9.01  In the United States, every reported case grappling with the validity of an arbitration clause in a donative trust¹ shares the assumption that state law governs.² The same is true of the vast majority of scholarship on the topic.³ Conversely, this chapter argues (p. 204) that the Federal Arbitration Act (FAA)⁴ applies to some trust arbitration provisions. The discussion then explores the consequences that flow from this conclusion.⁵

9.03  After Henderson died, Lingo allegedly abused his power by selling the trust’s real property to himself and his wife, Denise, for a fraction of its appraised value. Diggs and Roberts sued Lingo and Denise for breach of fiduciary duty. In response, Lingo sought to compel arbitration.

9.04  The Chancery Court denied Lingo’s motion for several reasons.⁹ For one, the court noted that Denise, who held an interest in the land, was a necessary party to the dispute but a stranger to the trust. Although Denise had volunteered to arbitrate, the beneficiaries had rejected this offer, meaning that she was not subject to an agreement to bypass the judicial system. In turn, mandating arbitration of Diggs’s and Roberts’s claims against Lingo while allowing their suit against Denise to unfold in court would create the risk of inconsistent rulings. Moreover, the arbitration clause in Henderson’s trust did not comply with a state law mandating that arbitration clauses in ‘contracts relating to farm property, structures or goods … shall be additionally signed or initialed by the parties’.¹⁰ Because no party had authenticated Henderson’s arbitration provision in this manner, it was invalid.¹¹

(p. 205) 9.05  Although the FAA never surfaced in the parties’ briefs or the court’s opinion, it most likely applied. Certainly, the FAA’s centrepiece, section 2, seems only to govern arbitration provisions in contracts that involve interstate commerce.¹² Yet the US Supreme Court has read section 2 broadly and has interpreted the FAA to stretch to the full extent of Congress’s Commerce Clause power.¹³ Thus, Henderson’s trust, which included bequests to various entities outside of Tennessee, such as the Louisiana State University Animal Science Center and the Eastern Band of the Cherokee Nation,¹⁴ probably featured the requisite nexus to interstate commerce. Furthermore, rather than insisting that an arbitration clause appear in a binding contract, the Court has merely asked whether the parties have agreed to arbitrate.¹⁵ Diggs, Roberts, and Lingo formed the requisite consensual relationship. By accepting distributions or fiduciaries’ fees from Henderson’s trust, they manifested their assent to the terms of that trust, including the arbitration clause.

9.06  The outcome would have been dramatically different under the FAA. First, the FAA requires courts to compel arbitration despite ‘the presence of other persons who are parties to the underlying dispute but not to the arbitration agreement’.¹⁶ Indeed, under the vigorous federal policy in favour of arbitration, ‘piecemeal resolution, although unfortunate, is unavoidable and necessary’.¹⁷ Thus, Denise’s status as a non-beneficiary would not have exempted Diggs and Roberts from the extrajudicial forum. Second, Diggs and Roberts probably would not have been able to refuse Denise’s offer to arbitrate in a case governed by the FAA because parties to an arbitration clause are ‘estopped from avoiding arbitration with a non-signatory “when the issues the nonsignatory is seeking to resolve … are intertwined with the [underlying] agreement”’.¹⁸ Accordingly, Diggs’s and Roberts’s lack of assent to arbitrate their claims against Denise would have been irrelevant because they were complaining about the administration of the very instrument that contained the arbitration provision. Third, the FAA prohibits (p. 206) states from ‘singling out arbitration provisions for suspect status’ by encumbering agreements to arbitrate with requirements that are ‘not applicable to contracts generally’.¹⁹ The Tennessee ‘separate signing’ statute is a textbook example of an impermissible attempt to regulate arbitration. Because the FAA would sweep the state law aside, the arbitration provision in Henderson’s trust would have been enforceable.

9.07  In fact, in the topsy-turvy world of the FAA, the Chancery Court might not have even possessed jurisdiction over the matters it decided. As noted earlier, Henderson expressly adopted arbitration rules developed by the American Arbitration Association (AAA) for commercial cases.²⁰ The AAA rules give arbitrators—not judges—the exclusive right to resolve ‘any objections with respect to the existence, scope or validity of the arbitration agreement’.²¹ Although the law is still evolving, most courts hold that a party’s selection of a particular dispute resolution institution’s internal standards incorporates them by reference into an agreement to arbitrate.²² Accordingly, Henderson’s trust could have been construed to contain a ‘delegation clause’ which passes the torch to the arbitrator to decide every major issue pertaining to the proceeding.

