interpretation or performance of this Agreement.”193 The issue before the court was whether the phrase “Courts of Texas” referred to the state and federal courts located in Texas or whether it referred only to the Texas state courts. The Fifth Circuit took the latter position:

Federal district courts may be in Texas, but they are not of Texas. Black’s Law Dictionary defines “of” as “denoting that from which anything proceeds; indicating origin, source, descent.” Federal courts indisputably proceed from, and find their origin in, the federal government, though located in particular geographic regions. By agreeing to litigate all relevant disputes solely in “the Courts of Texas,” [the defendant] waived its right to removal.194

This interpretive rule has since been adopted by a majority of U.S. courts.195 An agreement to litigate in the courts “of” a particular state is typically construed as an agreement to litigate the dispute exclusively in state court. As such, it constitutes a waiver of both the right to remove a case to federal court and the right to file suit in federal court in the first instance.196

B.     The “In” Canon

The second canon posits that the phrase “the courts in State A” signals an intent to select the state and the federal courts in that state. When an exclusive forum selection clause provides that all claims must be brought in the courts “in” a particular state, courts have generally held that the clause contemplates litigation in either state or federal court. As the Ninth Circuit has explained:

[A] forum selection clause referring to “courts in” a state imposes a geographic limitation, not one of sovereignty. . . . Hence the phrase “courts in” a state includes any court within the physical boundaries of the state, even if the court does not derive its power and authority from the sovereignty of the state. In short, the rule . . . is that a forum selection clause that specifies “courts of” a state limits jurisdiction to state courts, but specification of “courts in” a state includes both state and federal courts.197

This interpretive rule has also been adopted by most courts.198 An agreement to litigate in the courts “in” a particular state is an agreement to litigate the dispute in any court—whether state or federal—that is located in that state. As such, it does not constitute a waiver of the right to remove a case to federal court or of the right to file suit in federal court in the first instance.

When the chosen jurisdiction is a county, rather than a state, the availability of a federal forum under this rule will depend—somewhat arbitrarily—on whether there is a federal courthouse physically present in that county. In Alliance Health Group LLC v. Bridging Health Options LLC, the Fifth Circuit was asked to construe a clause which stated that “exclusive venue for any litigation related hereto shall occur in Harrison County, Mississippi.”199 The court concluded that the use of the term “in” encompassed both the state and federal courthouses located in Harrison County and that litigation could therefore proceed in federal court.200 In situations where the parties choose to litigate their disputes “in” a county that lacks a federal courthouse, however, the courts have held that any litigation must proceed in state court.201 The Second Circuit, for example, has held that a clause stipulating that “venue . . . shall be Nassau County” precluded litigation in federal court because no federal courthouse was located in Nassau County at the time the suit was brought.202

C.     The “County” Canon

The third and final canon relating to federal court is followed exclusively by the U.S. Court of Appeals for the Tenth Circuit. This court has held that when a forum selection clause names a county as the intended venue—as opposed to a judicial district—the parties must have intended to litigate their dispute exclusively in state court. The Tenth Circuit first adopted this rule in 1992.203 It reaffirmed its support for the rule five years later in Excell, Inc. v. Sterling Boiler & Mechanical, Inc.204 In Excell, the court was asked to determine whether a clause stating that “venue shall lie in the County of El Paso” precluded a motion to remove the case to federal court.205 In concluding that the clause did in fact prohibit removal—because it contemplated litigation exclusively in state court—the court made the following observation:

Although [defendant] argues the clause can be reasonably interpreted to allow removal of the case to federal district court that sits in El Paso County, we reject this argument. For federal court purposes, venue is not stated in terms of “counties.” Rather, it is stated in terms of “judicial districts.” See 28 U.S.C. § 1391. Because the language of the clause refers only to a specific county and not to a specific judicial district, we conclude venue is intended to lie only in state district court.206

In the Tenth Circuit, therefore, the prevailing interpretive rule used to distinguish between clauses that select state courts exclusively and clauses that select state and federal courts tracks the text of the federal venue statute, 28 U.S.C. § 1391.

While there are a few scattered district court decisions that have followed this canon of construction, the other federal courts of appeal have generally declined to adopt it.207 The most comprehensive critique of this interpretive rule can be found in a Fifth Circuit decision rendered in 2008. In that case, the court observed that the county canon “would be more persuasive were the federal courts organized in total disregard of state counties; if, for instance, federal judicial districts were defined by metes and bounds.”208 Instead, these federal “districts and their divisions are defined by specific reference to the counties they encompass.”209 Moreover, the court observed, even state courts are not always defined by county:

Within Harrison County, [Mississippi,] for example, there are two state judicial districts. Even in counties within a single judicial district, there are multiple fora because Mississippi retains separate courts of law and equity, as well as courts of limited jurisdiction. There are three courts whose geographic jurisdiction encompasses Harrison County: the Chancery Court, the Circuit Court, and the County Court. Although Mississippi has rules delineating these courts’ respective jurisdictions, it can hardly be said that a reference to “county” clearly suggests the Harrison County Circuit Court rather than the United States District Court when it has a courthouse in, and jurisdiction over, Harrison County.210

The Fifth Circuit’s analysis calls into question whether selecting a “county” as the sole venue for resolving a dispute clearly signals an intent to litigate exclusively in state court. To date, however, the Tenth Circuit has declined to revisit its earlier decision and federal district courts within that circuit continue to apply this approach in determining the meaning of ambiguous forum selection clauses.211

VI.     Evaluating the Canons

In developing the canons outlined above, the courts have made a number of guesses as to what most contracting parties probably want their forum selection clauses to mean. To date, however, no one has ever sought to ascertain whether these judicial guesses are accurate. It is long past time that these rules were subjected to scrutiny to determine whether they do, in fact, produce results that are consistent with the expectations of most contracting parties. This Part proposes a methodology by which to answer this question.

For better or worse, judges do not always have the same expectations as commercial actors.212 Many of the interpretive rules discussed in the preceding Parts trace their origins to doctrines that are much more familiar to judges than to businesspeople. One test that the courts routinely use to determine the scope of a clause, for example, is similar to a test long used by courts to determine whether to exercise supplemental jurisdiction over a claim.213 Similarly, a court’s analysis of whether and to what extent a claim “arises out of” a contract for purposes of determining the scope of a forum selection clause is often informed by a judge’s past experience interpreting this same phrase in determining whether the courts have specific jurisdiction over a defendant.214 When judges are called upon to construe an ambiguous forum selection clause, they understandably draw upon their own experiences as judges in assigning meaning to that clause. It is worth asking, however, whether contract drafters understand this language to mean the same thing.215

A.     Methodology

With all of this in mind, I developed a questionnaire with ten questions—reproduced in the Appendix—in an attempt to find answers to these questions. In the summer of 2017, I circulated the questionnaire to dozens of individuals at law firms and in-house counsel’s offices across the United States. The questionnaire identified interpretive issues that frequently arise in litigation relating to forum selection clauses and asked the respondents how they generally wanted these interpretive issues to be resolved. In many cases, individuals forwarded the questionnaire to their colleagues at the same firm or company, which resulted in additional questionnaires being returned to me.216

Ultimately, I received 78 responses. I received responses from 35 lawyers in North Carolina, 11 in Texas, eight in New York, four in Oklahoma, four in South Carolina, three in Colorado, three in the District of Columbia, two in Illinois, two in Georgia, and one apiece in Arizona, California, Kansas, Minnesota, Tennessee, and Utah. Fifty-six respondents worked at law firms. Sixteen worked as in-house counsel. Two respondents worked at a company that publishes materials to help practicing attorneys with contract drafting. Two respondents had recently moved into legal academia after working as transactional attorneys at major law firms. And two respondents had recently retired after long careers working at major law firms. The respondents had been in practice for an average of 22 years.217 Among the firm lawyers, I received responses from 46 transactional attorneys and 16 commercial litigators.

I also conducted phone interviews with an additional eight attorneys in which I posed open-ended questions about how they dealt with forum selection clauses in their practice. Seven of these attorneys were based in North Carolina. One was based in Texas. Seven of these interviewees worked as transactional lawyers at large law firms. One worked as in-house counsel. All told, I interviewed or entered into structured e-mail exchanges with 86 attorneys in an attempt to better understand the expectations of the lawyers who draft and litigate forum selection clauses in the real world. The information gleaned from these exchanges comprises the core of the empirical analysis set forth below.

B.     Caveats

As with all empirical projects, this one is subject to several caveats. The lawyers who completed the questionnaire or who agreed to be interviewed were not randomly drawn from a cross-section of U.S. lawyers; they were drawn from my preexisting contacts and from referrals made by these contacts.218 The respondents were not representative of all U.S. lawyers; they were much more likely to work at large law firms or large companies than the typical attorney and were far more likely to live in North Carolina. When evaluated against the baseline of a randomized survey of many thousands of attorneys across the United States, in summary, the present study presents obvious limitations.

When evaluated against the baseline of the prevailing status quo, however, the utility of the data presented in this article snaps into focus. As things stand, judges called upon to interpret ambiguous language in boilerplate forum selection clauses must either rely on (1) evidence presented to them by self-interested litigants, or (2) the judge’s own intuitions as to how the clause should be interpreted. This Article surveys dozens of attorneys with no rooting interest in any particular case to ask them how they would resolve these same interpretive questions. On the whole it seems unlikely that their answers to these questions are less reliable than the existing alternatives. Indeed, one could make a strong case that these answers are likely to be more reliable because they were given by disinterested attorneys. This survey also lays the groundwork for future studies in which these same questions—or improved versions thereof—may be posed to a larger population of attorneys across a broader geographic area to get a still better sense for how attorneys generally want courts to construe ambiguous contract boilerplate.

One final caveat is in order. The insights set forth in the following Parts are principally relevant when construing forum selection clauses in contracts between and among business entities. They are of far less relevance when construing forum selection clauses in consumer contracts. The lawyers surveyed and interviewed spend their time either drafting and negotiating business contracts on behalf of business entities or representing these entities in commercial litigation. Their responses to the survey were given with this type of agreement and this type of client in mind.

C.     Possible Objections

Critics of the methodological approach set forth above might raise two objections. First, one might argue that expert testimony would be a superior means of determining the meaning of ambiguous contract boilerplate. In theory, expert testimony—in which a disinterested third party offers an unbiased view as to a provision’s meaning—could help address the question of how courts should construe ambiguous contract boilerplate. In practice, there are two problems with using expert testimony to achieve this goal. First, expert testimony will frequently be inadmissible. The courts have consistently held that it is the job of the court—not the expert—to interpret ambiguous contract language.219 In order for experts to testify as to the meaning of boilerplate forum selection clauses, therefore, this evidentiary rule would need to change. Second, expert witnesses are hired guns. As such, they are unlikely to provide disinterested or unbiased views as to the interpretive expectations of the typical contract drafter.220 Instead, they will testify in support of the interpretation that favors the interests of the party that hired them. A survey interpretation approach is thus superior to an approach that relies on expert testimony because it looks to disinterested third parties—not hired guns—for guidance as to the meaning of ambiguous boilerplate. Particularly when the survey is conducted in advance of any litigation, it arguably offers a more reliable guide to the interpretive preferences of the broader contracting community.

A second possible objection to the survey interpretation approach is that it is the responsibility of the contract drafter to familiarize himself or herself with the relevant interpretive rules before sitting down to draft the agreement. When one of these rules produces a result that is contrary to the preferences of the parties, the drafter should contract around it. If the drafter fails to draft around that rule, so this argument goes, then the fault lies with the drafter.

The problem with this argument is that it is far more efficient for a single court to adopt an interpretive default rule that aligns with majority preferences than for tens of thousands of contract users to redraft hundreds of thousands of agreements to correct for a judicial mistake. The mere fact that the parties have the ability to contract around idiosyncratic interpretive rules, in short, does not absolve the courts of the responsibility to construe contract boilerplate in a way that avoids inflicting costs on third parties in the first place. While lawyers should obviously take steps to familiarize themselves with these interpretive rules, it is not unfair to ask the courts to take special care to ensure that they adopt interpretive default rules that generally track majoritarian preferences.

This second objection also overlooks the impact that idiosyncratic interpretations of boilerplate can have on existing contracts. When a boilerplate provision has been written into thousands of contracts, and when a judge then assigns an idiosyncratic meaning to that provision, the judge has effectively rewritten the thousands of contracts already in existence. While it is theoretically possible for the parties to these agreements to amend all of these contracts to correct for this interpretive decision, such an endeavor would be time consuming and costly. It would be far better, in the first instance, for the courts to interpret the boilerplate in a manner that is consistent with the preferences of most drafters.

VII.     Survey Results and Implications

The information about attorney preferences when it comes to forum selection clauses is set forth below. The analysis is organized by canon. It first details the survey responses that bear on whether a particular canon produces outcomes that are consistent with attorney preferences. It then discusses the implications that flow from these findings.

A.     The Canons Relating to Exclusivity

The canons relating to exclusivity, it will be recalled, help the court determine whether the parties have chosen to litigate their dispute exclusively in the forum named in the clause or whether they parties have merely consented to jurisdiction or venue in that forum. The survey results with respect to these canons—and the implications that flow from these results—are set forth below.

1.     Survey Results

The canons relating to exclusivity present two empirical questions. First, are practicing attorneys generally aware of these canons? Second, do these canons produce outcomes that are consistent with the expectations of most contracting parties? The survey of 78 attorneys conducted in connection with this Article provides tentative answers to each of these questions.

The survey first asked whether the lawyer was familiar with the distinction between “exclusive” and “non-exclusive” forum selection clauses. Approximately 90% of the respondents reported that they were familiar with this distinction. I then presented each of these individuals with four forum selection clauses—all drawn from actual cases—and asked them to state whether these clauses were exclusive or non-exclusive. In so doing, I sought to determine whether the respondents could recognize language of exclusivity in context and then apply the proper label to the clause.

Approximately 90% of respondents correctly identified the first clause as exclusive.221 Approximately 94% correctly identified the second clause as non-exclusive.222 Approximately 97% correctly identified the third clause as non-exclusive.223 And approximately 92% of respondents correctly identified the fourth and final clause as exclusive.224Overall, these responses suggest that the respondents were quite familiar with the interpretive rules that judges use to determine whether a clause is exclusive or non-exclusive.

The survey then asked respondents whether they generally wanted a clause to be exclusive or non-exclusive. A majority of respondents (62%) stated that they typically preferred exclusive clauses. Some respondents (22%) observed that their preference varied depending on the situation. A minority of respondents (6%) stated that they typically preferred non-exclusive clauses. The remaining respondents (10%) reported that they were unfamiliar with the distinction between the two types of clauses and did not provide a response to the question.

Among those respondents who expressed a general preference for exclusive clauses, a significant number stated that they valued the certainty that came with knowing the forum in which any disputes would be litigated.225 Others noted that they placed a high premium on consistent case outcomes and that these outcomes were most likely to be obtained by litigating in just one forum.226 Still other respondents who preferred exclusive clauses reported that it was important to them that the chosen forum and the chosen law be the same and that the only way to guarantee this outcome was to make the forum selection clause exclusive.227

Many of the respondents who answered “it depends” reported that they wanted a clause to be exclusive when the chosen forum was in their client’s home jurisdiction and non-exclusive when it was not. As one respondent explained:

I typically want it to be exclusive if I have the contracting leverage and can dictate the forum. If that is the case, I’m typically going to assert home court advantage. If I don’t have the leverage, I may try to keep the contract silent so I can have the flexibility to choose the forum to initiate litigation, recognizing the counter-party will have the same option.228

In defense of this position, most respondents cited the convenience—and cost savings—of litigating close to home and the “home court” advantage that comes from litigating in such a venue.229

Other respondents with contingent views stated that their preference varied based on whether their client was more likely to be a plaintiff or a defendant in a lawsuit.230 If a company was more likely to be a defendant, they wanted the clause to be exclusive. If a company was more likely to be a plaintiff, they wanted the clause to be non-exclusive.231 Still others stated that they placed a premium on litigating before sophisticated judges who understood commercial law. These attorneys typically insisted on exclusive forum selection clauses selecting the courts in New York or Delaware and non-exclusive clauses selecting the courts of other states.232 Finally, a few respondents mentioned that they would sometimes write “co-exclusive” forum selection clauses in their contracts whereby each party agreed that it will bring any lawsuit in the home jurisdiction of the other party.233

The minority of respondents who expressed a general preference for non-exclusive clauses typically justified this preference by citing a desire for flexibility. One stated that non-exclusive clauses gave his “client[s] the ability to sue in the designated jurisdiction, while preserving the ability to sue elsewhere if appropriate.”234 Another observed that such clauses gave “more flexibility to my clients, who are often negotiating with money-center banks.”235

2.     Implications

The survey results suggest that the canons relating to exclusivity produce outcomes that are broadly consistent with the expectations of most contracting parties in the United States. The respondents were generally familiar with these interpretive rules. They also reported that they can—and often do—negotiate the language in the clause with an eye to these rules. Accordingly, there is no pressing need for the courts to revisit their past decisions in this area. Nor is there any pressing need to educate attorneys as to the content of these rules.

