Almost one hundred years after the Federal Arbitration Act (FAA)’s enactment, a long-overlooked limit in its text is attracting the attention of federal judges. The FAA commands courts to enforce arbitration clauses in private contracts and empowers them to do so by staying judicial proceedings and ordering arbitration when a litigant shows that her opponent agreed to arbitrate the instant suit. But not all disputes fall within the FAA’s ambit. As section 2 of the Act makes clear, the FAA authorizes courts to enforce arbitration agreements only when the dispute to be arbitrated “aris[es] out of” the contract containing the clause or the transaction evidenced by that contract. This limitation has long lain dormant because until recently contracting parties rarely (if ever) prescribed arbitration for disputes without a fairly close relationship to the original transaction or agreement.
Times have changed. With corporations increasingly binding workers and consumers to arbitration provisions that purport to cover vast classes of claims and defendants, courts have begun asking whether the FAA actually authorizes them to compel arbitration in disputes that come within these contracts’ expansive language but bear little relation to their substance. Recently, in Calderon v. SIXT Rent a Car, LLC, an Eleventh Circuit panel became the first federal court to find that a dispute did not arise out of the contract on the basis of which a party sought to compel arbitration. While the court took a welcome step in acknowledging the FAA’s “arising out of” limitation, it drastically understated the legal consequences of that holding. In so doing, it created doctrinal confusion at the precise moment the federal judiciary is beginning to take the FAA’s built-in limits seriously.
In early 2019, Ancizar Marin rented a luxury car from SIXT Rent a Car (“Sixt”) and returned it two uneventful days later. Much to his surprise, he soon received an email from Sixt accusing him of damaging the car. Then, a month later, a second email arrived with a demand letter for over $700. Suspecting a “company-wide scheme to systematically defraud consumers,” Marin brought a class-action lawsuit in federal court in Florida alleging violations of Florida’s state consumer protection laws. Marin was joined by a second named plaintiff, Philippe Calderon, whose story differed from Marin’s in one key respect: whereas Marin reserved his car through Orbitz.com (“Orbitz”), a third-party travel site, Calderon booked his directly through Sixt. The difference mattered because Orbitz, unlike Sixt itself, required customers to assent to a broadly worded arbitration provision when completing their reservation. In particular, Orbitz’s arbitration clause purported to bind customers to arbitrate “any and all Claims,” including not only “Claims” asserted against Orbitz itself, but also those brought against its “subsidiaries, travel suppliers or any companies offering products or services through [Orbitz], including Suppliers, (which are beneficiaries of this arbitration agreement).” Citing this agreement, Sixt moved to compel arbitration in Marin’s case and that of any other class members who had booked through Orbitz.
The district court denied Sixt’s motion. First, the court held that Sixt was neither a party to the agreement containing the arbitration provision nor a third-party beneficiary entitled to enforce it. The court explained that the plain language of the contract repeatedly referred to agreements between “You”—that is, the customer—and “Orbitz,” “We,” or “Us.” Thus, the court concluded, “under no reasonable reading [could] it be said that Sixt was a party.” Nor was the court persuaded that the narrow definition of “Supplier” presented elsewhere in the contract could be read to cover Sixt such that the company could invoke the arbitration clause’s reference to “Suppliers.” Finally, turning to a different section of the agreement, the court reasoned that the term “Claims” referred only to allegations involving customers’ use of the Orbitz platform, not subsequent disputes between customers and service providers.
The Eleventh Circuit affirmed. Writing for a unanimous panel, Judge Newsom largely echoed the district court’s analysis of Orbitz’s arbitration agreement and placed particular weight on the contractually defined term “Claim,” which the court understood to refer only to “services or products provided . . . by Orbitz.” As such, the court held that “Sixt’s rental-car service [did not] fall within the category of arbitrable ‘Claims’” (emphasis added).
Sixt tried to “overcome that most natural reading” by invoking the Supreme Court’s oft-cited command that “any doubts concerning the scope of arbitrable issues should be resolved in favor of arbitration.” This presumption is sometimes called the “Moses H. Cone canon” after the case in which the Court derived it from the “liberal federal policy favoring arbitration agreements” that it understood the FAA to establish. In a portion of the Eleventh Circuit’s opinion joined by only two judges, the court held the presumption unavailable. Judge Newsom observed that section 2 provides that “[t]he Act governs an arbitration agreement only to the extent that it compels arbitration of ‘controvers[ies]’ that ‘aris[e] out of’ the ‘contract’ containing the arbitration agreement or the ‘transaction’ evidenced thereby.” And since Marin’s suit against Sixt did not arise out of his contract with Orbitz—because it was not “an immediate, foreseeable result of [its] performance”—the court concluded that it was not governed by the FAA. “[A]ccordingly, . . . Moses H. Cone’s pro-arbitration canon of construction d[id] not apply.”
