Arbitration Waiver and Prejudice

Timothy Leake* November, 2020

Arbitration agreements are common in commercial and consumer contracts. But two parties can litigate an arbitrable dispute in court if neither party seeks arbitration. That presents a problem if one party changes its mind and invokes its arbitration rights months or years after the lawsuit was filed and substantial litigation activity has taken place. Federal and state courts agree that a party can waive its arbitration rights by engaging in sufficient litigation activity without seeking arbitration, but they take different approaches to deciding how much litigation is too much. Two basic methods exist. Some courts say waiver requires the party opposing arbitration to show it would be prejudiced by the delay. Others say that waiver does not require a showing of prejudice.

This Note demonstrates that the presence or absence of a prejudice requirement does not accurately capture the disagreements between the federal circuit courts. Indeed, some circuits that impose a prejudice requirement will find waiver in circumstances where other courts that do not impose a prejudice requirement will not. These divergent approaches result in uncertainty, delay, and expense, undermining arbitration’s benefits. To resolve the circuit split, this Note proposes a bright-line standard under which engaging in litigation never supports a finding of waiver. It also shows that this approach is consistent with common law waiver doctrine and the Federal Arbitration Act.

Introduction

In 2017, the Shaw Group sued former employee Dorsey McCall in Louisiana state court for a breach of contract.¹ The contract included an arbitration agreement, and after Aptim Corporation acquired part of Shaw, it asked the state court to enforce the arbitration agreement.² The state court declined, finding that the litigation that had already taken place “constituted a waiver of the arbitration provision.”³ Undeterred, Aptim sought relief in federal court. Despite the fact that Louisiana state courts and the Fifth Circuit apply similar tests to determine whether a party has waived its arbitration rights,⁴ the federal court decided that Aptim could pursue arbitration and stayed the state court case.⁵

Aptim Corp. v. McCall illustrates the confused state of the law about the waiver of arbitration rights. The basic question is easy to state: What standard should determine whether a party to an arbitration agreement has waived its right to arbitrate by engaging in litigation? The answer is less clear. Various state and federal courts have adopted conflicting tests.⁶ In particular, the federal circuits have taken different views about whether prejudice to the party opposing arbitration is a necessary element of waiver.⁷ And despite the Supreme Court’s interest in arbitration waiver,⁸ it has not yet resolved the circuit split.⁹

The question is especially important since the use of arbitration agreements has skyrocketed over time. Once a little-known method for quickly resolving disputes between businesses, arbitration agreements are now commonplace in consumer and employment contracts.¹⁰ The Supreme Court has encouraged this trend by interpreting the Federal Arbitration Act to favor expansive arbitration rights.¹¹ In response, opponents of arbitration agreements have criticized their use in contexts ranging from employment¹² to banking.¹³ Media scrutiny, meanwhile, has focused on the extent to which businesses can use arbitration agreements to conceal allegations of sexual harassment from the public.¹⁴ Arbitration-waiver rules help to determine when these ubiquitous agreements are enforceable. For example, consumers’ ability to sue a business as a class can depend on whether the business waived its arbitration rights.¹⁵

This Note provides a framework to resolve disagreements about arbitration waiver. Part I explores how standards for arbitration waiver evolved from interactions between common law waiver and the Federal Arbitration Act. Part II examines existing arbitration-waiver rules and challenges the usefulness of classifying them by the presence or absence of a prejudice requirement. Finally, Part III proposes a bright-line approach in which a party’s litigation conduct does not support a finding of waiver. This bright-line approach is consistent with common law waiver doctrines and the Federal Arbitration Act.

I. Waiver in the Arbitration Context

The Federal Arbitration Act (FAA)¹⁶ determines when a party can ask a court to enforce an arbitration agreement. Under the FAA, a party to a suit involving an arbitrable claim is ordinarily entitled to a stay of the litigation to allow arbitration to proceed.¹⁷ But section 3 of the FAA says that a party “in default in proceeding with such arbitration” is not entitled to a stay.¹⁸ One way to default is to litigate rather than invoke a binding arbitration agreement.¹⁹ Just how much a party can litigate before it defaults under section 3 is a question that has long divided both federal and state courts. The answer depends on principles of common law waiver and the policy underlying the FAA.

A. Principles of Arbitration and Waiver

The FAA does not define “default,” so courts seek guidance in the common law doctrines of waiver and forfeiture. Most compare default to waiver.²⁰ Broadly speaking, waiver refers to the “voluntary relinquishment or abandonment—express or implied—of a legal right or advantage.”²¹ Waiver can be established by words or conduct without any particular formal requirements; however, it does require a “clear, unequivocal, and decisive act.”²² It does not rely on any acts by the opposing party, and a waiving party must know that it possesses an underlying right in order to waive it.²³ Finally, waiver is an equitable doctrine with the ultimate goal of “further[ing] the interests of justice.”²⁴Under this approach, litigation conduct inconsistent with arbitration supports waiver.²⁵ Unfortunately, no clear test exists to determine if litigation conduct is inconsistent with arbitration rights.²⁶

At least three federal circuits invoke doctrines of forfeiture to analyze defaults under section 3 in addition to, or instead of, waiver. Forfeiture, generally defined as “failure to make a timely assertion of a right,” does not require intent.²⁷ In Zuckerman Spaeder, LLP v. Auffenberg, for example, the D.C. Circuit embraced forfeiture as the proper mode of analysis for claims of default under section 3.²⁸ Similarly, the Seventh Circuit has distinguished waiver from forfeiture in the criminal context.²⁹ In the arbitration context, on the other hand, it decided to use waiver as an umbrella term encompassing both concepts.³⁰ The Fourth Circuit has also referred to both waiver and forfeiture of the right to arbitrate, but it did not distinguish them.³¹ Nevertheless, some decisions in the Fourth, Seventh, and D.C. Circuits have spoken in terms of waiver rather than forfeiture.³² The willingness of these courts to use forfeiture and waiver interchangeably suggests that the distinction has little practical effect. Consistent with the majority approach, this Note uses waiver as a general term to refer to defaults under section 3 caused by excessive litigation activity.