Why does the feedback loop between judges and drafters function so well in this context? There are several possible explanations. First, the case law explicating the distinction between exclusive and non-exclusive clauses dates back to the mid-1970s.236 This means that U.S. attorneys have had more than 40 years to acclimate themselves to the distinction. Second, the issue is highly salient to non-lawyers. While the business principals may not care all that much about whether they are litigating in state or federal court, they care a great deal about whether they are litigating in their home state or somewhere on the other side of the country. As one in-house attorney explained: “My basic bias is that a court where my institution has 35,000 employees is more likely to give me a fair hearing.”237 Lawyers are thus incentivized to learn about the relevant interpretive rules relating to exclusivity. Third, this particular issue is frequently litigated.238 There are hundreds of cases filed in the United States every year in which the court is called upon to decide whether a clause is exclusive or non-exclusive. This steady drumbeat of cases ensures that new lawyers entering the profession are exposed to this issue and made aware of the significance of particular words in their forum selection clauses.239

B.     The Canons Relating to Scope

The canons relating to scope, it will be recalled, help the court determine whether the forum selection clause applies exclusively to contract claims or whether it also applies to tort and statutory claims that relate in some way to the contract. The survey results with respect to these canons—and the implications that flow from these results—are set forth below.

1.       Survey Results

In an attempt to determine whether most lawyers want their forum selection clause to have a broad or narrow scope, I asked 78 attorneys about their preferences. Their responses are detailed in Figure 1 below.

Figure 1 Coyle

The most popular preference—expressed by 48% of respondents—was to have their forum selection clause apply to tort and statutory claims when these claims had some connection to the contract. The second-most popular preference—expressed by 45% of respondents—was for their clause to apply to any and all claims that related in some way to the business relationship between the parties. The least popular preference—stated by only 6% respondents—was for their clause to apply exclusively to contract claims. These findings suggest that the origination approach—which holds that generic forum selection clauses only apply to contract claims—is inconsistent with the preferences of most lawyers. They also suggest that a significant number of respondents want their clauses to apply even when the claim merely arises out of the business relationship between the parties.

When an attorney stated that he or she wanted forum selection clauses to apply to tort and statutory claims with some connection to the contract, I also asked that attorney to identify scenarios where such claims should be covered by a forum selection clause. The attorney was provided with a list of seven scenarios derived from the case law and asked to select those in which the clause should apply. Respondents were permitted to choose more than one scenario. These findings are detailed in Figure 2 below.

Figure 2 Coyle

The most popular answer—selected by 89% of respondents—was that forum selection clauses should cover non-contractual claims when these claims cannot be adjudicated without determining whether the defendant is in compliance with the contract. The second most popular answer—selected by 84% of respondents—was that these clauses should cover non-contractual claims where the court must construe the contract to resolve the claims. The third most popular answer—also selected by 84% of respondents—was that these clauses should cover non-contractual claims when they arise out of the same operative facts as a parallel contract claim. The fourth most popular answer—selected by 81% of respondents—was that these clauses should cover non-contractual claims when these claims relate in some way to the interpretation of the contract. In addition, 70% of respondents stated that a forum selection clause should apply to non-contractual claims when these claims ultimately depend on the existence of a contractual relationship between the parties.

I also asked the survey respondents whether they wanted their forum selection clauses to apply in two additional situations in which courts have sometimes held that parties do not want these clauses to apply. First, I asked whether they wanted their clauses to cover claims alleging that one party has made false or misleading statements about the other to third parties. Many courts have held that most parties do not want their generic forum selection clauses to cover claims for slander, defamation, or tortious interference.240 Second, I asked whether they wanted their clause to cover claims arising out of activities that predate the signing of the contract. Courts have generally held that most parties do not want their clauses to cover claims that predate the signing of the contract.241 By asking about lawyer preferences with respect to each of these issues, I was able to gain insights into whether these decisions were consistent with party expectations.

With respect to claims for slander, defamation, and tortious interference, only 38% of these respondents indicated that they generally wanted their forum selection clauses to cover these claims. With respect to claims arising out of activities that predate the contract, only 22% of these respondents expressed a desire for such coverage. Such results provide empirical support for the decisions in which courts have shied away from enforcing forum selection clauses in these situations. The results also suggest that the survey respondents were not mindlessly selecting among various interpretive rules presented to them. Instead, they expressed skepticism about applying forum selection clauses in precisely those situations where the courts have also expressed some skepticism.

Finally, where a respondent stated that he or she wanted a forum selection clause to apply where the claim depended on the existence of a contractual relationship, I asked whether he or she wanted the claims to apply when (1) the claim would not have arisen “but for” the contractual relationship between the parties, or (2) the contract was the “proximate cause” of these claims. Approximately 46% of these respondents preferred a “but for” causation test. Approximately 35% of these respondents preferred a “proximate cause” test. The remaining 19% of these respondents either stated that both tests should apply, that they did not know, or did not answer the questions. Although the differences between these two tests are significant in percentage terms, they are smaller in absolute terms. Out of a total of 26 respondents, 12 expressed a preference for a “but for” test. Nine respondents expressed a preference for a proximate cause test. Five respondents did not express a clear choice for either test.

2.     Implications

The foregoing analysis suggests several important insights. First, those courts that have held that non-contractual claims do not originate in the contract—and hence do not arise out of that contract—have adopted an interpretive default rule that is inconsistent with the expectations of most parties.242 Only 6% of the survey respondents expressed support for this approach. This finding suggests that those courts that have applied the origination test—a list that includes the Second Circuit—can and should revisit their past decisions on this issue.

Second, those courts that have traditionally applied just one of the interpretive approaches set forth above are probably taking too narrow an approach.243 When presented with seven possible scenarios as to when a claim might come within the scope of a contract’s forum selection clause, the overwhelming majority of survey respondents identified multiple scenarios when the claim should be covered. Courts that have historically applied just one of these tests to the exclusion of all others—a list that includes the First Circuit and the Ninth Circuit—should also rethink their approach to this issue.

Third, the hybrid approach pioneered by the Eighth Circuit would appear to be the one that most closely approximates the preferences of most contracting parties. Rather than focusing on just one test, the Eighth Circuit has held that courts should consider whether the non-contractual claims arise out of the same operative facts as a parallel contract claim and conduct an analysis of the contract and ask whether the claims depend on the existence of a contractual relationship.244 The survey respondents routinely identified multiple criteria from the various tests that the courts should use to determine whether a forum selection clause covered a non-contractual claim. In addition, since the de facto result of such an approach will result in these clauses being given a broader scope, this approach is likely to generate outcomes closer to the preferences of the significant number of respondents who wanted their clause to apply to all claims arising out of the business relationship.

Fourth, courts should err on the side of finding a non-contractual claim is covered in cases where the issue is close. Approximately 45% of respondents stated that they generally want their forum selection clauses to apply to all claims that relate in some way to the business relationship. Such an approach would address the concerns that some courts have raised about artful pleading and fragmented litigation proceedings.

Fifth, and finally, contract drafters who want their forum selection clauses to have a broad scope should take care to include phrases such as “related to” or “in connection with.” In the vast majority of cases, writing these phrases into the clause will make it unnecessary for the courts to apply any of the interpretive tests set forth above.

C.     The Canons Relating to Non-Signatories

The canons relating to non-signatories, it will be recalled, help the courts determine whether the forum selection clause binds parties that are affiliated with the contract signatory but who did not actually sign the contract. The survey results with respect to these canons—and the implications that flow from these results—are set forth below.

1.     Survey Results

To ascertain whether the “closely related” test produced results that were consistent with the expectations of most contracting parties, I posed the following question to the survey respondents:

Should a forum selection clause apply to a party that has not signed the contract if that party is “closely related” to a contract signatory and it is “foreseeable” that the non-signatory would be bound by the clause? For example, should a company president be entitled to invoke a forum selection clause in a case where the contract containing the clause was only signed by the company?

 

A narrow majority of respondents—51%—stated that non-signatories should not be covered by a forum selection clause on these facts. In support of this position, respondents cited the hornbook principle that contracts cannot bind parties who do not sign them.245 They also invoked third-party-beneficiary law and stated that, on these facts, the necessary conditions did not appear to have been satisfied for a non-signatory to benefit from the contract.246

Just over a quarter of respondents—26%—took the position that non-signatories should be covered by forum selection clauses under the scenario posed by the question. In their comments, these respondents emphasized the ability of plaintiffs to create mischief under the traditional (and stricter) law of third party beneficiaries.247 Some respondents also noted that the purpose of a forum selection clause was to provide predictability to the transaction and that having non-signatories be covered by the clause furthered that goal.248

Another 22% of respondents declined to answer “yes” or “no” to the question posed. One remarked that it would be “nice” if the closely related test covered non-signatories but “I don’t think that’s the law.”249 Another observed that “[t]his is a policy question and I don’t have an opinion on it. I have been on both sides of this issue in litigation.”250 Others opined that the closely related test was “too fuzzy” but expressed support for the general notion that a company employee being sued for acts taken within the scope of employment that somehow related to the contract should be covered by an exclusive forum selection clause.251 Still another observed that the clause “probably should apply for the benefit of the non-signatory if the underlying claim is somehow related to the contract” but went on to add that “it seems hard and perhaps unfair to bind the non-signatory if the claim is against him or her, without consent or some involvement in preparing the contract.”252

As discussed above, the courts have recognized the ability of the parties to contract around the canons relating to non-signatories by (1) stating explicitly that certain non-signatories are third-party beneficiaries to a contract or (2) stating explicitly that the contract is not intended to confer any rights on non-signatories.253 Accordingly, I also asked the survey respondents whether their standard contract includes a no-third-party-beneficiary clause. A substantial majority—79%—responded that they do typically include such a clause in their contracts. In these cases, the forum selection clause will not apply to non-signatories because it does not apply to any third-party beneficiaries. It is far from clear, however, that most attorneys recognize the connection between a forum selection clause and a no-third-party-beneficiary clause. As one survey respondent observed: “I doubt third party beneficiary rights [regarding] forum selection is always considered.”254 Approximately 15% of respondents reported that their standard contract does not contain a no-third-party-beneficiary clause. The remaining 8% of respondents either stated they did not know whether their standard contract contained a no-third-party-beneficiary clause or that it depended on the circumstances.

Finally, it is worth mentioning that several survey respondents commented that their no-third-party-beneficiary clauses frequently contain exceptions and carve-outs. In principle, this list of exceptions and carve-outs could be used to extend the scope of a forum selection clause to cover non-signatories. In practice, however, it appears that these exceptions are typically used to carve out third parties who are indemnitees under the agreement.255 No respondent indicated that he or she had ever expressly addressed the interplay between this clause and a forum selection clause in the context of analyzing the rights of non-signatories.

2.     Implications

The survey results suggest several insights into the canons relating to non-signatories. First, when lawyers are asked whether a forum selection clause covers a non-signatory, they are much more likely to reference third-party-beneficiary doctrine than the “closely related” test. This finding suggests that judges should rethink whether the “closely related” test produces results that are consistent with majoritarian expectations. While most lawyers want their clauses to have a broad scope in the sense that they want them to cover tort and statutory claims as well as contract claims, their survey responses indicate that they do not necessarily want these clauses to cover individuals who (1) never signed the contract in the first place, and (2) do not qualify as third-party beneficiaries under traditional doctrine. To the extent that the “closely related” test produces outcomes inconsistent with these expectations—and with longstanding third-party-beneficiary doctrine—it should be reexamined.

The second insight relates more to contract drafting than to judicial practice. A non-trivial number of the respondents—23%—reported that they wanted their forum selection clause to apply to non-signatories and that their standard contract included a no-third-party-beneficiary clause. Only one respondent acknowledged the tension between these two positions in his response. Going forward, lawyers should pay closer attention to the rights of non-signatories when drafting their contracts. If the parties want their forum selection clause to cover non-signatories, they should write language into the clause stating as much. Alternatively, this issue could be addressed by carving certain non-signatories out from a no-third-party-beneficiary clause. Either approach would give the courts better guidance as to party expectations and would mitigate the current disconnect between judicial practice and the reported expectations of the contract drafters who are often unaware of the interplay between their forum selection clause and their no-third-party-beneficiary clause.

D.     The Canons Relating to Federal Court

The canons relating to federal court, it will be recalled, help the courts determine whether the parties contemplated litigation in state court to the exclusion of federal court or whether they contemplated litigation in either state or federal court. The survey results with respect to these canons—and the implications that flow from these results—are set forth below.

1.     Survey Results

With respect to the canons relating to federal court, I asked the survey respondents whether they generally want to preserve the option of going to federal court when they are negotiating a forum selection clause. Approximately 79% of survey respondents stated that they always wanted to preserve the option of going to federal court.256 The most commonly cited justification for this position was that federal courts are less biased against out-of-state defendants than are state courts. As one respondent succinctly explained: “I would always want the federal court option, for fear of ‘home cooking.’”257

Other survey respondents—approximately 13%—were more equivocal in their responses. One respondent, for example, stated that he paid little attention to the state/federal distinction when negotiating forum selection clauses “unless there are concerns about the state court that would be involved.”258 Another stated that the answer to the question depended on the client: “if it is a ‘local’ entity, state court would be fine. If I represent a multi-state, multi-national entity . . . I want to have the option to go to [federal] court.”259 Still another observed that “it depends on the client and the types of potential claims that could be raised, but generally flexibility is preferred.”260

A minority of survey respondents—just 8%—stated that their general preference when negotiating forum selection clauses was to exclude altogether the possibility of going to federal court. Two respondents explained that they wanted to litigate exclusively in the Delaware Court of Chancery.261 One respondent stated that he “usually prefer[red] state courts for simplicity and costs reasons.”262 And one respondent stated that “[w]e select the state court, but reserve the option for federal court in the event the state court does not have jurisdiction.”263

Having ascertained the baseline preferences of the survey respondents, I then sought to ascertain their familiarity with the relevant interpretive rules by asking them to read two forum selection clauses patterned on clauses from actual cases. The first stated that: “The courts of Texas shall have sole and exclusive jurisdiction over all disputes arising out of this Agreement.” The second stated that: “You hereby consent to the sole and exclusive jurisdiction and venue of courts in King County, Washington in all disputes arising out of this Agreement.” When the canons relating to federal court are applied to construe the clauses, the latent ambiguities relating to a state or federal forum are easily answered. The first clause—which uses the word “of” instead of “in”—will be read to select the Texas state courts and to exclude the federal courts in Texas. The second clause—which uses the word “in” rather than “of”—will generally be read to permit litigation in the federal courts so long as a federal courthouse is physically located in King County. In the Tenth Circuit, however, the second clause will be read to exclude the federal courts because the clause refers to a county rather than a judicial district.

In construing these two clauses, approximately 21% of respondents stated that they perceived no legal distinction between the two clauses. These answers suggest that these respondents were generally unaware of the interpretive rules discussed above. An additional 49% of respondents commented on legal issues that were unrelated to the availability of a federal forum.264 These responses suggest that (1) a significant number of respondents were either unaware of the significance the courts have attached to the words “of” and “in” and “county” in forum selection clauses, or (2) they viewed the potential availability of a federal forum as less important than other issues presented by the precise wording of the clause.

The remaining 30% of respondents specifically addressed the question of whether litigation in a federal forum was possible under at least one of the clauses. With respect to the first clause, 14 respondents correctly predicted that the clause selected the Texas state courts exclusively. By contrast, seven respondents guessed incorrectly that the clause permitted litigation in federal court. With respect to the second clause, 17 respondents stated that the clause would permit litigation in the federal courts if King County contained a federal courthouse. By contrast, only one respondent stated that the clause selected the Washington state courts exclusively.

