Last week, we posted about two cases in which Circuit Courts tried to figure out when a party will be deemed to have waived its right to arbitrate through conduct inconsistent with the right, such as having “substantially invoked the judicial process.” As discussed last week, the Fifth Circuit held in an unpublished opinion, Global Advantech Resources Ltd. v. Brown, that defendants had waived their arbitration rights by repeatedly requesting dismissal with prejudice, by raising thirty-two affirmative defenses in their Answer without reserving a right to seek arbitration, and by engaging in merits-based discovery. However, in Parkin v. Avis Rent a Car System, LLC, the Third Circuit vacated the denial of a motion to compel arbitration of a putative class action, finding that filing a motion to dismiss and conducting discovery did not constitute an implied waiver of their right to arbitrate, mostly because Defendants needed to conduct discovery in order to determine whether they had a basis for invoking the arbitration clause.
I expressed skepticism about the Third Circuit approach. Defendants in Parkin got two of three claims dismissed before they began discovery. Courts could require limited discovery related to arbitration before considering dispositive motions, but the Third Circuit does not want to make that option a requirement, but in an unpublished opinion, the majority of a Ninth Circuit panel did. We’ll get there, but first, we start with another Third Circuit case.
In Salvatora v. XTO Energy Inc., the Third Circuit vacated the denial of a motion to compel arbitration despite the fact that defendant had litigated for 55 months before moving to compel arbitration.
In Uhl v. Roblox Corporation, the Ninth Circuit ruled, over a dissent, that the district court did not err in finding that defendant waived its right to compel arbitration by litigating for almost a year, removing the case to federal court, and moving to dismiss on the merits.
Salvatora is a class action filed in August 2019, alleging $5 million in damages from underpayments of royalties due under oil and gas leases. None of the six named plaintiffs’ leases included arbitration provisions. Defendant XTO Energy Inc. (XTO) answered the three amended complaints, asserting affirmative defenses, but never mentioning arbitration. The parties proceeded to discovery and mediation. During discovery, it emerged that as many as 16 potential class members’ leases (out of 500) included arbitration provisions. In its opposition to class certification, XTO expressly refused to waive its right to arbitration under leases containing arbitration provisions. The Magistrate Judge recommended certification, while leaving open the possibility that XTO could later raise an arbitration defense to certification. The District Court adopted that recommendation.
In March 2024, XTO moved to compel arbitration against class members whose leases included arbitration provisions. Plaintiffs countered that XTO had waived its right to compel arbitration by not mentioning arbitration in its Answers and by litigating for 55 months before moving to compel. The Magistrate Judge denied the motion. Relying on the District Court opinion in Valli v. Avis Budget Grp., Inc., which was vacated by the Third Circuit, as discussed in last week’s post, the Magistrate concluded that XTO had illustrated a “preference for litigation.” The District Court rejected XTO’s objections to the Magistrate Judge’s recommendations, and XTO took an interlocutory appeal.
One might think that 55 months of litigation illustrates XTO’s preference for litigation over arbitration, but Valli has a caveat that applies to class actions in which the named plaintiffs are not subject to arbitration provisions. The caveat means that XTO’s failure to seek to compel arbitration before class certification does not illustrate a preference for litigation over arbitration, but such a preference could be shown through other conduct. Plaintiffs pointed to five actions or inactions by XTO as evidence of such a preference. Long story short, according to the Third Circuit, none of them really establish such a preference, and that seems right to me.
The Third Circuit explained that before certification, unnamed class members are not parties, so XTO’s answers, discovery, and mediation were not litigation against them. Its opposition to certification expressly preserved arbitration and sought to exclude arbitration leaseholders from the class rather than litigate against them. Because the class was not yet defined, it was unclear which arbitration leaseholders would be included in the class. And XTO moved to compel arbitration within five weeks of the finalization of class membership, the timing the Magistrate Judge had contemplated.