9.08  The rest of this chapter elaborates on these topics. The chapter begins by comparing the evolution of contractual arbitration under the FAA to trust arbitration under state law. It then moves on to explain why these parallel developments are on a collision course. Specifically, the chapter argues that some arbitration provisions in trusts trigger section 2 of the FAA because they arise from contract-like instruments that affect commerce and because the parties’ conduct gives rise to an implicit agreement to arbitrate. Finally, the chapter pays special attention to two thorny problems that are likely to emerge in trust arbitration under the FAA: how to handle the in rem nature of some trust matters and the role of delegation clauses in trusts.

9.09  In the seventeenth and eighteenth centuries, it was next to impossible to obtain specific performance of an agreement to arbitrate a future controversy. Arbitration provisions were seen as illicitly ‘ousting’ the courts’ jurisdiction.²³ Likewise, under contract law’s revocability doctrine, a party could withdraw its assent to arbitrate (p. 207) at any time before the arbitrator ruled, sending the dispute back into the judicial sphere.²⁴

9.11  Despite the radical nature of the FAA, the statute’s immediate impact was muted for several reasons. First, the overwhelming consensus was that the statute arose from Congress’s Article III powers and thus applied exclusively in federal court.²⁷ Accordingly, state judges and lawmakers felt free to adopt anti-arbitration measures that echoed the ouster and revocability doctrines.²⁸ Similarly, courts exempted federal statutory claims from the FAA under a federal common law rule known as the non-arbitrability doctrine.²⁹ Courts also exempted public law claims from the statute, citing arbitration’s lay judges, abridged procedures, and narrow appellate rights.³⁰

9.12  But in the last half of the twentieth century, the Supreme Court reversed course. Announcing that the FAA expresses ‘a liberal federal policy favoring arbitration agreements, notwithstanding any state substantive or procedural policies to the contrary’,³¹ the Court began to expand the statute’s ambit. Three aspects of this jurisprudence are particularly relevant for my purposes. The first is the separability doctrine.³² Separability allocates power between courts and arbitrators, and proceeds from the legal fiction that every contract that features an arbitration clause is, in fact, two agreements: the larger ‘container’ contract and the arbitration clause, which is its own freestanding contract.³³ This bifurcated approach becomes relevant when a party seeks to invalidate all or part of an agreement that contains an arbitration (p. 208) clause. If the party’s legal theory focuses on the arbitration clause specifically—for instance, by asserting that the provision is unconscionable—then a judge resolves the matter.³⁴ But if the claim merely targets the container contract under standard contract defences such as fraud, duress, or mistake, then the sanctity of the independent agreement to arbitrate is unaffected and the dispute is sent to arbitration.³⁵

9.13  Second, despite strong historical evidence to the contrary, the Court held that Congress intended the FAA to apply in state court and override contrary state law.³⁶ Although the contours of FAA pre-emption have always been somewhat ambiguous, the touchstone is section 2’s ‘savings clause’, which declares that judges can refuse to enforce arbitration provisions ‘upon such grounds as exist at law or in equity for the revocation of any contract’.³⁷ According to the Court, this passage means that only ‘generally applicable contract defenses, such as fraud, duress, or unconscionability, may be applied to invalidate arbitration agreements’.³⁸ As a result, the FAA eclipses state statutes such as a Montana law that requires drafters to give conspicuous notice on the first page of a contract that it contains an arbitration clause³⁹ and California legislation nullifying arbitration provisions in franchise⁴⁰ or wage disputes.⁴¹ These are not state rules that govern all contracts; rather, they ‘singl[e] out arbitration provisions for suspect status’ and derive ‘meaning precisely from the fact that a contract to arbitrate is at issue’.⁴²

9.15  In sharp contrast to the FAA’s expansive treatment of arbitration clauses in contracts, the status of arbitration provisions in estate plans has never been clear. This section describes this long-standing disparity.

9.16  Although the ouster and revocability doctrines held sway in the field of contractual arbitration during the late nineteenth and early twentieth centuries, courts were less concerned about arbitration clauses in wills.⁴⁹ The tradition of testamentary freedom in the United States meant that judges were accustomed to allowing owners to leave their possessions as they wished. For instance, courts usually upheld conditional bequests allowing beneficiaries to inherit only if they marry particular individuals or achieve prescribed goals.⁵⁰ Likewise, most jurisdictions enforced no-contest clauses disinheriting anyone who filed litigation relating to the decedent’s property.⁵¹ To some judges, a mandate that beneficiaries arbitrate claims seemed no different than these standard estate planning techniques. For example, in the 1886 case of Moore v Harper, the West Virginia Supreme Court honoured an arbitration provision in a will, reasoning that ‘[t]he testator has full dominion over his property with the absolute right … to do with and dispose of it in any manner or to whomever his will or caprice may suggest.’⁵²