2.     Implications

The findings set forth above suggest three important insights with respect to the canons relating to federal court. First, most parties generally want to preserve their option of going to federal court when drafting a forum selection clause. This preference was expressed by 79% of survey respondents. Only 8% of survey respondents stated that they generally want their forum selection clauses to exclude the possibility of going to federal court. In close cases, therefore, courts should err on the side of finding that the parties did not intend to foreclose the possibility of litigating their dispute in federal court because this is the outcome that aligns with the expectations of most contracting parties. In practice, many courts give effect to this preference by requiring “clear and unequivocal” evidence that the defendant waived its right to remove.265 The survey suggests that this standard should be retained and vigorously applied by courts called upon to construe ambiguous forum selection clauses.

Second, many lawyers—even sophisticated lawyers engaged in transactional practice at major law firms—are unaware of the linguistic distinction between “of” and “in” when it comes to forum selection clauses. When presented with two forum selection clauses in which these words appeared, approximately 70% of the survey respondents were seemingly unaware of their significance in determining the availability of a federal forum. Where the survey respondents spotted this issue, they generally agreed with the courts that the word “in” connoted an intent to litigate in either state or federal courts. There was, however, disagreement among the respondents as to the legal significance of the word “of.”266 In the future, courts should take these disagreements as to meaning—as well as the general lack of awareness regarding the issue—into account in construing ambiguous clauses.

Third, and finally, not a single respondent—even those based in Oklahoma and Colorado, two states within the Tenth Circuit—invoked the distinction between “counties” and “judicial districts” drawn by the Tenth Circuit when commenting on the clauses above. With respect to this rule in particular, there is scant evidence that the canon of construction developed by the Tenth Circuit produces outcomes that are consistent with the expectations of the parties at the time of drafting. Accordingly, the Tenth Circuit can and should revisit its earlier decisions and adopt a new interpretive default rule that is more in line with party expectations and decisions rendered by the other federal circuits.

VIII.     Conclusion

Whenever one proposes a new approach to solving an old problem, the burden is on the innovator to show that the new approach is better than the old ones. The question of how courts should interpret ambiguous contract boilerplate is an old problem. The traditional solution required litigants to make their interpretive arguments to a judge who then drew upon his or her own intuitions as to what most parties would probably want the clause to mean. There is nothing inherently wrong with this approach. However, it sometimes results in idiosyncratic interpretations that inflict costs on third parties whose contracts contain similar language.  This Article suggests that an alternative approach—one that looks to interviews and surveys for insights into what most parties want the provision to mean—can reduce these costs. While the Article utilized this approach to interpret the forum selection clause, it could just as easily be utilized to interpret other common boilerplate provisions that are routinely written into commercial contracts. 

If a contract is clearly drafted, of course, there will be no need for the courts to invoke any of the canons discussed above. Contract drafters should therefore aspire to state their intentions clearly, thereby making it unnecessary for the courts to construe a clause. The table below seeks to assist them in this task.

Table 1 Coyle

The table distills the essence of the judicial decisions surveyed in this Article into a simple and straightforward guide to contract drafters. With any luck, future drafters will pay heed to this guidance and incorporate the insights into their agreements. Until the day when all forum selection clauses state the intentions of the parties in such clear and unequivocal language, however, the courts will continue to rely on the canons discussed above to construe forum selection clauses.

  

Appendix

Forum Selection Clause Questionnaire

 

  1. Are you familiar with the distinction between “exclusive” and “non-exclusive” forum selection clauses? (This is sometimes framed as a distinction between “mandatory” and “permissive” clauses.)

    If “no,” please skip ahead to Question 4.
    If “yes,” please proceed to the next question.

  2. When you are negotiating a forum selection clause, do you generally want the clause to be exclusive or non-exclusive? What factors play into this decision?

  3. Please review each of the forum selection clauses below. Without conducting any research, please state whether the clause is exclusive (E) or non-exclusive (N) in the adjacent box.

    Table 2 Coyle
  4. When you are negotiating a forum selection clause, do you generally want a clause that requires the parties to litigate disputes exclusively in state court? Or do you generally want to preserve the option of going to federal court?

  5. Do you perceive any legally significant distinction between the two forum selection clauses below? If so, please explain.

    A.   “The courts of Texas shall have sole and exclusive jurisdiction over all disputes arising out of this Agreement.”

    B.   “You hereby consent to the sole and exclusive jurisdiction and venue of courts in King County, Washington in all disputes arising out of this Agreement.”

  6. Please read the three statements below. Mark the statement that best captures your baseline preference when it comes to forum selection clause scope.

    A.    I want my forum selection clause to cover breach of contract claims exclusively. If I wish to bring a tort or statutory claim against my counterparty, I may ignore the clause. Conversely, if my counterparty wishes to bring a tort or statutory claim against me, it may ignore the clause.

    B.    I want my forum selection clause to cover claims for breach of contract as well as tort and statutory claims that have some connection to the contract.

    C.    I want my forum selection clause to cover all contract, tort, and statutory claims that relate in some way to the business relationship. Even if these claims have nothing to do with the contract, I still want them to be covered by my forum selection clause.

    If you selected statement A or statement C, please skip ahead to Question 9.

    If you selected statement B, please proceed to the next question.

  7. Please put an “X” next to each statement that describes a situation where you believe a tort or statutory claim asserted by one contract counterparty against the other should be covered by a forum selection clause. You may select more than one statement.

    Table 3 Coyle

  8. If you did not put an “X” next to the final statement listed above—expressing a preference for coverage when the claim depends on the existence of a contractual relationship between the parties—please proceed to Question 9.

    If you did put an “X” next to this final statement, how closely related should the claims be to the existence of the contractual relationship? Do you want the clause to cover tort and statutory claims where these claims would not have arisen “but for” the contractual relationship between the parties? Or do you want the clause to cover tort and statutory claims only when the contract is the “proximate cause” of these claims?

  9. Should a forum selection clause apply to a party that has not signed the contract if that party is “closely related” to a contract signatory and it is “foreseeable” that the non-signatory would be bound by the clause? For example, should a company president be entitled to invoke a forum selection clause in a case where the contract containing the clause was only signed by the company?

  10. Do you typically include in your contracts a clause stating that the contract is “not intended to confer any rights or remedies upon any person other than the parties and their permitted successors and assigns” or other language to that effect?

  1. [1]. See Matthew D. Cain & Steven M. Davidoff, Delaware’s Competitive Reach, 9 J. Empirical Legal Stud. 92, 94 (2012) (finding that 60% of the merger agreements in the sample contained forum selection clauses with Delaware as their choice of forum); Ya-Wei Li, Note, Dispute Resolution Clauses in International Contracts: An Empirical Study, 39 Cornell Int’l L.J. 789, 797–99 (2006) (finding that 67% of “merger, acquisition, stock exchange and share exchange, reorganization, and combination contracts filed with the [SEC] between January 1, 2002 and March 31, 2003 and involving at least one foreign party” contained a forum selection clause).

  2. [2]. Compare Samuel Williston & Richard A. Lord, A Treatise on the Law of Contracts § 15:15, at 290–301 (4th ed. 1997) (“Formerly, no agreement confining the right of a party to sue in a particular court or tribunal or in the courts or tribunals of a certain jurisdiction, or to determine the venue of a suit in such a way as to deprive the defendant of his statutory privileges as to place of trial was enforced, unless perhaps where the agreement was made after the cause of action had arisen and was part of a fair compromise.”), with Martinez v. Bloomberg LP,
    740 F.3d 211, 217 (2d Cir. 2014) (“If the forum clause was communicated to the resisting party, has mandatory force and covers the claims and parties involved in the dispute, it ispresumptively enforceable.” (quoting Phillips v. Audio Active Ltd., 494 F.3d 378, 383 (2d Cir. 2007))).

  3. [3]. See Scherk v. Alberto-Culver Co., 417 U.S. 506, 519 n.14 (1974); M/S Bremen v. Zapata Off-Shore Co., 407 U.S. 1, 11 (1972). In 1991, the Supreme Court held that a non-negotiated forum selection clause in a form ticket contract was enforceable even though the plaintiff resided in Washington and the clause required the suit to be brought in Florida. Carnival Cruise Lines, Inc. v. Shute, 499 U.S. 585, 595 (1991). In reaching this conclusion, the Court noted that the clause was reasonable, that Florida was not a remote alien forum, the plaintiff did not claim any lack of notice, and that there was no evidence of fraud or overreaching. Id. at 592–95.

  4. [4]. See, e.g., Hannah L. Buxbaum, Forum Selection in International Contract Litigation: The Role of Judicial Discretion, 12 Willamette J. Int’l L. & Disp. Resol. 185, 192–97 (2004); Robin J. Effron, Ousted: The New Dynamics of Privatized Procedure and Judicial Discretion, 98 B.U. L. Rev. 127, 160–61 (2018); Michael Gruson, Forum-Selection Clauses in International and Interstate Commercial Agreements, 1982 U. Ill. L. Rev. 133, 137; Linda S. Mullenix, Another Choice of Forum, Another Choice of Law: Consensual Adjudicatory Procedure in Federal Court,57 Fordham L. Rev. 291, 294 (1988); Leandra Lederman, Note, Viva Zapata!: Toward a Rational System of Forum-Selection Clause Enforcement in Diversity Cases, 66 N.Y.U. L. Rev. 422, 424 (1991); see also Gary B. Born
    & Peter B. Rutledge, International Civil Litigation in United States Courts 461–546
    (5th ed. 2011) (discussing role of forum selection clauses in international civil litigation).

  5. [5]. One threshold issue that frequently arises when a court is called upon to interpret a forum selection clause is what state’s interpretive law to apply. When the contract containing a forum selection clause also contains a choice-of-law clause—as it frequently does—the consensus among commentators is to apply the interpretive law of the state named in the choice-of-law clause. See Hannah L. Buxbaum, The Interpretation and Effect of Permissive Forum Selection Clauses Under U.S. Law, 66 Am. J. Comp. L. 127, 151–52 (2018); Kevin M. Clermont, Governing Law on Forum-Selection Agreements, 66 Hastings L.J. 643, 673 (2015); Symeon C. Symeonides, What Law Governs Forum Selection Clauses, 78 La. L. Rev. 1119, 1152 (2018); see also Collins v. Mary Kay, Inc., 874 F.3d 176, 183 (3d Cir. 2017). But see Sun v. Advanced China Healthcare, Inc., 901 F.3d 1081, 1086 (9th Cir. 2018) (applying federal common law to interpret forum selection clause). When the contract lacks a choice-of-law clause, the prevailing view is that the court should apply the interpretive law of the jurisdiction chosen after conducting a conflict-of-laws analysis. See Buxbaum, supra, at 149–51; Clermont, supra, at 673; Symeonides, supra, at 1152; cf. Barnett v. DynCorp Int’l, L.L.C., 831 F.3d 296, 302–03 (5th Cir. 2016) (discussing this issue).

  6. [6]. For one of the vanishingly few academic works discussing the question of how courts should interpret forum selection clauses, see Eric Sherby, Forum Selection Clauses in International Commerce, in 1 International Aspects of U.S. Litigation: A Practitioner’s Deskbook 267, 280–94 (James E. Berger ed., 2017); see also Peter Hay et al., Conflict of Laws 1146–49 (5th ed. 2010). See generally Buxbaum, supra note 5 (providing brief overview of interpretive issues relating to forum selection clauses).

  7. [7]. The term “boilerplate” is typically understood to refer to “[r]eady-made or all-purpose language that will fit in a variety of documents.” Boilerplate, Black’s Law Dictionary (10th ed. 2014).

  8. [8]. Hartford Cas. Ins. Co. v. Swift Distribution, Inc., 326 P.3d 253, 259 (Cal. 2014); Greenfield v. Philles Records, Inc., 780 N.E.2d 166, 170 (N.Y. 2002).

  9. [9]. Hartford Cas. Ins. Co., 326 P.3d at 259; Greenfield, 780 N.E.2d at 170.

  10. [10]. See Stephen J. Choi et al., The Black Hole Problem in Commercial Boilerplate, 67 Duke L.J. 1, 15 (2017); Stephen J. Choi et al., Variation in Boilerplate: Rational Design or Random Mutation?,
    20 Am. L. & Econ. Rev. 1, 3 (2018); John F. Coyle & W. Mark C. Weidemaier, Interpreting Contracts Without Context, 67 Am. U. L. Rev. 1673, 1676–78 (2018).

  11. [11]. See John F. Coyle, The Canons of Construction for Choice-of-Law Clauses, 92 Wash. L. Rev. 631, 682–87 (2017) (arguing that courts should interpret boilerplate contract provisions in a manner consistent with majoritarian expectations).

  12. [12]. The doctrine of contra proferentem, which holds that contracts should be construed against the drafter, will occasionally trump each of these canons. See, e.g., Stateline Power Corp. v. Kremer, 148 F. App’x 770, 771 (11th Cir. 2005); Glob. Satellite Commc’n Co. v. Starmill U.K. Ltd., 378 F.3d 1269, 1273–74 (11th Cir. 2004); Carolina Pole, Inc. v. FirstEnergy Corp., No. COA08-829, 2009 WL 1522696, at *8–11 (N.C. Ct. App. June 2, 2009). In other cases, the doctrine of contra proferentem is invoked to provide additional support for an interpretive decision that the court has already reached. See, e.g., Troy Corp. v. Schoon, No. C.A.1959-VCL, 2007 WL 949441, at *4–5 (Del. Ch. Mar. 26, 2007); Granados Quinones v. Swiss Bank Corp. (Overseas), S.A., 509 So. 2d 273, 274 (Fla. 1987); Boland v. George S. May Int’l Co., 969 N.E.2d 166, 174 (Mass. App. Ct. 2012). The precise impact of contra proferentem may also vary depending upon the court and the context. See Terra Int’l, Inc. v. Miss. Chem. Corp., 119 F.3d 688, 692 (8th Cir. 1997) (declining to apply the doctrine of contra proferentem to construe forum selection clause “due to the relatively equal bargaining strengths of both parties and the fact that Terra was represented by sophisticated legal counsel during the formation of the license agreement”); 11 Richard A. Lord, Williston on Contracts § 32.12 (4th ed. 2018) (“The rule of contra proferentem is generally said to be a rule of last resort and is applied only when other secondary rules ofinterpretation have failed to elucidate the contract’s meaning.”).

  13. [13]. See supra note 6 (surveying limited literature on the topic).

  14. [14]. See, e.g., Nedlloyd Lines B.V. v. Superior Court of San Mateo Cty., 834 P.2d 1148, 1154–55 (Cal. 1992) (interpreting boilerplate choice-of-law clause in a manner consistent with the expectation of a “rational businessperson”); Bank of N.Y. Mellon Tr. Co. v. Liberty Media Corp., 29 A.3d 225, 241 (Del. 2011) (“[I]n interpreting boilerplate indenture provisions, courts will not look to the intent of the parties, but rather the accepted common purpose of such provisions.” (citations omitted) (internal quotation marks omitted)); Fiona Trust & Holding Corp. v. Privalov [2007] UKHL 40 (referencing the expectations of “[o]rdinary businessmen” in concluding that the words “arising out of” in arbitration clause should be read to cover “every dispute except a dispute as to whether there was ever a contract at all”).

  15. [15]. Aaron D. Goldstein, The Public Meaning Rule: Reconciling Meaning, Intent, and Contract Interpretation, 53 Santa Clara L. Rev. 73, 90–91 (2013).

  16. [16]. See Stephen J. Choi & G. Mitu Gulati, Contract as Statute, 104 Mich. L. Rev. 1129,
    1131–32 (2006).

  17. [17]. This phenomenon may be usefully analogized to the “butterfly effect” described by chaos theorists. See James Gleick, Chaos: Making a New Science 8 (1987). Just as a butterfly flapping its wings in Tokyo can change the weather in London, so too can a judicial decision interpreting contract boilerplate change the meaning of the language in thousands of other contracts governed by the same law. To speak of the “butterfly effect” in boilerplate contract interpretation, therefore, is to describe the effect that a single interpretive decision can have on the interests of far-flung parties not involved in the litigation at hand.