Judge Newsom also concurred separately to “voice [his] skepticism of the Moses H. Cone rule.” He argued that the rule finds support neither in the FAA’s text nor any history that might inform that text’s meaning. Furthermore, he continued, this “especially made up” rule has become “especially potent” because unlike the many substantive canons of construction that are applied as a last resort, the Moses H. Cone rule “diverts courts from the best reading of the text at the first hint of uncertainty, and thereby works a massive alteration of written contracts in America.” The better way to apply the FAA, he concluded, would be to enforce arbitration agreements just like other contracts, including by interpreting them, “first and foremost, according to their plain text.”
Finally, Judge Jill Pryor filed a concurring opinion to explain her decision not to join the portion of the court’s opinion declining to apply the Moses H. Cone canon. Because she had “no doubt” that the Orbitz agreement’s arbitration clause did not cover Marin’s claims, she found it unnecessary to consider the canon’s applicability.
Most commentary on Calderon has focused on Judge Newsom’s forceful critique of the Moses H. Cone rule. Far less attention has been paid to the court’s recognition that the FAA only applies where a party seeks to compel arbitration of a dispute that “arises out of” the contract containing the arbitration clause or the transaction evidenced thereby. Following its initial appearance in a 2020 article by Professor David Horton, the observation that section 2 imposes this “contractual nexus requirement” found its way into a Fourth Circuit dissent and then a Ninth Circuit concurrence. The Calderon court’s finding that Marin’s suit against Sixt did not “arise out of” his agreement with Orbitz—and thus that it was not governed by the FAA—marks the first time a federal court majority has acknowledged and applied this feature of section 2. This is a welcome development because section 2’s contractual nexus requirement offers courts a way to reject overbroad applications of expansively worded arbitration agreements like Orbitz’s while remaining faithful to the FAA’s text and Supreme Court precedent.
Unfortunately, however, the Eleventh Circuit severely understated the consequences of its holding. According to Judge Newsom, the upshot of the FAA not applying was merely that Sixt could not benefit from the Moses H. Cone rule. Indeed, the Westlaw version of Calderon now bears a citable headnote summarizing its holding as “[s]trong pro-arbitration requirement emanating from [FAA] applies only to arbitration agreements within coverage of Act.” This is surely correct, but only as a somewhat trivial consequence of the more fundamental conclusion that the court overlooked: absent authorization from the FAA, a court has no power under federal law whatsoever to enforce an arbitration agreement.
As Judge Newsom noted in his concurrence, “American law made arbitration contracts functionally unenforceable” prior to the FAA’s enactment in 1925. For centuries, courts guarding against “ouster” of their jurisdiction and exploitation of vulnerable parties simply refused to enforce arbitration agreements—regardless of “whether the agreement was to arbitrate future disputes or to submit an existing dispute to arbitration.” There is broad consensus that the FAA’s purpose was to “reverse th[is] longstanding judicial hostility” to arbitration, at least in some contexts, but nonenforcement remains the default rule where the FAA or a parallel state arbitration act does not apply. For that reason, the Supreme Court has held as recently as 2019 that whenan arbitration contract falls into one of the exceptions in the FAA’s first section, federal courts “lack[] authority under the Act to order arbitration.” Thus, having found that the FAA did not govern Marin’s agreement with Orbitz in this dispute, the court should have acknowledged that it had no power to compel arbitration under that statute and identified a separate basis, if any existed, for that authority.
At first glance, this may seem like nitpicking. After all, the court did refuse to order arbitration, albeit on different grounds. And even if the FAA did not authorize the court to enforce Orbitz’s agreement, the Florida Arbitration Act, which does not contain a contractual nexus requirement, likely would have. However, precision is important here for two reasons. First, the Supreme Court has construed the FAA to preempt broad swaths of arbitration-limiting state statutory and common law. By keeping agreements under the rubric of the FAA when it concededly does not govern, courts following Calderon would unfairly deprive claimants of these more favorable rules, which should return to force in the Act’s absence. Second, the court’s reasoning could allow corporations to enforce egregiously overbroad, even “infinite,” arbitration clauses so long as they drafted them clearly enough to leave no doubt about their expansive scope. Under the court’s approach, a plaintiff’s protest that her dispute did not “arise out of” a given arbitration agreement would be no answer to the defendant’s attempt to enforce it because the lack of contractual nexus would do no more than deprive the defendant of the presumption that “doubts concerning the scope of arbitral issues should be resolved in favor of arbitration”—a presumption the defendant would not need to rely on if the agreement spoke in unambiguously infinite terms.
Calderon represents an important milestone in the federal judiciary’s growing appreciation for the FAA’s boundaries. Courts considering the next round of infinite arbitration agreements should embrace the Eleventh Circuit’s approach to section 2 but follow it to its true conclusion: when a dispute does not arise out of an arbitration agreement, federal courts cannot rely on the FAA for authority to compel arbitration. That authority must come, if at all, from state law, with whatever caveats and limitations the state has chosen to impose.