9.17  But after Congress passed the FAA, something curious happened: judges became increasingly hostile to testamentary arbitration. Some noted a tension between certain probate matters and arbitration. For example, issues relating to the validity (p. 210) of a testamentary instrument are in rem: rather than merely settling in personam rights of individuals, the judgment attaches to the decedent’s property and is ‘good against the world’.⁵³ As a result, probate courts generally cannot hold hearings unless the petitioner has published notice or attempted to inform all affected parties of the proceeding.⁵⁴ Some judges held that these time-consuming, formal requirements were inconsistent with the streamlined nature of private dispute resolution and refused to compel arbitration of incapacity and undue influence claims.⁵⁵ Others held that the state legislature had given probate courts exclusive jurisdiction over estate-related issues—a rationale that is indistinguishable from the ouster doctrine.⁵⁶

9.20  In sum, there has long been confusion about whether trust arbitration clauses are valid. But as discussed in the following paragraphs, most of these jurisdiction-specific authorities are beside the point. Many trust arbitration provisions need not satisfy the rigours of state law because they fall under the FAA.

9.21  On its face, the FAA applies to contracts that affect interstate and international commerce and that contain a written arbitration provision that complies with standard principles of contract law. However, as this section demonstrates, the statute really only requires some kind of underlying consensual relationship—even if it is not a full-fledged contract—that impacts interstate commerce and gives rise to an agreement to arbitrate. Notably, some trust arbitration clauses satisfy these elements.

(p. 212) 9.22  Section 2 of the FAA governs arbitration clauses in contracts ‘evidencing a transaction involving [interstate] commerce’⁶⁵ and thus is coextensive with Congress’s sweeping authority under the Commerce Clause of the US Constitution.⁶⁶ For instance, in Citizens Bank v Alafabco, the US Supreme Court held that the FAA applied to an arbitration clause in a loan restructuring agreement between an Alabama bank and an Alabama construction firm.⁶⁷ The Court explained that federal lawmakers may regulate intrinsically mercantile activity, such as banking.⁶⁸ In addition, the Court noted that the construction company had secured the loan with assets that were made from out-of-state materials and used the proceeds for projects in North Carolina and Tennessee.

9.24  Given this background, there are three possible ways that trusts could demonstrate a link to interstate commerce. First, as with the lending cases, a court might deem the multi-billion dollar business of trust administration to be inherently commercial.⁷⁴ Such a finding would make the management of settlors’ property ‘a general practice’ that is ‘subject to federal control’ even if there is no proof of an interstate dimension in any given case.⁷⁵ Second, a settlor who selects a corporate trustee likely brings the instrument within the scope of the Commerce Clause, since assets held by banks and trust companies are ‘constantly moving in and out of interstate commerce’.⁷⁶ Third, the fact a trustee or a beneficiary resides in a different state could be sufficient to bring the matter within the scope of the FAA. Of course, there would be tension between any such holding and the residential real estate cases. Nevertheless, the ongoing cross-border management and distribution of a settlor’s property arguably features a deeper economic imprint than a one-off sale of land. Thus, at bare minimum, at least some trusts ‘evidence a transaction involving commerce’ under section 2.⁷⁷

9.25  The next element—that is, the FAA’s multiple references to a ‘contract’—is even more critical. For example, section 2 begins by presumptively validating a ‘written [arbitration] provision in any … contract’.⁷⁸ How can trusts overcome this threshold textual hurdle?

9.26  The answer is that federal common law does not mandate that an arbitration clause appear in an overarching ‘contract’. As the Supreme Court has explained, section 2 ‘states that a “written provision” “to settle by arbitration a controversy” is “valid, irrevocable, and enforceable” without mention of the validity of the contract in which it is contained’.⁷⁹ The most prominent example of the way that the Court has jettisoned any requirement that an arbitration clause appear in a ‘contract’ is the separability doctrine. As noted, this rule permits arbitrators to decide challenges to the enforceability of the container contract. If the arbitrator strikes down this macro-agreement, arbitration has taken place even though there never was a contract between the parties. In fact, lower courts have gone further. A district judge in Massachusetts has enforced an arbitration provision in (p. 214) the bylaws of a real estate investment trust.⁸⁰ The California Supreme Court has compelled arbitration based on a term a developer included in a condominium’s covenants, conditions, and restrictions.⁸¹ The Eighth Circuit has upheld an arbitration clause in an employee handbook that expressly stated that it was ‘not intended to constitute a legal contract’.⁸² Likewise, the Seventh Circuit enforced an agreement to arbitrate in a tariff filed with the Federal Communications Commission, reasoning that the FAA only requires ‘an offer and acceptance that produces a legally binding document’.⁸³ Thus, the federal statute obliterates the sharp distinction that some state courts have drawn between contracts and other instruments.