  18. [18]. In a similar vein, a number of scholars have recently conducted interviews with individuals tasked with drafting statutes and agency rules at the federal level in order to determine whether they were familiar with the canons of construction that courts use to construe these statutes. See generally Abbe R. Gluck & Lisa Schultz Bressman, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part I, 65 Stan. L. Rev. 901 (2013) (interviewing a number of government officials); Lisa Schultz Bressman & Abbe R. Gluck, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part II, 66 Stan. L. Rev. 725 (2014) (conducting a survey of a number of government officials); Christopher J. Walker, Inside Agency Statutory Interpretation, 67 Stan. L. Rev. 999, 1013 (2015) (discussing survey methodology of government officials).

  19. [19]. This Article is not the first to note the advantages of an interpretive approach that looks to surveys and interviews to construe ambiguous contract language. See generally Omri Ben-Shahar & Lior Jacob Strahilevitz, Interpreting Contracts Via Surveys and Experiments, 92 N.Y.U. L. Rev. 1753 (2017) (describing a method of interpreting the language of contracts by polling relevant respondents in surveys). And Mitu Gulati and Bob Scott have sought to gain insight into the meaning of a particular piece of contract boilerplate—the pari passu clause—that appears in many sovereign debt contracts by interviewing the lawyers who draft these contracts. See generally Mitu Gulati & Robert E. Scott, The Three and a Half Minute Transaction: Boilerplate and the Limits of Contract Design (2013) (relying on lawyer interviews to show that contract boilerplate is sometimes resistant to change); Mark Weidemaier et al., Origin Myths, Contracts, and the Hunt for Pari Passu, 38 Law & Soc. Inquiry 72 (2013). This Article is, however, among the first to make the case for relying on lawyer surveys and interviews to construe “ordinary” boilerplate provisions such as forum selection clauses. See generally Coyle, supra note 11 (using a similar methodological approach to assist courts called upon to construe language in boilerplate choice-of-law clauses).

  20. [20]. The methodological approach utilized in this Article is broadly similar to the “survey interpretation method” recently proposed by Ben-Shahar and Strahilevitz. See generally Ben-Shahar & Strahilevitz, supra note 19. However, there are important differences between their approach and the one utilized here. Ben-Shahar and Strahilevitz are principally interested in surveying consumers in an attempt to determine how substantive provisions in consumer contracts should be interpreted. This Article, by comparison, is principally interested in surveying lawyers in an attempt to ascertain how boilerplate provisions relating to dispute resolution in business-to-business contracts should be interpreted. While each approach makes perfect sense in context, the differences between the two inquiries necessitate different methodological approaches. It is much easier to conduct online surveys of consumers, for example, than it is to conduct online surveys of practicing attorneys. See infra note 218 (discussing methodological challenges of surveying attorneys). Nevertheless, the method proposed by Ben-Shahar and Strahilevitz and the approach used in this Article share the same basic insight—the best way to determine what certain types of contracts mean may be to ask a great many people in the relevant interpretive community what, exactly, they think these contracts mean.

  21. [21]. See Atl. Marine Constr. Co. v. U.S. Dist. Court for the W. Dist. of Tex., 571 U.S. 49, 64 (2013); Stephen E. Sachs, The Forum Selection Defense, 10Duke J. Const. L. & Pub. Pol’y1, 1 (2015).

  22. [22]. Compare Am. Soda, LLP v. U.S. Filter Wastewater Grp., Inc., 428 F.3d 921, 927 (10th Cir. 2005) (concluding that a clause providing that the parties “submit to the jurisdiction of the Courts of the State of Colorado and agree that the Courts of the State of Colorado ... shall be the exclusive forum for the resolution of any disputes” was exclusive), with Animal Film, LLC v. D.E.J. Prods., Inc., 123 Cal. Rptr. 3d 72, 77 (Ct. App. 2011) (concluding that a clause providing that the parties “submit and consent to the jurisdiction of the courts present in the State of Texas” was non-exclusive).

  23. [23]. Compare Rafael Rodríguez Barril, Inc. v. Conbraco Indus., Inc., 619 F.3d 90, 92 (1st Cir. 2010) (concluding that a clause providing that the parties “consent and agree that jurisdiction for such action will lie only in the state and federal courts sitting in Mecklenburg County, North Carolina” was exclusive), with Luffey ex rel. Fredericksburg Props. of Tex., LP v. Fredericksburg Props. of Tex., LP, 862 So. 2d 403, 406 (La. Ct. App. 2003) (concluding that a clause providing that the parties agree that “Travis County, Texas shall be appropriate jurisdiction and venue for any litigation with respect to the Partnership” was non-exclusive).

  24. [24]. See BAE Sys. Tech. Sol. & Servs., Inc. v. Republic of Korea’s Def. Acquisition Program Admin., 884 F.3d 463, 472 (4th Cir. 2018); San Diego Gas & Elec. Co. v. Gilbert, 329 P.3d 1264, 1268 (Mont. 2014); Packers Printing & Publ’g Co. v. Anajet, LLC, No. COA13-1449, 2014 WL 4081765, at *4 (N.C. Ct. App. Aug. 19, 2014).

  25. [25]. See Starkey v. G Adventures, Inc., 796 F.3d 193, 196 (2d Cir. 2015); Verdugo v. Alliantgroup, L.P., 187 Cal. Rptr. 3d 613, 617 (Ct. App. 2015); Laibe Corp. v. Gen. Pump
    & Well, Inc., 733 S.E.2d 332, 336–37 (Ga. Ct. App. 2012).

  26. [26]. Rivera v. Centro Médico de Turabo, Inc., 575 F.3d 10, 14 (1st Cir. 2009).

  27. [27]. Cate St. Capital, Inc. v. DG Whitefield, LLC, No. 2182013CV00734, 2013 WL 10399704, at *2 (N.H. Super. Ct. Nov. 25, 2013) (emphasis added).

  28. [28]. Am. Soda, LLP v. U.S. Filter Wastewater Grp., Inc., 428 F.3d 921, 924 (10th Cir. 2005) (emphasis added).

  29. [29]. Rafael Rodriguez Barril, Inc. v. Conbraco Indus., Inc., 619 F.3d 90, 92 (1st Cir. 2010) (emphasis added).

  30. [30]. Mueller v. Sample, 93 P.3d 769, 772 (N.M. Ct. App. 2004) (emphasis added).

  31. [31]. See Caperton v. A.T. Massey Coal Co., 690 S.E.2d 322, 327 (W. Va. 2009) (“If jurisdiction is specified with mandatory terms such as ‘shall,’ ... the clause will be enforced as a mandatory forum-selection clause.”); see also Claudio-de León v. Sistema Universitario Ana G. Méndez,
    775 F.3d 41, 46 (1st Cir. 2014) (observing that “it is axiomatic that the word ‘shall’ has a mandatory connotation”).

  32. [32]. Slater v. Energy Servs. Grp. Int’l, Inc., 634 F.3d 1326, 1328–29 (11th Cir. 2011).

  33. [33]. Id. at 1330.

  34. [34]. Hunt Wesson Foods, Inc. v. Supreme Oil Co., 817 F.2d 75, 76 (9th Cir. 1987).

  35. [35]. Id. at 77; see also Weber v. PACT XPP Techs., AG, 811 F.3d 758, 768 (5th Cir. 2016) (stating that a forum selection clause “is mandatory only if it contains clear language specifying that litigationmust occur in the specified forum—and language merely indicating that the courts of a particular place ‘shall have jurisdiction’ (or similar) is insufficient”); Caldas & Sons, Inc. v. Willingham, 17 F.3d 123, 127 (5th Cir. 1994) (“Even though the clause now before us uses ‘shall,’ which is generally mandatory, this clause need not necessarily be classified as mandatory.”). The above list of exclusive words and phrases is not exhaustive; there are other ways for the parties to signal their intent to resolve their disputes exclusively in the chosen forum. The clause in question, however, must contain “clear” or “specific” language to this effect. See Weber, 811 F.3d at 768 (requiring “clear language”); Claudio-de León, 775 F.3d at 46 (same); Am. Soda, LLP v. U.S. Filter Wastewater Group, Inc., 428 F.3d 921, 926 (10th Cir. 2005) (same); N. Cal. Dist. Council of Laborers v. Pittsburg-Des Moines Steel Co., 69 F.3d 1034, 1037 (9th Cir. 1995) (requiring “clear[] language”); see also Global Seafood Inc. v. Bantry Bay Mussels Ltd., 659 F.3d 221, 221 (2d Cir. 2011) (requiring “specific language of exclusion”); IntraComm, Inc. v. Bajaj, 492 F.3d 285, 290 (4th Cir. 2007) (same); 14D Charles Alan Wright et al., Federal Practice and Procedure § 3803.1 (3d ed. 1998); cf. Troy Corp. v. Schoon, No. C.A.1959-VCL, 2007 WL 949441, at *9 (Del. Ch. Mar. 26, 2007) (requiring “express language”). In only a few jurisdictions do the courts apply “ordinary contract principles” to resolve this question. See, e.g., Liles v. Ginn-La West End, Ltd., 631 F.3d 1242, 1255 n.20 (11th Cir. 2011). While the simplest way of establishing exclusivity is to use one of the magic words—“exclusive,” “shall,” “sole,” or “must”—there are other possible words and phrases that will achieve this same end. Rivera v. Centro Médico de Turabo, Inc., 575 F.3d 10, 17 n.5 (1st Cir. 2009) (“We reject appellant’s argument that the absence of ‘typical mandatory terms’ such as ‘shall,’ ‘exclusive,’ ‘only,’ or ‘must’ requires a contrary result.”); Michaluk v. Credorax (USA), Inc., 164 So. 3d 719, 725 (Fla. Dist. Ct. App. 2015) (“Although we noted that the clause did not contain the ‘“magic words” “shall” or must,”’ the parties nevertheless employed language of exclusivity.” (quoting Celistics, LLC v. Gonzalez, 22 So. 3d 824, 826 (Fla. Dist. Ct. App. 2009))).

  36. [36]. In the years that followed the Supreme Court’s decision in M/S Bremen, a few courts took the position that all forum selection clauses were exclusive. In Republic Int’l Corp. v. Amco Eng’rs, Inc., 516 F.2d 161, 168 n.11 (9th Cir. 1975), for example, the Ninth Circuit concluded that a clause stating “the contracting parties place themselves under the jurisdiction and competence of the courts of the Republic of Uruguay” was a mandatory forum selection clause. Today, such a clause would likely be classified as permissive.

  37. [37]. See, e.g., Offshore Exploration & Prod., LLC v. Morgan Stanley Private Bank, N.A., 626 F. App’x 303, 306 (2d Cir. 2015) (holding a clause in which “parties[] consent[ed] to jurisdiction” was permissive).

  38. [38]. See Buxbaum, supra note 5, at 135–40 (discussing distinction between permissive and mandatory clauses).

  39. [39]. See Hoffman v. Blaski, 363 U.S. 335, 343 (1960) (observing that “venue, like jurisdiction over the person, may be waived”).

  40. [40]. See Davis v. Hall, 733 S.E.2d 878, 880 (N.C. Ct. App. 2012) (“[A] forum selection clause designates the venue, a consent to jurisdiction clause waives personal jurisdiction and venue
    ... .” (quoting Corbin Russwin, Inc. v. Alexander’s Hardware, Inc., 556 S.E.2d 592, 596 (N.C. Ct. App. 2001))).

  41. [41]. See supra notes 22–23 and accompanying text.

  42. [42]. Animal Film, LLC v. D.E.J. Prods., Inc., 123 Cal. Rptr. 3d 72, 75 (Ct. App. 2011) (emphasis added).

  43. [43]. State ex rel. Cordray v. Makedonija Tabak 2000, 937 N.E.2d 595, 600–01 (Ohio Ct. App. 2010) (emphasis added).

  44. [44]. Luffey ex rel. Fredericksburg Props. of Tex., LP v. Fredericksburg Props. of Tex., LP,
    862 So. 2d 403, 406 (La. Ct. App. 2003).

  45. [45]. There also exist so-called “hybrid” clauses that function as consent-to-jurisdiction or consent-to-venue clauses until suit is filed. Up until this point, the parties may sue wherever they want. Once a lawsuit is filed, however, the clause becomes mandatory because the parties have waived their right to seek a transfer of the suit to another court. See Ocwen Orlando Holdings Corp. v. Harvard Prop. Tr., LLC, 526 F.3d 1379, 1381 (11th Cir. 2008).

  46. [46]. John Wyeth & Brother Ltd. v. Cigna Int’l Corp., 119 F.3d 1070, 1075 (3d Cir. 1997) (“Whether or not a forum selection clause applies depends on what the specific clause at issue says.”); New Moon Shipping Co. v. MAN B & W Diesel AG, 121 F.3d 24, 33 (2d Cir. 1997) (“The scope of the forum selection clause is a contractual question that requires the courts to interpret the clause and, where ambiguous, to consider the intent of the parties.”). Whether the parties should be granted the autonomy to choose the law and forum for tort claims before the tort is committed is a question that has attracted little attention in the United States. In many other countries, the courts will only enforce such clauses when they are negotiated after the tort has been committed. Cf. Hay et al., supra note 6, at 1142 (“In the rest of the world, the prevailing view is that the parties may choose the applicable law only after the occurrence of the tort.”).

  47. [47]. The broadest possible version of the forum selection clause is one stipulating the forum for any claim arising out of the parties’ relationship. See Cascade Promotion Corp. v. AMA Sys., LLC, No. C 07-1513 CW, 2007 WL 1574544, at *1 (N.D. Cal. May 30, 2007) (concerning a clause stating that the parties “consent to theexclusivejurisdiction of thecourtsof the State of Maryland, USA for any disputearising out oftheirrelationship”); Hoes of Am., Inc. v. Hoes, 493 F. Supp. 1205, 1208 (C.D. Ill. 1979) (concerning a clause choosing a forum for disputes “arising out of the relationship between the parties”). Such a clause will cover claims that are completely unrelated to the contract so long as they relate in some way to the relationship between the parties—as they almost always will.

  48. [48]. Slater v. Energy Servs. Grp. Int’l, Inc., 634 F.3d 1326, 1328–29 (11th Cir. 2011) (emphasis added).

  49. [49]. Stewart Org., Inc. v. Ricoh Corp., 810 F.2d 1066, 1073 n.4 (11th Cir. 1987); see also
    In re
    McGraw-Hill Glob. Educ. Holdings LLC, 909 F.3d 48, 68 (3d Cir. 2018) (concluding that the word “regarding” was the equivalent of the phrase “relates to” and broader than the phrase “arising out of”); Sun v. Advanced China Healthcare, Inc., 901 F.3d 1081, 1086 (9th Cir. 2018) (concluding that phrase “related to” covered a tort claim for fraud); Bartels ex rel. Bartels v. Saber Healthcare Grp., LLC, 880 F.3d 668, 678 (4th Cir. 2018) (“The ‘in connection with’ language broadens the scope of the clause beyond pure contract claims and extends it to ‘every dispute between the parties having a significant relationship to the contract regardless of the label attached to the dispute.’” (quoting J.J. Ryan & Sons, Inc. v. Rhone Poulenc Textile, S.A., 863 F.2d 315, 321 (4th Cir. 1988))); WMW Mach., Inc. v. Werkzeugmaschinenhandel GmbH IM Aufbau, 960 F. Supp. 734, 747 (S.D.N.Y. 1997) (considering a forum selection clause governing “anydispute arising out of or in connection with” an agreement and holding that the clause covered a wrongful conversion claim); Caperton v. A.T. Massey Coal Co., 679 S.E.2d 223, 234, 239 (W. Va. 2008) (holding that tort claims were encompassed by a forum selection clause concerning “[a]ll actions brought in connection with” a contract), rev’d and remanded, 556 U.S. 868 (2009).

  50. [50]. S-Fer Int’l, Inc. v. Paladion Partners, Ltd., 906 F. Supp. 211, 213 (S.D.N.Y. 1995).

  51. [51]. Bancomer v. Superior Court, 52 Cal. Rptr. 2d 435, 437 (Ct. App. 1996)
    (emphasis added).