Regardless of terminology, there is broad agreement that the right to arbitrate can be waived in some fashion.³³ And courts have recognized both express and implied waiver.³⁴ Beyond these general principles, however, there is “no set rule as to what constitutes a waiver or abandonment of the arbitration agreement.”³⁵

B. Policy and Prejudice

Without a clear waiver test in the FAA’s text, courts look to the law’s purpose of promoting arbitration and the benefits that stem from it. Arbitration allows parties to resolve disputes quickly by limiting access to procedural devices like discovery and class certification.³⁶ Parties can also provide for arbitration by a subject-matter expert, which helps them resolve technical disputes outside the judiciary’s expertise.³⁷ Moreover, parties can protect proprietary information like trade secrets by imposing confidentiality requirements in arbitration.³⁸

Despite these benefits, courts were once wary of enforcing arbitration agreements.³⁹ In response, Congress passed the FAA to establish a “liberal federal policy favoring arbitration.”⁴⁰ The FAA has two specific goals: “enforcement of private agreements and encouragement of efficient and speedy dispute resolution.”⁴¹ Consistent with the first goal, arbitration is “a matter of contract,” so state contract law typically determines whether a dispute is arbitrable.⁴² However, the use of contract law doctrines like unconscionability to limit arbitration might undermine the FAA’s efficiency goals.⁴³ For this reason, the FAA preempts traditional contract law principles when those principles are “applied in a fashion that disfavors arbitration.”⁴⁴

To accommodate the FAA’s pro-arbitration stance, many courts have grafted prejudice requirements onto traditional waiver tests.⁴⁵ Prejudice refers to an “unfair tactical advantage” that one party would gain over another by engaging in litigation and then moving to compel arbitration.⁴⁶ For example, a party faces prejudice if its opponent moved for arbitration after the opponent obtained discovery unavailable in arbitration.⁴⁷ A party also faces prejudice if its opponent moved for arbitration after the opponent lost on a motion going to the merits of the dispute.⁴⁸ But it is often difficult to tell exactly what conduct is prejudicial in particular cases, especially because different circuits have adopted different kinds of prejudice requirements.⁴⁹ As a result, the question of waiver can become much more complicated when courts add prejudice as an element.

However, making it difficult to establish waiver could undermine the FAA’s efficiency goals. Allowing a party to test the waters in litigation and then jump to arbitration for a do-over if the litigation goes poorly has the potential to disincentivize parties from invoking arbitration at the earliest possible opportunity.⁵⁰ That could lead to duplicative proceedings, excessive costs, and needless delays.⁵¹ These countervailing considerations also sow confusion and inconsistency in the approaches of the disparate circuits.

Part II explores the different standards for waiver adopted by various circuit courts in more detail. It also shows that waiver is not always harder to establish in circuits that impose a prejudice requirement, which calls into question the utility of classifying arbitration-waiver rules based on whether or not prejudice is required.

II. The Role of Prejudice in Waiver Determinations

Arbitration, like other contractual rights, can be waived.⁵² Waiver, in turn, can be expressly asserted or inferred from a party’s conduct.⁵³ And litigating a dispute subject to arbitration is conduct that can support a finding of waiver.⁵⁴ But the federal courts of appeals disagree about the proper standard for determining when a party that engaged in litigation has waived its right to arbitrate. In particular, they split as to whether a party seeking to avoid arbitration must affirmatively show that it suffered prejudice from the opposing party’s delay.

A. The Circuit Split

Among the federal courts, there are three separate approaches to determining whether a litigant has waived arbitration rights. First, the Seventh and D.C. Circuits apply a traditional waiver test under which a party waives the right to arbitrate if its litigation conduct is inconsistent with arbitration, regardless of prejudice.⁵⁵ Second, the Tenth Circuit applies a multifactor test with prejudice as a nondispositive factor.⁵⁶ And third, the remaining circuits apply prejudice-based tests that allow a party to invoke its arbitration rights unless its litigation conduct is inconsistent with arbitration and prejudicial.⁵⁷

The Seventh and D.C. Circuits apply a traditional waiver test that presumes waiver once a party starts to litigate without demanding arbitration.⁵⁸ In Cabinetree of Wisconsin, Inc. v. Kraftmaid Cabinetry, Inc., Judge Posner explained that a party’s decision to litigate is inconsistent with arbitration, implying that the party intends to waive its arbitration rights.⁵⁹ This traditional approach derives from common law waiver, which depends on a party’s conduct rather than on how that party’s acts affect its opponent.⁶⁰ Courts applying the traditional test appeal to the idea that “Congress’s goal in enacting the Arbitration Act was to place arbitration agreements ‘upon the same footing as other contracts.’”⁶¹ As such, “the court is not to place its thumb on the scales” to favor arbitration by reading a prejudice requirement into its discussion of waiver.⁶² And as a practical matter, the Seventh and D.C. Circuits believe that their approach minimizes needless duplication.⁶³