9.27  Instead, section 2 simply requires an agreement to arbitrate. On this crucial issue, ‘a party’s acceptance may be implied in fact’.⁸⁴ Thus, one way that a person or entity can assent to settle disputes outside of the court system is by exploiting their rights under an instrument that contains an arbitration clause. For instance, the doctrine of direct benefits estoppel bars a litigant from receiving money or property under a document and yet disavowing its arbitration provision.⁸⁵ Courts have invoked this rule against beneficiaries asserting claims arising out of a decedent’s pension, health plan, or retirement account, reasoning that the arbitration clause governs not just the original signatory, but her ‘intended successors’.⁸⁶ As the California Court of Appeals put it, ‘[d]ecedents are able to bind their heirs through wills and other testamentary dispositions so the concept is not new or illogical.’⁸⁷

9.28  Likewise, in some contexts, parties acquiesce to a trust’s arbitration clause. Trustees are a particularly easy case. In many cases, the settlor agrees to pay a fee in return for the trustee’s promise to manage the corpus as instructed. Because the settlor dictates the terms and the trustee can either accept or decline them, courts and scholars have observed for over a century that their ‘relationship is … (p. 215) contractual’.⁸⁸ An arbitration clause is no less consensual than the other provisions that govern the trustee’s rights and duties.

9.29  Similarly, some beneficiaries also agree to arbitrate. State law generally gives individuals several months to disclaim their bequests.⁸⁹ As a result, the beneficiaries cannot be bound to the terms of an estate plan involuntarily. When someone chooses to inherit under a trust, she also accepts the other provisions of the instrument, such as what it says about how property is to be held, invested, and distributed. There is no reason to treat an arbitration clause differently.

9.30  But not every beneficiary falls within this paradigm. The critical question is whether a person has accepted benefits under the trust or is trying to exercise an entitlement that exists only because of the instrument. For instance, suppose a settlor, S, executes a trust that contains an arbitration clause and leaves one-third of her property to her son, X, one-third to her daughter, Y, and one-third to a friend, Z. X files a petition alleging that the bequest to Y is invalid on the grounds of incapacity and undue influence. Although X is seeking to overturn part of the estate plan, he is also simultaneously trying to enforce part of it. If he knocks out the gift to Y, he increases his bounty under the trust. As a result, X should be seen as tacitly agreeing to the trust’s provisions, including the language about arbitration. Conversely, if X alleges that the entire trust is unenforceable, then he has not acquiesced to the settlor’s scheme. Thus, he should be free to pursue his claim in court.

9.32  McArthur v McArthur, a California appellate decision from 2014, represents the flip side of this coin.⁹³ The settlor executed a trust in 2001 that split her estate equally among her daughters, Deborah, Kristi, and Pamela. A decade later, the settlor amended her trust to increase Kristi’s share and add an arbitration clause. Pamela then sued, contending that the 2011 amendment was invalid due to incapacity and Kristi’s undue influence. In response, Kristi sought to compel arbitration. The trial court denied the motion and the appellate panel affirmed. As the reviewing court explained, the case boiled down to one vital point: because Pamela was attempting to nullify the 2011 amendment—the only document that contained an arbitration provision—she did not agree to arbitrate.⁹⁴ Thus, like Rachal and unlike the spate of state court cases from the 2000s, McArthur did not rest on the formalistic divide between trusts and contracts. Indeed, as the court recognized, the pivotal issue is whether there is an agreement to arbitrate—an approach that would make the FAA applicable to any arbitration clause in a trust that involved interstate commerce.

9.33  However, courts seeking to apply federal legislation face a final obstacle: section 2’s savings clause. Because that component of the statute only permits courts to strike down agreements to arbitrate ‘upon grounds as exist at law or in equity for the revocation of any contract’,⁹⁵ it creates problems in the trust setting. Recall the cast of characters from Rachal: Andrew was the settlor, and John, a beneficiary, accused the drafting attorney and successor trustee, Rachal, of severe breaches of fiduciary duty.⁹⁶ Suppose John had conceded that he had assented to the arbitration provision by trying to enforce his rights under the trust, but asserted that any such agreement lacked consideration. On the one hand, to be legally sufficient, consideration merely requires an exchange of legal detriments.⁹⁷ Each party must promise to do (p. 217) something (or not do something) in return for the other party making a reciprocal pledge.⁹⁸ One could plausibly argue that the arbitration provision satisfied this criterion, since the relevant parties surrendered their freedom of action: Andrew by leaving part of his estate to John, and John by relinquishing his ability to access the judicial system. On the other hand, we typically think of bequests as donative and thus devoid of consideration.⁹⁹ Moreover, to conclude that the agreement to arbitrate is backed by consideration is to find that it is an industrial-strength contract. That would seem to eradicate the traditional boundary between conditional gifts and binding agreements.¹⁰⁰

9.34  Likewise, the savings clause might actually prevent judges from invalidating arbitration provisions in trusts under well-established probate principles. Imagine a variation on Rachal in which Rachal had not simply drafted a standard arbitration clause but prohibited the arbitrator from awarding punitive damages. Some states view provisions that exculpate a trustee from liability with suspicion, particularly when the trustee also doubles as the drafting attorney.¹⁰¹ Yet because this strand of trust law is not a ‘generally applicable contract defense[]’,¹⁰² it might be incapable of invalidating a damages waiver in cases involving an arbitration clause. These results suggest that it would be unwise to try to force the square peg of trusts into the round hole of the FAA.