  52. [52]. See, e.g., Hansa Consult of N. Am., LLC v. Hansaconsult Ingenieurgesellschaft mbH,
    35 A.3d 587, 593(N.H. 2011) (“The clause at issue here—‘Place of Jurisdiction is only Hamburg’
    —provides little guidance as to the scope of possible non-contract-based disputes the parties intended to litigate in Hamburg.”).

  53. [53]. Verdugo v. Alliantgroup, L.P., 187 Cal. Rptr. 3d 613, 617 (Ct. App. 2015), as modified on denial of reh’g (June 25, 2015) (emphasis added).

  54. [54]. Cheney v. IPD Analytics, LLC, 583 F. Supp. 2d 108, 114 (D.D.C. 2008) (emphasis added); see also Presbyterian Healthcare Servs. v. Goldman, Sachs & Co., 122 F. Supp. 3d 1157, 1166 (D.N.M. 2015) (“[A]ll actions and proceedings arising out of this Broker–Dealer Agreement or any of the transactions contemplated hereby shall be brought in the United States District Court in the County of New York ... .” (emphasis omitted)).

  55. [55]. Graham Tech. Sols., Inc. v. Thinking Pictures, Inc., 949 F. Supp. 1427, 1429 (N.D. Cal. 1997).

  56. [56]. Courts disagree as to whether the phrase “arising out of” should be included in the list of “magic words” that give a clause a broad scope. A few courts have held that this phrase is the functional equivalent of “relating to” or “in connection with” and that it should therefore be given a broad scope. See Roby v. Corp. of Lloyd’s, 996 F.2d 1353, 1361 (2d Cir. 1993) (“We find no substantive difference in the present context between the phrases ‘relating to,’ ‘in connection with’ or ‘arising from.’”); Credit Suisse Sec. (USA) LLC v. Hilliard, 469 F. Supp. 2d 103, 107 (S.D.N.Y. 2007) (“When ‘arising out of,’ ‘relating to,’ or similar words appear in a forum selection clause, such language is regularly construed to encompass ... tort claims associated with the underlying contract.”); cf. Autoridad de Energía Eléctrica v. Vitol S.A., 859 F.3d 140, 143 (1st Cir. 2017) (concluding that a clause giving the courts of Puerto Rico exclusive jurisdiction to hear “controversies ... regarding the terms and conditions of this Contract” covered statutory claims). However, most courts—including several other panels of judges sitting in the Second Circuit—have held that the phrase “arising out of” is narrower than the phrase “relating to.” See Huffington v. T.C. Grp. LLC, 637 F.3d 18, 23 (1st Cir. 2011)(discussing crucial differences between clauses with “embracing language” such as “with respect to,” “with reference to,” “relating to,” “in connection with,” and “associated with” which “have usually beenconstruedbroadly,” and clauses with terms like “to enforce” or “arising out of,” which necessitate a “narrower focus”); Manila Indus., Inc. v. Ondova Ltd. Co., 334 F. App’x 821, 823 (9th Cir. 2009) (Tallman, J., concurring) (arguing that the phrase “arising out of or resulting from” is narrower than the phrase “relates in some way”); Phillips v. Audio Active, Ltd., 494 F.3d 378, 389 (2d Cir. 2007) (“We do not understand the words ‘arise out of’ as encompassing all claims that have some possible relationship with the contract, including claims that may only ‘relate to,’ be ‘associated with,’ or ‘arise in connection with’ the contract.”); Coregis Ins. Co. v. Am. Health Found., Inc., 241 F.3d 123, 129 (2d Cir. 2001) (concluding that the phrases “relating to,” “in connection with,” “associated with,” “with respect to,” and “with reference to” were synonyms and that all connoted a broader scope that the phrase “arising out of”); Brown v. Federated Capital Corp., 991 F. Supp. 2d 857, 862 (S.D. Tex. 2014) (observing that “forum-selection clauses that apply to all disputes that ‘relate to’ or ‘are connected with’ the contractareconstruedbroadly, while clauses that apply to all disputes ‘arising out of’ or ‘the implementation and interpretation of’ the contract areconstruednarrowly”); cf. Cape Flattery Ltd. v. Titan Mar., LLC, 647 F.3d 914, 921–23 (9th Cir. 2011) (concluding that the phrase “arising out of” is narrower than the phrase “arising out of or relating to” in arbitration context).

  57. [57]. In Forrest v. Verizon Commc’ns, Inc., 805 A.2d 1007, 1009 (D.C. 2002), for example, the court was called upon to construe a clause stipulating that the parties “consent to the exclusive personal jurisdiction of and venue in a court of competent jurisdiction located in Fairfax County, Virginia.” The trial court held, and the court of appeals agreed, that this provision was susceptible to two equally plausible interpretations as to its scope: “[T]he clause could reasonably be interpreted as applying only to disputes concerning the Agreement or as applying to all claims between the parties.” Id. at 1014.­­

  58. [58]. See supra note 56 and accompanying text.

  59. [59]. See supra note 56 and accompanying text.

  60. [60]. Hansa Consult of N. Am., LLC v. Hansaconsult Ingenieurgesellschaft mbH, 35 A.3d 587, 593–94 (N.H. 2011) (“[A]pplying a contractual forum selection provision ... outside the scope of the contract would upset the legitimate expectations of the contracting parties and would encompass a potentially unforeseeable range of noncontractual disputes. Limiting the reach of such generic forum selection provisions to disputes involving the rights and obligations that the contract itself creates will give effect to the most likely intent of the parties and help foster stable and predictable commercial relationships.” (citations omitted)).

  61. [61]. Some courts have invoked the canons relating to scope where the clause in question contained the phrase “brought hereunder.” See Stiles v. Bankers Healthcare Grp., Inc., 637 F. App’x 556, 558, 559–61 (11th Cir. 2016); Ex parte Bad Toys Holdings, Inc., 958 So. 2d 852,
    855–60 (Ala. 2006). Other courts have applied them to determine the scope of a clause in which the parties “consent to jurisdiction” in a particular court. Kochert v. Adagen Med. Int’l, Inc., 491 F.3d 674, 677–79 (7th Cir. 2007); Tuxedo Int’l Inc. v. Rosenberg, 251 P.3d 690, 696–700 (Nev. 2011). Most courts have applied these canons to construe some variation of the phrase “arising out of.” See, e.g., Marinechance Shipping, Ltd. v. Sebastian, 143 F.3d 216, 220–23 (5th Cir. 1998).

  62. [62]. See, e.g., Hansa Consult of N. Am., 35 A.3d at 593.

  63. [63]. See Sherby, supra note 6, at 280–93.

  64. [64]. Occasionally, courts have invoked public policy as a rationale for construing generic forum selection clauses so as not to cover non-contractual claims. See, e.g., N. Star Gas Co. v. Pac. Gas & Elec. Co., No. 15-cv-02575-HSG, 2016 WL 5358590, at *16 (N.D. Cal. Sept. 26, 2016); Mgmt. Comput. Controls, Inc. v. Charles Perry Constr., Inc., 743 So. 2d 627, 632 (Fla. Dist. Ct. App. 1999).

  65. [65]. Compare Phillips v. Audio Active Ltd., 494 F.3d 378, 390–91 (2d Cir. 2007) (utilizing an origination approach), with Magi XXI, Inc. v. Stato della Città del Vaticano, 714 F.3d 714, 725 (2d Cir. 2013) (utilizing a contract compliance approach).

  66. [66]. This lack of predictability is frustrating and “runs against the very fabric of the common law because ‘the effectiveness of the law is seriously diminished when legal practitioners believe they cannot confidently advise what the law is.’” Glenn D. West & W. Benton Lewis, Jr., Contracting to Avoid Extra-Contractual Liability—Can Your Contractual Deal Ever Really Be the “Entire” Deal?, 64 Bus. Law. 999, 1035 (2009) (quoting Perre v. Apand Pty Ltd. (1999) 198 C.L.R. 180, 215 (Austl.)).

  67. [67]. Phillips, 494 F.3d at 390–91. But see Cheney v. IPD Analytics, LLC, 583 F. Supp. 2d 108, 122 (D.D.C. 2008) (“Substituting ‘originate’ for ‘arising out of’ is of little use in construing the scope of the forum selection clause because it simply substitutes a synonym.”).

  68. [68]. Phillips, 494 F.3d at 390–­91. The clause in question provided that “any legal proceedings that may arise out of [the agreement] are to be brought in England.” Id. at 382.

  69. [69]. Id. at 390–91.

  70. [70]. Id. (“[Plaintiff] does not rely on the recording contract to establish his ownership of the relevant copyrights, but on his authorship of the work, a status afforded him as the composerwho translates an idea into a fixed, tangible musical expression entitled to copyright protection.”). This approach has been followed by a number of other courts considering intellectual property cases. See Altvater Gessler-J.A. Baczewski Int’l (USA) Inc. v. Sobieski Destylarnia S.A., 572 F.3d 86, 91 (2d Cir. 2009) (applying the same analysis to copyright claims); Corcovado Music Corp. v. Hollis Music, Inc.,981 F.2d 679, 684–85 (2d Cir. 1993) (applying the same analysis to copyright claims); Xiao Wei Yang Catering Linkage in Inner Mong. Co., Ltd. v. Inner Mong. Xiao Wei Yang USA, Inc., 150 F. Supp. 3d 71, 80–81 (D. Mass. 2015) (applying the same analysis to trademark claims).

  71. [71]. Phillips, 494 F.3d at 390–91. The court also noted that the copyright claims did not “arise out of” the contract because the contract was relevant only as a defense in the litigation.
    Id. at 382. This approach has been followed by at least one other federal court of appeals. See Reading Health Sys. v. Bear Stearns & Co., 900 F.3d 87, 100 (3d Cir. 2018) (“Where, as here, the clause encompasses only disputes ‘arising out of’ the contract, courts have rejected the argument that a contractual defense alone is sufficient to bring the dispute within the scope of the clause.”). It has, however, also been rejected by at least two other federal appeals courts. See John Wyeth
    & Brother Ltd. v. Cigna Int’l Corp., 119 F.3d 1070, 1076 (3d Cir. 1997) (concluding that the phrase “arising ... in relation to” encompasses a dispute in which a contract is raised as a defense); Omron Healthcare, Inc. v. Maclaren Exps. Ltd., 28 F.3d 600, 602 (7th Cir. 1994) (reaching same conclusion with respect to the phrase “arises out of”). In 2015, the Texas Court of Appeals adopted an analytical approach that was similar to the one utilized by the Second Circuit in Phillips, but this decision was reversed on appeal. See Sheldon v. Pinto Tech. Ventures, L.P., 477 S.W.3d 411, 420 (Tex. App. 2015), rev’d, 526 S.W.3d 428 (Tex. 2017).

  72. [72]. AMA Multimedia LLC v. Sagan Ltd., No. CV-16-01269-PHX-DGC, 2017 WL 372970,
    at *5 (D. Ariz. Jan. 26, 2017) (declining to follow this approach because it was inconsistent with Ninth Circuit precedent and observing that “[a] claim that may be brought without reference to a contract may nonetheless require interpretation of the contract, particularly if the contract presents a possible defense to the claim.”), vacated and remanded, 720 F. App’x 873 (9th Cir. 2018).

  73. [73]. Hansa Consult of N. Am. v. Hansaconsult Ingenieurgesellschaft mbH, 35 A.3d 587, 595 (N.H. 2011).

  74. [74]. Phillips, 494 F.3d at 393.

  75. [75]. See Pinto Tech. Ventures, L.P. v. Sheldon, 526 S.W.3d 428, 440 (Tex. 2017) (observing that this approach promotes “artful pleading” and noting that “[a] plaintiff could characterize its claim as a statutory or common-law tort claim to evade the agreed-upon forum despite essential allegations that are ‘inextricably enmeshed’ or ‘factually intertwined’ with the underlying contract” (quoting AutoNation USA Corp. v. Leroy, 105 S.W.3d 190, 195 (Tex. App. 2003)));
    see also Ashall Homes Ltd. v. ROK Entm’t Grp. Inc., 992 A.2d 1239, 1252 n.62 (Del. Ch. 2010) (“A forum selection clause should not be defeated by artful pleading of claims not based on the contract containing the clause if those claims grow out of the contractual relationship ... .” (quoting Direct Mail Prod. Servs. Ltd. v. MBNA Corp., No. 99 Civ. 10550(SHS), 2000 WL 1277597, at *6 (S.D.N.Y. Sept. 7, 2000))). See generally Ferens v. John Deere Co., 494 U.S. 516 (1989) (“To permit a situation in which two cases involving precisely the same issues are simultaneously pending in different District Courts leads to ... wastefulness of time, energy, and money.” (citations omitted) (internal quotation marks omitted)).

  76. [76]. See, e.g., Rivera v. Centro Médico de Turabo, Inc., 575 F.3d 10, 24 (1st Cir. 2009); Terra Int’l, Inc. v. Miss. Chem. Corp., 119 F.3d 688, 694 (8th Cir. 1997); Sagittarius Sporting Goods Co. v. LG Sourcing, Inc., 162 F. Supp. 3d 531, 539 (D.S.C. 2016); Lawler v. Schumacher Filters Am., Inc., 832 F. Supp. 1044, 1052 (E.D. Va. 1993); Pinto Tech. Ventures, 526 S.W.3d at 439; Tuxedo Int’l Inc. v. Rosenberg, 251 P.3d 690, 699 (Nev. 2011); Hansa Consult of N. Am., 35 A.3d at 595; Forrest v. Verizon Commc’ns., Inc., 805 A.2d 1007, 1014 (D.C. 2002); see also Stiles v. Bankers Healthcare Grp., Inc., 637 F. App’x 556, 561 (11th Cir. 2016); Bancomer v. Superior Court, 52 Cal. Rptr. 2d 435, 442 (Ct. App. 1996) (concluding that the claims in question were not covered by the clause). In this same vein, some courts have held that non-contractual claims are covered where they “are ‘integrally related’ to the” contract. Paduano v. Express Scripts, Inc., 55 F. Supp. 3d 400, 433 (E.D.N.Y. 2014).

  77. [77]. Terra Int’l, 119 F.3d at 695.

  78. [78]. Hansa Consult of N. Am., 35 A.3d at 595.

  79. [79]. Id. at 590.

  80. [80]. Id.

  81. [81]. Id.

  82. [82]. Id.

  83. [83]. Id. at 592.

  84. [84]. Id.

  85. [85]. Id.

  86. [86]. Id.

  87. [87]. Id. at 595.

  88. [88]. Id.

  89. [89]. Id.

  90. [90]. Id.

  91. [91]. Id. at 597.

  92. [92]. Id. at 596.

  93. [93]. See infra Section III.C.1.

  94. [94]. See infra Section III.C.2.

  95. [95]. See infra Section III.C.3.

  96. [96]. See In re Orange, S.A., 818 F.3d 956, 962 (9th Cir. 2016) (“Nothing in the claims required the district court to interpret, let alone reference, the [contract] to issue a ruling on [plaintiff’s] claims.”); see also Manetti-Farrow, Inc. v. Gucci Am., Inc., 858 F.2d 509, 514 (9th Cir. 1988) (stating that the court “must, therefore, determine if Manetti-Farrow’s claims require interpretation of the contract”); Oracle Am., Inc. v. Or. Health Ins. Exch. Corp., 145 F. Supp. 3d 1018, 1027 (D. Or. 2015) (concluding that “resolution of Oracle’s copyright-infringement claim against DHS/OHA will require the Court to analyze the terms and provisions of the [contract]”).

  97. [97]. Graham Tech. Sols., Inc. v. Thinking Pictures, Inc., 949 F. Supp. 1427, 1432–33 (N.D. Cal. 1997).

  98. [98]. Id. at 1429.

  99. [99]. Id.

  100. [100]. Id.

  101. [101]. Id.

  102. [102]. Id. at 1431–32.

  103. [103]. Id.

  104. [104]. Id. at 1432.

  105. [105]. Id. at 1434 (emphasis added).

  106. [106]. Id. at 1433.

  107. [107]. Omron Healthcare, Inc. v. Maclaren Exps. Ltd., 28 F.3d 600, 603 (7th Cir. 1994). For a case with similar facts that came out differently on the same test, see Payne v. N. Tool & Equip. Co., No. 2:13-CV-109 JD, 2013 WL 6019299, at *18–20 (N.D. Ind. Nov. 12, 2013).