Meanwhile, the Tenth Circuit considers prejudice as part of a multifactor test (the Peterson test).⁶⁴ The Peterson test does not derive from any statutory text but rather combines factors that federal courts frequently use to decide waiver cases.⁶⁵ The test is flexible: courts can add or discount factors, and they need not adhere to a strict prejudice-based or traditional waiver approach.⁶⁶

In contrast, the remaining circuits apply prejudice-based tests.⁶⁷ The use of prejudice as a prerequisite for arbitration waiver differs from waiver in other contexts as a result of the “liberal federal policy favoring arbitration agreements.”⁶⁸ The Supreme Court has declared that “any doubts concerning the scope of arbitrable issues should be resolved in favor of arbitration,” including doubts arising from “an allegation of waiver, delay, or a like defense to arbitrability.”⁶⁹ Imposing a prejudice requirement as a prerequisite to a finding of waiver gives effect to the federal policy favoring arbitration of disputes.⁷⁰

Widespread agreement that prejudice is an element of waiver obscures significant differences between the circuits that use prejudice-based tests regarding exactly what constitutes prejudice. At one end of the spectrum, the prejudice requirement is minimal; nearly any litigation expense incurred by the party opposing arbitration is sufficient.⁷¹ For example, in Joca-Roca Real Estate, LLC v. Brennan, the First Circuit found that the expense of conducting discovery for eight months was prejudicial, even if similar discovery would have been available in arbitration, because the delay meant that the party opposing arbitration was unable to “tailor his discovery strategy” to account for the possibility of arbitration.⁷² In practice, this weak prejudice requirement resembles a presumption of waiver if arbitration is not invoked at the earliest possible opportunity.

At the other end of the spectrum, a party must show prejudice by producing unequivocal evidence that its opponent obtained an unfair advantage through litigation that would have been unavailable in arbitration. For example, in MicroStrategy, Inc. v. Lauricia, the Fourth Circuit reasoned that no prejudice resulted where the party opposing arbitration had provided information during discovery in litigation that would have been available during discovery in arbitration as well.⁷³ The existence of different kinds of prejudice is unsurprising because neither the FAA’s text nor the Supreme Court’s jurisprudence provides specific guidance about how to apply the FAA’s pro-arbitration policy in the arbitration-waiver context.⁷⁴

The circuit split is longstanding and intractable. Conflicting approaches have existed for at least fifty years.⁷⁵ And the Supreme Court has not yet been able to resolve the dispute despite its apparent interest in doing so.⁷⁶

Commentators have stepped into the void, proposing a variety of tests to resolve the circuits’ current disagreements.⁷⁷ They tend to divide into two camps, either endorsing or rejecting a prejudice-based test. For example, James Savage suggests that courts should draw from doctrines of estoppel rather than waiver because engaging in litigation does not necessarily show an intention to relinquish arbitration rights; estoppel requires prejudice.⁷⁸ Paul Bennett IV takes the milder view that prejudice should be considered, but only in the context of a multifactor test.⁷⁹ On the other side, Professor Richard Frankel suggests that imposing a prejudice requirement undermines the FAA’s efficiency goals by enabling duplicative proceedings in court and in arbitration.⁸⁰ Similarly, Professor Thomas J. Lilly, Jr. endorses a bright-line approach under which arbitration rights are waived if not asserted in an answer, regardless of whether prejudice exists.⁸¹ Notwithstanding disagreement about the best solution, there is widespread consensus that prejudice is the key dividing line between existing waiver tests.⁸²

B. Is Prejudice Useful?

It is tempting to think of prejudice as an extra hurdle a party must surmount to establish waiver. Indeed, the absence of prejudice can be fatal to waiver claims regardless of whether a party’s litigation conduct shows that it intended to waive its arbitration rights.⁸³ But this view is overly simplistic. Courts that do not require a formal showing of prejudice weigh the same factors to decide if a party has waived its contractual right to arbitrate that other courts use to figure out whether prejudice—and thus waiver—exists. A closer analysis of three factors reveals that waiver is not necessarily more difficult to obtain in circuits that apply prejudice-based tests. These factors are (1) motion practice, (2) discovery, and (3) delay.

First, both courts that require prejudice and those that do not determine waiver partly by considering the extent to which parties have engaged in motion practice. Under the traditional waiver test, a party’s participation in litigation is inconsistent with its right to arbitrate.⁸⁴ This conclusion is especially strong once the party asks the court for a decision on the merits.⁸⁵ Similarly, engaging in litigation can be prejudicial. Under an expansive conception of prejudice, the expenses inherent in motion practice, which would have been avoided in arbitration, prejudice the party opposing arbitration.⁸⁶ An even more exacting prejudice standard is satisfied when a party learns an opponent’s litigation strategy through motion practice or prevails on a merits issue in court before the opponent invokes the arbitration agreement.⁸⁷

In fact, traditional waiver tests can be more difficult to satisfy than prejudice-based formulations with respect to the extent of motion practice required to find waiver. For instance, the Seventh Circuit has not required prejudice but has also held that engaging in non-merits litigation cannot support a finding of waiver since it does not show that a party wants the court to resolve the dispute on the merits.⁸⁸ In prejudice-based circuits, engaging in non-merits litigation can be prejudicial once enough time and money have been expended.⁸⁹ Thus, engaging in non-merits litigation may establish waiver under a prejudice-based test without satisfying a traditional test.