9.35  Nevertheless, these are probably not insurmountable barriers. First, it is possible to conclude that agreements to arbitrate possess their own consideration without collapsing the distinction between trusts and contracts completely since the separability doctrine already conceptualizes arbitration provisions as sovereign agreements.¹⁰³ Although other trust terms—such as no-contest clauses and conditional bequests—also seem to involve a quid pro quo and thus demonstrate an element of consideration, they lack the pedigree of the separability doctrine. Thus, the doctrinal infrastructure exists to treat testamentary arbitration clauses as sui generis.

(p. 218) 9.36  Second, although trust and contract law are not coterminous, they do overlap. For instance, they share the defences of incapacity, undue influence, mistake, and duress.¹⁰⁴ And although a court might not be able to overcome a one-sided arbitration clause under trust principles, the contract defence of unconscionability should be sufficient to invalidate such a provision.¹⁰⁵ In fact, several courts that have extended the FAA to non-contracts—specifically, corporate by-laws and condominium covenants—have also entertained unconscionability challenges.¹⁰⁶ Accordingly, even without protective trust principles, beneficiaries would have weapons to combat overreaching by the drafters.

9.37  To conclude, the FAA applies if there is an agreement to arbitrate that springs from a contract-like relationship that affects interstate commerce. Some trusts meet both prongs of this test, although applying the statute raises new questions. The next section considers two particularly vexing problems: the in rem status of some trust litigation and the use of delegation clauses.

9.38  Over the years, a number of concerns have been enunciated with respect to the arbitration of claims that exist in rem. Some of these issues may be exacerbated by recent cases from the US Supreme Court concerning class arbitration.¹⁰⁷ Nevertheless, the problems posed by in rem arbitration can be solved.

9.39  The traditional unease about arbitrating in rem cases has faded from view. In the early twentieth century, judges were deeply concerned about the fact that in rem decisions are ‘binding on the whole world’ and include parties who never learn about—much less participate in—the arbitration.¹⁰⁸ That seemed fundamentally at odds with the fundamental concept of arbitration as a consensual proceeding. Thus, as the California Supreme Court proclaimed in 1900 while refusing to order a will contest to arbitration, ‘[a] few individuals, claiming to be the heirs, cannot by stipulation, determine such [a] controversy.’¹⁰⁹ Modern judges, however, have expressed fewer concerns. For instance, in 2009, a Michigan appellate court had (p. 219) no qualms about an arbitrator deciding undue influence and incapacity allegations, noting that ‘all interested parties agreed to submit their dispute to binding arbitration’.¹¹⁰ In fact, as far as this author is aware, the in rem objection has not appeared in the case law for roughly half a century.¹¹¹

9.40  But the US Supreme Court’s class arbitration jurisprudence may have changed the analysis. In 2011, the Court decided the watershed pre-emption case of AT&T Mobility LLC v Concepcion.¹¹² Without getting too deep into the details, Concepcion presented the question of whether the FAA allowed judges to find class arbitration waivers—terms that require plaintiffs to arbitrate on an individual, rather than an aggregate, basis—to be unconscionable.¹¹³ The Court held that the federal statute precluded any such ruling.¹¹⁴ Central to this conclusion was the Justices’ belief that ‘class arbitration’ is a contradiction in terms. Unlike conventional two-party arbitration, class arbitration must be designed with absent parties in mind, and therefore includes laborious and expensive notice schemes that make ‘[c]onfidentiality … more difficult’, and ‘sacrifice[] the principal advantage of arbitration—its informality.’¹¹⁵ In a similar, albeit less dramatic vein, in 2013’s Oxford Health Plans LLC v Sutter, Justice Alito’s concurrence, which Justice Thomas joined, expressed grave reservations about the efficacy of class arbitration.¹¹⁶ According to Justice Alito, absent class members would not be bound by an arbitrator’s resolution of the case because there is ‘no reason to think that [they] … ever agreed to class arbitration’.¹¹⁷ Thus, the Court has hinted that cases that require intricate procedures or run the risk of excluding many interested stakeholders may not be arbitrable.