  108. [108]. Omron Healthcare, 28 F.3d at 603.

  109. [109]. Id. at 601.

  110. [110]. Id.

  111. [111]. Id.

  112. [112]. Id.

  113. [113]. Id.

  114. [114]. Id. at 601–02.

  115. [115]. Id. at 602.

  116. [116]. Id.

  117. [117]. Id.; see also Wellogix, Inc. v. SAP Am., Inc., 58 F. Supp. 3d 766, 779 (S.D. Tex. 2014) ([S]ince the NetWeaver Agreement governs the confidentiality of Wellogix’s trade secrets, the resolution of Wellogix’s claims ‘arguably depends’ on construction of the NetWeaver Agreement.” (footnote omitted)).

  118. [118]. The Omron court specifically rejected the notion that the plaintiff’s intellectual property claims were not covered by the forum selection clause merely because these claims originated in a federal statute. Omron Healthcare, 28 F.3d at 602.

  119. [119]. See, e.g., Int’l Software Sys., Inc. v. Amplicon, Inc., 77 F.3d 112, 116 (5th Cir. 1996); Manetti-Farrow, Inc. v. Gucci Am., Inc., 858 F.2d 509, 514 (9th Cir. 1988); Sebascodegan Enters., LLC v. Petland, Inc., 647 F. Supp. 2d 71, 75 (D. Me. 2009); S. Union Co. v. Sw. Gas Corp., 165 F. Supp. 2d 1010, 1030 (D. Ariz. 2001); Tuxedo Int’l Inc. v. Rosenberg, 251 P.3d 690, 695 (Nev. 2011).

  120. [120]. Magi XXI, Inc. v. Stato della Città del Vaticano, 714 F.3d 714, 725 (2d Cir. 2013).

  121. [121]. Id. at 717.

  122. [122]. Id. at 718.

  123. [123]. Id. at 725 (emphasis added).

  124. [124]. Id. at 725.

  125. [125]. Coastal Steel Corp. v. Tilghman Wheelabrator Ltd., 709 F.2d 190, 203 (3d Cir. 1983), overruling recognized, In re McGraw-Hill Glob. Educ. Holdings LLC, 909 F.3d 48 (3d Cir. 2018).

  126. [126]. See, e.g., YWCA v. HMC Entm’t, Inc., No. 91 Civ. 7943 (KMW), 1992 WL 279361, at *4 (S.D.N.Y. Sept. 25, 1992) (“Plaintiff’s entire business relation with defendant with respect to the proposed award program stemmed from the contract between the two parties.”).

  127. [127]. Pinto Tech. Ventures, L.P. v. Sheldon, 526 S.W.3d 428, 438 (Tex. 2017) (discussing this narrower approach and framing it as an issue of “proximate caus[e]”).

  128. [128]. Some courts have also held that a forum selection clause will cover non-contractual claims where the “gist” of those claims is a breach of a contractual relationship. Cfirstclass Corp. v. Silverjet PLC, 560 F. Supp. 2d 324, 329 (S.D.N.Y. 2008).

  129. [129]. Doe v. Seacamp Ass’n, Inc., 276 F. Supp. 2d 222, 228 (D. Mass. 2003).

  130. [130]. Id.; see also Starkey v. G Adventures, Inc., 796 F.3d 193, 196 (2d Cir. 2015) (concluding that sexual assault claim was covered by forum selection clause).

  131. [131]. Seacamp Ass’n, 276 F. Supp. 2d at 228.

  132. [132]. The Texas Supreme Court recently adopted a version of this approach. Rather than focusing on the existence of a contractual relationship between the parties, however, the court instead focused on the existence of specific provisions within the shareholder agreement that were relevant to the suit. Pinto Tech. Ventures, 526 S.W.3d at 442–43. This approach is more narrowly tailored than the approach utilized by the court in Doe. The Texas Supreme Court did not imagine a world in which the parties had never entered into any contractual relationship. Instead, it imagined a world in which certain provisions had never been written into the shareholder agreement. After determining that the dispute would not exist in such a world, it then concluded that the plaintiff’s tort and statutory claims must necessarily be covered by the agreement’s forum selection clause.

  133. [133]. Morgan Trailer Mfg. Co. v. Hydraroll, Ltd., 759 A.2d 926, 932 (Pa. Super. Ct. 2000) (“While Morgan certainly had a contract with appellees, that does not mean that all future relations with appellees are somehow connected to that contract.”).

  134. [134]. Omron Healthcare, Inc. v. Maclaren Exps. Ltd., 28 F.3d 600, 602 (7th Cir. 1994); see also Abbott Labs. v. Takeda Pharm. Co., 476 F.3d 421, 425–26 (7th Cir. 2007) (“Abbott has one good interpretive argument, which is that a ‘but for’ test of ‘arising out of’ or ‘related to’ would be unsound. For then if Abbott and Takeda had a dispute over an unrelated agreement made 20 years later, Takeda could argue that it arose from the 1985 agreement because if it hadn’t been for what the parties learned about each other in operating under the agreement they would not have included in the later agreement the clause giving rise to the dispute. Takeda’slawyer at argument readily conceded that theforum selectionclause should not be stretched that far, that is, to its literal limits.”).

  135. [135]. Serv. Vending Co. v. Wal-Mart Stores, Inc., 93 S.W.3d 764, 768–69 (Mo. Ct. App. 2002); see also Bah. Sales Assoc., LLC v. Byers, 701 F.3d 1335, 1341 (11th Cir. 2012) (“[T]he fact that a dispute could not have arisen but for an agreement does not mean that the dispute necessarily ‘relates to’ that agreement.”).

  136. [136]. Bailey v. ERG Enters., LP, 705 F.3d 1311, 1318 (11th Cir. 2013).

  137. [137]. Id.

  138. [138]. Id.

  139. [139]. Id.

  140. [140]. Bah. Sales Assoc., 701 F.3d at 1341 (“The dispute between Byers and the Mortgage Entities is not ‘a fairly direct result of the performance of contractual duties’ under the lot purchase contract.”).

  141. [141]. Bailey, 705 F.3d at 1323.

  142. [142]. Terra Int’l, Inc. v. Miss. Chem. Corp., 119 F.3d 688, 693 (8th Cir. 1997).

  143. [143]. Id. at 694.

  144. [144]. Id. at 694–95.

  145. [145]. Id. at 697.

  146. [146]. See Stiles v. Bankers Healthcare Grp., Inc., 637 F. App’x 556, 561 (11th Cir. 2016) (concluding that tort claims fell with the scope of the forum selection clause both because they related to the interpretation of the contract and because they arose out of the same set of operative facts as the contract claim); Magi XXI, Inc. v. Stato della Città del Vaticano, 714 F.3d 714, 724 (2d Cir. 2013) (same); Worldwide Network Servs., LLC v. DynCorp Int’l, 496 F. Supp. 2d 59, 63 (D.D.C. 2007) (same); BioCapital, LLC v. BioSystem Sols., Inc., No. FSTCV085009331S, 2009 WL 1815056, at *7 (Conn. Super. Ct. June 1, 2009) (same).

  147. [147]. The origination approach is not incorporated into any of these analyses.

  148. [148]. In Worldwide Network Services, LLC v. DynCorp International, for example, the U.S. District Court for the District of Columbia held that the plaintiff’s non-contractual clauses were covered because (1) they arose out of the same set of operative facts as a parallel contract claim, and
    (2) the adjudication of those claims required the court to interpret the contract. Worldwide Network Servs., 496 F. Supp. 2d at 63.

  149. [149]. Adams v. Unione Mediterranea di Sicurta, 364 F.3d 646, 652 (5th Cir. 2004) (“Under the general principles of contract law, it is axiomatic that courts cannot bind anon-partyto a contract, because that party never agreed to the terms set forth therein.” (quoting EEOC v. Frank’s Nursery & Crafts, Inc., 177 F.3d 448, 460 (6th Cir. 1999))); cf. AT&T Techs., Inc. v. Commc’ns Workers of Am., 475 U.S. 643, 648 (1986) (“[A] party cannot be required to submit to arbitration any dispute which he has not agreed so to submit.” (quoting United Steelworkers v. Warrior & Gulf Navigation Co., 363 U.S. 574, 582 (1960))).

  150. [150]. See infra note 153 (collecting cases).

  151. [151]. See infra notes 179–80.

  152. [152]. Several of these rationales sound in equity. Some courts, for example, have held that the doctrine of “direct benefit estoppel” may be used to bind a non-signatory to a forum selection clause. See, e.g., In re Lloyd’s Register N. Am., Inc., 780 F.3d 283, 291 (5th Cir. 2015); Liles v. Ginn-La W. End, Ltd., 631 F.3d 1242, 1256–57 (11th Cir. 2011); LaRoss Partners, LLC v. Contact 911 Inc., 874 F. Supp. 2d 147, 156 (E.D.N.Y. 2012); see also Cajun Glob. LLC v. Swati Enters., Inc., 283 F. Supp. 3d 1325, 1330 (N.D. Ga. 2017). In such cases, the courts do not purport to be “interpreting” the contract to reach their decision.

  153. [153]. See, e.g., Firefly Equities LLC v. Ultimate Combustion Co., 736 F. Supp. 2d 797, 800 (S.D.N.Y. 2010) (holding that forum selection clause binds company president who signed contract on behalf of the company). The following are non-signatories that have been found to be “closely related” to the contracting party:(1) “parentcorporations to the contracting party;”(2) “spousal guarantors of the contracting party;”(3) “directors of the contracting party;”(4) “corporations controlled by the contracting party;”(5) “agents of the contracting party;”(6) “successor corporations to the contracting party;”and (7) “corporations affiliated with the contracting party.” See Venard v. Jackson Hole Paragliding, LLC, 292 P.3d 165, 172 (Wyo. 2013) (collecting cases); see also Wylie v. Kerzner Int’l Bah. Ltd., 706 F. App’x 577, 579 (11th Cir. 2017) (considering whether non-signatory wife was bound to a forum selection clause signed by her husband).

  154. [154]. Lu v. Dryclean-U.S.A. of Cal., Inc., 14 Cal. Rptr. 2d 906, 907–08 (Ct. App. 1992).

  155. [155]. Id.

  156. [156]. Id.

  157. [157]. Id.

  158. [158]. Id.

  159. [159]. Id.

  160. [160]. Id.

  161. [161]. Bugna v. Fike, 95 Cal. Rptr. 2d 161, 165 (Ct. App. 2000).

  162. [162]. Some courts have adopted a “totality of the circumstances” test in determining whether one party is closely related to another. See Synthes, Inc. v. Emerge Med., Inc., 887 F. Supp. 2d 598, 607 (E.D. Pa. 2012) (“[C]ourts considering this question of whether anon-signatorymay be bound by aforum selectionclause take a common sense, totality of the circumstances approach that essentially inquires into whether, in light of those circumstances, it is fair and reasonable to bind a non-party to theforum selectionclause.” (quoting Regions Bank v. Wyndham Hotel Mgmt., Inc., No. 3:09-1054, 2010 WL 908753, at *6 (M.D. Tenn. Mar. 12, 2010))). A few courts have applied a version of this test which holds that so-called “transaction participants” may enforce a forum selection clause even if they are not signatories to the contract containing that clause. See Pinto Tech. Ventures, L.P. v. Sheldon, 526 S.W.3d 428, 444 (Tex. 2017) (observing “that ‘transaction participants’ may enforce a valid forum-selection clause even if they are not actually signatories to the contract”). In practice, the “closely related” test subsumes the “transaction participants” test because transaction participants are by definition “closely related” to the contract. See id. (noting similarities between the two tests).

  163. [163]. Magi XXI, Inc. v. Stato della Città del Vaticano, 714 F.3d 714, 723 (2d Cir. 2013). The California Court of Appeal has opined that the judicial inquiry in this area should focus on the closeness to the contractual relationship rather than the closeness to the contracting parties. Bugna, 95 Cal. Rptr. 2d at 165.

  164. [164]. Magi XXI, 714 F.3d at 723 (quoting M/S Bremen v. Zapata Off-Shore Co., 407 U.S. 1, 12 (1972)). For a rare case concluding that the “foreseeable” requirement had not been satisfied, see In re Howmedica Osteonics Corp., 867 F.3d 390, 407 n.13 (3d Cir. 2017).

  165. [165]. See generally LaRoss Partners, LLC v. Contact 911 Inc., 874 F. Supp. 2d 147, 160 (E.D.N.Y. 2012) (collecting cases).

  166. [166]. Marano Enters. of Kan. v. Z-Teca Rests., L.P., 254 F.3d 753, 754 (8th Cir. 2001).

  167. [167]. Id. at 754–55.

  168. [168]. Id. at 757–58.

  169. [169]. Id.

  170. [170]. Id.

  171. [171]. Id. (citation omitted); see also Manetti-Farrow, Inc. v. Gucci Am., Inc.,858 F.2d 509, 514 n.5 (9th Cir. 1988) (stating “that the alleged conduct of the non-parties is so closely related to the contractual relationship that the forum selection clause applies to all defendants”).

  172. [172]. Compare Marano Enters. of Kan., 254 F.3d at 757 (holding that a non-signatory plaintiff was “properly bound by the forum-selection provisions”), with Medtronic, Inc. v. Endologix, Inc.,530 F. Supp. 2d 1054, 1056–57 (D. Minn. 2008) (agreeing that a non-signatory defendant “is bound by the forum selection clauses in the employment agreements”).

  173. [173]. Dos Santos v. Bell Helicopter Textron, Inc., 651 F. Supp. 2d 550, 556 (N.D. Tex. 2009).

  174. [174]. BNY AIS Nominees Ltd. v. Quan, 609 F. Supp. 2d 269, 275 (D. Conn. 2009) (“One situation where a non-party may invoke a contractual forum selection clause, or it can be invoked against the non-party, is where the non-party is a third-party beneficiary of the contract.”); Thorworks Indus. v. E.I. DuPont De Nemours & Co., 606 F. Supp. 2d 691, 695–97 (N.D. Ohio 2008) (concluding defendant was not a third-party beneficiary to a contract and could therefore not invoke a forum selection clause); Try Switch, Ltd. v. Endurance Int’l Grp., No. 12-P-1535, 2013 WL 2096605, at *1 (Mass. App. Ct. May 16, 2013) (concluding Massachusetts defendant was not a third-party beneficiary and was hence ineligible to invoke forum selection clause).

  175. [175]. Dos Santos, 651 F. Supp. 2d at 556.

  176. [176]. See, e.g., Hugel v. Corp. of Lloyd’s, 999 F.2d 206, 209–10 n.7 (7th Cir. 1993). Some courts have used the fact that a non-signatory was a third-party beneficiary to the agreement to justify a conclusion that that party was “closely related” to the transaction. See, e.g., Freeford Ltd. v. Pendleton, 857 N.Y.S.2d 62, 66–67 (App. Div. 2008); Solargenix Energy, LLC v. Acciona, S.A., 17 N.E.3d 171, 183 (Ill. App. Ct. 2014). See generally George S. Geis, Broadcast Contracting, 106 Nw. U. L. Rev. 1153, 1193 (2012) (arguing that “the default rule for third-party rights should be off, not maybe”).

  177. [177]. Carlyle Inv. Mgmt. LLC v. Moonmouth Co. SA, 779 F.3d 214, 218 (3d Cir. 2015) (observing that, under Delaware law, anon-signatoryto an agreement is bound by itsforum selectionclause when “thenon-signatory [is]athird-party beneficiaryof the agreement or closely related to the agreement”).

  178. [178]. Hadley v. Shaffer, No. Civ.A. 99-144-JJF, 2003 WL 21960406, at *6 (D. Del. Aug. 12, 2003) (“Even if the Hadleys are not third-party beneficiaries to the Merger Agreement, the Court concludes that they are closely related to the Merger Agreement, and therefore, reasonably bound by its forum selection clause.”); Sealord Marine Co. v. Am. Bureau of Shipping, 220 F. Supp. 2d 260, 270 (S.D.N.Y. 2002) (observing “that third party beneficiarystatus is not required in order for a non-party to invoke aforum-selection provision”); cf. In re McGraw-Hill Glob. Educ. Holdings LLC, 909 F.3d 48, 59 (3d Cir. 2018) (concluding that plaintiffs were not bound by forum selection clause under either third-party beneficiary doctrine or the “closely related” test); Coastal Steel Corp. v. Tilghman Wheelabrator Ltd., 709 F.2d 190, 202–04 (3d Cir. 1983) (holding that third-party-beneficiary status was not a ground for disregarding an otherwise valid forum selection clause).