Second, participating in discovery may support a finding of waiver in both courts that apply a prejudice-based test and those that do not. A party’s pursuit of discovery through judicial processes rather than arbitration suggests that the party does not want to pursue arbitration.⁹⁰ Even acquiescence to court-ordered discovery suggests abandonment of arbitration, since a party could otherwise move to stay discovery and pursue arbitration.⁹¹ Again, these considerations go hand in hand with prejudice. Discovery in court is expensive; needless expense creates prejudice.⁹² Discovery also serves as the paradigmatic example of prejudice: it might yield information that would not be available through more restrictive discovery on offer in arbitration.⁹³ And even when the same discovery is available in arbitration and litigation, quicker procedures in arbitration can prejudice a party forced to rely on slower, more expensive processes in litigation.⁹⁴

As with motion practice, engaging in discovery may be prejudicial and still fail to justify waiver in courts that apply traditional waiver tests. In Cabinetree of Wisconsin, Inc. v. Kraftmaid Cabinetry, Inc., the Seventh Circuit found waiver but noted that engaging in discovery does not necessarily foreclose arbitration rights.⁹⁵ Later, the Seventh Circuit held that “participation in discovery [that] was not as extensive as that of the defendant in Cabinetree” did not support a finding of waiver.⁹⁶ However, even minimal discovery can support waiver under lax prejudice-based tests. In In re Citigroup, Inc., the First Circuit emphasized that the defendant “failed to timely invoke its rights,” allowing the litigation to “proceed[] into discovery.” This discovery justified a finding of waiver even though the only discovery relevant to the claims at issue consisted of “five narrowly tailored . . . interrogatories” that the defendant asserted would have been available in arbitration.⁹⁷ Other circuits using prejudice-based tests have also found waiver despite limited discovery.⁹⁸ While the distinction between prejudice-based and traditional waiver rules in the context of discovery may not be entirely clear-cut, it is nevertheless sufficiently porous to call into question the usefulness of categorizing courts based on a prejudice / no prejudice distinction.

Finally, unnecessary delay supports a finding of waiver under both traditional and prejudice-based tests. A party’s failure to assert its right to arbitrate at the earliest possible opportunity supports an inference of disinterest in arbitration,⁹⁹ an inference that strengthens over time as trial approaches.¹⁰⁰ Courts that require prejudice also take delay into account. Needless delays from the start of a lawsuit can, but do not necessarily, cause prejudice sufficient to support a finding of waiver.¹⁰¹ And proximity to trial provides strong evidence of prejudice because the party opposing arbitration will have wasted time and money on trial preparations,¹⁰² especially when the costs would be duplicated in arbitration.¹⁰³ As with motion practice and discovery, lengthy delays in jurisdictions with traditional tests do not always lead to waiver findings, but short delays in jurisdictions with prejudice-based tests can.¹⁰⁴ These results show that imposing a prejudice requirement does not necessarily make waiver more difficult to obtain in practice.

Formal similarities between the Third and Tenth Circuits’ arbitration-waiver tests also suggest that prejudice does little independent work in the waiver analysis. In the Third Circuit, waiver depends solely on prejudice, which is defined by reference to several factors.¹⁰⁵ Meanwhile, the Tenth Circuit’s multifactor approach treats prejudice as a nondispositive factor.¹⁰⁶ The two circuits’ tests are strikingly similar. Each includes factors that capture (a) the timeliness of the motion to arbitrate; (b) the extent of discovery; (c) the extent of motion practice; and (d) whether the party seeking arbitration warned the opposing party.¹⁰⁷ Moreover, the Tenth Circuit’s use of prejudice largely overlaps with its treatment of delay.¹⁰⁸

Overall, courts look to the same facts to make waiver determinations regardless of how they characterize their prejudice requirements. As a result, some courts can apply prejudice-based tests and find waiver while others apply traditional waiver tests in similar circumstances yet conclude there was no waiver. These findings cast doubt on the utility of prejudice as a classification tool for arbitration-waiver tests.

C. Other Roles for Prejudice

Even if prejudice-based rules are not always more demanding than traditional waiver tests, there is at least a conceptual difference between the two: traditional tests focus only on intent expressed through a party’s conduct, while prejudice-based tests focus on the consequences of that conduct for the other party. Some courts have adopted rules that appear to capture this distinction. However, these rules rarely result in different outcomes in practice.

First, the Fourth Circuit distinguishes discovery available in both arbitration and litigation from discovery available only in litigation because “[i]f the same information could have been obtained in [arbitration], then [the party opposing arbitration] would have suffered no prejudice.”¹⁰⁹ Under the Fourth Circuit’s approach, a party could invoke its right to arbitrate after it requests discovery from a court, or even after it obtains discovery that it also could have gotten in arbitration.¹¹⁰ In contrast, a court that does not require a showing of prejudice could find waiver based merely on a party’s pursuit of discovery.¹¹¹

But the Fourth Circuit’s approach is not widely followed, even in other circuits that require a showing of prejudice. For example, obtaining discovery going to the merits constitutes prejudice in the Fifth Circuit’s prejudice-based test.¹¹² Moreover, engaging in litigation that would be duplicated in arbitration establishes an intent to waive under a traditional test.¹¹³ Conversely, a court applying the traditional test would likely give little weight to a party’s pursuit of discovery that would be also usable in arbitration when it determines whether that party intended to waive its arbitration rights.¹¹⁴