So here, unlike in Parkin, I think the Third Circuit got things right. In both cases, defendants faced a choice of not whether to litigate but how to litigate. However, in Parkin, I think defendants acted strategically, using dispositive motions to dispose of claims prior to arbitration before moving to compel. In Salvatora, XTO participated in discovery and the certification process while reserving its right to arbitration and awaiting class members’ individual decisions about whether to opt out of the putative classes. As XTO is only seeking to arbitrate with 15 or 16 out of 500 potential class members, it is far easier to see why it would have proceeded with litigation as long as it did before determining that a motion to compel was needed.
Uhl v. Roblox Corporation is an unpublished Ninth Circuit decision. The Ninth Circuit test for waiver requires only (1) knowledge of an existing right to compel arbitration and (2) intentional acts inconsistent with that existing right. The first part requires only that the party has knowledge of and knows how to assert its right to compel arbitration. That part is met because plaintiff Damien Uhl’s children had been playing Roblox Corporation (Roblox) games since 2017, and Roblox’s Terms have included a requirement of parent or guardian consent and an arbitration clause since that time. Roblox thus had knowledge of its right to arbitrate based on the complaint. As to the second part, Roblox litigated for months and filed a motion to dismiss. Only after the District Court declined to dismiss the claims with prejudice did it move to compel.
Roblox argued that it could not move to compel without Mr. Uhl’s daughter’s username. In stark contrast to the Third Circuit approach that we saw in Parkin, the Ninth Circuit provides that “when a party has knowledge of its right to compel arbitration but does not have all the evidence required to win a motion to compel, the proper procedure is to move to compel, and then, if necessary, seek arbitration-related discovery.” Yup, that’s the rule I like. Judge Bumatay, dissenting, does not think that is the rule in the Ninth Circuit. “[N]o case advises that the failure to seek discovery on arbitration issues constitutes a waiver of the right to arbitrate.” Judge Bumatay acknowledges that it would be good to have a rule requiring that a party seek discovery to determine whether to file a motion to compel arbitration, but no such rule currently exists, and he thinks it would be unfair to apply such a rule in this case for the first time. That also seems like a good argument.
In the Ninth Circuit, a motion to dismiss on the merits is generally regarded as inconsistent with intent to arbitrate. Roblox did not only do that. It also removed the case to federal court, litigated a motion to remand, and never in eleven months of litigation sought to compel arbitration. Judge Bumatay thinks none of that is enough to illustrate a preference for litigation over arbitration, and he persuades me that the matter is close. However, waiver is a mixed question of fact and law. In a close case, I would give more weight to the District Court’s conclusion than Judge Bumatay did.
Judge Bumatay closed by suggesting that Roblox’s conduct sounds more like forfeiture than waiver. It’s not clear why that would matter here. In Morgan v. Sundance, Inc., the U.S. Supreme Court did not resolve whether arbitration rights could be forfeited in addition to being waived. Nothing in the opinion suggests that the issue was preserved here, but it would be nice to know whether late assertion, without proof of intent to relinquish, should cost a party its right to arbitrate. But if the Ninth Circuit could have addressed that issue, it would have had to make this a published opinion.
What conclusions can we draw from these cases? While all courts use a holistic approach to decide the question, it seems that the Third Circuit is more forgiving of parties that delay in bringing their motions to compel arbitration than other Circuit Courts. The test is whether a court can infer from the party’s conduct that it prefers litigation to arbitration. Moving to dismiss with prejudice is a strong indicator but need not be dispositive. Expressly reserving the right to arbitrate throughout the litigation process is a strong indicator in the other direction, but that can be overridden if the party bypasses multiple off-ramps from litigation into arbitration. I would like courts to adopt the procedure the majority invoked in Uhl v. Roblox Corporation, which the Uhl dissent questioned and the Third Circuit in Parkin declined to require. Under that approach, a court could treat it as evidence of a preference for litigation over arbitration when a party that knows of its right moves to dismiss on the merits though it could have first sought discovery in support of a motion to compel arbitration.
Recent Comments