  179. [179]. Harland Clarke Holdings Corp. v. Milken, 997 F. Supp. 2d 561, 581 (W.D. Tex. 2014).

  180. [180]. Magi XXI, Inc. v. Stato della Città del Vaticano, 714 F.3d 714, 723 (2d Cir. 2013) (quoting M/S Bremen v. Zapata Off-Shore Co., 407 U.S. 1, 12 (1972)).

  181. [181]. Pinto Tech. Ventures, L.P. v. Sheldon, 526 S.W.3d 438, 445 (2017) (alteration in original) (emphasis omitted).

  182. [182]. Id.; see also Black v. Diamond Offshore Drilling, Inc., 551 S.W.3d 346, 352–53 (Tex. App. 2018) (noting contractual limits in holding that non-signatories were not covered by forum selection clause).

  183. [183]. Casville Invs., Ltd. v. Kates, No. 12 Civ. 6968(RA), 2013 WL 3465816, at *6 (S.D.N.Y. July 8, 2013).

  184. [184]. Id.; see also Bensinger v. Denbury Res. Inc., No. 10-CV-1917 (JG), 2011 WL 3648277,
    at *6 (E.D.N.Y. Aug. 17, 2011) (construing similar clause to reach same result); APA Excelsior III L.P. v. Premiere Techs., Inc., 49 F. Supp. 2d 664, 672 (S.D.N.Y. 1999) (same).

  185. [185]. See Team Y&R Holdings H.K. Ltd. v. Ghossoub [2017] EWHC, 2401, para. 82 (Comm) (Rabinowitz, J.) (“[W]here contracting parties intend that any claim relating to the contract be subject to the exclusive jurisdiction clause even where it is one brought by or against a non-contracting party, clear words should be used expressly setting out this intention, the parties to be affected and, if relevant, the manner in which submission of any non-contracting parties to the jurisdiction of the chosen court is to be ensured.”).

  186. [186]. See Glenn West, On Naval Ramming Bows and Contractual Boilerplate—Are Standard “No Third-Party Beneficiary” Clauses Always a Good Thing?, Global Priv. Equity Watch (June 19, 2017), https://privateequity.weil.com/insights/naval-ramming-bows-contractual-boilerplate-standard-no-third-party-beneficiary-clauses-always-good-thing (urging parties to “[c]arefully determine which provisions of your contract are actually intended to benefit and be enforceable by a nonparty
    ... and carve those provisions out from the standard ‘no third-party beneficiary’ provision, else you may find yourself trying to explain how the ‘no third-party beneficiary’ provision does not trump the otherwise clear intent to benefit nonparty affiliates”).

  187. [187]. See generally David S. Coale et. al., Contractual Waiver of the Right to Remove to Federal Court: How Policy Judgments Guide Contract Interpretation, 29 Rev. Litig. 327 (2010) (providing detailed analysis of the case law in this area).

  188. [188]. 1A James W.M. Moore & Brett A. Ringle,Moore’s Federal Practicepara. 0.157[9], at 152 (2d ed. 1996) (“Waiver of the right to remove is ... possible but defendant’s intent must be clear and unequivocal.”); see also New Jersey v. Merrill Lynch & Co., 640 F.3d 545, 548 (3d Cir. 2011) (looking to “plain and ordinary meaning” of the clause); Snapper, Inc. v. Redan, 171 F.3d 1249, 1261 (11th Cir. 1999) (observing that “in the context of removal based solely on diversity jurisdiction, ordinary contract principles govern a contractual waiver” (footnote omitted)); Regis Assocs. v. Rank Hotels (Mgmt.) Ltd., 894 F.2d 193, 195 (6th Cir. 1990) (adopting clear and unequivocal standard). In some cases, a party’s ability to remove a case to federal court may be limited by other language in the forum selection clause. In iNet Directories, LLC v. Developershed, Inc., for example, the Eighth Circuit held that a clause stating that the parties “waive any and all objections ... to the laying of venue of any ... suit, action or proceeding brought in any ... federalor state court in the State of Missouri” constituted a clear and unequivocal waiver of the right to remove a case to federal court notwithstanding the fact that the clause specifically referred to “federal” courts. iNet Directories, LLC v. Developershed, Inc., 394 F.3d 1081, 1081–82 (8th Cir. 2005); accord Waters v. Browning-Ferris Indus., Inc., 252 F.3d 796, 797–98 (5th Cir. 2001) (“A reading of this provision leads this court to the inescapable conclusion that the plaintiff negotiated with the defendant a clear right to establish ‘irrevocably’ the place where his suit could be filed and heard.”).

  189. [189]. See infra note 190.

  190. [190]. Several courts have addressed clauses that clearly contemplated the choice of a state court to the exclusion of a federal forum. See, e.g., Push Pedal Pull, Inc. v. Casperson, 971 F. Supp. 2d 918, 928 (D.S.D. 2013) (“[T]he state court situated in Minnehaha County, South Dakota, shall be the exclusive jurisdiction of any dispute relating to this Agreement.” (emphasis added)); Cty. of Jefferson v. Tyler Techs., Inc., No. 4:09CV554SNLJ, 2009 WL 1811804, at *2 (E.D. Mo. June 25, 2009) (“Contractor further consents and agrees venue for State court proceedings shall be in the County of Jefferson, Missouri, and no court actions commenced in Missouri shall be transferred or removed to any other State or Federal court.” (emphasis added)); see also Emerald Grande, Inc. v. Junkin, 334 F. App’x 973, 975 (11th Cir. 2009); Gamma Constr. Co. v. Werhan Folkers & Monihan, Inc., 11 F. App’x 814, 814 (9th Cir. 2001); Roberts & Schaefer Co. v. Merit Contracting, Inc., 99 F.3d 248, 254 (7th Cir. 1996); OHM Hotel Grp., LLC v. Dewberry Consultants, LLC, No. 4:15-CV-1705 CAS, 2016 WL 427959, at *2 (E.D. Mo. Feb. 4, 2016). For clauses that clearly contemplated the choice of either a state or federal forum, see Bonanno v. VTB Holdings, Inc., No. 10681-VCN, 2016 WL 614412, at *2 (Del. Ch. Feb. 8, 2016) (“Each of the parties hereto irrevocably agrees that all claims ... shall be heard and determined in ... a New York State or federal court, and that such jurisdiction of such courts with respect thereto shall be exclusive.” (emphasis added)); and Ex parte B2K Systems, LLC, 162 So. 3d 896, 901–02 (Ala. 2014) (“Any dispute shall be brought in the appropriate state or federal court in Kent County, Michigan.” (second emphasis added)).

  191. [191]. See Sherby, supra note 6, at 267 (“For counsel representing a non-American party to a contract that selects an American forum, the lesson of these cases is clear—in an exclusive [forum selection clause], never use the phrase ‘the courts of a state.’”).

  192. [192]. Dixon v. TSE Int’l Inc., 330 F.3d 396, 397–98 (5th Cir. 2003).

  193. [193]. Id. at 397 (emphasis added).

  194. [194]. Id. at 398 (quoting of, Black’s Law Dictionary 1232 (4th ed. 1968)).

  195. [195]. See New Jersey v. Merrill Lynch & Co., 640 F.3d 545, 548–49 (3d Cir. 2011); FindWhere Holdings, Inc. v. Sys. Env’t Optimization, LLC, 626 F.3d 752, 755 (4th Cir. 2010); Doe 1 v. AOL LLC, 552 F.3d 1077, 1081–82 (9th Cir. 2009); Am. Soda, LLP v. U.S. Filter Wastewater Grp., Inc., 428 F.3d 921, 926 (10th Cir. 2005); LFC Lessors, Inc. v. Pac. Sewer Maint. Corp., 739 F.2d 4, 7
    (1st Cir. 1984); Setzer v. Natixis Real Estate Capital, Inc., 537 F. Supp. 2d 876, 879 (E.D. Ky. 2008); see also supra note 188 (stipulating that waivers of the right to remove must be “clear and unequivocal”). But see Stateline Power Corp. v. Kremer, 148 F. App’x 770, 771 (11th Cir. 2005) (concluding that the phrase “the courts of the State of Florida” was ambiguous and declining to assign the word “of” a constructive meaning).

  196. [196]. As discussed below, when the chosen venue is a county—rather than a state—the courts follow a somewhat different line of reasoning. In these cases, the courts will generally look to whether there is a federal courthouse located “in” the chosen county even where the clause in question refers to the courts “of” that county. See Coale et al., supra note 187, at 348–49; infra notes 199–202 and accompanying text.

  197. [197]. Simonoff v. Expedia, Inc., 643 F.3d 1202, 1205–06 (9th Cir. 2011) (citations omitted).

  198. [198]. See, e.g., Asphalt Paving Sys., Inc. v. Gen. Combustion Corp., No. 13-7318, 2015 WL 167378, at *5 n.5 (D.N.J. Jan. 13, 2015) (collecting cases).

  199. [199]. All. Health Grp., LLC v. Bridging Health Options, LLC, 553 F.3d 397, 399 (5th Cir. 2008).

  200. [200]. Id.; see also Bartels ex rel. Bartels v. Saber Healthcare Grp., LLC, 880 F.3d 668, 676 (4th Cir. 2018) (“Because there is no federal court in Franklin County, the plain language of the forum-selection clause precludes removal.”); City of W. Palm Beach v. VisionAIR, Inc., 199 F. App’x 768, 770 (11th Cir. 2006).

  201. [201]. Cornett v. Carrithers, 465 F. App’x 841, 843 (11th Cir. 2012); Progressive Publ’ns, Inc. v. Capitol Color Mail, Inc., 500 F. Supp. 2d 1004, 1005–06 (N.D. Ill. 2007).

  202. [202]. Yakin v. Tyler Hill Corp., 566 F.3d 72, 76 (2d Cir. 2009) (“Had there been a federal court located in Nassau County at the time of this litigation, remand would have been improper. But there was none.”); see also Coale et al., supra note 187, at 349 (“The presence of a federal courthouse helps preserve the right to remove, while the absence of one can support a finding of waiver.”).

  203. [203]. Top of FormMilk ‘N’ More, Inc. v. Beavert, 963 F.2d 1342, 1345–46 (10th Cir. 1992).

  204. [204]. Excell, Inc. v. Sterling Boiler & Mech., Inc., 106 F.3d 318, 321 (10th Cir. 1997).

  205. [205]. Id.

  206. [206]. Id.

  207. [207]. For federal district court decisions following this rule, see Ondova Ltd. Co. v. Manila Indus., Inc., 513 F. Supp. 2d 762, 769 (N.D. Tex. 2007); and Progressive Publ’ns, Inc. v. Capitol Color Mail, Inc., 500 F. Supp. 2d 1004, 1005–06 (N.D. Ill. 2007). For decisions by other federal courts of appeal rejecting it, see Simonoff v. Expedia, Inc., 643 F.3d 1202, 1206 (9th Cir. 2011); and Glob. Satellite Comm’n Co. v. Starmill U.K. Ltd., 378 F.3d 1269, 1271 (11th Cir. 2004). But see Bartels v. Saber Healthcare Group, LLC, 880 F.3d 668, 684 (4th Cir. 2018) (Floyd, J., concurring) (criticizing the majority for not adopting this rule).

  208. [208]. All. Health Grp., LLC v. Bridging Health Options, LLC, 553 F.3d 397, 400–01 (5th Cir. 2008).

  209. [209]. Id. at 401.

  210. [210]. Id. (citations omitted).

  211. [211]. See Mozingo v. Trend Pers. Servs., 504 F. App’x 753, 758 n.2 (10th Cir. 2012); see also Irsik & Doll Feed Servs., Inc. v. Roberts Enters. Invs. Inc., No. 6:16-1018-EFM-GEB, 2016 WL 3405175, at *26 (D. Kan. June 21, 2016) (“[T]he language ‘venue ... shall be in Gray County, Kansas,’ specifies state court, not federal court... . The Court finds this language unambiguously refers to the Gray County District Court of Kansas.”).

  212. [212]. Cf. Mark Kelman, Intuitions, 65 Stan. L. Rev. 1291, 1293 (2013) (“[W]hile armchair philosophers might believe they can intuit what commonplace intuitions really are––mistaking introspection for population sampling––they might well be wrong: experimental empiricists could therefore advance philosophical inquiry by helping philosophers and philosophically inclined legal academics both learn how widespread an intuition about a particular legally relevant philosophical proposition might be and explore how commonplace intuitions could best be specified.”).

  213. [213]. Lawless v. Steward Health Care Sys., LLC, 894 F.3d 9, 19–20 (1st Cir. 2018) (concluding that the court could exercise supplemental jurisdiction over the plaintiff’s state law claim because they arose out of “thesame nucleus of operative facts as her” federal claim); D’Onofrio v. Vacation Publ’ns, Inc., 888 F.3d 197, 206–07 (5th Cir. 2018) (same).

  214. [214]. See Lea Brilmayer et al., Conflict of Laws 425 (7th ed. 2015) (collecting cases addressing whether a claim “arises from” local contacts in disputes relating to personal jurisdiction).

  215. [215]. See Taylor v. State Farm Mut. Auto. Ins. Co., 854 P.2d 1134, 1139 (Ariz. 1993) (en banc) (“The judge ... must avoid the often irresistible temptation to automatically interpret contract language as he or she would understand the words.”).

  216. [216]. In practice, the survey relied on snowball sampling to identify respondents. See Michele DeStefano, Creating a Culture of Compliance: Why Departmentalization May Not Be the Answer, 10 Hastings Bus. L.J. 71, 171 n.454 (2014) (“Snowball sampling essentially means that initial participants provide connections to other people who meet the study criteria and might be willing to be interviewed by the researcher.”); see also Leo A. Goodman, Snowball Sampling, 32 Annals Mathematical Stat. 148, 148–49 (1961) (defining snowball sampling); Charles Kadushin, Power, Influence and Social Circles: A New Methodology for Studying Opinion Makers, 33 Am. Soc. Rev. 685, 694–96 (1968) (discussing the strengths and weaknesses of snowball sampling). In the past, a number of researchers have relied on a snowball sample approach to gain insight into the work performed by individuals at large law firms. See, e.g., Elizabeth Chambliss & David B. Wilkins, The Emerging Role of Ethics Advisors, General Counsel, and Other Compliance Specialists in Large Law Firms,44 Ariz. L. Rev. 559, 563–66 (2002) (using snowball sampling to explain the role of compliance specialists); Kimberly Kirkland, Ethics in Large Law Firms: The Principle of Pragmatism,35 U. Mem. L. Rev. 631, 632–34 (2005) (relying on in-person interviews to gain insights into lawyers’ workplace and ethical consciousness, with a specific focus on law firms).

  217. [217]. This number was calculated by subtracting the year in which the respondents first obtained their JDs from 2017. I was unable to ascertain the year when six of the respondents obtained their JDs. This average is therefore based on information from only 72 respondents.

  218. [218]. Scholars working in this area have long bemoaned the difficulty in getting attorneys to sit for interviews or fill out questionnaires. See Stewart Macaulay, Contracts, New Legal Realism, and Improving the Navigation of The Yellow Submarine, 80 Tul. L. Rev. 1161, 1184 (2006) (“We seldom will be able to send important people a questionnaire and have any real expectation that they will take the time to fill it out. Busy professionals may not have time to talk with us or to fill out our questionnaire.”). It is worth noting that one of the seminal works in modern contract scholarship—a survey of Wisconsin businessmen and lawyers in the early 1960s—drew upon on interviews conducted with only 68 people. See Stewart Macaulay, Non-Contractual Relations in Business: A Preliminary Study, 28 Am. Soc. Rev. 55, 55 (1963). Other important studies exploring contracting practice in the real world were similarly based on relatively small numbers of respondents. See, e.g., Daniel Keating, Exploring the Battle of the Forms in Action,98 Mich. L. Rev. 2678, 2681 (2000)(recounting impressions derived from interviews with 25 company representatives); Russell J. Weintraub, A Survey of Contract Practice and Policy,1992 Wis. L. Rev. 1, 1n.1 (discussing results of a survey to which 83 in-house attorneys provided reliable responses).