Next, the Ninth Circuit has articulated a “self-inflicted wound[]” theory, holding that a party opposing arbitration cannot support a finding of prejudice by pointing to costs it incurred because of its decision to litigate.¹¹⁵ Under this method, a court may allow a party to assert its arbitration rights even after the opposing party has paid substantial litigation expenses associated with preparing a complaint, serving notice, and arguing non-merits issues.¹¹⁶ In theory, prejudice—and thus waiver—is harder to establish under the Ninth Circuit’s approach than an ordinary prejudice-based test. In practice, however, the Ninth Circuit’s approach still overlaps with other arbitration-waiver tests. It is similar to those prejudice-based tests that do not consider non-merits litigation prejudicial.¹¹⁷ It is also similar to traditional waiver tests. Actions by the party opposing arbitration are not relevant to decide if the party seeking arbitration relinquished its arbitration rights under traditional waiver tests, so costs associated with service and litigating non-merits issues incurred by the party opposing arbitration do not support a finding of waiver.¹¹⁸ As a result, the Ninth Circuit’s approach is just as similar to traditional waiver tests as other prejudice-based tests.

Finally, the Seventh and D.C. Circuits’ traditional waiver tests state that a party presumptively waives arbitration rights it fails to invoke at the earliest opportunity.¹¹⁹ Traditional tests with strong waiver presumptions could lead to different outcomes than prejudice-based tests, which lack waiver presumptions, even if minimal prejudice is required. But today, the Seventh and D.C. Circuits’ waiver presumptions appear to be largely toothless. The Seventh Circuit has treated the presumption as mere dicta and even explicitly rejected it in recent cases.¹²⁰ Similarly, the body of case law in the D.C. Circuit suggests that the presumption has hardly any effect on the actual outcomes of waiver questions. While the D.C. Circuit still discusses presumptive waiver, its decisions rest on factors in common with those of the other circuits, including the extent of discovery and litigation; which party bears the burden of proof does not control the analysis.¹²¹ With the waiver presumption waning in the Seventh Circuit and possessing little importance in the D.C. Circuit, litigants must rely on the same arguments about motion practice, discovery, and delay as those in courts that apply prejudice-based tests.

These distinctions show that differences within prejudice-based and traditional waiver rules can be as substantial as differences between them. Consequently, prejudice is not an especially useful classification tool in attempting to solve the deficiencies and confusion in current approaches to arbitration waiver. Part III proposes an alternative standard that minimizes confusion and promotes consistent results.

III. The No-Waiver Alternative

Existing approaches to arbitration waiver foster uncertainty and foment time-consuming disputes about the extent of litigation required to establish waiver, whether or not courts impose a formal prejudice requirement. A better approach discards these context-dependent inquiries to promote certainty while remaining faithful to common law waiver doctrines, achieve consistency with the FAA’s text and purpose, and foster fair resolution of disputes.

A. Alternative Standard

Courts struggle to establish consistent tests for waiver because they erroneously treat litigation activity as evidence of an intent to relinquish arbitration rights. Nevertheless, some courts have recognized specific circumstances in which litigation does not establish the intent to waive arbitration rights under either traditional or prejudice-based tests. For example, in these courts, neither engaging in discovery¹²² nor litigating jurisdictional issues¹²³ necessarily establishes intent to waive arbitration rights. Even filing a lawsuit—at first glance, a clear sign that a party wants a court to decide a dispute—can be merely “a step towards choosing litigation” that does not support waiver if the plaintiff does not pursue the suit.¹²⁴

The real concern underlying both traditional and prejudice-based waiver tests is fairness rather than intent.¹²⁵ One fear is that a party might abuse the system by litigating and then abruptly changing course to pursue arbitration if the litigation goes poorly.¹²⁶ Judge Posner characterized that kind of late move to arbitration as an attempt to play “heads I win, tails you lose” with the other party.¹²⁷ Others have viewed attempts to compel arbitration as examples of “[w]illfulness and [b]ad [f]aith” designed to “manipulate[] the situation”¹²⁸ or mislead the court.¹²⁹

Concerns about fairness also push courts to consider how the nonmoving party will be affected by a motion to compel arbitration. That explains why some courts may decline to find waiver, even if a party has acted inconsistently with its arbitration rights, where there was no prejudice to the party opposing arbitration.¹³⁰ It also explains why courts create standards like “actual prejudice”¹³¹ and “unfair prejudice”¹³² to describe how much delay and litigation expense is too much, rather than follow an inflexible rule that requires a party to immediately assert its arbitration rights.¹³³ Unfortunately, waiver’s focus on determining the intent of the parties makes the doctrine ill-suited to address general fairness concerns.

A better approach recognizes that litigation activity is not the sort of unequivocal conduct that supports a finding of implied waiver. Courts should grant stays under section 3 only if a party has expressly waived its right to arbitration or engaged in unequivocal conduct beyond mere participation in litigation that establishes its intent to waive. Examples of express or unequivocal conduct that should be required to find waiver include a party saying it is “better off” in litigation,¹³⁴ refusing to arbitrate,¹³⁵ or denying that an arbitration agreement exists.¹³⁶ This approach furthers the FAA’s text and purpose, adheres to traditional waiver doctrines, and promotes fair dispute resolution. Moreover, it discourages wasteful litigation by reducing the unpredictability inherent in contemporary interest-balancing approaches. For simplicity, the rule proposed in this paragraph will be referred to as the “bright-line rule.”