9.41  There are striking parallels between this reasoning and the traditional resistance to arbitrating in rem disputes. To be sure, this friction is more pronounced in intestacies and will administrations, where personal representatives must take elaborate measures to keep heirs and creditors informed,¹¹⁸ yet at least in some jurisdictions, trustees occasionally must comply with similar mandates, bringing the streamlined arbitral machine to a screeching halt.¹¹⁹ Moreover, as with class arbitration, there is no guarantee that absent parties have consented to the arbitrator’s ruling. (p. 220) These gaps in the preclusive effect of arbitral awards would have to be filled with satellite litigation, making trust arbitration an exercise in futility.

9.42  Ultimately, though, one doubts that there is more than mere smoke here. In the first instance, the Supreme Court’s class arbitration opinions are unique. One need not be a legal realist to see these decisions as the by-product of the conservative Justices’ loathing of the class mechanism. In fact, Concepcion’s insistence that ‘[a]rbitration is poorly suited to the higher stakes of class litigation’ might as well have been taken from the pages of one of the now-overruled non-arbitrability cases from the middle of the twentieth century.¹²⁰ This singular, gloomy view of arbitration is unlikely to spill over into different contexts.

9.43  In addition, because of the comparatively narrow domain of trust arbitration, few cases with absent parties will ever make it to private tribunals. As this author has argued previously, arbitration clauses only apply to individuals who have taken advantage of the trust. Accordingly, it is difficult to conjure up scenarios in which some interested parties are before the arbitrator but others are not.¹²¹ For instance, suppose a settlor leaves her property to two of her five children, and one of these beneficiaries alleges that the trustee has breached her fiduciary duties. The outcome of this arbitration could not affect the three disinherited children, who have no stake in the management of property that does not belong to them. Moreover, claims that do have the potential to affect absent parties are unlikely to end up in arbitration. For instance, a challenge to the validity of a trust instrument could affect the rights of any disinherited child. But these very individuals have never taken any action that could be perceived as consenting to the arbitration clause in the trust. Thus, they would be entitled to bring their lawsuit in court. As a practical matter, then, the absent party problem may be more apparent than real.

9.44  Recently, courts in contractual arbitration cases have struggled with ‘delegation clauses’ which entrust the arbitrator—not a court—with determining the scope and validity of the arbitration provision. This section addresses how delegation clauses should work in the trust arena.

9.45  The point of departure for understanding delegation clauses is the separability doctrine. That curious feature of federal common law emanates from section 4 of the FAA, which outlines how courts should evaluate motions to compel arbitration, stating:

(p. 221) upon being satisfied that the making of the agreement for arbitration … is not in issue, … the court shall make an order directing the parties to proceed to arbitration in accordance with the terms of the agreement …. If the making of the arbitration agreement … be in issue, the court shall proceed summarily to the trial thereof.¹²²

9.46  In the seminal case of Prima Paint Corp v Flood & Conklin Manufacturing Co, the US Supreme Court held that the words ‘arbitration agreement’ meant the arbitration clause itself, rather than the container contract.¹²³ As a result, if a lawsuit specifically seeks to invalidate the arbitration provision, the petitioner has made an ‘issue’ of ‘the making of the arbitration agreement’, and a judge—not an arbitrator—presides.¹²⁴ Conversely, if a party merely argues that the container contract is invalid under a defence such as fraud, duress, or mistake, the matter goes to arbitration.¹²⁵

9.47  However, this dichotomy is subject to an important caveat. Some claims can upend both the container contract and the arbitration clause. Most importantly, under what will here be called the ‘formation exception’ to the separability doctrine, a party who never actually manifested assent to the container contract should be entitled to pursue her allegations in a judicial forum.¹²⁶ Even though her complaint targets the container contract, the gravamen of the claim is that the petitioner never agreed to anything, including the freestanding arbitration clause.¹²⁷

9.48  When viewed against this backdrop, some separability issues in trust cases will be relatively straightforward. If a trustee or beneficiary argues that the arbitration provision itself is the product of fraud, duress, mistake, or is unconscionable, the party has placed ‘the making of the arbitration agreement … in issue’ and therefore brings the dispute within the jurisdiction of the courts.¹²⁸ But judges may also sometimes entertain legal theories that are not aimed squarely at the arbitration clause. Recall the facts from McArthur, in which Pamela, a beneficiary, brought incapacity and undue influence claims against a 2011 trust amendment that contained an arbitration clause.¹²⁹ Although Pamela’s lawsuit challenged the container instrument—not the arbitration provision—it nevertheless would have fallen within the formation exception to the separability doctrine because there (p. 222) is no basis for deeming Pamela to have agreed to the 2011 amendment. To the contrary, she actively sought to resist its terms. And when a party asserts that she has never actually manifested assent to the document that contains an arbitration clause, a court—rather than an arbitrator—should wield authority.