  219. [219]. See Am. Jur., Expert and Opinion Evidence § 354 (“Generally, unless the words or phrases as used in a writing which is the subject of controversy are terms of art, science, or trade, or there is something to show that they were not used in the ordinary and plain meaning, expert testimony to interpret the language is not admissible.”); see also Westfield Ins. v. Sheehan Constr. Co., 564 F.3d 817, 819 n.† (7th Cir. 2009) (“[T]he interpretation of contracts is for the judge.”).

  220. [220]. See Judge Harvey Brown, Eight Gates for Expert Witnesses, 36 Hous. L. Rev. 743, 795 (1999).

  221. [221]. Rafael Rodrìguez Barril, Inc. v. Conbraco Indus., 619 F.3d 90, 92 (1st Cir. P.R. 2010) (“In the event that either party brings suit to enforce the terms of this [a]greement both [parties] consent and agree that jurisdiction for such action will lie only in the state and federal courts sitting in Mecklenburg County, North Carolina.”).

  222. [222]. Mueller v. Sample, 93 P.3d 769, 772 (N.M. Ct. App. 2004) (“Any suit, action or proceeding arising out of or relating to this Agreement may be commenced and maintained in any court of competent subject matter jurisdiction in Miami-Dade County, Florida and each party waives objection to such jurisdiction and venue.”).

  223. [223]. Animal Film, LLC v. D.E.J. Prods., Inc., 123 Cal. Rptr. 3d. 72, 75 (Ct. App. 2011) (“[T]he parties hereto submit and consent to the jurisdiction of the courts present in the state of Texas in any action brought to enforce (or otherwise relating to) this agreement.”).

  224. [224]. Mueller, 93 P.3d at 772 (“The parties specifically agree and consent that any causes of action or suits related to this Agreement must be filed in the Second Judicial District Court, Albuquerque, New Mexico, USA.”).

  225. [225]. Eighteen respondents specifically mentioned certainty of forum.

  226. [226]. Nine respondents specifically mentioned predictability of case outcomes.

  227. [227]. Eight respondents specifically mentioned a desire to match forum to law.

  228. [228]. E-mail from Kan. Firm Attorney, to Author (July 21, 2017, 2:01 PM) (on file with Author); see also E-mail from N.C. Firm Attorney, to Author (July 28, 2017, 11:08 AM) (on file with Author) (“Exclusive if on my own home turf, non-exclusive if not on my home turf.”); E-mail from Minn. Firm Attorney, to Author (July 21, 2017, 8:10 AM) (on file with Author) (“I generally would prefer exclusive if this is limited to venues convenient to my client... . If not convenient for my client, I would want it permissive.”); E-mail from N.Y. Firm Attorney, to Author (July 8, 2017, 1:31 PM) (on file with Author) (“If the venue is somewhere we want, ideally it’s exclusive. If it’s neutral, we don’t care; if it’s bad, we’d like [non-]exclusive.”).

  229. [229]. Sixteen respondents specifically mentioned convenience. Eleven cited a home court advantage. Nine cited cost savings.

  230. [230]. Four respondents specifically referenced this preference.

  231. [231]. E-mail from N.C. Firm Attorney, to Author (Aug. 2, 2017, 12:53 PM) (on file with Author) (“It depends on whether I’m representing the party who I anticipate is more likely to bring a claim ... .”); E-mail from Tex. Firm Attorney, to Author (July 5, 2017, 11:51 AM) (on file with Author) (“If possible, I generally want it to be non-exclusive for suits brought by my client and exclusive for suits brought against my client.”); E-mail from N.C. Firm Attorney, to Author (July 28, 2017, 11:08 AM) (on file with Author) (“If I am more likely to be sued ... I’d prefer to have it be exclusive.”). But see E-mail from Ga. Firm Attorney, to Author (Aug. 9, 2017, 7:24 PM) (on file with Author) (“If we are representing the party more likely to bring a suit, we will typically prefer that the forum selection clause be exclusive.”).

  232. [232]. Seven respondents specifically mentioned the advantages of litigating in New York
    or Delaware.

  233. [233]. Four respondents specifically mentioned co-exclusive clauses.

  234. [234]. E-mail from N.Y. Firm Attorney, to Author (Aug. 23, 2017, 6:17 AM) (on file with Author).

  235. [235]. E-mail from N.C. Firm Attorney, to Author (July 10, 2017, 12:06 PM) (on file with Author); see also E-mail from Tex. Firm Attorney, to Author (Jul. 28, 2017, 11:03 AM) (on file with Author) (“I prefer to have the other parties submit to the preferred forum for my client, to preempt an argument that the forum is not convenient, but not to bind my client to a particular forum.”).

  236. [236]. See Republic Int’l Corp. v. Amco Eng’rs, Inc., 516 F.2d 161, 168 (9th Cir. 1975).

  237. [237]. Interview with In-House Attorney, in N.C. (May 18, 2017).

  238. [238]. See supra note 35 and accompanying text.

  239. [239]. If the United States were to formally accede to the Hague Convention on Choice of Court Agreements—a treaty that it has signed but not ratified—then a different set of canons relating to exclusivity would apply to international contracts that fall within the scope of the Convention. Under current U.S. law and practice, forum selection clauses are generally presumed to be non-exclusive unless the clause contains the requisite language of exclusivity. See supra Part II. Under the Hague Convention, by contrast, forum selection clauses are generally presumed to be exclusive. See Hague Convention on Choice of Court Agreements art. 3(b), June 30, 2005, 44 I.L.M. 1294 (stating that “a choice of court agreement which designates the courts of one Contracting State ... shall be deemed to be exclusive unless the parties have expressly provided otherwise”). If the Hague Convention were to enter into force, therefore, one set of interpretive rules would be applied to construe forum selection clauses in international commercial contracts and a different set of interpretive rules would be applied to construe these same clauses in domestic contracts.

  240. [240]. When the defendant is alleged to have made false or misleading statements about the plaintiff to third parties, some courts have held that a generic forum selection clause does not cover claims relating to these third-party communications. In one case, the defendant was alleged to have contacted large numbers of the plaintiff’s clients to (1) encourage them not to renew their contracts with the plaintiff, and (2) tell them that the plaintiff was engaged in fraudulent business practices. Berrett v. Life Ins. Co. of the Sw., 623 F. Supp. 946, 949 (D. Utah 1985). On these facts, the court concluded that it was “highly unlikely that in entering into the
    ... agreement the parties contemplated that tort claims such as are here alleged would be governed by the forum selection clause” and held that the claims were not covered. Id. While the courts do not invariably hold that claims relating to the defendant’s representations to third parties fall outside of generic forum selection clauses, there are a number of cases in which they reached precisely that conclusion when presented with claims for slander, defamation, or tortious interference. See, e.g., Ingenieria Alimentaria Del Matatipac, S.A. de C.V. v. Ocean Garden Prods., Inc., 320 F. App’x 548, 549–50 (9th Cir. 2009); Duffield v. MPC Pipelines, Inc., No. 17-cv00115-MSK-STV, 2017 WL 7311884, at *13–15 (D. Colo. Apr. 28, 2017); Serv. Vending Co. v. Wal-Mart Stores, Inc., 93 S.W.3d 764, 768–69 (Mo. Ct. App. 2002); Leatherwood v. Cardservice Int’l, Inc., 885 So. 2d 997, 998–99 (Fla. Dist. Ct. App. 2004); Hansa Consult of N. Am., L.L.C. v. Hansaconsult Ingenieurgesellschaft,mbH, 35 A.3d 587, 593 (N.H. 2011).

  241. [241]. A number of courts have held that tort and statutory claims alleging bad acts that predate the signing of a contract containing a forum selection clause are not covered by the clause. See, e.g., Imation Corp. v. Quantum Corp., No. CIV. 01-1798(RHK/JMM), 2002 WL 385550, at *4 (D. Minn. Mar. 8, 2002); Armco Inc. v. N. Atl. Ins. Co., 68 F. Supp. 2d 330, 338 (S.D.N.Y. 1999); Anselmo v. Univision Station Grp., Inc., No. 92 Civ. 1471(RLC), 1993 WL 17173, at *2–3 (S.D.N.Y. Jan. 15, 1993); Gen. Envtl. Sci. Corp. v. Horsfall, 753 F. Supp. 664, 667–68 (N.D. Ohio 1990); Hansa Consult, 35 A.3d at 593–94. These decisions notwithstanding, most courts have held that fraudulent inducement claims are covered by forum selection clauses unless the plaintiff can prove that “the inclusion of that clause in the contract was the product of fraud or coercion.” Scherk v. Alberto-Culver Co., 417 U.S. 506, 519 n.14 (1974) (emphasis omitted); see also Am. Patriot Ins. Agency, Inc. v. Mut. Risk Mgmt., Ltd., 364 F.3d 884, 889 (7th Cir. 2004) (“[A] dispute over a contract does not cease to be such merely because instead of charging breach of contract the plaintiff charges a fraudulent breach, or fraudulent inducement, or fraudulent performance.”); Marano Enters. of Kan. v. Z-Teca Rests., L.P., 254 F.3d 753, 757 (8th Cir. 2001) (“The general allegation by Marano that it was induced by fraud to enter into the franchise and development agreements is insufficient to raise an issue that the forum-selection clauses within those agreements may be unenforceable because of fraud, and so Marano’s argument must fail.”); Moses v. Bus. Card Exp., Inc.,929 F.2d 1131, 1138 (6th Cir. 1991) (“[U]nless there is a showing that the alleged fraud or misrepresentation induced the party opposing a forum selection clause to agree to inclusion of that clause in the contract, a general claim of fraud or misrepresentation as to the entire contract does not affect the validity of the forum selection clause.”). But see Beckley v. Auto Profit Masters, L.L.C., 266 F. Supp. 2d 1001, 1004–06 (S.D. Iowa 2003) (concluding that fraudulent inducement claim was not covered by clause); Johnson v. Key Equip. Fin., 627 S.E.2d 740, 742 (S.C. 2006) (“Generally, when wrongs arise inducing a party to execute a contract and not directly from the breach of that contract, the remedies and limitations specified by the contract do not apply.”).

  242. [242]. See supra note 61 (collecting cases).

  243. [243]. Courts in the First Circuit, for example, have historically utilized the same-operative-facts test to the exclusion of the others. Similarly, courts in the Ninth Circuit have typically relied on the contract-analysis test to the exclusion of the others.

  244. [244]. Terra Int’l, Inc. v. Miss. Chem. Corp., 119 F.3d 688, 693 (8th Cir. 1997).

  245. [245]. E-mail from Kan. In-House Attorney, to Author (July 21, 2017, 2:01 PM) (on file with Author) (“No, parties should be expressly stated inthe contract.”); E-mail from N.Y. Firm Attorney, to Author (July 21, 2017, 7:57 AM) (on file with Author) (“I would think not, I would think only the parties would be bound ... .”); E-mail from Okla. Firm Attorney, to Author (July 12, 2017, 10:17 AM) (on file with Author) (“No[,] the contract terms would only bind parties to the contract.”); E-mail from N.C. In-House Attorney, to Author (July 21, 2017, 8:28 AM) (on file with Author) (“Generally[,] I would say no since the person is not a party to the contract and most contracts have a no third party beneficiary clause.”).

  246. [246]. E-mail from N.C. Firm Attorney, to Author (July 7, 2017, 8:56 AM) (on file with Author) (“No, unless explicitly made a third party beneficiary.”); E-mail from N.C. Firm Attorney, to Author (Aug. 2, 2017, 12:56 PM) (on file with Author) (“Unless there is a third party beneficiary/equivalent party (i.e., officer/director action is the [company]’s action), I would want a party that would be bound by the clause to sign the [agreement].”).

  247. [247]. E-mail from S.C. Firm Attorney, to Author (July 12, 2017, 3:09 PM) (on file with Author) (“Otherwise you invite mischief by plaintiffs.”).

  248. [248]. E-mail from S.C. Firm Attorney, to Author (July 18, 2017, 4:33 PM) (on file with Author) (“Yes. In my view, the point of a forum-selection clause is to provide predictability to the transaction, and it seems consistent with this goal for an agent/employee/officer of a contracting party to be able to enforce a forum-selection clause.”); E-mail from Tex. Firm Attorney, to Author (Aug. 4, 2017, 2:55 PM) (on file with Author) (“I would want all matters in that forum so ‘should’ apply to non-parties ... .”).

  249. [249]. E-mail from D.C. Firm Attorney, to Author (Aug. 7, 2017, 8:58 AM) (on file with Author).

  250. [250]. E-mail from N.C. Firm Attorney, to Author (July 7, 2017, 12:32 PM) (on file with Author).

  251. [251]. E-mail from N.C. Firm Attorney, to Author (July 28, 2017, 1:38 PM) (on file with Author); see also E-mail from Tex. Firm Attorney, to Author (July 7, 2017, 4:01 PM) (on file with Author) (“The forum selection clause should apply to closely related parties to a reasonable extent (i.e., the clause should apply to any party acting under the terms of the agreement, even if they did not sign the agreement). With that said, the clause is not all encompassing.”); E-mail from Tex. Firm Attorney, to Author (July 5, 2017, 3:17 PM) (on file with Author) (observing that “‘closely related’ and ‘foreseeability’ is quite ambiguous and so I would not base on that”).

  252. [252]. E-mail from Ill. Firm Attorney, to Author (July 17, 2017, 10:52 AM) (on file with Author).

  253. [253]. See supra Part IV.

  254. [254]. E-mail from Tex. Firm Attorney, to Author (July 10, 2017, 3:59 PM) (on file with Author).

  255. [255]. E-mail from N.C. Firm Attorney, to Author (Aug. 4, 2017, 1:36 PM) (on file with Author) (stating that no-third-party-beneficiary clauses are standard but that they “often except third parties who are indemnitees”).

  256. [256]. This answer was given by 62 respondents.

  257. [257]. E-mail from N.C. Attorney, to Author (July 13, 2017, 5:57 AM) (on file with Author).

  258. [258]. E-mail from Minn. Firm Attorney, to Author (July 21, 2017, 8:10 AM) (on file with Author).

  259. [259]. E-mail from N.C. Firm Attorney, to Author (Aug. 2, 2017, 12:56 PM) (on file with Author).

  260. [260]. E-mail from Tex. Firm Attorney, to Author (July 5, 2017, 3:17 PM) (on file with Author).

  261. [261]. E-mail from N.Y. Firm Attorney, to Author (July 31, 2017, 12:37 PM) (on file with Author) (“My preference is always the DE state courts, with [Chancery] being the preferred state court.”); E-mail from Tex. Firm Attorney, to Author (July 18, 2017, 5:37 PM) (on file with Author) (“Depends, but usually say state court ... if we’re talking about Delaware.”).

  262. [262]. E-mail from Tex. Firm Attorney, to Author (July 7, 2017, 4:01 PM) (on file with Author).

  263. [263]. E-mail from N.C. Firm Attorney, to Author (July 31, 2017, 9:52 AM) (on file with Author).

  264. [264]. Several respondents, for example, noted that the second clause referenced jurisdiction and venue whereas the first clause only made reference to jurisdiction. Other respondents noted that the first clause was more strongly worded than the second because it contained the word “shall” rather than the word “consent.” Still others noted that the second clause—by utilizing the word “You”—appeared to bind one party but not the other.

  265. [265]. See supra note 188 and accompanying text (collecting cases adopting “clear and unequivocal” test).

  266. [266]. As discussed above, seven respondents stated that the phrase “courts of Texas” permitted litigation in federal courts. Fourteen respondents concluded that this same phrase precluded litigation in federal courts.

*

Reef C. Ivey II Term Professor of Law, University of North Carolina at Chapel Hill.


I would like to thank Hayward Armstrong, Bill Dodge, Mitu Gulati, Andy Hessick, Mike Hoffheimer, Dave Hoffman, Ralf Michaels, Giesela Rühl, Steve Sachs, Aaron Simowitz, Kathleen Thomas, Mark Weidemaier, Mark Weisburd, Glenn West, and workshop participants at the University of North Carolina School of Law, Wake Forest University School of Law, and Duke Law School for comments on an earlier draft of this Article.