B. Consistency with the FAA

The FAA’s text and purpose support the bright-line rule. Section 3 instructs courts to grant stays if “the applicant for the stay is not in default in proceeding with . . . arbitration.”¹³⁷ The definition of “default” around the time Congress enacted the FAA was the “omission or failure to fulfill a duty, observe a promise, discharge an obligation, or perform an agreement.”¹³⁸ Default did not require harm to the other party. Moreover, courts that applied the FAA shortly after its adoption read “in default” to mean refusal to proceed with arbitration.¹³⁹ The same understanding persists today. For example, a party is “in default” under section 3 when it fails to pay required arbitration fees.¹⁴⁰ Thus, there is no textual basis to find a party “in default” just because its opponent faces prejudice.

Adopting a narrow interpretation of “default” also supports the FAA’s policy goals. At common law, judges were hostile to arbitration, and parties could easily repudiate their arbitration agreements.¹⁴¹ In response, Congress passed the FAA to enshrine arbitration as a preferred method of dispute resolution.¹⁴²Under modern Supreme Court jurisprudence, arbitration agreements must be enforced “according to their terms,”¹⁴³ and any doubts about the enforceability of arbitration agreements are resolved in favor of arbitration.¹⁴⁴ In particular, the Court has explained that arbitration should typically prevail over “an allegation of waiver” and similar defenses.¹⁴⁵ Moreover, efforts to carve out exceptions to the enforceability of arbitration agreements are disfavored.¹⁴⁶ Applying the FAA’s pro-arbitration policy to arbitration waiver indicates that judges should not hold a party “in default” even if its opponent might be prejudiced.

Existing prejudice requirements attempt to follow the Supreme Court’s command that “an allegation of waiver” should be resolved in favor of arbitration.¹⁴⁷ That language inspired courts to say that a party could only be “in default” if its opponent faced prejudice.¹⁴⁸ At first glance, imposing a prejudice requirement furthers the FAA’s pro-arbitration policy, but courts affixed prejudice requirements to overly broad conceptions of “default” based on erroneous interpretations of common law waiver. Unfortunately, the combination fails to fully respect either the full scope of the FAA’s pro-arbitration policy or the principles underlying waiver at common law.

C. Consistency with Waiver Doctrine

A rule that does not count litigation conduct toward a finding of waiver furthers the goals of common law waiver doctrine. At common law, there is a presumption against waiver, which is especially strong when the right in question is favored.¹⁴⁹ As a result, implied waiver must be “clearly demonstrated,”¹⁵⁰ making it “nearly impossible for courts to imply waiver . . . based on conduct alone.”¹⁵¹

In the arbitration context, engaging in litigation is not the unequivocal conduct that traditional waiver rules require; there are many reasons that a party might engage in litigation without intending to give up its right to arbitrate. First, a defendant with a strong jurisdictional or procedural defense might gain little by moving to compel arbitration rather than obtaining a quick dismissal in court. For example, a party sued in federal court over a contract dispute could easily obtain dismissal for want of subject-matter jurisdiction while preserving the right to arbitrate on the merits.¹⁵² Indeed, judges are well equipped to resolve the narrow legal questions presented by jurisdictional disputes, unlike arbitrators who are experts in the subject matter of the dispute. In fact, many courts already recognize that filing procedural motions does not establish intent to waive arbitration rights.¹⁵³ That logic also extends to dispositive motions on the merits. A party’s decision to seek the court’s resolution of a particular motion does not show that the party wants to stay in court if the motion fails. The early stages of litigation also provide opportunities for settlement talks. A party might reasonably engage in settlement negotiations after the lawsuit was filed but not intend to lose its arbitration rights should the settlement talks fail.¹⁵⁴

Second, the parties may intend to narrow the scope of an arbitration agreement without jettisoning it entirely. For example, arbitration agreements may restrict the scope of discovery available to the parties.¹⁵⁵ But it is possible to use discovery obtained from litigation in subsequent arbitration.¹⁵⁶ That matters under current approaches to waiver because discovery obtained in litigation that would also have been obtainable in arbitration is less likely to support waiver.¹⁵⁷ However, both parties may want access to additional discovery beyond that provided under their arbitration agreement.¹⁵⁸ Under the bright-line approach, the parties could pursue discovery in court and then apply it in arbitration. This method is flexible and preserves many of arbitration’s benefits, such as the ability to appoint a subject-matter expert as the arbitrator to resolve complex technical disputes on the merits.¹⁵⁹ Moreover, it allows parties to decide whether to pursue arbitration with the benefit of the additional information obtained through the discovery process.¹⁶⁰

Third, parties may engage in litigation despite intending to arbitrate because there are questions about whether a dispute is arbitrable, and erroneously pursuing arbitration risks expiration of the statute of limitations.¹⁶¹ In other cases, a dispute might not be arbitrable when a suit is filed, but becomes arbitrable after a later change of law.¹⁶² For example, the Supreme Court determined that claims could be arbitrable even if “intertwined factually and legally” with nonarbitrable claims in 1985, reversing decisions in the Fifth, Ninth, and Eleventh Circuits.¹⁶³ As a result, parties in these circuits that had litigated arbitrable claims intertwined with nonarbitrable claims before 1985 based on circuit precedent may not have intended to waive their arbitration rights. In fact, some pursued arbitration after the Supreme Court rejected the intertwining doctrine.¹⁶⁴ Finally, litigation allows for the use of procedural devices like class actions that may be unavailable in arbitration. A party might attempt to proceed as a class in court but seek individual arbitration if its class action is dismissed.¹⁶⁵

These examples show that courts should be reluctant to find that a party’s litigation conduct demonstrates an intent to waive its arbitration right. Instead, a party should be able to choose when to exercise its right to arbitrate without risking the loss of an arbitrator’s skill and expertise and the cost of a lengthy trial.