9.49  Delegation clauses compound the complexity in this intricate doctrinal niche. They capitalize on the fact that separability is a default rule: the parties can rearrange section 4’s division of dominion between courts and arbitrators if they desire. The first glimmer of this phenomenon occurred in First Options v Kaplan.¹³⁰ Manuel Kaplan owned a company called MKI Investments, Inc. Manuel, his wife, Carol, and MKI entered into a series of contracts with First Options of Chicago, Inc. One of these agreements contained an arbitration clause, and it was between First Options and MKI. When a dispute arose, First Options sought to arbitrate not only against MKI but also against Manuel and Carol Kaplan in their individual capacities. Although MKI complied, the Kaplans objected to the arbitrator’s assertion of power over them, noting that they had never signed a contract that contained an arbitration provision. First Options replied that the Kaplans had tacitly agreed to arbitrate the issue of whether they had agreed to arbitrate the merits by filing briefs within the arbitral forum.

9.50  The Supreme Court concurred with First Options that parties can agree to arbitrate ‘arbitrability’: that is, parties can ask the arbitrator to decide matters that section 4 of the FAA assigns to judges. The Court explained that the parties’ intent was paramount and that there was no reason to restrict the varieties of issues that they could to submit to arbitration. However, the Court also created a heavy presumption against finding any such arrangement, declaring that it would take ‘clear and unmistakable evidence’ to overcome the statute’s apportionment of power between judges and arbitrators.¹³¹ First Options’ proof on this issue fell well short of the mark. Indeed, the Kaplans had only appeared before the arbitrator to argue that they had never agreed to arbitrate at all.¹³²

9.51  Nevertheless, First Options opened the door for drafters to use delegation clauses that meet the ‘clear and unmistakable’ standard by expressly assigning to arbitrators controversies that judges would normally decide. The Court considered such a provision in 2010 in Rent-A-Center, West, Inc v Jackson.¹³³ Antonio Jackson, an employee of the automobile leasing giant, signed an arbitration agreement that included a delegation clause that gave the arbitrator the ‘exclusive authority to resolve … any claim that all or any part of this [a]greement is void or voidable’.¹³⁴ Later, Jackson sued for race discrimination. In the district court, he argued that (p. 223) the arbitration provision was unconscionable when applied to the merits of his civil rights claims because it required him to pay half of the arbitrator’s fees and limited his discovery rights.¹³⁵ The Supreme Court first held that the delegation provision unambiguously entrusted challenges to the arbitration agreement to the arbitrator, thus fulfilling the mandate from First Options.¹³⁶

9.52  The Supreme Court then viewed delegation provisions through the prism of the separability doctrine. The Justices reasoned that just as arbitration clauses are autonomous agreements within larger container contracts, delegation clauses are (1) an agreement to arbitrate the validity of the arbitration clause (2) within an agreement to arbitrate the merits of the dispute (3) within the container contract.¹³⁷ This meant that Jackson had missed the proverbial boat by arguing that the arbitration clause (as opposed to the delegation provision) was unconscionable. Just as he could not escape arbitration by contending that the container contract was invalid, he could not escape the delegation clause by attacking the agreement to arbitrate the merits of his case. Although a court could have probed whether the delegation clause itself was unconscionable, Jackson had made no such argument. Thus, he needed to arbitrate his argument that the arbitration clause was unconscionable.¹³⁸

9.53  This expansion of arbitral autonomy has the potential to wreak havoc in the trust context. Three separate issues will be discussed here. First, there is a colourable argument that delegation clauses in trusts do not satisfy the ‘clear and unmistakable’ rule. After all, First Options held that the Kaplans’ participation in the arbitration was not enough to reveal that they had agreed to arbitrate the question of whether they were covered by the arbitration clause in the contract between First Options and MKI. Likewise, trust arbitration also hinges on how the parties behave. Unlike the parties to a contract, most trustees and beneficiaries never sign a document that contains an arbitration or delegation clause; rather, they manifest assent through their actions. Perhaps conduct alone can never be ‘clear and unmistakable’ evidence of the parties’ intent to allow the arbitrator to decide whether the arbitration clause is valid.