D. Fair Dispute Resolution

The bright-line rule also helps to ensure that arbitration agreements are enforced fairly. Both federal and state judges have raised concerns that enforcing arbitration agreements between parties of disparate bargaining strength can be unjust. For example, the California Supreme Court has held that class action waivers in “consumer contract[s] of adhesion” are unconscionable when a “party with the superior bargaining power” defrauded large numbers of customers of individually small amounts of money.¹⁶⁶ And Justice Ginsburg’s dissent in Epic Systems Corp. v. Lewis criticized the majority’s decision to enforce collective-action waivers in adhesive employment contracts without addressing the power imbalance between employers and employees.¹⁶⁷ Furthermore, binding arbitration makes noncompetition agreements easier to enforce.¹⁶⁸ Moreover, worries about the extensive use of adhesive arbitration agreements have grown substantially as of late, even beginning to reach the general public.¹⁶⁹

Despite these criticisms, the Supreme Court has dismissed arguments that equitable concerns justify the nonenforcement of arbitration agreements. Instead, it requires arbitration agreements to be enforced as written.¹⁷⁰ The Court has also specifically rejected the application of tools “based on public policy considerations” rather than the intent of the parties to ameliorate harsh consequences of arbitration agreements.¹⁷¹ To do otherwise, the Court explained, would undermine both the parties’ agreement and arbitration’s promise of quick and efficient dispute resolution by a subject-matter expert.¹⁷²

But the Supreme Court’s reluctance to use equitable concerns when interpreting the FAA does not necessarily mean that arbitration is unfair. Rather, whether arbitration is fair depends on how it is implemented.¹⁷³ Professor Jean R. Sternlight identifies two features of arbitration in the United States that undermine its potential as a just dispute-resolution system: lack of subjective consent and insufficient transparency.¹⁷⁴ The bright-line rule ameliorates both of these equitable concerns in the arbitration-waiver context.

First, the bright-line rule promotes subjective consent because it incentivizes the parties to discuss the potential for arbitration at the start of litigation. Under current arbitration-waiver rules, powerful parties can engage in protracted litigation to drain their opponents’ resources before invoking their arbitration rights. For example, the court in MicroStrategy, Inc. v. Lauricia compelled arbitration of an employee’s age and sex discrimination claims even after the employer undertook costly litigation, apparently “for the sole purpose of wearing [the employee] out, both emotionally and financially.”¹⁷⁵ The ability of a party with superior bargaining power to drain an opponent’s resources under the guise of litigation undermines the fundamental principle that “[a]rbitration is strictly a matter of consent.”¹⁷⁶

In contrast, the bright-line rule puts both parties on notice that arbitration is always possible. A party that wants to avoid arbitration will then have a strong incentive to discuss the possibility of waiver early in the lawsuit.¹⁷⁷ The other party may agree to expressly waive arbitration, which would promote the intent of both parties¹⁷⁸ and keep the court informed.¹⁷⁹ Alternatively, the other party might insist on immediate arbitration, which would ensure that no party is misled into wasting money on litigation. Finally, the opposing party may refuse to come to any agreement, in which case the party seeking to avoid arbitration would be fully aware of the danger of a late motion to stay. As a result, the party opposing arbitration could accurately weigh the costs and benefits of remaining in litigation and decide whether to preemptively move to compel arbitration to reduce unnecessary litigation costs. Regardless of the outcome, the party opposing arbitration can make an informed, affirmative choice about how best to protect its interests. And making informed choices promotes the kind of subjective consent that adhesive contracts might otherwise lack.¹⁸⁰ Moreover, trial courts are well positioned to facilitate discussions between the parties about their arbitration rights before substantial discovery or motion practice takes place because they supervise case management.¹⁸¹

The bright-line rule also promotes subjective consent by discouraging unfair gamesmanship. Under the current system, parties—including plaintiffs—file late motions to stay opportunistically.¹⁸² And existing waiver rules make it difficult to predict when motions to stay will be granted after litigation has begun. This uncertainty makes it difficult for parties to plan and exposes them to prejudice if they guess wrong,¹⁸³ undermining the informed decisionmaking that subjective consent requires.¹⁸⁴ The need for a uniform rule is especially acute because certain states have adopted a different test for waiver than the federal circuits in which they lie.¹⁸⁵ In these jurisdictions, the losing party of a state court waiver battle might file a federal lawsuit designed to reach the opposite result.¹⁸⁶ Forum shopping undermines the finality of the state court’s decision and causes needless expense and delay.¹⁸⁷

However, it is fair to order a stay when the party opposing arbitration was aware that arbitration was a possibility because that party can protect itself in litigation.¹⁸⁸ A litigant can avoid prejudice by seeking an express waiver or immediately invoking the arbitration agreement. Of course, the party could still decide to take its chances in litigation, thereby making an informed choice to litigate and bear the costs of late motions to stay. Indeed, the Ninth Circuit has explained that costs resulting directly from a party’s decision to litigate rather than arbitrate do not support prejudice.¹⁸⁹ The Ninth Circuit’s approach is qualified because it still counts costs incurred due to the other party’s litigation conduct toward prejudice,¹⁹⁰ but its logic extends to support the bright-line rule: a party should not be insulated from litigation costs that result from its choice to forego arbitration. Overall, the bright-line rule facilitates subjective consent that protects vulnerable parties and minimizes gamesmanship more effectively than an ad hoc prejudice inquiry.¹⁹¹

Second, the bright-line rule promotes the kind of transparency and rule-of-law values that Professor Sternlight argues fair arbitration systems must protect because it offers a rubric for clear guidance and consistent decisions for courts.¹⁹² Current arbitration-waiver rules are unclear and inconsistent. They fail not only because courts in different jurisdictions apply different standards but also because the amorphous nature of existing tests makes it difficult to predict the outcome of particular cases. For example, a thirteen-month delay encompassing discovery did not justify waiver in one case, while an eight-month delay did in another case—all in the First Circuit.¹⁹³ A system that produces such inconsistent results undermines transparency because the public cannot predict how courts will treat future cases. In contrast, the bright-line rule is easy to apply and will produce consistent, predictable results. This consistency promotes transparent decisionmaking and rule-of-law values by treating similarly situated litigants alike.¹⁹⁴

The bright-line rule does not, of course, address broader questions involving consent in contracts of adhesion or transparency within the arbitration process. But it promotes fair enforcement of arbitration-waiver rules in a manner that is consistent with the FAA and that properly leaves broader structural questions about arbitration to Congress, which actively considers reforms and is better positioned than courts to weigh arbitration’s advantages and disadvantages.¹⁹⁵

E. Risks of the Bright-Line Approach

One potential disadvantage of the bright-line rule is the risk of duplicative proceedings in arbitration following a late motion to stay litigation.¹⁹⁶ Excessive costs and delays from duplication would undermine the FAA’s efficiency goals.¹⁹⁷ But the bright-line rule incentivizes parties to either quickly agree to arbitration to avoid a switch to arbitration later in litigation or to expressly waive arbitration rights by shifting the risk of prejudice to the party that might otherwise try to avoid arbitration entirely.¹⁹⁸ That reduces the chance that disputes regarding the use of arbitration will break out after extended litigation. Should a dispute nevertheless emerge, the bright-line rule would provide clear and quick answers because courts would order stays regardless of how much litigation activity took place. Current approaches to waiver that rely on multifactor tests¹⁹⁹ or broad language²⁰⁰ invite litigation because both parties can often present colorable arguments in their favor regardless of the exact wording of the controlling legal test. Moreover, orders denying stays under section 3 are immediately appealable, exacerbating the expense of litigation under the current approach by enabling lengthy appeals.²⁰¹

Even if the bright-line rule would result in some duplication in individual cases, its costs are unlikely to significantly exceed those incurred under current rules. Discovery obtained in litigation may be usable in arbitration.²⁰² Some of the work required to craft a litigation strategy and to prepare motions in litigation could also be applied in the arbitration context. And litigation expenses may already be duplicated under the current tests.²⁰³ Overall, excess costs that might arise in individual cases from the bright-line rule are likely to be small and outweighed by the structural advantages it would provide.

Another potential argument against the bright-line rule is its heavy reliance on waiver doctrine. Indeed, attempts to interpret section 3 according to common law waiver principles would be fundamentally misguided if “in default” did not refer to waiver. Moreover, there is widespread disagreement about which common law doctrines, if any, should be used.²⁰⁴ Two alternatives include estoppel²⁰⁵ and forfeiture.²⁰⁶

Both of these alternatives are seriously flawed. First, estoppel rests on inequitable conduct and is “invoked to prevent injustice” or “promote the ends of justice.”²⁰⁷ But the bright-line rule avoids injustice by placing a party subject to an arbitration agreement on notice that the other party could invoke the arbitration agreement at any time, which allows the party to take steps in litigation to protect itself by immediately discussing with the other party whether to expressly waive the arbitration agreement or by preemptively invoking it.²⁰⁸ In addition, estoppel generally requires that the party being estopped “intended to influence the other” through its words or actions.²⁰⁹ In principle, a party could intend to mislead the other by engaging in litigation.²¹⁰ However, courts deciding waiver questions tend to consider whether a party moving for a stay intended to litigate rather than whether it intended to deceive the other party.²¹¹

Forfeiture fares little better. There is little case law that clearly or consistently distinguishes waiver from forfeiture.²¹² The D.C. Circuit has declared that forfeiture (unlike waiver) “entails no element of intent”;²¹³ however, its approach has yet to catch on.²¹⁴ Regardless, the D.C. Circuit’s test still looks at familiar factors like the cost imposed on the other party, the amount of discovery, the length of the delay, and the volume of filings to determine whether the party forfeited its right to arbitration.²¹⁵ That undermines the idea that the D.C. Circuit’s test is truly different from a test based on common law waiver doctrines. Worse, forfeiture-based tests fail to adequately respect the FAA’s pro-arbitration policy.²¹⁶ Grounding the analysis of default under section 3 in forfeiture is therefore unlikely to solve the problems of existing waiver-based approaches.

Conclusion

Current approaches to waiver in the arbitration context are unworkable. These approaches have created a federal circuit split and conflicts between state and federal courts with overlapping jurisdictions, thereby undermining the FAA’s promise of consistent enforcement of arbitration agreements and efficient dispute resolution. Moreover, due to disagreements about the definition of prejudice, adopting a prejudice requirement would fail to provide clarity. Rejecting a prejudice requirement similarly fails to provide adequate guidance to courts. Instead, the bright-line approach that enforces arbitration agreements as written and without regard to litigation conduct would best vindicate the parties’ intent and promote speedy and efficient dispute resolution.

* J.D. Candidate, May 2021, University of Michigan Law School. I would like to thank Professor Adam C. Pritchard for his suggestions and advice on this Note. I am also grateful for the thoughtful feedback and careful editing provided by the Michigan Law Review.