9.54  Ultimately, this approach is not persuasive. In First Options, no contract expressly allowed the arbitrator to decide whether the Kaplans fell within the scope of the arbitration clause in the First Options/MKI agreement. Accordingly, First Options’ claim that the Kaplans had agreed to arbitrate the issue of whether they had agreed to arbitrate the merits was a stretch: it would have been extraordinary if the Court had found an implied agreement to arbitrate based on nothing more than the Kaplans’ vociferous objections to proceeding in the extrajudicial tribunal. The dynamic in trust arbitration is different. If a settlor includes a delegation clause, she provides the key variable that was missing in First Options: a writing that empowers (p. 224) the arbitrator to hear matters that traditionally are the prerogative of courts. The beneficiaries either accept the terms of the trust, including the delegation clause, or they do not. But the fact that the trust beneficiaries express their agreement through ex post conduct rather than an ex ante signature is irrelevant. Absent unusual circumstances, assent is assent—clearly and unmistakably.

9.55  A second issue relates to the use of what will here be called ‘ghost’ delegation clauses. Contracts occasionally incorporate a particular dispute resolution provider’s rules by reference. Some of these arbitral institutions (including the AAA, Judicial Arbitration and Mediation Services, Inc. (JAMS), and the International Chamber of Commerce (ICC)) allow an arbitrator to ‘rule on his or her own jurisdiction, including any objections with respect to the existence, scope or validity of the arbitration agreement’.¹³⁹ In matters involving sophisticated parties, courts have uniformly found that opting into one of these entities’ internal rules is sufficient to satisfy the First Options ‘clear and unmistakable’ rule.¹⁴⁰ But judges have differed over whether the same result flows from adhesion contracts. A few have reasoned that because consumers and employees do not read or understand the fine print of a contract, merely mentioning a specific arbitration administrator does not meet the standard necessary to create a delegation clause.¹⁴¹

9.56  Some trust arbitration provisions also contain ghost delegation clauses. For instance, in Diggs v Lingo, mentioned previously, the settlor specified that arbitration should occur under the AAA Commercial Rules, which allow arbitrators to define the ambit and decide the enforceability of the arbitration provision.¹⁴² Notably, the AAA’s newly promulgated rules for wills and trust arbitration contain precisely the same tenet.¹⁴³ Thus, any testamentary instrument that incorporates either of these rule (p. 225) sets contains a ghost delegation clause. Should courts defer to these phantom terms (as is the case in negotiated deals between peers) or be sceptical of them (as is the case with adhesion contracts)?

9.57  One interpretation is that trusts are more like transactions between sophisticated parties than they are like boilerplate relationships. For one, being named as a trust beneficiary is a relatively unusual occurrence as compared to acting as a consumer or employee. Thus, beneficiaries are more likely to read and maybe even scrutinize trust arbitration clauses. In addition, beneficiaries have several months to decide whether to accept or disclaim a gift.¹⁴⁴ That is a far more leisurely timeline than the spur-of-the-moment decisions that often give rise to adhesion contracts. At least in the abstract, then, ghost delegation clauses may be less troubling in trusts than they are in consumer and employment contracts.

9.59  Not everyone on the appellate panel was persuaded. Dissenting Judge Dennis chastised the majority for ignoring the spirit of the FAA.¹⁴⁷

9.60  Trust arbitration under the FAA compounds the confusion relating to questions of scope. Suppose a settlor permits the arbitrator to define the boundaries of the agreement to arbitrate. A beneficiary then seeks to invalidate the entire trust. Should a court honour the plain language of the delegation clause and permit the arbitrator to decide whether the beneficiary has agreed to arbitrate the merits of her petition? Or, as in Douglas, should a court retain jurisdiction if it determines that the assertion of arbitral authority is ‘wholly groundless’? Or is there some other path out of this maze?

9.61  The best solution, in this author’s view, revolves around the formation exception to the separability doctrine. As described earlier, a party who contends that he or she never agreed to an instrument that includes an arbitration provision should be entitled to pursue that theory in court. Similarly, a beneficiary who attempts to strike down a trust instrument that contains a delegation clause has no business being sent to arbitration. In Douglas, the plaintiff had assented to something (ie an arbitration provision with a delegation clause when she opened her Union account) and therefore the question was how far this agreement extended. Conversely, the hypothetical beneficiary who wishes to nullify the entire trust has not agreed to a single term in the instrument: not its dispositive provisions, not its investment instructions, and certainly not its language allowing arbitrators to decide who has agreed to arbitrate. Even Rent-A-Center acknowledged that courts should preside over a party’s argument that a delegation clause is unconscionable.¹⁴⁸ The same should hold true when a trust beneficiary contends that she never actually agreed to an instrument that contains the delegation clause in the first place.

9.62  The debate over arbitration clauses in trusts has been confined to state law. This chapter has argued that some such provisions fall within the FAA. Indeed, they (p. 227) involve interstate commerce, arise from contract-like relationships, and generate agreements to arbitrate. In addition, although this development may raise new questions about the arbitration of in rem claims or the scope of delegation clauses, the answers can be derived from existing law. It is only a matter of time before some attentive litigant or court brings trust arbitration under the FAA’s broad umbrella.